Non-vagal reflex effects on medullary inspiratory neurons during inspiratory loading

193

Respiration Physiology (1985) 60, 193-204 Elsevier

NON-VAGAL REFLEX EFFECTS ON MEDULLARY INSPIRATORY NEURONS DURING INSPIRATORY LOADING

R. SHANNON, W.T. SHEAR, A.R. MERCAK, D.C. BOLSER and B.G. LINDSEY Department of Physiology, College of Medicine, University of South Florida, Tampa, FL 33612, U.S.A.

Abstract. Studies were conducted to compare the first-breath responses of medullary Dorsal and Ventral

Respiratory Group inspiratory (I) neurons to the mechanical loading (tracheal occlusion, TO) of inspiration in unanesthetized (decerebrate) and anesthetized (Dial ®) vagotomized cats, and to determine the sources of the sensory activity causing the changes in I-neuron activity. In decerebrate cats, TO resulted in a prolongation of the firing duration in 49~ of the 1-neurons. There was a delayed onset of firing in 7~o of the I-neurons. The responses of 1-neurons to TO in anesthetized cats were similar to the responses in decerebrate cats. Changes in 1-neuron activity with TO were still present in cats with their cervical (C3-7) or thoracic (T1-9) dorsal roots cut, and absent when both cervical and thoracic dorsal roots were cut. The most probable sources of the cervical and thoracic afferent information altering medullary I-neuron activity during loading are the diaphragm and inspiratory intercostal muscles. Afferent information Cat Diaphragm

Dorsal root Inspiratory neurons Medulla oblongata

Respiratory centers Respiratory loading Vagus nerve

It is well known that mechanoreceptors located in intercostal muscles can influence the brainstem mechanisms controlling the respiratory muscles. Reports from this laboratory (Shannon etaL, 1972; Shannon, 1980; Shannon and Freeman, 1981)have concluded that the predominant action of mechanoreceptors in the thoracic wall on brainstem respiratory activity is to reduce or terminate inspiratory drive. These studies suggested there is little if any enhancement of medullary inspiratory drive by mechanoreceptors in intercostal muscles. One of these studies (Shannon et al., 1972) utilized external mechanical loading of inspiration as a method for increasing the activity of inspiratory intercostal muscle spindle endings and tendon organs in an effort to study their effects on brainstem respiratory control. The only change noted in medullary Ventral Respiratory Group inspiratory neuron activity during the first-loaded breath was a small Accepted for publication 23 February 1985 0034-5687/85/$03.30 © 1985 Elsevier Science Publishers B.V. (Biomedical Division)

194

R. SHANNON et al.

reduction in firing rate and/or delay in onset of activity in some neurons. It was suggested that this neural reflex effect was most probably due to stimulation of tendon organs in the inspiratory intercostal muscles and/or diaphragm. In a recent preliminary study, we observed that there was an increase in activity of some medullary inspiratory neurons during inspiratory loading (tracheal occlusion) which was not observed in the earlier study (Shannon et aL, 1972). The only obvious difference in the animal preparations was that anesthetized (Dial ®) cats were used in the earlier study and cats in the latter study were decerebrated (unanesthetized). Thus, we wondered whether the depressing effects of anesthesia might have blocked the action of the mechanoreceptors responsible for the facilitation of medullary inspiratory neuron activity during loading. Experiments were conducted to answer this question and to further investigate the extravagal reflex(es) in more detail. This reflex response was particularly interesting because it was the first evidence, that we are aware of, suggesting that inspiratory muscle mechanoreceptors may enhance or prolong the activity of medullary inspiratory neurons. The first-breath response of medullary Dorsal and Ventral Respiratory Group inspiratory neurons to inspiratory loading was examined in decerebrated, vagotomized, spontaneously breathing cats. Changes in neuron activity during the first-loaded breath result from rapid neural reflexes and not changes in chemical drive. Subsequent experiments were conducted to determine the source(s) of the afferent information altering medullary inspiratory neuron activity during loading. The most likely source(s) were thought to be the inspiratory intercostal muscles and/or diaphragm. Sensory information from these potential sources was interrupted by sectioning the appropriate spinal dorsal roots. These studies further help characterize reflexes that can influence central respiratory control.

