CHAPTER 1
The Upper airway motor system, Obstructive sleep apnea/hypopnea syndrome and Pulmonary vagal afferents
Introduction
The upper airway (UAW) of the mammals is the airflow passage extending from the external nares, pharynx to the larynx (Bailey and Fregosi, 2006; Bartlett, 1986;
Pierce and Worsnop, 1999). It is surrounded by several skeletal muscles, which are primarily innervated by the cranial nerve, including the facial, hypoglossal, superior laryngeal and recurrent laryngeal nerve (Miller, 2002; Pierce and Worsnop, 1999).
The UAW muscles or motor nerves display respiratory-related activity, and are involved in respiratory and non-respiratory functions. Activity of the UAW motor outputs can regulate the airflow passing through the UAW and thus may relate to the modulation of the UAW patency. Clinically, dysfunction of respiratory-related activity of the UAW motor outputs has been demonstrated to be involved in obstructive sleep apnea/hypopnea syndrome (OSAHS), which may further induce some cardiopulmonary diseases. Therefore, understanding neural control mechanism of the UAW motor outputs is necessary and important. In this regard, pulmonary vagal afferents play an important role in the regulation of breathing patterns. However, influence of pulmonary vagal afferents on the UAW motor outputs has not been fully understood in the rat. This dissertation included a series of experiments in attempt to investigate that pulmonary vagal afferents may produce a reflex control of the UAW motor outputs. An overview of the UAW motor system, obstructive sleep apnea/hypopnea and pulmonary vagal afferents were provided in the chapter 1.
The upper airway motor system
Neural control of the alae nasi
The nasal muscles
The nasal cavity is the first entrance of inspiratory airflow during breathing and consists of various skeletal muscles with a very complicated arrangement (Table 1-1, Bruintjes et al., 1998). Studies on the nasal muscles primarily focus on the alae nasi muscles. Nostril flaring evoked by voluntary contraction of the alae nasi muscles could reduce the resistance of the nasal airway in human subjects (Strohl et al., 1982;
Shi et al., 1998). This is mainly due to dilatation of the nasal airway by shortening of the nasal muscle length (van Lunteren et al., 1985). Haight and Cole (1983) documented that paralysis of the facial nerve by local anesthetics evoked a collapse of the alae. In addition, clinical study has shown that activity of the alae nasi muscles was reduced during phasic rapid-eye-movement (REM) sleep and thus induced an occlusion of the nasal airway leading to hypoventilation (Wiegand et al., 1991).
Therefore, activity of the alae nasi may be critical for maintaining a nasal airway patency.
Respiratory-related activity of the facial nerve and alae nasi muscles
The facial nerve and alae nasi exhibit respiratory-related activity. Onimaru et al., (2006) observed that the facial nerve display pre-inspiratory (Pre-I), inspiratory (I) and post-inspiratory (Post-I) activity in the newborn rat. A detail study by Hwang and St. John (1988) demonstrated that the facial nerve of the cat was composed of various types of respiratory-related motoneurons. Respiratory-related rhythmic activity of the facial nerve was synchronized with rhythm of the phrenic nerve (Hwang et al., 1988;
Jacquin et al., 1999). Jacquin et al. (1999) indicated that rhythmic facial activity may
be generated in the brainstem and may be mostly from the ventrolateral medulla (VLM), an area that presumed to have respiratory rhythm generators (preBötzinger complex and para facial respiratory group; pFRG) (Feldman and Del Negro, 2006;
Onimaru and Homma, 2003; Smith et al., 1991). In addition, pons may also play an important role in modulation of rhythmic facial nerve activity (Jacquin et al., 1999;
Onimaru et al., 2006; Tanabe et al., 2005).
Respiratory-related activity of the facial nerve, facial motoneurons or alae nasi muscles could be modulated by the peripheral and central stimuli. In this regard, hypoxia and hypercapnia elicited excitation of the alae nasi muscles/facial nerve (Martin et al., 1990; Patrick et al., 1982; Strohl, 1985). Intralaryngeal CO2 evoked an enhancement of the facial nerve activity with a concomitant depression of the phrenic burst (Bartlett et al., 1992). Negative pressure in the UAW induced a reflex increase in the UAW muscle activity (alae nasi and posterior cricoartenoid muscles) and inhibition on the diaphragm (Mathew, 1984; van Lunteren et al., 1984). Static and phasic lung inflation produced inhibition on inspiratory activity of the alae nasi musclesand facial nerve, suggesting that both tonic and phasic discharge of pulmonary stretch receptors may contribute to modulate activity of the alae nasi muscles and facial nerve (Hwang et al., 1988; Hwang and St. John, 1988; St. John and Zhou, 1992; van Lunteren et al., 1984). Electrical and chemical stimulation delivered to the rostral ventrolateral medulla (RVLM) and nucleus of the tractus solitarius (NTS) could also modulate respiratory-related activity of the facial nerve (Lin and Hwang, 1994; Wu and Hwang, 1997). Interestingly, the facial activity was diminished during bladder contraction (Gdovin et al., 1994), indicating the nasal airway patency is not only influenced by respiratory-related stimuli but also respond to some non-respiratory afferents inputs.
Neural control of the tongue muscles
The tongue muscles and hypoglossal nerve
The tongue is involved in many physiology functions, including feeding, swallowing, licking, chewing, speech and respiration (Lowe, 1981; Miller, 2002;
Sawczuk and Mosier, 2001). The tongue muscles are composed of the extrinsic and intrinsic muscles. The extrinsic tongue muscles, which originated from an external bone and insert into the tongue base, include the genioglossus (GG), styloglossus (SG) and hyoglossus (HG) muscles (Fig. 1-1A). The GG originates from the mandible and inserts into the ventromedial base of the tongue. The GG could be further divided into the horizontal and oblique compartments according to different anatomical specialization, including arrangement and composition of the muscle fibers, and innervation pattern (Mu and Sanders, 2000). The SG attaches to the mastoid bone via the styloid process and inserts into the ventrolateral base of the tongue. The HG enters into the superiorlateral base of the tongue from its origin on the hyoid bone. The intrinsic tongue muscles are the tongue itself and include the transverse, vertical, superior and inferior longitudinal muscles. The vertical and transverse muscles are mainly located in the central of the tongue. The longitudinal muscle fibers run along the peripheral of the tongue and are encircled by the transverse muscles in the rostral part of the tongue (McClung and Goldberg, 2000a,b ; Mu and Sanders, 1999).
The extrinsic and intrinsic tongue muscles are all innervated by the hypoglossal nerve, which is composed of the medial and lateral hypoglossal branch (MHN and LHN, respectively) (Fig. 1-1B). The MHN innervates the extrinsic GG, intrinsic vertical and transverse muscles and the geniohyoid (GH) muscle. The LHN innervates the extrinsic SG, HG and the intrinsic longitudinal muscles (McClung and Goldberg, 2000a, b).
The extrinsic and intrinsic tongue muscles work cooperatively to produce
retraction and protrusion of the tongue. Protrusion of the tongue is accomplished by contraction of the extrinsic GG and intrinsic vertical and transverse muscles.
Contraction of the extrinsic SG and HG, and intrinsic superior and inferior longitudinal muscles can retract the tongue. Vertical movements of the tongue are mainly controlled by the extrinsic tongue muscles, of which activation of the GG and HG muscles depress the tongue, while contraction of the SG serves to elevation the tongue (Gestreau et al., 2005; McClung and Goldberg, 2000). Shape of the tongue and movements of the tongue tip are entirely controlled by the intrinsic muscles (Sawczuk and Mosier, 2001).
The tongue muscles are involved in various oral and pharyngeal reflexes.
Contractions of the tongue muscles are also correlated with the pharyngeal patency.
Electrical stimulation of the LHN induced retraction of the tongue, whereas stimulation of the MHN produced protrusion of the tongue (Gilliam and Goldberg, 1995). Retractive force elicited by LHN stimulation was stronger than protrusive force induced by stimulation of the MHN. Hence, stimulation of the whole hypoglossal nerve caused retraction of the tongue instead of protrusion (Fuller et al., 1998; Gilliam and Goldberg, 1995). Clinical study has also revealed that stimulation of the whole hypoglossal nerve retract the tongue in the human subject (Eisele et al., 1997). The movement of the tongue certainly influences the airway airflow mechanics through regulation of the UAW patency. In this regard, stimulation of the whole hypoglossal nerve increased the nasal airflow rate and decreased pharyngeal airway collapsibility by stiffening the tongue. Selective stimulation of the MHN could promote the oral airflow by dilating the oropharynx and decreasing oral resistance, while independent stimulation of the LHN would rather reduce the airflow rate (Bailey and Fregosi, 2003; Fuller et al., 1999). In addition, stimulation of the MHN could relieve pharyngeal airway obstruction induced by negative UAW pressure in the
canine (Goding et al., 1998). According to magnetic resonance imaging (MRI) study of the pharyngeal airway, the nasopharyngeal and oropharyngeal volume could be enlarged by stimulation of the whole hypoglossal nerve or MHN due to the depression of the tongue (Brennick et al., 2001). These studies suggested that activation of the whole hypoglossal nerve or its MHN may have benefit to the UAW patency and could be applied to clinical therapy for OSAHS.
