- Top of page
- MATERIAL AND METHODS
- LITERATURE CITED
Neurons innervating the intrinsic muscles of the larynx are located within the nucleus ambiguus but the precise distribution of the neurons for each muscle is still a matter for debate. The purpose of this study was to finely determine the position and the number of the neurons innervating the intrinsic laryngeal muscles cricothyroid, posterior cricoarytenoid, and thyroarytenoid in the rat. The study was carried out in a total of 28 Sprague Dawley rats. The B subunit of the cholera toxin was employed as a retrograde tracer to determine the locations, within the nucleus ambiguus, of the neurons of these intrinsic laryngeal muscles following intramuscular injection. The labelled neurons were found ipsilaterally in the nucleus ambiguus grouped in discrete populations with reproducible rostrocaudal and dorsoventral locations among the sample of animals. Neurons innervating the cricothyroid muscle were located the most rostral of the three populations, neurons innervating the posterior cricoarytenoid were found more caudal, though there was a region of rostrocaudal overlap between these two populations. The most caudal were the neurons innervating the thyroarytenoid muscle, presenting a variable degree of overlap with the posterior cricoarytenoid muscle. The mean number (±SD) of labelled neurons was found to be 41 ± 9 for the cricothyroid, 39 ± 10 for the posterior cricoarytenoid and 33 ± 12 for the thyroarytenoid. Anat Rec, 296:470–479, 2013. © 2013 Wiley Periodicals, Inc.
The larynx receives its motor innervation from two branches of the vagus nerve, the superior laryngeal nerve (SLN), and the recurrent laryngeal nerve (RLN). It is generally accepted that the SLN innervates the cricothyroid muscle (CT) whereas RLN innervates the remaining intrinsic laryngeal muscles (Sasaki, 2006; McHanwell, 2008). The cell bodies of the neurons that innervate the intrinsic laryngeal muscles are located within the nucleus ambiguus in the brainstem. This is a long rostrocaudally oriented column of neurons lying in the ventrolateral portion of the medulla oblongata, extending between the motor nucleus of the facial nerve to, at least, the level of the pyramidal decussation (Ramon y Cajal, 1909; Lawn, 1966b). In addition to innervating the muscles of the larynx, the nucleus ambiguus also contains neurons that provide motor innervation to the oesophagus, and the pharynx (Lawn, 1966a, 1966b; Bieger and Hopkins, 1987). It is generally accepted that laryngeal neurons are located in the caudal part of the nucleus ambiguus, the so-called loose and semicompact formations (Bieger and Hopkins, 1987) but some differences on the distribution of the neurons that innervate each individual intrinsic laryngeal muscle remain a matter of debate.
First, a number of studies describe laryngeal neurons as being located as a discrete single column of neurons extending along the rostrocaudal axis of the nucleus ambiguus in the brainstem (Szentagothai, 1943; Kalia and Mesulam, 1980; Hinrichsen and Ryan, 1981; Pásaro et al., 1981, 1983; Yoshida et al., 1982, 1984, 1985, 1998; Davis and Nail, 1984; Basterra et al., 1987; Okubo et al., 1987; Portillo and Pásaro, 1988; Van Daele and Cassell, 2009; Weissbrod et al., 2011; Pascual-Font et al., 2011). Other studies describe laryngeal neurons as being clustered along the nucleus ambiguus into several defined columns each columns being separated by gaps where no laryngeal neurons are present (Lawn, 1966a, 1966b; Gacek, 1975; Wetzel et al., 1980; Hisa et al., 1984; Bieger and Hopkins, 1987; Okubo et al., 1987; Patrickson et al., 1991; Núñez-Abades et al., 1992; Hirasugi et al., 2007). Moreover, while the majority of studies have focused on the rostrocaudal organization of laryngeal neurons in the nucleus ambiguus other authors have claimed patterns of dorsoventral (Gacek, 1975; Yoshida et al., 1985, 1998) or mediolateral (Flint et al., 1991) somatotopy within the nucleus ambiguus.
