NeuropharmacologyVol. 31. No. 10, pp. 1049-1058,1992

0028-3908/92$5.00+ 0.00 PergamonPress Ltd

Printed in Great Britain

EFFECTS OF HALOTHANE ON MEMBRANE POTENTIAL AND DISCHARGE ACTIVITY IN PAIRS OF BULBAR RESPIRATORY NEURONS OF DECEREBRATE CATS R. TAKEDAand A. HAJI Department of Pharmacology, Faculty of Medicine, Toyama Medical and Pharmaceutical University, 2630 Sugitani, Toyama 930-01, Japan

(Accepted 10 March 1992) Summary--In aiming to test the possibility of synaptic interactions, the effect of inhalation of halothane (2% for 90 see) was studied on 45 out of 88 pairs of respiratory neurons, simultaneously recorded with intracellular and extracellular microelectrodes,in both sides of the ventral respiratory group of decerebrate cats. Halothane produced various effects on these respiratory neurons; namely, depolarization (n = 30) or hyperpolarization (n = 15) of intracellularly recorded neurons, an increase (n = 7) or decrease (n = 38) in the firing of extracellularly recorded neurons. However, with repeated application, the agent produced a consistent effect in a given cell. Spike-triggered averaging of synaptic noise, using spikes of nonantidromically-activated respiratory units, did not reveal any unitary postsynaptic potential but a symmetric synaptic wave of medium-frequency-oscillation(35-50 Hz) in 7 pairs. In addition, power spectral analysis of the membrane potential and spike-intervalhistogram of the paired neuron, displayed no correlated activity suggestiveof synaptic interactions. For all the neuronal pairs examined, halothane produced random effects on their patterns of firing and synaptic waves. The present results suggest that halothane exerts a selectiveeffect on each respiratory neuron and that the lack of a correlated response to application of halothane reflects the lack of synaptic interaction between pairs of bilaterally sampled neurons of the ventral respiratory group.

Key words--respiratory neuron, halothane, action potential, postsynaptic potential, medium-frequencyoscillation.

Application of an anesthetic agent alters the discharge activity of buibar and pontine respiratory neurons, in a manner characteristic of each functional group (Hukuhara, 1974; Caille, Vibert, Bertrand, Gromysz and Hugelin, 1979; Grelot and Bianchi, 1987). Recently, it was shown that a small dose of halothane or thiopental produced a consistent effect on respiratory neurons of the ventral respiratory group, characterized by an important decline of synaptic noise and an increase in input resistance (Takeda, Haji and Hukuhara, 1990). Hence, the primary effect of anesthetic agents appears to be depression or disruption of synaptic interactions among the respiratory network of neurons. This may lead to the suggestion that the agents should affect the patterns of firing and synaptic waves, which could either deny or confirm the interconnections between pairs of neurons. Therefore, the effects of inhalation of halothane was investigated on the rhythmic firing and synaptic noise in pairs of respiratory neurons, recorded simultaneously in the medullary respiratory complex. Electrophysiological and morphological examinations of the medullary axonal projections indicate that some of the neurons of the ventral respiratory group send axons or axon collaterals to the contralateral ventral respiratory group (see reviews; Long and Duffin, 1986; Ezure, 1990). Feldman, Sommer

and Cohen (1980) have found that a short-time correlation, which was suggestive of pauci-synaptic interactions, was extremely rare in bilaterally located neuronal pairs in the ventral respiratory group, the firing periods of which overlapped. Using spiketriggered averaging of intracellularly recorded synaptic noise, Ezure and Manabe (1989) demonstrated monosynaptic excitatory projections to some inspiratory neurons and none of the expiratory neurons of the ventral respiratory group, from bulbospinal inspiratory neurons of the contralateral ventral respiratory group. However, this part of ventral respiratory group contains a population of respiratory neurons, having no spinal and no vagal axons (Long and Duffin, 1986; Ezure, 1990). The second purpose of this study was to examine whether such synaptic interactions might be demonstrated during the application of anesthetic between the non-vagal, nonbulbospinal respiratory neurons and contralateral neurons in the ventral respiratory group. METHODS

Surgical procedures A total of 24 cats of either sex, weighing 2.5-3.9 kg, was anesthetized with halothane (1.8-2.2%) for tracheostomy and catheterization of the urethra and the femoral artery and vein. The carotid sinus

