- Email: [email protected]
Motor and sensory centers for the innervation of mandibular and sublingual salivary glands: A horseradish peroxidase study in the dog
Brain Research, 189 (1980) 301-313 © Elsevier/North-Holland Biomedical Press
301
Research Reports
M O T O R A N D S E N S O R Y C E N T E R S F O R T H E I N N E R V A T I O N OF M A N D I BULAR AND SUBLINGUAL SALIVARY GLANDS: A HORSERADISH PEROXIDASE STUDY IN THE DOG
GREGORY A. CHIBUZO* and JOHN F. CUMMINGS Department of Anatomy, N YS College of Veterinary Medicine, Cornell University, Ithaca, N. Y. 14853 (U.S.A.)
(Accepted October 1lth, 1979)
Key words: salivary gland innervation - - horseradish peroxidase - - dog - - cranial nerves - - reticular
formation - - brain stem
SUMMARY Horseradish peroxidase was injected at multiple sites in the mandibular and sublingual salivary glands in order to label the preganglionic salivatory neurons in the brain stem. The same injections resulted in retrograde labeling of the sympathetic and sensory neurons that project to these glands. Labeled fusiform and multipolar salivatory neurons were found ipsilaterally in the lateral reticular formation of the medulla where they extended over the rostral four-fifths of the facial nucleus and the caudal one-third of the dorsal nucleus of the trapezoid body. The vast majority of the small and medium-sized, labeled neurons appeared in the nucleus reticularis parvocellularis. Labeled neurons also appeared occasionally at the ventral and lateral aspects of the facial nucleus. Enzyme injections into these glands labeled sympathetic neurons that were concentrated in the caudal one-third of the ipsilateral cranial cervical ganglion. Labeled sensory neurons were distributed randomly in the ipsilateral proximal vagal and geniculate ganglia. Large numbers of sensory neurons were concentrated ventromedially within the mandibular zone of the trigeminal ganglion.
* Address for all correspondence: Department of Anatomy, School of Veterinary Medicine, Tuskegee Institute, Ala. 36088, U.S.A.
302 INTRODUCTION Previous studies of the innervation of the major salivary glands have focused mainly on the preganglionic parasympathetic neurons of the brain stem, i.e. the salivatory nucleus or nuclei. Anatomic studies of these neurons have been based largely upon retrograde cell change induced by interruption of preganglionic axons at various peripheral levels. This technique has led to divergent results regarding the size and location of the somata and the course of their preganglionic axons14,~3,26,29. Neurophysiologic studies on the salivatory center have relied upon the identification of medullary stimulation sites that evoke salivary gland secretion. It has been difficult to draw firm conclusions on the location of the preganglionic neurons from these studies since it was often problematical as to whether cell bodies or fibers were activated or whether stimulated sites were afferent or efferent5A9,27. In employing the horseradish peroxidase (HRP) retrograde labeling technique, the present study eliminated the interpretation problems that were encountered in earlier neuroanatomic and physiologic experiments. Since many of the preganglionic neurons supplying the mandibular and sublingual glands terminate in ganglia associated with these gland masses 18, diffuse injections of the glands consistently and specifically labeled the medullary parasympathetic neurons. These intragtandular injections also labeled sympathetic and sensory neurons serving these glands. MATERIALS AND METHODS Nine neonatal dogs (2-6 weeks of age) were available for the experiment. Each received an intramuscular injection of promazine hydrochloride (2-6 mg/kg) and/or a subcutaneous injection of atropine sulfate (0.04 mg/kg) preanesthetically. After 10-15 rain, fentanyl citrate-droperidol (0.2-0.4 ml/kg) (Innovar Vet., Pitman-Moore, Inc., Washington Crossing, N. J. 08560) was injected intraperitoneally, and after a 10-min delay, this was followed by pentobarbital sodium (4-8 mg/kg) administered intravenously. Supplemental doses of pentobarbital sodium for extension of analgesia beyond 35 rain were given to effect. The preceding anesthetic regime was modified from schedules developed for use in adult dogs 9,1zAzAT,24. After the animals were anesthetized, the left mandibular (submaxillary) and/or sublingual salivary glands were exposed surgically. Twenty to thirty microliters of a 30 ~ aqueous solution of HRP (Sigma type IV) were injected at multiple sites in the mandibular salivary gland in 3 dogs, in the sublingual salivary gland in 3 dogs, and in both glands in 3 dogs. The incisions were closed and the animals were placed in a warm recovery room. At 24 or 48 h postoperatively, the animals received an intravenous overdose of pentobarbital sodium and were then perfused via the left cardiac ventricle with a solution containing 0.5 ~o paraformaldehyde, 2.5 ~ glutaraldehyde, and 2.5 ~ sucrose in 0.1 M phosphate buffer (pH 7.4). The brain was removed and stored in chilled fixative for 24 h. The trigeminal, geniculate, glossopharyngeal, vagal, otic and cranial cervical ganglia were removed bilaterally and also immersed in refrigerated fixative for 24 h. The tissues were then stored in refrigerated 0.1 M
303 phosphate buffer (pH 7.4) prepared in a 30 ~o sucrose solution. Serial frozen sections of the brain stem and ganglia were cut at 60 #m. All sections were incubated with 3,3'diaminobenzidine and hydrogen peroxide, counterstained and mounted according to the method of Nauta et al. 21. Labeled neurons were detected by bright-field or darkfield microscopy. RESULTS
Preganglionic parasympathetic salivatory neurons Injections of horseradish peroxidase (HRP) into the mandibular and/or sublingual salivary glands consistently labeled nerve cell bodies in the lateral medullary reticular formation. The labeled multipolar somata were confined to the side of the injection. The cell bodies were small or medium sized and presented fusiform or polygonal outlines (Figs. 1 and 2). They ranged in size from 28 x 19 # m to 55 x 17 /~m, and most contained discrete Nissl granules. The great bulk of the labeled population was situated dorsal to the rostral four-fifths of the facial nucleus and continued rostrally over the caudal one-third of the dorsal nucleus of the trapezoid body (Fig. 3). The descending root o f the facial nerve marked the rostral limit of the labeled neurons. Occasionally, labeled neurons appeared at the medial or ventral aspect of the facial nucleus (Fig. 3). These were rare, however, and most of the labeled cells were
Fig. 1. Labeled small and medium-sized multipolar somata in the salivatory center. The few medium sized neurons are indicated by arrows. Dark-field, 200 x.
