Brain Research, 524 (1990) 282-290
282
Elsevier BRES 15757
Crossed forelimb extension produced in thalamic cats by injection of putative transmitter substances into the paralemniscal pontine reticular formation Muneo Shimamura, Tatsu Fuwa and Ikuko Tanaka Department of Neurophysiology, Tokyo Metropolitan Institutefor Neurosciences, Fuchu-city, Tokyo (Japan) (Accepted 20 February 1990)
Key words: Reticulospinal tract neuron; Extensor muscle; Glutamate; Forelimb movement; Thalamic cat; Crossed limb extension
To analyze the descending pathways of the paralemniscal pontine reticular formation (PLRF), a technique was used for the selective activation of cell bodies by localized injection of putative neurotransmitters in the PLRF. When a small amount (less than 0.1 pl) of 0.1 M glutamate was injected into the PLRF unilaterally in thalamic cats, the forelimb contralateral (c-forelimb) to the injection was extended, and occasionally the ipsilateral forelimb was flexed. These responses were similar to those obtained by electrical stimulation of the PLRF, but were relatively weaker. Unit spikes of PLRF neurons were increased in frequency following administration of glutamate. The latent periods and durations of increases in spike frequency varied depending on the concentration and quantity of the glutamate solution, and were roughly similar to those of the extensor EMG in the c-forelimb. Since the firing of PLRF neurons preceded the EMG with 11 ms latency, the unit spike of PLRF neurons could be used as a triggering signal to observe a spike triggered averaged EMG response in the extensor muscle of the c-forelimb. Results similar to those with glutamate were observed upon administration of quisqualate, kainate and aspartate. The most effective compound was quisqualate. Application to the PLRF of 1-naphthylacetyl spermine (1-NA-Spm), an analogue of the natural spider toxin JSTX-3 and an antagonist of glutamate, suppressed both the PLRF neuron activity and the extensor EMG of the c-forelimb. These observations suggest that extensor muscles of the forelimb are excited by the contralateral PLRF, perhaps via the crossed reticulospinal tract from the PLRF. PLRF neurons may be activated by glutamate (quisqualate) receptors. INTRODUCTION Recently we have r e p o r t e d that the paralemniscal pontine reticular formation ( P L R F ) exerts a bilateral control of forelimb muscles and plays a role in initiating and maintaining forelimb stepping; the P L R F receives afferent input bilaterally from the forelimb nerves, and electrical stimulation of the P L R F in thalamic cats causes s t e r e o t y p e d locomotor-like forelimb movements ~5. The forelimb ipsilateral to stimulation (i-forelimb) is flexed while the contralateral forelimb (c-forelimb) is extended. This m a y be caused by impulses descending through the reticulospinal tracts directly and/or indirectly to spinal m o t o r pools. Unilateral transection of the ventral quadrant at the cervical cord (C3) abolished PLRF-elicited flexor E M G s and associated extensor E M G suppression ipsilateral to the transection, while the side contralateral to the transection r e m a i n e d unchanged 15. Nevertheless, there has been no detailed analysis of the relationship b e t w e e n reticulospinal tracts from P L R F and s t e r e o t y p e d forelimb movements. In o r d e r to confirm the presence in the P L R F of neuron somata capable of producing
s t e r e o t y p e d forelimb m o v e m e n t s when stimulated and to map their distribution in the brainstem, we used a technique for selective activation of cell bodies by the localized injection of small quantities of putative neurotransmitters 6. In this p a p e r , we d e m o n s t r a t e that the selective activation of cell bodies in the P L R F by a variety of neuroactive substances causes c-forelimb extension in thalamic cats. To analyze the effects of drug administration, we used several putative transmitter substances, glutamate, quisqualate, kainate, aspartate, and the glutamate antagonist, 1-naphthylacetyl-spermine (1-NASpm), and p e r f o r m e d kinematic analyses of forelimb movements, bilateral E M G recordings in forelimb extensor and flexor muscles, and recordings of cutaneous e v o k e d P L R F potentials, P L R F - i n d u c e d forelimb muscle responses and unit discharges in the P L R F neurons.
MATERIALS AND METHODS Twenty-five adults cats (2.3-3.0 kg b. wt.) were studied. A detailed description of the animal preparation has been reported
Correspondence: M. Shimamura, Department of Neurophysiology, Tokyo Metropolitan Institute for Neurosciences, 2-6 Musashidai, Fuchu-city, Tokyo 183, Japan. 0006-8993/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
283 elsewhere 15. Under ether anesthesia, tracheal cannulation, decerebration at the stereotaxic A12 level and spinal transection at the T H segment were performed. Immediately after the decerebration, ether was withdrawn. The head was fixed in a stereotaxic device, the T 2 spinal process clamped to a metal frame and the lumbar region suspended by a hammock so that all limbs touched the floor of the treadmill belt. Electrical stimulation was applied to the PLRF, using a coaxial needle electrode and a glass-coated tungsten microelectrode. Pulses of 0.1-0.3 ms in duration, 50 Hz in frequency and of less than 50 p A intensity were used. The threshold of electrical stimulation for evoking EMGs of forelimb muscles was measured. Evoked potentials were recorded from the PLRF following stimulation of the superficial radial nerve with a cuff electrodeTM. EMGs were recorded on a polygraph using appropriate amplifiers. The electrodes consisted of paired enamel-coated copper wire inserted bilaterally into extensors (m. triceps brachii) and flexors (m. biceps brachii) of the forelimb. Chemical stimulation of the PLRF was achieved by administering substances by pressure microinjection, using a multi- (mostly 6) or double-barreled glass micropipene. In one barrel, a tungsten micro-wire (tip 10 pm diameter) was inserted for recording of neural activity and for electrical stimulation of the PLRF while the other barrel (tip about 6 /tin diameter) was filled with a solution containing one of the following: glutamate (sodium L-glutamate, mono), quisqualic acid (fl-(3,5-dioxo-l,2,4-oxadiazolidin-2-yl)-Lalanine), aspartate (L-aspartic acid), kainic acid, serotonin (serotonin-creatinine sulfate), noradrenaline (4-(2-amino-l-hydroxyethyl)1,2-benzene-diol), dopamine (3,4-dihydroxyphenyl-ethylamine), GABA (4-amino-N-butyric acid), carbachol, picrotoxin and 1naphthylacetyl spermine (1-NA-Spm). These drugs were dissolved in normal saline except for noradrenaline and dopamine which were dissolved in 5 x 10-5 M ascorbic acid in D-H20 to prevent oxidation. Every drug was used at concentrations between 1-100 raM. Concentration ranges were chosen on the basis of previous successful experiments using drug injections to initiate locomotion5' lo. The drugs were pressure injected at a rate of 0.06 pl/min with the micropipette attached to a micrometer-driven oil pressure pump with polyethylene tubing. We used a multi-barreled glass micropipene for examination of the approximate responsiveness of PLRF neuron activity to a variety of solutions, and a double-barreled one for detailed analysis of the effects of each solution, because of possible cross-contamination between barrels. When chemical substances were injected into the PLRF, the following phenomena were used for monitoring the effects of drug administration; unit discharges of PLRF neurons, leg movements observed directly and through EMG recordings, cutaneous evoked PLRF potentials and the threshold PLRF stimulation for inducing stereotyped leg movement. In several cases, unit discharges were identified as antidromic spikes from the spinal cord in reticulospinal neurons, recorded upon electrical stimulation of the lateral funiculus of the cervical cord at the 3rd and 7th segments bilaterally. To identify interactions between the unit spike of PLRF neurons and the EMG of forelimb muscles, a spike-triggered averaged EMG was constructed by computer. Leg movements were monitored by a TV recording system which ran synchronously with the EMG recording system. EMGs were recorded from the triceps and biceps brachii muscles of bilateral forelimbs. Cutaneously evoked PLRF potentials were recorded by the glass-coated tungsten microelectrode following stimulation of the superficial radial nerve. Stimulation and recordings of electrical activity were started 2 h after cessation of ether anesthesia. After electrophysiologicaland pharmacological observations were completed, a DC current was applied to the electrode appropriate to create a small lesion. The cat was then injected with a large dose of pentobarbitai sodium. The brain was removed and 100-pm transverse frozen sections were cut to identify the position of the tip of the electrode, The location of the electrode tip in the brainstem was plotted in relation to the anatomical features described in the atlas of Berman 3.