1049

1050

R. TAKEDAand A. HArt

and cervical vagus nerves were cut bilaterally to eliminate the presumed peripheral effects of halothane (Coleridge, Coleridge, Luck and Borman, 1968). The head of the animal was mounted on a stereotaxic frame and decerebration was performed at the midcollicular level. Then, muscular paralysis was instituted with gallamine triethiodide. The animals were subjected to a bilateral pneumothorax and artificially ventilated with oxygen-enriched air. A positive post-expiratory pressure of l - 2 c m H20 was applied to prevent collapse of the lung. The fractional end-tidal concentrations of CO2 were kept at 0.044).05. Rectal temperature was maintained at 37-39°C by a heating pad. Systemic blood pressure was kept at more than 80mmHg by infusing a lactated Ringer solution, as required. Arterial PaCO2 and PaO2 were measured in 4 randomly selected animals and ranged from 27 to 32mmHg (pH 7.37-7.40) and from 232 to 269 mmHg, respectively, throughout the experiments. The most cranial root of the phrenic nerve was cut, desheathed and placed on a bipolar silver electrode. Bilaterally incised vagus nerves and superior laryngeal nerves were mounted on bipolar stimulating electrodes. The medulla oblongata was exposed through an occipital craniotomy. A C2~23 laminectomy was performed and an array of 5 coaxial stimulating electrodes was inserted into the ventrolateral part of the spinal cord. Halothane anesthesia was then discontinued.

Electrical recording Phrenic nerve activity was rectified and integrated by a leaky integrator with a 0.1 sec time constant. Extracellular action potentials of a respiratory neuron and the membrane potential of another neuron, were recorded simultaneously in both sides of the ventral respiratory group, the former with metal microelectrodes (1-5 Mf~ at 1000 Hz) and the latter with glass micropipettes, filled with 2 M potassium citrate (20-30 MI)). All the neurons examined lay in the region extending 0-2.5 mm rostral to the obex, 3.24.3 mm lateral to the midline and 2.7-4.6mm below the dorsal surface. Augmenting inspiratory, decrementing expiratory or post-inspiratory and augmenting expiratory neurons were identified, based upon the temporal relationship of the membrane potential trajectory or of the action potential firing to the phrenic discharge (Richter, 1982). In the following text, these intracellularly and extracellularly recorded neurons are denoted by the abbreviations using upper and lower case letters, respectively; inspiratory (I, i), post-inspiratory (PI, pi) and expiratory (E, e) neurons. The decrementing type of inspiratory neuron was not encountered at the present study. The late inspiratory neurons, the firing period of which started late in inspiration and extended to an early part of stage 1 expiration, were recorded extracellularly in 2 cases. These were classified as the i-unit group (Tables 1 and 2). According to the response to electrical stimulation of the ipsilateral vagus nerve

(1.0-1.5V, 0.2msec) and the spinal cord (5-10V, 0.2 msec), neurons were identified as laryngeal motor (LM), bulbospinal (BS) or non-antidromicallyactivated (NAA) neurons (Richter, 1982). None of these neurons was antidromically activated by stimulation of the ipsilateral superior laryngeal nerve (1.0-1.5 V, 0.2 msec). Only spikes from NAA respiratory units and membrane potentials of greater than - 50 mV during the phase of maximum hyperpolarization were used. Thus, while keeping extracellular recording of one such N A A unit, several respiratory neurons (up to 4 neurons) were impaled in the contralateral ventral respiratory group. Electrical activity of the paired respiratory neurons and phrenic nerve were stored on magnetic tape and played back later for computer analyses.

Application of drugs The first examination of the effect of a drug was started 4 h r after withdrawal of the anesthesia for surgery. Halothane was evaporated through a precalibrated vaporizer and administered through the endotracheal tube, using an artificial respirator. As was shown in the previous study (Takeda et al., 1990), a 90 sec inhalation of 2% halothane was adequate to yield a brief, consistent effect on the phrenic and bulbar neuronal activity. Larger and longer doses were avoided, because they caused a severe fall in blood pressure, which often made the stable impalement of the microelectrode impossible. Since the peak effect was achieved at approximately 1 min after the onset of inhalation and remained fairly steady in the subsequent 2min, the effect of halothane was evaluated during the period of 1.5-3 min after the onset of inhalation. At least 20 min elapsed between successive doses, to eliminate a possible residual effect of the preceding application.