304
Fig. 2. Isolated field with two labeled medium sized somata (arrows) among smaller neurons in the medullary reticular formation ventral to medial vestibular nucleus. Dark-field, 400 .:. scattered dorsal to this nucleus in an area that was bounded laterally by the nucleus of the spinal tract of the trigeminal nerve, medially by the nucleus reticularis gigantocellularis, and dorsally by the descending, medial and lateral vestibular nuclei (Fig. 3). Within these boundaries, most of the labeled cells were small. Medium-sized neurons were less frequent and appeared usually in the region immediately ventral to the medial vestibular nucleus. The majority of the labeled neurons resided within the nucleus reticularis parvocellularis. Neurons labeling after mandibular gland injections were coextensive with those that labeled after sublingual gland injections. Mandibular gland injections, however, resulted in greater numbers of labeled neurons. Although injections into both salivary glands produced the greatest number of labeled cells, the distribution of the neurons in these cases conformed to that observed after injection of a single gland.
Autonomic ganglia Numerous labeled neurons were observed in the ipsilateral cranial cervical ganglion following mandibular and/or sublingual salivary gland H R P injections (Fig. 4). These labeled sympathetic neurons were concentrated at the caudal one-third of this ganglion (Fig. 5). Labeled cell bodies were not observed in otic ganglia after sublingual and/or mandibular gland injections.
Sensory ganglia Labeled neurons were observed in the ipsilateral trigeminal ganglion after gland
305 injections (Fig. 6). Mandibular gland injections produced greater numbers of labeled neurons than did sublingual gland injections, and the most extensive labeling occurred in those cases in which both glands were injected. In all animals, the labeled somata appeared in the lateral or mandibular zone of the ganglion, and most were localized ventromedially within the mandibular zone (Fig. 7). Labeled neurons were also present in the ipsilateral geniculate and proximal vagal (jugular) ganglia (Fig. 8). In both ganglia a small proportion of the cells labeled and there was no detectable somatotopic organization.
e~ YP~ , ~ / " / C
B
AUDAL END
"
E
. .:-.;'. .:.,.
.©
H~
"
ROSTRALEND
Fig. 3. Brain stem sections arranged in a caudal (A) to rostral (I) sequence to show distribution of labeled salivatory neurons (solid dots) in the medullary reticular formation. Abbreviations: NVC, caudal vestibular nucleus; NVM, medial vestibular nucleus; NVL, lateral vestibular nucleus; R VIII, vestibulocochlear nerve; NRV, nucleus and spinal tract of trigeminal nerve; N VII, facial nucleus; PY, pyramid; N VIII, cochlear nucleus; NTD, dorsal nucleus of trapezoid body (superior olive); G, genu of facial nerve; NVI, abdueens nucleus; R VII, descending root of facial nerve; TB, trapezoid body.
306
Fig. 4. Labeled neuronal s o m a t a in the cranial cervical ganglion. Dark-field, 367
.
Proximal
Laterar
Medial
Distal
g
y
p
thetic trunk
Fig. 5. A chart of labeled neurons (solid dots) in a horizontal section of the cranial (superior) ccrvica! ganglion following H R P injection of mandibular and sublingual salivary glands. Labeled cell bodies were localized towards the caudal end of this ganglion.
307
Fig. 6. Labeled neuronal cell bodies from the mandibular or lateral division of trigeminal ganglion. Dark-field, 375 ×.
ateral
A B
Caudal
Dorsal
M e dial
Lateral
Ventral
Medial B
~
Dorsal Lateral
Ventral
Fig. 7. A chart of the distribution of labeled somata (solid dots) in the trigeminal ganglion following HRP injections of mandibular and sublingual salivary glands. Labeled cell bodies were localized in the ventromedial part of the mandibular division of the trigeminal ganglion. Abbreviations: O, ophthalmic nerve; M, maxillary nerve; T, mandibular nerve, A and B: rostral and caudal levels of transverse sections from trigeminal ganglion.
308
Fig. 8. Labeled neuronal cell bodies in the proximal vaga[ (jugular) ganglion following HRP injection of mandibular salivary gland. Dark-field, 500 : . Neither the glossopharyngeal nor the distal vagal (nodose) ganglia labeled as a result of mandibular and/or sublingual gland injections. DISCUSSION
Preganglionic parasympathetic salivatory center Kohnstamm, after transecting chorda tympani fibers to the mandibular gland, identified retrograde cell changes in a rather diffuse salivatory nucleus in the dog 14. The majority of reactive neurons were found contralateral to the operation. He described reactive large motor neurons that extended longitudinally from a point just rostral to the caudal pole of the facial nucleus to the rostral pole of the motor nucleus of the trigeminal nerve. These neurons were bordered by the raphe, lateral vestibular nucleus and the ventricular base. In subsequent papers15,16, he referred to these cells as the superior salivatory nucleus. He also described an inferior salivatory nucleus consisting of smaller neurons located between the inferior olivary nucleus and the nucleus ambiguus. He stated that many of the inferior salivatory neurons projected to terminations in the ipsilateral otic ganglion, from which arose the postganglionic fibers to the parotid gland. He observed, however, that a proportion of inferior salivatory neurons reacted to contralateral lingual nerve section and surmised that these crossed under the ventricular floor and projected through the otic ganglion to cell bodies in the mandibular ganglia. Solomowicz found reactive neurons bilaterally in the brain stems of two dogs that had undergone unilateral mandibular gland extirpations za. He reported a slight
309 predominance of reactive neurons ipsilaterally in the region of the lateral vestibular nucleus. These neurons were large. Most resembled motor neurons although others were oval and resembled the cells of Clarke's column. Solomowicz observed that most chromatolytic neurons were in the region of Deiter's nucleus with only a few scattered in the adjoining reticular formation. Yagita and Hayama, after studying Nissl-stained brain sections from unoperated dogs, suggested that Kohnstamm's identification of the inferior salivatory nucleus may have been confounded by the occurrence of normal neurons with chromatolytic appearances2L After transecting the chorda tympani at various levels, these investigators observed reactive somata in the reticular formation at the level of the facial nucleus. The size and distribution of the affected neurons differed significantly from some of the earlier descriptions of the superior salivatory nucleus14-tr, ~3. The reactive cells approximated the neurons of the dorsal motor nucleus of the vagus in size and shape, and they occurred almost exclusively on the operated side. Affected and unaffected cell bodies were found in small groups near the ventromedial border of Deiter's nucleus; others were scattered in the reticular formation midway between Deiter's nucleus and the facial nucleus and a third group was situated over the dorsal aspect of the middle third of the facial nucleus. Caudally, the reactive neurons appeared just rostral to the caudal pole of the facial nucleus and then extended rostraUy to the point where the descending radix of the facial nerve was formed. All were located lateral to the ascending fibers of the facial nerve. Later, after transecting the ramus tympanicus of the glossopharyngeal nerve, Yagita identified the inferior salivatory nucleus as a direct caudal continuation of the cell group supplying mandibular and sublingual glands 28. Torvik, after disrupting the chorda tympani, and possibly the tympanic nerve, in the middle ear of neonatal cats, described the salivatory center on the basis of retrograde cell change 26. His findings were similar to those of Yagita and Hayama29. Reactive neurons were diffusely arranged in the lateral part of the medullary reticular formation in a region corresponding to the nucleus reticularis