Brain Research, 533 (1990) 141-146
141
Elsevier
BRES 24366
Respiratory interneurons in the Cs segment of the spinal cord of the cat Mark C. Bellingham
and Janusz Lipski*
Division of Neuroscience, John Curtin School of Medical Research, The Australian National University, Canberra, ACT (Australia)
(Accepted 31 July 1990) Key words: Respiration; Spinal cord; Intemeuron; Phrenic motoneuron; Cat; Superior laryngeal nerve
Extracellular recordingswere made in the C5 segment of the spinal cord of anaesthetised cats from 129 units which showed respiratory phased discharge. The majority of recordings (88%) were thought to arise from the somata of respiratory spinal interneurons. Inspiratory units and expiratory units comprised 42% and 52% of all recorded units. A small number of postiospiratory units were also found (n = 5). Most units did not respond to electrical stimulation of the ipsilateral superior laryngeal (SLN) and phrenic nerves (PN), but a few expiratory (n = 2) and postinspiratory units (n = 1) were excited by SLN stimulation, while 6 inspiratory units had their discharge suppressed by the same stimulus. PN stimulation evoked a long latency (2-7 ms) burst of firing in 2 inspiratoryand I expiratory intemeurons. It is concluded that these respiratory interneurons may provide a segmental input to phrenic motoneurons, in addition to synaptic drives mediated by bulbospinal pathways.
The majority of synaptic input to a-motoneurons in the spinal cord is from segmental and propriospinal interneurons rather than long descending pathways or peripheral afferents9. However, little evidence has been found for the participation of segmental and suprasegmental interneurons in the regulation of phrenic motoneurons (PMNs) in the cat 4'8'13. This is in contrast to a number of reports on the thoracic spinal cord which, in this species, appears to contain a large number of propriospinal and segmental interneurons thought to be responsible for integrating inputs to intercostal motoneurons1,19,20, 26. It is also difficult to reconcile the view of PMNs as a motor pool which is regulated by monosynaptic inputs from bulbospinal neurons 16 with the level of monosynaptic connectivity between these medullary respiratory neurons and PMNs, estimated to be 20-30% 1L12. It thus seems likely that spinal interneurons also play a significant part in the regulation of PMN discharge. The aim of this study was to examine the intermediate and ventral grey matter of the C 5 segment for respiratory interneurons, as this segment contains the majority of PMNs in the cat an. These results have already been published as a preliminary note 6. Experiments were done on 39 cats (1.5-2.6 kg). In 31 experiments, anaesthesia was induced with sodium pentobarbitone (Nembutal, 35-40 mg/kg, i.p.) and maintained by supplementary doses of 3-6 mg/h i.v. In the
remaining 8 experiments, anaesthesia was induced with halothane, followed by a-chloralose (60-70 mg/kg, i.v.; maintenance 5 mg/kg/h, i.v.). Dexamethasone (2 mg i.m.) and atropine sulphate (0.04 mg/kg, i.m.) were given in a single dose. The femoral artery and vein were catheterised to monitor arterial blood pressure and heart rate, and to administer drugs. Arterial blood gases and pH were measured (Coming 168 pH/blood gas analyser) and corrected if necessary by adjustment of ventilation or administration of HCO 3- (1 mmol/ml, i.v.). Following muscle paralysis with gallamine triethiodide (Flaxedil, 8 mg/kg, i.v.), the animals were artificially ventilated through a tracheal cannula, and bilateral pneumothoraces were done. An end-expiratory pressure of 2-3 cm H20 was maintained. End-tidal CO2 (Datex, Normocap) was kept between 4-6%. The rectal temperature was monitored and kept at 38 + 1 °C. One or both C 5 branches of the phrenic nerve (PN) and both superior laryngeal nerves (SLNs) were prepared for standard bipolar stimulation and recording. The animal was fixed in a stereotaxic frame and the C3 to the C7 spinal cord segments exposed. The SLN was stimulated with single pulses or bursts of 2-3 shocks at 0.1 ms intervals, using constant voltage stimulation (Digitimer DS2, stimulus voltage 1.0-2.0 V, pulse duration 0.1 ms, at 50 Hz). The threshold voltage (0.25-1.0 V; mean = 0,48 V) for inhibition of ipsilateral PN discharge was determined at the start of the experiment and SLN
* Present address: Department of Physiology, Medical School, University of Auckland, Private bag, Auckland, New Zealand. Correspondence: M.C. Beilingham. Present address: Zentrum Physiologieund Pathophysiologie, Universit/it G6ttingen, Humboldtallee 23,
D-3400 G6ttingen, ER.G.
142 stimulation was subsequently set at 1.5-3.0 times threshold level. Extracellular recordings were made with glass microelectrodes, pulled from 1.5 mm thin walled glass tubing with an internal filament (Clark Electromedieal Instruments, GC150TF), broken back to 1.5-3.0 #m tip diameter and filled with either 4 M NaCI or 1 M NaC2H30 2 with 3% (w/v) Pontamine sky blue 6BX dye (George T. Gurr Ltd.) 17. In 13 experiments, a systematic search for units firing in phase with parts of the respiratory cycle was made in the C 5 segment, with a series of tracks (lateral intervals of 100/~m) from 200 to 300 #m lateral to the dorsal root entry zone to the midline. The search for respiratory activity began at a depth of 1.5-2.0 mm below the dorsal surface. The microelectrode was advanced in steps of approximately 10-25 pm and held at each position for at least one complete respiratory cycle. When a respiratory unit was detected, the ipsilateral C 5 branch of the PN was stimulated (2.5 V, 0.1 ms, 50 Hz) to determine whether the unit was activated, and whether an antidromic field potential was present. Some units were also tested for a response to stimulation of the ipsilateral SLN. Traces of extracellular spike activity and ipsilateral C 5 PN activity were recorded on a storage oscilloscope screen (Tektronix D13) and photographed. The spike activity of the units was discriminated from background noise by a time/amplitude window discriminator, and the unit's firing rate counted by a digital frequency monitor and displayed on a pen recorder, together with full wave rectified and integrated (time constant = 100 ms) PN output, arterial blood pressure and tracheal pressure. The unit's maximal firing rate was the average of peak firing rates (counted as the number of spikes in 150 ms bins) for 5 consecutive respiratory cycles. The position of the PMN pool was localised by recording the maximal amplitude of the PMN antidromic field potential, pressure injection of Pontamine blue dye in these points, and subsequent histological analysis. The location of the PMN pool, as determined by the depth and mediolateral position of the maximal antidromic field potential, was normalised to a standard position (mean location was at a depth of 3.5 + 0.4 mm, and 0.4 + 0.16 mm medial to the dorsal root entry zone) and the normalised dorsoventral and mediolateral coordinates of recordings from each experiment were then plotted relative to this position. A total of 129 extracellular recordings were made from units showing phasic discharge related to the respiratory cycle. Units were classified into 3 types - expiratory (E), inspiratory (I) and postinspiratory (PI), according to the times of their maximum firing rate. I units were defined as those having their maximal firing rate during the main (ramp) discharge period of the PN, while E units had
