- Email: [email protected]
Autonomic innervation of the tongue: A horseradish peroxidase study in the dog
Jouvnol o? the Autonomic Nervoum System, 2 (1980) 117--129 ~) Elsevier/North-Holland Biomedical Press
117
AUTONOMIC INNERVATION OF THE TONGUE: A H~.,~.~ERADISH PEROXIDASE STUDY IN THE DOG
GREGORY A. CHIBUZO I, JOHN F. CUMMINGS and HOWARD E. EVANS Depar~.m~n: c f Avmtomy, N Y S Coliege o f Veterinary Medicine, Cornell Univereity, Ithaca, N.Y. 14853 (U.8.A.) (Received February 20th, 1980) (Accepted April 22nd, 1980)
Keywords: lingua! autonomic innervation -- horseradish peroxidase -- salivatory nucleus
ABSTRACT
Autonomic ganglia have been found along the lingual nerve in the rostral two-thirds of the canine tongue and along the glossopharyngeal nerve in the caudal glandular third of the tongue [4,17,18]. A 30% horseradish peroxidase (HRP) solution was injected throughout the~.~ gar.glionated areas in order to identify the origin of the pregangliontc fibe~.s to the lingual ganglia. These injections resulted in ipsilateral retrograde labeling of small multipolar neurons in the lateral reticular formation of the medulla oblongata. The same injections labeled neurons in the ipsilat~ral cranial cer;ical ganglion, but preganglionic sympathetic neurons in the thoracic spinal cord were not labeled. These findings indicated that the lingual ganglia consist of parasympathetic neurons which receive preganglionic projections from the medulla. The lingual preganglionic neurons were located within the nucleus reticularis parvicellularis and, in this location, were coextensive with salivatory neurons that labeled after HRP injections ir~ the mandibular and sublingual salivary glands. A degree of somatotopic orgenization within the lingual preganglionic group was indicated by the rescOts of regional injections of enzyme and was confirmed by performing unilateral chorda tympani and glossopharyngeal neurectomies prior to extensive biliateral injections of HRP.
I Present adclrus: Department of Anatomy, School of Vvterinary Medicine, Tuskegee Institute, Ala. 36088, U.S.A.
118
INTRODUCTION
Previous anatomic and physiologic studies on the motor innervation of the tongue have concentrated on the somatic efferent neurons to the lingual muscles [1,12,14,2.t]. The v~:eral motor innervation of the glands and blood vessels of the tongue h ~ received less attention. Histologlc studies of serial sections of the canine brogue have revealed an abundance of lingual ganglia associated with the glossopharyngeal and lingual nerves [4,17,18]. Many of the ganglia along the branches of the glossopharyngeal nerve in the caudal one-third of the tongue are associated closely with the lingual and gustatory glands. This proximity has suggested that the postganglionic fibers subserve v secretomotor fupcti,)n. Glands are absent in the rostral two-thirds of the canine tongue, yet gangiia are encountered frequently here along the ramifications of the lingual nerve. The studies of Brown [3] and Pritchard and Daniel [16] on the dog's tongue have rev¢~!c ~ v a abundance of arterio~enous anastomoses. The an~stomotic segment.~ : , e richly innervated [3] and it may be that the fibers from the more rostral ganglia serve largely in their vasomotor control. The categorization of the autonomic ganglia of the tongue has b,~e.n debated. Despite their peripheral location, some investigators have suggested that they contain sympathetic neurons that have migrated from the cervical ..~ympathetic trunk [18,22]; others have maintained that they are accumula.'.ions of parasympathetic cell bodies [ 6,17 ]. The origin and course of the preganglionic fibers to the lingual ganglia have not been studied experimentally. It has been stated that preganglionic parasympathetic fibers to the tip of the tongue travel via the chorda tympani, and those to the caudal cne-third, via the glossopharyngeal nerve [5]. The visceral efferent fibers in these nerves are said to originate from the superior and inferior salivator,,, nuclei. However, the location of the salivatory neurons and the existence of two discrete nuclei have been the subjects of conside-'able debate [ 8--11,19--21,23,25,26]. The present experimental investigation was undertaken in order to define the autonomic innervation of the tongue. More specifically, the objectives were to determine: (1) the nature of the lingual ganglia (i.e. whether they consis~d of parasympathetic and/or sympathetic neurons); (2) the pre.~ise origin of the preganglionic fibels that project to the lingual ganglia; and (3) the route of these preganglionic axons to the tongue. MATEI:~IALS A N D M E T H O D S