parvocellularis. Affected neurons appeared at the caudal pole of the facial nucleus and extended rostrally to the level of the descending root of the facial nerve. Although the operative procedure destroyed the nervus intermedius (i. e. the preganglionic supply to the mandibular and sublingual glands), the extent of damage to the tympanic nerve (i. e. the preganglionic supply to the parotid gland) was not defined fully. Nevertheless, he suggested that the cells of origin of the motor fibers in the n. intermedius and n. tympanicus form a single loose aggregate rather than distinct superior and inferior salivatory nuclei. The brain stem stimulation studies of some investigators5,19,27 supported the confluence of the mandibular-sublingual neurons and the parotid neurons. Although no separation occurred between the two groups, mandibular-sublingual neurons were distributed rostral, and parotid neurons caudal to an intermediate area of overlapping distribution. Schute and Lewis, employing a modified Koelle technique for cholinesterase in the rat, identified two cell groups on the central course of the secretory motor component of the nervus intermedius22. These groups, identified as the medial and
310 lateral salivatory nuclei, flanked the rostral end of the genu of the lhcial nerve root. They suggested that previous identifications of the salivatory center in the lateral reticular formation by means of induced chromatolysis may not have demonstrated secretomotor preganglionic neurons, but, instead, reticular neurons that project to the salivatory center and react transynaptically to lesions of the chorda tympani or nervus intermedius. They perceived these medial and lateral nuclei as the sole source of preganglionic salivatory and lacrimal fibers. The divergent results obtained in earlier studies of the salivatory neurons by means of induced retrograde cell change may be attributed to lack of selectivity in nerve lesions and, perhaps to a greater degree, to difficulty in distinguishing reactive neurons from normal neurons with sparse Nissl granulation. These problems did not exist in the present study since the injected marker was confined to the salivary gland masses and the enzyme reaction product provided a means to distinguish clearly between affected and unaffected cell bodies. It is problematical, however, as to what proportion of the mandibular and sublingual neurons were labeled. The gland injections undoubtedly labeled preganglionic neurons that terminate in many of the salivary ganglia described by Langley is. Such ganglia would include those scattered over the sublingual gland, as well as those at the hilus (the submaxillary ganglion of Langley) and along the lobular ducts of the mandibular gland. The large ganglion rostral to the sublingual gland in the angle between the chorda tyrnpani and the lingual nerve (i. e. the sublingual ganglion of Langley) was not injected. Thus, the proportionate representation of labeled sublingual preganglionic neurons was probably less than that of mandibular preganglionic neurons. The preganglionic neurons labeling after mandibular gland injections were coextensive with those labeling after sublingual gland injection. Thus, discrete somatotopic organization of these preganglionic neurons was not evident within the salivatory center. The observed size and distribution of the salivatory neurons conformed most closely to the earlier description of Torvik '~6. Labeled neurons appeared dorsal to the facial nucleus just rostral to its caudal pole and extended rostrally over the caudal third of the dorsal nucleus of the trapezoid body. Labeled neurons were bounded laterally by the spinal nucleus of trigeminal nerve; dorsally by the medial, descending, and lateral vestibular nuclei: and medially by the nucleus reticularis gigantocellularis. As noted by Torvik 2~, many of the salivatory neurons were within the nucleus reticularis parvocellularis. The salivatory center for the mandibular and sublingual glands appeared to be coextensive with the more rostral neurons of origin of the greater petrosal nerve as determined recently by Gomez 1°. Thus, the preganglionic neurons supplying these major salivary glands would seem to be partially overlapped by neurons supplying the lacrimal gland and glands of the nasal and palatine mucosa. In the past, these neurons were designated collectively as the superior or rostral salivatory nucleus.
Autonomic ganglion The cranial cervical ganglion had been identified in early investigations as the source of postganglionic sympathetic fibers to the salivary glands is. The present H R P
311 study confirmed this, and, in addition, indicated that the cell bodies, supplying the mandibular and sublingual glands, were localized caudally within the ganglion.
Sensory ganglia Prior studies of salivary gland innervation have concentrated largely on the motor nerve supply with little consideration of the afferent innervation. Although sensory fibers and receptors have been identified within the glands 25, the location of the afferent cell bodies seems to have received little attention. The present HRP study afforded a means to identify the location of afferent neurons serving the mandibular and sublingual glands. Following HRP injections into these glands, afferent cell bodies were identified in the trigeminal, the geniculate, and the proximal vagal ganglia. Previous anatomic and physiologic studies of the trigeminal ganglion in various species have revealed a medial-lateral somatotopic organization corresponding to the three afferent divisions of the nerve 1,a,7,11,20. Thus, somata related to the ophthalmic division are arranged medially, while somata related to the mandibular division are located laterally. Maxillary division cell bodies occupy an intermediate position. In the present study, HRP injections of the mandibular and/or sublingual glands labeled cells in the lateral or mandibular division of the ganglion. Within this division, the labeled cells were concentrated ventromedially. The ventral predominance was seemingly in correspondence with the results of the neurophysiologic study of Beaudreau and Jerge who found that perioral structures were innervated by cells near the ventral surface of the ganglion in the cat a. Not all investigators, however, have found a dorsoventral distribution pattern in this or other species7. The labeling of the proximal vagal or jugular ganglion as a consequence of ipsilateral HRP injections into the mandibular and sublingual glands was not expected. Classically, the neurons of this ganglion have been associated with exteroceptive fibers 6. Du Bois and Foley found that an average of 73 ~ of the jugular neurons underwent chromatolysis after transection of the auricular branch of the vagus in the cat s. Although the jugular ganglion has been associated almost exclusively with cutaneous innervation of the external ear, the present results and studies in progress indicate additional receptive fields. The course of the salivary afferent fibers to the proximal vagal ganglion was not resolved in the present study. Presumably, enzyme deposited in the glands was taken up and conveyed along fibers of the chords tympani to the facial nerve and then via the auricular branch of the vagus to the proximal vagal ganglion. The classification of the afferent neurons to the salivary glands also deserves consideration. More specifically, should these be classified as somatic or visceral afferent neurons? Of the three labeled sensory ganglia, only the geniculate ganglion has been recognized as containing visceral afferent components4,6. The mandibular and sublingual glands have been regarded as ectodermal derivatives2. Thus, on a developmental basis, their sensory fibers would be considered appropriately as general somatic afferent neurons. By proposing this somatic classification, the afferent supply to the salivary glands would be considered in the same manner as the trigeminal sensory innervation of the oral and nasal mucosa.