The hypoglossal nucleus
The hypoglossal nucleus is located in the dorsal medulla and exhibits two major subdivisions. Medial and lateral branches of the hypoglossal nerve are originated from the ventral and dorsal subdivisions of the hypoglossal nucleus, respectively (Dobbins and Feldman, 1995). Motoneurons that innervate the horizontal compartment of GG are located in the lateral part of the ventral hypoglossal nucleus, while motoneurons that supply to the oblique compartment of GG are situated at the centrolateral part of the ventral hypoglossal nucleus (Fig. 1-2, Gestreau et al., 2005; McClung and Goldberg, 2002). The caudolateral and rostrolateral part of the dorsal hypoglossal nucleus project axons to the HG and SG, respectively. Motoneurons that innervate the superior and inferior longitudinal muscles are clustered around central and medial part of the dorsal hypoglossal nucleus (Fig. 1-2, Gestreau et al., 2005; Guo et al., 1996;
McClung and Goldberg, 1999).
Motoneurons from the MHN and LHN receive convergent inputs from the raphe nucleus, Kölliker-Fuse (KF) nucleus, nucleus subcoeruleus and gigantocellular reticular formation in the medulla and also receive divergent projections from the lateral tegmental field. Premotor neurons of the MHN in the lateral tegmental field are located more ventral and ventrolateral than which project to the LHN. The motoneurons of the MHN receive projection from the ipislateral parabrachial nuclei
and retrotrapezoid nuclei. However, the motoneurons of the LHN receive projections from the ventral and medial portion of the nucleus tractus solitarius (NTS), facial nuclei and spinal trigeminal nucleus (Dobbins and Feldman, 1995). These convergent and divergent projections to the hypoglossal motoneurons may coordinate the tongue muscles and constitute complex behavior of the tongue movement (Dobbins and Feldman, 1995; Travers and Rinaman, 2002).
In addition to anatomy heterogeneity of the hypoglossal nuclei, the hypoglossal motoneurons were classified into distinct groups according to changes of membrane potential and discharge pattern during fictive breathing, coughing and swallowing (Roda et al., 2002). Interestingly, some hypoglossal motoneurons could be recruited during above-mentioned two or three fictive behaviors. These results indicate that neural networks of respiration, coughing and swallowing generators may collectively control some portions of the hypoglossal motoneurons.
Respiratory-related activity of the tongue muscles and the hypoglossal nerve
The tongue muscles and the hypoglossal nerve display obvious respiratory-related activity (Fregosi and Fuller, 1997, Hwang et al., 1983a; Haxhiu et al., 1988; Pillar et al., 2001). Inspiratory activity of the tongue muscles could prevent the tongue from being sucked into the pharynx by the negative transpulmonary pressure during inspiration. Harper and Sauerland (1978) have observed that decrease in inspiratory activity of the tongue muscles and/or the hypoglossal nerve might induce the UAW collapsibility and/or obstruction, and thus increase risk of being suffered from OSAHS. Hence, inspiratory activity of the tongue muscles and hypoglossal nerve are critical for the UAW patency.
Rhythmic hypoglossal activity display in various experimental conditions, such
as awake, decerebrated, anesthetized, in situ and in vitro (Fuller et al., 1998; Hwang et al., 1983; Leiter and St-John, 2004; Smith et al., 1991). In the brain slice preparation, rhythmic activity of the hypoglossal root may be transmitted via the pauci-synaptic pathway from the pre-Bötzinger complex, which is considered as the respiratory rhythm generator (Feldman and Del Negro, 2006; Smith et al., 1991). This rhythmic hypoglossal activity coincides with the phrenic nerve bursting in in vivo and in situ.
Ono et al., (1994) has reported that some hypoglossal premotor neurons also project to both the phrenic and hypoglossal nucleus in the cat. However, cross-correlation analysis indicated that hypoglossal motoneurons could not be excited by the premotor neuron of the phrenic nerve in the VRG in the rat (Peever et al., 2001). Furthermore, cessation of hypoglossal activity during expiratory duration is resulted from disfacilitation of excitatory inputs to the hypoglossal nucleus but not inhibitory effects of the Bötzinger complex (Peever et al., 2002). These results suggested that neural control of respiratory activity of the hypoglossal nerve might be different from that of the phrenic nerve. Inspiratory activity of the hypoglossal nerve is proposed to be driven by the inspiratory premotor neurons in the hypoglossal nucleus and the lateral tegmental field, which including dorsal respiratory groups, neurons located dorsomedial to ventral respiratory group, and ventrolateral and lateral to hypoglossal nucleus (Peever et al., 2002).
Various signals coming from the peripheral afferents can influence activity of the hypoglossal nerve and the tongue muscles. Phasic and static lung volume feedback induced decrease in activity of the GG and hypoglossal nerve (Bartlett and St. John, 1988; St. John and Zhou, 1992; van Lunteren et al., 1984), which was abolished by bilateral vagotomy. These results indicated that the vagus nerve transmitted signal from lung inflation and then restrain neural drives to the tongue muscles. UAW negative pressure augmented hypoglossal activity but decreased phrenic discahrge via
activation of the superior laryngeal nerve, glossopharyngeal nerve and trigeminal nerve (Hwang et al., 1984; Hwang and Young, 1989; Mathew, 1984; van Lunteren et al., 1984). Stimulation of the central and peripheral chemoreceptors by hypercapnia and hypoxia could enhance hypoglossal activity (Bruce et al., 1982; Hwang et al., 1983; Kuna, 1987; Weiner et al., 1982). Rhythmic activitiy of the hypoglossal nerve decreased and even disappeared after hyperventilation-induced hypocapnia (Fukada and Honda, 1983; Hwang et al., 1988), suggesting that changes in CO2 concentration may provide an important driving force to evoke respiratory rhythm of the hypoglossal nerve.
Sleep-wake state also affects activity of the GG and hypoglossal nerve.
Inspiratory and tonic GG activity was suppressed during non-REM and REM sleep in the freely moving rat. However, GG activity was augmented by hypercapnia only during awake and non-REM sleep but still reduced during REM sleep (Horner et al., 2002). It is documented that decreased GG activity in non-REM sleep may be due to activation of glycine and GABA (γ-aminobutyric acid) transmission (Morrison et al., 2003a and 2003b). Furthermore, recent study provided evidences that decline of endogenous drive of noradrenaline to the hypoglossal nucleus may contribute to reduction in GG activity during REM sleep (Chan et al., 2006).
As mentioned above, the hypoglossal nerve and tongue muscles display obvious inspiratory activity. Onset of inspiratory activity of the hypoglossal nerve precedes the phrenic nerve bursting (Hwang et al., 1983; Saito et al., 2002; van Lunteren et al., 1984). This temporal difference in onset of the hypoglossal nerve and the phrenic nerve may result from lower threshold of the hypoglossal premotor neurons or hypoglossal motoneurons, and/or stronger excitatory drives to these neurons (Sica et al., 1984). Another possibility for preceding onset of the hypoglossal activity is that differential inhibitory influences on the phrenic and hypoglossal motoneurons during
expiratory duration (Fukada and Honda, 1998). This has been demonstrated by Peever’s study (2001), which reported that expiratory neurons in the Bötzinger complex produce inhibition impinging on the phrenic motoneurons but not on the hypoglossal motoneurons.
These preceding or Pre-I activity of the hypoglossal nerve was increased by hypercapnia, hypoxia and bilateral vagotomy, and inhibited by lung inflation (Bartlett and St. John, 1988; Fukuda and Honda, 1982; Sica et al., 1984). Saito et al., (2002) observed that Pre-I activity of the hypoglossal nerve could be augmented and transformed into the uncoupled activity (persisrent rhythmic discharge during expiration) with increasing positive end-expired pressure (PEEP). Similarly, this Pre-I discharge of the hypoglossal nerve was uncoupled from the phrenic discharge during hypocapnia or hypothermia in the perfused rat preparation (St. John et al., 2004).