Methods

All experiments, except one series, were performed on mid-collicular decerebrated, vagotomized (cervical), spontaneously breathing cats (2.5-4.0 kg). The decerebration was performed under sodium thiopental anesthesia (30 mg/kg). The experiments were begun at least 2 h after the decerebration in order to allow the potential effects of the anesthetic on the reflexes to subside. One series of experiments was conducted on Dial ® (0.6 ml/kg) anesthetized cats. Rectal temperature was maintained at 37.5 + 0.5 °C by a hot-water heating pad. Arterial blood pressure was monitored and arterial blood was periodically analyzed for Po2, Pco2 and pH. These values were used as an index of the animal's general cardiovascular and respiratory status; all experimental data presented are from animals with normal blood pressure and blood gases. Arterial hypotension was sometimes observed after the decerebration and phenylephrine hydrochloride (vasoconstrictor) and Dextran were infused intravenously to maintain arterial blood pressure in the normal range (100-125 mm Hg). Metabolic acidosis occurred in some animals

MEDULLARYINSPIRATORYNEURON RESPONSES TO LOADING

195

and sodium bicarbonate was infused to adjust the arterial pH. The animals were vagotomized to eliminate the effects of vagal afferent information on medullary respiratory activity, and thereby expose the weaker effects of respiratory muscle afferent activity. Five series of experiments were conducted in the following vagotomized preparations: (a) Mid-collicular decerebration, 9 cats. (b) Decerebrate, thoracic dorsal rhizotomies (T1-T9), 6 cats. (c) Decerebrate, cervical dorsal rhizotomies (C3-C7), 8 cats. (d) Decerebrate, cervical plus thoracic dorsal rhizotomies (C3-T9), 10 cats. (e) Dial ® anesthetized, 4 cats. Bilateral thoracic and cervical dorsal rhizotomies interrupted proprioceptive afferent pathways from the intercostal muscles and diaphragm, respectively. Dorsal roots were exposed by dorsal laminectomies. Following all surgery, the animal was allowed to recover from the trauma for 1 h. Only those animals with stable breathing patterns were used for these experiments. The external intercostal muscles and diaphragm were active in all experimental animals, as determined by visual observation. It is well known that there is a diminution of external intercostal motor activity in cats following thoracic dorsal rhizotomies, which returns within about an hour (Shannon and Zechman, 1972). Diaphragm motor activity in the cat appears to be unaffected by cervical dorsal rhizotomies (Euler, 1973). A two-way breathing valve (dead space 2.7 ml) was connected to a tracheal cannula. Tracheal pressure changes were used as an indicator of the respiratory phases. The mechanical loading of inspiration (tracheal occlusion) was accomplished by blocking the inspiratory port of the breathing valve during the expiratory pause (at FRC) and during the subsequent inspiration. Examining the first-loaded inspiration allows one to investigate the neural reflex responses to loading before significant changes in chemical respiratory drive have time to develop. The dorsal surface of the medulla was surgically exposed and the cerebellum partially removed by suctioning. Extracellular recordings of single inspiratory neurons in the medullary Dorsal and Ventral Respiratory Groups were made with tungsten microelectrodes (10-15 M~). Coordinates for the DRG neurons were from the level of the obex to 2 mm rostral, 1.5-2.3 mm lateral to the midline, and 1.0-2.5 mm below the dorsal surface. Coordinates for the VRG neurons were from 1.0 mm caudal to 2.0 mm rostral to the obex, 2.8-4.5 mm lateral to the midline, and 2.8-4.5 mm below the dorsal surface. The neurons were identified as inspiratory by their relationship to the tracheal pressure changes. In three cats (decerebrated, vagotomized), Ventral Respiratory Group I-neurons with axonal projections down the spinal cord (bulbospinal) were antidromically identified by collision testing (collision of ortho- and antidromic spikes). Antidromic spikes were generated by stimulating the spinal cord at T2. Following laminectomies, two stainlesssteel electrodes were inserted via a micromanipulator through the dorsal surface of the cord. The electrodes were contralaterally positioned approximately 1.5 mm from the