312 ACKNOWLEDGEMENTS This research was supported by N I H G r a n t 5-F34-FM06013-02. The a u t h o r s t h a n k Drs. A. de L a h u n t a , H. E. Evans, a n d J. M. Petras for their valuable directions during the experiment a n d for editing the manuscript. The authors are also grateful to Mr. W. P. H a m i l t o n IV for the illustrations, to Dr. Isak Foss for his surgical assistance, to Mrs. Helen L e h m a n a n d Miss B a r b a r a L o m a n for their technical help, a n d to Miss Alice C u r r a n for secretarial support.
REFERENCES t Allen, W. F., Localization in the ganglion semilunare of the cat, J. comp. Neurok, 38 (1924) 1-24. 2 Arey, L. B., Developmental Anatomy, W. B. Saunders, Philadelphia, 1965, p. 227. 3 Beaudreau, D. E. and Jerge, C. R., Somatotopic representation in the Gasserian ganglion of tactile peripheral fields in the cat, Arch. oral Biol., 13 (1968) 247-255. 4 Brodal, A., Neurological Anatomy, Oxford University Press, London, 1969, pp. 350-460. 5 Chatfield, P. O., Salivation in response to localized stimulation of the medulla, Amer. J. Physiol., 133 (1941) 637-641. 6 Crosby, E. C., Humphrey, T. and Lauer, E. W., Correlative Anatomy of the Nervous System, Macmillan, New York, 1962, p. 138-165. 7 Darian-Smith, I., Mutton, P. and Proctor, R., Functional organization of tactile cutaneou~ afferents within the semilunar ganglion and trigeminal spinal tract of the cat, J. Neurophysiol., 28 (1965) 682-694. 8 DuBois, F. S. and Foley, J. O., Quantitative studies of the vagus nerve in the cat. II. The ratio of jugular to nodose fibers, J. comp. Neurol., 67 (1937) 69-87. 9 Franklin, I. I. and Reid, J. S., Clinical use of a combination of fentanyl and droperidol in dogs. Vet. Med. small Anim. Clin., 60 (1965) 927-930. 10 Gomez, M. M., Central origin of greater superficial petrosal nerve efferents, Anat. Rec., 193 (1979) 539. 11 Gregg, J. M. and Dixon, A. D., Somatotopic organization of the trigeminal ganglion in the rat, Arch. oral Biol., 18 (1973) 487-498. 12 Hamlin, R. L., Ginaven, S. M. and Smith, C. R., Fentanyl citrate-droperidol and pentobarbital for intravenous anesthesia in dogs, J. Amer. vet. reed. Ass., 152 (1968) 360-364. 13 Jones, L. M., Booth, N. H. and McDonald, L. E., Veterinary Pharmacology and Therapeutics, The Iowa State University Press, Ames, Iowa, 1977, pp. 318-321. 14 Kohnstamm, O., Der Nucleus salivatorius chordae tympani (nervi intermedii), Anat. Anz., 21 (1902) 362-363. 15 Kohnstamm, O., Der Nucleus salivatorius und das cranio-viscerale System (mit Demonstration), Arch. Psychiat. Nervenkr., 37 (1903) 660-662. 16 Kohnstamm, O., Der Nucleus salivatorius inferior und des cranio-viscerale System, Neurol. Zentralbl., 22 (1903) 699. 17 Krahwinkel, D. J., Sawyer, D. C. and Evans, A. T., Neuroleptanalgesia and neuroleptanesthesia, J. Amer. Anita. Hosp. Ass., 8 (1972) 368-370. 18 Langley, J. N., On the physiology of the salivary secretion. Part VI. Chiefly upon the connections of peripheral nerve cells with the nerve fibers which run to sub-lingual and sub-maxillary glands, J. Physiol. (Lond.), 11 (1890) 123-158. 19 Magoun, H. W. and Beaton, L. E., The salivatory motor nuclei in the monkey, Amer. J. Physiol., 136 (1942) 720-725. 20 Mazza, J. P. and Dixon, A. D., A histological study of chromatolytic cell groups in the trigeminal ganglion of the rat, Arch. oral Biol., 17 (1972) 377-387. 21 Nauta, H. J. W., Pritz, M. B. and Lasek, R. J., Afferents to the rat caudoputamen studied with horseradish peroxidase. An evaluation of a retrograde neuroanatomical research method, Brain Research, 67 (1974) 219-238. 22 Shute, C. C. D. and Lewis, P. R., The salivatory center in the rat, J. Anat. (Lond.), 94 (1960) 59-81.
313 23 Solomowicz, J., Von Centrum der Submaxillardriise, Neurol. Zbl., 27 (1908) 724-727. 24 Soma, L. R. and Shields, D. R., Neuroleptanalgesia produced by fentanyl and droperidol, J. Amer. vet med. Ass., 145 (1964) 897-902. 25 Stohr, P., Nervensystem. In W. Mollendorff and W. Bargmann, (Eds.), Handbuch der Mikroskopischen Anatomie des Menschen, Vol. 4, Springer-Verlag, Berlin, 1957, p. 297. 26 Torvik, A., Die Lokalisation des 'Speichelzentrums' bei der Katze, Z. Mikr.-anat. Forsch., 63 (1957) 317-326. 27 Wang, S. C., Localization of the salivatory center in the medulla of the cat, J. Neurophysiol., 6 (1943) 195-202. 28 Yagita, K., Weitere Untersuchungen fiber das Speichelzentrum, Anat. Ariz., 35 (1909) 70-75. 29 Yagita, K. and Hayama, S., Ober des Speichelsekretioncentrum, Neurol. Zbl., 28 (1909) 738-753.