RESULTS
Forelimb movements elicited by injection of glutamate into the PLRF The location of the P L R F has been r e p o r t e d 15 as P 2.0-3.5 mm, 3 . 0 - 4 . 0 m m lateral from the midline, and 5.5-6.5 mm in d e p t h from the surface of the I V ventricle. The physiological criteria for the identification of P L R F are the followingl~: neuronal discharges of the P L R F are increased by limb m o v e m e n t and by touching the skin over the forelimbs bilaterally, particularly on the forelimb contralateral to discharge recording. Electrical stimulation of the P L R F causes s t e r e o t y p e d forelimb movements; the i-forelimb is flexed and the c-forelimb extended; E M G activity a p p e a r s in the flexor and is suppressed in the extensor of the i-forelimb, while the E M G of the c-forelimb extensor was increased. W h e n a small a m o u n t (less than 0.1 /d) of 0.1 M glutamate solution was a d m i n i s t e r e d from one barrel of the d o u b l e - b a r r e l e d glass m i c r o p i p e t t e unilaterally to the P L R F of the thalamic cats, s t e r e o t y p e d forelimb movements were e v o k e d , such that the c-forelimb increased the extension posture while the i-forelimb was lifted slightly, similar to results o b t a i n e d by electrical stimulation of the PLRF. The extensor E M G of the c-forelimb was increased while E M G a p p e a r e d in the ipsilateral flexor and was suppressed in the extensor of the i-forelimb. A s shown in Fig. 1A, a small a m o u n t of quisqualate (10 m M in saline solution, 0.01 /A total volume applied over 5 s) was injected into the right P L R E A few seconds after starting the injection, neuronal discharges from the P L R F a d j a c e n t to the region of microinjection of quisqualate were increased. The extensor E M G s in the c-forelimb (left triceps) were also increased simultaneously with the increase in the P L R F neuron activity. T h e flexor E M G of the i-forelimb (right biceps), on the o t h e r h a n d , a p p e a r e d at once and only for a brief period, and the extensor E M G was suppressed in the i-forelimb (right triceps). A f t e r cessation of injection, neuronal discharges were still increased in frequency and reached maximal frequency after 20-30 s, then decreased and recovered to the control level. During increase of neuronal discharge, E M G amplitudes were still elevated in the left triceps brachii muscles, while E M G amplitude in the triceps brachii muscle of the right forelimb decreased, and an E M G a p p e a r e d transiently in the biceps brachii muscle of the right forelimb. Fig. 1B shows E M G s of extensors and flexors in the forelimb bilaterally induced by electrical stimulation of the right P L R F with 50 I-Iz frequency and 3 5 / t A intensity using a glass-coated tungsten microelectrode. E M G s were markedly increased in the left triceps, and E M G activity a p p e a r e d in the right biceps but was suppressed in the right triceps.
284
A I l-biceps
,ii,,=,=J............................................................................. J ............I
r-triceps
il
:~:::
r-biceps
~
I
IO.5mV
~, . . . . .
0.2mV
r-PLRF ls
lOmM Qulsqulnc acid
C
.....
i................... ii
elect.stim. 2s
Fig. 1. Effects of application of quisqualic acid in the PLRF on EMGs of forelimb muscles and on PLRF neuronal activity. EMGs were recorded bilaterally from the biceps and triceps brachii muscles and unit spikes of the right PLRF neuron were recorded at the same time. A: when a small amount (0.01/A) of 10 mM quisqualic acid solution was injected into the right PLRF, the EMG of left triceps brachii muscle and PLRF neuronal discharges increased a few seconds after the injection and lasted for about 40 s. A burst EMG appeared twice from the r-biceps muscle, and was associated with suppression of r-triceps EMG. B: when 50 Hz stimulation was applied to the r-PLRF, EMGs were increased in amplitude from the ipsilateral flexor (r-biceps) and from the contralateral extensor (l-triceps), while the EMG from the ipsilateral extensor (r-triceps) was decreased. C: a histological sketch of location of the electrode tip.
Regions affected by glutamate injection were very narrow in the PLRE When the injection site was moved dorso-ventrally in the PLRF, the resulting phenomena varied in degree depending on the position of the injection site. The most effective region was about 0.2 mm within the PLRF, adjoining the medial aspect of the rubrospinal tract. If the electrode was moved upward or downward about 0.2 mm, no effect on spike frequency could be elicited by drug administration. This region corresponded to the region where stereotyped limb movements could clearly be obtained by electrical stimulation of the PLRF at low intensity. This means that regions in the PLRF which excite the contralaterai extensor are slightly narrower than implied by electrical stimulation (e.g. about 0.5 mm in diameter15). When glutamate was injected into the more lateral and caudal regions of the PLRF, adjoining the medial aspect of the trigeminal spinal tract nucleus, no increase in neuronal activity was observed. When large amounts of glutamate were administered into the medial regions of the PLRF, step-like alternating limb movements were occasionally evoked.
The latency, duration and degree of the glutamate effects on the reticular neuron activity and EMG activity of C-forelimb extensors varied according to the quantity and concentration of the solution. Injections of 1 mM glutamate proved ineffective in discernibly increasing neural discharges of the PLRF neurons and EMG activity of c-forelimb extensors, however, 10 mM glutamate clearly generated neural discharges and E M G activity of C-forelimb extensors. The latency of neural discharge increases after the injection was essentially constant with no significant differences observed between various concentrations and amounts of glutamate (10--100 mM). The duration of excitatory effects depended on the concentration and the amount of glutamate solution. For 10 mM solution (0.01 /~1), neural activity and EMG activity of C-forelimb extensors increased for only a few minutes. When 100 mM solution was injected (0.01/~1) to the PLRF, a longer period of up to 30 min was required for the recovery to the normal level of discharge and EMG activity of C-forelimb extensors. At this high concentration (100 mM) of solution administration, the rate of discharge was also higher than that obtained at lower concentrations.
285 Similar effects on reticular neuron activity and cforelimb extensor EMGs were observed upon administration of quisqualate (1.0-10 mM), aspartate (1.0-10 mM) and kainate (0.5-5 mM). The most effective drug was quisqualate, follcwed by kainate, glutamate and aspartate in that order. Each of the following drugs (concentrations 0.5-10 mM): carbachol, serotonin, dopamine, noradrenaline, G A B A and picrotoxin were injected individually into the PLRF, however, no increase in PLRF neuron activity and no movements of the bilateral forelimbs were observed.
Unit discharges of the PLRF neurons underlying extensor muscle activity of the contralateral forelimb The amplitudes and frequencies of unit discharges recorded from the PLRF varied between individual neurons. The majority of PLRF neurons showed spontaneous firing at low frequency and of about 100aV spike height, and the spikes were modulated in frequency in a manner associated with the limb movements. Unit spikes of the PLRF were increased in frequency when a small amount of glutamate was injected into the unilateral PLRF. Fig. 2 shows the unit discharge of PLRF neurons after injection of a small amount of quisqualate into I-PLRF, 3 times at 30 min intervals. Unit spikes appeared when 10 mM quisqualate was injected into the PLRF
with a slow (0.005 pl/s) rate of injection. In this case, about 0.05 BI of 10 mM quisqualate solution was required to produce firing. Spike frequency increased, reached a maximal frequency, then decreased, and firing stopped after about 20-30 s. Thirty minutes after cessation of firing, quisqualate was again injected in a similar manner to the first administration, and unit spikes also appeared several seconds following the start of the injection. As shown in Fig. 2-2, spike frequency was higher than for the first injection. Quisqualate was administered once more after a 30-min interval, and spikes appeared with even higher frequency and lower spike amplitudes (Fig. 2-3). This sensitization with repetitive induction of firing was also observed with glutamate. Dose-dependent behavior was observed when various concentrations and volumes of glutamate were injected into the PLRE With volumes of 10 mM glutamate solution from 0.01 to 0.5 BI, the following effects were observed on the time course of firing frequency. With large amounts (0.5 /~1) of glutamate, a higher spike frequency, prolongation of the firing period and a shortening of the latent period were evident. In contrast, with small amounts (less than 0.1 ~1), less firing was obtained, and it decayed more quickly. About one-half of PLRF neurons (37/78) examined showed increased firing frequency with glutamate (below
2
I1
3
quisqualic acid
2s
Fig. 2. Effects of quisqualate on PLRF neuronal activity. When a small amount (0.05/xl) of 10 mM quisqualate was injected into the PLRF with the micropipette, unit spikes were obtained several seconds after starting the injection. The response was obtained again by injections repeated at 30-min intervals, and showed a clear sensitization. The baseline activity also showed increases for each successive application.
286
B
A T
I-biceps
b.,:.;
" r-~
=~
r-triceps
b. ,J,l i::;:::'_:~:; C".:;?---::
. . . .
"~ ' " "
-7 ~_~
: ":~ . ~. I [ IOMV
r-biceps
b.
.....
'
~
[0.5mV
l
O.ImV
0.1mV
r-PLRF I
2s
I
r - P L R F stim.