Computer analysis and data acquisition A power spectral analysis was performed on synaptic noise during the non-spiking phase of the respiratory cycle, using a signal processing computer (Nihon-Kohden, ATAC 450). A "raw" signal from each intracellularly recorded neuron was filtered (bandwidth 1-300 Hz) and sampled at 250 Hz. Power spectral density was computed on 1.024 sec during inspiration for E and PI neurons, using a pulse derived from the onset of the phrenic burst as a trigger (1-trigger) and during stage 1 expiration or post-inspiration (Richter, Ballantyne and Remmers, 1986) for I neurons, triggered at the rapid fall of the phrenic burst (E-trigger). These were averaged for 20 consecutive respiratory cycles before and during the peak effect after inhalation of halothane. Concomitantly, power spectra of the phrenic discharge were averaged in the corresponding respiratory cycles. Spike-triggered averaging was attempted for all neuronal pairs, the firing periods of which were not overlapped. The recording of membrane potential was filtered (10-1000 Hz) and averaged, using

Halothane on bulbar discriminated spikes from a contralateral N A A respiratory unit as trigger events. In most cases, averaging of 1000 sweeps, using 1024 points of 40 gsec bin width, was employed. In addition, a spike-interval histogram was computed for the all extraccllularly recorded units by counting 1000 spikes, which were usually obtained in 10-25 respiratory cycles before and after inhalation of halothane. Values of the membrane potential were evaluated at the most hyperpolarized point, which occurred during the inactive phase of the respiratory cycle. Mean spike frequencies of extracellularly recorded units were calculated from the spike-interval histograms. These values were averaged for each group of respiratory neurons. Mean values ( + SEM) of these variables, observed before and during the peak effect after halothane, were compared using a paired t-test (two-sided). Statistical significance was assumed at P < 0.05. RESULTS

Power-spectral analysis and spike-triggered averaging of synaptic noise

respiratory

neurons

1051

fired predominantly at the frequency of M F O (Fig. 2B). Average power spectra of the phrenic neurogram displayed a predominant peak at a similar frequency (40-50 Hz) with a broader bandwidth (Fig. 2A). These results indicated that inhibitory synaptic waves of certain PI and E neurons oscillated in synchrony with spikes of contralateral i-units at a similar rate of M F O in the phrenic neurogram. For these 7 neurons, the occurrence of action potentials in i-units, with respect to M F O in membrane potential, was distributed in a relatively short interval, beginning 5.7 msec before and extending to 2.5 msec after peak depolarization of the MFO. Mean values for the duration and amplitude of MFOs in membrane potential, were 20.2 + 1.5 msec and 310 + 62 #V (n = 7), respectively. The membrane potential H F O appeared in only one cell (PI-LM neuron), where the peak power spectrum occurred at about 90 Hz (Fig. 3). In the rest of the whole population, both power spectral densities and spike-interval histograms showed no correlated activity, suggestive of synaptic interactions between pairs of bilaterally recorded neurons of the ventral respiratory group.

Stable, simultaneous recordings of intracellular membrane potential and extracellular action potentials were achieved in 88 instances (Table 1). For all neuronal pairs examined, spike-triggered averagings did not reveal any short-term unitary post-synaptic potential, such as has been demonstrated in certain bulbar respiratory neurons (Merrill, Lipski, Kubin and Fedorko, 1983). In 2 out of 7 E neurons and in 5 out of 15 PI neurons, which were paired with contralateral i-units, spike-triggered averages of synaptic noise revealed a symmetrical wave. Figure 1 illustrates such an example for a PI-i neuronal pair. The shape of the symmetrical wave resembled the high-frequency-oscillation (HFO) in the membrane potential of bulbar respiratory neurons (Mitchell and Herbert, 1974), with the oscillating rate (41 Hz) corresponding to that of the medium-frequency-oscillation (MFO) in the respiratory motor activity (Cohen, See, Christakos and Sica, 1987). The I-triggered averages of power spectra, taken from these PI and E neurons, revealed a prominent peak at the frequency range between 35 and 50 Hz (Fig. 2A). The paired i-unit