301
Research Reports
M O T O R A N D S E N S O R Y C E N T E R S F O R T H E I N N E R V A T I O N OF M A N D I BULAR AND SUBLINGUAL SALIVARY GLANDS: A HORSERADISH PEROXIDASE STUDY IN THE DOG
GREGORY A. CHIBUZO* and JOHN F. CUMMINGS Department of Anatomy, N YS College of Veterinary Medicine, Cornell University, Ithaca, N. Y. 14853 (U.S.A.)
(Accepted October 1lth, 1979)
Key words: salivary gland innervation - - horseradish peroxidase - - dog - - cranial nerves - - reticular
formation - - brain stem
SUMMARY Horseradish peroxidase was injected at multiple sites in the mandibular and sublingual salivary glands in order to label the preganglionic salivatory neurons in the brain stem. The same injections resulted in retrograde labeling of the sympathetic and sensory neurons that project to these glands. Labeled fusiform and multipolar salivatory neurons were found ipsilaterally in the lateral reticular formation of the medulla where they extended over the rostral four-fifths of the facial nucleus and the caudal one-third of the dorsal nucleus of the trapezoid body. The vast majority of the small and medium-sized, labeled neurons appeared in the nucleus reticularis parvocellularis. Labeled neurons also appeared occasionally at the ventral and lateral aspects of the facial nucleus. Enzyme injections into these glands labeled sympathetic neurons that were concentrated in the caudal one-third of the ipsilateral cranial cervical ganglion. Labeled sensory neurons were distributed randomly in the ipsilateral proximal vagal and geniculate ganglia. Large numbers of sensory neurons were concentrated ventromedially within the mandibular zone of the trigeminal ganglion.
* Address for all correspondence: Department of Anatomy, School of Veterinary Medicine, Tuskegee Institute, Ala. 36088, U.S.A.
302 INTRODUCTION Previous studies of the innervation of the major salivary glands have focused mainly on the preganglionic parasympathetic neurons of the brain stem, i.e. the salivatory nucleus or nuclei. Anatomic studies of these neurons have been based largely upon retrograde cell change induced by interruption of preganglionic axons at various peripheral levels. This technique has led to divergent results regarding the size and location of the somata and the course of their preganglionic axons14,~3,26,29. Neurophysiologic studies on the salivatory center have relied upon the identification of medullary stimulation sites that evoke salivary gland secretion. It has been difficult to draw firm conclusions on the location of the preganglionic neurons from these studies since it was often problematical as to whether cell bodies or fibers were activated or whether stimulated sites were afferent or efferent5A9,27. In employing the horseradish peroxidase (HRP) retrograde labeling technique, the present study eliminated the interpretation problems that were encountered in earlier neuroanatomic and physiologic experiments. Since many of the preganglionic neurons supplying the mandibular and sublingual glands terminate in ganglia associated with these gland masses 18, diffuse injections of the glands consistently and specifically labeled the medullary parasympathetic neurons. These intragtandular injections also labeled sympathetic and sensory neurons serving these glands. MATERIALS AND METHODS Nine neonatal dogs (2-6 weeks of age) were available for the experiment. Each received an intramuscular injection of promazine hydrochloride (2-6 mg/kg) and/or a subcutaneous injection of atropine sulfate (0.04 mg/kg) preanesthetically. After 10-15 rain, fentanyl citrate-droperidol (0.2-0.4 ml/kg) (Innovar Vet., Pitman-Moore, Inc., Washington Crossing, N. J. 08560) was injected intraperitoneally, and after a 10-min delay, this was followed by pentobarbital sodium (4-8 mg/kg) administered intravenously. Supplemental doses of pentobarbital sodium for extension of analgesia beyond 35 rain were given to effect. The preceding anesthetic regime was modified from schedules developed for use in adult dogs 9,1zAzAT,24. After the animals were anesthetized, the left mandibular (submaxillary) and/or sublingual salivary glands were exposed surgically. Twenty to thirty microliters of a 30 ~ aqueous solution of HRP (Sigma type IV) were injected at multiple sites in the mandibular salivary gland in 3 dogs, in the sublingual salivary gland in 3 dogs, and in both glands in 3 dogs. The incisions were closed and the animals were placed in a warm recovery room. At 24 or 48 h postoperatively, the animals received an intravenous overdose of pentobarbital sodium and were then perfused via the left cardiac ventricle with a solution containing 0.5 ~o paraformaldehyde, 2.5 ~ glutaraldehyde, and 2.5 ~ sucrose in 0.1 M phosphate buffer (pH 7.4). The brain was removed and stored in chilled fixative for 24 h. The trigeminal, geniculate, glossopharyngeal, vagal, otic and cranial cervical ganglia were removed bilaterally and also immersed in refrigerated fixative for 24 h. The tissues were then stored in refrigerated 0.1 M
303 phosphate buffer (pH 7.4) prepared in a 30 ~o sucrose solution. Serial frozen sections of the brain stem and ganglia were cut at 60 #m. All sections were incubated with 3,3'diaminobenzidine and hydrogen peroxide, counterstained and mounted according to the method of Nauta et al. 21. Labeled neurons were detected by bright-field or darkfield microscopy. RESULTS
Preganglionic parasympathetic salivatory neurons Injections of horseradish peroxidase (HRP) into the mandibular and/or sublingual salivary glands consistently labeled nerve cell bodies in the lateral medullary reticular formation. The labeled multipolar somata were confined to the side of the injection. The cell bodies were small or medium sized and presented fusiform or polygonal outlines (Figs. 1 and 2). They ranged in size from 28 x 19 # m to 55 x 17 /~m, and most contained discrete Nissl granules. The great bulk of the labeled population was situated dorsal to the rostral four-fifths of the facial nucleus and continued rostrally over the caudal one-third of the dorsal nucleus of the trapezoid body (Fig. 3). The descending root o f the facial nerve marked the rostral limit of the labeled neurons. Occasionally, labeled neurons appeared at the medial or ventral aspect of the facial nucleus (Fig. 3). These were rare, however, and most of the labeled cells were
Fig. 1. Labeled small and medium-sized multipolar somata in the salivatory center. The few medium sized neurons are indicated by arrows. Dark-field, 200 x.
