THE ANATOMICAL RECORD 230551-556 (1991) Motor Neurons of the Laryngeal Nerves J.W. PATRICKSON, T.E. SMITH, AND S.-S.ZHOU Department of Anatomy, Loma Linda University, Loma Linda, California (J.W.P., T.E.S.); Department of Anatomy, Suzhou Medical College, Suzhou, Jiangsu, People’s Republic of China (S.-S.Z.) ABSTRACT The present study was undertaken to determine the relationship between the motor neurons of the superior and recurrent laryngeal nerves within the nucleus ambiguus. The retrograde transport of horseradish peroxidase was utilized to identify the motor neurons subsequent to its application to the proximal transected end of the superior and recurrent laryngeal nerves. Labeled superior laryngeal motor neurons were distributed ventrolaterally in the rostral portion of the nucleus. The recurrent laryngeal motor neurons were distributed throughout the nucleus with two distinct populations: a rostral group and a caudal group. The rostral group overlaps the motor neurons of the superior laryngeal nerve. The caudal group occupies that portion of the nucleus that is classically described for the recurrent laryngeal nerve. Additional superior laryngeal nerve labeled perikarya were found in the dorsal motor nucleus of the vagus. This study defines the rostral distribution of the recurrent laryngeal nerve motor neurons and suggests that this rostral group is a component of the neuroanatomical substrate that is involved in the co-activation of the laryngeal abductors controlling the laryngeal aperture. The articulation of the larynx by the intrinsic muscles of the larynx plays a n important role in the modulation of respiration. This is achieved by changing the aperture of the airway and thus the resistance to air flow. Of the laryngeal muscles, the posterior cricoarytenoid and the cricothyroid muscles are the most dominant in this function. The contraction of the posterior cricoarytenoid muscles results in the abduction of the vocal cords, increasing the glottic chink. The cricothyroid muscles when contracted tense the vocal cords and increase the anteroposterior diameter of the laryngeal aperture (for review, see Sasaki and Isaacson, 1988). During the inspiratory phase of respiration, there is first a n increase of the laryngeal aperture primarily by the co-activation of the posterior cricoarytenoid and the cricothyroid muscles followed by the activation of the diaphragm. The cricothyroid muscle is innervated by the superior laryngeal nerve (SLN) while the posterior cricoarytenoid and the remaining intrinsic laryngeal muscles (thyroarytenoid, lateral cricoarytenoid, and the interarytenoid) are innervated by the recurrent laryngeal nerve (RLN). The motor neurons innervating these muscles are somatotopically distributed in the nucleus ambiguus (Gacek, 1975; Kalia and Mesulam, 1980a,b; Hinrichsen and Ryan, 1981; Yoshida et al., 1982; Pasaro e t al., 1983; Bieger and Hopkins, 1987). Of the two nerves, SLN and RLN, the motor neurons of the SLN (cricothyroid) are the most rostrally placed. The motor neuron pool of the RLN, a s one would expect based on the number of muscles supplied, is larger and more complex than that of the SLN. The motor neurons of the posterior cricoarytenoid are the most rostrally placed. The remaining motor neurons of the intrinsic laryngeal muscles are distributed progressively caudal 0 1991 WILEY-LISS. INC within the nucleus with extensive overlapping (Hinrichsen and Ryan, 1981; Hisa et al., 1984; Bieger and Hopkins, 1987; Okubo et al., 1987). However, there are conflicting reports regarding the distribution of the SLN versus the RLN motor neuron pools. The cricothyroid motor neurons are consistently reported as being localized in the rostral ambiguus. The posterior cricoarytenoid motor neurons, on the other hand, are described by some investigators as extending caudally from the middle of the nucleus (Bieger and Hopkins, 1987; Hisa et al., 1984; Yoshida e t al., 1982) while others localize these neurons a s extending caudally from the most rostral limits of the nucleus (Hinrichsen and Ryan, 1981; Gacek, 1975). Recently, Yajima and Hayashi (1989) recorded evoked antidromic potentials from motor neurons within the nucleus ambiguus in response to both superior and recurrent laryngeal nerve stimulation. This observation would suggest a unique relationship must exist between the motor neurons of the recurrent and superior laryngeal nerves. Based on these divergent opinions, the present study was undertaken to examine the distribution of, and the relationship between, the SLN and RLN motor neuron pools. MATERIALS AND METHODS Experiments were performed on 29 Sprague Dawley rats weighing 200-310 g. The animals were anesthe- Received July 16, 1990; accepted January 10, 1991. Address reprint requests to J.W. Patrickson, Ph.D., Dept. of Anatomy, Lorna Linda University, Lorna Linda, CA 92350. 