,
5ms
,
Fig. 3. Interactions between PLRF neuron activity and EMGs of forelimb muscles. A: unit spikes were recorded from the PLRF on the right side together with EMGs of extensors (triceps) and flexors (biceps) in the bilateral forelimbs, when the neuronal firing rate was increased by administration of quisqualate to the right PLRE B: spike-triggered averaged EMGs by 80 spikes averaging. The averaged EMG was observed in the l-triceps brachii muscle with a latency of 11 ms. C: evoked EMGs following electrical stimulation of the r-PLRF with 10 times averaged. The evoked EMGs obtained from l-triceps and r-biceps brachii muscles had latencies of 6 and 7.5 ms, respectively.
0.1 M concentration) injection, while the remaining neurons were not affected by glutamate. However, at a high concentration (1 M) of glutamate a much greater proportion of neurons fired. As shown in Fig. 3B, several neurons of the P L R F discharged upon quisqualate injection in the P L R F at the same time, E M G s were increased in the extensor of the c-forelimb. A spike-triggered averaged E M G response was measured from the extensor of the c-forelimb with the P L R F neuron spike used for triggering signals. The compilation time of averaged E M G was different in individual neurons: the majority of neurons required an average of about 50 times, some required as much as 500 times. The spike-triggered averaged E M G was about 11 ms latency and about 30 MV in amplitude. A b o u t 80% of those P L R F neurons (30/37) which were responsive to glutamate had a discernible averaged extensor E M G in the c-forelimbs. Latencies of the averaged E M G varied for individual neuron spikes: the mean latency was 11.3 + 1.8 ms (+ S.D.), which ranged between 8 and 17 ms. In contrast, when electrical stimulation was applied to the right PLRF, an evoked E M G response was observed both from the ipsilateral flexor (right biceps) and the contra-
lateral extensor (left biceps) of 7.5 and 6 ms latency, respectively, as shown in Fig. 3C. The latency of electrically evoked contralateral extensor responses was constant even when the intensity and frequency of the P L R F stimulation was changed. It is noteworthy that an evoked E M G in the ipsilateral flexor was observed only by electrical stimulation, but was not obtained by chemical stimulation. Contralateral extensor responses had different latencies and amplitudes when evoked electrically and chemically (latencies were 6 and 11 ms, and amplitudes were 1 mV and 30 MV, respectively). The threshold for limb movements evoked by electrical stimulation of the P L R F was decreased when small amounts of glutamate were administered. For example, before administration of glutamate, the threshold was about 50 ktA, but decreased to about 30 # A (by 3 0 - 6 0 % ) after administration of glutamate. Upon bilateral electrical stimulation of the superficial radial nerve, an evoked potential was obtained from the PLRF, a positive spike-like potential with short latency, sharp negative potentials, prolonged negative deflection and late positivity. As shown in Fig. 4, when glutamate was applied to the PLRF, cutaneously evoked P L R F
287
A
) .
:
;
;
I
I
l
;
;
I
)
.
I-sup.R.N.stim.
.
.
.
.
.
.
.
I
I
!
5ms/div.
r-sup.R.N.stim.
Fig. 4. Effects of quisqualate administration to the I-PLRF on cutaneously evoked PLRF potentials. Electrical potentials were recorded from the I-PLRF following electrical stimulation applied to the superficial radial nerve of the left (A) and right (B) forelimb. The uppermost records (1) show potentials before administration of quisqualate into the I-PLRE while (2) is during administration, and (3) is 30 min after administration. Each record was averaged 10 times.
and then disappeared for several minutes. 1-NA-Spm is an analogue of JSTX-3, and is an antagonist of the action of glutamate 1. Fig. 5 shows unit discharges of PLRF neurons and evoked potentials of the PLRF following stimulation of the superficial radial nerves in bilateral forelimb, before, during and after 1-NA-Spm administration in the PLRE Spontaneous unit discharges were obtained and spike frequency was modulated with limb movements. The unit spike was related to the excitation of the extensor muscle in the contralateral forelimb to the
potentials were not markedly changed, except that the sharp negative potentials, evoked by stimulation of contralateral (right) forelimb, were prolonged and increased in amplitude.
Effects of 1-NA-Spm on P L R F neuron activity and leg movements When small amounts (0.01 ~1) of 2-10 mM 1-NA-Spm solution were injected into the unilateral PLRF, unit discharges of the PLRF neurons decreased in frequency
1-NA-Spm
after 25rain.
I-PLRF
Io.ImV 5$
B1
A1
I-sup.R.N.
II
2
,,u
- P''/
C1
/
t
Fig. 5. Effects of 1-NA-Spm in the PLRF upon unit spikes and cutaneously evoked PLRF potentials. Unit spikes were recorded from the 1-PLRF as well as evoked potentials produced by stimulation of the superficial radial nerve in the left (A1, B1, C1) and right forelimb (A2, B2, C2). When 1-NA-Spm (2 mM solution, 0.01 pl) was injected in the PLRF, unit spikes were abolished 20 s after starting the injection; at 25 min after cessation of the injection, spikes reappeared and evoked potentials were recovered.
288
I-triceps i-biceps
I'
i , , ,,
I ] ! , I
j,
r-trlcep .r-biceps
,
....... r
' '. : ~
I
,i~'"
~ ~.~T~
._i ,~=Li.
" Y w-y-~-y
-
= ~ VkJlL'k&i~[o.smv ~ w-r V I jlO.lmV
I-PLRF
1-NA-Spm
lOs
Fig. 6. Effects of 1-NA-Spm on EMG activity of forelimb muscles. EMGs were recorded from extensors and flexors in bilateral forelimbs and neuronal discharges were recorded from the PLRF before, during, and after administration of 1-NA-Spm into the I-PLRE In this preparation, glutamate was administered into the I-PLRF 30 min before administration of 1-NA-Spm. EMGs with elevated amplitude appeared in the bilateral forelimb extensors with 2 stepping movements of the right forelimb. When a small amount (0.05/d, 10 mM solution) of 1-NA-Spm was injected into the I-PLRF, PLRF neuronal discharges were decreased in amplitude and EMGs were suppressed in the contralateral (right) triceps and augmented in the ipsilateral (left) triceps. About 30 s, step-like EMGs appeared transiently, followed by continuous EMG activity. See text for details.