196

R. SHANNON et al.

lateral borders of the cord, at a depth which insured that the ventrolateral columns on both sides of the cord were stimulated. Square wave pulses (3-10 V) of either polarity with a duration of 0.05-0.1 msec were delivered through an isolation transformer. The data were quantitatively analyzed by comparing cycle triggered histograms (CTHs) of I-neuron activity obtained during control and loaded breaths. Inspiratory neuron activity and voltage pulses indicating the onset of the control and loaded breaths were recorded on magnetic tape for later generation of CTHs by computer. The voltage pulses marking the onset of inspiratory activity were generated by passing the tracheal pressure signal through a time/voltage window discriminator. The CTHs were generated by averaging multiple (20-30) control and loaded breaths. The sequence for collecting data was as follows: control breath, loaded breath, five to ten recovery breaths, control breath, etc. The firing rates of the neurons returned to control values within five recovery breaths.

Results

Decerebrated, vagotomized. The activity of some Dorsal and Ventral Respiratory Group I-neurons was altered during the mechanical loading of inspiration (tracheal occlusion at FRC). The response most often observed was a prolonged firing duration (mean increase in duration = 26.5~, range = 5-65~o) which is illustrated in four neurons in fig. 1. The percentage of neurons showing this response varied between animals (n = 9) and ranged from none in two animals to all of them in three others. Twenty of 41 neurons tested showed this response (table 1). In most neurons showing an increase in duration of activity, the differences between the unloaded and loaded breaths were obvious on the audio monitor and in the oscilloscope tracings of the amplified signals. To further insure that the prolonged activity present in the CTH of all loading trials was not due to a random prolonged breath, CTHs generated from different segments of the data were examined for prolonged activity of similar magnitude as seen for all the trials. Only two of the neurons judged as having prolonged activity when all trials were first examined were then excluded when the segmented data were examined. There was a hint of an increased firing rate in a few neurons and a decreased firing rate in others during the rising phase of l-neuron activity with TO. These differences between the control and loaded CTHs were tested for statistical significance by comparing the mean and standard error of the firing rates of the individual control and test cycles during the time when differences were suggested by visual inspection of the CTHs. These differences were statistically significant (Student's t-test, P < 0.05) in only one of the neurons showing a decrease in activity. Three I-neurons that began firing after the onset of the inspiratory phase showed a delay in their onset of activity during TO (fig. 2). When CTHs of segments of the data and the total number of trials were compared, the shift in activity was of similar magnitude; this confirms that the shift was real and not due to random variation.

M E D U L L A R Y I N S P I R A T O R Y N E U R O N R E S P O N S E S TO L O A D I N G CONTROL

_

197

LOAD

5 0 . 7 5PlrS_5/g~C I

_

~ . 8 ~llt.f..$/~C I

A

B

5.0 8EC

0.0

52.0

0.0

5P lt,~/.~C

6.5 SEC

5 0 . 7 .SP[ItF2i/.SEC I

I

D

C

0.0

5.0

0.0

8EC

4.5 8EC

Fig. 1. Cycle triggered histograms of four medullary 1-neurons with different firing patterns showing augmentation of activity during tracheal occlusion. Y-axis is neuron firing rate. X-axis is time in sec. (A) VRG, neuron, 18 control and test cycles and 50 msec bin widths. ( B ) V R G neuron, 20 cycles and 65 msec bin widths. (C) D R G neuron, 15 cycles and 50 msec bin widths. (D) D R G neuron, 20 cycles and 45 msec bin widths.