Effects of inhalation of halothane on individual neurons of the ventral respiratory group Inhalation of halothane produced a decrease in phrenic inspiratory discharge and shortened the inspiratory and expiratory intervals. The peak amplitude of the integrated phrenic neurogram was reduced to 75 + 2% of the control in 45 trials with halothane. Also, the agent invariably decreased the phrenic post-inspiratory activity (Figs 1, 3-5). On 45 pairs of neurons of the ventral respiratory group, effects of halothane were successfully examined (Table 2). During and succeeding 2-3 min after the period of inhalation, three types of responses were observed in membrane potential and spike discharge of the intracellularly recorded neurons; namely, depolarization associated with an increase of action potential firing (n = 5), depolarization with a decrease or arrest of firing (n = 25) and hyperpolarization with a decrease or complete cessation of firing (n = 15). These responses were observed similarly in I, PI and E neuron groups (Table 2A). Moreover, halothane decreased the respiratory fluctuations

Table 1. A combination matrix of respiratory neuronal pairs, recorded simultaneously from both sides of the ventral respiratory group lntracellularly recorded neuron Neuron type

I

P1

E

n

i 13/20 7/15 3/7 23/41 Extracellularlyrecorded unit pi 8/21 7/12 1/5 16/38 e 2/2 4/4 0/3 6/9 n 23/43 18/31 4/15 45/88 Neuron type; inspiratory(I, i), post-inspiratory(PI, pi) and expiratory(E, e). Data are expressed as the number of neuronal pairs tested with halothan¢per (/) the number of simultaneouslyrecordedpairs,n: A total number of cellcombination.Allextracellularly recorded neurons are non-antidromically-activated(NAA) cells. The intracellularly recorded group consists of laryngealmotor (n = 16/28), bulbo-spinal(n ~ 9/12) and NAA neurons (n = 20/48).

1052

R. TAKEDAand A. HAJ[

(C)

(A) BEFORE

SPIKE-TRIGGERED AVERAGES

~

-6,=

i I III I

al[k~bbJ"II--

,=

.. •

L=

L"

.. . . . . . . . .

,. . . . . . . . . .

- - x

1000

BEFORE

rll

2

(B) HALOTHANE 2rain

HALOTHANE 2rain

3

x 1000

lOmV

-68-

0.2mV

x500 I

I

= =

.

I ....

ii

lmV 0.2mV

i

2sec

10msec

Fig. 1. (A, B) Simultaneous recordings of intraccllular membrane potential of a PI-NAA neuron (trace 1), extracellular action potentials of an i-NAA unit of the contralateral ventral respiratory group (trace 2) and phrenic nerve activity (trace 3). Records taken before (A) and 2 min after the onset of a 90 sec inhalation of 2% halothane (B) are illustrated. Reference membrane potentials are shown in mV on the left side of each trace. (C) Trace 1; averages of 1000 samples of trigger spikes of the i-unit taken during the control period. Traces 2-4 illustrate, respectively, spike-triggered averages of synaptic noise in the PI neuron taken before, 2 and 15 min after the onset of halothane inhalation. Traces 2 and 3 derived from 1000 samples, and trace 4 from 500 samples. A vertical dotted line denotes the start of average coinciding with the occurrence of trigger spikes.

(A) POWER SPECTRUM (B) SPIKE-INTERVAL HISTOGRAM

1 BEFORE I

1 BEFORE MP

32 0 I

16

PN 0

3

M~pIo

.

2 HALOTHANE

1:t

PN

2 HALOTHANE

20

0

0

50

100

28

36

44

52

INTERVAL(msec)

FREQUENCY (Hz)

Fig. 2. (A) Power-spectral densities of synaptic noise (MP) in the PI neuron, presented in Fig. 1 and of phrenic discharge (PN). These were averaged for 20 consecutive respiratory cycles using I-trigger pulses before (1) and 2 rain after the onset of inhalation of halothane (2). (B) Interval-histograms of spike train of the i-unit shown in Fig. 1. Histograms I and 2 were taken before and 2 rain after the onset of halothane inhalation, respectively. A total of 1000 spikes were counted for each analysis. Parts of the histogram, at the interval range of longer than 52 msec, are not shown. No spike was counted at the range of 1-12 msec.

Halothane on bulbar respiratory neurons in membrane potential and synaptic noise in all intracellularly recorded neurons (Figs 1, 3-5). Likewise, halothane produced either a reduction or augmentation of spike discharge in extracellularly recorded i-, pi- and e-neurons (Table 2B). In 7 units, halothane caused a complete arrest of firing, which lasted 5-10 min after the onset of the application of drug. However, repeated application of halothane produced a consistent effect on a given cell. This effect was observed in 7 i-, 2 e- and 6 pi-units; an increase of firing in 2 units and a decrease or arrest of firing in 13 units. There was no respiratory unit that showed an opposite response to repeated application of halothane. These results suggest that the agent produced a specific and selective effect on an individual respiratory neuron (Takeda et al., 1990).