A total of 26 neonatal dogs (2--4 weeks of age) were available for this study. Surgical anesthesia was effected by i.p. injection of fanl~Lnyl citratedroperidol (lnno~,ar Vet Pitman-Moore, Washington Crossing, N.J. 08560, U.S.A.) (0.2--0.4 ml/kg) followed by i.v. administration of pentobarbital sodium (4--3 mg/kg). A 30% aqueous solution of horseradish peroxida:;e (Sigma type IV) was prepared jt st prior to administration. The enzyme w~s
119
delivered by means of a 10/~I I-fatal}tonsyrinse. In the first group of 4 pups 80--160 p l of H R P solution were injected deep to the mucosa at multiple sims along the left half of the tongue. In each of 3 other pups, injections of H R P (80 pl) were confined to the left rostral, middle or caudal one-third of the tongue. The left lingual mucosa was scm'ified with a glass knife prior to topical application of H R P in 6 animals. In 2 animals, the left dorsal aspect of the tongue was scarified before 100 .alof H R P were applied; in the remaining animals, the sc~ication and enzyme applicati.)n (50 #I) were limited to the area of the foliate papillae (2 dog~) and the w~llate papillae (2 dogs). Extrinsic lingual muscles were exposed surgically on the left and injected with 80--130 #I of H R P in a total of 7 animals. Multiple injections of enzyme were restricted to single muscles in 6 of these animals (styloglossus in 2 dogs, hyoglossus in 2 dogs, genioglossus in 2 dogs). All 3 of these muscles were injected in the remaining animal. In two animals, H R P (80 /~I) was injected into the caudal one-third of the tongue (i.e. the region of the lingual glands) on the right after the mandibular and sublingual salivary glands had been exposed and injected with H R P (100 #I) at multiple sites on tke left. In 4 animals, bilateral intralingual injections of H R P (130/~I) were preceded by unilateral cranial nerve transections. In two cases, both the left chorda tympani and glossopharyngeal nerve were transected prior to enzyme injections. In the third animal, the bilateral injections were preceded by transection of the left chorda tympani, and in the fourth by transection of the left glossophalTngeal nerve. After post-injection ~m~rival periods of 24--48 h, the animals were euthanatized by barbiturate overdose and perfused via the left cardiac ventricle with a chilled Karnovsky-type fixative contmning 0.5% paraformaldehyde, 2.5% glutaraldehyde, and 2.5% sucrose in 0.1 M phosphate buffer (pH 7.4). In all cases, the brains were removed promptly, cut transversely into blocks, and stc,red in chilled fixative overnight. The spinal cords and cranial cervical ganglia were also removed from the first group of 4 pups. The brain stem blocks from all animals, the cranial cervical ganglia from the first 4 animals, and cranial thoracic spinal segments from 2 of the~e 4 animals were transferred to 0.1 M phosphate buffer (pH 7.4) prepsrei in a 3 0 % sucrose solution. After these tissues had settled in the sucrose solution, they were removed, frozen, and sectioned serially at 60 /~m. Sections were collected se:ially and the H R P reaction product was demonstrated by treatment with hydrogen peroxide and 3',3'.diaminobenzidine according to the method of Nauta et al. [15]. The sections were mounted and counterstained with cresyl violet. The slides were examined with bright- and dark-field optics. The intralingual dispersion of injected H R P was studied on 60/~m frozen sections of the tongue. These sections were reacted and stained in the same way as the nervous tissues.