552 J.W. PATRICKSON ET AL. tized with a 12% aqueous solution of chloral hydrate (0.3 mlilOO g body weight i.p.). Surgical Procedure A ventral cervical midline incision was made, and the larynx and trachea were exposed by blunt dissection. With the aid of a dissecting microscope, the laryngeal nerves were dissected free from the surrounding tissue using hand-pulled glass micro-probes. For optimum tracer uptake, it is imperative that the transected nerve be free of blood. The vascular tissue, where possible, was therefore dissected away from the nerves prior to transection. Care was taken to prevent trauma to the nerves during these manipulations. At a point immediately proximal to the larynx, yet before any branching may have occurred, the nerve was obliquely transected. HRP Application Wheat germ agglutin-horseradish peroxidase (WGAHRP) was applied to the transected nerve by one of two methods, each yielding similar results. In ten animals, the nerve was isolated from the surrounding tissue by packing the area with gel-foam. Dry crystals of WGAHRP were applied directly to the transected nerve for 30 minutes. At the end of the allotted time, the excess tracer was sponged from the nerve and the area flushed with saline. In 13 animals, the isolated nerve was inserted into a small-diameter plastic tubing and the proximal end sealed with Gi-Mask (Coltene Inc.), a non-toxic dental impression material. The tube was then filled with a 20% WGA-HRP saline solution and sealed. This technique allows for the continuous bathing of the transected nerve stump for the allotted transport time of 48 hours. In most cases, the procedure was applied to one nerve per animal. In six animals, the tracer was applied to both nerves using the tube method, the SLN on one side and the RLN on the other. The wound was closed with stainless-steel wound clips. Tissue Preparation and Processing Forty-eight hours post HRP application, the animal was deeply anesthetized and perfused through the aorta with 250 ml of saline at room temperature (25°C). This was followed with 500 ml of fixative consisting of 2% paraformaldehyde and 0.75% glutaraldehyde in 0.1 M phosphate buffer (pH 7.4) at 25°C. For cryoprotection, the fixative was followed with 250 ml 10% sucrose in 0.1 M phosphate buffer (pH 7.4) a t 4°C. The brain was removed and placed in sucrose-buffer solution a t 4°C for approximately 24 hours. Frozen 30 pm serial sections in the coronal, sagittal, or horizontal planes were made. The tissues were incubated using tetramethylbenzidine (TMB) a s the chromagen following the protocol of Mesulam (1978) with minor modifications. The sections were mounted on gelatin-coated slides, air dried, dehydrated, and cover slipped. The specimens were examined under light- and darkfield illumination for the presence of HRP positive neurons. RESULTS Both methods of exposing the central end of the transected laryngeal nerves to WGA-HRP resulted in a n intense labeling of the motor neurons and presented a near Golgi-type staining, thus allowing for the visu- alization of the cell processes. There were two cell types based on their size and processes: First, there were small cells of approximately 21 pm in diameter with short and thin dendritic processes that were observed consistently in the rostral area of nucleus ambiguus. The other were larger (approximately 35 pm in diameter) multipolar cells with rather large dendritic processes. These cells were distributed along the entire length of the nucleus ambiguus. Exposure of the superior laryngeal nerve to the neuronal tracer resulted in HRP positive neurons in the ipsilateral rostral nucleus ambiguus (Fig. 