These observations were compatible with the notion that peak frequency the hypoglossal nerve during Pre-I duration was lower than those during inspiratory duration. These results strongly suggested that controlling mechanisms of Pre-I activity and of inspiratory activity of the hypoglossal nerve might be different from each other (Leiter and St.-John, 2004).
Co-activation of the tongue muscles
In most of the previous studies, functions of the tongue muscles in terms of the UAW patency were primarily focused on the protrusion of the GG. However, Fuller et al. (1998) has demonstrated that inspiratory activity of the tongue retractor (SG and HG) and protrudor (GG) muscles are co-activated by the trachea occlusion and stimulation of chemoreceptors. Other studies also revealed that lung inflation and negative upper airway pressure induced similar responses of neural drives to the tongue protrudor and retractor muscles (Bailey et al., 2001; Ryan et al., 2001).
Hypoxia-induced augmented breath has also evoked a long-term reduction in inspiratory motor outputs to the both tongue protrudor and retractor tongue muscles (Janssen et al., 2000). In addition, episodic hypoxia evoked a long-term facilitation in both MHN and LHN (Fuller, 2005). These results indicated that the tongue protrudor and retractor muscles might display a similar discharge pattern and responses to various stimuli. This co-activation of the tongue protrudor and retractor muscles may stiff the tongue and prevent collapsibility of the UAW. Thus, the hypoglossal motoneurons supplied to the tongue protrucdor and retractor muscles may have received the same premotor neuron pools (Peever et al., 2001).
Role of the intrinsic tongue muscles during respiration has been focused recently.
Bailey and her colleagues reported that inspiratory-related activity of the intrinsic tongue muscles could be activated during hypoxia and hypercapnia (Bailey and Fregosi, 2004; Bailey et al., 2005). Furthermore, patency of the velopharyngeal airway was reduced after denervating the intrinsic tongue muscles during atmospheric and positive airway pressure (Bailey et al., 2006). These data provided a furthermore evidence that co-activation of the extrinsic and intrinsic tongue protrudor and retractor tongue muscles may play a critical role in the maintenance of the UAW patency.
Excitability of the hypoglossal motoneurons
There are various types of receptors observed on the hypoglossal motoneurons.
Activation of these receptors may have affected excitability of the hypoglossal motoneurons. In fact, rhythmic activity of the hypoglossal motoneurons are primarily mediated by activation of non-N-methy-D-asparate (non-NMDA) receptors in vitro.
In addition, NMDA receptors, nicotine receptors and gap junction also constituted excitatory drives to the hypoglossal motoneurons during inspiration (Bou-flores and Berger, 2001; Wang et al., 2002). Perfusion of non-NMDA and NMDA receptor
agonists into the hypoglossal nucleus evoked tonic activity of the GG. Data from the in vivo study has been demonstrated that antagonists of both NMDA and non-NMDA receptors suppressed inspiratory activity of the GG (Steenland et al., 2006). These results suggest that respiratory-related rhythmic hypoglossal activity is mainly driven by glutamatergic inputs from the respiratory rhythm generator. In addition, activation of NK1 receptors could enhance excitability of the hypoglossal motoneurons in neonatal rats (Yasuda et al., 2001). Application α1 and β noradrenergic receptor agonist (phenylephrine and isoproterenol, respectively) on the hypoglossal nucleus potentiated inspiratory activity of hypoglossal root in the mouse brain slice preparation. However, inspiratory hypoglossal activity were inhibited by α2
noradrenergic receptor agonist (clonidine) (Selvaratnam et al., 1998). Indeed, the GG activity was differentially modulated by the cholinergic agents. Microdialysis application of acetylcholine, cholinergic agonists (carbachol and muscarine) and the acetylcholinesterase inhibitor (eserine) into the hypoglossal nucleus could suppress the GG activity, which was mainly mediated by activation of the muscarinic acetylcholine receptors (mAChRs). Furthermore, selective activation of the nicotinic acetylcholine receptors (nAChRs) would evoke inspiratory and tonic GG activity (Chamberlin et al., 2002; Liu et al., 2005). Glycine and GABA are the major inhibitory neurotransmitters in the central nervous system. Infusion glycine and GABAA agonist (muscimol) produced an inhibition on the GG activity (Morrison et al., 2002). 5-hydroxytryptamine (5-HT, serotonin) also exerts effects on the hypoglossal activity. Intravenous injection of L-5-hydroxytryptophan (5-HT precursor) and 5-HT1A receptor agonist could increase inspiratory activity of both hypoglossal and phrenic nerve. While activation of 5-HT2 receptors induced tonic activity of the hypoglossal nerve (Richmonds and Hudgel, 1996). Microdialysis of 5-HT into the hypoglossal motor nucleus increased the GG activity in anesthetized, awake and sleep
rats (Jelev et al., 2001; Sood et al., 2003). However, GG activity was not decreased by infusion of 5-HT antagonist in the hypoglossal nucleus during awake or sleeping.
Moreover, the GG activity was also not influenced by inhibition of the medullary raphe obscurus, where project serotonergic axons to the hypoglossal nucleus, in the conscious rat. These results suggest that respiratory modulatory effects of 5-HT and medullary raphe on the GG activity were probably minor in the conscious rat (Sood et al., 2005; Sood et al., 2006). In addition, thyrotropin-releasing hormone (TRH) has also reported to be able to excite inspiratory hypoglossal activity (Funk et al., 1994)
Intracellular signal pathways may have also involved in modulating hypoglossal motoneuronal excitability. Data from the in vitro study revealed that both excitatory and inhibitory currents of the hypoglossal motoneuron were potentiated in the presence of protein kinase A (PKA) (Feldman et al., 2005). Aoki et al., (2006) has shown that application of the cyclic adenosine-3’-5’-monophosphate (cAMP) analog and adenylyl cyclase activator to the hypoglossal nucleus could enhanced inspiratory and tonic GG activity in vivo, suggesting that cAMP may relate to the excitability of the hypoglossal motoeneurons. In contrast, the cyclic guanosine-3’-5’-monophosphate (cGMP) analog blocked excitatory effects of 5-HT and phenylephrine on the hypoglossal motoneurons.
Modulatory effects of the KF nucleus on the hypoglossal activity
Kuna and Remmers (1999) have observed that activation of ipsilateral hypoglossal premotor neurons located in the KF nucleus by glutamate injection could induce bilateral excitation or inhibition of the hypoglossal nerve activity, or ipsilateral excitation and contralateral inhibition. Data from the working-heart brainstem preparation also revealed that chemical stimulation of KF nucleus could evoke increase and/or decrease of the hypoglossal activity (Gestreau et al., 2005). In addition,
increasing inspiratory activity of the hypoglossal nerve during halothane treatment is also accompanied by elevation of c-fos expression in the KF nucleus (Roda et al., 2004). Moreover, expiratory-inspiratory (EI) and inspiratory (I) neurons of the KF nucleus were demonstrated to project neural pathway to the hypoglossal nucleus and showed a similar discharge pattern with the hypoglossal activity during sustained lung inflation (Ezure et al., 2003; Ezure and Tanaka, 2006). These results indicated that KF nucleus might have been involved in regulation of inspiratory activity of the hypoglossal nerve.
Activity of the hypoglossal nerve during oropharyngeal behaviors
In addition to respiratory-related activity, the hypoglossal nerve also discharged with various behaviors. GH did not display respiratory-related activity, while thyrohyoid (TH) slightly discharged at Post-I duration during eupnea in cats (Umezaki et al., 1998). During swallowing and vomiting, these two hypoglossal hyoid nerves/muscles was excited and displayed obvious activity to protect the lower airway by closing the laryngeal vestibule through elevation of the larynx (Umezaki et al., 1998). The tongue retractors (HG and SG) were activated, but tongue protrudor (GG) was silent during swallowing in the sheep (Amri et al., 1989). Both GG and SG were excited initially, however, the GG was then inhibited and the SG continued firing during swallowing (Ono et al., 1998; Tomomune and Takata, 1988). These sequences of the tongue muscle discharges indicated that contraction of the GG would induce the food to move backward, and contraction of the SG would subsequently elevate the tongue and then push the food into the pharynx. During sneezing, the SG or LHN could be excited to retract and elevate the tongue, which may increase the oral airway resistance and then cause airflow to pass through the nasal cavity (Satoh et al., 1998).