304
Fig. 2. Isolated field with two labeled medium sized somata (arrows) among smaller neurons in the medullary reticular formation ventral to medial vestibular nucleus. Dark-field, 400 .:. scattered dorsal to this nucleus in an area that was bounded laterally by the nucleus of the spinal tract of the trigeminal nerve, medially by the nucleus reticularis gigantocellularis, and dorsally by the descending, medial and lateral vestibular nuclei (Fig. 3). Within these boundaries, most of the labeled cells were small. Medium-sized neurons were less frequent and appeared usually in the region immediately ventral to the medial vestibular nucleus. The majority of the labeled neurons resided within the nucleus reticularis parvocellularis. Neurons labeling after mandibular gland injections were coextensive with those that labeled after sublingual gland injections. Mandibular gland injections, however, resulted in greater numbers of labeled neurons. Although injections into both salivary glands produced the greatest number of labeled cells, the distribution of the neurons in these cases conformed to that observed after injection of a single gland.
Autonomic ganglia Numerous labeled neurons were observed in the ipsilateral cranial cervical ganglion following mandibular and/or sublingual salivary gland H R P injections (Fig. 4). These labeled sympathetic neurons were concentrated at the caudal one-third of this ganglion (Fig. 5). Labeled cell bodies were not observed in otic ganglia after sublingual and/or mandibular gland injections.
Sensory ganglia Labeled neurons were observed in the ipsilateral trigeminal ganglion after gland
305 injections (Fig. 6). Mandibular gland injections produced greater numbers of labeled neurons than did sublingual gland injections, and the most extensive labeling occurred in those cases in which both glands were injected. In all animals, the labeled somata appeared in the lateral or mandibular zone of the ganglion, and most were localized ventromedially within the mandibular zone (Fig. 7). Labeled neurons were also present in the ipsilateral geniculate and proximal vagal (jugular) ganglia (Fig. 8). In both ganglia a small proportion of the cells labeled and there was no detectable somatotopic organization.
e~ YP~ , ~ / " / C
B
AUDAL END
"
E
. .:-.;'. .:.,.
.©
H~
"
ROSTRALEND
Fig. 3. Brain stem sections arranged in a caudal (A) to rostral (I) sequence to show distribution of labeled salivatory neurons (solid dots) in the medullary reticular formation. Abbreviations: NVC, caudal vestibular nucleus; NVM, medial vestibular nucleus; NVL, lateral vestibular nucleus; R VIII, vestibulocochlear nerve; NRV, nucleus and spinal tract of trigeminal nerve; N VII, facial nucleus; PY, pyramid; N VIII, cochlear nucleus; NTD, dorsal nucleus of trapezoid body (superior olive); G, genu of facial nerve; NVI, abdueens nucleus; R VII, descending root of facial nerve; TB, trapezoid body.
306
Fig. 4. Labeled neuronal s o m a t a in the cranial cervical ganglion. Dark-field, 367
.
Proximal
Laterar
Medial
Distal
g
y
p
thetic trunk
Fig. 5. A chart of labeled neurons (solid dots) in a horizontal section of the cranial (superior) ccrvica! ganglion following H R P injection of mandibular and sublingual salivary glands. Labeled cell bodies were localized towards the caudal end of this ganglion.
307
Fig. 6. Labeled neuronal cell bodies from the mandibular or lateral division of trigeminal ganglion. Dark-field, 375 ×.
ateral
A B
Caudal
Dorsal
M e dial
Lateral
Ventral
Medial B
~
Dorsal Lateral
Ventral
Fig. 7. A chart of the distribution of labeled somata (solid dots) in the trigeminal ganglion following HRP injections of mandibular and sublingual salivary glands. Labeled cell bodies were localized in the ventromedial part of the mandibular division of the trigeminal ganglion. Abbreviations: O, ophthalmic nerve; M, maxillary nerve; T, mandibular nerve, A and B: rostral and caudal levels of transverse sections from trigeminal ganglion.
308
Fig. 8. Labeled neuronal cell bodies in the proximal vaga[ (jugular) ganglion following HRP injection of mandibular salivary gland. Dark-field, 500 : . Neither the glossopharyngeal nor the distal vagal (nodose) ganglia labeled as a result of mandibular and/or sublingual gland injections. DISCUSSION
Preganglionic parasympathetic salivatory center Kohnstamm, after transecting chorda tympani fibers to the mandibular gland, identified retrograde cell changes in a rather diffuse salivatory nucleus in the dog 14. The majority of reactive neurons were found contralateral to the operation. He described reactive large motor neurons that extended longitudinally from a point just rostral to the caudal pole of the facial nucleus to the rostral pole of the motor nucleus of the trigeminal nerve. These neurons were bordered by the raphe, lateral vestibular nucleus and the ventricular base. In subsequent papers15,16, he referred to these cells as the superior salivatory nucleus. He also described an inferior salivatory nucleus consisting of smaller neurons located between the inferior olivary nucleus and the nucleus ambiguus. He stated that many of the inferior salivatory neurons projected to terminations in the ipsilateral otic ganglion, from which arose the postganglionic fibers to the parotid gland. He observed, however, that a proportion of inferior salivatory neurons reacted to contralateral lingual nerve section and surmised that these crossed under the ventricular floor and projected through the otic ganglion to cell bodies in the mandibular ganglia. Solomowicz found reactive neurons bilaterally in the brain stems of two dogs that had undergone unilateral mandibular gland extirpations za. He reported a slight
309 predominance of reactive neurons ipsilaterally in the region of the lateral vestibular nucleus. These neurons were large. Most resembled motor neurons although others were oval and resembled the cells of Clarke's column. Solomowicz observed that most chromatolytic neurons were in the region of Deiter's nucleus with only a few scattered in the adjoining reticular formation. Yagita and Hayama, after studying Nissl-stained brain sections from unoperated dogs, suggested that Kohnstamm's identification of the inferior salivatory nucleus may have been confounded by the occurrence of normal neurons with chromatolytic appearances2L After transecting the chorda tympani at various levels, these investigators observed reactive somata in the reticular formation at the level of the facial nucleus. The size and distribution of the affected neurons differed significantly from some of the earlier descriptions of the superior salivatory nucleus14-tr, ~3. The reactive cells approximated the neurons of the dorsal motor nucleus of the vagus in size and shape, and they occurred almost exclusively on the operated side. Affected and unaffected cell bodies were found in small groups near the ventromedial border of Deiter's nucleus; others were scattered in the reticular formation midway between Deiter's nucleus and the facial nucleus and a third group was situated over the dorsal aspect of the middle third of the facial nucleus. Caudally, the reactive neurons appeared just rostral to the caudal pole of the facial nucleus and then extended rostraUy to the point where the descending radix of the facial nerve was formed. All were located lateral to the ascending fibers of the facial nerve. Later, after transecting the ramus tympanicus of the glossopharyngeal nerve, Yagita identified the inferior salivatory nucleus as a direct caudal continuation of the cell group supplying mandibular and sublingual glands 28. Torvik, after disrupting the chorda tympani, and possibly the tympanic nerve, in the middle ear of neonatal cats, described the salivatory center on the basis of retrograde cell change 26. His findings were similar to those of Yagita and Hayama29. Reactive neurons were diffusely arranged in the lateral part of the medullary reticular formation in a region corresponding to the nucleus reticularis