their maximal firing rate between these periods. PI units were defined as those which had their maximal firing rate at the transition from I to E, or during the early part of expiration (approximately the first third), when the PN discharge often showed a 'tail' of discharge. Attempts were made to classify action potentials as axonal or somatic. Recordings likely to arise from axons usually showed either very brief (approximately 0.1 ms), negative going spikes, or wider (up to 0.3 ms) positive going spikes; recordings of such spikes were sensitive to movement of the recording microelectrode over a short distance, usually vanishing abruptly 19. In contrast, recordings from the somas of neurons showed biphasic (usually negative/positive) spikes and could be held over a relatively long tracking distance (about 100/tm), with steadily increasing and then decreasing spike amplitude as the microelectrode was advanced ventrally28. Fifty-six I units were recorded, consisting of 6 axons, 48 interneurons and 2 PMNs. All showed an incrementing pattern of discharge. At the transition between inspiration and expiration, some then showed an abrupt decline in firing rate, while others declined more slowly. A few I units showed a discharge, at a lower rate during most of E (e.g. Fig. 1E,H). Examples of recordings from I units are shown in Fig. 1D-F,H. Maximal firing rates ranged from 8 to 138 impulses/s (all I units). The mean value of the maximal firing rate for I axons was 92 + 20 impulses/s (n = 4), while that for I interneurons was 43 + 34 impulses/s (n = 38); these means were significantly different (Student's t-test, P < 0.01). Fourteen I units (3 axons and 11 interneurons) were tested for response to stimulation of the ipsilateral SLN. The discharge of 6 units (2 axons and 4 interneurons) was suppressed, while 1 axon was excited. Twenty-four interneurons were tested for response to stimulation of the ipsilateral C5 PN. Two showed a burst of activity; the latency between the stimulus and the beginning of the response was approximately 2 and 6 ms. These I interneurons may have been either Renshaw interneurons 18'23, or interneurons activated by slow conducting afferents in the PN. Examples of the response of I interneurons to SLN and PN stimulation are shown in Fig. 2B,C. Sixty-seven E units were recorded (10 axons, 57 interneurons), which showed discharge patterns which were incremental, decremental or relatively constant throughout the expiratory phase. Some E units also discharged during inspiration at much lower rates. Examples of recordings are shown in Fig. 1A-C,G. Maximal firing rates ranged from 10 to 207 impulses/s (all E units). E axons had a mean maximal firing rate of 80 + 53 impulses/s (n = 7), while that of E interneurons was 37 + 27 (n = 33). The mean value for E axons was
143
A
I
I,
,
",
]
. -'-"
-"-"-"
,
,
.
.
~
I
I
I
!'
p
B
.
S0
-'-:- :"-" ----.
-""
"
Hz
1 rovE. 5 v
,00.vt__
C
"ILl l i ~
Him
ls
I~
'"
"2 ~J
"'
6 see
D
E
I
F
A_J_J_l I__
oo.vL
;0.5 s
S Jl;_;_12: i
."
'.=!
6 Iwc t's = ~tl
r° "
ii m
H
20mVL 200 .u.V O.5s
I
.~
" "
:,-
'~':
" ~'"
Fig. 1. Examples of recordings from respiratory interneurones and axons of the C5 segment. All photographs are of single sweeps. Negatwtty is downwards for all traces except the upper trace in B. A (bottom trace): the firing rate of an E interneurone; the top trace is integrated C5 PN discharge. B (top trace): a photograph of the discharge pattern of the same interneurone, bottom trace is C5 PN discharge. C (top trace): the discharge pattern of another E interneurone; bottom trace is the T I ~ pulses from the discriminated C~ PN discharge. D: the firing rate of an I interneurone; top and bottom traces are as for A. E: a photograph of the discharge of the same interneurone shown in D; top and bottom traces as for B. F: a photograph of the discharge pattern of another I interneurone; top and bottom traces are as for B. G (top trace): the discharge of an E axon; bottom trace is C5 PN discharge. H: the discharge of an I axon; top and bottom traces as for G. I: the discharge of a PI axon; top and bottom traces as for G. Note that bottom trace of I, showing the PN activity, has been cut off by the frame of the camera. Large positive spikes in H and I indicate partial impalement of the axons.
significantly higher than that for E interneurons ( P < 0.05). T h e r e were no significant differences between the m e a n values of maximal firing frequency of E and I axons, or b e t w e e n those of E and I interneurons ( P > 0.05). Thirteen E interneurons were tested for response to stimulation of the ipsilateral SLN; 2 showed an excitatory response, an e x a m p l e of which is seen in Fig. 2A. O n e E interneuron r e s p o n d e d to stimulation of the ipsilateral C 5 PN with a burst of spike activity beginning 6 - 7 ms after the stimulus. Five PI units were recorded. A l l showed a decrementing p a t t e r n o f discharge, with an a b r u p t onset at, or slightly after, the p e a k discharge of the PN. O n e recording was thought to be from an axon, and 4 from interneurons. A n e x a m p l e of an extracellular recording of the discharge p a t t e r n of a PI axon is shown in Fig. 1I. The
PI axon had a maximal F R of 61 impulses/s, while 3 PI interneurons had maximal F R s of 20, 61 and 100 impulses/s. O n e PI i n t e r n e u r o n was excited by stimulation of the ipsilateral SLN. Recordings from respiratory units were distributed throughout the length of the C 5 segment. T h e normalised locations are shown in Fig. 3. T h e locations of these recordings were not confined to the region occupied by the P M N pool 33 or to the areas d o r s o m e d i a l and dorsolateral to the P M N somas through which P M N dendrites are known to r u n 10'2a. T h e paucity of recordings from the region of the phrenic nucleus was due to the presence of a large antidromic field potential which m a d e positive identification of i n t e r n e u r o n s in this region impossible. The activity classified as ' a x o n a l ' m a y have originated from descending axons o r axon collaterals of bulbospinal
144
A
tt! i
i
SLN stimulation """
.".
....
........
:::-
• "-• •
•'-"'
"'°
:
" . . . . . . . .
~.
