Respiration in vitro: I. Spontaneous activity.

Respiration In Vitro: I. Spontaneous Activity 0. Hamada, E. Garcia-Rill,’ and R. D. Skinner

Somatosens Mot Res Downloaded from informahealthcare.com by UB Heidelberg on 11/15/14 For personal use only.

Department of Anatomy, University of Arkansas for Medical Sciences, Little Rock, Arkansas 72205 Abstract The present report describes respiratory-like activity recorded from intercostal muscles in the neonatal rat in vitro brain stem-spinal cord, rib-attached preparation. In this preparation from 1- to 4-day-old rats, spontaneous rhythmic and synchronized upward movements of the rib cage coincided with the recorded muscle activity. Spontaneous respiratory-like activity showed a frequency in the range of 0.05-0.2 Hz,with single-, double-, and mixed-burst patterns. Spontaneous activity declined over time, but increased in frequency as temperature increased. Multilevel recordings showed a cephalocaudal order of bursting of intercostal muscles. Brain stem transections at the prepontine level did not affect spontaneous fiequency, whereas premedullary transections resulted in an increase in spontaneous respiratory frequency. High spinal transections eliminated spontaneous respiratorylike activity. These results suggest that there is a well-organized pontomedullary pattern generator for respiratory-like activity in this preparation, which can be modulated by temperature. The characteristics of these electromyographic (EMG) recordings allow comparison with previous in vitro studies of respiratory-like activity using nerve activity and in vivo studies using EMG activity. These results provide basic information on the spontaneous activity of this preparation as a prelude to the study of the effects of electrical stimulation of the spinal cord to induce respiratory-like activity, as described in the companion article.

Key words respiration, in vitro brain stem-spinal cord, neonatal rat

1. To whom all correspondence should be addressed, at Department of Anatomy, University of Arkansas for Medical Sciences, 4301 West Markham Street, Little Rock, Arkansas 72205.

et al., 1985; Smith et al., 1986; Onimaru and Homma, 1987; Smith and Feldman, 1987a,b; Hilaire el al., 1989; Liu et al., 1990). In most of these studies, neurographic recordings from cranial and spinal nerves were used to assess respiration-related activity. In the present study, we used the rib-attached brain stem-spinal cord preparation to carry out a more detailed analysis of the respiratory movements induced by recording from intercostal muscles. Recently, an elegant study described the spatial and temporal patterns of activity of spinal and cranial nerves and of single medullary neurons in the brain stem-spinal cord preparation (Smith et al., 1990). This study compared the activity of nerve and cell recordings in vitro to electromyographic(EMG) activity obtained in the adult and neonatal rat in vivo. The present study provides some of the necessary complementary information for comparing EMG activity in the in vitro preparation directly to in vivo experimental paradigms. Moreover, a thorough study of the spontaneous patterns of activity present in the in vitro preparation is provided,

Somatosensory and Motor Research, Vol. 9, No. 4, 1992, pp. 313-326

Accepted June 26, 1992

In vitro preparations from lower vertebrates have been used with great success to assess rhythmic motor functions such as locomotion and respiration in adult (Lennard and Stein, 1977; Rovainen, 1977; Grillner et al., 1983; McClellan, 1984) and developing (Bekoff, 1976; Jacobson and Hollyday, 1982; Stehouwer and Farel, 1985) systems. Mammalian in vitro preparations have been developed (Otsuka and Konishi, 1974; Llinas et al., 1981) for the study of motoneuronal and reflex pathways (Otsuka and Konishi, 1974; Preston and Wallis, 1979; Bagust et al., 1985), as well as of respiratory activity (Suzue, 1984). Since the first report by Suzue (1984) demonstrating that the neonatal rat brain stemspinal cord preparation retained the ability to generate spontaneous and rhythmic respiratory-like activity, several studies have been carried out describing the nature of this activity (Harada et al., 1985; Murakoshi

313

HAMADA, GARCIA-RILL, AND SKINNER


including a detailed description of the effects of time and temperature on such activity. This information is crucial in assessing the viability of this preparation and the ways in which experimental conditions may affect the respiratory activity being observed. In addition, the present study provides a description of the effects of transecting the brain stem at various levels on the manifestation of spontaneous respiratory activity. These findings help localize neuronal aggregates that excite or inhibit the spontaneous respiratory drive in vitro. Finally, all of this information is.essential background for the study of the effects of electrical stimulation on respiratory movements, described in the companion article.

A

METHODS Surgery

A total of 73 neonatal Sprague-Dawley rats aged 1-4 days were deeply anesthetized with Penthrane until the withdrawal reflex to noxious stimulation was abolished. During anesthesia, the mean spontaneous respiratory rate was 33.1 -t 6.7/min, or a frequency of 0.55 Hz. The scalp was reflected, the skull cartilage was cut coronally, and the brain stem was initially transected caudal to bregma. The skull was cut sagitally along the midline, and the colliculi were visualized in order to make a precollicular transection. The skin over the head and back was removed together with the forelimbs, and the preparation was immersed in the bathing chamber. All the laminae from cervical to lumbar segments were removed with microscissors under a dissecting microscope ( x 12-25). The skin and muscles over the chest were removed, and the diaphragm was cut adjacent to the ribs and removed together with the viscera. The lungs and heart were removed, and the pelvis and hindlimbs were cut off by transecting the lower lumbar spine. Thoracic and lumbar vertebral bodies were removed by cutting the pedicles caudally, and the skull base and cervical vertebral bodies were removed via a rostral approach. This procedure exposed the entire spinal cord to artificial cerebrospinal fluid (aCSF) while maintaining the sternum and ribs attached, along with the intercostal muscles and their innervation (Fig. 1A). Recording

Double, hooked, fine wire (100-pm-diameter insulated tungsten) electrodes were inserted into the intercostal muscles. Pups aged 0 days (day of birth) were not used because of the fragility of these muscles. For most experiments, EMG recordings were carried out bilat314