parvocellularis. Affected neurons appeared at the caudal pole of the facial nucleus and extended rostrally to the level of the descending root of the facial nerve. Although the operative procedure destroyed the nervus intermedius (i. e. the preganglionic supply to the mandibular and sublingual glands), the extent of damage to the tympanic nerve (i. e. the preganglionic supply to the parotid gland) was not defined fully. Nevertheless, he suggested that the cells of origin of the motor fibers in the n. intermedius and n. tympanicus form a single loose aggregate rather than distinct superior and inferior salivatory nuclei. The brain stem stimulation studies of some investigators5,19,27 supported the confluence of the mandibular-sublingual neurons and the parotid neurons. Although no separation occurred between the two groups, mandibular-sublingual neurons were distributed rostral, and parotid neurons caudal to an intermediate area of overlapping distribution. Schute and Lewis, employing a modified Koelle technique for cholinesterase in the rat, identified two cell groups on the central course of the secretory motor component of the nervus intermedius22. These groups, identified as the medial and
310 lateral salivatory nuclei, flanked the rostral end of the genu of the lhcial nerve root. They suggested that previous identifications of the salivatory center in the lateral reticular formation by means of induced chromatolysis may not have demonstrated secretomotor preganglionic neurons, but, instead, reticular neurons that project to the salivatory center and react transynaptically to lesions of the chorda tympani or nervus intermedius. They perceived these medial and lateral nuclei as the sole source of preganglionic salivatory and lacrimal fibers. The divergent results obtained in earlier studies of the salivatory neurons by means of induced retrograde cell change may be attributed to lack of selectivity in nerve lesions and, perhaps to a greater degree, to difficulty in distinguishing reactive neurons from normal neurons with sparse Nissl granulation. These problems did not exist in the present study since the injected marker was confined to the salivary gland masses and the enzyme reaction product provided a means to distinguish clearly between affected and unaffected cell bodies. It is problematical, however, as to what proportion of the mandibular and sublingual neurons were labeled. The gland injections undoubtedly labeled preganglionic neurons that terminate in many of the salivary ganglia described by Langley is. Such ganglia would include those scattered over the sublingual gland, as well as those at the hilus (the submaxillary ganglion of Langley) and along the lobular ducts of the mandibular gland. The large ganglion rostral to the sublingual gland in the angle between the chorda tyrnpani and the lingual nerve (i. e. the sublingual ganglion of Langley) was not injected. Thus, the proportionate representation of labeled sublingual preganglionic neurons was probably less than that of mandibular preganglionic neurons. The preganglionic neurons labeling after mandibular gland injections were coextensive with those labeling after sublingual gland injection. Thus, discrete somatotopic organization of these preganglionic neurons was not evident within the salivatory center. The observed size and distribution of the salivatory neurons conformed most closely to the earlier description of Torvik '~6. Labeled neurons appeared dorsal to the facial nucleus just rostral to its caudal pole and extended rostrally over the caudal third of the dorsal nucleus of the trapezoid body. Labeled neurons were bounded laterally by the spinal nucleus of trigeminal nerve; dorsally by the medial, descending, and lateral vestibular nuclei: and medially by the nucleus reticularis gigantocellularis. As noted by Torvik 2~, many of the salivatory neurons were within the nucleus reticularis parvocellularis. The salivatory center for the mandibular and sublingual glands appeared to be coextensive with the more rostral neurons of origin of the greater petrosal nerve as determined recently by Gomez 1°. Thus, the preganglionic neurons supplying these major salivary glands would seem to be partially overlapped by neurons supplying the lacrimal gland and glands of the nasal and palatine mucosa. In the past, these neurons were designated collectively as the superior or rostral salivatory nucleus.
Autonomic ganglion The cranial cervical ganglion had been identified in early investigations as the source of postganglionic sympathetic fibers to the salivary glands is. The present H R P
311 study confirmed this, and, in addition, indicated that the cell bodies, supplying the mandibular and sublingual glands, were localized caudally within the ganglion.
Sensory ganglia Prior studies of salivary gland innervation have concentrated largely on the motor nerve supply with little consideration of the afferent innervation. Although sensory fibers and receptors have been identified within the glands 25, the location of the afferent cell bodies seems to have received little attention. The present HRP study afforded a means to identify the location of afferent neurons serving the mandibular and sublingual glands. Following HRP injections into these glands, afferent cell bodies were identified in the trigeminal, the geniculate, and the proximal vagal ganglia. Previous anatomic and physiologic studies of the trigeminal ganglion in various species have revealed a medial-lateral somatotopic organization corresponding to the three afferent divisions of the nerve 1,a,7,11,20. Thus, somata related to the ophthalmic division are arranged medially, while somata related to the mandibular division are located laterally. Maxillary division cell bodies occupy an intermediate position. In the present study, HRP injections of the mandibular and/or sublingual glands labeled cells in the lateral or mandibular division of the ganglion. Within this division, the labeled cells were concentrated ventromedially. The ventral predominance was seemingly in correspondence with the results of the neurophysiologic study of Beaudreau and Jerge who found that perioral structures were innervated by cells near the ventral surface of the ganglion in the cat a. Not all investigators, however, have found a dorsoventral distribution pattern in this or other species7. The labeling of the proximal vagal or jugular ganglion as a consequence of ipsilateral HRP injections into the mandibular and sublingual glands was not expected. Classically, the neurons of this ganglion have been associated with exteroceptive fibers 6. Du Bois and Foley found that an average of 73 ~ of the jugular neurons underwent chromatolysis after transection of the auricular branch of the vagus in the cat s. Although the jugular ganglion has been associated almost exclusively with cutaneous innervation of the external ear, the present results and studies in progress indicate additional receptive fields. The course of the salivary afferent fibers to the proximal vagal ganglion was not resolved in the present study. Presumably, enzyme deposited in the glands was taken up and conveyed along fibers of the chords tympani to the facial nerve and then via the auricular branch of the vagus to the proximal vagal ganglion. The classification of the afferent neurons to the salivary glands also deserves consideration. More specifically, should these be classified as somatic or visceral afferent neurons? Of the three labeled sensory ganglia, only the geniculate ganglion has been recognized as containing visceral afferent components4,6. The mandibular and sublingual glands have been regarded as ectodermal derivatives2. Thus, on a developmental basis, their sensory fibers would be considered appropriately as general somatic afferent neurons. By proposing this somatic classification, the afferent supply to the salivary glands would be considered in the same manner as the trigeminal sensory innervation of the oral and nasal mucosa.