recording, since a spike-triggered averaged E M G could be obtained from the muscle. When a small amount (0.01-0.05/~1, 2-10 mM solution) of 1-NA-Spm was injected slowly into the I-PLRF, unit spikes were diminished and then abolished, and the background discharges of the PLRF were also depressed for several minutes. The duration of inhibition depended on the concentration and the amount of 1-NA-Spm. For 2 mM solution (0.01/ll), neural activity ceased for only a few minutes. When 10 mM solution was injected (0.01 /A) to the PLRF, a longer period of up to 40 min was required for recovery of unit spikes. At low concentrations (1-2 mM) large amounts were required for effects to appear on PLRF neuron activity. Cutaneously evoked PLRF potentials induced by superficial radial nerve stimulation at the forelimb bilaterally were depressed in amplitude by 1-NA-Spm administration. A positive spike-like potential with short latency, negative spike-like potentials, prolonged negative deflection and late positivity could be obtained by stimulation of the superficial radial nerve. The positive short-latency, spike-like potential and negative sharp potentials were depressed in amplitude after administration of 1-NA-Spm, and unit spikes were absent from the PLRF neurons (Fig. 5A-C). These changes showed recovery after several minutes. The threshold of PLRFinduced stereotaxic forelimb movements was increased slightly after 1-NA-Spm administration in the PLRF, e.g. from 35 to 45/~A. Forelimb movements occasionally appeared during 1-NA-Spm administration; the ipsilateral forelimb showed a slight increase in extension posture, while the
contralateral forelimb showed a relaxation of extension posture, with occasional flexion posture. E M G records of these effects were observable, but not constant and rather weak. The patterns and step cycle of stepping on a moving treadmill were not markedly changed by administration of 1-NA-Spm in the PLRF, but a slight decrease in the duration of the step cycle was observed. The suppression of the forelimb movements induced by 1-NA-Spm administration depended markedly on the pretreatment of glutamate administration to the P L R E Fig. 6 shows E M G activity and the PLRF neural discharge after administration of glutamate solution (0.1 M, 0.01 #1) into the left PLRF. Elevated neural discharges appeared in the PLRF and EMGs with higher amplitude were observed bilaterally in the forelimb extensors with occasional appearance of alternating forelimb movements. When a small amount (0.05/d, 10 mM solution) of 1-NA-Spm was injected into the I-PLRF through one barrel of the multibarreled glass micropipette, PLRF neural discharges decreased in amplitude and the background tonic E M G activity of the contralateral extensor (r-triceps) disappeared over 30 s. Then EMGs reappeared together with step-like forelimb movements. PLRF neural discharges recovered after a few minutes, with the duration of the suppression depending on the concentration and amount of 1-NA-Spm, as described above. DISCUSSION From the present experiments, it appears that the PLRF induces extension in the contralateral forelimb via
289 a crossed reticulospinal tract projecting to forelimb motoneurons. Several results support this conclusion. Electrical stimulation of the PLRF increases extensor muscle activity of the contralateral forelimb 15, and chemical substances, the putative transmitters glutamate and quisqualate, increase extensor activity in the contralateral forelimb when applied to the PLRF. The number of PLRF neuron spikes was increased by administration of either glutamate or quisqualate, and these changes paralleled an increase in extensor EMG of the cforelimb. From the averaged EMG response triggered by spikes in the PLRF, it appears that about 40% of examined PLRF neurons exert an excitatory effect on contralateral extensor motoneurons. About half of the neurons examined showed increased neural activity, and about 80% of these excited neurons showed averaged responses in the contralateral extensors. Some PLRF neurons were fired antidromically when the dorsal part of the ventrolateral funiculus was stimulated contralaterally at the upper cervical cord. This implies that some axons of PLRF neurons descend through the ventrolateral funiculus of the cervical cord 16. These results support the view that some PLRF neurons project to the cervical cord via crossed reticulospinal tracts. The presence of this descending pathway is supported by several other observations 4'6's'9'1~. For example, a partial transection of the cervical cord abolishes the EMG responses of the extensor muscles on the same side as the transection ~5. A crossed pontine reticulospinal tract is also suggested by experiments in which HRP is applied to the cervical cord and labeled cells appear in several regions of the pontine tegmentum contralateral to the HRP administration. This region has been called the ventro-lateral pontine tegmentum (VLPT) by Kuypers 8'9 and others 11. The VLPT may correspond physiologically to certain parts of the PLRF 4' 15. The crossed reticulospinal tract from the PLRF is independent of the rubrospinal tract 2'9, descending fibers from the mesencephalic tegmentum 2'1x and the reticulospinal tracts 1°-13'16. The reticulospinal tract may project to the extensor motoneurons in the c-forelimb indirectly via spinal interneurons, since the latency of the spike triggered averaged response was about 11 ms, which seems to be a slightly longer period than would be observed for a direct connection to the motoneurons. The latency of the electrically induced extensor muscle response was 6 ms, indicating that one or more interneurons are involved in the descending pathway from REFERENCES 1 Asami, T., Kagechika, H., Hashimoto, Y., Shudo, K., Miwa, A., Kawai, N. and Nakajima, T., Acylpolyamines mimic the action of Joro spider toxin (JSTX) on crustacean muscle glutamate receptors, Biomed. Res., 10 (1989) 185-189.
PLRF to spinal motoneurons. The crossed pontine reticulospinal tract is thought to play an important role in the coordination of stereotyped locomotion forelimb movements. Glutamate is an excitatory transmitter substance 7'19 and receptors are of 3 main subtypes: quisqualate, kainate and aspartate. In this series of experiments, quisqualate was the most effective for increasing the excitability of the PLRE Therefore, it seems likely that glutamate is a transmitter acting on PLRF neurons, and particularly on receptors of the quisqualate type. This idea is supported by the results using 1-NA-Spm, an analogue of the natural spider toxin, JSTX-31, which has an antagonistic action on glutamate receptors. Electrical stimulation applied to the PLRF obtained ipsilateral forelimb flexion and contralateral forelimb extension 15. Chemical stimulation of the PLRF gave rise only to the contralateral extension, while ipsilateral forelimb movement was weak and only occasionally observed. In other words, the contralateral forelimb extension could be produced by stimulation both electrically and chemically, however, the ipsilateral forelimb flexion was not generated by chemical stimulation. It is expected that chemicals selective for postsynaptic receptors would selectively affect neuron cell bodies, and it is probable that the observed responses were induced by increases in activity level of the neurons in the P L R E On the other hand, electrical stimulation could excite both somas and fibers of passage from anywhere in the central nervous system, which could converge in the PLRF and produce flexion of the forelimb on the ipsilateral side to stimulation. This possibility requires further investigation. When large amounts of glutamate were administered into the mesencephalic tegmentum, the more medial region of the PLRF, locomotor-like stepping was produced in this series of experiments. Similar results have been described by Noga et al. 1°, who observed that injection of glutamic acid (500 nM) into the magnocellular reticular formation produces locomotion. This implies that glutamate infusion may distribute over a wide area in the pontine reticular formation. Acknowledgements. The authors thank Prof. V.R. Edgerton and Dr. H. Robinson for their helpful comments on preparation of the manuscript, l-Naphthylacetylspermine was kindly supplied by Dr. N. Kawai. This research was supported by a Grant-in-Aid (60304044) for Scientific Research from the Japanese Ministry of Education, Science and Culture.
2 Baldissera, E, Lundberg, A. and Udo, M., Activity evoked from the mesencephalic tegmentum in descending pathways other than the rubrospinal tract, Exp. Brain Res., 15 (1972) 133-150. 3 Berman, A.L., The Brainstem of the Cat. A Cytoarchitectonic Atlas with Stereotaxic Coordinates, Univ. of Wisconsin Press, Madison, 1968.
290 4 Carlton, S.M., Chung, J.M., Leonard, R.B. and Willis, W.D., Funicular trajections of brainstem neurons projecting to the lumbar spinal cord in the monkey (Macaca fascicularis): a retrograde labeling study, J. Comp. Neurol., 241 (1985) 382404. 5 Garcia-Rill, E., Skinner, R.D. and Fitzgerald, J.A., Chemical activation of the mesencephalic locomotor region, Brain Research, 411 (1985) 43-54. 6 Goodchild, A.K., Dampney, R.A.L. and Bandler, R., A method for evoking physiological responses by stimulation of cell bodies, but not axons of passage, within localized regions of the central nervous system, J. Neurosci. Methods, 6 (1982) 351-363. 7 Hayashi, T., Effects of sodium glutamate on the nervous system, Keio J. Med., 3 (1952) 183-192. 8 Holstage, G. and Kuypers, H.G.J.M., The anatomy of brainstem pathways to the spinal cord in cat. A labeled amino acid tracing study, Prog. Brain Res., 57 (1982) 145-183. 9 Kuypers, H.G.J.M. and Maisky, N.A., Funicular trajectories of descending brainstem pathways in cat, Brain Research, 136 (1977) 159--165. 10 Noga, B.R., Kettler, J. and Jordan, L.M., Locomotion produced in mesencephalic cats by injections of putative transmitter substances and antagonists into the medial reticular formation and pontomedullary locomotor strip, J. Neurosci., 8 (1988) 2074-2086. 11 Nyberg-Hansen, R., Sites and mode of termination of reticulospinal fibers in the cat. An experimental study with silver impregnation methods, J. Comp. Neurol., 124 (1965) 71-100.
12 Orlovsky, G.N., Work of the reticulospinal neurons during locomotion, Biophysics, 15 (1970) 761-771. 13 Orlovsky, G.N., The effects of different descending systems of flexor and extensor activity during locomotion, Brain Research, 40(1972) 359-371. 14 Shimamura, M., Kogure, I. and Fuwa, T., Discharge patterns of reticulospinal neurons corresponding with quadrupedal leg movements in thalamic cats, Brain Research, 260 (1983) 27-34. 15 Shimamura, M., Kogure, I. and Fuwa, T., The role of the paralemniscal pontine reticular formation in forelimb stepping of thalamic cats, Neurosci. Res., 1 (1984) 393-410. 16 Shimamura, M., Fuwa, T. and Kogure, I., Burst discharges of pontine reticular neurons in relation to forelimb stepping of thalamic and high spinal cats, Brain Research, 346 (1985) 363-367. 17 Shimamura, M., Kogure, I. and Fuwa, T., Locomotion of forelimbs in thalamic cats: joint afferent and supraspinal contributions. In S. Grillner et al. (Eds.), Neurobiology of Vertebrate Locomotion, Macmillan, England, 1986. 18 Shimamura, M., Tanaka, I. and Fuwa, T., Comparison between spino-bulbo-spinal and propriospinal reflexes in thalamic cats during stepping, Neurosci. Res., 7 (1990) 358-368. 19 Watkins, L.R., Griffin, G.R., Leichnetz, G.R. and Mayer, D.J., The somatotopic organization of the NRM and surrounding brainstem structures as revealed by HRP slow-release gels, Brain Research, 181 (1980) 1-15. 20 Zieglg~insberger,W. and Puil, E.A., Actions of glutamic acid on spinal neurons, Exp. Brain Res., 17 (1973) 35-49.