120
I(ESULTS
The nature oi ti~e lingual ganglia and the origin of their preganglionic fibers Multiple injections of HRP were m a d e into the left half of the t o n g u e m 4 dogs in an effort to instill e n z y m e t h r o u g h o u t the interior of the t o n g u e . ~ubsequent inspecti(m of serial frozen sections of the tongues revealed diffuse (leposiL~ of e n z y m e reaction p r o d u c t e n c o m p a s s i n g most of the lingual ganglia and glands and mucil of the lingual muscle mass. These injections resulted in consistent retrograde labeling of cell bodies in the ipsilateral cranial cerw, cal ganglion (Fig. 1). The majority of the labeled cell bodies were located in the caudal half of the ganglion. Serial Iransverse frozen secti¢,ns were prepared from the first 4 thoracic spinal .,egm,,nt.q in 2 of the 4 t,,gs. r~:¢aml:,atmn "-" ' " of these spinal sections revealed a ¢omFlete lack of neur(~:,,,l l~bt,,~.g. The absence of label in the zona intermedia of the thoracic segments and the presence of substantial label in the cranial cervical ganglion indicated t h a t the lingual ganglia were n o t in receipt of proganglionic s y m p a t h e t i c fibers, but rather that the s y m p a t h e t i c projections to t h , tongue consist of postganglionic fibers. The preceding observations, t)y elimination, suggested that the lingual ganglia were c o m p o s e d largely, if not exclusively, of p a r a s y m p a t h e t i c neur(ms, l hus, the brain stems of these pups were sectioned serially and
Fig. 1. Dark-field rnierograph showing labeled neuronal cell bodies in the cranial cervical ganglion following multiple HI(P inwetions into tile tongue. 430"~.
121
Fig. 2. Dark-field micrograph s h o w i n g labeled fusiform neurons in the medullary reticular formation after HRP injection into the tongue. 450x.
e x a m i n e d in an e f f o r t to i d e n t i f y the. origin of preganglionic p a r a s y m p a t h e t ' c fibers to t h e t o n g u e . After unilateral injections o f H R P into t h e t o n g u e , labeled s o m a t a a p p e a r e d c o n s i s t e n t l y , as e x p e c t e d , in t h e hypoglo~qal nucleus. A n o t h e r g r o u p of labeled cells, h o w e v e r , was a,-ranged d i f f u s e l y in t h e lateral reticular f o r m a t i o n . T h e ~ , f u s i f o r m or m u l t i p o l a r n e u r o n s (Fig. 2) c o n t a i n e d finer Nissi flak(.s a n d were n o t a b l y smaller than the hypoglossal s o m a t a (Fig. 3). N e u r o n s in these s a m e t w o lo~,ations also labeled w h e n d e e p unilateral inj,.,ctions of e n z y m e were restricted ~(~ the caudal o n e - t h i r d , m i d d l e o n e - t h i r d or rostra~ o n e - t h i r d of the t o n g u e . Superficial applit:ations of H R P to scarified expan.ses ()f t h e lingual m u c o s a in 6 dogs failed to lal)(,I n e u r o n s w i t h i n t h e brain s ~ m . It was c o n t : l u d e d , t h e r e f o r e , t h a t t h e medullary n e u r o n s supplied somali(: or visceral m o t o r fibers to t h e s u b m u c o s : ' l areas of t h e t o n g u e . Ad(litional e x p e r i m e n t s wt,re pt,rforme(l in an e f f o r t to distinguish t h e
122
Fig. 3. Dark-field micrograph of labeled neurons in the hypogl _o$~__1 nucleus following linRual HRP injectior~s. 440x.