TABLE 1 Responses of medullary l-neurons to tracheal occlusion in decerebrated, vagotomized cats. Symbols for change in neuron activity are: PA = prolonged activity, D = decrease, DO = delay in onset, and NC = no change. D R G = dorsal respiratory group. V R G = ventral respiratory group. Cats

No. of neurons DRG

A Activity VRG

PA

D

DO

NC

1

5

3

0

0

2

2 3 4 5 6 7 8 9

3 8 4 5 5

0 3 4 5 1 0 1 3

0 0 0 0 0 1 0 0

1 1 0 0 0 1 0 0

3 5 0 0 4 5 2 0

30

20

1

3

21

Total

5 3 3 11

198

R. S H A N N O N et al. CONTROL

LOAD

5 8 . 6 5PIKE.5/SEC I

52.B sPlr,.Fr.S/'c~E I

=

A

0.0

3.5

SEC

0.0

3.0

SEG

Fig. 2. Medullary I-neurons showing a reduction and a shift in activity during tracheal occlusion. (A) D R G neuron, 13 control and test cycles and 25 msec bin widths. (B) V R G neuron, 10 cycles and 45 msec bin widths. (C) VRG neuron, 24 cycles and 35 msec bin widths. (D) D R G neuron, 25 cycles and 50 msec bin widths.

There was no change in activity in 21 of 41 neurons during TO. Some examples are shown in fig. 3. The axonal projections of these 41 neurons were not determined; however, the areas from which these I-neurons were recorded are known to contain large numbers of CONTROL

_

LOAD

84.a

51.2 5PIr,.f.S/SIEC

5PIr,E.S/SEE I

A

o.o

4.5

8EC

0.0

3 2 . 4 5PIr,.E.S/SEr I

4.0 8EC

10"2.9 .sf'IrJ~/SEC I

.:

~

=

~'-~

C •

0.0

5.0

SEC

0.0

!

!

2.0 SEC

Fig. 3. Medullary I-neurons showing no change in activity during tracheal occlusion. (A) VRG neuron, 25 control and test cycles and 45 msec bin widths. ( B ) V R G neuron, 20 cycles and 40 msec bin widths. (C) VRG neuron, 20 cycles and 50 msec bin widths. (D) D R G neuron, 14 cycles and 20 msec bin widths.

MEDULLARY INSPIRATORY NEURON RESPONSES TO LOADING

199

premotor neurons whose axons descend down the spinal cord to drive phrenic and inspiratory intercostal motoneurons (Euler, 1983). Thus, it is likely that many of the neurons recorded were of this type (bulbospinal). In order to confirm that the most common change observed during TO occurred in bulbospinal I-neurons, antidromic identification of bulbospinal I-neurons was carried out in one series of experiments. In three animals, six Ventral Respiratory Group 1-neurons were antidromically identified with positive collision tests as having axons projecting down the spinal cord. There was a prolonged fu'ing time in three of these neurons during TO, whereas there was no effect on the other three neurons. These results show that TO elicits an increase in activity of some medullary I-neurons that are involved in the control of inspiratory muscle activity. None of these six neurons showed a delay in the onset of activity during TO. The lack of this type of change in activity in these bulbospinal neurons does not imply that it does not occur in some bulbospinal neurons. Only a limited population of identified bulbospinal 1-neurons was tested because these experiments were extremely time consuming. The animals had to be paralyzed during electrical stimulation of the spinal cord and once an 1-neuron was identified as bulbospinal, the neuron response to TO could not be determined until the effects of the paralyzing agent had completely disappeared and normal spontaneous breathing had returned. It usually took at least 1 h for normal breathing to return.

Decerebrated, vagotomized, thoracic dorsal rhizotomies (T1-T9).

There were still changes in I-neuron activity in this preparation (table 2). Eleven of 34 neurons showed a prolonged f'n'ing duration (mean increase in duration = 18 %, range = 7-33 %). Two neurons showed a delay in the onset of firing, similar to the neurons in fig. 2. No neurons responded to TO with a statistically significant decrease or increase in firing rate.

