Bonora, M., and M. Vizek. Role of vagal fibers in the hypoxia-induced increases in end-expiratory lung volume and diaphragmatic activity. J. Appl. Physiol. 83(3): 700–706, 1997.—The possible role of pulmonary C fibers in the hypoxia-induced concomitant increases in end-expiratory lung volume (EELV) and in the activity of the diaphragm at the end of expiration (De) were evaluated by measuring the effects of hypoxia (10% O2) on ventilation, EELV, and De in eight chloralose-urethan anesthetized rats. Recordings were made before and after blocking vagal C fibers and after bilateral vagotomy. C-fiber conduction was blocked by applying capsaicin perineurally to the cervical vagi. The efficiency of C-fiber blockade was tested with intravenous capsaicin and its selectivity by the Hering-Breuer reflex. Perineural capsaicin abolished the reflex apnea induced by intravenous capsaicin and transiently reduced Hering-Breuer reflex. Perineural capsaicin affected neither ventilation, De, and EELV in air nor the hypoxia-induced increases in these parameters. Vagotomy caused the typical changes of breathing pattern in air, but the ventilatory response to hypoxia was unchanged. Vagotomy performed during hypoxia resulted in large decreases in De and EELV. Hypoxia increased De and EELV in vagotomized rats but less than in intact rats. We conclude that the hypoxia-induced increases in EELV and diaphragmatic activity are probably not mediated by vagal C fibers and that vagal afferents are involved but not fully responsible for this phenomenon.
- vagal C fibers
- Hering-Breuer reflex
hypocapnic hypoxia has been found to prolong the activity of the diaphragm during expiration (3-5, 30, 31). As this activity may persist until the end of expiration, it can enlarge the end-expiratory lung volume (EELV) by preventing the thorax from collapsing to its relaxed position. Hypoxia-induced increases in EELV have been reported in several studies (2, 6, 12, 32), and a clear relationship between concomitant changes in end-expiratory diaphragmatic activity (De) and EELV has been recently shown in anesthetized cats (5).
The mechanism responsible for this phenomenon is not clear, although several studies suggest that vagal afferents are involved. Vizek et al. (32) showed that vagotomy prevented EELV from increasing during hypoxia. In other respects, the pulmonary neuroendocrine cells, which are able to transduce an hypoxic stimulus via an oxygen-sensing mechanism similar to that of carotid bodies (34), are in close contact with sensory unmyelinated vagal fibers (20). Also, electrical or pharmacological stimulation of pulmonary unmyelinated C fibers induces an increase in the postinspiratory inspiratory activity of the diaphragm (15, 16), similar to that previously observed during hypoxia (5). Moreover, the ventilatory response to hypoxia is markedly reduced in rats with degenerated unmyelinated fibers (9). Hence, we speculated that the hypoxia-induced increase in diaphragmatic activity during expiration and the subsequent enlargement of EELV may be mediated through vagal nerve afferents, in particular, unmyelinated C fibers.
The goal of the present study was, therefore, to examine whether the selective block of vagal C fibers will affect the changes in ventilation, diaphragmatic activity, and EELV induced by hypoxia. The C-fiber block was obtained by topical application of capsaicin to both cervical vagi, a technique reported to block the respiratory reflexes resulting from stimulation of C-fiber nerve endings (15, 17,18, 21, 28) while leaving myelinated fibers intact (15, 28, 33). We, finally, compared the results with those obtained after bilateral vagotomy.
Studies were performed on eight adult male Wistar rats [mean body weight 323 ± 5 (SE) g]. They were anesthetized with an initial intraperitoneal injection of chloralose (100 mg/kg) and urethan (500 mg/kg). Supplemental doses of anesthetic were administered when necessary. The stability of anesthesia was judged by immobility or lack of response to tactile stimuli and by the regularity of the respiratory pattern.
The abdomen was opened with a midline incision, and two electrodes (multistrand Teflon-coated wires) were implanted in the rostral part of the right hemi-diaphragm. A third electrode (ground) was tied into the neck muscles. All the wires were tunneled subcutaneously and exteriorized at the back of the neck.