somatic motor n,:urons to lingual muscles from the visceral neurons projecting to th ~ ling aal ganglia. In one series of experiments, enzyme injections were confii: ~1 to the extrinsic lingual muscles. These muscles, in contrast to the intrinsic lingual muscles, lack proximate autonomic ganglia, and it was anticipated ~.hat enzyme injections into these muscles would label only somatic motor neurons in the medulla. After multiple injections of enzyme into the individual extrinsic muscles (i.e. the left styloglossus, the left hyoglossus or the left genioglossus), labeled neurons appeared in aggregates in the hypc,gloss~l nucleus. The enzyme injections into the extrinsic muscles differed from those into the tongue in that they failed to produce retrograde labelir~g within the lateral reticular formation of the medulla. These findings suggested that the somatic efferent fibers to lingual mus,~les, both intrinsic and extrinsic, arose from the hypoglossal nucleus, while the lingual ganglia received preganglionic fiber projections from the smaller neurons in the lateral reticular formation. The reticular neurons that labeled after unilateral intralingual injections were situated dorsal to the rostral four-fifths of the ipsilateral facial nucteus and extended over the caudal one-third of the superior olivary nucleus. These small fu~iform or multipolar cells were diffusely arranged over an area that was b o u r d e d laterally by the spinal nucleus of the trigeminal nerve, medially by the nucleus reticularis gigantocellularis and dorsally by the vestibular nucl .'i.
123
Another set of experiments was performed in an effort to define the anatomic relationship between the preganglionic neurons to the lingual gland areas and those supplying the major salivary glands. A prior HRP study in the dog [4] had revealed that the preganglionic parasympathetic fibers to the mandibular (submaxillary) and sublingual salivary glands originated from the ipsilateral lateral reticular formation nf the medulla. In order to achieve a precise topographic comparison between these salivatory neurons and those projecting to the lingual glands, HRP was injected into the gland area of the caudal one-third of the tongue on the right and ~lso into the left mandibular and sublingual salivary glands in two dogs. Left--right comparison of the labcled medullary reticular neurons revealed that the preganglionic neurons to the glands of the tongue, for the most part, were co-extensive with those to the two major salivary glands (Fig. 4). The majority of labeled cells on both sides of the brain stem were located in the nucleus reticularis parvicellularis. Both labeled populations consisted of small fusiform or multipolar neurons (Figs. 2 and 5), yet those labeling as a result of intralingual injections were smaller on the average of 100 cell measurements (29.4 × 21.7 pm) than those labeling after injections into the major salivary glands (38.2 × 21.3 ~m). Some medium sized neurons (55 × 20 /am) also labeled in the area immediately ventral to the medial vestibular nucleus as a result of injections into the nlandibular and sublingual glands. Similar medium-sized neurons were not labeled on the side of the intralingual injections. Although more cell bodies labeled as a result of injections into the major salivary glands on the left (Figs. 2, 4 and 5), the labeled preganglionic neurons on both sides were defined by similar boundaries. Both labeled cell populations shared the same caudal limit (i.e. a point just rostral to the caudal end of the facial nucleus). However, the preganglionic neurons to the major salivary glands extended beyond the rostral end of the facial nucleus and disappeared over the superior olivary nucleus at a point just caudal to the descending root of the facial nerve (Fig. 4). Thus, the preganglionic neurons to the glandular region of the tongue fell caudally within the slightly larger salivatory center containing the preganglionic neurons to the mandibular and sublingual glands, i.e. lhe so-called superior or rostral salivatory nucleus.