Second, some authors maintain that laryngeal neurons innervating individual laryngeal muscles are diffusely distributed along nucleus ambiguus while others maintain that a clear somatotopic organization can be seen. Most studies are in agreement in locating the neurons innervating the CT more rostral in the nucleus ambiguus than the remaining laryngeal muscles neurons (Szentagothai, 1943; Lawn, 1966a, 1966b; Hinrichsen and Ryan, 1981; Yoshida et al., 1982, 1984, 1985, 1998; Pásaro et al., 1983; Hisa et al., 1984; Davis and Nail, 1984; Basterra et al., 1987; Núñez-Abades et al., 1992; Hirasugi et al., 2007) in the semicompact formation of the nucleus ambiguus (Bieger and Hopkins, 1987). On the other hand, other published studies established that the neurons of the loose formation that innervate intrinsic laryngeal muscles of the larynx, but the CT, are intermingled within the nucleus ambiguus (Gacek, 1975; Hinrichsen and Ryan, 1981; Pásaro et al., 1983; Davis and Nail, 1984; Hisa et al., 1984; Basterra et al., 1987; Flint et al., 1991; Nahm et al., 1990, 1993).
The extent to which the somatotopic organization of laryngeal neurons in the nucleus ambiguus is conserved between species is also contested. Some authors maintain that the pattern is strongly conserved between species (Szentagothai, 1943; Yoshida et al., 1982, 1985, 1998; Okubo et al., 1987). However, because of the uncertainties in homology between laryngeal muscles in different species this conservation of somatotopic pattern is not always easy to establish. In the rat, for example, the subject of this study, the morphology of the larynx is different to other species. In the rat the arytenoid muscle is absent and instead its function is carried out by the superior cricoarytenoid muscle, a muscle found only in the rat (Kobler et al., 1994; Inagi et al., 1998). This is one of the differences that could explain the reported differences in somatotopic organization of laryngeal neurons in the rat (Portillo and Pásaro, 1988; Hirasugi et al., 2007).
From this brief review it can be seen that there have been a number of studies that have attempted to determine the somatotopic organization of the neurons innervating the intrinsic muscles of the larynx in a wide range of species. These studies have produced a number of discrepant results. However, not only the anatomical differences between species are the source of these discrepancies, a second source of problems is likely to be technical. A variety of different techniques and surgical protocols have been used to determine the location of laryngeal neurons within the nucleus ambiguus. The small size of the larynx, especially in some of the commonly used laboratory species such as rat, presents significant technical challenges to carrying out procedures reproducibly. This is particularly the case with studies that employ retrograde tracers to identify neurons where preventing tracer spread from the small intrinsic muscles is a particular problem.
Thus, the aim of our study is to examine the precise location and the number of the neurons innervating three intrinsic laryngeal muscles in the rat: CT, posterior cricoarytenoid (PCA) and thyroarytenoid (TA), using a less problematic retrograde tracer such as cholera toxin and a larger sample size to acquire statistically more significant data.
- Top of page
- MATERIAL AND METHODS
- LITERATURE CITED
Our results demonstrate that neurons that innervate the intrinsic laryngeal muscles CT, PCA, and TA, in the rat, are located in a continuous rostrocaudal column within the ipsilateral nucleus ambiguus, somatotopically arranged from the most rostral population, the CT neurons, to the most caudal population, the TA neurons. Each neuronal population has most of the half of their neurons clustered in a central region in the column, whereas the remaining ones are loosely distributed in the rostral and caudal poles, where small overlapping areas between adjacent neuron populations occur. This model is sustained by the fact that, in all the experiments, injection of CtB in the rat muscles CT, PCA and TA results exclusively in ipsilateral labelling of neuronal bodies within the nucleus ambiguus. Labelled neurons are clustered in populations corresponding to each injected muscle and rostrocaudally organized in columns extending over 1 mm in length, being the CT population the most rostral, followed caudalward by the PCA group, and, the most caudal, the set corresponding to the TA muscle. The constitution of each column is quite similar, with most of the half of the labelled cell bodies clustered in a rostrocaudal extension of 0.3–0.4 mm, whereas the remaining labelled somata are more loosely distributed within the rostral and caudal poles of the column. That arrangement gives continuity to the nucleus ambiguus column, since it allows that small territories of overlap, 0.3–0.1 mm in length, occurs between adjacent labelled neuronal populations. To our knowledge, this arrangement with most of the neurons grouped in a more compact core and the remaining ones scattered within the rostral and caudal poles has never been previously described.