1A). These neurons were distributed in the ventrolateral region of the nucleus. The labeled area was found to extend for a n average of 430 pm rostrally from a point 420 pm rostral to obex. No labeled cells were found caudal to this area. The cells (average 71 in number) consisted primarily of small spindle-shaped and small (but more typical looking) multipolar motor neurons. The spindle cells were always dorsomedial to the latter, the majority of which were found to occupy the most rostral portion of the nucleus. In addition to the nucleus ambiguus, labeled neurons were identified within the dorsal motor nucleus of the vagus (Fig. 1B). These vagal neurons (average 69 in number) were found a t a level 338 pm rostral to obex and to extend rostrally for approximately 454 pm. Labeled neurons of the recurrent laryngeal nerve were identified over a much wider area of the ipsilatera1 nucleus ambiguus. Two distinct populations of neurons were identified within the nucleus, a rostral and caudal group. The rostral labeling was found ventrolaterally within the nucleus and extended for a n average of 260 pm rostrally from a point 525 pm rostral to obex and consisted primarily of the small spindle-shaped cells (Fig. 2C). The caudal group of cells extended for a n average of 740 pm and were distributed rostrally from a point 409 pm caudal to obex or a t the level of pyramidal decussation. The neurons of this group consisted primarily of the larger multipolar-type motor neurons (Figs. 2A,B). Labeling of the dorsal motor nucleus of the vagus via the recurrent laryngeal nerve was found to be virtually nonexistent, with a n average of three labeled cells per animal. The dimensions of the motor neuron pool of the recurrent laryngeal nerve within nucleus ambiguus reveal two interesting phenomena: First, there is a gap in the labeling of recurrent laryngeal motor neurons within the nucleus ambiguus, thereby creating a rostral and caudal group. This gap (void in labeling) was found to extend rostrally for approximately 200 pm from a level 330 pm rostral to obex. Second, the dimensions further reveal a n area of overlap in the ventrolateral pole of nucleus ambiguus between the two laryngeal nerves. To recapitulate, the motor neurons contributing to the superior laryngeal nerves began 420 pm rostral to obex and continued rostrally for 430 pm. The neurons contributing to the recurrent laryngeal nerve were located 525 pm rostral to obex and continued rostrally for 260 pm. These data suggest a n area of overlap between the neuron pools of the superior and recurrent laryngeal nerves spanning 260 pm. To further demonstrate this gap and overlap, horizontal sections were made following the bilateral exposure of laryngeal nerves alternatively and contralateral to MOTORNEURONSOFTHELARYNGEALNERVES 553 Fig. 1. Brightfield photomicrographs taken in the coronal plane demonstrating HRP labeled motor neurons of the superior laryngeal nerve. A: Ventrolateral region of rostral nucleus ambiguus. Bar = 40 pm. B: Dorsal motor nucleus of vagus. Dorsal = top; medial = left; Bar = 30 pm. each other (i.e., the recurrent on the right and the superior on the left). This arrangement provided for a more precise method of comparison (Fig. 31, which confirmed our previous observations. DISCUSSION In the Results we described a n area within the nucleus ambiguus that is common to both the superior and recurrent laryngeal nerves. The methodology of whole nerve exposure to the tracer material utilized in this study maximizes the possibility of identifying the total motor neuron population and the distribution for each laryngeal nerve. The superior laryngeal nerve in addition to its efferent projection to the cricothyroid muscle contains a small component to the inferior constrictor muscles of the pharynx and to the proximal muscularis externa of the esophagus (Bieger and Hopkins, 1987), the motor neurons of which are distributed with those of the cricothyroid. The location of the superior laryngeal nerve motor neurons described in this study are similar to that described for the cricothyroid in the rabbit (Okubo et al., 1987); cat (Gacek, 1975; Davis and Nail, 1984; Pasaro e t al., 1983); dog (Hisa et al., 1984); and rat (Hinrichsen and Ryan, 1981; Bieger and