A polyethylene catheter (ID 0.6, OD 1 mm) was inserted into the right jugular vein with the tip close to the right atrium for the intravenous injection of capsaicin. The neck was opened in the midline, and a length of ∼3 mm of each cervical vagus was carefully separated from the carotid artery and cleared of surrounding connective tissue. The nerves were prevented from drying with cottonwool soaked in saline. Finally, a short cannula (ID 1.7, OD 2.3 mm) was inserted into the trachea just below the larynx.
Measurements. Minute ventilation (V˙e) was measured with a body plethysmograph. The tracheal tube was connected to an external circuit. Changes in plethysmograph pressure representing tidal volume (Vt) were detected by a differential pressure transducer (Validyne MP45). Tracheal pressure was measured with a Statham transducer (Gould P23). The electromyographic (EMG) acticity of the diaphragm was amplified (Disa 15C01; input impedance 250 MΩ, noise level 0.7 μV), filtered (30–1,000 Hz), and integrated by a resistance-capacitance circuit with a time constant of 100 ms (moving time average). Raw and integrated diaphragmatic signals were displayed on an oscilloscope and recorded on a polygraph together with the respiratory and pressure signals. Inspired oxygen concentration was measured by using a Beckman OM11 analyser.
EELV was computed by the manometric method of Dubois et al. (10). The tracheal tube was occluded at the end of expiration. EELV was calculated by using Boyle’s law from the changes in tracheal pressure and Vt induced by three consecutive efforts.
A digitizer connected to a computer was used to measure Vt, inspiratory (Ti) and expiratory (Te) durations, respiratory frequency (f),V˙e, and instantaneous values of integrated diaphragmatic EMG at the peak activity, which corresponds to the end of neural inspiration, and at the trough, which corresponds to the end of expiration (De).
The small unmyelinated vagal C fibers were blocked with perineural application of capsaicin. A stock solution (0.3 mg/ml) was prepared by dissolving capsaicin (Sigma Chemical) in saline, ethanol, and Tween 80 and stored at 4°C (8). The exposed parts of the cervical vagi were surrounded with a cotton strip soaked in freshly prepared capsaicin solution (30 μg/ml) for 30 min.
The efficiency and the specificity of the vagal C-fiber block was tested by two procedures.
1) The C fibers were stimulated by injecting 0.2 ml capsaicin (1 μg/kg) into the jugular venous catheter. Ventilation and diaphragmatic activity were recorded and measured 5 min before the injection (control), immediately preceding and following the injection, and again 5 min after the injection (recovery). EELV was measured during the control and recovery periods. This test was performed before and immediately after perineural capsaicin and later on, after the hypoxic test, to check the persistence of the vagal block throughout the experiment.
2) The myelinated A fibers were activated by inflating the lungs to 10 cmH2O pressure at the end of inspiration or by occluding the airways at the end of inspiration; the duration of the characteristic apnea was measured. These tests were performed before, during (at 10, 20, and 30 min), and after (at 20 min) perineural capsaicin treatment and, finally, after bilateral vagotomy.
Protocol. The rat was placed prone in the plethysmograph, and the colon temperature was continuously monitored. V˙e, diaphragmatic activity, and EELV were measured during air breathing [0.21 inspired O2 fraction ( )] 5 and 10 min after its breathing had became stable. was then abruptly switched to 0.10, and the recordings were repeated 5 and 10 min after the beginning of hypoxia. Finally, was returned to 0.21, and recordings were made 5 and 10 min later. A similar protocol was repeated after perineural capsaicin treatment of the vagus nerves and after bilateral vagotomy.
Data analysis and statistics. Each variable of ventilation and diaphragmatic activity was averaged over 10 consecutive respiratory cycles. For reflex apnea induced by occlusions and inflations, duration of the occluded breath (TTo) was compared with the control breath (TTc) immediately preceding the maneuver.
Results are presented as means ± SE. Statistical analysis was done by using nonparametric tests, since they do not assume a normal distribution. The Wilcoxon test and Friedman two-way analysis of variance were used to compare the data with their own control values. Differences were considered significant whenP < 0.05.