Neural control of the laryngeal muscles
Innervation of the laryngeal muscles
The laryngeal muscles are responsible for various functions, such as phonation, respiration, swallow and cough. The laryngeal musculature is complex and could be divided into the intrinsic and extrinsic laryngeal muscles (Ludlow, 2005). The intrinsic laryngeal muscles are confined to the larynx and composed of the lateral cricoarytenoid (LCA), thyroarytenoid (TA) and interarytenoid (IA), cricothyroid (CT) and posterior cricoarytenoid (PCA) (Fig. 1-3). Activation of the CT and PCA abduct the vocal folds, which maintain the laryngeal patency and inspiratory airflow.
Contrastly, contraction of the LCA, TA and IA adducts the vocal folds and reduces the expiratory airflow and maintains the end-expiratory lung volume, which would prevent the lung collapse (Bartlett, 1989). The extrinsic laryngeal muscles include the thyrohyoid and sternothyroid, which connect the thyroid cartilage with the hyoid and sternum, respectively. Contraction of these extrinsic laryngeal muscles could alter position of the larynx by moving the thyroid cartilage. Other extrinsic laryngeal muscles (stylohyoid, omohyoid, diagstric, mylohyoid, stylopharyngeus, pharyngeal constrictors and palatopharyngeus) are also involved in regulating location of the larynx.
The intrinsic laryngeal muscles are innervated by the efferent of the superior laryngeal nerve (SLN) and recurrent laryngeal nerve (RLN). The CT is controlled by external branch of the SLN, which is originated from the ventral formation of nucleus ambiguous (NA) and dorsal motor nucleus of vagus (DMV) in the rat (Pascual-Font et al., 2006a; Yajima and Hayashi, 1983). While other intrinsic laryngeal muscles (LCA, TA, IA and PCA) of the rat, supplied by the RLN, receive projections from the dorsal portion of NA (Pascual-Font et al., 2006b; Yajima and Hayashi, 1983). Transneuronal labelling study revealed that the ventral medullary surface (central chemorepceotors,
Mulkey et al., 2004), preBözinger complex (putative respiratory/inspiratory rhythm generator, Feldman and Del Negro, 2006; Smith et al., 1991) and NTS (termination site of pulmonary afferents) project to the PCA motoneurons (Waldbaum et al., 2001).
In addition to brainstem, pontine areas (KF nucleus, locus coeruleus (LC) and lateral dorsal tegmental area is also involved in regulation of the PCA muscles (Waldbaum et al., 2001).
Respiratory-related activity of the laryngeal motor outputs
The laryngeal muscles are innervated by the SLN and RLN, which display respiratory-related activity including Pre-I, inspiratory and Post-I activity in the neonatal and adult rat and in the cat (Dutschmann et al., 2000; Huang et al., 1989; Lu et al., 2005; Zhou et al., 1989). There are two main nerve branch of the RLN. One innervates the TA muscle (adducent branch) and primarily displays Post-I activity.
The other branch of RLN (abducent branch) innervated the PCA and mainly discharged during inspiration (Zhou et al., 1989). Similar discharge pattern of adducent and abducent branch of the RLN was also observed in the rat (Lu et al., 2005). Single motor unit recording from the RLN in the rabbit and cat revealed that there are several types of respiratory-related motoneurons, including I, Post-I and EI motoneurons (Ryan et al., 2002; Sica et al., 1985). Moreover, Post-I motoneurons could be divided into two types; one became silent and the other display tonic activity during hyperventilation-induced apnea (Sun et al., 2005). These laryngeal motoneurons have demonstrated to be modulated by synaptic inputs from the VLM.
The inspiratory laryngeal motoneurons might be excited by the augmenting inspiratory neurons, and inhibited by the decrementing inspiratory neurons and decrementing expiratory neurons. The expiratory laryngeal motoneuons might receive inhibitory inputs from the decrementing inspiratory neurons and augmenting
expiratory neurons (Baekey et al., 2001; Ono et al., 2006). Post-I activity of the RLN was enhanced by microinjection of glutamate into the intermediate part of the KF nucleus, while GABA injection suppressed discharge of the RLN during Post-I duration (Dutschmann and Herbert, 2006).
Respiratory-related activity of the RLN was altered during apneusis and gasping.
During eupnea, inspiratory RLN activity discharged earlier than that of the phrenic nerve, while expiratory RLN activity was after the phrenic bursting, termed Post-I activity. During apneusis, onset of the RLN activity was still prior to the phrenic nerve bursting and yet its Post-I activity was reduced. During gasping, onset of inspiratory RLN activity was after the phrenic bursting while Post-I activity was much reduced (St. John et al., 1981; St. John et al., 1989). This temporal coordination of rhythmic burst of the RLN relative to the phrenic nerve bursting has been demonstrated to be dependent on the glycinergic modulation. In this regard, Dutschmann and Paton (2002) observed that blockade of glycine receptor or trigeminal stimulation shifted Post-I activity of the RLN to advance during inspiration, which caused a paradoxical adduction of the glottis during inspiration in the neonatal rat. Intravenous injection of strychnine (glycine receptor antagonist) also evoked a shifting of Post-I laryngeal motoneurons discharged during inspiration in the adult rat (Berkowitz et al., 2005).
However, direct microinjection of strychnine into the NA did not induce phase switching of the laryngeal motoneuorns. These results strongly suggest that temporal coordination of glycine inhibition may be important in various functions and might occur at premotor system of the laryngeal motoneurons.
Chemoreceptor activation with hypercapnia and hypoxia enhanced inspiratory activity of the SLN, and abducent branch of the RLN (Huang et al., 1989; Kuna, 1987;
Zhou et al., 1989). Post-I activity of adducent branch of the RLN was also enhanced by hypocapnia and hypoxia (Huang et al., 1989). Excitation of peripheral
chemoreceptors (carotid and aortic body) by administration of NaCN, lobeline and isoproterenol resulted in augment of the RLN activity (Paton et al., 1999; van Lunteren et al., 1984). Interesting, distension of the pharyngeal-oesophageal junction also excited Post-I discharge of the RLN (Paton et al., 1999). Paton et al. (1999) demonstrated that some neurons in NTS receive both signals from peripheral chemoreceptor and pharyngoesophageal receptor, suggesting that NTS may integrate peripheral inputs and regulate the laryngeal patency.
Both inspiratory and Post-I activity of the RLN was increased after withholding lung inflation at end-expiration (Huang et al., 1989; Kuna, 1987; St. John and Zhou, 1990). Excitation of tonic activity of the pulmonary stretch receptors by sustained lung inflation produces inhibition on discharge of RLN during inspiration and expiration (St. John and Zhou, 1990 and 1992). Elevation of end-expired lung volume by increasing PEEP produced a progressive reduction in adducent branch of the RLN activity in the cat (St. John and Zhou, 1990).
Activation of pulmonary vagal C-fiber (PCF) receptors by intra-jugular capsaicin administration produced enhancement of adducent branch but inhibition of abducent branch of the RLN. Simultaneously, Post-I activity of the RLN was transformed into tonic discharge pattern (Lu et al., 2005). Holmes and Remmers (1989) demonstrated similar responses of the PCA and TA muscles to selectively electrical stimulation of PCF receptors in the decerebrate cat. These reflexive alterations in discharge of the RLN or laryngeal muscles suggest that activation of PCF receptors under the pathological conditions may induce a reflexively glottal closure (Diaz et al., 1999; Lu et al., 2006). In addition, Post-I activity of RLN/adductor muscles were excited and became tonic after clonidine (α2 adrenergic receptor agonist) treatment in the goat (Hedrick et al., 1995; O’Halloran et al., 1999).
Obstructive sleep apnea/hypopnea syndrome
Obstructive sleep apnea/hypopnea syndrome (OSAHS) is a common sleep breathing disorder and characterized by periodic complete or partial occlusion of the UAW during sleep. Complete UAW occlusion could induce apnea, which is defined as cessation of the airflow for more than 10 seconds (American Academy of sleep medicine, 1999). However, partial UAW occlusion may evoke hypopnea, which is defined as a decrease in airflow at least 30 % for 10 seconds (American Academy of sleep medicine, 1999). OSAHS has been characterized by apnea-hypopnea index (AHI), representing numbers of apneas and/or hypopneas per hour during sleep.
Severity of OSAHS could be classified into three levels: mild (5 ≦ AHI < 15), moderate (15≧ AHI ≦ 30) and severe (AHI > 30) (American Academy of sleep medicine, 1999). Young et al., (2002) estimated that 20 % and 6.7 % of the white people might have mild and moderate OSAHS, respectively. Syndromes of OSAHS included snoring, witnessed apneas, daytime sleepiness and so on (Malhotra and White, 2002; Qureshi and Ballard, 2003). The correct features of OSAHS could be diagnosed by the overnight polysomnogram.
