INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

RECORDINGS from the superior laryngeal nerve (SLN) have demonstrated that laryngeal receptors are responsive to a wide range of physical and chemical stimuli (2, 14, 15, 22-25, 28). Boushey et al. (6) showed that laryngeal mechanoreceptors are sensitive to gas mixtures containing 5 and 10% CO₂ blown over the exposed laryngeal mucosa of anesthetized cats. This finding has been confirmed by the addition of CO₂ to a constant rostral flow through the isolated larynx in anesthetized dogs (1) and anesthetized or decerebrate cats (3). We have previously shown that the application of 5 and 9% CO₂ in 21% O₂ and balance N₂ during artificial ventilation of the isolated larynx of anesthetized cats can significantly alter the discharge of SLN afferents (10, 11). In the cat, it has also been shown that application of CO₂ continuously to the upper airway (UA) reflexly inhibits breathing (4, 5, 18) and excites UA muscle (18) and motor nerve (4) activity through a SLN-dependent reflex.

In all of the above studies, the laryngeal mucosa was exposed to a continuous level of CO₂. However, during normal breathing, the laryngeal lumen is exposed to CO₂ phasically rather than continuously because CO₂ passes over the larynx mainly during expiration. Bartlett and Knuth (3) showed that alternating through the larynx unidirectional flow between room air and mixtures containing CO₂ can alter the activity of laryngeal afferents. However, although the briefest duration of CO₂ application used in their study was comparable with the duration of expiration in eupnea, the receptors involved were not adequately identified and responses were studied in the absence of the cyclical changes in pressure, flow, temperature, and humidity to which the receptors are normally subject.

The purpose of the present investigation was to study the responses of laryngeal receptors to CO₂ applied during simulated expirations and to compare them with responses to continuously applied CO₂, i.e., applied during inspiration and expiration. To do this, we used an artificially ventilated laryngeal preparation in anesthetized, paralyzed cats in which the receptors were exposed to cyclical variations in pressure, flow, temperature, and humidity resembling those in the larynx during a normal respiratory cycle. Additionally, we used a servo respirator to artificially ventilate the larynx in spontaneously breathing cats to study the effects of CO₂ in the absence of muscle paralysis. This was done because it has been shown that paralysis can greatly affect laryngeal mechanoreceptor discharge (24). The servo respirator was used because it was necessary to match the artificial ventilation of the larynx with that of spontaneous breathing. Some of these results have been reported in preliminary form (9, 12).

	METHODS

Top Abstract Introduction Methods Results Discussion References

Experiments were carried out in 13 adult cats of either sex. Two preparations were used. In the first preparation, eight animals were anesthetized with pentobarbital sodium (Sagatal, May & Baker, Dagenham, Essex, UK; 40-48 mg/kg ip initially, maintenance 6-12 mg iv as required), paralyzed with pancuronium bromide (Sigma Chemical; 0.8 mg iv as required), and artificially ventilated through a low-cervical tracheostomy. In the second preparation, five animals were anesthetized with 1.5 ml/kg Saffan im (alphaxalone 0.9% wt/vol and alphadalone acetate 0.3% wt/vol) and maintained with pentobarbital sodium (20 mg/kg iv initially and 6-12 mg iv as required) and allowed to breath room air spontaneously through a low-cervical tracheostomy. Saffan was given to reduce the amount of pentobarbital sodium used because the latter has been shown to greatly affect laryngeal muscle activity (26). In the paralyzed animals, the depth of anesthesia was assessed by continuously monitoring blood pressure and heart rate.

A femoral artery was cannulated to allow measurement of arterial blood pressure (Statham P23, Hato Rey, PR) and periodic sampling of arterial blood for analysis of blood PO₂, PCO₂, and pH (Ciba Corning Diagnostics, Halstead, Essex, UK). A femoral vein was cannulated for the administration of supplementary anesthesia and other drugs. Atropine sulfate (0.3 mg/kg; Antigen, Roscrea, Tipperary, Ireland) was administered intravenously to reduce airway secretions. Body temperature was maintained at 37-38°C by using a rectal probe and a thermostatically controlled heating blanket (Harvard Apparatus, Edenbridge, Kent, UK). Tracheal airflow was recorded by using a pneumotachograph (Fleisch 00) and differential pressure transducer (model PT 5, Grass Instruments, Quincy, MA) attached to the tracheostomy cannula. End-tidal CO₂ was continuously monitored by using an infrared CO₂ analyzer (Engstrom Eliza Duo, Gambro, Sweden).

