INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

PREVIOUS STUDIES have demonstrated that negative pressure applied to the upper airway of animals and humans causes reflex activation of a number of upper airway dilator muscles, including the alae nasi (22), genioglossus (10, 15, 16, 22), cricothyroid (14), posterior cricoarytenoid (14, 22), sternohyoid and sternothyroid (14), and soft palate muscles (21). It has been proposed that upper airway muscle recruitment by negative pressure protects the upper airway against collapse from negative intraluminal pressures (14, 22). Although the precise afferent pathways for these reflexes have not been determined, pressure-sensitive mechanoreceptors in the upper airway have been demonstrated to respond to negative airway pressure (11, 12, 15, 16, 19). Furthermore, many of these receptors are superficial (located in the mucosal and submucosal tissues), because they are affected by local anesthesia (9, 14). However, a role for more deeply located pressure receptors (i.e., within muscle tissue) is not precluded. Negative upper airway pressure (NUAP) may activate muscle mechanoreceptors by stretching and deforming muscle tissue. Hence, an alternate group of nonsuperficial upper airway receptors could participate in the detection of NUAP. In addition, these muscle mechanoreceptors may be important in the reflex response of individual upper airway muscles to NUAP. This mechanism would permit upper airway muscles to respond differentially to the same negative pressure, with the stimulated activity partially dependent on the level of local mechanoreceptor activation. However, there is no information concerning the afferent nerve traffic arising from such mechanoreceptors in upper airway muscles when they are stimulated by NUAP.

By recording from the hypoglossal nerve (motor-sensory nerve to the tongue muscles), we examined the role of NUAP in stimulation of muscle receptors from the tongue. The tongue muscle was selected for study on the basis of its recognized importance in the control and maintenance of upper airway patency, because of the relative accessibility of the hypoglossal nerve, and because the hypoglossal nerve does not carry sensory information from mucosal surfaces in the upper airway.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Studies were performed on six anesthetized, paralyzed (gallamine trithiodide, 4 mg/kg iv) and mechanically ventilated supine adult cats [weight 3.0 ± 0.8 (±SD) kg]. After animals were premedicated with acetylpromazine (0.2 mg/kg im), general anesthesia was induced by placing the animal in a box filled with a mixture of halothane and 100% oxygen delivered by a ventilatory pump. After anesthesia was induced, the halothane circuit was switched to a face mask so that anesthesia could be maintained while femoral venous and arterial catheters were inserted for the systemic administration of saline, anesthetics, and paralyzing agents and for the measurement of arterial blood pressure. Thiopentone sodium (10-20 mg/kg iv) was then injected to allow insertion of an endotracheal tube, which was used for the subsequent continued administration of halothane. Rectal temperature was monitored and maintained at 37.5-38.5°C by using an infrared lamp. End-tidal CO₂ was monitored with an infrared gas analyzer (Morgan 901) and was maintained at 4-5% by adjusting ventilation. Arterial blood pressure (±200 cmH₂O pressure transducer; Celesco) and heart rate were continuously monitored as indexes of the depth of anesthesia. The investigation was approved by the University of Sydney Animal Care and Ethics Committee.

A tracheostomy was performed, and the caudal tracheal end was connected to a halothane gas-anesthetic circuit and ventilator, while the laryngeal end was connected to a pressure transducer (±50 cmH₂O; Celesco) to record upper airway pressure (Fig. 1). The esophagus was ligated, and the upper airway was sealed by placing a specially constructed plastic mask (made airtight with petroleum jelly) over the cat's nose and mouth. Negative pressure was delivered to the sealed upper airway via a perforated catheter placed through the face mask to enter the mouth. The other end of the catheter was connected to a large volume (23 liters) pressure reservoir via a solenoid valve. A suction pump was used to generate negative pressure within the reservoir; the level of negative pressure was monitored by using a differential pressure transducer (±50 cmH₂O; Celesco). Activation of the solenoid valve delivered a step change of negative pressure to the isolated and sealed upper airway. The system caused a step change in upper airway pressure, with a response time of 10 ms, to achieve 90% equilibration of pressure. A nylon ligature was passed through the face mask via an airtight rubber seal and was sutured to the tip of the tongue. The distal end of the ligature was then connected to a displacement transducer (model 305B, Cambridge Technology, Watertown, MA) to permit the measurement of tongue stretching.

View larger version (19K): [in this window] [in a new window]   Fig. 1.   Schematic diagram of isolated and sealed upper airway in supine cat. Animals were mechanically ventilated via the tracheostomy. Negative pressure was applied to upper airway via a perforated tube placed in the mouth. Tongue was stretched by using a displacement transducer.