B

C

FIGURE1. (A) Drawing of rib-attached brain stem-spinal cord preparation. Decerebration was made precollicularly. All ribs and intercostal muscles were left attached, maintaining their innervation. (B, C) Brain stem transection levels. Transections were made (a) between the midbrain and the pons (prepontine transection), (b) between the pons and the medulla (premedullary transection), and (c) at the junction of the medulla and the spinal cord (C, transection). SC, superior colliculus; IC, inferior colliculus; V, trigeminal nerve.

erally from the fourth intercostal muscles, except for additional recordings from the first and sixth intercostal muscles in certain experiments (see below). Recordings below the sixth intercostal muscle proved unreliable because of the fragile nature of these muscles. EMG recordings were amplified 5000- 10,000 times, stored on FM tape, and computer-analyzed. The description of the chamber and composition of the aCSF have been published previously (Atsuta et al., 1988, 1990; Iwahara et al., 1991). Briefly, the composition of the aCSF (in millimolar concentrations) was as follows: NaC1, 129;KCl, 3; NaH#04,0.5; CaCl?, 1.5;MgS04, 1; NaHC03, 20;dextrose, 30. Temperature

RESPIRATION ZN VZTRO: I. SPONTANEOUS ACTIVITY

and pH were measured with an indwelling temperature probe and a temperature-compensated pH probe, respectively. Temperature and pH were maintained at 25" ? 1°C and 7.4 k 0.1, respectively, except where otherwise specified (see below).


Data Analysis

Stored EMG recordings were digitized at a rate of lo00 Hz using an Amiga-based data analysis system (SNIP; Digital Dynamics, Santa Monica, CA). Durations and intervals between muscle bursts were analyzed with an accuracy of k 100 msec, and delays between bursts of muscles at different levels with an accuracy of ? 1 msec. All of the results are expressed as the mean plus or minus standard deviation. Differences were considered to be significant at p values under 0.05; paired or nonpaired Student's t tests were used.

and 7.50, while a constant temperature of 25°C was maintained. Each preparation was exposed to a certain pH in random order.

EFFECTSOF TRANSECTION A total of eight preparations underwent a series of transections. After recording of spontaneous activity in the standard precollicular preparation, the bmin stem was transected in sequence as shown in Figures 1B and 1C. The first transection was made at the rostral level of the pons (prepontine transection), the second at the rostral margin of the medulla (premedullary transection), and the third at the medullospinal junction (C1 spinal transection). EMGs were recorded after each transection (every 30 min). Each transection was accomplished in several steps; microscissors were used to minimize the effects of surgery. No significant variability in the level of the transections was detected.

Experimental Paradigms SPONTANEOUS ACTIVITY

STATISTICS

Intercostal muscle activity was recorded prior to each manipulation for 1-2 min. A total of 16 preparations, 4 from each age group (1-4 days), were observed for at least 3 hr in the standard environment (25" ? 1"C, pH 7.4 ? 0.1). The aCSF was changed every 15 min. EMGs were recorded every 30 min to assess the natural changes in spontaneous respiration, given constant conditions. In another six preparations, EMGs from the first and sixth intercostal muscles were recorded in addition to the fourth intercostal muscle, in order to assess the multilevel sequence of spontaneous contractions.

Statistical comparisons of multiple groups were carried out with the Kruskal-Wallis test, a nonparametric equivalent to the analysis of variance; comparisons of pairs of groups were carried out with the Wilcoxon two-sample test, a nonparametric equivalent to a oneway analysis of variance. Differences on these conservative tests were considered significant at the 0.05 level. The protocols for the experiments described herein were approved by the University of Arkansas for Medical Sciences Animal Care Committee.

EFFECTS OF TEMPERATURE

A total of five preparations were studied by changing the temperature of the aCSF from 25°C to 30°C. In this case, recordings at 25°C were carried out; the bath was changed quickly (5 sec) to 30°C; and then activity was measured as the temperature slowly returned to 25°C (approximately 10 min). In another six preparations, the temperature was changed from 25°C to 28°C to 30°C to 33"C, and EMGs were recorded at each temperature. In an additional four preparations, the aCSF was cooled to 22°C to compare spontaneous EMG activity at lower temperatures. pH was maintained at 7.4 throughout these experiments. EFFECTS OF pH

A total of seven preparations were exposed to small changes in the pH of the aCSF, 7.30, 7.35, 7.40, 7.45,

RESULTS General Characteristics of Spontaneous Activity RESPIRATORY PA'ITERN AND FREQUENCY

All of the preparations exhibited spontaneous, rhythmic, and synchronized upward movements of the rib cage. EMG activity recorded from intercostal muscles coincided with these upward movements. The basic patterns of spontaneous respiration consisted of single-, double-, and mixed-burst activity. The single-burst pattern exhibited a regularly repeated bilaterally synchronous muscle burst occurring with each movement of the rib cage. The double-burst pattern consisted of two sequential muscle bursts occurring in conjunction with each respiratory movement. The mixed-burst pattern contained single, double, and on occasion triple bursts of EMG activity. Table 1 shows the frequency of occurrence of single-, mixed-, and double-burst patterns and their respective mean frequencies. The frequency was measured as the time between the beginning of 315


TABLE1. Relationship between Respiratory Pattern and Rate (n = Number of Rats) ~

Pattern

Rate (Hz)


Single-burst pattern ( n = 46) Mixed-burst pattern (n = 16) Double-burst pattern (n = 11)

0.088 k 0.013 0.088 k 0.011 0.094 2 0.009

each burst in the single-burst pattern, or between the beginning of each burst of a group in the mixed- or double-burst pattern. Table 1 shows that the kquencies of single and mixed patterns were virtually the same, while the double-burst frequency was slightly higher but not statistically different. Figure 2 demonstrates that the same preparation could periodically show different patterns (although most preparations showed a single-burst pattern most of the time). In this case, a single-burst pattern (Fig. 2a) was observed to change into a short triple-burst and then a double-burst pattern (Fig. 2b), and to revert to a single-burst pattern (Fig. 2c). These transitions were evident within a 15-min period under constant conditions. Figure 3 shows spontaneous respiratory frequencies in another preparation, which showed a single-burst and a double-burst pattern. In these interburst interval plots, the single-burst mode showed one peak at about 11 sec, whereas the interburst interval plot of the double-

b.