282
Elsevier BRES 15757
Crossed forelimb extension produced in thalamic cats by injection of putative transmitter substances into the paralemniscal pontine reticular formation Muneo Shimamura, Tatsu Fuwa and Ikuko Tanaka Department of Neurophysiology, Tokyo Metropolitan Institutefor Neurosciences, Fuchu-city, Tokyo (Japan) (Accepted 20 February 1990)
Key words: Reticulospinal tract neuron; Extensor muscle; Glutamate; Forelimb movement; Thalamic cat; Crossed limb extension
To analyze the descending pathways of the paralemniscal pontine reticular formation (PLRF), a technique was used for the selective activation of cell bodies by localized injection of putative neurotransmitters in the PLRF. When a small amount (less than 0.1 pl) of 0.1 M glutamate was injected into the PLRF unilaterally in thalamic cats, the forelimb contralateral (c-forelimb) to the injection was extended, and occasionally the ipsilateral forelimb was flexed. These responses were similar to those obtained by electrical stimulation of the PLRF, but were relatively weaker. Unit spikes of PLRF neurons were increased in frequency following administration of glutamate. The latent periods and durations of increases in spike frequency varied depending on the concentration and quantity of the glutamate solution, and were roughly similar to those of the extensor EMG in the c-forelimb. Since the firing of PLRF neurons preceded the EMG with 11 ms latency, the unit spike of PLRF neurons could be used as a triggering signal to observe a spike triggered averaged EMG response in the extensor muscle of the c-forelimb. Results similar to those with glutamate were observed upon administration of quisqualate, kainate and aspartate. The most effective compound was quisqualate. Application to the PLRF of 1-naphthylacetyl spermine (1-NA-Spm), an analogue of the natural spider toxin JSTX-3 and an antagonist of glutamate, suppressed both the PLRF neuron activity and the extensor EMG of the c-forelimb. These observations suggest that extensor muscles of the forelimb are excited by the contralateral PLRF, perhaps via the crossed reticulospinal tract from the PLRF. PLRF neurons may be activated by glutamate (quisqualate) receptors. INTRODUCTION Recently we have r e p o r t e d that the paralemniscal pontine reticular formation ( P L R F ) exerts a bilateral control of forelimb muscles and plays a role in initiating and maintaining forelimb stepping; the P L R F receives afferent input bilaterally from the forelimb nerves, and electrical stimulation of the P L R F in thalamic cats causes s t e r e o t y p e d locomotor-like forelimb movements ~5. The forelimb ipsilateral to stimulation (i-forelimb) is flexed while the contralateral forelimb (c-forelimb) is extended. This m a y be caused by impulses descending through the reticulospinal tracts directly and/or indirectly to spinal m o t o r pools. Unilateral transection of the ventral quadrant at the cervical cord (C3) abolished PLRF-elicited flexor E M G s and associated extensor E M G suppression ipsilateral to the transection, while the side contralateral to the transection r e m a i n e d unchanged 15. Nevertheless, there has been no detailed analysis of the relationship b e t w e e n reticulospinal tracts from P L R F and s t e r e o t y p e d forelimb movements. In o r d e r to confirm the presence in the P L R F of neuron somata capable of producing
s t e r e o t y p e d forelimb m o v e m e n t s when stimulated and to map their distribution in the brainstem, we used a technique for selective activation of cell bodies by the localized injection of small quantities of putative neurotransmitters 6. In this p a p e r , we d e m o n s t r a t e that the selective activation of cell bodies in the P L R F by a variety of neuroactive substances causes c-forelimb extension in thalamic cats. To analyze the effects of drug administration, we used several putative transmitter substances, glutamate, quisqualate, kainate, aspartate, and the glutamate antagonist, 1-naphthylacetyl-spermine (1-NASpm), and p e r f o r m e d kinematic analyses of forelimb movements, bilateral E M G recordings in forelimb extensor and flexor muscles, and recordings of cutaneous e v o k e d P L R F potentials, P L R F - i n d u c e d forelimb muscle responses and unit discharges in the P L R F neurons.
MATERIALS AND METHODS Twenty-five adults cats (2.3-3.0 kg b. wt.) were studied. A detailed description of the animal preparation has been reported
Correspondence: M. Shimamura, Department of Neurophysiology, Tokyo Metropolitan Institute for Neurosciences, 2-6 Musashidai, Fuchu-city, Tokyo 183, Japan. 0006-8993/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
283 elsewhere 15. Under ether anesthesia, tracheal cannulation, decerebration at the stereotaxic A12 level and spinal transection at the T H segment were performed. Immediately after the decerebration, ether was withdrawn. The head was fixed in a stereotaxic device, the T 2 spinal process clamped to a metal frame and the lumbar region suspended by a hammock so that all limbs touched the floor of the treadmill belt. Electrical stimulation was applied to the PLRF, using a coaxial needle electrode and a glass-coated tungsten microelectrode. Pulses of 0.1-0.3 ms in duration, 50 Hz in frequency and of less than 50 p A intensity were used. The threshold of electrical stimulation for evoking EMGs of forelimb muscles was measured. Evoked potentials were recorded from the PLRF following stimulation of the superficial radial nerve with a cuff electrodeTM. EMGs were recorded on a polygraph using appropriate amplifiers. The electrodes consisted of paired enamel-coated copper wire inserted bilaterally into extensors (m. triceps brachii) and flexors (m. biceps brachii) of the forelimb. Chemical stimulation of the PLRF was achieved by administering substances by pressure microinjection, using a multi- (mostly 6) or double-barreled glass micropipene. In one barrel, a tungsten micro-wire (tip 10 pm diameter) was inserted for recording of neural activity and for electrical stimulation of the PLRF while the other barrel (tip about 6 /tin diameter) was filled with a solution containing one of the following: glutamate (sodium L-glutamate, mono), quisqualic acid (fl-(3,5-dioxo-l,2,4-oxadiazolidin-2-yl)-Lalanine), aspartate (L-aspartic acid), kainic acid, serotonin (serotonin-creatinine sulfate), noradrenaline (4-(2-amino-l-hydroxyethyl)1,2-benzene-diol), dopamine (3,4-dihydroxyphenyl-ethylamine), GABA (4-amino-N-butyric acid), carbachol, picrotoxin and 1naphthylacetyl spermine (1-NA-Spm). These drugs were dissolved in normal saline except for noradrenaline and dopamine which were dissolved in 5 x 10-5 M ascorbic acid in D-H20 to prevent oxidation. Every drug was used at concentrations between 1-100 raM. Concentration ranges were chosen on the basis of previous successful experiments using drug injections to initiate locomotion5' lo. The drugs were pressure injected at a rate of 0.06 pl/min with the micropipette attached to a micrometer-driven oil pressure pump with polyethylene tubing. We used a multi-barreled glass micropipene for examination of the approximate responsiveness of PLRF neuron activity to a variety of solutions, and a double-barreled one for detailed analysis of the effects of each solution, because of possible cross-contamination between barrels. When chemical substances were injected into the PLRF, the following phenomena were used for monitoring the effects of drug administration; unit discharges of PLRF neurons, leg movements observed directly and through EMG recordings, cutaneous evoked PLRF potentials and the threshold PLRF stimulation for inducing stereotyped leg movement. In several cases, unit discharges were identified as antidromic spikes from the spinal cord in reticulospinal neurons, recorded upon electrical stimulation of the lateral funiculus of the cervical cord at the 3rd and 7th segments bilaterally. To identify interactions between the unit spike of PLRF neurons and the EMG of forelimb muscles, a spike-triggered averaged EMG was constructed by computer. Leg movements were monitored by a TV recording system which ran synchronously with the EMG recording system. EMGs were recorded from the triceps and biceps brachii muscles of bilateral forelimbs. Cutaneously evoked PLRF potentials were recorded by the glass-coated tungsten microelectrode following stimulation of the superficial radial nerve. Stimulation and recordings of electrical activity were started 2 h after cessation of ether anesthesia. After electrophysiologicaland pharmacological observations were completed, a DC current was applied to the electrode appropriate to create a small lesion. The cat was then injected with a large dose of pentobarbitai sodium. The brain was removed and 100-pm transverse frozen sections were cut to identify the position of the tip of the electrode, The location of the electrode tip in the brainstem was plotted in relation to the anatomical features described in the atlas of Berman 3.