1053

averaging detected no MFO wave in these 3 neurons (Fig. IC, trace 2), whereas the MFO-related peak was still present in power spectra of the membrane potential (Fig. 2A). For the paired i-unit, the distribution pattern of spike-interval histogram became broader and shifted to a longer interval range (Fig. 2B). In other 2 PI-i and 2 E-i neuronal pairs, showing MFOs in control, halothane caused a complete arrest of the spike discharge in the i-units, which made the spiketriggered averaging impossible. For these PI and E neurons, inhalation of halothane decreased but did not eliminate the peak spectra of MFOs of synaptic noise, observed during inspiration. In one PI-LM neuron, where a prominent peak appeared at the HFO range (90 Hz) in the average power spectral density, halothane severely depressed this peak spectrum (Fig. 3C). For the other neuron pairs in which either power spectrum, spike-interval histogram or spike-triggered averaging detected no correlated activity, suggestive of synaptic interactions, application of halothane produced various combinations of changes in membrane potential and discharge activity. Figure 3 illustrates such an example for a PI-LM neuron and an e-NAA unit. The PI neuron displayed distinct hyperpolarization during stage 2 expiration, following the post-inspiratory depolarization and spiking. The contralateral e-unit fired only during stage 2 expiration. The timing of action potential firing in the e-unit and the hyperpolarizing wave during stage 2 expiration ot the PI neuron led to the suggestion that the former is a presynaptic source for inhibition in the latter

Response of paired respiratory neurons to inhalation of halothane The influence of inhalation of halothane on the spike-triggered MFO waves was successfully examined on 3 PI-i neuronal pairs, where the trigger spikes of i-units continued to occur after halothane. For 2 PI neurons, the anesthetic agent hyperpolarized the membrane and decreased the respiratory fluctuations of membrane potential, synaptic noise and periodic spiking. Concomitantly, it decreased the rate of firing in the accompanying i-units (Figs 1B and 2B). In one case, the PI neuron was depolarized while a paired i-unit showed a slight decrease in the discharge rate. During inhalation of halothane, spike-triggered

Table 2. Effects of halothane inhalation on membrane potential and discharge activity in the ventral respiratory group neurons (A) Intracellularly recorded neuron MP (mV) Neuron type I (n = 23) PI

(n = 18) E (n = 4)

Depolarization Before

After

Hyperpolarization Before

After

-64.7_+1.7 -55.4_+2.1"* (n = 15)

-65.4_+2.6 -68.6_+2.0* (n = 8)

-68.9_+2.8

-64.6_+3.4

-59.3_+2.4**

(n = 13) -67.5 + 1.8 -59.5 _+5.3 (n = 2)

-69.8_+3.3*

(n = 5) -64.5 _+3.9 -67.5 + 5.3 (n = 2)

(B) Extracellularly recorded neuron AP (Hz) Neuron type i (n = 23)

Increase Before

After

48.6+8.4 54.4+8.1" (n = 5)

pi (n = 16)

80

e

38

(n = 6)

Decrease After

88

48.6_+5.3 34.3_+6.1"* (n = 15)

40

37.6 + 8.8

(n = 1)

(n = 1)

Before

47.4_+5.2 32.5_+5.2** (n = 18)

24.2 + 9.7**

(n = 5)

Neuron type: inspiratory (I, i), post-inspiratory (PI, pi) and expiratory (E, e) neurons. (A) Average membrane potentials (MP) measured at the most hyperpolarized point in each respiratory cycle before and during the peak effect at 1.5-3 rain after the onset of a 90 sec inhalation of 2% halothane. (B) Mean frequencies of the firing of action potentials in a burst (AP) taken before and during the peak effect after halothane inhalation. All values are means + SEM. The number of neurons are in parentheses. *P < 0.05, **P < 0.01; significantly different from the "before" value (paired t-test). NP 31/10--G

1054

R. TAKEDA and A. HAJI

(A)

(C) POWER SPECTRUM

BEFORE

PN 1

i

0

(B)

1

HALOTHANE 10mV

2 HALOTHANE

MP

-500 1

0.2mV

PN lmV

2sec

0 0

50 100 FREQUENCY (Hz)

Fig. 3. (A, B) Simultaneous recordings of membrane potential of a PI-LM neuron (trace 1) and extracellular spikes of an e-NAA unit (trace 2), together with integrated (trace 3) and raw phrenic neurogram (trace 4). Records taken before (A) and 2 min after the onset of inhalation of halothane (B) are shown. (C) I-Triggered averages of power-spectral density of the PI neuron (MP) and of phrenic neurogram (PN), taken before (1) and 2 min after the onset of inhalation of halothane (2). Note that prominent peaks appeared at 90-92 Hz in both MP and PN power spectra.