Hypoglossal nerve isolation. The hypoglossal nerve was dissected free from where it crossed the carotid artery, and it was sectioned just caudal to where it branches to the tongue muscles. The distal end of the medial division of the hypoglossal nerve was placed on a small platform in a warm (37°C) paraffin pool. The length of nerve was then desheathed by using watchmaker's forceps, with the aid of a dissecting microscope, and small strands of the nerve were dissected free. Strands were placed on a pair of platinum electrodes connected to an alternating current-coupled amplifier (JRAK Biosignals, Sydney, Australia). The signal was filtered (1 Hz to 10 kHz), amplified, and displayed on an oscilloscope with audio output from a loudspeaker. Progressively smaller strands were teased from the nerve until a filament containing one active fiber was obtained.

Experimental protocol. Single fibers were selected for study only if they were sensitive to manual stretching of the tongue via its attached ligature. Tongue tip displacements of known magnitude (between 1 and 3 cm) were then made (beginning from the resting position of the tongue) by using the displacement transducer attached to the tongue ligature. After this, the same mechanoreceptive fiber was tested for its sensitivity to step changes in negative pressure, ranging from 5 to 30 cmH₂O, that were applied to the upper airway. The duration of tongue displacement and negative pressure varied from 2 to 10 s. At least three runs were performed for each fiber at each displacement and negative pressure level. In one cat, the study was repeated for two fibers after topical anesthesia (10% lignocaine) was applied liberally to the entire upper airway mucosa. This application consisted of the upper airway's being completely filled with 10 ml of lignocaine for at least 10 min.

Data analysis. The ENG and other signals recorded on the pulse-code modulated system were analyzed off-line after the study. The frequency of ENG discharge was measured at increasing tongue displacements, ranging from 1 to 3 cm, and at different levels of negative pressure, ranging from 5 to 30 cmH₂O, by using a custom-written computer program (ASYST 3.0, Clinical Engineering Solutions) which detected interspike intervals. Values of firing frequency were obtained during steady-state fiber discharge. For each fiber, duplicate runs were analyzed for each displacement and negative pressure level.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In six cats, 12 afferent nerve fibers were selected for study on the basis of a positive ENG response to passive stretch (PS) of the tongue. Seven of the 12 fibers were tonically active (Fig. 2) and the remaining five were silent in the absence of stimulation (Fig. 3). In response to a constant stretch of the tongue, there was an initial burst of activity which then decreased to a steady rate of discharge above the control level (Fig. 2). The initial burst of activity occurred more frequently at the larger displacements. During steady-state PS of between 1 and 3 cm, firing frequency increased in the group as a whole from 17 ± 7 to 40 ± 11 Hz (Fig. 4; P < 0.01). Afferent activity was transiently inhibited after the removal of PS but then returned to control levels (Fig. 2).

View larger version (35K): [in this window] [in a new window]   Fig. 2.   Example of afferent activity of a single hypoglossal nerve fiber recorded in response to a step increase in tongue displacement. Top: raw electroneurographic (ENG) activity from the fiber. Middle: instantaneous firing frequency of the fiber. Bottom: tongue displacement. Note tonic discharge of single fiber before stimulus. In response to a constant displacement of the tongue, there was an initial burst of activity which then decreased to a steady rate of discharge above the control level. After reversal of the displacement, afferent activity was briefly inhibited but then returned to control levels.

View larger version (23K): [in this window] [in a new window]   Fig. 3.   Example of afferent activity of a single fiber recorded during a step change in negative upper airway pressure (NUAP). Top: raw ENG activity from the fiber. Middle: instantaneous firing frequency of the fiber. Bottom: NUAP. Note that this fiber was silent before negative pressure. In response to NUAP, there was an initial burst of activity which then decreased to a steady rate of discharge above the control value for the duration of the negative pressure. In addition, note that gradual change in negative pressure after removal of the pressure source was associated with a progressive decrease in firing frequency.

View larger version (10K): [in this window] [in a new window]   Fig. 4.   Mean firing frequency of 12 single fibers before (control) and during tongue displacement (1-3 cm) (left), and mean firing frequency of 8 single fibers before (control) and during NUAP of 10 to 30 cmH2O (right). Values of firing frequency were obtained during steady-state fiber discharge. Bars, SE. * P < 0.01. Note increase in afferent discharge in response to both displacement of the tongue and NUAP.

Step changes in negative pressure applied to the upper airway led to an immediate increase in ENG activity (Fig. 3). In addition, at the higher NUAPs, there was an initial burst of activity which then decreased to a steady rate of discharge above the control value for the duration of the negative pressure. Of the 12 fibers that responded to PS, 8 responded to NUAP between 10 and 30 cmH₂O, with firing frequency increasing from 23 ± 9 to 41 ± 9 Hz (Fig. 4; P < 0.001).