Somatotopic organization of the lingual parasympathetic the cranial nerves conveying their preganglionic fibers
neuror~s and
In 3 dogs, deep unilateral injections of enzyme were restricted to the cranial, the middle, or the caudal one-third of the tongue. The labeling of lingual preganglionic neurons in the nucleus reticularis parvicellularis was most abundant as a result of injections into the caudal or glandular one:;hird of the tongue. The labeled neurons in this case, however, did not exter:d to the rostrai end of the facial nucleus as they had after injections that were either limited to or included the rostral one-third of the tongue. In l hese latter instances, the labeled neurons were found rostral to the facial nu(leus and reached the rostral end of the salivatory center. These results suggested a
124
I1
I
t
t
VI
Fig. 4. D!agrammatic illustrations of brain s t e m sections to show the comparative distribution of labeled cell bodies (solid dots) in the medullary reticular f o r m a t i o n following caudal intralingual HRP injections on the right (R) and injections o f H R P into the man-
125
Fig. 5. Dark-field micrograph showir.~ labeled fusiform neuronal cell bodies in the medullary reticular formation following rr.andibular-sublingual HRP injections. degree of rostrocaudal somatotopic organization of the lingual preganglionic neurons. Additional experiments were performed to explore f,~ther the suggested s o m a t o t o p y and also to determine the routes of the preganglionic fibers to the tongue. In these experiments, bilaterally symmetrical intralingual injections of HRP were preceded by selective unilateral neurectomies. '~ ,~; extent of preganglionic neuronal labeling on the operated and unoperated si0~s was then contrasted on serial sections of the medulla. In one dog, bilateral deep injections of HRP along the entire length of the tongue were preceded by transection of the left chorda tympani. In this case, the labeled preganglionic neurons on the operated side were n o t found rostral to the facial nucleus, whereas those on the intact side e x t e n d e d over the caudal one-third of the superior olivary nucleus. The caudal extent of the labeled l~eurons was similar on both sides of the medulla. When left glossopharyngeal n e u r e c t o m y preceded bilateral intralingual injections dibular and sublingual salivary glands on the left (L). The brain stem sections are arranged in a caudal (A) to rostral (J) .sequence. Abbreviations: NVC, caudal vestibular nucleus; NVM, medial vestibular nucleus; NCL, lateral vestibular nucleus; RVIII, vestibulo-c(~chlear nerve; NRV, nucleus and spinal tract of trigeminal nerve; NVII, facial nucleus; PY, pyramid; NVIII, cochlear nucleus; SO, superior olivary nucleus; G, genu of facial ne'..~e; NVI, abducens nucleus; RVII, descending root of facial nerve; TB, trapezoid t;ody.
126 of the enzyme, the lingual neurons at the rostral end of the salivatory center were labeled bilaterally; however, the labeling on the operated side did n o t extend as far caudally over the facial nucleus as it did on the intact side. When both the left chorda tympani and glossopharyngeal nerve were transected prior to bilateral HRP injections, lingual preganglionic neurons appeared to label in usual abundance in the salivatory center on the right, but were completely absent on the left. From these experiments it was concluded that the preganglionic neurons t o the tongue project from the nucleus reticularis parvicellularis to the tongue via the chorda tympani and glossopharyngeal nerves. Although tbe cell bodies of origin of the chorda tympani and glossopharyngeal fibers were overlapped throughout much of this r, ucleus, the most rostral of the lingual preganglionic neurons projected exclusively ~Ja the chorda tympani and the most caudal ~xclusively via the glossopharyngeal nerve.