The anthropometric factors (age, sex, race and genetics) have reported to contribute to pathogenesis of OSAHS. Young et al., (2002) reported that numbers of male patients with OSAHS were 2 to 3 times greater than female patients. This phenomenon may be due to difference of sexual hormone and/or fat distribution.
Possibility of suffering OSAHS was increased with aging in a limited range (Bixler et al., 1998). Prevalence of OSAHS is higher and earlier developed in African-Americans than Caucasians (Redline et al., 1997). Moreover, Some studies also indicated that OSAHS is a kind of familial diseases (Gislason et al., 2002;
Redline et al., 1995). Redline et al., (1995) reported that 35 % family have two or more members suffering OSAHS.
Smoking is also a risk factor for OSAHS. Many studies observed that OSAHS or snoring is more frequent in smokers than non-smokers (Franklin et al., 2004; Kashyap et al., 2001; Wetter et al., 1994). Possible mechanisms of higher prevalence of OSAHS in smokers may be resulted from the airway inflammation and/or increase of sleep instability (Young et al., 2002). Study from Vgontzas et al., (2004) confirmed that neutralizing tumor necrosis factor-alpha (TNF-α), a kind of inflammatory mediators, could reduce severity of sleep apnea, suggesting an interaction between the inflammation and OSAHS.
Pathogenesis of obstructive sleep apnea/hypopnea syndrome
Collapse of the UAW is the most common characteristic of OSAHS. Patency of the UAW is influenced by the UAW anatomy, activity of the UAW muscles and inspiratory negative pressure. Obstruction of the UAW usually occurs at the level of the pharynx. The pharynx is surrounded by various bones and soft tissues, and could be divided into four compartments, including the nasopharynx, velopharynx, oropharynx and hypopharynx. The nasopharynx lies between the posterior nasal septum and the soft palate. The velopharynx is located behind the soft palate. The oropharynx is situated between the soft palate and epiglottis. The hypopharynx extends from the tongue base to the larynx (Ayappa and Rapoport, 2003; Pierce and Worsnop, 1999).
Various factors may contribute to obstruction of the pharynx. Reduction of pharyngeal airway diameter is one of these factors and probably due to abnormal structure of the soft tissue and skeleton in the pharynx. Schwab et al. (2003) observed that people who suffered from sleep apnea have a large size of the tongue, lateral pharyngeal walls and soft tissues. Enlarged soft palate and uvula also increased risk of suffering OSAHS (Stauffer et al., 1989; Yu et al., 2003). Furthermore, displacement of
the mandible, maxilla and hyoid bone may be predisposing factors inducing narrowing of the pharynx (Riha et al., 2005; Watanabe et al., 2002). Fat deposition in the neck may also press the pharynx and reduce diameter of the pharynx (Deegan and McNicholas, 1995). In addition to anatomic factors, activity of the pharyngeal dilator muscles may play an important role in modulating the UAW patency. The pharyngeal dimension depends on a balance between collapsing and dilating force on the pharynx.
Collapsing force is primarily generated by inspiratory negative pressure produced by contraction of the diaphragm and chest muscles. Dilating force is resulted from activation of the pharyngeal dilating muscles, which comprise the palatal, GG and hyoid muscles (Pierce and Worsnop, 1999). Most of these pharyngeal dilator muscles display respiratory-related or tonic activity, which are related to the UAW patency.
(Bianchi et al., 1995; Fregosi and Fuller, 1997; O’Halloran et al., 1996; St John, 1998;
Tangel et al., 1995).
The palatal muscles (tensor palatine, palatopharyngeus, levator palatini and palatoglossus) could regulate position of the palate and determine patency of the velopharynx. Many studies have shown that Patients with OSAHS and normal subjects display lower activity of these palatal muscles during sleep (Carlson et al., 1995; Mezzanotte et al., 1996; Tangel et al., 1991). The GG, the tongue protrudor muscles, could induce anterior movement of the tongue and increase area of the oropharynx. Thus, decrease in activity of the GG has been reported to coincide with occurrence of apnea or hypopnea (Carlson et al., 1995; Katz and White, 2004;
Remmers et al., 1978;). Position of the hyoid arch is mainly controlled by the activity of the suprahyoid (geniohyoid) and infrahyoid (thyrohyoid and sternohyoid) muscles.
Stimulation of the hyoid muscles and/or its motor nerves has been demonstrated to decrease the UAW resistance at the hypopharyngeal level in dogs (van de Graaff et al., 1984; Yoo and Durand, 2005). Clinic studies revealed that activity of the GH reduced
during sleep (Wiegand et al., 1990a and 1990b). Thus, decrementing activity of the hyoid muscles may increase pharyngeal collapsibility.
Based on the above-mentioned studies, the pharyngeal dilating muscles are involved in maintenance of the UAW patency. Activity of the pharyngeal dilators could be modulated by various respiratory-related stimuli, including hypercapnia, hypoxia, lung volume changes and negative UAW pressure (Amis et al., 1999; Bailey et al., 2001; Fregosi and Fuller, 1997; van de Graaff et al., 1984; van der Touw et al., 1994). Impaired reflexes of the pharyngeal dilators to these stimuli during sleep may have a risk of obstruction of the UAW (Horner et al., 1994; Horner et al., 2002;
Mortimore and Douglas, 1997; Shea et al., 1999; Wheatley et al., 1993).
Consequence of obstructive sleep apnea/hypopnea syndrome
Obstruction of the UAW in patients with OSAHS always display hypercapnia, hypoxia and overexcitation of sympathetic activity and may evoke a variety of cardiopulmonary diseases. Young et al., (1997) and Peppard et al., (2000) found that there is a positive relation between sleep apnea/hypopnea and hypertension. Other study also indicated that OSAHS may associate with the myocardial infarction, congestive heart failure and stroke (Shahar et al., 2001). Pulmonary diseases such as pulmonary hypertension and asthma are always occurred in patients with OSAHS (Chan et al., 1988; Hetzel et al., 2003). Recently, there is a hypothesis that cardiopulmonary diseases may be caused by OSAHS-induced inflammation (Bergeron et al., 2005). Patients with OSAHS have a higher level of circulating pro-inflammatory mediators, such as tumor necrosis factor-alpha (TNF-α), interleukin-8 (IL-8), intercellular adhesion molecule-1 (ICAM-1), monocyte chemo-attractant protein-1 (MCP-1) and reactive oxygen species (ROS) (Hatipoglu and Rubinstein, 2003; Ohga et al., 2003; Ryan et al., 2005). These pro-inflammatory
mediators may have been related to various cardiopulmonary diseases (Granger et al., 2004). Continuous positive airway pressure therapy, an effectively well-known treatment of OSAHS, has been reported to decrease pro-inflammatory mediators and may decrease possibility of suffering OSAHS-induced disorders (Ohga et al., 2003).
Pulmonary vagal afferents
Pulmonary vagal afferents can detect physiological and pathological conditions of the lungs and relay these signals to the central nervous system. Activity of pulmonary vagal afferents are considered to be involved in regulation of cardiorespiratory functions. Pulmonary vagal afferents can be primarily divided into three categories, pulmonary vagal C-fiber (PCF) receptors, slowly adapting pulmonary stretch receptors (SARs) and rapidly adapting receptors (RARs).
Pulmonary vagal C-fiber receptors
Types of pulmonary vagal C-fiber receptors
Unmyelinated PCF receptors are major sensory nerve fibers of the vagal afferent in the cat (Agostoni et al., 1957) formerly named as J receptors, which were discovered by Paintal (1955). PCF receptors are the primary chemosensitive afferents in the lung (Ho et al., 2001). Coleridge and Coleridge (1977 and 1986) considered that PCF receptors of the dog could be further divided into the pulmonary C-fiber and bronchial C-fiber receptors according to their blood supply via pulmonary and bronchial circulation, respectively. Both pulmonary and bronchial C-fiber receptors could be excited by capsaicin. However, only bronchial C-fiber receptors could be stimulated by phenyldiguanide (PDG) (Coleridge and Coleridge, 1977). In addition to different sensitivities to chemical stimuli, Chen et al., (1999) observed that there are two groups of PCF receptors in the dog, the so-called high-resistance and low-resistance C-fibers which could be blocked by different concentration of perivagal capsiacin treatment and temperature. Usually, this classification of PCF receptors does not apply to the rat (Bergren and Peterson, 1993). Indeed, pulmonary C-fiber receptors in the rat’s lung are sensitive to PDG (Bonham and Joad, 1991).