RESULTS
Forelimb movements elicited by injection of glutamate into the PLRF The location of the P L R F has been r e p o r t e d 15 as P 2.0-3.5 mm, 3 . 0 - 4 . 0 m m lateral from the midline, and 5.5-6.5 mm in d e p t h from the surface of the I V ventricle. The physiological criteria for the identification of P L R F are the followingl~: neuronal discharges of the P L R F are increased by limb m o v e m e n t and by touching the skin over the forelimbs bilaterally, particularly on the forelimb contralateral to discharge recording. Electrical stimulation of the P L R F causes s t e r e o t y p e d forelimb movements; the i-forelimb is flexed and the c-forelimb extended; E M G activity a p p e a r s in the flexor and is suppressed in the extensor of the i-forelimb, while the E M G of the c-forelimb extensor was increased. W h e n a small a m o u n t (less than 0.1 /d) of 0.1 M glutamate solution was a d m i n i s t e r e d from one barrel of the d o u b l e - b a r r e l e d glass m i c r o p i p e t t e unilaterally to the P L R F of the thalamic cats, s t e r e o t y p e d forelimb movements were e v o k e d , such that the c-forelimb increased the extension posture while the i-forelimb was lifted slightly, similar to results o b t a i n e d by electrical stimulation of the PLRF. The extensor E M G of the c-forelimb was increased while E M G a p p e a r e d in the ipsilateral flexor and was suppressed in the extensor of the i-forelimb. A s shown in Fig. 1A, a small a m o u n t of quisqualate (10 m M in saline solution, 0.01 /A total volume applied over 5 s) was injected into the right P L R E A few seconds after starting the injection, neuronal discharges from the P L R F a d j a c e n t to the region of microinjection of quisqualate were increased. The extensor E M G s in the c-forelimb (left triceps) were also increased simultaneously with the increase in the P L R F neuron activity. T h e flexor E M G of the i-forelimb (right biceps), on the o t h e r h a n d , a p p e a r e d at once and only for a brief period, and the extensor E M G was suppressed in the i-forelimb (right triceps). A f t e r cessation of injection, neuronal discharges were still increased in frequency and reached maximal frequency after 20-30 s, then decreased and recovered to the control level. During increase of neuronal discharge, E M G amplitudes were still elevated in the left triceps brachii muscles, while E M G amplitude in the triceps brachii muscle of the right forelimb decreased, and an E M G a p p e a r e d transiently in the biceps brachii muscle of the right forelimb. Fig. 1B shows E M G s of extensors and flexors in the forelimb bilaterally induced by electrical stimulation of the right P L R F with 50 I-Iz frequency and 3 5 / t A intensity using a glass-coated tungsten microelectrode. E M G s were markedly increased in the left triceps, and E M G activity a p p e a r e d in the right biceps but was suppressed in the right triceps.
284
A I l-biceps
,ii,,=,=J............................................................................. J ............I
r-triceps
il
:~:::
r-biceps
~
I
IO.5mV
~, . . . . .
0.2mV
r-PLRF ls
lOmM Qulsqulnc acid
C
.....
i................... ii
elect.stim. 2s
Fig. 1. Effects of application of quisqualic acid in the PLRF on EMGs of forelimb muscles and on PLRF neuronal activity. EMGs were recorded bilaterally from the biceps and triceps brachii muscles and unit spikes of the right PLRF neuron were recorded at the same time. A: when a small amount (0.01/A) of 10 mM quisqualic acid solution was injected into the right PLRF, the EMG of left triceps brachii muscle and PLRF neuronal discharges increased a few seconds after the injection and lasted for about 40 s. A burst EMG appeared twice from the r-biceps muscle, and was associated with suppression of r-triceps EMG. B: when 50 Hz stimulation was applied to the r-PLRF, EMGs were increased in amplitude from the ipsilateral flexor (r-biceps) and from the contralateral extensor (l-triceps), while the EMG from the ipsilateral extensor (r-triceps) was decreased. C: a histological sketch of location of the electrode tip.
Regions affected by glutamate injection were very narrow in the PLRE When the injection site was moved dorso-ventrally in the PLRF, the resulting phenomena varied in degree depending on the position of the injection site. The most effective region was about 0.2 mm within the PLRF, adjoining the medial aspect of the rubrospinal tract. If the electrode was moved upward or downward about 0.2 mm, no effect on spike frequency could be elicited by drug administration. This region corresponded to the region where stereotyped limb movements could clearly be obtained by electrical stimulation of the PLRF at low intensity. This means that regions in the PLRF which excite the contralaterai extensor are slightly narrower than implied by electrical stimulation (e.g. about 0.5 mm in diameter15). When glutamate was injected into the more lateral and caudal regions of the PLRF, adjoining the medial aspect of the trigeminal spinal tract nucleus, no increase in neuronal activity was observed. When large amounts of glutamate were administered into the medial regions of the PLRF, step-like alternating limb movements were occasionally evoked.
The latency, duration and degree of the glutamate effects on the reticular neuron activity and EMG activity of C-forelimb extensors varied according to the quantity and concentration of the solution. Injections of 1 mM glutamate proved ineffective in discernibly increasing neural discharges of the PLRF neurons and EMG activity of c-forelimb extensors, however, 10 mM glutamate clearly generated neural discharges and E M G activity of C-forelimb extensors. The latency of neural discharge increases after the injection was essentially constant with no significant differences observed between various concentrations and amounts of glutamate (10--100 mM). The duration of excitatory effects depended on the concentration and the amount of glutamate solution. For 10 mM solution (0.01 /~1), neural activity and EMG activity of C-forelimb extensors increased for only a few minutes. When 100 mM solution was injected (0.01/~1) to the PLRF, a longer period of up to 30 min was required for the recovery to the normal level of discharge and EMG activity of C-forelimb extensors. At this high concentration (100 mM) of solution administration, the rate of discharge was also higher than that obtained at lower concentrations.
285 Similar effects on reticular neuron activity and cforelimb extensor EMGs were observed upon administration of quisqualate (1.0-10 mM), aspartate (1.0-10 mM) and kainate (0.5-5 mM). The most effective drug was quisqualate, follcwed by kainate, glutamate and aspartate in that order. Each of the following drugs (concentrations 0.5-10 mM): carbachol, serotonin, dopamine, noradrenaline, G A B A and picrotoxin were injected individually into the PLRF, however, no increase in PLRF neuron activity and no movements of the bilateral forelimbs were observed.
Unit discharges of the PLRF neurons underlying extensor muscle activity of the contralateral forelimb The amplitudes and frequencies of unit discharges recorded from the PLRF varied between individual neurons. The majority of PLRF neurons showed spontaneous firing at low frequency and of about 100aV spike height, and the spikes were modulated in frequency in a manner associated with the limb movements. Unit spikes of the PLRF were increased in frequency when a small amount of glutamate was injected into the unilateral PLRF. Fig. 2 shows the unit discharge of PLRF neurons after injection of a small amount of quisqualate into I-PLRF, 3 times at 30 min intervals. Unit spikes appeared when 10 mM quisqualate was injected into the PLRF
with a slow (0.005 pl/s) rate of injection. In this case, about 0.05 BI of 10 mM quisqualate solution was required to produce firing. Spike frequency increased, reached a maximal frequency, then decreased, and firing stopped after about 20-30 s. Thirty minutes after cessation of firing, quisqualate was again injected in a similar manner to the first administration, and unit spikes also appeared several seconds following the start of the injection. As shown in Fig. 2-2, spike frequency was higher than for the first injection. Quisqualate was administered once more after a 30-min interval, and spikes appeared with even higher frequency and lower spike amplitudes (Fig. 2-3). This sensitization with repetitive induction of firing was also observed with glutamate. Dose-dependent behavior was observed when various concentrations and volumes of glutamate were injected into the PLRE With volumes of 10 mM glutamate solution from 0.01 to 0.5 BI, the following effects were observed on the time course of firing frequency. With large amounts (0.5 /~1) of glutamate, a higher spike frequency, prolongation of the firing period and a shortening of the latent period were evident. In contrast, with small amounts (less than 0.1 ~1), less firing was obtained, and it decayed more quickly. About one-half of PLRF neurons (37/78) examined showed increased firing frequency with glutamate (below
2
I1
3
quisqualic acid
2s
Fig. 2. Effects of quisqualate on PLRF neuronal activity. When a small amount (0.05/xl) of 10 mM quisqualate was injected into the PLRF with the micropipette, unit spikes were obtained several seconds after starting the injection. The response was obtained again by injections repeated at 30-min intervals, and showed a clear sensitization. The baseline activity also showed increases for each successive application.
286
B
A T
I-biceps
b.,:.;
" r-~
=~
r-triceps
b. ,J,l i::;:::'_:~:; C".:;?---::
. . . .
"~ ' " "
-7 ~_~
: ":~ . ~. I [ IOMV
r-biceps
b.
.....
'
~
[0.5mV
l
O.ImV
0.1mV
r-PLRF I
2s
I
r - P L R F stim.