burst pattern showed two peaks-one at about 2 sec representing the short intervals within each doublet, and a second peak at about 9 sec representing the interdoublet interval. This latter interval was slightly shorter than the single-burst interburst interval, although the difference was not statistically significant. Table 2 shows the relationship between age and respiratory frequency immediately after surgery. Basically, there was no difference in the spontaneous respiratory frequency across the first 4 postnatal days. EFFECT OF TIME ON RESPIRATORY RATE

Figure 4 shows the gradual changes evident in mean respiratory-likefrequency in the four different age groups across time. All of the 1-day and 2-day preparations maintained a stable frequency in the range of 0.050.12 Hz for over 3 hr, with a slow decrement in frequency over time. There was no significant difference between the 1-day and 2-day age groups until 2 hr. After that, the frequency observed in the 1-day preparations was significantly higher (p s 0.01) than in the 2-day preparations. On the other hand, two of four of the 3-day preparations ceased spontaneously bursting in 2.5 hr, while the others maintained frequencies below 0.04 Hz.In the May group, three of four preparations ceased bursting within 1.5 hr and the last after 2 hr. Basically, then, the most stable recording period for this preparation, at least under the given conditions, was within

H-d I '

I

C.

10 sec

FIGURE2. Sample records of spontaneous intercostal muscle EMGs. This preparation showed a single-burst pattern of respiration at a frequency of 0.076 Hz immediately after surgery (a). Ten minutes later, the pattern turned to a triple-burst pattern, followed by a double-burst pattern of respiration at a frequency of 0.088 Hz (b). Five minutes later, the pattern returned to a single-burst pattern at a frequency of 0.086 Hz (c). 316

RESPIRATION IN VZTRO: I. SPONTANEOUS ACTIVITY

RESPIRATION PATTERN COMPARISON


SINGLE

. o

.

i

1

i

i

ii

s

i

i

b

1'0

1'21'3

1'1

the first 1.5 hr after the end of surgery. Preparations from younger animals appeared to maintain their spontaneous bursting frequency for longer periods. Figure 5 depicts the relationship between age and the occurrence of the double-burst pattern over time. The percentage of double bursts out of the total number of bursts in each preparation appeared to be age-related and to be af€ected by time after surgery. In the I-day preparations, the percentage of double bursts decreased from 60% to 50% to 30% to 0% within 1.5 hr after surgery. Decreased rates of decay were observed in the 2-day and 3-day preparations, while there was little double bursting in 4-day preparations even early in the experiment. These findings indicate a greater incidence of double bursting in younger preparations and a general decrease in double bursting over time after surgery.

INTERVAL (sec)

ORDER OF ACTIVATION OF INTERCOSTAL MUSCLES

All of the preparations (n = 6) tested showed the same

0

1

z

3

4

5

B

7

8

a

1

0

1

1

1

2

1

3

INTERVAL (sec)

FIGURE 3. Interburst intervals of single- and double-burst patterns. All intervals were obtained from the same preparation. The intervals during the single-burst pattern were about 1 1 sec, while the double-burst pattern showed two peaks-one at about 2 sec representing the intervals within the double burst, and one at about 9 sec representing the intervals between each doublet.

sequence of intercostal muscle contractions when multiple levels of muscles were recorded. Figure 6 shows that the sequence of contractions was from upper to lower both in single trials (Fig. 6A) and in integrated, averaged records (Fig. 6B). Interestingly, the durations of muscle bursts appeared greater in the first intercostals than in the fourth than in the sixth. Table 3 shows the mean durations of first, fourth, and sixth intercostal muscle bursts. Statistical comparisons across the three muscle groups using the Kruskal- Wallis test suggested that the durations of the first and fourth intercostal muscle bursts were different ( p d 0.001) from that of the sixth intercostal muscle burst. Table 4 shows the mean interburst intervals between the first and fourth (shorter) and between the fourth and sixth (longer). Comparisons using a Wilcoxon twesample test showed that the interval was significantly greater (p d 0.005) between the fourth and sixth intercostal muscle bursts than between the first and fourth. E f e c t of Temperature

TABLE2. Relationship between Age and Respiratory Rate (n = Number of Rats) Age 1 day (n = 15) 2 days (n = 18) 3 days (n = 29) 4 days (n = 11)

Rate (Hz) 0.090 f 0.007 0.088 f 0.014 0.089 ? 0.011 0.090 f 0.016

Figure 7A shows the changes in spontaneous respiratory frequency observed at different temperatures. In the experimental paradigm in which the temperature was quickly changed from 25°C to 30°C and then allowed to cool slowly to 25"C, the initial respiratory frequency of 0.084 f 0.012 Hz increased significantly ( p S 0.01) to 0.197 f 0.025 Hz and gradually deceased to a frequency similar to the initial one (0.078 ? 0.013 Hz). Although the final frequency (following exposure to 30°C) was slightly lower (but not statistically significant) than initially, it remained stable for at least 1 hr. 317


NATURAL DEGRADING COURSE

* *

-

DAY 1

DAY 2 DAY 3


DAY 4

1 .o

0

2.0

3.0

TIME ( HRS

FIGURE 4. Changes in respiratory frequency in 1- to 4-day-old preparations within 3 hr postsurgery. The mean respiratory frequency gradually decreased throughout this period. Three-day-old preparations did not maintain their activity as well as 1- and 2-day-old preparations, while the 4-day-old group deteriorated within 2 hr.