Nevertheless, Undem et al., (2004) suggested that PCF receptors in the guinea pig
could be classified into two distinct groups based on different embryonic development.
Those C-fibers originated from the jugular ganglion, which derived from the neural crest, innervate the intrapulmonary and extrapulmonary airways and contain substance P (SP). However, C-fibers originated from the nodose ganglia, which derived from the epibranchial placode, project almost to the intrapulmonary airways and do not express SP (Undem et al., 2004). Both subtypes of C-fibers could be activated by capsaicin. Yet, in contrast with the nodose C-fibers, the jugular C-fibers were poorly responsive to ATP, 5-hydroxytryptamine (5-HT) and anandamide (Chuaychoo et al., 2005; Lee et al., 2005; Undem et al., 2004). In the mouse, PCF receptors are subgrouped into capsaicin-sensitve and capsaicin-insensitve types.
Capsaicin-sensitive C-fibers could be also excited by bradykinin via B2 receptor activation and display lower conduction velocities (< 0.7 m/s) as compared with capsaicin-insensitive C-fibers (0.7 ~ 1.5 m/s) (Kollarik et al., 2003).
Sensitivity of pulmonary vagal C-fiber receptors
There are many external and internal stimuli producing activation of PCF receptors. Inhalation of volatile anesthetics and exogenous irritants, such as wood smoke, cigarette smoke, sulfur dioxide, ammonia and ozone, could excite PCF receptors and induce cardiorespiratory reflex (Ho and Lee, 1998; Lai and Kou, 1998;
Lee et al., 1990; Mutoh et al., 1998; Schelegle et al., 2001; Vesely et al., 1999; Wang et al., 1996). Pathological condition that occurred in the airways and lungs also activate PCF receptors. In this regard, endotoxemia could evoke long lasting excitation of PCF receptors via hydroxyl radical and cyclooxygenase metabolites (Lai et al., 2002; Lai et al., 2005). Intratracheal injection of synthetic cationic proteins (poly-L-lysine and poly-L-arginine) or endogenous cationic proteins, which secreted by the inflammatory cells, would excite and sensitize PCF receptors (Lee and Gu,
2003; Lee et al., 2001; Gu and Lee, 2001). Neuropeptide released from the small cell lung cancer could induce sensitization of PCF receptors as well (Gu and Lee, 2005).
Pulmonary air embolism and edema has also been demonstrated to activate PCF receptors (Chen and Kou, 2000; Chen et al., 2000; Diaz et al., 1999).
Capsaicin, the main pungent ingredient of the red peppers, could activate the transient receptor potential channel vanilloid subfamily member 1 (TRPV1), to induce excitation of PCF receptors (Caterina et al., 1997; Ho et al., 2001; Kollarik and Undem, 2004; Lee and Pisarri, 2001). TRPV1 contains six transmembrane domains and is a ligand-gated and non-selective cation channel (Caterina et al., 1997; Cortright and Szallasi, 2004). Heat, proton and endogenous arachidonic acid metabolites could directly activate TRPV1. Hence, these stimuli could activate and/or sensitize PCF receptors via TRPV1 activation (Gu and Lee, 2002; Ho et al., 2000; Kollarik et al., 2003; Kwong and Lee, 2002; Lin and Lee, 2002; Lin et al., 2005; Ruan et al., 2005).
Other receptors also contribute to activation of PCF receptors. Adenosine generated from degradation of ATP during hypoxia and ischemia could bind to the adenosine A1 receptor and then sensitize and/or excite PCF receptors (Gu et al., 2003; Hong et al., 1998). P2X receptors could mediate stimulatory effects of smoke, ROS and ATP on PCF receptors (Pelleg and Hurt, 1996; Ruan et al., 2005). 5-HT3 receptor agonist (phenylbiguanide; PBG) could also evoke excitation of PCF receptors (Veelken et al., 1997). Recently study indicated that nicotine could induce excitation of the vagal pulmonary sensory neurons via the neuronal nicotine acetylcholine receptors (NnAChRs) in vitro. In addition, PCF receptors could be also excited by intra right-atrial injection of nicotine in anesthetized rats (Xu et al., 2006). Acid-sensing ion channels (ASICs) as well as TRPV1 are involved in hydrogen-induced excitation of PCF receptors (Gu and Lee, 2006; Kollarik M and Undem). Bradykinin B2 receptor, a kind of G-protein-coupled receptor, could also contribute to activation of PCF
receptors via at least two pathways. First, activation of B2 receptor may produce metabolites of arachidonic acid by lipoxygenase. These metabolites could bind to TRPV1 and then activate PCF receptors (Carr et al., 2003). Second, Bradykinin could excite vagal afferent by closing K+ channel and activating Ca2+-activated Cl- channel, which induce efflux but not influx of Cl- in the afferent neuron (Lee et al., 2005) Various receptors (Histamine H3 receptor, dopamine D2-likereceptor and opioid receptors) exert inhibitory effects on PCF receptors (Groneberg et al., 2004; Lin et al., 2003; Nemmar et al., 1999). Although PCF receptors are more sensitive to chemical stimuli, some mechanical stimuli could also excite PCF receptors. Bronchial C-fibers of the rabbit could monitor the lung compliance (Ma et al., 2003). Hyperinflation and changes of PEEP could induce slightly excitation of PCF receptors (Ho et al., 2001;
Kaufman et al., 1985).
Reflexes induced by pulmonary vagal C-fiber receptor activation
Activation of PCF receptors could induce central-mediated pulmonary chemoreflexes and peripheral axon reflexes (Lee and Pisarri, 2001; Lee et al., 2003).
Signals from PCFs transmit to the caudal-medial part of the commissural NTS via activation of Non-NMDA, NMDA and 5-HT3 receptors (Bonham and Joad, 1991;
Jeggo et al., 2005; Jordan, 2001; Wilson et al., 1996). These NTS relay neurons may further project to cardiorespiratory related region in the central nervous system and then induce pulmonary chemoreflex (Bonham and Joad, 1991; Kawai and Senba, 2000; Neff et al., 1998). Pulmonary chemoreflex characterized by bradycardia, hypotension and apnea, sometimes may followed by rapid shallow breathing (Coleridge and Coleridge, 1977). Additionally, the glottis may tightly close during the period of apnea because of excitation of abducent branch of the RLN (Lu et al., 2005, 2006). Stimulation of PCF receptors also evoked cholinergic-mediated
bronchoconstriction and mucus secretion (Barnes, 2001). In addition, activiy of limb muscles could be inhibited by PCFs activation and resulted in stopping exercise (Gandevia et al., 2000).
Local axon reflex evoked by stimulation of PCF receptors induce release of neuropeptide (substance P; SP, calcitonin gene-related peptide; CGRP and neurokinin A; NKA) from C-fiber endings. These neuropeptides could bind to G protein-coupled receptors and evoke bronchoconstriction, vasodilatation, extravasation and hypersecretion, so-called neurogenic inflammation (Barnes, 2001; Belvisi, 2003). SP and NKA are major neuropeptide in the lung and airway. SP mainly binds to neurokinin 1 (NK1) receptors, which are situated in the airway epithelium, submucosal glands and vessels. NKA primarily activates NK2 receptor, which located on the airway smooth muscles (Groneberg et al., 2004). SP could further binds to presynaptic NK1 receptor on the parasympathetic nerve in the lung and then facilitate release of acetylcholine (Nemmar et al., 1999). In addition, SP has immunomodulatory effects on various types of immune cells (Joos et al., 2000), suggesting a linkage between immune and nervous system.
Slowly adapting pulmonary stretch receptors
Types of slowly adapting pulmonary stretch receptors
Slowly adapting pulmonary stretch receptors (SARs) was found by Adrian in 1933 and characterized by Knowlton and Larrabee in 1946. Discharge pattern of SARs is correlated with the level of lung volume and adapts slowly during sustained lung inflation. SARs are located in the airway smooth muscles and mostly distributed in the intrapulmonary airways. Some SARs are located in the intrathoracic and extrathoracic trachea, and extrapulmonary bronchus (Bartlett et al., 1979; Keller et al., 1989; Krauhs, 1984; Ravi, 1986). SARs exhibit different discharge pattern and
respond differently to lung volume change. Briefly, the intrathoracic SARs display rhythmic discharge with respiration, while extrathoracic SARs have more irregular discharge pattern. According to discharge pattern, SARs was divided into high threshold SARs and low threshold SARs. High threshold SARs active only during inspiration, whereas low threshold SARs discharged throughout whole respiratory cycle and exhibited higher discharge rate during inspiration at end expiratory lung volume (Sant’Ambrogio, 1982; Schelegle and Green, 2001). Bergren and Peterson (1993) further grouped SARs into the inflationary SARs, deflation SARs and biphasic SARs (most inflationary and most deflationary SARs) in the rat. Whether these different types of SARs exerted different influence on control of breathing is still remained to be determined. Recently reports further found that a single SAR has multiple receptive fields, which has its own encoders (Yu, 2005; Yu and Zhang, 2004).