The position of the labelled neuronal populations fits that observed for the projections described in the rat for the SLN and RLN (Pascual-Font et al., 2011; Weissbrod et al., 2011) and are in agreement with some previous studies that show that in the rat the neurons innervating the intrinsic laryngeal muscles constitute a continuous column with a rostrocaudal somatotopic organization (Hinrichsen and Ryan, 1981; Portillo and Pásaro, 1988; Van Daele and Cassell, 2009), as it has also been described in other animal models (Szentagothai, 1943; Yoshida et al., 1982, 1985; Davis and Nail, 1984; Okubo et al., 1987). This fact can be consistent with the double innervation of the PCA muscle, both by the superior and the recurrent laryngeal nerves, proposed by some authors (Bieger and Hopkins, 1987; Kobler et al., 1994; Furusawa et al., 1996; Hydman and Mattson, 2008). Furthermore, the pattern we have described, the dense core with the loose poles, can explain those descriptions of discontinuity between the neuronal populations of the CT and PCA muscles (Bieger and Hopkins, 1987; Hirasugi et al., 2007), or those for the PCA and TA muscles (Hirasugi et al., 2007), or even the opposite description of long overlapping territories for the PCA and TA muscles (Hinrichsen and Ryan, 1981; Flint et al., 1991; Van Daele and Cassell, 2009). In other animal models these discrepancies have also been depicted, thus a gap between the pools of neurons belonging to the CT and PCA muscles has been shown in the rabbit (Lawn, 1966a; Okubo et al., 1987; Kitamura et al., 1987), and the dog (Hisa et al., 1984; Hisa and Sato, 1987), as well as the lengthy overlap of the neuronal populations of the PCA and the TA muscles both in the cat (Yoshida et al., 1982, 1998; Pásaro et al., 1983; Davis and Nail, 1984) and the monkey (Yoshida et al., 1985, 1987). Far beyond the interspecies variations, these differences could also be due to the “dense core loose poles” arrangement of the neuronal populations. We have not observed, in any of the rats studied, other somatotopic map than the rostrocaudal, thus nor the mediolateral described also in the rat (Flint et al., 1991), nor the dorsoventral depicted in the kitten (Gacek, 1975). We speculate that the use of CtB versus HRP can influence the presence of these differences since in the dorsoventral somatotopic map described for the nucleus ambiguus of the kitten two discrete neuronal populations appears for each injected muscle (Gacek, 1975), which is probably due to uncontrolled diffusion of HRP to neighbor structures (Hinrichsen and Ryan, 1981; Bieger and Hopkins, 1987; Patrickson et al., 1991; Furusawa et al., 1996). Our experiments strongly support the ipsilateral origin of the efferent ambigual projections in the rat (Hinrichsen and Ryan, 1981; Bieger and Hopkins, 1987; Patrickson et al., 1991; Furusawa et al., 1996; Pascual-Font et al., 2006a, 2006b, 2011) but not the bilateral innervation (Wetzel et al., 1980; Lobera et al., 1981; Pásaro et al., 1981; Van Daele and Cassell, 2009), and, once again, the unrestrained spreading of the tracer can be the reason that produces artifactual bilateral tracings. On the other hand, CtB has the advantage of the low risk of spurious labelling by uncontrolled spreading of the tracer to tissues adjacent to the injection site. To confirm that spurious labelling was not a problem in our study, control experiments were carried out releasing tracer randomly over the larynx and the trachea; in these experiments no labelled perikarya could be observed in the brainstem. We have established 3 days as survival period following the injection of CtB because it is an optimal time to a retrogradely labelling of neural soma (Vercelli et al., 2000). However, in previous laryngeal muscle CtB tracing experiments, animals were allowed to survive longer times, 5 days, but no differences were observed.