Decerebrated, vagotomized, cervical dorsal rhizotomies (C3-C7).

The response of Ineurons to TO in this preparation was qualitatively similar to that in animals with cervical dorsal roots intact (table 2). Ten of 45 neurons showed a prolonged frring duration (mean increase in duration = 17.6%, range = 11-28%). Two neurons responded to TO with a delay in the onset of firing. There was no change in activity in 35 neurons.

Decerebrated, vagotomized, cervical plus thoracic dorsal rhizotomies (C3-Tg).

None of

the 43 I-neurons tested changed their activity during TO (table 2).

Anesthetized, vagotomized. Since the results in unanesthetized, decerebrated cats were different (i.e., evidence of increased I-neuron activity with TO) than a previous study (Shannon et al., 1972) in anesthetized (Dial ®) cats, we re-examined the response of VRG I-neurons to TO in anesthetized (Dial ®) cats. Since the response of Dorsal and Ventral Respiratory Group I-neurons to TO are similar, no dorsal neurons were tested. Unlike the previous study in anesthetized cats, there was a prolonged firing time in some I-neurons (8 of 26 neurons) following TO (table 1). Tracheal occlusion also caused a delay in the onset of fn-ing in three late 1-neurons.

200

R. S H A N N O N et al.

TABLE 2 Summary of responses of medullary I-neurons to tracheal occlusion in vagotomized cats. Symbols for change in neuron activity are: PA = prolonged activity, D = decrease, DO = delay in onset, and NC = no change. CDR = cervical dorsal rhizotomies. TDR = thoracic dorsal rhizotomies. D R G = dorsal respiratory group. VRG = ventral respiratory group. No. of cats

No. of neurons

A Activity PA

D

DO

NC

Decerebrate DRG VRG

3 6

11 30

4 16

1 0

1 2

7 14

Decerebrate + TDR DRG VRG

3 3

14 20

3 8

0 0

1 1

11 12

Decerebrate + CDR DRG VRG

4 4

24 21

7 3

0 0

2 0

17 18

Decerebrate + CDR + TDR DRG VRG

3 7

14 29

0 0

0 0

0 0

14 29

Anesthetized DRG VRG

0 4

26

8

0

3

18

Discussion Results from this study in anesthetized and unanesthetized vagotomized cats showed that medullary Dorsal and Ventral Respiratory Group inspiratory neuron activity was reflexly altered when inspiration was impeded by tracheal occlusion (TO). The strength of these reflexes varied between animals and not all neurons were affected by the mechanical load. The response observed most often was a prolonged duration of firing. The neurons were recorded from regions which are known to contain large numbers of I-neurons that project to spinal motoneurons (bulbospinal). Some I-neurons specifically identified as bulbospinal by antidromic stimulation techniques showed a prolonged duration of activity with TO. Thus, TO results in an alteration in mechanoreceptor activity leading to augmentation of some medullary I-neurons that drive inspiratory muscle motoneurons. There was a delay in the onset of fLring in a few late inspiratory neurons during TO. This response could be due to the same mechanisms responsible for the prolonged duration. For example, if these late I-neurons inhibit bulbospinal I-neurons (Ballantyne and Richter, 1984), then a delay in their onset of activity would allow bulbospinal