Hopkins, 1987), with minor species differences. These neurons are located in the ventrolateral region of the nucleus. The motor neurons of the recurrent laryngeal nerve in its most rostral distribution are within the ventrolateral region of nucleus ambiguus overlapping those of the superior laryngeal nerve. These neurons are seen as a discrete population of the RLN that is distinct from those classically described. The distribution of these more rostrally placed recurrent laryngeal nerve motor neurons using neuronal tracers was first described by Gacek (1975) in the kitten and later by Hinrichsen and Ryan (1981) in the rat. Subsequent to these studies, other reports on the laryngeal motor neuron failed to describe this rostral population of cells (Hisa e t al., 1984; Bieger and Hopkins, 1987; Yoshida et al., 1982). The uniqueness of this study is the exposure of each laryngeal nerve to the tracer contralateral to one another and visualizing the labeled neurons in the horizontal plane. In so doing, the distribution of the neurons for each nerve could be compared in the same animal. No attempts were made in this study to identify the somatotopic organization of nucleus ambiguus in reference to each laryngeal muscle since numerous studies have been conducted on this subject. In brief, the cricothyroid motor neurons (via the superior laryngeal nerve) are most rostral, followed by the posterior cricoarytenoid and the remaining intrinsic laryngeal motor neurons, the thyroarytenoid, lateral cricoarytenoid, and interarytenoid more caudally and overlapping each other (Hinrichsen and Ryan, 1981; Bieger and Hopkins, 1987). Somatotopically, the rostral group of recurrent laryngeal motor neurons is a component of the posterior cricoarytenoid motor neuron pool, the larger population of which is within the more rostral portion of the caudal group (Hinrichsen and Ryan, 1981). Because of the somatotopic organization of the cricothyroid and the posterior cricoarytenoid motor neurons relative to superior and recurrent laryngeal nerves, the whole nerve preparation was utilized in this study in lieu of the individual muscle injections which are always at risk to the undesirable spread of tracer material to neighboring structures. The recent electrophysiological study on the motor neurons of the recurrent and superior laryngeal nerve by Yajima and Hayashi (1989) reported on neurons whose axons traverse both laryngeal nerves to innervate the intrinsic muscles of the larynx. They have suggested that these axons branch midway between 554 J.W. PATRICKSON E T AL. Fig. 2. A series of brightfield photomicrographs taken in the coronal plane demonstrating recurrent laryngeal motor neurons and their arrangement a t various levels within the nucleus ambiguus. A. Caudalmost region of the nucleus. B: At the level of obex. C: Ventrolateral area of the rostral area of the nucleus. Motor neurons in panels A and B are from the caudal group, while those of panel C are from the rostral group. Dorsal = top; medial = left; bar = 40 pm. cated ventrolaterally within the nucleus ambiguus. The remaining intrinsic muscles of the larynx are adductors. The motor neuron pools of the adductors are located dorsally and are more caudally placed with extensive overlapping of the motor neuron pool for each muscle (Hisa et al., 1984; Okubo et al., 1987; Hinrichsen and Ryan, 1981, Gacek, 1975). Motor neuron size may be of functional significance relative to its effector organ. The intrinsic laryngeal muscles are fast contracting (Syrovy and Gutmann, 1971) and contain no muscle spindles (Raman and Devanandan, 1989). Similar to most skeletal muscles, there are two types of muscle fibers found in the laryngeal muscles, one with low actomyosin ATPase activity (type I) and the other with high actomyosin ATPase activity (type 11) (Rosenfield et al., 1982; Sahgal and Hast, 1974). Type I muscle fibers are slow contracting and fatigue resistant. In contrast, the type I1 fibers are the soma and the effector. The functional significance fast contracting and fatigue rapidly. Type I fibers are of these laryngeal motor neurons are yet to be deter- innervated by small motor neurons. These fibers are mined; however, this neuronal arrangement along easily recruited (low