(Richter, 1982; Richter et al., 1986). However, inhalation of halothane suppressed the hyperpolarizing wave and released spikes during stage 2 expiration in the PI neuron, while it increased the spike activity in the e-unit (Fig. 3B). Thus, the changes in spike discharge and synaptic wave, induced by halothane could not confirm the presumed reciprocal interaction between the two neurons. In addition, phrenic post-inspiratory activity was suppressed by halothane, while the spike discharge of the PI neuron was increased, contrasting with the results shown in Fig. 1. This means that halothane exerted a uniform depressant effect on phrenic post-inspiratory activity, whereas it caused different and selective effects on individual PI or pi neurons. Figure 4 shows another example for an I - N A A neuron and a late i - N A A unit. The I neuron started to fire at low frequency after the post-inspiratory pause and progressively increased the firing rate during inspiration. The firing frequency reached the peak shortly (200--300 msec) before the decline of the phrenic discharge. Spike discharge of a simultaneously recorded i-unit began in late inspiration and extended to the early phase of stage 1 expiration. These patterns of firing in the late i-unit imply a notion that this type of inspiratory neuron plays a key role to switch off inspiration, leading to a

rapid hyperpolarization in I neurons (Richter et al., 1986). During inhalation of halothane, however, the firing of the late i-unit increased, whereas the hyperpolarizing shift of membrane potential became less pronounced in the I neuron at the transition from inspiration to post-inspiration (Fig. 4B). Figure 5 shows an example where two unitary activities were recorded with an extracellular electrode, in conjunction with the recording of membrane potential in a contralateral neuron of the ventral respiratory group. Each of the two extracellular units, a late i-unit and an e-unit, could be discriminated by the difference in their spike height (trace 2 in Figs 5A and B). These three neurons displayed different responses to inhalation of halothane. The late i-unit increased the duration of burst discharge, while the e-unit stopped firing completely (Fig. 5B). Thus, facilitatory and inhibitory responses to halothane occurred concomitantly in two neighboring respiratory neurons, recorded by one single microelectrode. For the paired P I - N A A neuron, in which the MFO wave was clearly visible during inspiration before application of drug (Figs 5A and C), inhalation of halothane caused hyperpolarization, a complete arrest of firing and reduction in the MFO wave (Figs 5B and C). In this case, the inspiratory time was prolonged by inhalation of halothane (Fig. 5B).

Halothane on bulbar respiratory neurons

(A)

1055

BEFORE

_,10_ I

I

i§l!l

2

!III![III

3 4 5

(B)

HALOTHANE lOmV llIB l

!|]!

Iilii=

mlm) O,mV

~I,,.A~[IL ~ • -r

*,.,rlrrr

,,

........

....

I..,,Jm.

' •

"l~

"

~j ,,JIJ ~I

•

-i-[~r¢¢ l

]0.5mV J

2sec Fig. 4. (A, B) Simultaneous recordings of membrane potential of an I - N A A neuron (trace I) and extracellular action potentials of a late i - N A A unit (trace 3). Trace 2 shows standardized pulses derived

from the discriminated spikes of the late i-unit. Traces 4 and 5 represent integrated and raw phrenic neurograms. In 6 instances of I-i pairs, excitatory and inhibitory effects occurred coincidentally in the two paired neurons, during inhalation of halothane. In the other 7 I-i neuronal pairs, halothane produced similar depressant effects on both of the paired neurons. Also, halothane provoked concomitant excitatory and inhibitory responses in 3 PI-pi neuronal pairs and similar inhibitory effects in 4 PI-pi pairs. Thus, the responses to halothane were at random in the pairs of neurons, randomly selected from both sides of the ventral respiratory group. DISCUSSION The concentration of halothane may hardly reach a steady-state level in the alveolar air, when inspired in such a short period (90 see). However, this dose of halothane produced a fairly steady effect, lasting from 1 to 3 min after the onset of inhalation. It was shown in the previous study that, as compared at their peak effects on bulbar respiratory neurons, a 90 sec inhalation of 2% halothane was equipotent