DISCUSSION In the present study, retrograde transport of H R P from sites throughout the ganglionated areas of the tongue afforded the means to clarify the nature of the lingual ganglia by defining the origin of their preganglionic fibers. The classification of the lingual ganglia along the ramifications ~f the lingual and glossopharyngeai nerves has been subject to vario,Js interpretations. The ganglia accompanying the bl~mches of the glossopha.'yngeal nerve in the caudal one-third of the tongue have been perceived commonly as a source of secretory motor fibers to the lingual glands [18]. This view was supported by Simonetta's extirpatlon of seemingly homologous ganglia in the sheep and by the proximity of these ganglia to tlte lingual glands [18]. Most previous investigations of the lingual ganglia, including Sarbach's study on the distribution of the ganglia in the tongue of the dog, cat, guinea pig, rat and rabbit, have been descriptive rather than experimental. As such, these studies lacked a means of determining whether the ganglia along lingual and glossopharyngeal nerves in the tongue were composed of sympathetic [18~2] or parasympathetic [6,17] cell bodies. Since intralingual injections of enzyme labeled postganglionic sympathetic neurons in the crania} cervical ganglion, but failed to label any preganglionic neurons in the thoracic spinal cord, it seemed unlikely that the lingualganglia were sympathetic in nature. If, as the suggested alternative, the ganglia were parasympathetic, the same enzyme injections would be expected to label preganglionic neurons in the brain stem. These injections, in fact, resulted in consistent ipsilateral neuronal labeling in the hypoglossal nucleus, and the parvicellular reticular nucleus. Since the injected enzyme had diffused widely within the tongue, its transport from somatic motor mid preganglionic terminals would result in "etrograde labeling of both somatic and visc~ra} neuronal groups in the brain stem. Retrograde labeling of hypoglossal neurons as a result of intralinguaJ injections of H R P had been demonstrated
127 previously [12,13]. In an effort to distinguish clearly the origins of the somatic and visceral motor projections to the tongue, enzyme was injected into the extrinsic lingual muscles. Since the extrinsic muscles lacked associated autonomic ganglia, itwas assumed that these injections would label only the somatic motor neurons in the medulla. The injections of the extrinsic muscles resembled the intralingual injections in that they also produced labeling of many hypoglossal neurons. However, they differed it, that they failed to label any neurons within the nucleus re~icularis parvicellularis. On comparing and contrasting the~e results, it was concluded that somatic efferent projections to the lingual muscle fibers originated from the hypoglossal nucleus, and that the lingual ganglia receive preganglionic parasympathetic projections from the nucleus reticularis parvicellularis. Textbooks state that parasympathetic preganglionic fibers ploiect via the chorda tympani to the rostral two-thirds of the tongue and via the glossopharyngeal nerve to the caudal one-third, and these lingual preganglionic fibers have been assumed to originate from the superior and inferior salivatory nuclei [2,5]. Yet attempts to locate these ~alivatory nuclei by means of induced retrograde chromatolysis have }ed to widely varying results. Kohnstamm's early investigations in the dog indicated that there were superior and inferior salivatory nuclei that formed mainly contralateral projections from the brain stem [10,11]. The superior nucleus consisted of large motor neurons that extended as far as the rostral pole of the motor nucleus of the tTigeminal nerve; the inferior nucleus consisted of smaller neurons at the level of the caudal olivary nucleus. Solomowicz, however, found that the preganglionic fibers to the mandibular salivary gland in the dog originated bilaterally from large neurons in the region of the lateral vestibular nucleus [21]. Yagita and Hsyama identified the superior salivatory nucleus as a loose group of small neurons situated in the medullary reticular formation at the level of the facial nucleus [26]. Yagita later described the inferior ~iivatory nucleus as a direct caudal continuation of the superior nucleus [25]. Torvik failed to find two discrete salivary nuclei in the kitten [ 23]. He concluded that the preganglionic facial and glossopharyngeal fibers arose ipsilaterally from a single loose aggregate of neurons in the nucleus reticularis parvicellularis. Using a modified