DOUBLE BURST ACTIVITY

' 80

W (II

0

°

8

60

DAY1

< lz

DAY2

W

DAY 3

0 K

W

n

40

DAY4

20

0 0

30 TIME

60

90

(rnlnuter)

FIGURE 5. Percentage of double-burst activity. Younger preparations (1-2 days) showed increased frequencies of doubleburst patterns, and this pattern occurred more often within 30 min of surgery than later. Older preparations (3-4 days) showed a lower incidence of double-burst patterns. 318

RESPIRATION ZN VZTRO: I. SPONTANEOUS ACTIVITY

TABLE3. Burst Duration (n = 6) Muscle First intercostal Fourth intercostal Sixth intercostal

1st I n t

Duration (msec) 342.8 ~f.46.4 328.7 f 15.8 207.5 f 48.2*

* p s 0.001. TABLE4. Multilevel Burst Interval (n = 6)

1' 1 1121

First-fourth intercostal Fourth-sixth intercostal

200mr


*p

4 t h Int

23.8 f 9.8 55.2 f 13.3*

d 0.005.

Figure 7B shows the types of activity observed in the same preparation as the temperature was increased from 25°C to 28°C to 30°C to 33"C, essentially leading to an increase in the frequency of spontaneous muscle bursting. Figure 7C shows the mean frequencies of spontaneous respiration observed as the preparations were heated from 25°C in a stepwise fashion to 33°C. Within a few minutes of reaching 33"C, respiratory activity ceased irreversibly (three of five preparations) or showed a decreased mean frequency (two of five). Also shown are the mean frequencies observed following cooling to 23°C and 22°C. All spontaneous respiratory activity ceased at 22"C, but returned to "normal" mean frequency if exposed to aCSF at 25°C. These findings indicate that stable, reversible respiratory frequencies ranging from 0.05 to 0.2 Hz are evident between 23°C and 30°C.

4

61h Int

l-u

I I 12

200mr

FIGURE 6. Sample records of the spontaneous respiratory pattern. (A) Multiple EMG recordings were carried out in the first, fourth, and sixth intercostal muscles on the same side. Activation of the intercostal muscles occurred in a sequential fashion, from upper to lower, in the spontaneous condition. The duration of the first intercostal burst in this record @1) was 276 msec, of the fourth (D2) 254 msec, and of the sixth 0 3 ) 161 msec. The intervals between the beginning of the first and fourth (11) were 35 msec, and between the fourth and sixth (12) 40 msec. When the EMG burst durations of six preparations were averaged, they were as follows: D1 = 342.8 f 46.4; D2 = 328.7 ? 15.8; D3 = 207 48.2 msec. The mean EMG interval durations of six preparations were as follows: I1 = 23.8 * 9.8; I2 = 55.2 ? 13.3. (B) Rectified and averaged intercostal EMGs from the first, fourth, and sixth levels. Averages of 10 successive spontaneous bursts in a typical preparation also showed a cephalocaudal gradient in onset and duration. Because of the time constant of integration, the averaged durations and intervals differed slightly from those measured from the raw EMG. However, integrated records were clearer regarding onset of activity. The means and standard deviations of 10 consecutive bursts in this preparation were as follows: D1 = 371.7 ? 38.3; D2 = 334.0 f 43.8; D3 = 309.7 2 27.0; I1 = 18.3 f 11.8; I2 = 35.3 ? 11.4 msec.

Effect of p H

Figure 8 shows the mean spontaneous respiratory frequency when the pH of the aCSF was varied randomly across a narrow range (7.30-7.50). There was no significant respiratory frequency change in this group of subjects, although the frequency tended to decrease slightly at higher pH values. All of the preparations ( n = 5) in this group maintained stable respiratory frequencies throughout this 1.5-hr study. Effect of Brain Stem Transection Level

Figure 9 illustrates the effects of brain stem transections at various levels on the spontaneous respiratory burst frequency. There were no obvious changes in respiratory burst frequency following a prepontine transection (see transection levels in Figs. 1B and 1C). In some cases (two of eight), there was a change from a single-burst 319


B

25-c

A

;

EFFECTS OF TEMPERATUnE

-

0.a

0.n

0.11-


aw-

10 sec

C am

EFFECT OF TEMPERATURE

-

O M 0.-

0.00

-

-

. . . . . 30'. . . . . 25'. 25.

(belore)

(aller) lLYPERAlunE I % 1

.

1 1 ...................................

....................................

I

iEuCEnAiunE

I'CI

FIGURE 7. (A) Mean respiratory frequency at 25°C and 30°C. The frequency at 25°C was 0.084 * 0.012 Hz;this increased to 0.197 f 0.025 Hz at 30°C (p < 0.01). All five preparations maintained a stable but slightly slower (not significant) respiration frequency (0.078 ? 0.013 Hz)after being exposed at 30°C. (B) Sample records of respiration at four different temperatures indicated on the left. The preparation showed a regular rhythm at each temperature. The respiration became faster as the temperature increased with burst duration, decreasing markedly at 33°C. (C) Mean respiratory frequency at different temperatures in the range of 22-33°C. The respiratory frequency decreased in linear fashion as the temperature was lowered, increased as it was raised, and ceased at about 33°C.

to a double-burst pattern in one case, and Erom a doubleburst to a single-burst pattern in the other case. All of the other preparations retained their initial pattern following transection. Figure 9A shows recordings from a representative preparation in which the frequency observed in the precollicular preparation was similar to that seen following a prepontine transection. However, there was a significant increase in frequency following a premedullary transection and complete cessation of bursting following a C1 spinal transection. Figure 9B shows the mean spontaneous respiratory bursting frequency in this group of subjects. The frequency in the precollicular preparation (0.081 ? 0.009 320

Hz) was unchanged following a prepontine transection (0.078 t 0.012 Hz), but doubled following a premedullary transection (0.188 ? 0.031 Hz, p G 0.01). All of the preparations ceased respiratory activity following a C1spinal transection. There was no spontaneous EMG activity even after 1.5 hr posttransection at this level. DISCUSSION Characteristics of Spontaneous Respiration

A major contribution of the present experiments is to provide essential information on the actual respiratory


EFFECTS OF pH

..........................................................................

0.12 O

.

’

l

i

0.10 -1..........................................................................

0.08

.................