,
5ms
,
Fig. 3. Interactions between PLRF neuron activity and EMGs of forelimb muscles. A: unit spikes were recorded from the PLRF on the right side together with EMGs of extensors (triceps) and flexors (biceps) in the bilateral forelimbs, when the neuronal firing rate was increased by administration of quisqualate to the right PLRE B: spike-triggered averaged EMGs by 80 spikes averaging. The averaged EMG was observed in the l-triceps brachii muscle with a latency of 11 ms. C: evoked EMGs following electrical stimulation of the r-PLRF with 10 times averaged. The evoked EMGs obtained from l-triceps and r-biceps brachii muscles had latencies of 6 and 7.5 ms, respectively.
0.1 M concentration) injection, while the remaining neurons were not affected by glutamate. However, at a high concentration (1 M) of glutamate a much greater proportion of neurons fired. As shown in Fig. 3B, several neurons of the P L R F discharged upon quisqualate injection in the P L R F at the same time, E M G s were increased in the extensor of the c-forelimb. A spike-triggered averaged E M G response was measured from the extensor of the c-forelimb with the P L R F neuron spike used for triggering signals. The compilation time of averaged E M G was different in individual neurons: the majority of neurons required an average of about 50 times, some required as much as 500 times. The spike-triggered averaged E M G was about 11 ms latency and about 30 MV in amplitude. A b o u t 80% of those P L R F neurons (30/37) which were responsive to glutamate had a discernible averaged extensor E M G in the c-forelimbs. Latencies of the averaged E M G varied for individual neuron spikes: the mean latency was 11.3 + 1.8 ms (+ S.D.), which ranged between 8 and 17 ms. In contrast, when electrical stimulation was applied to the right PLRF, an evoked E M G response was observed both from the ipsilateral flexor (right biceps) and the contra-
lateral extensor (left biceps) of 7.5 and 6 ms latency, respectively, as shown in Fig. 3C. The latency of electrically evoked contralateral extensor responses was constant even when the intensity and frequency of the P L R F stimulation was changed. It is noteworthy that an evoked E M G in the ipsilateral flexor was observed only by electrical stimulation, but was not obtained by chemical stimulation. Contralateral extensor responses had different latencies and amplitudes when evoked electrically and chemically (latencies were 6 and 11 ms, and amplitudes were 1 mV and 30 MV, respectively). The threshold for limb movements evoked by electrical stimulation of the P L R F was decreased when small amounts of glutamate were administered. For example, before administration of glutamate, the threshold was about 50 ktA, but decreased to about 30 # A (by 3 0 - 6 0 % ) after administration of glutamate. Upon bilateral electrical stimulation of the superficial radial nerve, an evoked potential was obtained from the PLRF, a positive spike-like potential with short latency, sharp negative potentials, prolonged negative deflection and late positivity. As shown in Fig. 4, when glutamate was applied to the PLRF, cutaneously evoked P L R F
287
A
) .
:
;
;
I
I
l
;
;
I
)
.
I-sup.R.N.stim.
.
.
.
.
.
.
.
I
I
!
5ms/div.
r-sup.R.N.stim.
Fig. 4. Effects of quisqualate administration to the I-PLRF on cutaneously evoked PLRF potentials. Electrical potentials were recorded from the I-PLRF following electrical stimulation applied to the superficial radial nerve of the left (A) and right (B) forelimb. The uppermost records (1) show potentials before administration of quisqualate into the I-PLRE while (2) is during administration, and (3) is 30 min after administration. Each record was averaged 10 times.
and then disappeared for several minutes. 1-NA-Spm is an analogue of JSTX-3, and is an antagonist of the action of glutamate 1. Fig. 5 shows unit discharges of PLRF neurons and evoked potentials of the PLRF following stimulation of the superficial radial nerves in bilateral forelimb, before, during and after 1-NA-Spm administration in the PLRE Spontaneous unit discharges were obtained and spike frequency was modulated with limb movements. The unit spike was related to the excitation of the extensor muscle in the contralateral forelimb to the
potentials were not markedly changed, except that the sharp negative potentials, evoked by stimulation of contralateral (right) forelimb, were prolonged and increased in amplitude.
Effects of 1-NA-Spm on P L R F neuron activity and leg movements When small amounts (0.01 ~1) of 2-10 mM 1-NA-Spm solution were injected into the unilateral PLRF, unit discharges of the PLRF neurons decreased in frequency
1-NA-Spm
after 25rain.
I-PLRF
Io.ImV 5$
B1
A1
I-sup.R.N.
II
2
,,u
- P''/
C1
/
t
Fig. 5. Effects of 1-NA-Spm in the PLRF upon unit spikes and cutaneously evoked PLRF potentials. Unit spikes were recorded from the 1-PLRF as well as evoked potentials produced by stimulation of the superficial radial nerve in the left (A1, B1, C1) and right forelimb (A2, B2, C2). When 1-NA-Spm (2 mM solution, 0.01 pl) was injected in the PLRF, unit spikes were abolished 20 s after starting the injection; at 25 min after cessation of the injection, spikes reappeared and evoked potentials were recovered.
288
I-triceps i-biceps
I'
i , , ,,
I ] ! , I
j,
r-trlcep .r-biceps
,
....... r
' '. : ~
I
,i~'"
~ ~.~T~
._i ,~=Li.
" Y w-y-~-y
-
= ~ VkJlL'k&i~[o.smv ~ w-r V I jlO.lmV
I-PLRF
1-NA-Spm
lOs
Fig. 6. Effects of 1-NA-Spm on EMG activity of forelimb muscles. EMGs were recorded from extensors and flexors in bilateral forelimbs and neuronal discharges were recorded from the PLRF before, during, and after administration of 1-NA-Spm into the I-PLRE In this preparation, glutamate was administered into the I-PLRF 30 min before administration of 1-NA-Spm. EMGs with elevated amplitude appeared in the bilateral forelimb extensors with 2 stepping movements of the right forelimb. When a small amount (0.05/d, 10 mM solution) of 1-NA-Spm was injected into the I-PLRF, PLRF neuronal discharges were decreased in amplitude and EMGs were suppressed in the contralateral (right) triceps and augmented in the ipsilateral (left) triceps. About 30 s, step-like EMGs appeared transiently, followed by continuous EMG activity. See text for details.