MEDULLARYINSPIRATORYNEURON RESPONSES TO LOADING

201

neurons to discharge for a longer time. This response also may be due to inhibitory information from mechanoreceptors stimulated by the TO (see later discussion). In a previous study in anesthetized cats (Shannon et al., 1972), no augmentation of medullary I-neuron activity was observed. It is not clear why I-neurons were observed to show an increase in activity with TO in this study and not in the previous anesthetized preparation. Clearly the reason was not depression of the reflexes by anesthesia. The I-neuron responses to loading were eliminated following cervical (C3-C7) plus thoracic (T l-T9) dorsal rhizotomies, but were still present with only cervical or thoracic dorsal roots intact. These results indicate that afferent information from mechanoreceptors with afferent fibers in these dorsal roots are responsible for the changes in medullary respiratory activity. These results are particularly interesting because the augmentation of I-neuron activity by afferents from either of the sources has not been previously reported. The specific mechanoreceptors involved in these reflexes can not be determined from this study; however, we can speculate on the possibilities. Since inspiratory (external and parasternal) intercostal muscle spindle endings and tendon organs increase their activity during TO (Critchlow and Euler, 1963; Sears, 1964; Corda et aL, 1965a), they might be considered as the most likely source(s) of the thoracic afferent information. The increase in fn-ing rate of tendon organs during TO is due to the increased muscle tension. Intercostal muscle tendon organs have been shown to have an inhibitory effect on medullary inspiratory activity (Bolser et aL, 1983) and could be responsible for the delay in onset of activity in some late I-neurons during TO. The increase in firing rate of muscle spindle endings during TO is due to misalignment between extrafusal muscle fibers and the intrafusal muscle fibers of the muscle spindles, which results from the decreased rate ofinspiratory intercostal muscle shortening during TO. Since inspiratory intercostal muscle spindle endings facilitate inspiratory intercostal alpha motoneurons via spinal pathways, one might conclude that the muscle spindle endings are responsible for the augmentation of medullary I-neuron activity via ascending pathways. However, there are no studies showing that these muscle spindle endings augment medullary I-neuron activity. In fact, it has been suggested that intercostal muscle spindle endings have an inhibitory effect on medullary I-neuron activity (Remmers, 1970; Remmers and Marttila, 1975), but no direct proof has been presented. Results from our laboratory (Bolser et al., 1984) strongly suggest that intercostal muscle spindle endings have no effect on medullary inspiratory neurons. Other possible sources of sensory activity responsible for the augmentation of Ineuron activity include other mechanosensitive receptors in inspiratory intercostal muscles, costovertebral joint mechanoreceptors and lung mechanoreceptors. There are force sensitive mechanoreceptors (e.g., free nerve endings and paciniform corpuscles) other than tendon organs in limb muscles (Cleland et al., 1982). These receptors in limb muscle, whose afferent fibers are primarily in the group III and IV range, may be involved in the neurogenic increase in ventilation accompanying exercise (Kalia et al., 1972; McCloskey and Mitchell, 1972). Whether these receptors in intercostal muscles influence brainstem respiratory activity has not been studied. Costovertebral joint

202

R. S H A N N O N et al.

mechanoreceptor afferent fibers travel in thoracic dorsal roots and they have been shown to be active during spontaneous breathing (Godwin-Austen, 1969) and have an inhibitory effect on medullary inspiratory drive during joint distortion and chest compression (Shannon, 1980), but whether they influence respiratory drive during eupnea is not known. If these joint receptors were responsible for the augmentation of I-neuron activity, it would have to be due to a decrease in their inhibitory effect (disinhibition) which could result from a decrease in the movement of the joints because of the reduced chest expansion with tracheal occlusion. Pulmonary mechanoreceptors with afferent fibers travelling in the thoracic sympathetic chain (Kostreva etal., 1975) could be stimulated by the large negative pressure change in intrathora¢ic pressure occurring with loading of inspiration. The cervical afferent information causing changes in I-neuron activity during loading may arise from mechanoreceptors in the diaphragm. Tendon organs in the diaphragm are considered to be responsible for the inhibition of phrenic alpha-motoneuron activity sometimes observed during tracheal occlusion (Corda et al., 1965a), and may be responsible for the delay in onset of activity of some medullary I-neurons. Whether diaphragm tendon organ afferent information ascends to medullary respiratory neurons is unknown. Diaphragm muscle spindle endings are not likely to contribute to the changes in I-neuron activity during loading. There are only a few (about nine) muscle spindles in the cat diaphragm (Corda etal., 1965b; Duron et al., 1978), and they produce no significant facilitation ofphrenic motoneuron activity during mechanical loading (Corda etaL, 1965a; Sant'Ambrogio and Widdicombe, 1965; Bradley, 1972). The augmentation of I-neuron activity during TO may have resulted in part from the stimulation of diaphragm force/sensitive mechanoreceptors other than tendon organs (e.g., free nerve endings, paciniform corpuscles). These receptors have not been studied in the diaphragm. Other mechanoreceptors with afferent fibers in cervical dorsal roots that may be involved in the I-neuron response to loading include receptors in the pleura and peritoneum (Duron and Marlot, 1980). Although the effects of these receptors on respiration has not been studied, they might be excited by changes in intrathoracic pressure accompanying tracheal occlusion. An obvious question is what is the net effect of the TO induced reflex changes in medullary I-neuron activity on inspiratory muscle motor activity? The most likely change we might predict from the results of this study would be a slight prolongation of motor activity in some animals. In the experiments on decerebrate vagotomized cats, there was a prolongation of activity in all I-neurons tested in only two of the nine cats. Thus, in most animals there was no significant change in descending drive from the medulla. We did not monitor diaphragm or intercostal motor activity in this study primarily because their motor activity is also affected by the segmental effects of their muscle spindle endings and tendon organs during TO, which would tend to mask the effects of altered descending inspiratory drive. Numerous previous studies (Corda et al., 1965a; Sant'Ambrogio and Widdicombe, 1965; Remmers, 1970; Bradley, 1972;