[ - -

I

I

..................................

o,s........................................................................


z

2

1

0.04

a

0.02

0.00

..........................................................................

- .......................................................................... I

I

I

1

PH FIGURE 8. Mean respiratory frequency with pH varied in the range of 7.30-7.50. There was no significant change in the spontaneous frequency, with a tendency for higher pH to decrese the frequency slightly.

movements being generated in the neonatal rat in vitro preparation by recording intercostal EMGs and correlating them with rib cage movement. These results provide a measure that can be compared directly with recordings in adults and neonates in vivo, which entail EMG and not nerve activity recordings. Several previous studies have reported the general characteristics of nerve activity in the in vitro preparation (Suzue, 1984; Harada et al., 1985; Smith et al., 1986; Onimaru and Homma, 1987; Smith and Feldman, 1987b; Feldman and Smith, 1989; Hilaire et al., 1989; McCrimmon et al., 1989; Liu et al., 1990). However, none have described the characteristics of EMG activity, although some reports have observed movements of the ribs. A recent study undertook a comparison of in vitro neurographic activity with in vivo EMG recordings (Smith et al., 1990). The main difference between in vivo and in vitro respiratory activity was identified as a lack of postinspiratory discharge in spinal motoneurons both in neonatal rats in vitro and in anesthetized neonatal rats in vivo, compared to adult in vivo preparations. This difference was attributed to developmental factors rather than to in vitro versus in vivo differences. Our results show that intercostal EMG activity consists of long-duration decrementing envelopes of activity (Fig. 6). This pattern agrees well with that observed by recording phrenic

ventral roots (see Fig. 7 of Smith et al., 1990). This is in contrast to the short-duration augmenting ramp observed in diaphragmatic EMG activity in vivo. At the present time, since no one has reported recording diaphragmatic activity, it is not known whether this pattern represents a difference between in vitro and in vivo patterns of activation, or a difference between diaphragmatic and intercostal EMGs regardless of preparation. Another possible source of this difference may be the absence of vagal afferent input. In an in vivo vagotomized preparation, the pattern of the diaphragmatic EMG increases in duration, although it does not clearly show a decrementing pattern, becoming more similar to the neurographic (Smith et al., 1990) and intercostal EMG (present report) patterns observed in vitro. Therefore, at least the long duration of the intercostal EMG burst may be attributed to a lack of vagal afferent input in this preparation. The absence of vagal afferent input may also be responsible for other basic differences in the characteristics of spontaneous respiratory activity in vivo versus in vitro. The mean respiratory frequency of rat pups was 0.55 Hz during anesthesia. Once the in vitro dissection was completed, the respiratory frequency was in the 0.1-Hz range (at 25°C). Murakoshi et al. (1985) suggested that the most important cause of the low 321

HAMADA, GARCIA-RILL, A N D SKINNER

A

I Pontlnr

2


Wedullary

Spinal

"

9 10 sec.

B TRANSECTION STUDY

TRANSECTION LEVEL

FIGURE 9. (A) Effects of brain stem transections at different levels. There was no obvious change in respiratory frequency after a prepontine transection, whereas a premedullary transection resulted in a doubling (0.078 0.012 Hz to 0.188 0.031 Hz)of the spontaneous respiratory frequency ( p d 0.01). Spontaneous respiration ceased after a spinal (C,) transection. (B) Changes in the mean frequency of spontaneous respiration showed a release or disinhibition following a premedullary transection.

*

322

*



respiratory frequency in the in vitro vagotomized preparation was the lack of afferent input, especially via the vagus. Intercostal muscle EMGs have been recorded in cats (Anderson and Lindsley, 1935; Greer, 1990), in dogs (DeTroyer and Ninane, 1986), and in humans (Campbell, 1955; Koepke et al., 1958; Taylor, 1960; DeTroyer and Sampson, 1982; DeTroyer and Estenne, 1984; Whitelaw, 1989). In the past, Suzue (1984) and Smith et al. (1988) used preparations with part of the rib cage attached in order to observe spontaneous movements caused by intercostal muscle contractions, but no EMG activity was recorded. Murakoshi et al. (1985) recorded thoracic movements by means of a force-displacement transducer connected to a part of the thorax. As our previous studies on locomotion have demonstrated, EMG recordings provide a more representative picture of the characteristics of actual movement. Stable EMG recordings from intercostal muscles, especially with the rib cage in a “normal” position such as that used in the present studies, permit us to conclude that respiratory-like activity is being represented by the EMGs being recorded. Our results from multilevel intercostal EMG recordings show a cephalocaudal sequence of muscle contractions. DeTroyer and Ninane (1986) found that in the in vivo dog, the external intercostals showed a cephalocaudal gradient of contractions, and the internal intercostals showed a caudocephalic order in both degree of activation and timing. Koepke et al. (1958) described the sequence of human intercostal muscle contractions as occumng from upper to lower interspaces as a deep b m t h was taken. The diaphragm was found to contract in advance of the first intercostal muscle. Judging by the cephalocaudal order of bursting, we believe that our recordings were from external intercostal (not internal intercostal) muscles, since these activities coincided with an upward movement of the rib cage (i.e., inspiration). Another interesting characteristic of these contractions was the longer duration of upper intercostal muscle bursts compared to lower. Figure 6B shows this relationship clearly, and serves to emphasize the cephalocaudal order of recruitment observed. It would be expected that the upper intercostals would remain contracted as long as the inspiration occurred. Deeper breaths would be expected to recruit lower intercostals for longer periods, but the upper intercostal must still remain contracted in order to execute a full inspiration. Suzue (1984) compared the motor pattern in vitro to gasping in anoxic states, mostly because of the lowfrequency and brief discharges during inspiration. However, Smith et al. (1990) argued that such a pattern may be accounted for by the absence of afferent input,