The efficiency and specificity of the vagal C-fiber block. 1) Before perineural capsaicin, the intravenous injection of capsaicin, which activates the vagal afferent C fibers, induced a characteristic reflex apnea followed by an increase in De (Fig.1). The mean values (±SE) of Te increased from 0.71 ± 0.04 to 2.16 ± 0.3 s, and Deincreased transiently (+49.6 ± 19.1%), returning to baseline 5 min after the injection.
After perineural capsaicin, intravenous injection of capsaicin did not induce the typical reflex apnea (Fig. 1), and the increase in De was smaller than before capsaicin treatment (+18.7 ± 3.2%). When the capsaicin test was repeated after exposure to hypoxia, the apneic response was still absent, and the increase in Dewas of the same magnitude as before hypoxia (+21.3 ± 4.0%).
2) To verify that A fibers supplying pulmonary stretch receptors were not blocked by perineural application of capsaicin, the Hering-Breuer reflex was tested by airway occlusion at the end of inspiration or inflation of the lungs.
After occlusion of the airways, there was an apneic pause of variable duration in all rats. The TTo/TTcratios measured before, during, and after removal of the perineural capsaicin are shown in Fig.2 A. The control TTo/TTcwas 2.1 ± 0.43; TTo/TTcdecreased transiently during treatment because of a shortening of the occluded breath, but at the end of the treatment and 20 min later, TTo/TTcwas not significantly different from control. Bilateral vagotomy abolished the Hering-Breuer reflex.
Inflation of the lungs also caused a characteristic apnea. The mean control TTo/TTcratio was 6.4 ± 1.1; 10 and 20 min after the beginning of perineural capsaicin, the apnea durations were significantly shortened, but they gradually recovered so that at the end of the treatment and later on TTo/TTcwas not significantly different from control (Fig.2 B). After bilateral vagotomy, inflation induced no apnea.
These results indicate that perineural application of capsaicin blocked the response to activation of vagal C fibers, whereas the function of the A fibers was only transiently affected.
Effects of perineural capsaicin treatment. The ventilatory parameters (V˙e, Vt, and f) during air breathing and in response to hypoxia were not significantly affected by perineural capsaicin (Table1).
Under control conditions, hypoxia induced a significant increase in De and EELV (Fig.3). Capsaicin treatment did not change significantly the mean values of these two parameters, in air and in response to hypoxia. Therefore, blockage of the C fibers did not affect the responses ofV˙e, De, and EELV to hypoxia.
Effects of bilateral vagotomy. The following results were obtained from six rats (two rats died shortly after vagotomy) during the measurement of EELV in hypoxia. TheV˙e of vagotomized rats breathing air did not significantly differ from that before vagotomy (Table 1); however, Vt increased markedly, whereas f decreased because of a large increase in Ti (from 0.39 ± 0.01 to 0.50 ± 0.03 s) and Te (from 0.64 ± 0.08 to 0.95 ± 0.11 s). Hypoxia increasedV˙e, Vt, and f; their relative changes from air control values were similar to those of intact animals.
Bilateral vagotomy performed during hypoxia resulted in large decreases in De and EELV (Fig.4). In fact, vagotomy almost completely removed the increases in De and EELV initially induced by hypoxia. During normoxia, the De and EELV values were not significantly different from those of intact animals. In vagotomized rats, the response to hypoxia showed significant increases in both De and EELV, although they were slightly smaller than before vagotomy in four out of six animals (Fig.3).
This study was designed to determine the role of vagal afferents, and in particular vagal C fibers, in hypoxia-induced changes in De and EELV. Our hypothesis that vagal C fibers may mediate this increase was based on studies showing an increase in postinspiratory diaphragmatic activity caused by electrical or pharmacological activation of C fibers (15, 16) and on our recent observation of a clear relationship between the changes in De and EELV induced by different degrees of hypoxia in cats. The results of the present study show that the increases in De and EELV induced by hypoxia also occur in rats. They also show that blocking of the vagal C fibers by the perineural application of capsaicin does not alter the hypoxia-induced changes in De or EELV. Bilateral vagotomy performed under hypoxia substantially reduces both De and EELV. The transition from normoxia to hypoxia caused these two parameters to increase in vagotomized animals, although the increases were less pronounced than in intact animals. These results are consistent with the involvement of vagal afferents in the hypoxia-induced increases in De and EELV but suggest that vagal fibers other than C fibers play a determinant role in this phenomenon.