: :_ ..-
'
-':
""
:-"
.-:"
.
.
~
.
.
"--
.--.." •
-"
- °-
2. ~ ; " . . ~
...................
-.-'.
-2.'"
..
--''._.'Y'-."
. . .'."
_
"
I|
imp/s i,lll
~°°~vL5
ms
III
C I
~00"~v °ri0I s
~ .'........
2 mV
m
-
•
Fig. 2. Responses from C5 respiratory intemeurones to stimulation of the ipsilateral SLN or C 5 branch of the PN. Stimuli are marked with arrows in B and C. A: a pen trace recording of the response of an E interncurone to stimulation of the ipsilateral SLN (1.5 V, 0.1 ms, 50 Hz) during the period indicated by the horizontal bar; the top trace is integrated C5 PN discharge, bottom trace is firing rate of unit (150 ms bins). The period in which the pen recorder ran at the faster speed is between the two vertical arrows. B: 40 superimposed responses of an I interneurone to stimulation of the ipsilateral SLN only during inspiration (1.7 V, 3 × 0.1 ms at 2 ms intervals, 2 Hz); the top trace is unit activity (negativity upwards), bottom trace is activity of the ipsilateral C5 PN. C: 4 superimposed responses of an I intemeurone to stimulation of the ipsilateral C5 PN (1.5 V, 0.1 ms, 4 Hz); the top trace is the potential recorded from the cord dorsum with a silver/silver chloride ball electrode, and the bottom trace is an extracellular recording of the unit's response; negativity downwards. respiratory neurons 13, from propriospinal axons of inspiratory neurons in the upper cervical cord 2"3'22, or from axons of local respiratory interneurons 13, while 'neuronal' activity was most probably from neuronal somas located close to the recording electrode 19'28. However, since no attempt was made to antidromically activate these units from other regions of the spinal cord, we cannot confirm the somai origin of these recordings, or define the direction of their axonal projections• The latter information would be useful in ascertaining the connections between medullary, high cervical (C1-C2) and lower cervical (C5-C6) respiratory neurons. Recordings of respiratory phased interneurons in the C 5 and C 6 segments of the cat have previously been occasionally noted 4's'13, but the present study is the first to characterize a large number of such interneurons in the cat. Recently, extracellular recordings in the cervical cord of the rabbit have shown that a large population of similar respiratory phased interneurons also exists in this species, in segments containing the phrenic nucleus 29. The firing patterns and general location of respiratory interneurons recorded in the rabbit are similar to those of the respiratory interneurons found in the cat, despite
differences in the method of anaesthesia used. It should be noted that intracellular recordings of C 5 respiratory interneurons in the cat often did not exhibit firing even when stable impalements with resting m e m b r a n e potentials o f - 5 0 mV or less were achieved a n d large respiratory phased shifts of m e m b r a n e potential were present 5. This implies that the number of respiratory units recorded extracellularly in the C 5 segment of the pentobarbitone anaesthetised cat may be an underestimate of the actual number of respiratory interneurons present. The presence of recordings which showed discharge or depolarisation during the PI period was surprising, as the PI neurons of the medulla have not been reported to have spinal projections 3°'31. In a recent study 32, one PI medullary neuron was found which had a spinal projection to the C 2 level. Extracellular recordings have also been made from a small number of respiratory units in the thoracic spinal cord which showed a PI discharge 19. This pattern of discharge may result from an interaction between I and E phased inhibitory inputs to t h e same interneuron, resulting in a short period of disinhibition during the PI phase, or may be due to excitatory synaptic
145
gl lira
¢
laaon
II
Evn
O
E~
I
Pl,~
D
PI ~ o n
(~
Prver~ mJ~s
A
E E
v
m
o o
-2
.-3
o ,12 1-
tll o
.5
TM
.6" -1
0
1
2
Mediolateral distance from the DREZ (turn) Fig. 3. Location of extracellular recording points from respiratory units of the C5 spinal segment. The normalised locations are plotted on an outline of a transverse section of the C5 spinal cord taken from an animal in which the location of the phrenic nucleus corresponded to the mean location of the nucleus from all experiments. The phrenic nucleus in the diagram is represented by an ellipsoid with axes equal to the standard deviation of measurements of phrenic nucleus location in both directions.
input from medullary PI neurons via rare spinal projections. The failure of most respiratory interneurons to respond to SLN stimulation was unexpected, as it was thought that disfacilitation of bulbospinal medullary neurons would have at least resulted in a decrease in discharge of the interneurons, as was seen in a few inspiratory interneurons (e.g. Fig. 2B). The effects of SLN stimulation on the firing of non-responsive inter-
1 Aminoff, M.J. and Sears, T.A., Spinal integration of segmental, cortical and breathing inputs to thoracic respiratory motoneurones, J. Physiol., 215 (1971) 537-575. 2 Aoki, M., Kasaba, T., Kurosawa, Y., Ohtsura, K. and Satomi, H., The projections of cervical respiratory neurons to the phrenic nucleus in the cat, Neurosci. Left., Suppl. 17 (1984) $49. 3 Aoki, M., Mori, S., Kawahara, K., Watanabe, H. and Ebata, N., Generation of spontaneous respiratory rhythm in high spinal cats, Brain Research, 202 (1980) 51-63. 4 Baumgarten, R. von, Schmiedt. H. and Dodich, N., Microelectrode studies of phrenic motoneurones, Ann. N.Y. Acad. Sci.,
neurons may have been weak due to divergence of descending respiratory drive, and hence not observed by our recording methods. This could be resolved by the use of more sensitive techniques, such as post stimulus time histograms. It was thought that some interneurons (particularly expiratory) might contribute to the synaptic inhibition of PMNs evoked by SLN stimulation 7 but interneurons which were excited by such stimulation were also uncommon, indicating that any interneuronal link in this inhibitory pathway is mainly located elsewhere, most probably in the medulla. The sources of respiratory phased inputs to PMNs have been intensively investigated, and several groups of medullary respiratory bulbospinal neurons have been shown to make monosynaptic connections with PMNs 14' 15,18,24,25,27. Quantitative studies lm2 have indicated that monosynaptic excitatory inputs to these motoneurons are insufficient to provide the necessary level of depolarisation. The respiratory interneurons described here are potentially able to provide additional excitatory and inhibitory synaptic drive to PMNs, as well as to integrate inputs from non-respiratory systems, such as the sensorimotor cortex TM. Further work is necessary to investigate the synaptic connections between descending respiratory axons and these interneurons, and between the interneurons and PMNs. In summary, extracellular recordings have been made from a number of respiration phased axons and neurons in the C 5 segment of the cat. It is most likely that the recorded spike activity was from descending axons or axon collaterals of respiratory neurons located in the medulla or upper cervical cord and from neuronal somas located in the C5 segment. While there is, as yet, no direct evidence to suggest that these interneurons receive inputs from the respiratory bulbospinal or upper cervical cord neurons, or that they make synaptic connections to PMNs, their discharge pattern and location suggests that they act as important nodal points for the convergence of multiple inputs prior to the final stage of integration in phrenic motoneurons. We would like to thank Terrina Thompson for superb technical assistance.