Koelle technique, Schute and Lewis disputed the findings based on retrograde cell change and identified salivatory nuclei medial and lateral to the genu of the facial nerve root [20]. Recent HRP investigations on the innervation of the mandibular and sublingual gland~ in the rat [8], cat [19], and dog [4] confirmed Torvik's location of t~ e salivatory center in the nucleus reticularis parvicellularis. The present study reaffirmed the parvicellular reticular nucleus as a salivary center and demonstrated that the origin of the preganglionic fibers to the lingual ganglia also lies within this nucleus. These lingual preganglionic neurons, although slightly smal'er on the average and fewer, were co,extensive with the preganglionic neurons to the mandibular and sublingual glands. A degree of somatotopic organization of the lingual preganglionic neurons was detected through experiments in which unilateral chorda tympani
128 or glossopharyngeal nerve transection preceded bilateral intraling, ml injections of HRP. The results indicated that the most rostral lingual preganglionic neurons (i.e. those rostral to the facial nucleus) projected to the tongue via the chorda tympani and the most caudal neurons (i.e. those closest to the caudal pole of the facial nucleus) projected via the glossopharyngeal nerve. Between these extremes, however, the lingual preganglionic neurons were not clearly separable into glossopharyngeal and chorda tympani groups. The preceding results also seemed pertinent to the question of two discrete salivatory nuclei. Torvik, in contrast to earlier workers, had observed that the preganglionic projections through the chorda tympani and glossopharyngeal nerves arose from a diffuse but confluent group of cells in the nucleus reticularis parvicellularis [23]. The present study supported this observation in specific regard to the lingual preganglionic neurons. These formed a short, diffuse, but uninterrupted column in which the chorda tympani and glossopharyngeal units were overlapped except at the rostral and caudal extremes. While the present study indicated that the lingual ganglia contain parasympathetic neurons which receive preganglionic projections from the nucleus reticularis parvicellularis via the chorda tympani and glossopharyngeal nerves, the functions served by th~se autonomic neurons have not been resolved. As noted previously, the proximity of the caudal (glossopharyngeal) ganglia to the lingual glands has suggested a secretory motor function. The effectors se~wed by the ganglia in the non-glandular, rostral two.thirds of the tongue are a matter of greater speculation. In the absence of glands, vessels appear as the predominant, autonomically innervated effectors. Prior studies have called attention to the abundance of richly innervated arteriovenous anastomoses in the canine tongue and their possible role in thermoregulation [3,16]. It seems, therefore, reasonable to speculate that the arteriovenous anastomotic segments receive substantial postganglionic projections from the lingual ganglia. This speculation is in accord with Hellekant's [7] recent demonstration of lingual vasodilator fibers within the chorda tympani of the rat. ACKNOWLEDGEMENTS
This research was supported by NIH Grant 5-F34-FM06013-02. The authors thank Drs. A. de Lahunta, B.P. Halpern and J.M. Petras for their valuable directions during the experiment and for editing the manuscript. The authors are also grateful to Mr. W.P. Hamilton IV for the illustrations, to Dr. Isak Foss for his surgical assistance, to Mrs. Helen Lehman and Miss Barbara Loman for their technical help, and to Miss Alice CkLwan for secretarial support. REFERENCES
1 Barnard, J.W., The hypoglossalcomplex of v,~.rtebrates, J. comp. Neuro]., 72 (1940) 489--524,