For two fibers tested at multiple pressure levels and degrees of tongue displacement in the same cat, there was a positive linear relationship between the increase in frequency of fiber discharge (above control) and both the degree of tongue stretch (Fig. 5), and the level of NUAP (Fig. 6).

View larger version (11K): [in this window] [in a new window]   Fig. 5.   Plots of %increase in firing frequency (above control) as function of tongue displacement for 2 fibers. Each point represents a single value. Two displacement runs were performed. Linear regression lines and correlation coefficients (R) are shown. Note that in both fibers there was a positive linear relationship between the increase in frequency of fiber discharge and the degree of tongue displacement.

View larger version (12K): [in this window] [in a new window]   Fig. 6.   Plots of %increase in firing frequency (above control) as function of NUAP for same 2 fibers as in Fig. 5. Linear regression lines and R values are shown. Each point represents a single value. Two NUAP runs were performed. Note that in both fibers there was a positive linear relationship between the increase in frequency of fiber discharge and the level of NUAP.

In the two fibers tested after local anesthesia was administered, the increase in firing frequency in response to 14 cmH₂O NUAP was not altered by topical anesthesia of the upper airway mucosa. In three to five trials of NUAP, the increase in firing frequency of these two fibers did not change: from 24.6 ± 0 and 13.0 ± 0.5 Hz before local anesthesia, to 20.7 ± 3.0 and 15.4 ± 0.9 Hz after local anesthesia, respectively (Fig. 7).

View larger version (13K): [in this window] [in a new window]   Fig. 7.   Mean increase in firing frequency of 2 fibers during NUAP of 14 cmH2O before and after appliciation of topical anesthesia to entire upper airway (UA) mucosa. Bars, SD. Note that UA anesthesia did not alter the increase in afferent discharge elicited by NUAP.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The principal findings of this study are that in anesthetized, supine, tracheostomized cats afferent discharges from the hypoglossal nerve are elicited by 1) mechanical displacement of the tongue and 2) NUAP. In addition, topical upper airway anesthesia does not abolish the afferent discharges elicited by NUAP.

A number of upper airway muscles in both animals and humans are recruited in response to NUAP within the upper airway lumen (10, 16, 22). The genioglossus and hypoglossal nerve efferent activity have been extensively studied, and negative pressure recruitment of genioglossus muscle activity is predominantly mediated via stimulation of pressure-sensitive upper airway mucosal mechanoreceptors (11, 12, 15, 16). This is supported by studies in which topical anesthesia of the upper airway mucosa can abolish or diminish the genioglossus response to NUAP in both animals and humans (9, 15, 16). However, in animals, sectioning the internal branches of the superior laryngeal nerve markedly reduced but did not eliminate the response to NUAP. This suggests that alternate afferent pathways are involved in this reflex response. The glossopharyngeal and trigeminal nerves have been suggested as possible afferent pathways for reflex recruitment of upper airway muscles in response to NUAP (9, 11, 12, 14), because the nerves are involved in detecting mucosal sensation in the nose, nasopharynx, and oropharynx. In addition to mucosal mechanoreceptors, there are several other possible upper airway receptors. These include slowly and rapidly adapting receptors of the upper airway, proprioceptive nerve endings of the tongue, laryngeal muscle and joint receptors, and other muscle mechanoreceptors.

Critique of method. There were a number of limitations to the methodology employed in the current study. First, passive distension of a fiber and characteristics of the receptive field of the fibers could not be tested. Second, the degree of movement associated with application of the face mask and negative pressure was such that it was not always possible to keep recording from a single fiber by using a teased fiber preparation. Consequently, data sets from individual fibers were frequently incomplete. Third, the experimental setup was unable to measure the degree of tongue displacement induced by negative pressure. Fourth, we were unable to determine whether the afferent activity from the hypoglossal nerve was derived from the intrinsic or extrinsic musculature of the tongue. However, muscle spindles have been found in both extrinsic and intrinsic tongue muscles (2, 5, 7). Moreover, either anterior or posterior tongue stretching is likely to lengthen the extrinsic (i.e., genioglossus) muscles because of the anatomic direction of the fibers. In addition, both intrinsic and extrinsic muscles are likely to be stretched or deformed by negative pressure due to the complex orientation of the fibers within the tongue, although this has not been tested. In support of this observation, Brennick et al. (4) have recently demonstrated that NUAP does stretch the genioglossus muscle. Last, it was unlikely that the afferent signals recorded from the hypoglossal nerve were artifacts from the solenoid valve. Both Figs. 2 and 3 show the typical initial bursts of activity and gradual fall in firing frequency with the onset and cessation of the pressure stimulus that are characteristic of muscle spindles.