109 (1963) 536-546. 5 Bellingham, M.C., Integration of Inhibitory Inputs to Phrenic Motoneurones, PhD Thesis, The Australian National University, Canberra, Australia, 1989. 6 Bellingham, M.C. and Lipski, J., Respiratory intemeurones in the region of the phrenic nucleus, Neurosci. Lett., Suppl. 27 (1987) $54. 7 Bellingham, M.C., Lipski, J. and Voss, M.D., Synaptic inhibition of phrenic motoneurones evoked by stimulation of the superior laryngeal nerve, Brain Research, 486 (1989) 391-395. 8 Biscoe, T.J. and Sampson, S.R., Analysis of the inhibition of
146 phrenic motoneurones which occur on stimulation of some cranial nerve afferents, J. Physiol., 209 (1970) 375-393. 9 Burke, R.E. and Rudomin, P., Spinal neurons and synapses. In E.R. Kandel (Ed.), Handbook of Physiology, Section 1: The Nervous System, Vol. I: Cellular Biology of Neurons, Part 2, American Physiological Society, Bethesda, MD, 1977, pp. 877-944. 10 Cameron, W.E., Averill, D.B. and Berger, A.J., Morphology of cat phrenic motoneurons as revealed by intracellular injection of horseradish peroxidase, J. Comp. Neurol.., 219 (1983) 70-80. 11 Davies, J.G.McE, Kirkwood, P.A. and Sears, T.A., The detection of monosynaptic connexions from inspiratory bulbospinal neurones to inspiratory neurones, J. Physiol., 368 (1985) 33-62. 12 Davies, J.G.McE, Kirkwood, P.A. and Sears, T.A., The distribution of monosynaptic connexions from inspiratory bulbospinal neurones to inspiratory motoneurones in the cat, J. Physiol., 368 (1985) 63-87. 13 Dick, T.E., Jodkowski, J.S., Viana, E and Berger, A.J., Projections and terminations of single respiratory axons in the cervical spinal cord of the cat, Brain Research, 449 (1988) 201-212. 14 Fedorko, L., Hoskin, R.W. and Duffin, J., Projections from inspiratory neurons of the nucleus retroambigualis to phrenic motoneurons in the cat, Exp. Neurol., 105 (1989) 306-310. 15 Fedorko, L., Merrill, E.G. and Lipski, J., Two descending medullary inspiratory pathways to phrenic motoneurones, Neurosci. Lett., 43 (1983) 285-291. 16 Feldman, J.L., Smith, J.C., McCrimmon, D.R., Ellenberger, H.H. and Speck, D.F., Generation of respiratory pattern in mammals. In A. Cohen (Ed.), Neural Control of Rhythmic Movements in Vertebrates, Wiley, 1988, pp. 73-100. 17 Hellon, R.E, The marking of electrode tip positions in nervous tissue, J. Physiol., 214 (1971) 12P. 18 Hilaire, G,, Khatib, M. and Monteau, R., Central drive on Renshaw cells coupled with phrenic motoneurones, Brain Research, 376 (1986) 133-139. 19 Kirkwood, P.A., Munson, J.B., Sears, T.A. and Westgaard, R.H., Respiratory interneurones in the thoracic spinal cord of the cat, J. Physiol., 395 (1988) 161-192. 20 Kirkwood, P.A. and Sears, T.A., Field potentials in the contralateral ventral horn evoked by individual thoracic respi-
ratory interneurones in the anaesthetized cat, J. Physiol., 418 (1989) 106P. 21 Lipski, J., Bektas, A. and Porter, R., Short latency inputs to phrenic motoneurones from the sensorimotor cortex in the cat, Exp. Brain Res., 61 (1986) 280-290. 22 Lipski, J. and Duffin, J., An electrophysiological investigation of propriospinal inspiratory neurons in the upper cervical cord of the cat, Exp. Brain Res., 61 (1986) 625-637. 23 Lipski, J., Fyffe, R.E.W. and Jodkowski, J.S., Recurrent inhibition of cat phrenic motoneurons, J. Neurosci., 5 (1985) 1545-1555. 24 Lipski, J., Kubin, L. and Jodkowski, J.S., Synaptic action of Rfl neurones on phrenic motoneurones studied with spike-triggered averaging, Brain Research, 288 (1983) 105-118. 25 Merrill, E.G. and Fedorko, L., Monosynaptic inhibition of phrenic motoneurones: a long descending projection from BOtzinger neurons, J. Neurosci., 4 (1984) 2350--2353. 26 Merrill, E.G. and Lipski, J., Inputs to intercostal motoneurons from ventrolateral medullary respiratory neurons in the cat, J. Neurophysiol., 57 (1987) 1837-1853. 27 Monteau, R., Khatib, M. and Hilaire, G., Central determination of recruitment order: intracellular study of phrenic motoneurones, Neurosci. Lett., 56 (1985) 341-346. 28 Nelson, J.R., Single unit activity in medullary respiratory centers of cat, J. Neurophysiol., 22 (1959) 590-598. 29 Palisses, R., Pers6gol, L. and Viala, D., Evidence for respiratory interneurones in the C3-C5 cervical spinal cord in the decorticate rabbit, Exp. Brain Res., 78 (1989) 624-632. 30 Remmers, J.E., Richter, D.W., Ballantyne, D., Bainton, C.R. and Klein, J.E, Reflex prolongation of stage I of expiration, Pfl~gers Arch., 407 (1986) 190-198. 31 Richter, D.W., Ballantyne, D. and Returners, J.E., The differential organization of medullary post-inspiratory activities, Pfl~gers Arch., 410 (1987) 420-427. 32 Richter, D.W., Lawson, E.E., Lalley, P.M. and BaUantyne, D., Peripheral chemoreceptor inputs to medullary inspiratory and postinspiratory neurons of cats, Pflagers Arch., 414 (1989) 523-533. 33 Webber, C.L.Jr., Wurster, R.D. and Chung, J.M., Cat phrenic nucleus architecture as revealed by horseradish peroxidase mapping, Exp. Brain Res., 35 (1979) 395-406.