threshold), control fine movewith those of the presently described overlapping ros- ments, and are capable of prolonged, slow rhythmic tral group may represent a component of the efferent discharge. The fiber ratio per neuron is small. The type system that is involved in the coordination and co-ac- I1 muscle fibers discharge a t higher thresholds, have tivation of the respective laryngeal muscles, the abduc- high discharge rates, and develop more tension but fatigue faster. These fibers are innervated by large motor tors, during the respiratory cycle. In support of this concept, the overlapping of the mo- neurons with a high fiber ratio per neuron (Warmolts, tor neuron populations to the cricothyroid and the pos- 1981; Burke et al., 1982; Burke et al., 1973; Burke and terior cricoarytenoid muscles appears to be a functional Tsairis, 1974). grouping of the motor neurons relative to that of its Although the laryngeal muscles contain both fiber effector. These muscles are abductors; when co- types, there are more type I fibers in the abductors and activated, they enlarge the aperture of the laryngeal a preponderance of type I1 in the adductors (Teig et al., airway. The motor neurons of these abductors are lo- 1978).Functionally, the fast action of the adductors are MOTOR NEURONS OF TH ELARYNGEALNERVES OBEX 555 to the laryngeal area. This visceral component is not visualized when the muscles of the larynx are exposed to neuronal tracers (Hinrichsen and Ryan, 19811, but is, rather, with injections into the laryngeopharyngeal areas including the mucosa or, a s in this study, with the exposure of the proximal end of the transected nerves. Contrary to the findings of Kalia and Mesulam (1980a,b), the present results show a substantial contribution of the dorsal motor nucleus of the vagus to the larynx, primarily via the superior laryngeal nerve. This visceromotor component is unilateral. This observation is supported by similar observations in the monkey (Yoshida et al., 1982) and guinea pig (Basterra e t al., 1988). Kalia and Mesulam (1980a,b) reported bilateral innervation to the larynx of the cat. This apparent contradiction, if not due to species differences, is resolved in that relatively large volumes of neuronal tracer was injected into the laryngeal mucosa; thus, the possibility of diffusion across the midline. If this is the case, then the results would be mistaken as a bilateral projection. To conclude, the present study has affirmed the rostral placement of the superior laryngeal nerve motor neurons within the ventrolateral subdivision of nucleus ambiguus. There are two distinct populations of recurrent laryngeal nerve motor neurons within nucleus ambiguus: a rostral and a caudal. The most rostral component overlaps those of the superior laryngeal nerve and may represent a component of the neuroanatomical substrate for the co-activation of the laryngeal abductors. LITERATURE CITED Fig. 3. A projected tracing of the lower brain stem and spinal cord in the horizontal plane illustrating labeled laryngeal motor neurons (stippled area) within nucleus ambiguus following the exposure of the laryngeal nerves (superior, left; recurrent, right) to HRP in the same animal. Arrow indicates gap between the rostral and caudal groups of the recurrent laryngeal nerve motor neurons. lfp: longitudinal fasciculus pons; ml: medial lemniscus; mlf: medial longitudinal fasciculus; m5: motor trigeminal nucleus; s5: sensory root trigeminal nerve. Bar = 1 mm. essential for the protection of the airway. The cricothyroid muscle tenses the vocal cord and requires prolonged contraction. Both the posterior cricoarytenoid and the cricothyroid muscles are tonically active during the inspiratory phase of respiration (Woodson et al., 1989; Horiuchi and Sasaki, 1978; Konrad and Rattenborg, 1969). This rhythmic co-activation of these muscles is best suited to a larger ratio of type I fibers. The neurons of overlap are the small type which also innervate the type I fibers of the abductors. This arrangement of the overlapping neurons may represent the neuroanatomical substrate for the co-activation of the abductors during eupnea. The presence of labeled neurons in the dorsal motor nucleus of the vagus via the laryngeal nerves demonstrates that