Methodology. Ventilation, EELV, and respiratory muscle activity all depend on the type of anesthetic agent used (35) and the depth of anesthesia (19). We used chloralose-urethan anesthesia, because volatile anesthetics and barbiturates have been shown to affect markedly the activity of respiratory muscles (11, 35). The depth of anesthesia was relatively stable in this study, as indicated by relatively stable baselineV˙e throughout the experiment.
The position of the rats could also influence the control values of EELV (19), but these values were comparable to those reported by Vizek et al. (32) and Barer et al. (2) for supine rats. Finally, the position of the electrodes in the costal part of the diaphragm could have affected the baseline diaphragmatic activity (5, 31), but neither the posture of the animal nor the position of the electrodes changed during the experiment and, thus, could not have affected the relative changes induced by hypoxia.
The efficiency of the perineural capsaicin treatment of the vagi was demonstrated by the effect of intravenous capsaicin on ventilation. The dose injected was probably appropriate, since it evoked a typical pulmonary chemoreflex apnea of the same duration as reported by others in rats (23, 27). This apnea was completely abolished after perineural capsaicin treatment of the vagi (15, 17, 18, 21, 28) while the conduction in the large myelinated fibers was not disrupted (28).
Effects of perineural capsaicin treatment. Capsaicin applied for 30 min should induce a stable and long-lasting block of C fibers (1, 28). It remained effective throughout the hypoxic test performed after the treatment, as shown by the absence of response to repeated intravenous injections of capsaicin.
Perineural capsaicin produced no change in baselineV˙e, in agreement with the reports of Kou et al. (18) and Lee et al. (21) in rats. However, changes of pattern of breathing have been reported in other species with a higher dose of capsaicin (15, 28). The ventilatory response of our rats to hypoxia remained unchanged after perineural capsaicin, in contrast to studies showing a significantly reduced ventilatory response to hypoxia in rats treated with capsaicin neonatally (9). This difference might be due to the fact that only vagal C fibers were blocked in our experiments, whereas the neonatal treatment disrupts all afferent C fibers, including those from peripheral chemoreceptors.
The hypoxia-induced increase in De is in agreement with our previous findings (4, 5) and with the report of prolonged diaphragmatic postinspiratory inspiratory activity in dogs (31). The enlargement of EELV during acute hypoxia has also been reported for both experimental animals and humans (2, 6, 12, 32), although it is not a consistent finding in humans (13). The mechanism of the hypoxia-induced enlargement of EELV is not clear. Possible causes include shortening of expiratory duration or a decrease in expiratory airflow resulting from a decrease in the elastic recoil of the system, an increase in airway resistance, or a change in respiratory muscle activity during expiration. In a previous study (5), we have shown that in anesthetized cats the increase in EELV during hypocapnic hypoxia resulted from a decrease of expiratory airflow, mainly caused by the persistent activity of the diaphragm up to the end of expiration, De, although a decrease of expiratory muscle activity may also have played a role. However, the present study was designed to test whether hypoxia-induced increases in De and EELV are mediated through vagal C fibers.
Baseline values of De and EELV remained unchanged after capsaicin treatment, and the increases in De and EELV in response to hypoxia were the same. Thus vagal C fibers were probably not responsible for this increase, but this assumption is valid only if perineural capsaicin completely blocks C fibers. Topical application of capsaicin has been widely used to selectively block C fibers (1, 15,17, 28, 33). The concentration of capsaicin we used was lower than in some studies (18, 21), but it was reported to be effective for rat nerves (1) and it did block the apneic pause induced by intravenous injection of capsaicin. However, there may be a capsaicin-resistant population of small nonmyelinated vagal fibers, as reported in some studies (1, 28), although by using a similar dose of capsaicin, Hatridge et al. (15) found the complete disappearence of C-fiber compound action potentials after the treatment. In addition to the concentration, the efficiency of the block may depend on the diameter of the nerve, thickness of the sheath, or tissue perfusion. Nevertheless, the resistant fibers to capsaicin are likely efferent preganglionic ones (28), which should not be important for the mechanism we were studying. Moreover, a higher concentration may have drastically affected the conduction of large myelinated A fibers.