141
Elsevier
BRES 24366
Respiratory interneurons in the Cs segment of the spinal cord of the cat Mark C. Bellingham
and Janusz Lipski*
Division of Neuroscience, John Curtin School of Medical Research, The Australian National University, Canberra, ACT (Australia)
(Accepted 31 July 1990) Key words: Respiration; Spinal cord; Intemeuron; Phrenic motoneuron; Cat; Superior laryngeal nerve
Extracellular recordingswere made in the C5 segment of the spinal cord of anaesthetised cats from 129 units which showed respiratory phased discharge. The majority of recordings (88%) were thought to arise from the somata of respiratory spinal interneurons. Inspiratory units and expiratory units comprised 42% and 52% of all recorded units. A small number of postiospiratory units were also found (n = 5). Most units did not respond to electrical stimulation of the ipsilateral superior laryngeal (SLN) and phrenic nerves (PN), but a few expiratory (n = 2) and postinspiratory units (n = 1) were excited by SLN stimulation, while 6 inspiratory units had their discharge suppressed by the same stimulus. PN stimulation evoked a long latency (2-7 ms) burst of firing in 2 inspiratoryand I expiratory intemeurons. It is concluded that these respiratory interneurons may provide a segmental input to phrenic motoneurons, in addition to synaptic drives mediated by bulbospinal pathways.
The majority of synaptic input to a-motoneurons in the spinal cord is from segmental and propriospinal interneurons rather than long descending pathways or peripheral afferents9. However, little evidence has been found for the participation of segmental and suprasegmental interneurons in the regulation of phrenic motoneurons (PMNs) in the cat 4'8'13. This is in contrast to a number of reports on the thoracic spinal cord which, in this species, appears to contain a large number of propriospinal and segmental interneurons thought to be responsible for integrating inputs to intercostal motoneurons1,19,20, 26. It is also difficult to reconcile the view of PMNs as a motor pool which is regulated by monosynaptic inputs from bulbospinal neurons 16 with the level of monosynaptic connectivity between these medullary respiratory neurons and PMNs, estimated to be 20-30% 1L12. It thus seems likely that spinal interneurons also play a significant part in the regulation of PMN discharge. The aim of this study was to examine the intermediate and ventral grey matter of the C 5 segment for respiratory interneurons, as this segment contains the majority of PMNs in the cat an. These results have already been published as a preliminary note 6. Experiments were done on 39 cats (1.5-2.6 kg). In 31 experiments, anaesthesia was induced with sodium pentobarbitone (Nembutal, 35-40 mg/kg, i.p.) and maintained by supplementary doses of 3-6 mg/h i.v. In the
remaining 8 experiments, anaesthesia was induced with halothane, followed by a-chloralose (60-70 mg/kg, i.v.; maintenance 5 mg/kg/h, i.v.). Dexamethasone (2 mg i.m.) and atropine sulphate (0.04 mg/kg, i.m.) were given in a single dose. The femoral artery and vein were catheterised to monitor arterial blood pressure and heart rate, and to administer drugs. Arterial blood gases and pH were measured (Coming 168 pH/blood gas analyser) and corrected if necessary by adjustment of ventilation or administration of HCO 3- (1 mmol/ml, i.v.). Following muscle paralysis with gallamine triethiodide (Flaxedil, 8 mg/kg, i.v.), the animals were artificially ventilated through a tracheal cannula, and bilateral pneumothoraces were done. An end-expiratory pressure of 2-3 cm H20 was maintained. End-tidal CO2 (Datex, Normocap) was kept between 4-6%. The rectal temperature was monitored and kept at 38 + 1 °C. One or both C 5 branches of the phrenic nerve (PN) and both superior laryngeal nerves (SLNs) were prepared for standard bipolar stimulation and recording. The animal was fixed in a stereotaxic frame and the C3 to the C7 spinal cord segments exposed. The SLN was stimulated with single pulses or bursts of 2-3 shocks at 0.1 ms intervals, using constant voltage stimulation (Digitimer DS2, stimulus voltage 1.0-2.0 V, pulse duration 0.1 ms, at 50 Hz). The threshold voltage (0.25-1.0 V; mean = 0,48 V) for inhibition of ipsilateral PN discharge was determined at the start of the experiment and SLN
* Present address: Department of Physiology, Medical School, University of Auckland, Private bag, Auckland, New Zealand. Correspondence: M.C. Beilingham. Present address: Zentrum Physiologieund Pathophysiologie, Universit/it G6ttingen, Humboldtallee 23,
D-3400 G6ttingen, ER.G.
142 stimulation was subsequently set at 1.5-3.0 times threshold level. Extracellular recordings were made with glass microelectrodes, pulled from 1.5 mm thin walled glass tubing with an internal filament (Clark Electromedieal Instruments, GC150TF), broken back to 1.5-3.0 #m tip diameter and filled with either 4 M NaCI or 1 M NaC2H30 2 with 3% (w/v) Pontamine sky blue 6BX dye (George T. Gurr Ltd.) 17. In 13 experiments, a systematic search for units firing in phase with parts of the respiratory cycle was made in the C 5 segment, with a series of tracks (lateral intervals of 100/~m) from 200 to 300 #m lateral to the dorsal root entry zone to the midline. The search for respiratory activity began at a depth of 1.5-2.0 mm below the dorsal surface. The microelectrode was advanced in steps of approximately 10-25 pm and held at each position for at least one complete respiratory cycle. When a respiratory unit was detected, the ipsilateral C 5 branch of the PN was stimulated (2.5 V, 0.1 ms, 50 Hz) to determine whether the unit was activated, and whether an antidromic field potential was present. Some units were also tested for a response to stimulation of the ipsilateral SLN. Traces of extracellular spike activity and ipsilateral C 5 PN activity were recorded on a storage oscilloscope screen (Tektronix D13) and photographed. The spike activity of the units was discriminated from background noise by a time/amplitude window discriminator, and the unit's firing rate counted by a digital frequency monitor and displayed on a pen recorder, together with full wave rectified and integrated (time constant = 100 ms) PN output, arterial blood pressure and tracheal pressure. The unit's maximal firing rate was the average of peak firing rates (counted as the number of spikes in 150 ms bins) for 5 consecutive respiratory cycles. The position of the PMN pool was localised by recording the maximal amplitude of the PMN antidromic field potential, pressure injection of Pontamine blue dye in these points, and subsequent histological analysis. The location of the PMN pool, as determined by the depth and mediolateral position of the maximal antidromic field potential, was normalised to a standard position (mean location was at a depth of 3.5 + 0.4 mm, and 0.4 + 0.16 mm medial to the dorsal root entry zone) and the normalised dorsoventral and mediolateral coordinates of recordings from each experiment were then plotted relative to this position. A total of 129 extracellular recordings were made from units showing phasic discharge related to the respiratory cycle. Units were classified into 3 types - expiratory (E), inspiratory (I) and postinspiratory (PI), according to the times of their maximum firing rate. I units were defined as those having their maximal firing rate during the main (ramp) discharge period of the PN, while E units had