312 ACKNOWLEDGEMENTS This research was supported by N I H G r a n t 5-F34-FM06013-02. The a u t h o r s t h a n k Drs. A. de L a h u n t a , H. E. Evans, a n d J. M. Petras for their valuable directions during the experiment a n d for editing the manuscript. The authors are also grateful to Mr. W. P. H a m i l t o n IV for the illustrations, to Dr. Isak Foss for his surgical assistance, to Mrs. Helen L e h m a n a n d Miss B a r b a r a L o m a n for their technical help, a n d to Miss Alice C u r r a n for secretarial support.
REFERENCES t Allen, W. F., Localization in the ganglion semilunare of the cat, J. comp. Neurok, 38 (1924) 1-24. 2 Arey, L. B., Developmental Anatomy, W. B. Saunders, Philadelphia, 1965, p. 227. 3 Beaudreau, D. E. and Jerge, C. R., Somatotopic representation in the Gasserian ganglion of tactile peripheral fields in the cat, Arch. oral Biol., 13 (1968) 247-255. 4 Brodal, A., Neurological Anatomy, Oxford University Press, London, 1969, pp. 350-460. 5 Chatfield, P. O., Salivation in response to localized stimulation of the medulla, Amer. J. Physiol., 133 (1941) 637-641. 6 Crosby, E. C., Humphrey, T. and Lauer, E. W., Correlative Anatomy of the Nervous System, Macmillan, New York, 1962, p. 138-165. 7 Darian-Smith, I., Mutton, P. and Proctor, R., Functional organization of tactile cutaneou~ afferents within the semilunar ganglion and trigeminal spinal tract of the cat, J. Neurophysiol., 28 (1965) 682-694. 8 DuBois, F. S. and Foley, J. O., Quantitative studies of the vagus nerve in the cat. II. The ratio of jugular to nodose fibers, J. comp. Neurol., 67 (1937) 69-87. 9 Franklin, I. I. and Reid, J. S., Clinical use of a combination of fentanyl and droperidol in dogs. Vet. Med. small Anim. Clin., 60 (1965) 927-930. 10 Gomez, M. M., Central origin of greater superficial petrosal nerve efferents, Anat. Rec., 193 (1979) 539. 11 Gregg, J. M. and Dixon, A. D., Somatotopic organization of the trigeminal ganglion in the rat, Arch. oral Biol., 18 (1973) 487-498. 12 Hamlin, R. L., Ginaven, S. M. and Smith, C. R., Fentanyl citrate-droperidol and pentobarbital for intravenous anesthesia in dogs, J. Amer. vet. reed. Ass., 152 (1968) 360-364. 13 Jones, L. M., Booth, N. H. and McDonald, L. E., Veterinary Pharmacology and Therapeutics, The Iowa State University Press, Ames, Iowa, 1977, pp. 318-321. 14 Kohnstamm, O., Der Nucleus salivatorius chordae tympani (nervi intermedii), Anat. Anz., 21 (1902) 362-363. 15 Kohnstamm, O., Der Nucleus salivatorius und das cranio-viscerale System (mit Demonstration), Arch. Psychiat. Nervenkr., 37 (1903) 660-662. 16 Kohnstamm, O., Der Nucleus salivatorius inferior und des cranio-viscerale System, Neurol. Zentralbl., 22 (1903) 699. 17 Krahwinkel, D. J., Sawyer, D. C. and Evans, A. T., Neuroleptanalgesia and neuroleptanesthesia, J. Amer. Anita. Hosp. Ass., 8 (1972) 368-370. 18 Langley, J. N., On the physiology of the salivary secretion. Part VI. Chiefly upon the connections of peripheral nerve cells with the nerve fibers which run to sub-lingual and sub-maxillary glands, J. Physiol. (Lond.), 11 (1890) 123-158. 19 Magoun, H. W. and Beaton, L. E., The salivatory motor nuclei in the monkey, Amer. J. Physiol., 136 (1942) 720-725. 20 Mazza, J. P. and Dixon, A. D., A histological study of chromatolytic cell groups in the trigeminal ganglion of the rat, Arch. oral Biol., 17 (1972) 377-387. 21 Nauta, H. J. W., Pritz, M. B. and Lasek, R. J., Afferents to the rat caudoputamen studied with horseradish peroxidase. An evaluation of a retrograde neuroanatomical research method, Brain Research, 67 (1974) 219-238. 22 Shute, C. C. D. and Lewis, P. R., The salivatory center in the rat, J. Anat. (Lond.), 94 (1960) 59-81.
313 23 Solomowicz, J., Von Centrum der Submaxillardriise, Neurol. Zbl., 27 (1908) 724-727. 24 Soma, L. R. and Shields, D. R., Neuroleptanalgesia produced by fentanyl and droperidol, J. Amer. vet med. Ass., 145 (1964) 897-902. 25 Stohr, P., Nervensystem. In W. Mollendorff and W. Bargmann, (Eds.), Handbuch der Mikroskopischen Anatomie des Menschen, Vol. 4, Springer-Verlag, Berlin, 1957, p. 297. 26 Torvik, A., Die Lokalisation des 'Speichelzentrums' bei der Katze, Z. Mikr.-anat. Forsch., 63 (1957) 317-326. 27 Wang, S. C., Localization of the salivatory center in the medulla of the cat, J. Neurophysiol., 6 (1943) 195-202. 28 Yagita, K., Weitere Untersuchungen fiber das Speichelzentrum, Anat. Ariz., 35 (1909) 70-75. 29 Yagita, K. and Hayama, S., Ober des Speichelsekretioncentrum, Neurol. Zbl., 28 (1909) 738-753.