To locate the labelled neurons we have followed the definition of the obex given by Hamilton and Norgren (1984), meanwhile other authors refer to the obex as the level where central canal opens into the fourth ventricle (Portillo and Pásaro, 1988; Hirasugi et al., 2007); as this landmark is 0.8 mm rostral to the true obex in the rat, this difference is carried to the measured positions of the labelled neurons along the nucleus ambiguus. In addition to the later methodological consideration, we have developed and applied a correction factor for the obliquity of the histological sections, a useful tool which allows gaining in precision in the location of the laryngeal neurons as we have observed in the reduction in the standard deviation appreciated after applying the correction.
Labelled neurons are mainly stellate in shape but, for the CT and PCA populations, some fusiform somata were observed. The largest labelled neurons correspond to the TA muscle, followed in size by those for the CT muscle, and being the smallest those for the PCA muscle. It is generally accepted that there are differences in somata size within the nucleus ambiguus (Lawn, 1966a, 1966b; Hinrichsen and Ryan, 1981; Pásaro et al., 1983; Portillo and Pásaro, 1988; Patrickson et al., 1991; Saxon et al., 1996; Hayakawa et al., 1999) and most of the authors agree that the TA neurons are the largest and the PCA neurons are the smallest (Davis and Nail, 1984; Portillo and Pásaro, 1988; Hirasugi et al., 2007). This variability in size has been related with the roles played by these neurons in laryngeal physiology, thus the larger size of the TA neurons is associated to the rapid adductor activity during breathing cycle, meanwhile the slower speed of contraction of the CT and PCA muscles could be linked to the smaller size of their innervating neurons (Hinrichsen and Ryan, 1981; Davis and Nail, 1984; Patrickson et al., 1991; Yoshida et al., 1998). Hayakawa et al. (1999) identified two types of neurons innervating the PCA muscle; the largest ones probably receiving excitatory inputs and the small ones receiving inhibitory inputs. We have not identified these two different sizes in the neurons labelled from the PCA muscle.
The main limitation in our study is due to the technical approach. It is not possible to guarantee that all the synaptic terminals innervating one muscle can uptake the tracer, and this means that surely some neurons are not labelled and, hence, not taken into account when the total number of neurons innervating one muscle was measured. When CtB is injected into one muscle it bounds to GM1 ganglioside embedded in the lipid matrix of the synaptic membrane, later incorporated inside the neuron, and finally retrogradely transported to the soma (Vercelli et al., 2000; Conte et al., 2009). Although we cannot expect that in all the terminals these processes will be fulfilled, the consistency in the measurements of the number of labelled neurons along the processed animals make us confident that the quantity of identified cells is closer to the real number of neurons innervating one muscle. In addition, the number of identified neurons for the PCA and the TA muscles fit well with the number of myelinated axons found in the recurrent laryngeal nerve of the rat (Pascual-Font et al., 2006c), and with the data from other quantitative studies in the rat (Hinrichsen and Ryan, 1981; Portillo and Pásaro, 1988). On the other hand, and regarding to the morphometrical analysis, the sizes obtained in the present work, performed by fluorescent tracers, are in the range of those obtained in previous morphometrical analysis developed in bright-field microscopy (Wetzel et al., 1980; Hinrichsen and Ryan, 1981; Bieger and Hopkins, 1987; Portillo and Pásaro, 1988; Patrickson et al., 1991).
In conclusion, our report precisely locates the neurons innervating significant intrinsic laryngeal muscles and, additionally, establishes a well defined model that permits to develop studies on the reorganization of the laryngeal somatotopy both when the laryngeal nerves are injured and allowed to reinnervate the larynx.