MEDULLARY INSPIRATORY NEURON RESPONSES TO LOADING

203

S h a n n o n and Z e c h m a n , 1972) have r e p o r t e d either no significant change or a decrease in phrenic m o t o r activity during mechanical loading in v a g o t o m i z e d cats. W h y is the T O i n d u c e d increase in activity o f D R G b u l b o s p i n a l I-neurons in some animals not reflected by an increase in phrenic activity? O n e possible explanation is that the spinal inhibition o f phrenic m o t o n e u r o n activity by d i a p h r a g m t e n d o n organs m a t c h e s or is greater than the increased descending drive from the medulla. The functional i m p o r t a n c e o f the extravagal reflexes d e m o n s t r a t e d in this study and the receptors involved are not clear. The conclusion from other studies using external mechanical loads is that extravagal reflexes, including the segmental a n d medullary effects o f inspiratory muscle p r o p r i o c e p t o r s , are not significant ' l o a d c o m p e n s a t o r y ' m e c h a n i s m s during the first-breath ( L y n n e - D a v i e s et al., 1971 ; S h a n n o n and Z e c h m a n , 1972) or during sustained loading (Younes e t a l . , 1973; Bruce et al., 1974) in anesthetized animals. The results o f this study are n o t necessarily inconsistent with this conclusion. These reflexes m a y not be i m p o r t a n t under the present experimental conditions, but they d o exist and could be i m p o r t a n t u n d e r other as yet undefined conditions.

Acknowledgements This study was s u p p o r t e d by a Public H e a l t h Service R e s e a r c h G r a n t ( H L 17715) from the N a t i o n a l Heart, Lung and B l o o d Institute.