their maximal firing rate between these periods. PI units were defined as those which had their maximal firing rate at the transition from I to E, or during the early part of expiration (approximately the first third), when the PN discharge often showed a 'tail' of discharge. Attempts were made to classify action potentials as axonal or somatic. Recordings likely to arise from axons usually showed either very brief (approximately 0.1 ms), negative going spikes, or wider (up to 0.3 ms) positive going spikes; recordings of such spikes were sensitive to movement of the recording microelectrode over a short distance, usually vanishing abruptly 19. In contrast, recordings from the somas of neurons showed biphasic (usually negative/positive) spikes and could be held over a relatively long tracking distance (about 100/tm), with steadily increasing and then decreasing spike amplitude as the microelectrode was advanced ventrally28. Fifty-six I units were recorded, consisting of 6 axons, 48 interneurons and 2 PMNs. All showed an incrementing pattern of discharge. At the transition between inspiration and expiration, some then showed an abrupt decline in firing rate, while others declined more slowly. A few I units showed a discharge, at a lower rate during most of E (e.g. Fig. 1E,H). Examples of recordings from I units are shown in Fig. 1D-F,H. Maximal firing rates ranged from 8 to 138 impulses/s (all I units). The mean value of the maximal firing rate for I axons was 92 + 20 impulses/s (n = 4), while that for I interneurons was 43 + 34 impulses/s (n = 38); these means were significantly different (Student's t-test, P < 0.01). Fourteen I units (3 axons and 11 interneurons) were tested for response to stimulation of the ipsilateral SLN. The discharge of 6 units (2 axons and 4 interneurons) was suppressed, while 1 axon was excited. Twenty-four interneurons were tested for response to stimulation of the ipsilateral C5 PN. Two showed a burst of activity; the latency between the stimulus and the beginning of the response was approximately 2 and 6 ms. These I interneurons may have been either Renshaw interneurons 18'23, or interneurons activated by slow conducting afferents in the PN. Examples of the response of I interneurons to SLN and PN stimulation are shown in Fig. 2B,C. Sixty-seven E units were recorded (10 axons, 57 interneurons), which showed discharge patterns which were incremental, decremental or relatively constant throughout the expiratory phase. Some E units also discharged during inspiration at much lower rates. Examples of recordings are shown in Fig. 1A-C,G. Maximal firing rates ranged from 10 to 207 impulses/s (all E units). E axons had a mean maximal firing rate of 80 + 53 impulses/s (n = 7), while that of E interneurons was 37 + 27 (n = 33). The mean value for E axons was
143
A
I
I,
,
",
]
. -'-"
-"-"-"
,
,
.
.
~
I
I
I
!'
p
B
.
S0
-'-:- :"-" ----.
-""
"
Hz
1 rovE. 5 v
,00.vt__
C
"ILl l i ~
Him
ls
I~
'"
"2 ~J
"'
6 see
D
E
I
F
A_J_J_l I__
oo.vL
;0.5 s
S Jl;_;_12: i
."
'.=!
6 Iwc t's = ~tl
r° "
ii m
H
20mVL 200 .u.V O.5s
I
.~
" "
:,-
'~':
" ~'"
Fig. 1. Examples of recordings from respiratory interneurones and axons of the C5 segment. All photographs are of single sweeps. Negatwtty is downwards for all traces except the upper trace in B. A (bottom trace): the firing rate of an E interneurone; the top trace is integrated C5 PN discharge. B (top trace): a photograph of the discharge pattern of the same interneurone, bottom trace is C5 PN discharge. C (top trace): the discharge pattern of another E interneurone; bottom trace is the T I ~ pulses from the discriminated C~ PN discharge. D: the firing rate of an I interneurone; top and bottom traces are as for A. E: a photograph of the discharge of the same interneurone shown in D; top and bottom traces as for B. F: a photograph of the discharge pattern of another I interneurone; top and bottom traces are as for B. G (top trace): the discharge of an E axon; bottom trace is C5 PN discharge. H: the discharge of an I axon; top and bottom traces as for G. I: the discharge of a PI axon; top and bottom traces as for G. Note that bottom trace of I, showing the PN activity, has been cut off by the frame of the camera. Large positive spikes in H and I indicate partial impalement of the axons.
significantly higher than that for E interneurons ( P < 0.05). T h e r e were no significant differences between the m e a n values of maximal firing frequency of E and I axons, or b e t w e e n those of E and I interneurons ( P > 0.05). Thirteen E interneurons were tested for response to stimulation of the ipsilateral SLN; 2 showed an excitatory response, an e x a m p l e of which is seen in Fig. 2A. O n e E interneuron r e s p o n d e d to stimulation of the ipsilateral C 5 PN with a burst of spike activity beginning 6 - 7 ms after the stimulus. Five PI units were recorded. A l l showed a decrementing p a t t e r n o f discharge, with an a b r u p t onset at, or slightly after, the p e a k discharge of the PN. O n e recording was thought to be from an axon, and 4 from interneurons. A n e x a m p l e of an extracellular recording of the discharge p a t t e r n of a PI axon is shown in Fig. 1I. The
PI axon had a maximal F R of 61 impulses/s, while 3 PI interneurons had maximal F R s of 20, 61 and 100 impulses/s. O n e PI i n t e r n e u r o n was excited by stimulation of the ipsilateral SLN. Recordings from respiratory units were distributed throughout the length of the C 5 segment. T h e normalised locations are shown in Fig. 3. T h e locations of these recordings were not confined to the region occupied by the P M N pool 33 or to the areas d o r s o m e d i a l and dorsolateral to the P M N somas through which P M N dendrites are known to r u n 10'2a. T h e paucity of recordings from the region of the phrenic nucleus was due to the presence of a large antidromic field potential which m a d e positive identification of i n t e r n e u r o n s in this region impossible. The activity classified as ' a x o n a l ' m a y have originated from descending axons o r axon collaterals of bulbospinal
144
A
tt! i
i
SLN stimulation """
.".
....
........
:::-
• "-• •
•'-"'
"'°
:
" . . . . . . . .
~.
: :_ ..-
'
-':
""
:-"
.-:"
.