to intravenously injected 2.5 mg/kg thiopental, the dose being approximately 1/10 of that necessary for surgical anesthesia in the cat (Takeda et al., 1990). Inhalation of halothane decreased synaptic noise and the respiratory fluctuations in membrane potential, in all neurons examined. In addition, it was shown that this dose of halothane produced an increase in input resistance and a decrease in the amplitudes of electrically-stimulated excitatory postsynaptic potentials (EPSPs) and inhibitory postsynaptic potentials (IPSPs) in neurons of the ventral respiratory group (Takeda et al., 1990). These results suggest that small doses of anesthetic agents depress or disrupt the synaptic interactions in the bulbar respiratory network. The response to halothane of a PI--e neuronal pair, shown in Fig. 3, did not support the presumed mutual interaction between the two types of expiratory neurons. Also, the behavior of a late i-unit, during inhalation of halothane (Fig. 4), did not confirm an assumption that this type of respiratory neuron might inhibit a population of I neurons to switch o f f

Halothane on bulbar respiratory neurons spectra taken from these PI and E neurons. At the same time, the contralateral i-units continued to fire with the reduced rates. This means that haiothane impaired the occurrence of trigger spikes of the i-unit in synchrony with inhibitory drives, provoking M F O s in membrane potential of PI or E neurons. Alternatively, though haiothane exerts either inhibitory or excitatory effects among the presynaptic neurons, its overall effect seems to be generalized depression of the synchronized drives in the respiratory neuronal network. This may result in a substantial decrease in the compound synaptic waves, such as inspiratory M F O waves in the PI and E neurons (Takeda et al., 1990). This can further account for the generalized depression of phrenic discharge, including H F O and M F O waves during anesthetic application (Cohen, 1972; Cohen et al., 1987; Richardson and Mitchell, 1982). For the neuronal pairs that showed no correlated activity in spike-triggered average, power spectral density and spike-interval histogram, halothane exerted either excitatory or inhibitory effects in a random fashion. This was observed even in the two neighboring respiratory units recorded with a single, extracellular microelectrode. These findings would indicate that the respiratory neuronal responses were not attributable to any particular experimental condition or location of recording. In addition, repeated application of the anesthetic agent produced a consistent effect on a given neuron of the ventral respiratory group. This result well agreed with the previous study which demonstrated a consistent and selective response of each neuron of the ventral respiratory group to both thiopental and halothane (Takeda et al., 1990). Hence, it is presumed that the response of a neuron of the ventral respiratory group to the anesthetic agent, is dependent upon the functional role it bears in the respiratory neuronal network (Hukuhara, 1974; Caille et al., 1979; Grelot and Bianchi, 1987; Takeda et al., 1990). The lack of correlated response to application of anesthetic may reflect the lack of functional interconnections between sampled pairs of bilateral neurons of the ventral respiratory group.

REFERENCES

Anders K., Ballantyne D., Bischoff A. M,, Lalley P. M. and Richter D. W. (1991) Inhibition of caudal medullary expiratory neurones by retrofacial inspiratory neurones in the cat. J. Physiol., Lond. 437: 1-25. Caille D., Vibert J. F., Bertrand F., Gromysz H. and Hugelin A. (1979) Pentobarbitone effects on respiratory related units; selective depression of bulbopontine reticular neurones. Resp. Physiol. 36: 201-216. Cohen M. I. (1972) Synchronization of discharge, spontaneous and evoked, between inspiratory neurons. Acta neurobiol, exp. 33: 189-218. Cohen M. I., See W. R., Christakos C. N. and Sica A. L. (1987) High-frequency and medium-frequency components of different inspiratory nerve discharges and their modification by various inputs. Brain Res. 417: 148-152.