although it may share some common mechanisms with gasping. We agree with this conclusion. The pattern observed in the present experiments would not be in keeping with that produced during gasping. The relatively long duration of the upper intercostals and the orderly recruitment of these muscles are more in keeping with a pattern of deep breathing than with one of gasping (Koepkeetal., 1958).Interestingly,theorderly sequence of EMG onset described herein is similar to the proximodistal delay in the activity of agonists acting at different joints in the same limb, which is one of the main characteristics of a well-organized locomotor pattern (Atsuta et al., 1990). The respiratory frequency is strongly affected by the temperature of the bathing medium. Murakoshi et al. (1985) reported that the respiratory frequency in vitro was highest at 27-28”C, and that both raising and lowering the temperature resulted in a decrease in frequency. We have demonstrated in the present research that the respiratory frequency kept increasing until reaching a maximum frequency of 0.3 Hz at 33”C, but there was rapid degradation of the preparation despite the fact that this is well below body temperature. Our findings at lower temperatures (0.1 Hz at 25°C and 0.21 Hz at 30°C) correspond well with previous reports (Smith and Fkldman, 1987a, 0.1-0.2 Hz at 27-28°C; Murakoshi et al., 1985, 0.15 Hz at 28”C), but are lower than the spontaneouslkquencies reported by Harada et al. (1985, 0.1-0.23 Hz at 25°C). The fundamental basis for the hyperpnea of hyperthermia has not been established, but several tentative explanations have been advanced. They include (1) a physical effect of increased temperature on the cells of the respiratory center andor the peripheral chemoreceptors to exaggerate the reactivity of these respiratory control mechanisms (Cunningham and O’Riordan, 1957); (2) an increase in the central stimulus level because of accelerated cellular metabolism; (3) diminished efficiency of the buffering of COz by body fluids(Stadie et al., 1925);and (4) an increase in activity of hydrogen ions (Alexander et al., 1%l). As there was no hypothalamus or peripheral chemoreceptors (e.g., carotid bodies) in our in vitro preparation, the effects observed seem attributable to the respiratory generator in the brain stem itself in playing a major role in accelerating the respiratory cycle. The vulnerability of the in vitro preparation, especially to hypoxia, made it difficult to maintain its viability at a high temperature. Increased metabolism seems to have resulted in a relatively hypoxic condition and to have shortened the useful span of the preparation. Our studies on the effect of pH on respiratory frequency only explored a narrow range of values (7.3-73, which represents pH values usually observed 323



arations. The pattern of respiration generally did not change after transection, and the double-burst pattern, when present, was maintained. On the other hand, the respiratory frequency doubled after a premedullary transection. There was no double-burst pattern evident in this type of preparation. Hilaire et al. (1989) showed that the medullary respiratory generator was inhibited by the caudal ventrolateral pons in their in vitro transection study. Our data (0.078 Hz in prepontine; 0.188 Hz in medullary) are similar to Hilaire et al.'s (0.073 Hz and 0.186 Hz, respectively). They concluded that the respiratory pattern generator in the medulla is inhibited by the pons, and that a premedullary transection will release or disinhibit the respiratory generator in the medulla. Aoki et al. (1978, 1980) reported spontaneous rhythmic breathing within 2 hr after a C1 transection in 13 of 18 adult cats. There appears to be a group of upper cervical respiratory neurons (Duffin and Hoskin, 1987; Nonaka and Miller, 1991) that may be mediating this effect. We did not observe spontaneous breathing following a C1transection, possibly because the spinal cord may have been in shock and/or not enough time may have elapsed to allow recovery of such neurons to induce respiratory activity. On the other hand, perhaps the upper cervical inspiratory group is not fully functional at this age. The respiratory pattern observed in the present study consisted of single, double, and occasionally triple bursts of EMG activity. These patterns could be seen within preparations and in varying mixes. Moreover, younger preparations appeared to show more multiple bursting than older preparations, and the frequency of occurrence of multiple bursting decreased over time

in the bathing medium throughout an experiment. There was no obvious difference in respiratory frequency in this pH range. However, Harada et al. (1985) and Murakoshi et al. (1985) studied the effects of pH in the 7.04-7.8 and 7.1-8.1 ranges, respectively. These authors found that low pH values led to increased, and high pH values to decreased, respiratory frequency. Many investigators in the field of respiratory neurophysiology have been using the in vifro brain stemspinal cord preparation because it has a number of advantages and is particularly suited to the study of the respiratory motor system (Berger, 1990). However, one must recognize the disadvantages and limits of such preparations. The most notable limitation of the in v i m method is the age of the preparation. In order to have a viable preparation, only very young animals can be used, mostly in the 0- to 4-day-old range. There are few descriptions of the viability of these preparations. Smith et al. (1990) recently suggested that the frequency of spontaneous nerve activity was stable over the initial 4 hr, followed by a decline in frequency and irregularities in the rhythm. Our data suggest that this degradation may be faster in older animals (3-4 days) than in younger preparations (1 -2 days). Although our preparations are different from those previously reported in that these have the rib cage attached, the exposure of the spinal cord to aCSF should not differ greatly from that in other preparations. Our results do suggest that it is essential to note the age and time after surgery in relationship to the gathering of important data related to respiratory frequency. There was no significant difference in respiratory frequency between precollicular and prepontine prepI

I

I

I

respiratory burst

t hrcshold level of transmitter

111

111

111

1 1 1

FIGURE10. Hypothesis to explain the changes of respiratory pattern observed in the rib-attached in vitro brain stem-spinal cord preparation. In this model, the level of a certain transmitter possesses a fairly regular cycle, and the relationship between this cycle and a threshold for respiratory firing fluctuates in this condition. The respiratory pattern changes from single (a) to double (b) to triple (c) bursting as the period during which the level of the transmitter exceeds the threshold increases (i.e., longer availability of transmitter). 324