recording, since a spike-triggered averaged E M G could be obtained from the muscle. When a small amount (0.01-0.05/~1, 2-10 mM solution) of 1-NA-Spm was injected slowly into the I-PLRF, unit spikes were diminished and then abolished, and the background discharges of the PLRF were also depressed for several minutes. The duration of inhibition depended on the concentration and the amount of 1-NA-Spm. For 2 mM solution (0.01/ll), neural activity ceased for only a few minutes. When 10 mM solution was injected (0.01 /A) to the PLRF, a longer period of up to 40 min was required for recovery of unit spikes. At low concentrations (1-2 mM) large amounts were required for effects to appear on PLRF neuron activity. Cutaneously evoked PLRF potentials induced by superficial radial nerve stimulation at the forelimb bilaterally were depressed in amplitude by 1-NA-Spm administration. A positive spike-like potential with short latency, negative spike-like potentials, prolonged negative deflection and late positivity could be obtained by stimulation of the superficial radial nerve. The positive short-latency, spike-like potential and negative sharp potentials were depressed in amplitude after administration of 1-NA-Spm, and unit spikes were absent from the PLRF neurons (Fig. 5A-C). These changes showed recovery after several minutes. The threshold of PLRFinduced stereotaxic forelimb movements was increased slightly after 1-NA-Spm administration in the PLRF, e.g. from 35 to 45/~A. Forelimb movements occasionally appeared during 1-NA-Spm administration; the ipsilateral forelimb showed a slight increase in extension posture, while the
contralateral forelimb showed a relaxation of extension posture, with occasional flexion posture. E M G records of these effects were observable, but not constant and rather weak. The patterns and step cycle of stepping on a moving treadmill were not markedly changed by administration of 1-NA-Spm in the PLRF, but a slight decrease in the duration of the step cycle was observed. The suppression of the forelimb movements induced by 1-NA-Spm administration depended markedly on the pretreatment of glutamate administration to the P L R E Fig. 6 shows E M G activity and the PLRF neural discharge after administration of glutamate solution (0.1 M, 0.01 #1) into the left PLRF. Elevated neural discharges appeared in the PLRF and EMGs with higher amplitude were observed bilaterally in the forelimb extensors with occasional appearance of alternating forelimb movements. When a small amount (0.05/d, 10 mM solution) of 1-NA-Spm was injected into the I-PLRF through one barrel of the multibarreled glass micropipette, PLRF neural discharges decreased in amplitude and the background tonic E M G activity of the contralateral extensor (r-triceps) disappeared over 30 s. Then EMGs reappeared together with step-like forelimb movements. PLRF neural discharges recovered after a few minutes, with the duration of the suppression depending on the concentration and amount of 1-NA-Spm, as described above. DISCUSSION From the present experiments, it appears that the PLRF induces extension in the contralateral forelimb via
289 a crossed reticulospinal tract projecting to forelimb motoneurons. Several results support this conclusion. Electrical stimulation of the PLRF increases extensor muscle activity of the contralateral forelimb 15, and chemical substances, the putative transmitters glutamate and quisqualate, increase extensor activity in the contralateral forelimb when applied to the PLRF. The number of PLRF neuron spikes was increased by administration of either glutamate or quisqualate, and these changes paralleled an increase in extensor EMG of the cforelimb. From the averaged EMG response triggered by spikes in the PLRF, it appears that about 40% of examined PLRF neurons exert an excitatory effect on contralateral extensor motoneurons. About half of the neurons examined showed increased neural activity, and about 80% of these excited neurons showed averaged responses in the contralateral extensors. Some PLRF neurons were fired antidromically when the dorsal part of the ventrolateral funiculus was stimulated contralaterally at the upper cervical cord. This implies that some axons of PLRF neurons descend through the ventrolateral funiculus of the cervical cord 16. These results support the view that some PLRF neurons project to the cervical cord via crossed reticulospinal tracts. The presence of this descending pathway is supported by several other observations 4'6's'9'1~. For example, a partial transection of the cervical cord abolishes the EMG responses of the extensor muscles on the same side as the transection ~5. A crossed pontine reticulospinal tract is also suggested by experiments in which HRP is applied to the cervical cord and labeled cells appear in several regions of the pontine tegmentum contralateral to the HRP administration. This region has been called the ventro-lateral pontine tegmentum (VLPT) by Kuypers 8'9 and others 11. The VLPT may correspond physiologically to certain parts of the PLRF 4' 15. The crossed reticulospinal tract from the PLRF is independent of the rubrospinal tract 2'9, descending fibers from the mesencephalic tegmentum 2'1x and the reticulospinal tracts 1°-13'16. The reticulospinal tract may project to the extensor motoneurons in the c-forelimb indirectly via spinal interneurons, since the latency of the spike triggered averaged response was about 11 ms, which seems to be a slightly longer period than would be observed for a direct connection to the motoneurons. The latency of the electrically induced extensor muscle response was 6 ms, indicating that one or more interneurons are involved in the descending pathway from REFERENCES 1 Asami, T., Kagechika, H., Hashimoto, Y., Shudo, K., Miwa, A., Kawai, N. and Nakajima, T., Acylpolyamines mimic the action of Joro spider toxin (JSTX) on crustacean muscle glutamate receptors, Biomed. Res., 10 (1989) 185-189.
PLRF to spinal motoneurons. The crossed pontine reticulospinal tract is thought to play an important role in the coordination of stereotyped locomotion forelimb movements. Glutamate is an excitatory transmitter substance 7'19 and receptors are of 3 main subtypes: quisqualate, kainate and aspartate. In this series of experiments, quisqualate was the most effective for increasing the excitability of the PLRE Therefore, it seems likely that glutamate is a transmitter acting on PLRF neurons, and particularly on receptors of the quisqualate type. This idea is supported by the results using 1-NA-Spm, an analogue of the natural spider toxin, JSTX-31, which has an antagonistic action on glutamate receptors. Electrical stimulation applied to the PLRF obtained ipsilateral forelimb flexion and contralateral forelimb extension 15. Chemical stimulation of the PLRF gave rise only to the contralateral extension, while ipsilateral forelimb movement was weak and only occasionally observed. In other words, the contralateral forelimb extension could be produced by stimulation both electrically and chemically, however, the ipsilateral forelimb flexion was not generated by chemical stimulation. It is expected that chemicals selective for postsynaptic receptors would selectively affect neuron cell bodies, and it is probable that the observed responses were induced by increases in activity level of the neurons in the P L R E On the other hand, electrical stimulation could excite both somas and fibers of passage from anywhere in the central nervous system, which could converge in the PLRF and produce flexion of the forelimb on the ipsilateral side to stimulation. This possibility requires further investigation. When large amounts of glutamate were administered into the mesencephalic tegmentum, the more medial region of the PLRF, locomotor-like stepping was produced in this series of experiments. Similar results have been described by Noga et al. 1°, who observed that injection of glutamic acid (500 nM) into the magnocellular reticular formation produces locomotion. This implies that glutamate infusion may distribute over a wide area in the pontine reticular formation. Acknowledgements. The authors thank Prof. V.R. Edgerton and Dr. H. Robinson for their helpful comments on preparation of the manuscript, l-Naphthylacetylspermine was kindly supplied by Dr. N. Kawai. This research was supported by a Grant-in-Aid (60304044) for Scientific Research from the Japanese Ministry of Education, Science and Culture.
2 Baldissera, E, Lundberg, A. and Udo, M., Activity evoked from the mesencephalic tegmentum in descending pathways other than the rubrospinal tract, Exp. Brain Res., 15 (1972) 133-150. 3 Berman, A.L., The Brainstem of the Cat. A Cytoarchitectonic Atlas with Stereotaxic Coordinates, Univ. of Wisconsin Press, Madison, 1968.
290 4 Carlton, S.M., Chung, J.M., Leonard, R.B. and Willis, W.D., Funicular trajections of brainstem neurons projecting to the lumbar spinal cord in the monkey (Macaca fascicularis): a retrograde labeling study, J. Comp. Neurol., 241 (1985) 382404. 5 Garcia-Rill, E., Skinner, R.D. and Fitzgerald, J.A., Chemical activation of the mesencephalic locomotor region, Brain Research, 411 (1985) 43-54. 6 Goodchild, A.K., Dampney, R.A.L. and Bandler, R., A method for evoking physiological responses by stimulation of cell bodies, but not axons of passage, within localized regions of the central nervous system, J. Neurosci. Methods, 6 (1982) 351-363. 7 Hayashi, T., Effects of sodium glutamate on the nervous system, Keio J. Med., 3 (1952) 183-192. 8 Holstage, G. and Kuypers, H.G.J.M., The anatomy of brainstem pathways to the spinal cord in cat. A labeled amino acid tracing study, Prog. Brain Res., 57 (1982) 145-183. 9 Kuypers, H.G.J.M. and Maisky, N.A., Funicular trajectories of descending brainstem pathways in cat, Brain Research, 136 (1977) 159--165. 10 Noga, B.R., Kettler, J. and Jordan, L.M., Locomotion produced in mesencephalic cats by injections of putative transmitter substances and antagonists into the medial reticular formation and pontomedullary locomotor strip, J. Neurosci., 8 (1988) 2074-2086. 11 Nyberg-Hansen, R., Sites and mode of termination of reticulospinal fibers in the cat. An experimental study with silver impregnation methods, J. Comp. Neurol., 124 (1965) 71-100.
12 Orlovsky, G.N., Work of the reticulospinal neurons during locomotion, Biophysics, 15 (1970) 761-771. 13 Orlovsky, G.N., The effects of different descending systems of flexor and extensor activity during locomotion, Brain Research, 40(1972) 359-371. 14 Shimamura, M., Kogure, I. and Fuwa, T., Discharge patterns of reticulospinal neurons corresponding with quadrupedal leg movements in thalamic cats, Brain Research, 260 (1983) 27-34. 15 Shimamura, M., Kogure, I. and Fuwa, T., The role of the paralemniscal pontine reticular formation in forelimb stepping of thalamic cats, Neurosci. Res., 1 (1984) 393-410. 16 Shimamura, M., Fuwa, T. and Kogure, I., Burst discharges of pontine reticular neurons in relation to forelimb stepping of thalamic and high spinal cats, Brain Research, 346 (1985) 363-367. 17 Shimamura, M., Kogure, I. and Fuwa, T., Locomotion of forelimbs in thalamic cats: joint afferent and supraspinal contributions. In S. Grillner et al. (Eds.), Neurobiology of Vertebrate Locomotion, Macmillan, England, 1986. 18 Shimamura, M., Tanaka, I. and Fuwa, T., Comparison between spino-bulbo-spinal and propriospinal reflexes in thalamic cats during stepping, Neurosci. Res., 7 (1990) 358-368. 19 Watkins, L.R., Griffin, G.R., Leichnetz, G.R. and Mayer, D.J., The somatotopic organization of the NRM and surrounding brainstem structures as revealed by HRP slow-release gels, Brain Research, 181 (1980) 1-15. 20 Zieglg~insberger,W. and Puil, E.A., Actions of glutamic acid on spinal neurons, Exp. Brain Res., 17 (1973) 35-49.