Indeed, the selectivity of this block may be questioned, since some authors (1, 15) have found that high doses of perineural capsaicin diminished the A-fiber conduction. Also, the Hering-Breuer reflex mediated by large myelinated A fibers (supplying slow-adapting receptors) was in the present study, as in others (15, 18), slightly weaker after the capsaicin treatment. However, the hypoxia-induced increase in EELV should not be affected by a partial block of A fibers in our experiments, since Barer et al. (2) showed that this effect is still observed after complete block of Hering-Breuer reflex.
The specificity of perineural capsaicin treatment can also be questioned because of its possible effect on small myelinated fibers (17), which includes the fibers supplying rapidly adapting receptors (RARs). Because augmented breaths initiated by RARs were still observed after capsaicin treatment (Ref. 18 and present paper), we believe that the RARs were not blocked. Thus the perineural capsaicin treatment probably blocked the decisive portion of vagal C fibers and slightly affected the supplying slow-adapting receptors but not the RAR afferentation.
Mechanisms. The vagal-dependent mechanism of hypoxia-induced increases in De and EELV could involve the pulmonary RARs. There is good evidence that activation of RARs is a powerful stimulus for the increased De induced by continuous negative airway pressure and histamine infusion (22). However, there is no direct data indicating that hypoxia also stimulates the RARs. Nevertheless, indirect stimulation could occur via the increased ventilation during hypoxia, since RARs respond to hyperpnea (29). RARs are also stimulated by bronchoconstriction (36), which has been described as part of the reflex response to hypoxia (24). The bronchoconstriction induced by hypoxia could stimulate RARs and initiate an increase in De via a central pathway. The consequent increase in EELV would reduce the resistance of small intrapulmonary airways and so counteract the effect of the bronchoconstriction on pulmonary resistance. The hypocapnia due to hyperventilation may also play an important role in this mechanism. Hypocapnia may stimulate RARs directly (7), and it has been shown that hypocapnia induces an increase in flow resistance (25) and bronchoconstriction (26). The contribution of hypocapnia fits well with our previous study showing that the large increases in De and EELV that occur during severe hypoxia are significantly reduced when normocapnia is restored (5).
However, if the mechanism underlying the increases in De and EELV depends on the activation of vagal small myelinated fibers, then the increases should be completely abolished by bilateral vagotomy. Indeed, vagotomy has blocked the increase in EELV during hypoxia in dogs and rats (14, 32), but in the present study such increase was diminished but not blocked. This discrepancy in the results may be partly explained by different anesthesia and position of the animal. Nevertheless, a vagal-independent mechanism should be considered. The results of Barer et al. (2), who found that the hypoxia-induced increase in EELV was completely abolished only by combined vagal and carotid sinus nerve blockade, are in favor of an involvement of carotid body chemoreceptors. Their contribution is also indicated by the fact that the carotid body denervation suppresses the hypoxia-induced increase in EELV in dogs (6). It should be noted that the suppression of the hypoxia-induced increase in EELV after carotid body denervation may result from the suppression of the overall ventilatory response to hypoxia. However, the changes in ventilation are not always accompanied by changes in EELV, as it occurs during hypercapnia and hyperoxia (5,6, 12).
In conclusion, we showed that the increases in De and EELV induced by hypoxia depend partly on vagal nerve activity but are probably not mediated by C fibers. We suggest that this phenomenon may involve small myelinated vagal fibers supplying RARs and a vagal-independent mechanism.
The authors thank H. Gautier for his support and helpful criticism and J. Chandellier for preparing the illustrations.
Address for reprint requests: M. Bonora, Laboratoire de Physiologie Respiratoire, Faculté de Médecine St-Antoine, 27 rue de Chaligny, 75012 Paris, France (E-mail:).
M. Vizek was the recipient of a fellowship as part of joint program between l’Université Pierre et Marie Curie, Paris, and Charles University, Prague. This work was also supported by French Embassy in Prague and by Grant 027/96 from the Grant Agency of the Charles University.
- Copyright © 1997 the American Physiological Society