RESPIRATION IN VITRO: I. SPONTANEOUS ACTIVITY

after surgery. These changes are interpreted as a difference in excitability, in which the respiratory threshold can change but the frequency of the pattern remains an integral part of the circuit. This explains the lack of difference in the frequency of single bursts compared to that of groups of multiple bursts. What may generate multiple bursts is not a change in the respiratory cycle, but the relationship between transmitter availability and cycle phase. If the threshold increases, only enough availability of transmitter to generate one burst may be manifested. If the threshold decreases, higher availability of transmitter may be present, and multiple bursts may result. Figure 10 is a model of the hypothesis proposed, but is incomplete in explainingwhy the burst does not increase in duration (i.e., become more “tonic”) as a result of transmitter availability, rather than generating multiple, similar-duration bursts. In summary, these results present significant new evidence regarding the characteristics of respiratory activity in the neonatal rat in vitro preparation. Much additional evidence is needed to bridge the gap between the in vitro and in vivo preparations and between neonatal and adult paradigms. However, the evidence provided herein helps establish the limits, advantages, and disadvantages of the in vitro rib-attached brain stemspinal cord preparation. The basic characteristics of spontaneous respiratory activity in this preparation form a foundation for studying the effects of electrical modulation and control of respiratory activity, as described in the companion article. ACKNOWLEDGMENTS

This research was supported by U.S. Public Health Service Grant No. NS20246.

REFERENCES

ALEXANDER, S. C., R. GELFAND, and C. J. LAMBERSTSEN (1961) The pK’ of carbonic acid in cerebrospinal fluid. J. Biol. Chem. 236: 592-5%.

ANDERSON, F. M., and D. B . LINDSLEY (1935) Action potentials from intercostal muscles before and after unilateral pneumectomy. J. Lab. Clin. Med. 20: 623-628. AOKI,M., H. WATANABE, N. EBATA, and S. MOM(1978) Spontaneous respiratory activity in spinal cats. Brain Res. 157 376-380. ATSUTA, Y.,E. GARCIA-RILL, and R. D. SKINNER (1988) Electrically induced locomotion in the in virro brainstem-spinal cord preparation. Dev. Brain Res. 42: 309-312. ATSUTA, Y.,E. GARCIA-RILL, and R. D. SKINNER(1990) Characteristics of electrically induced locomotion in rat in virro brain stem-spinal cord preparation. J. Neurophysiol. 64: 727-735. ANDERSON, F. M., and D. B. LINDSLEY (1935) Action potentials from intercostal muscles before and after unilateral pneumectomy. J. Lab. Clin. Med. 20: 623-628. BAGUST, J., M. E. M. KELY,and G. A. KERKUT(1985) An isolated

mammalian brainstem-spinal cord preparation suitable for the investigation of descending control of motor activity. Brain Res. 327 370-374.

BEKOFF,A. (1976) Ontogeny of leg motor output in the chick embryo: A neural analysis. Brain Res. 106: 271-291. BERGER,A. J. (1990) Recent advances in respiratory neurobiology using in virro methods. Am. J. Physiol. 259: L24-L29. CAMPBELL,E. J. M. (1955) An electromyographic examination of the role of the intercostal muscles in breathing in man. J. Physiol. 129: 12-26.

CUNNINGHAM, D. J. C., and J. L. H. O’RIORDAN (1957) The effect of a rise in the temperature of the body on the respiratory response to carbon dioxide at rest. Quart. J. Exp. Physiol. 42: 329-345. DETROYER, A., and A. ESTENNE(1984) Coordination between rib cage muscles and diaphragm during quiet breathing in humans. J. Appl. Physiol. 5 7 899-906. DETROYER, A., and V. NINANE(1986) Respiratory function of intercostal muscles in supine dog: An electromyographic study. J. Appl. Physiol. 60: 1692-1699. DETROYER, A., and M. G. S ~ h f P s (1982) o ~ Activation of the parasternal intercostals during breathing efforts in human subjects. J. Appl. Physiol. 52: 524-529. DUFFIN,J., and R. W. HOSKIN(1987) Intracellular recordings from upper cervical inspiratory neurons in the cat. Brain Res. 435: 351-354.

FELDMAN,J. L., and J. C. SMITH(1989) Cellular mechanisms underlying modulation of breathing pattern in mammals. Ann. NY Acad. Sci. 563: 114-130. GREER,J. J. (1990). Distribution of muscle fiber types and EMG activity in cat intercostal muscles. J. App. Physiol. 69: 1208- 1211. GRILLNER, S., P. WALLEN,A. MCCLELLAN, K. SIGVARDT,T. WILLIAMS,and J. FELDMAN (1983) The neural generation of locomotion in the lamprey: An incomplete account. In Neural Origin of Rhythmic Movements, A. Roberts and B. Roberts, eds., pp. 285303, Society for Experimental Biology, London. HARADA,Y., M. KUNO,and Y. Z. WANG(1985) Differential effects of carbon dioxide and pH on central chemoreceptors in the rat in virro. J. Physiol. 368: 679-693. HILAIRE. G., R. MONTEAU,and S. ERRCHIDI(1989) Possible modulation of the medullary respiratory rhythm generator by the notadrenergic As area: An in virro study in the newborn rat. Brain Res. 485: 325-332.

IWAHARA, T.,Y. ATSUTA, E. GARCIA-RILL, and R. D. SKINNER (1991) Locomotion induced by spinal cord stimulation in the neonate rat in virro. Somatosens. Mot. Res. 8: 281-287. JACOBSON, R. D., and M. HOLLYDAY (1982) Electrically evoked walking and fictive locomotion in the chick. J. Neurophysiol. 48: 257-270. KOEPKE,G. H.,E. M. SMITH,A. J. MURPHY, andD. G. DICKINSON (1958) Sequence of action of the diaphragm and intercostal muscles during respiration: I. Inspiration. Arch. Phys. Med. Rehab. 39: 426-430.

LENNARD, P. R., and P. S. G. STEIN(1977) Swimming movements elicited by electrical stimulation of turtle spinal cord: I. Low-spinal and intact preparations. J. Neurophysiol. 40: 768-778. LIU, G., J. L. FELDMAN,and J. C. SMITH(1990) Excitatory amino acid mediated transmission of inspiratory drive to phrenic m e toneurons. J. Neurophysiol. 64: 423-436. LLINAS,R., Y.YAROM, and M. SUGIMOIU (1981) Isolated mammalian brain in virro: New technique for analysis of electrical activity of neuronal circuit function. Fed.Proc. 4 0 224-2245. MCCLELLAN, A. D. (1984) Descending control and sensory gaiting of “fictive” swimming turning responses elicited in an in virro brainstem-spinal cord preparation of the lamprey. Brain Res. 302: 151-162.