UA preparation. The larynx and trachea were exposed by means of a ventral midline incision. A cannula was inserted into the trachea and directed rostrally to the level of the cricoid cartilage (cricoid cannula). A second cannula was inserted through the mouth to the tip of the epiglottis, and the mouth and nose were sealed with dental cement. A cannula attached to a pressure transducer (model PT 5, Grass Instruments), and a thermistor probe (model 402, Yellow Springs Instruments, Yellow Springs, OH) were inserted through the oral cannula to the tip of the epiglottis to measure UA pressure and temperature, respectively. The UA CO₂ concentration was measured from a sidearm of the cricoid cannula by using an infrared CO₂ analyzer (Engstrom Eliza).

Nerve recording. The SLN were cut close to the nodose ganglion and dissected back toward the larynx. The peripheral ends were desheathed, and single- and few-fiber preparations were recorded by using bipolar tungsten electrodes. Activity was amplified (type P16, Grass Instruments) and monitored by using an audio monitor (type AM7, Grass Instruments) and oscilloscope (type 5013, Tektronix Guernsey, Guernsey, Channel Islands). Nerve discharge; blood pressure; UA temperature, pressure, airflow, and CO₂ concentration; pulmonary airflow; and end-tidal CO₂ concentration were recorded on a Gould electrostatic recorder (type ES100, Gould Electronics, Cleveland, OH) and/or an ink-writing oscillograph (type 7WC 12PA, Grass Instruments).

Protocol. Nerve fibers were classified as quiescent or tonic. When UA ventilation was stopped, quiescent fiber activity was irregular and infrequent (<3 impulses/s) and tonic fiber activity was continuous and steady. We also observed fibers that continued to have a respiratory rhythm when UA ventilation was stopped in spontaneously breathing animals. However, the number of fibers examined was small and the effects of CO₂ applied in the expiratory phase were not examined. Therefore, the effects of CO₂ on this category of fiber are not reported. Quiescent and tonic fibers were classified as positive- or negative-pressure receptors on the basis of their responses to pulses of positive and negative pressure. Quiescent fibers with no response to pressure stimuli were identified as cold receptors if their activity was increased by drawing cold air continuously through the larynx in a caudal direction (although a cranial direction could also have been used). Finally, there were a group of fibers that were quiescent, had no respiratory-related activity during UA ventilation, and did not respond to pressure or cold stimuli. These were termed fibers of unknown modality.

Data analysis. Single-unit activity was quantified as the number of impulses per UA ventilatory cycle or its inspiratory or expiratory components before, during, and after CO₂ trials when CO₂ concentration and nerve discharge had reached a steady state. Activity during the trial period was also quantified as percentage of activity before the trial. Differences in mean discharge before and during trials were tested for significance by using Student's t-test, with P < 0.05 taken as significant. In a small number of trials, there was a significant difference between activity in the period before the application of CO₂ and activity in the period after the CO₂ had been removed; i.e., there was no recovery to pretrial values after the CO₂ was removed. For each category of fiber, the overall mean discharge ± SE before and during CO₂ application and the overall mean percent change from control ± SE were calculated for those fibers for which a significant effect was observed.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

In all, 66 receptors were studied, of which 21 were identified as quiescent negative-pressure receptors, 16 as quiescent positive-pressure receptors, 6 as tonic negative-pressure receptors, 9 as tonic positive-pressure receptors, 9 as cold receptors, and 5 as receptors of unknown modality. Responses to CO₂ were quantitatively and qualitatively similar for the two preparations so that data from both experiments were pooled. In both preparations, switching from UA ventilation with room air to ventilation with CO₂-containing gas mixtures did not alter UA pressure, temperature and airflow (see Figs. 2 and 3), or systemic arterial blood pressure, PO₂, PCO₂, and pH. Because of difficulties associated with performing single-fiber recordings in spontaneously breathing animals, the protocol for testing responses to both continuous and "expired" 5 and 9% CO₂ was not completed for all fibers.

Quiescent negative-pressure receptors. Seventeen of twenty-one fibers in this category were excited by both 5 and 9% CO₂ when applied either continuously or during simulated expiration only (Fig. 1A). The remain- ing four fibers were unaffected by CO₂. For the analysis of individual fibers, the responses to 9% CO₂ were significantly greater than the responses to 5% CO₂ for both continuous and expired CO₂ in 7 of the 17 fibers. Although the overall values for 9 vs. 5% CO₂ were also greater, the differences did not reach statistical significance for either continuous or expired CO₂. Similarly, there were no significant differences between the overall values for continuous vs. expired CO₂ for either 5 or 9% CO₂. An example of the excitatory effect of 5 and 9% CO₂ applied during the expiratory phase is shown in Fig. 2.