.
~
.
.
"--
.--.." •
-"
- °-
2. ~ ; " . . ~
...................
-.-'.
-2.'"
..
--''._.'Y'-."
. . .'."
_
"
I|
imp/s i,lll
~°°~vL5
ms
III
C I
~00"~v °ri0I s
~ .'........
2 mV
m
-
•
Fig. 2. Responses from C5 respiratory intemeurones to stimulation of the ipsilateral SLN or C 5 branch of the PN. Stimuli are marked with arrows in B and C. A: a pen trace recording of the response of an E interncurone to stimulation of the ipsilateral SLN (1.5 V, 0.1 ms, 50 Hz) during the period indicated by the horizontal bar; the top trace is integrated C5 PN discharge, bottom trace is firing rate of unit (150 ms bins). The period in which the pen recorder ran at the faster speed is between the two vertical arrows. B: 40 superimposed responses of an I interneurone to stimulation of the ipsilateral SLN only during inspiration (1.7 V, 3 × 0.1 ms at 2 ms intervals, 2 Hz); the top trace is unit activity (negativity upwards), bottom trace is activity of the ipsilateral C5 PN. C: 4 superimposed responses of an I intemeurone to stimulation of the ipsilateral C5 PN (1.5 V, 0.1 ms, 4 Hz); the top trace is the potential recorded from the cord dorsum with a silver/silver chloride ball electrode, and the bottom trace is an extracellular recording of the unit's response; negativity downwards. respiratory neurons 13, from propriospinal axons of inspiratory neurons in the upper cervical cord 2"3'22, or from axons of local respiratory interneurons 13, while 'neuronal' activity was most probably from neuronal somas located close to the recording electrode 19'28. However, since no attempt was made to antidromically activate these units from other regions of the spinal cord, we cannot confirm the somai origin of these recordings, or define the direction of their axonal projections• The latter information would be useful in ascertaining the connections between medullary, high cervical (C1-C2) and lower cervical (C5-C6) respiratory neurons. Recordings of respiratory phased interneurons in the C 5 and C 6 segments of the cat have previously been occasionally noted 4's'13, but the present study is the first to characterize a large number of such interneurons in the cat. Recently, extracellular recordings in the cervical cord of the rabbit have shown that a large population of similar respiratory phased interneurons also exists in this species, in segments containing the phrenic nucleus 29. The firing patterns and general location of respiratory interneurons recorded in the rabbit are similar to those of the respiratory interneurons found in the cat, despite
differences in the method of anaesthesia used. It should be noted that intracellular recordings of C 5 respiratory interneurons in the cat often did not exhibit firing even when stable impalements with resting m e m b r a n e potentials o f - 5 0 mV or less were achieved a n d large respiratory phased shifts of m e m b r a n e potential were present 5. This implies that the number of respiratory units recorded extracellularly in the C 5 segment of the pentobarbitone anaesthetised cat may be an underestimate of the actual number of respiratory interneurons present. The presence of recordings which showed discharge or depolarisation during the PI period was surprising, as the PI neurons of the medulla have not been reported to have spinal projections 3°'31. In a recent study 32, one PI medullary neuron was found which had a spinal projection to the C 2 level. Extracellular recordings have also been made from a small number of respiratory units in the thoracic spinal cord which showed a PI discharge 19. This pattern of discharge may result from an interaction between I and E phased inhibitory inputs to t h e same interneuron, resulting in a short period of disinhibition during the PI phase, or may be due to excitatory synaptic
145
gl lira
¢
laaon
II
Evn
O
E~
I
Pl,~
D
PI ~ o n
(~
Prver~ mJ~s
A
E E
v
m
o o
-2
.-3
o ,12 1-
tll o
.5
TM
.6" -1
0
1
2
Mediolateral distance from the DREZ (turn) Fig. 3. Location of extracellular recording points from respiratory units of the C5 spinal segment. The normalised locations are plotted on an outline of a transverse section of the C5 spinal cord taken from an animal in which the location of the phrenic nucleus corresponded to the mean location of the nucleus from all experiments. The phrenic nucleus in the diagram is represented by an ellipsoid with axes equal to the standard deviation of measurements of phrenic nucleus location in both directions.
input from medullary PI neurons via rare spinal projections. The failure of most respiratory interneurons to respond to SLN stimulation was unexpected, as it was thought that disfacilitation of bulbospinal medullary neurons would have at least resulted in a decrease in discharge of the interneurons, as was seen in a few inspiratory interneurons (e.g. Fig. 2B). The effects of SLN stimulation on the firing of non-responsive inter-
1 Aminoff, M.J. and Sears, T.A., Spinal integration of segmental, cortical and breathing inputs to thoracic respiratory motoneurones, J. Physiol., 215 (1971) 537-575. 2 Aoki, M., Kasaba, T., Kurosawa, Y., Ohtsura, K. and Satomi, H., The projections of cervical respiratory neurons to the phrenic nucleus in the cat, Neurosci. Left., Suppl. 17 (1984) $49. 3 Aoki, M., Mori, S., Kawahara, K., Watanabe, H. and Ebata, N., Generation of spontaneous respiratory rhythm in high spinal cats, Brain Research, 202 (1980) 51-63. 4 Baumgarten, R. von, Schmiedt. H. and Dodich, N., Microelectrode studies of phrenic motoneurones, Ann. N.Y. Acad. Sci.,
neurons may have been weak due to divergence of descending respiratory drive, and hence not observed by our recording methods. This could be resolved by the use of more sensitive techniques, such as post stimulus time histograms. It was thought that some interneurons (particularly expiratory) might contribute to the synaptic inhibition of PMNs evoked by SLN stimulation 7 but interneurons which were excited by such stimulation were also uncommon, indicating that any interneuronal link in this inhibitory pathway is mainly located elsewhere, most probably in the medulla. The sources of respiratory phased inputs to PMNs have been intensively investigated, and several groups of medullary respiratory bulbospinal neurons have been shown to make monosynaptic connections with PMNs 14' 15,18,24,25,27. Quantitative studies lm2 have indicated that monosynaptic excitatory inputs to these motoneurons are insufficient to provide the necessary level of depolarisation. The respiratory interneurons described here are potentially able to provide additional excitatory and inhibitory synaptic drive to PMNs, as well as to integrate inputs from non-respiratory systems, such as the sensorimotor cortex TM. Further work is necessary to investigate the synaptic connections between descending respiratory axons and these interneurons, and between the interneurons and PMNs. In summary, extracellular recordings have been made from a number of respiration phased axons and neurons in the C 5 segment of the cat. It is most likely that the recorded spike activity was from descending axons or axon collaterals of respiratory neurons located in the medulla or upper cervical cord and from neuronal somas located in the C5 segment. While there is, as yet, no direct evidence to suggest that these interneurons receive inputs from the respiratory bulbospinal or upper cervical cord neurons, or that they make synaptic connections to PMNs, their discharge pattern and location suggests that they act as important nodal points for the convergence of multiple inputs prior to the final stage of integration in phrenic motoneurons. We would like to thank Terrina Thompson for superb technical assistance.