1057

Coleridge H. M., Coleridge J. C. G., Luck J. C. and Borman J. (1968) The effect of four volatile anesthetic agents on impulse activity of two types of pulmonary receptors. Br. J. Anaesth. 40: 484-492. Ezure K. (1990) Synaptic connections between medullary respiratory neurons and considerations on the genesis of respiratory rhythm. Progr. Neurobiol. 35: 429-450. Ezure K. and Manabe M. (1988) Decrementing expiratory neurons of the Boetzinger complex. II. Direct inhibitory synaptic linkage with ventral respiratory group neurons. Expl Brain Res. 72: 159-166. Ezure K. and Manabe M. (1989) Monosynaptic excitation of medullary inspiratory neurons by bulbospinal inspiratory neurons of the ventral respiratory group in the cat. Expl Brain Res. 74:501-51 I. Fedorko L. and Merrill E. G. (1984) Axonal projections from the rostral expiratory neurones of the Boetzinger complex to medulla and spinal cord in the cat. J. Physiol., Lond. 350: 487-496. Feldman J. L. (1986) Neurophysiology of breathing in mammals. In: Handbook of Physiology, Sect. l, Vol. IV, pp. 463 524. American Physiol. Soc., Bethesda. Feldman J. L. and Speck D. F. (1983) Interactions among inspiratory neurons in dorsal and ventral respiratory groups in cat medulla. J. Neurophysiol. 49: 472-490. Feldman J. L., Sommer D. and Cohen M. I. (1980) Short time scale correlations between discharges of medullary respiratory neurons. J. Neurophysiol. 43: 1284-1295. Grelot L. and Bianchi A. L. (1987) Differential effects of halothane anesthesia on the pattern of discharge of inspiratory and expiratory neurons in the region of the retrofacial nucleus. Brain Res. 404: 335-338. Hilaire G., Monteau R. and Bianchi A. L. (1984) A cross-correlation study of interactions among respiratory neurons of dorsal, ventral and retrofacial groups in cat medulla. Brain Res. 302: 19-31. Hukuhara T. (1974) Functional organization of brain stem respiratory neurons and rhythmogenesis. In: Central Rhythmic and Regulation (Umbach W. and Koepchen H. P., Eds), pp. 35-49. Hippokrates, Stuttgart. Linsey B. G., Segers L. S. and Shannon R. (1987) Functional associations among simultaneously monitored lateral medullary respiratory neurons in the cat. II. Evidence for inhibitory actions of expiratory neurons. J. Neurophysiol. 57:1101-1117. Long S. E. and Duffin J. (1984) Cross-correlation of ventrolateral inspiratory neurons in the cat. Expl Neurol. 83: 233-253. Long S. E. and Duffin J. (1986) The neuronal determinants of respiratory rhythm. Progr, Neurobiol. 27: 101-182. Madden K. P. and Returners J. E. (1982) Short time scale correlations between spike activity of neighboring respiratory neurons of nucleus tractus solitarius. J. Neurophysiol. 48: 749-760. Merrill E. G., Lipski J., Kubin L. and Fedorko L. (1983) Origin of the expiratory inhibition of nucleus tractus solitarius inspiratory neurones. Brain Res. 263: 43-50. Mitchell R. A. and Herbert D. A. (1974) Synchronized high frequency synaptic potentials in medullary respiratory neurons. Brain Res. 75: 350-355. Remmers J. E., Takeda R., Schultz S. A. and Haji A. (1985) Relationship of membrane potential of ventral respiratory group neurons to action potentials of retro-facial respiratory units. In: Neurogenesis of Central Respiratory Rhythm (Bianchi A. L. and Denavit-Saubie M., Eds), pp. 117-120. MTA, Lancaster. Richardson C. A. and Mitchell R. A. (1982) Power spectral analysis on inspiratory nerve activity in the decerebrate cat. Brain Res. 233: 317-336. Richter D. W. (1982) Generation and maintenance of the respiratory rhythm. J. exp. Biol. 100: 93-107.

1058

R. TAKEDAand A. HAIl

Richter D. W., Ballantyne D. and Remmers J. E. (1986) How is the respiratory rhythm generated? A model. News Physiol. Sci. (NIPS) 1: 109-112. Sehmid K., Boemer G. and Weichel T. (1990) Concurrent fast and slow synchronized efferent phrenic activities in time and frequency domain. Brain Res. 528:1-11. Segers L. S., Shannon R., Saporta S. and Lindsey B. G. (1987) Functional associations among simultaneously monitored lateral medullary respiratory neurons in the

cat. I. Evidence for excitatory and inhibitory actions of inspiratory neurons. J. Neurophysiol. 57: 1078-I 100. Takeda R., Haji A. and Hukuhara T. (1990) Selective actions of anesthetic agents on membrane potential trajectory in bulbar respiratory neurons of cats. Pfliigers Arch. 416: 375-384. Vachon B. R. and Duflin J. (1978) Cross-correlation of medullary respiratory neurons in the cat. Expl Neurol. 6 h 15-30.