MCCIUMMON, D. R., J. C. SMITH,and I. L. FELDMAN (1989) In325

HAMADA, GARCIA-RILL, AND SKINNER volvement of excitatory amino acids in neurotransmission of inspiratory drive to spinal respiratory motoneurons. J. Neurosci. 9:


1910-1921.

MURAKOSHI, T., T. SUZUE,and S. TAMAI(1985) A pharmacological study on respiratory rhythm in the isolated brainstem-spinal cord preparation of the newborn rat. Br. J. Pharmacol. 86: 95- 104. NONAKA, S., and A. D. MILLER(1991) Behavior of upper cervical inspiratory propriospinal neurons during fictive vomiting. N. Neurophysiol. 65: 1492-1500. ONIMARU, H., and I. HOMMA (1987) Respiratory rhythm generator neurons in medulla of brainstem-spinal cord preparation from newborn rat. Brain Res. 403: 380-384. O ~ ~ U KM., A , and S. KONISHI (1974) Electrophysiology of mammalian spinal cord in vitro. Nature 252: 733-734. PRESTON,P. R., and D. I. WALLIS(1979) Dorsal root potentials recorded in the isolated spinal cord of the neonatal rat. J. Physiol. (Lond.) 289: 84P. ROVAINEN, C. M. (1977) Neural control of ventilation in the lamprey. Fed. Proc. 36: 2386-2389. SMITH,J. C., and J. L. FELDMAN (1987a) In vitro brainstem-spinal cord preparations for study of motor systems for mammalian respiration and locomotion. J. Neurosci. Meth. 21: 321-333. SMITH,J. C., and J. L. FELDMAN (1987b) Central respiratory pattern generation studied in an in vitro mammalian brainstem-spinal cord preparation. In Respiratory Muscles and Their Neuromotor

326

Conrrol, G. Sieck, S. C. Gandevia, and W. E. Cameron, eds., pp. 27-36, Alan R. Liss, New York. and J. L. FELDMAN (1986) Chemical SMTH,J. C., E. GARCIA-RILL, activation of mammalian locomotion in vitro. Soc. Neurosci. Abstr. 12: 386. SMITH,J. C., J. J. GREER,G. LIU,and J. L. FELDMAN (1990) Neural mechanisms generating respiratory pattern in mammalian brain stem-spinal cord in vitro: I. Spatiotemporal patterns of motor and medullary neuron activity. J. Neurophysiol. 64: 1149- 1169. SMITH,J. C., G. LIU,and J. L. FELDMAN (1988) Inhcellular reconing from phrenic motoneurons receiving respiratory drive in vitro. Neurosci. Lett. 88: 27-32. STADIE,W. C., J. H. AUSTIN,and H. W. ROBINSON (1925) The effect of temperature on the acid-base protein equilibrium and its influence on the CO2 absorption curve of whole blood, true and separated serum. J. Biol. Chem. 66: 901-920. STEHOUWER, D. J., and P. B. FAREL(1985) Development of locomotor mechanisms in the frog. J. Neurophysiol. 53: 1453-1466. SUZUE, T. (1984) Respiratory rhythm generation in the in vitro brain stem-spinal cord preparation of the neonatal rat. J. Physiol. (Lond.) 354: 173-183.

TAYLOR, A. (1960) The contribution of the intercostal muscles to the effort of respiration in man. J. Physiol. (Lond.) 151: 390-402. WHITELAW, W. A. (1989) Patterns of intercostal muscle activity in humans. J. Appl. Physiol. 6 7 2087-2094.

Spontaneous respiration with the Bain breathing circuit.

Spontaneous electrical activity of the rat epididymis in vitro.

Respiration in vitro: II. Electrical stimulation.

Letter: Sequelae of delayed spontaneous respiration in breech infants.

Ligandin reverses bilirubin inhibition of liver mitochondrial respiration in vitro.

Regulation of spontaneous activity and oscillatory spike firing in rat midbrain dopamine neurons recorded in vitro.

Spontaneous EEG activity and spontaneous emotion regulation.

A comparative statistical study of hippocampal neuronal spontaneous spike activity in situ and in vitro.

Spontaneous Gamma Activity in Schizophrenia.

Analysis of spontaneous activity of superficial dorsal horn neurons in vitro: neuropathy-induced changes.

In vitro cardiac models of dog Purkinje fibre triggered and spontaneous electrical activity: effects of nicorandil.

Differential modulation of spontaneous and evoked thalamocortical network activity by acetylcholine level in vitro.

Investigations into the spontaneous contractile activity of isolated human uterine arteries in vitro.

Effects on respiration of nonintubated anesthesia in thoracoscopic surgery under spontaneous ventilation.

Controlled ventilation or spontaneous respiration in anesthesia for tracheobronchial foreign body removal: a meta-analysis.

Rapid Return of Spontaneous Respiration after General Anesthesia with Sugammadex in a Patient with Myasthenia Gravis.

Effects of mechanical ventilation and spontaneous respiration on hemodynamics in calves with total artificial hearts.

Miconazole in coccidioidomycosis. I. Assays of activity in mice and in vitro.

Blowing off the Retained CO2 During Spontaneous Respiration in Neurosurgery: A Novel Way!

Brexpiprazole I: in vitro and in vivo characterization of a novel serotonin-dopamine activity modulator.

Evaluation of RNA chaperone activity in vivo and in vitro using misfolded group I ribozymes.

Spontaneous mechanical activity in depolarized frog ventricle.

Spontaneous activity in denervated mouse diaphragm muscle.

Spontaneous Parahisian Accessory Pathway Activity.