View larger version (46K): [in this window] [in a new window]   Fig. 1.   Laryngeal receptor responses to 5 and 9% CO2 applied continuously and during expiration. Responses of laryngeal receptors to CO2 applied continuously (crosshatched bars) and in expiration (hatched bars) are expressed as mean percent changes from air control (± SE) for those fibers for which significant effects were observed. A: quiescent negative-pressure receptors (n = 17). B: cold receptors (n = 9). C: tonic negative-pressure receptors (n = 3). D: tonic positive-pressure receptors (n = 4). E: quiescent positive-pressure receptors (n = 11).

View larger version (44K): [in this window] [in a new window]   Fig. 2.   Effect of 5 and 9% CO2 applied during expiration on activity of a laryngeal quiescent negative-pressure receptor in a paralyzed, artificially ventilated cat. A: nerve discharge increases when 5% CO2 is applied during expiration. B: when CO2 concentration is increased to 9%, there is a further increase in nerve discharge, followed by recovery when CO2 is removed. Records A and B are continuous. PUA, upper airway pressure; UA, upper airway airflow; FUA CO2, upper airway fractional concentration of CO2; TUA, upper airway temperature.

Quiescent positive-pressure receptors. Continuous and expired CO₂ inhibited 11 of the 16 fibers in this category (Fig. 1E). Of the remaining five, four were unaffected and one was excited. There were no significant differences between 5 vs. 9% CO₂ values or between continuous vs. expired CO₂ values. An example of the inhibitory effect of 5% CO₂ applied during the expiratory phase is shown in Fig. 3.

View larger version (29K): [in this window] [in a new window]   Fig. 3.   Effect of 5% CO2 applied during expiration on activity of a laryngeal quiescent positive-pressure receptor in a paralyzed, artificially ventilated cat. A: nerve discharge is inhibited when 5% CO2 is applied during expiration. B: nerve discharge recovers when CO2 is removed. Records A and B are continuous.

Tonic negative- and positive-pressure receptors. Three of the six tonic negative-pressure receptors were inhibited by both continuous and expired CO₂ (Fig. 1C), and the remainder were unaffected. There were no significant differences between responses to 5 and 9% CO₂ values or between continuous vs. expired CO₂ values.

Cold receptors. All nine receptors studied were excited by continuously applied 5 and 9% CO₂ (Fig. 1B). The 9% CO₂ response was larger, but the overall difference did not reach statistical significance. Analysis of individual fibers showed significantly larger effects for 9 vs. 5% CO₂ in five of the nine fibers, with no significant differences in the remaining four fibers. Responses to expired CO₂ were tested in five of the nine fibers, all five being excited by both 5 and 9% CO₂. There were no significant differences between responses to 5 and 9% CO₂ values or between continuous vs. expired CO₂ values.

Receptors of unknown modality. The effects of 5 and 9% CO₂ applied continuously and during simulated expiration were inconsistent.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

It is well established that laryngeal receptors of different modalities are sensitive to changes in UA CO₂ concentrations. The new finding in this investigation is that these receptors respond to changes not only in UA CO₂ applied continuously during inspiration and expiration but also when CO₂ is applied only during the expiratory phase of artificial ventilation of the UA. We used 5% CO₂ to approximate normal end-tidal concentrations and 9% CO₂ to test hypercapnic responses. Normal end-tidal CO₂ levels are generally <5% in anesthetized cats, but we also observed, in a small number of fibers tested in the present experiments (not described), that 3% CO₂ applied during the expiratory phase also greatly affected receptor activity. Furthermore, Bartlett and Knuth (3) showed that pulses of 3% CO₂ alternating with air affected laryngeal-receptor discharge in cats.

Our results with CO₂ applied continuously during inspiration and expiration were broadly consistent with our previous findings (8, 11). Responses to CO₂ applied during the expiratory phase of artificial ventilation of the UA were very similar.

The present results confirm that quiescent negative- and positive-pressure receptors are excited and inhibited, respectively, by CO₂. In the present study, 81% of quiescent negative-pressure receptors and 69% of quiescent positive-pressure receptors were affected compared with 81 and 50%, respectively, in a previous study (11). Responses to CO₂ applied during expiration only were qualitatively similar, and, although there was a tendency toward larger responses for CO₂ applied continuously, this did not reach statistical significance. Bartlett and Knuth (3) reported predominantly inhibitory effects of CO₂ on SLN afferents in cats by using a preparation that allowed alteration of a continuous flow through the larynx between warmed humidified room air and CO₂-containing mixtures. This effect was obtained at durations of CO₂ application similar to those in the present work, but, when the CO₂ was applied continuously, i.e., for longer than 10 s, the effect was greater. We did not observe larger effects with continuous compared with phasically applied CO₂, but direct comparison is not feasible because the identity of the fibers described by Bartlett and Knuth (3) was not adequately established and, unlike the present experiments, the receptors were not simultaneously subjected to the mechanical and thermal stimuli similar to those of a normal respiratory cycle. In the Bartlett and Knuth study, the effects of phasic CO₂ on fibers that were excited by continuous CO₂ were not examined. The fact that we did not observe greater effects with continuous compared with phasic CO₂ may suggest that the CO₂ concentration in the local environment of the receptor is similar under both conditions. This might be possible if the rate of CO₂ diffusion from the receptor environment to the UA lumen during the inspiratory phase was slow relative to the duration of the inspiratory phase.