References Ballantyne, D. and D.W. Richter (1984). Post-synaptic inhibition of bulbar inspiratory neurons in the cat. J. Physiol. (London) 348: 67-87. Bolser, D. C., B. G. Lindsey and R. Shannon (1983). Inhibition of medullary inspiratory drive by intercostal muscle tendon organs. Soc. Neurosci. Abst. 9: 1160. Bolser, D.C., B.G. Lindsey and R. Shannon (1984). Evidence that intercostal muscle spindle endings do not reflexly modulate medullary inspiratory drive. Fed. Proc. 43: 433. Bradley, G.W. (1972). The response of the respiratory system to elastic loading in cats. Respir. Physiol. 16: 142-160. Bruce, E. N., J. D. Smith and F. S. Grodins (1974). Chemical and reflex drives to breathing during resistance loading in cats. J. Appl. Physiol. 37: 176-182. Cleland, C.L, W.Z. Rymer and F.R. Edwards (1982). Force sensitive interneurons in the spinal cord of the cat. Science 217: 652-655. Corda, M., G. Eklund and C. von Euler (1965a). External intercostal and phrenic or-motor responses to changes in respiratory load. Acta Physiol. Scand. 63: 391-400. Corda, M., C. von Euler and G. Lennerstrand (1965b). Proprioceptive innervation of the diaphragm. J. Physiol. (London) 178: 161-177. Critchlow, V. and C. yon Euler (1963). Intercostal muscle spindle activity and its y-motor control. J. Physiol. (London) 168: 820-847. Duron, B., M. C. Jung-Caillol and D. Marlot (1978). Myelinated nerve fiber supply and muscle spindles in the respiratory muscles of cat: quantitative study. Anat. Embryol. 152: 171-192. Duron, B. and D. Marlot (1980). The non-myelinated fibers of the phrenic and the intercostal nerves in the cat. Z. Mikrosk. Anat. Forsch. 94: 257-268.

204

R. SHANNON et al.

Euler, C. von (1973). The role of proprioceptive afferents in the control of respiratory muscles. Acta Neurobiol. Exp. 33: 329-341. Euler, C. von (1983). On the central pattern generator for the basic breathing rhythmicity. J. Appl. PhysioL 55: 1647-1659. Godwin-Austen, R. B. (1969). The mechanoreceptors of the costo-vertebral joints. J. PhysioL (London) 202: 737-753. Kalia, M., J.M. S enapti, B. Parida and A. Panda (1972). Reflex increase in ventilation by muscle receptors with nonmedullated fibers (C fibers). J. Appl. Physiol. 32: 189-193. Kostreva, D.R., E.J. Zuperku, G.L. Hess and J.P. Kampine (1975). Pulmonary afferent activity recorded from sympathetic nerves. J. Appl. Physiol. 31: 37-40. Lynne-Davies, P., J. Couture, L. D. Pengelly, D. West, P. R. Bromage and J. Milic-Emili (1971). Partitioning of immediate ventilatory stability to added elastic loads in cats. J. Appl. Physiol. 30: 814-819. McCloskey, D.I. and J.H. Mitchell (1972). Reflex cardiovascular and respiratory responses originating in exercising muscle. J. Physiol. (London) 224: 173-176. Remmers, J.E. (1970). Inhibition of inspiratory activity by intercostal muscle afferents. Respir. Physiol. 10: 358-383. Remmers, J. E. and I. Marttila (1975). Action of intercostal muscle afferents on the respiratory rhythm of anesthetized cats. Respir. Physiol. 24: 31-41. S ant'Ambrogio, G. and J. G. Widdicombe (1965). Respiratory reflexes acting on the diaphragm and inspiratory intercostal muscles of the rabbit. J. Physiol. (London) 180: 766-779. Sears, T. A. (1964). Some properties and reflex connexions of respiratory motoneurones of the cat's thoracic spinal cord. J. Physiol. (London) 175: 386-403. Shannon, R. and F.W. Zechman (1972). The reflex and mechanical response of the inspiratory muscles to an increased airflow resistance. Respir. Physiol. 16:51-69. Shannon, R., F.W. Zechman and D.T. Frazier (1972). First-breath response of medullary inspiratory neurons to the mechanical loading of inspiration. Respir. Physiol. 16: 70-78. Shannon, R. (1980). Respiratory pattern changes during costovertebral joint movement. J. Appl. Physiol. 48: 862-867. Shannon, R. and D. Freeman ( 1981). Nucleus retroambigualis respiratory neurons: responses to intercostal and abdominal muscle afferents. Respir. Physiol. 45: 357-375. Younes, M., W. Arkinstall and J. Milic-Emili (1973). Mechanisms of rapid ventilatory compensation to added elastic loads in cats. J. Appl. Physiol. 35: 443-453.