these nerves provide visceral innervation Basterra, J., C.C. Chumbley, and P.N. Dilly 1988 The superior laryngeal nerve: Its projection to the dorsal motor nucleus of the vagus in the guinea pig. Laryngoscope, 98539-92. Bieger, D., and D.A. Hopkins 1987 Viscerotropic representation of the upper alimentary tract in the medulla oblongata in the rat: The nucleus ambiguus. J. Comp. Neurol., 262546-562. Burke, R.E., and P. Tsairis 1974 The correlation of physiological properties with histochemical characteristics in single motor units. N.Y. Acad. Sci., 228:145-159. Burke, R.E., D.N. Levine, P. Tsairis, and F.E. Zasac 1973 Physiological types of histochemical profiles in motor units of the cat gastrocnemius. J. Physiol. (Lond.), 234:723-748. Burke, R.E., R.P. Dum, J.W. Fleshman, L.L. Glenn, A. Lev-Tov, M.J. ODonovan, and M.J. Pinter 1982 An HRP study of the relation between cell size and motor unit type in cat ankle extensor motoneurons. J. Comp. Neural., 209t17-28. Davis, P.J., and B.S. Nail 1984 On the location and size of laryngeal motor neurons in the cat and rabbit. J . Comp. Neurol., 230t1332. Gacek, R.R. 1975 Localization of laryngeal motor neurons in the kitten. Laryngoscope, 85:1841-1 861. Hinrichsen, C.F.L., and A.T. Ryan 1981 Localization of laryngeal motor neurons in the rat: Morphologic evidence for dual innervation? Exp. Neurol., 74:341-355. Hisa, Y., F. Sato, K. Fukui, Y. Ibata, and 0. Mizuokoshi 1984 Nucleus ambiguus motor neurons innverating the canine intrinsic laryngeal muscle by the fluorescent labelling technique. Exp. Neurol., 84:441-449. Horiuchi, M., and C.T. Sasaki 1978 Cricothyroid muscle in respiration. Ann. Otol., 87:386-391. Kalia, M., and M.M. Mesulam 1980a Brain stem projections of sensory and motor components of the vagus complex in the cat: I. The cervical vagus and nodose ganglion. J. Comp. Neural., 193,435465. Kalia, M., and M.M. Mesulam 1980b Brain stem projections of sensory and motor components of the vagus complex in the cat: 11. Laryngeal, trachea-bronchial, pulmonary, cardiac, and gastrointestinal branches. J. Comp. Neural., 193,467-508. 556 J.W. PATRICKSON ET AL. Konrad, H.R., and C.C. Rattenborg 1969 Combined action of laryngeal muscles. Acta. Otolaryngol. (Stockh),67546-649. Mesulam, M.M. 1978 Tetramethylbenzidine for horseradish peroxidase neurohistochemistry: Non-carcinogenic blue reaction product with superior sensitivity for visualizing neural afferents and efferents. J. Histochem. Cytochem., 26t106-117. Okubo, J., S. Kitamura, K. Ogata, and A. Sakai 1987 Localization of rabbit laryngeal motor neurons in the nucleus ambiguus. Exp. Neurol., 96t528-539. Pasaro, R., B. Lobera, S. Gonzalez-Baron, and J.M. Delgado-Garcia 1983 Cytoarchitectonic organization of laryngeal motor neurons within the nucleus ambiguus of the cat. Exp. Neurol., 82,623-634. Raman, R., and M.S. Devanandan 1989 Muscle receptors: Content of some of the extrinsic muscles of the larynx in the bonnet monkey (Macaca radiatai. Anat. Rec., 223t433-436. Rosenfield, D.B., R.H. Miller, R.B. Sessions, and B.M. Patten 1982 Morphological and histochemical characteristics of laryngeal muscle. Arch. Otolaryngol., 108:662-666. Sahgal, V., and M.H. Hast 1974 Histochemistry of primate laryngeal muscles. Acta. Otolaryngol., 78:277-281. Sasaki, C.T., and G. Isaacson 1988 Functional anatomy of the larynx. Otolaryngol. Clin. North Am., 21t595-612. Syrovy, I., and E. Gutmann 1971 ATPase activity of two laryngeal muscles. Experienta, 27t248. Teig, E., H.A. Dahl, and H. Thorkelsen 1978 Actomyosin ATPase activity of human laryngeal muscles. Acta. Otolaryngol., 85:272281. Warmolts, J.R. 1981 Electrodiagnosis in neuromuscular disorders. Ann. Intern. Med., 95:599-608. Woodson, G.E., 0. Mathew, F. Sant’Ambrogio, and G. Sant’Ambrogio 1989 Effects of cricothyroid muscle contraction on laryngeal resistance and glottic area. Ann. Otol. Rhinol. Laryngol., 98:119124. Yajima, Y., and Y. Hayashi 1989 Electrophysiological evidence for axonal branching of ambiguous laryngeal motor neurons. Brain Res., 478:309-314. Yoshida, Y., T. Miyazaki, M. Hirano, T. Shin, and T. Kanaseki 1982 Arrangement of motor neurons innverating the intrinsic laryngeal muscles of the cat as demonstrated by horseradish peroxidase. Acta. Otolaryngol., 94t329-334.