As in our previous study (11), tonic receptors that responded to either negative or positive pressure were inhibited by continuous CO₂. These receptors were also inhibited to a comparable extent by CO₂ applied during the expiratory phase of artificial ventilation of the UA. Bartlett and Knuth (3) have also shown that tonically active fibers are inhibited by phasic CO₂.

We have previously shown that cold receptors are strongly excited by CO₂ (11), although Anderson et al. (1) observed no effect of 10% CO₂ applied continuously to the UA in dogs. In the latter study, CO₂ was applied when the receptors were already being stimulated by cold air, which may have masked any stimulatory effect by CO₂. We initially identified cold receptors by applying cold air, but then we applied CO₂ during artificial ventilation of the UA when receptors were no longer exposed to cold air. This is because the room temperature gas mixtures that were drawn over the larynx during simulated inspiration had warmed to 35°C or more at the larynx so that the temperature was probably above the threshold for cold-receptor activation (25). In the present experiments, all of the cold receptors studied were stimulated by continuous CO₂ and were equally excited by CO₂ applied during the expiratory phase of artificial ventilation of the UA.

The use of the servo respirator allowed examination of the effects of CO₂ on receptors while the larynx was subjected to laryngeal motion and laryngeal muscle contraction in synchrony with spontaneous breathing. Synchronization of UA artificial ventilation with spontaneous breathing was necessary so that laryngeal motion and laryngeal muscle contraction occurred at the same time as the mechanical, thermal, and chemical changes produced by the UA artificial ventilation. Laryngeal paralysis has been reported to markedly affect laryngeal-receptor activity in anesthetized dogs (24). However, we were not able to directly assess the effects of laryngeal paralysis on baseline receptor activity because receptor activity was never recorded before and after paralysis in the same experiment. We found that the responses of laryngeal receptors to CO₂ were the same in paralyzed and nonparalyzed animals.

A small number of receptors were recorded, the modality of which we were unable to determine. These may be a heterogeneous group, but it is likely that some at least are receptors responsive to mechanical probing of the laryngeal mucosa because it has been shown that "nonmodulated" mechanoreceptors are activated by such stimulation (1). In our previous work (11) and that of Anderson et al. in the dog (1), afferents of this type have been shown to be excited by CO₂. In the present study, responses to CO₂ were more variable and less pronounced.

It has been shown that intralaryngeal CO₂ evokes important reflex effects on breathing and UA muscle activity that are SLN mediated (4, 5, 18). Topical anesthesia of the airway mucosa alters the ventilatory pattern in response to inspired CO₂ (29) as does laryngeal denervation (7); UA topical anesthesia also alters ventilatory pattern of air-breathing dogs (13). These results suggest that fluctuations in UA CO₂ affect UA-receptor activity. The present results show that the fluctuations in UA CO₂ during breathing significantly affect laryngeal-receptor responses to other respiration-related stimuli. Bartlett et al. (4) have shown that phasic intralaryngeal CO₂ causes weaker reflex effects on phrenic nerve activity and has minimal effects on hypoglossal nerve activity compared with continuously applied CO₂. This was consistent with their finding that phasic CO₂ produced quantitatively lesser effects on SLN activity. On the other hand, the addition of CO₂ to the airflow only during the expiratory phase increases tidal volume in conscious ponies (21), whereas addition of CO₂ during inspiration does not (27).

Our results suggest that alterations in the CO₂ concentration of expired gas could produce substantial reflex effects because there was little difference between SLN afferent responses to continuously applied and phasically applied CO₂. The reflex effects of CO₂ applied continuously to the UA include an inhibition of breathing (4, 5, 18), an excitation of UA dilator muscle and motor nerve activity (4, 18), increased laryngeal resistance (20), and excitation of laryngeal adductor muscle activity (4). UA negative pressure also reflexly excites UA dilator muscle activity (17), and we have previously proposed that the excitatory effect of UA CO₂ on laryngeal negative-pressure receptors would act together with negative pressure during UA occlusion to excite UA muscle activity and alleviate the occlusion (18). This proposal has been supported by the finding that UA CO₂ enhances laryngeal negative-pressure activity during UA occlusion (16). The fact that the activity of negative-pressure receptors is also dependent on the levels of CO₂ that occur in the UA during inspiration and expiration, as shown in the present experiments, is further evidence for a role of UA CO₂ in regulating UA patency.