These observations may explain why some discharge pattern of afferents ranged from typical SARs and RARs.
Sensitivity of slowly adapting pulmonary stretch receptors
Pulmonary SARs could detect changes of lung mechanics. Different types of SARs responded to lung inflation and/or lung deflation. Increase of the tracheal pressure evoked by increasing tidal volume would enhance discharge of SARs during inflation phase (Yu et al., 1991). Reduction of the lung compliance by regulating end-expired pressure would eliminate SARs activity during deflation phase of respiration (Ma et al., 2003; Yu et al., 1991). SARs activity was enhanced during smooth muscle contraction induced by vagal stimulation or administration of histamine (Matsumoto, 1996; Matsumoto et al., 1992). Although mechanism of SARs activation is not fully understood. Some reports have showed that activity of SARs could be also modulated by the stretch-activated and voltage-sensitive channels. Ma et
al., (2004) observed that hyperinflation-induced activation of inflationary SARs could be attenuated by application of gadolinium chloride (inhibitor of the stretch-activated channels) and inhibition of the sodium channels but not potassium channel (Matsumoto et al., 2000), suggesting that the increase in lung volume could activate the stretch channels and sodium channels on nerve endings of SARs. Similarly, lung deflation-evoked excitation of deflationary SARs is mediated by the flecainide-sensitive sodium channels (Matsumoto et al., 2002). Interestingly, activity of SARs would be reduced after hyperinflation-induced excitation. This phenomenon is resulted from slow afterhyperpolarization of SARs mediated by activation of the Na+-K+ pump (Matsumoto et al., 2000).
In addition to respond to mechanical stimuli, SARs could also detect a few of chemical stimuli. Administration of veratridine (a kind of veratrum alkaloids) induced tonic discharge of SARs through action on the sodium channel and Na+-K+ pump (Matsumoto and Shimizu, 1995; Matsumoto et al., 1998). CO2 has been demonstrated to inhibit SARs activity (Green et al., 1986). This inhibitory effect of CO2 may be resulted from increasing concentration of hydrogen and activation of 4-AP sensitive voltage-gated K+ channel (Matsumoto et al., 2000 and 2004). Inhalation of SO2
attenuated discharge of SARs in the rabbit but not in other species (Callanan et al., 1975; Coleridge and Coleridge, 1986). Some volatile anesthetics could depress SARs activity and attenuate Hering-Breuer reflex (Mutoh et al., 1998).
Activity of SARs have been shown to be changed under pathological conditions.
Pulmoanry air embolism exerted distinct influence on SARs of different location.
Pulmoanry air embolism inhibits SARs located at smaller airways (< 2mm), which may be due to collapse of the small airway, but excite SARs located within large airway (> 2mm) in the dog (Lee et al., 1994). During bleomycin-induced pulmonary fibrosis, sensitivity of SARs to lung volume changes was enhanced, but responses to
transpulmonary pressure were reduced (Schelegle and Green, 2001).
Reflexes induced by slowly adapting pulmonary stretch receptors
Activity of SARs may play an important role in regulation of breathing pattern.
Signals from SARs transmitted to pump cells of NTS are via activation of non-NMDA receptors (Bonham et al., 1993). Pump cells that drove by SARs would then project to various regions in the central nervous system to regulate breathing pattern (Ezure and Tanaka, 1996). The typical reflexes induced by activation of SARs are Hering-Breuer reflexes. Both phasic and tonic activity of SARs has been considered to regulate respiratory pattern. Elevation of static airway pressure produced inhibition on phrenic bursting and respiratory frequency in the dog (Mitchell et al., 1980). Phasic lung inflation or vagal stimulation during inspiration would inhibit inspiratory duration and phrenic discharge. Expiratory duration was elongated when lung inflation was introduced during the expiration by cycle trigger pump (Feldman and Gautier, 1975).
These changes of breathing pattern could be abolished by a bilateral vagotomy, suggesting effects of lung volume changes and vagal stimulation were mediated by the vagus nerve. These phasic and tonic SARs activity could also modulate breathing pattern in the perfused rat preparation in both eupnea and gasping (Harris and St. John, 2003 and 2005). Moreover, SARs also regulate the UAW patency by modulating UAW motor outputs and airway smooth muscles (Kuna, 1987; Schelegle and Green, 2001; St. John and Zhou , 1992; van Lunteren et al., 1984).
Rapidly adapting receptors
Physiology of rapidly adapting receptors
RARs were first explored by Keller and Loeser (1929) and also called irritant receptors. RARs displayed irregular discharge pattern and adapted rapidly to sustained
lung inflation (Coleridge and Coleridge, 1986; Sant’Ambrogio and Widdicomne, 2001). RARs are Aδ myelinated fibers and located throughout the airway from the nasopharynx to intrapulmonary airway, and respond to various types of mechanical and chemical stimuli (Mortola et al., 1975; Widdicombe, 2003). RARs were excited by lung inflation at different rate of inflation. At the same inflation volume, the larger inflationary rate produced more excitation on RARs (Pack et al., 1983). Excitation of RARs during lung inflation may be mediated thought activation of the stretch-activated channel (Ma et al., 2004). Moreover, Matsumoto et al. (2001) revealed that responses of RARs to hyperinflation were potentiated by 4-AP (a potassium channel blocker), which also reduced adaptation index of RARs, suggesting the potassium channel could regulate excitability of RARs. Reduced lung compliance by restricting lung inflation in the spontaneously breathing dog or by altering end-expired pressure in the artificially ventilated rabbit could also activate RARs (Ma et al., 2003; Pissarri et al., 1990). Furthermore, Kappagoda and Ravi (2006) revealed that activity of RARs was enhanced by increasing extravascular fluid in the airway during elevation of left atrial pressure, microvascular permeability and obstruction of pulmonary lymphatic drainage.
RARs are also sensitive to stimuli of some external and internal chemicals.
Capsaicin has been reported to excite 89% of PCF receptors and 14.5 % of RARs in the rat. RARs also slightly responded to PBG and inhalation of SO2 (Ho et al., 2001).
Acid has been demonstrated to activate RARs via action of hydrogen on ASICs (Undem et al., 2002). Tachykinins (substance P and NKA) released from C-fiber endings, could also induce excitation of RARs via action of NK1 and NK2 receptors (Matsumoto et al., 1997). Inhalation of ammonia could cause an increase in RARs activity by directly acting on nerve terminals. On the contrary, increases of RARs activity following administration of histamine were significantly attenuated by
atropine and isoprenaline, suggesting that excitatory responses of RARs were secondly mediated by bronchoconstriction (Matsumoto, 1989). Inhalation of cigarette smoke could stimulate RARs via directly action of nicotine on nicotinic acetylcholine receptors on nerve terminals and indirectly effects by nicotine-induced smooth muscle contraction (Lee et al., 2007). Additionally, wood smoke exerted excitatory effects on RARs by smoke particulate, hydroxyl radical and cyclooxygenase metabolites (Lai and Kou, 1998). The same research group also revealed that excitatory responses of RARs to circulatory endotoxin were mediated by cyclooxygenase metabolites and hydroxyl radical induced bronchoconstriction (Lai et al., 2002 and 2005).
Reflexes from RARs activation
RARs afferents terminated at the medial and commissural NTS and released glutamate to activate non-NMDA receptors of RAR relay neurons (Ezure et al., 1999).
RAR relay neurons not only receive afferent inputs from RARs but also integrated signals from pump cells (relay neurons of SARs) and central inspiratory rhythm generator. In addition, RAR relay neurons were modulated by tonic GABAergic inputs (Ezure et al., 1999). Most RAR relay neurons projected to pontine and ventral respiratory group (VRG) to regulate breathing patterns (Otake et al., 2001). Davies and Roumy (1982) observed that augmented breath was evoked by applying positive and negative pulse pressure during inspiration in rabbits with blockade of SARs. In addition, expiratory duration was shortened when introducing negative pulse pressure during expiration. RARs also contributed to enhance inspiratory activity during deflation reflex (Davies et al., 1978). Tachypnea induced by inhaling ROS was resulted from stimulation of RARs (Ruan et al., 2003). RARs arising from the trachea and bronchi were considered to evoke cough during stimulation (Sant’Ambrogio and Widdicomne, 2001).