109 (1963) 536-546. 5 Bellingham, M.C., Integration of Inhibitory Inputs to Phrenic Motoneurones, PhD Thesis, The Australian National University, Canberra, Australia, 1989. 6 Bellingham, M.C. and Lipski, J., Respiratory intemeurones in the region of the phrenic nucleus, Neurosci. Lett., Suppl. 27 (1987) $54. 7 Bellingham, M.C., Lipski, J. and Voss, M.D., Synaptic inhibition of phrenic motoneurones evoked by stimulation of the superior laryngeal nerve, Brain Research, 486 (1989) 391-395. 8 Biscoe, T.J. and Sampson, S.R., Analysis of the inhibition of
146 phrenic motoneurones which occur on stimulation of some cranial nerve afferents, J. Physiol., 209 (1970) 375-393. 9 Burke, R.E. and Rudomin, P., Spinal neurons and synapses. In E.R. Kandel (Ed.), Handbook of Physiology, Section 1: The Nervous System, Vol. I: Cellular Biology of Neurons, Part 2, American Physiological Society, Bethesda, MD, 1977, pp. 877-944. 10 Cameron, W.E., Averill, D.B. and Berger, A.J., Morphology of cat phrenic motoneurons as revealed by intracellular injection of horseradish peroxidase, J. Comp. Neurol.., 219 (1983) 70-80. 11 Davies, J.G.McE, Kirkwood, P.A. and Sears, T.A., The detection of monosynaptic connexions from inspiratory bulbospinal neurones to inspiratory neurones, J. Physiol., 368 (1985) 33-62. 12 Davies, J.G.McE, Kirkwood, P.A. and Sears, T.A., The distribution of monosynaptic connexions from inspiratory bulbospinal neurones to inspiratory motoneurones in the cat, J. Physiol., 368 (1985) 63-87. 13 Dick, T.E., Jodkowski, J.S., Viana, E and Berger, A.J., Projections and terminations of single respiratory axons in the cervical spinal cord of the cat, Brain Research, 449 (1988) 201-212. 14 Fedorko, L., Hoskin, R.W. and Duffin, J., Projections from inspiratory neurons of the nucleus retroambigualis to phrenic motoneurons in the cat, Exp. Neurol., 105 (1989) 306-310. 15 Fedorko, L., Merrill, E.G. and Lipski, J., Two descending medullary inspiratory pathways to phrenic motoneurones, Neurosci. Lett., 43 (1983) 285-291. 16 Feldman, J.L., Smith, J.C., McCrimmon, D.R., Ellenberger, H.H. and Speck, D.F., Generation of respiratory pattern in mammals. In A. Cohen (Ed.), Neural Control of Rhythmic Movements in Vertebrates, Wiley, 1988, pp. 73-100. 17 Hellon, R.E, The marking of electrode tip positions in nervous tissue, J. Physiol., 214 (1971) 12P. 18 Hilaire, G,, Khatib, M. and Monteau, R., Central drive on Renshaw cells coupled with phrenic motoneurones, Brain Research, 376 (1986) 133-139. 19 Kirkwood, P.A., Munson, J.B., Sears, T.A. and Westgaard, R.H., Respiratory interneurones in the thoracic spinal cord of the cat, J. Physiol., 395 (1988) 161-192. 20 Kirkwood, P.A. and Sears, T.A., Field potentials in the contralateral ventral horn evoked by individual thoracic respi-
ratory interneurones in the anaesthetized cat, J. Physiol., 418 (1989) 106P. 21 Lipski, J., Bektas, A. and Porter, R., Short latency inputs to phrenic motoneurones from the sensorimotor cortex in the cat, Exp. Brain Res., 61 (1986) 280-290. 22 Lipski, J. and Duffin, J., An electrophysiological investigation of propriospinal inspiratory neurons in the upper cervical cord of the cat, Exp. Brain Res., 61 (1986) 625-637. 23 Lipski, J., Fyffe, R.E.W. and Jodkowski, J.S., Recurrent inhibition of cat phrenic motoneurons, J. Neurosci., 5 (1985) 1545-1555. 24 Lipski, J., Kubin, L. and Jodkowski, J.S., Synaptic action of Rfl neurones on phrenic motoneurones studied with spike-triggered averaging, Brain Research, 288 (1983) 105-118. 25 Merrill, E.G. and Fedorko, L., Monosynaptic inhibition of phrenic motoneurones: a long descending projection from BOtzinger neurons, J. Neurosci., 4 (1984) 2350--2353. 26 Merrill, E.G. and Lipski, J., Inputs to intercostal motoneurons from ventrolateral medullary respiratory neurons in the cat, J. Neurophysiol., 57 (1987) 1837-1853. 27 Monteau, R., Khatib, M. and Hilaire, G., Central determination of recruitment order: intracellular study of phrenic motoneurones, Neurosci. Lett., 56 (1985) 341-346. 28 Nelson, J.R., Single unit activity in medullary respiratory centers of cat, J. Neurophysiol., 22 (1959) 590-598. 29 Palisses, R., Pers6gol, L. and Viala, D., Evidence for respiratory interneurones in the C3-C5 cervical spinal cord in the decorticate rabbit, Exp. Brain Res., 78 (1989) 624-632. 30 Remmers, J.E., Richter, D.W., Ballantyne, D., Bainton, C.R. and Klein, J.E, Reflex prolongation of stage I of expiration, Pfl~gers Arch., 407 (1986) 190-198. 31 Richter, D.W., Ballantyne, D. and Returners, J.E., The differential organization of medullary post-inspiratory activities, Pfl~gers Arch., 410 (1987) 420-427. 32 Richter, D.W., Lawson, E.E., Lalley, P.M. and BaUantyne, D., Peripheral chemoreceptor inputs to medullary inspiratory and postinspiratory neurons of cats, Pflagers Arch., 414 (1989) 523-533. 33 Webber, C.L.Jr., Wurster, R.D. and Chung, J.M., Cat phrenic nucleus architecture as revealed by horseradish peroxidase mapping, Exp. Brain Res., 35 (1979) 395-406.