The significance and purpose of the present study
Respiratory-related activity of the UAW motor outputs plays an important role in the upper airway patency. Dysfunction of these UAW motor activities may contribute to the OSAHS. Young et al. (2002) estimated that there are about 20 % adults having been suffered from mild OSAHS. Patients with OSAHS usually also suffer from a variety of the cardiovascular and pulmonary disorders (Malhorta and White, 2002;
Qureshi and Ballard, 2003; Young et al., 2002). In addition, some studies report that OSAHS are associated with the impairment of cognition (Engleman and Douglas, 2004; Peng et al., 2004). Furthermore, patients with OSAHS have higher traffic accident rates (George, 2004). These consequences of OSAHS will consume many medical resources and society costs. Hence, it is critical to understand the pathogenesis of OSAHS. Several clinic studies have indicated that reduction of respiratory-related activities of the tongue muscles might be a risk factor for OSAHS (Deegan and McNicholas, 1995; Harper and Sauerland, 1978; Remmers et al., 1978;
Ryan and Bradley, 2005). These observations suggested that the tongue muscles play an important role in the maintenance of UAW patency. PCF receptors are the main chemosensitive afferents in the lung and could be excited and/or sensitized by various external irritants and internal chemicals. Activation of PCF receptors induces central-mediated pulmonary chemoreflex, characterized by apnea, hypotension and bradycardia (Coleridge and Coleridge, 1986). In addition, our previous studies also indicated that activation of PCF receptors could evoke the glottal closure by shifting post-inspiratory activity to discharge earlier during inspiration and inhibiting inspiratory activity of the recurrent laryngeal nerve (Lu et al., 2002; Lu et al., 2005;
Lu et al., 2006). However, it is unclear whether PCF receptor activation could modulate the pharyngeal patency by influencing activities of the hypoglossal nerve or tongue muscles. It is critical to understand the response of the hypoglossal nerve
and/or other UAW motor nerves to the activation of pulmonary vagal afferents. To reveal the effect of activation of pulmonary vagal afferents on hypoglossal activity or other UAW motor discharge, there are six projects conducted in this dissertation.
The first project was to investigate responses of the hypoglossal nerve to capsaicin-induced pulmonary vagal C-fiber receptors activation. Since inspiratory activity of the recurrent laryngeal and superior laryngeal nerve was inhibited by PCF receptors activation, I therefore proposed that inspiratory-related activity of the hypoglossal nerve might be also reduced by intra-jugular capsaicin administration. To test this hypothesis, the whole hypoglossal nerve and phrenic nerve were simultaneously recorded in response to capsaicin treatment. The second project was to examine whether response of two hypoglossal branches to capsaicin-induced activation of PCF receptor was similar or dissimilar. Many reports indicated that inspiratory-related activity of the tongue protrudor and retractor muscles responded similarly to the activation of the chemoreceptors and pulmonary stretch receptors. My hypothesis was that capsaicin administration may also produce a similar inhibition on the neural drives to the tongue protrudor and retractor muscles. The third project was to study the alterations of hypoglossal motoneuron discharge in response to capsaicin-induced activation of PCF receptors in rats. There are two mechanisms, the increase in discharge rate and recruited principle, for the motor increase to various stimuli. According to my hypothesis in the first two projects, activation of PCF receptors might produce reflexive inhibition on inspiratory activity of both hypoglossal branches, this effect would reflect from their decrease in motoneuron discharges. Therefore, data from single fiber recording on the hypoglossal branches might underlie the response of the hypoglossal branches to PCF receptor activation.
My hypothesis was that discharge rate of the hypoglossal motoneurons might decrease if the whole hypoglossal nerves or two branches were inhibited by capsaicin
administration. The fourth project was to examine whether response of the hypogloosal nerves to PCF receptor activation might reflect on the tongue force development. Since the hypoglossal nerve innervates the tongue muscles, reflex response of the hyplgossal nnerve to PCF receptor activation should reflect on the development of the tongue muscle force. This is very important to understand if PCF activation produced an advantage or disadvantage for UAW patency. Therefore, changes of the protrusive and retractive tongue force development were induced by airway occlusion and then evaluated with capsaicin-induced PCF receptor activation in this project. When tongue muscle force development was inhibited by PCF receptor activation, a disadvantage for UAW patency might be concluded and vice versa.
In addition, modulatory effects of other pulmonary vagal afferents, the SARs and RARs, on rhythmic activity of the UAW motor nerves were also examined in my dissertation. SARs-mediated Hering-Breuer reflexes have been well documented since 1868. Most of the previous studies focused on alterations in the amplitude of the UAW motor outputs during Hering-Breuer reflexes. These studies considered that rhythm of the UAW motor nerve are synchronized with the phrenic nerve. However, if the phrenic nerve and the UAW motor nerves receive a single rhythmic input, then rhythmic activity of the phrenic nerve and UAW motor nerves should commence at the same time. Indeed, discharge onset of the UAW motor nerves precedes the phrenic nerve bursting and has been established the notion that rhythmic hypoglossal activity can be uncoupled from the phrenic burst during elevation of PEEP, which may activate both SARs and RARs activity (Ezure et al., 2003). These observations mean that hypoglossal rhythm remains during phrenic apnea induced by PEEP, suggesting a possibility that respiratory rhythm of the hypoglossal nerve and the phrenic nerve bursting may have been differentially modulated by peripheral inputs. Therefore, neural control of rhythmic hypoglossal activity may differ from the phrenic bursting
when SARs and RARs are activated by changing lung volume. I proposed herein that uncoupling might be a common property for UAW motor outputs and thus respiratory rhythmic activity of the UAW motor nerves could be also uncoupled during phrenic apnea induced by increase of PEEP. If true, mechanism for uncoupled activity of the UAW motor nerves was studied by single motoneuron discharge to reveal whether uncoupling was due to the recruitment of expiratory motoneurons or a persistent discharge of inspiratory motoneuron activity. Hence, the fifth project was undertaken to examine my hypothesis that UAW motor activity could be uncoupled during phrenic apnea by positive end-expired pressure in the rat by simultaneous recording of discharge of the facial, hypoglossal, superior laryngeal, and recurrent laryngeal nerve and also of their motoenurons during application of PEEP.
Uncoupling is primarily because PEEP produced inhibition on the phrenic bursting and also on the inspiratory activity of the hypoglossal nerve (Ezure et al., 2003) and is probably contributed by the preceding activity, termed Pre-I, of the hypoglossal nerve. Pre-I UAW motor activity may play a role in the maintenance of UAW patency. However, which neurotransmitter is responsible for Pre-I activity of UAW motor nerves is still unclear. It has been well documented that glycinergic inhibition participated in modulating respiratory rhythm and also coordinating respiratory activity of the cranial and spinal nerve. Blockade of glycine receptor or reduction of chloride concentration induced shift forward of post-inspiratory activity of the recurrent laryngeal nerve from expiratory stage to inspiratory duration. In other words, reduction of synaptic inhibition may synchronize the inspiratory and expiratory activity at inspiratory phase. Accordingly, I proposed that glycinergic inhibition might participate in regulating Pre-I activity. The sixth project was designed to examine this possibility that glycinergic transmission plays a role in modulating pre-inspiratory hypoglossal nerve activity in the rat. The result obtained
showed that application of strychnine not only attenuated Pre-I activity of the hypoglossal nerve during control condition and also the earlier onset of hypoglossl activity induced by sustained lung inflation and PEEP manipulation.
Tables Table 1-1
Origin and insertion of the nasal muscles
Name Origin Insertion M. nasalis, pars transversa Maxilla Aponeurosis on nasal
dorsum
M. nasalis, pars alaris Maxilla Alar skin, minor alar cartilages
M. depressor septi Maxilla Greater alar cartilage M. levator labii superioris
alaeque nasi
Maxilla Upper lip, nasolabial fold
M. procerus Occipitofrontalis muscle Aponeurosis on nasal dorsum
M. dilatator naris Greater alar cartilage Alar skin
M. apicis nasi Greater alar cartilage Skin of nasal tip From Laryngoscope 108(7): 1025-1032, 1998.