129 2 Brodal, A., Neurological Anatomy, Oxford Universit) [:÷_-% London, 1969. 3 Brown, M.E., The occurrence of arterio-venous anastomoses in the tongue of the dog, Anat. Rec., 69 (1937) 287--292. 4 Chibuzo, G A., The Location of Cell Bodies of Primary Motor and Ser, sory Neurons that Innervate the Tongue of the Dog, Ph.D. Thesis, Cornell University, Ithaca, 1979. 5 Crosby, E.C., Humphrey, T. and Lauer, E.W., Correlative Anatomy o the Nervous System, MacMillan, New York, 1962. 6 Gomez, H., The innervation of lingual salivary glands, Anat. Rec., 139 (I 961 ) 69--76. 7 Hellekant, G., Vasodilator fibers to the tongue in the chorda tympani proper nerve, Acta physiol, scand., 99 (1977) 292--299. 8 Hiura, T., Sa]ivatory neurons innervate the submandibular and sublingual glands in the rat: horseradish peroxidase study, Brain Res., 137 (1977) 145--149. 9 Kohrustamm, O., Der Nucleus salivatorius Chorda tympani (Nervi intermedii), Anat. Anz., 21 (1902) 362--363. I0 Kohnstamm, O., Der Nucleus salivatorius und das cranioviscerale System (mit Demonstration), Arch. Psychiat. Nervenkr., 37 (1903) 660 G62. 11 K o h n s t ~ . O., Der Nucleus salivatorius inferior und das cranio-viscerale System, Neurol. Zbl., ~2 (1903) 699. 12 Kristensson, K and Olsson, Y., Uptake and retrograde axona] transport of protein tracers in hypo.,flossal neurons: fate of the tracer and reaction of the nerve cell bodies, Acta neuropatl~. (Berl.), 23 (1973) 43--47. 13 Kristensson, K., Olsson, Y. and Sj~strand, J., Axonal uptake and retrograde transport of exogenous proteins in the hypoglossal nerve, Brain Res., 32 (1971) 399--406. 14 Kristensson, K. and Sj6strand, J., Retrograde transport of protein tracer in the rabbit hypoglo~al nerve during regenermtir,,, Brain Res., 45 (197~.) 175--181. 15 Nauta, H.J.W., Pritz, M.B. and Llsek, R.J., .~tterents to the rat caudo-putamen studied with horseradish peroxidas..~. An evaluation of a retrograde neuroanatomical research method, Brain Res., 67 (1974) 219--238. 16 Pritchard, M.M.L. and Daniel. P.M., Arterio-venous anastomoses in the tongue of the dog, J. Anat. (Lond.), 87 (1953) 66--74. 17 Rakhawy, M.T., Phosphatases in the nervous tissue (the nature of the ganglionic nerve cells in the tongue), Acta Anat. (Basel), 83 (1972) 356--366. ~ Sarbach, K., De la r~partition de~ ganglions nervcux dans la langue de rongeurs et de carnivores, Schwelz. Mschr. Zahnheilk., 65 (1955) 1--17. !9 Satomi, H., Takahashi, K., Ise, H. and Yamamoto, T., Identification of the superior salivatory nucleus in the cat as studied by the HRP method, Neurosci. Lett. 14 (1979) 135--139. 20 Shute, C.C.D. and Lewis, P.R., The salivatory center in the rat, J. Anat. (Lond.), 94 (1960) 59--81. 21 Solomowicz, J., Vom Centrum der Submaxillardr/ise, Neurol. Zbl., 27 (1908) 724--727. 22 Stewart, F.W., The development of the cranial sympathetic ganglia in rat, J. comp. Neurol., 31 (1920) 163--218. 23 Torvik, A., Die Lokalisation des Speichelzertrum bei der Katze, Z. mikr.-anat. Forsch., 63 (1957) 317--326. 24 Wstkins, D.W, Hypoglossal nucleus: localizatio,~ of motoneuronal pools, Anat. Rec., 190 (1978) 5"6---577. 25 Ya~i,~, K., W.-~itere Untersuchungen fiber das So..~ichelzentrum, Anat. Ariz., 35 (1909) 70--'!5. 26 Ysbi'-a, K. ard ,ttayama, S., Uber des Speichelsekretioncentrum, Neurol. Zbl., 28 (1909) 738--~ 53.
117
AUTONOMIC INNERVATION OF THE TONGUE: A H~.,~.~ERADISH PEROXIDASE STUDY IN THE DOG
GREGORY A. CHIBUZO I, JOHN F. CUMMINGS and HOWARD E. EVANS Depar~.m~n: c f Avmtomy, N Y S Coliege o f Veterinary Medicine, Cornell Univereity, Ithaca, N.Y. 14853 (U.8.A.) (Received February 20th, 1980) (Accepted April 22nd, 1980)
Keywords: lingua! autonomic innervation -- horseradish peroxidase -- salivatory nucleus
ABSTRACT
Autonomic ganglia have been found along the lingual nerve in the rostral two-thirds of the canine tongue and along the glossopharyngeal nerve in the caudal glandular third of the tongue [4,17,18]. A 30% horseradish peroxidase (HRP) solution was injected throughout the~.~ gar.glionated areas in order to identify the origin of the pregangliontc fibe~.s to the lingual ganglia. These injections resulted in ipsilateral retrograde labeling of small multipolar neurons in the lateral reticular formation of the medulla oblongata. The same injections labeled neurons in the ipsilat~ral cranial cer;ical ganglion, but preganglionic sympathetic neurons in the thoracic spinal cord were not labeled. These findings indicated that the lingual ganglia consist of parasympathetic neurons which receive preganglionic projections from the medulla. The lingual preganglionic neurons were located within the nucleus reticularis parvicellularis and, in this location, were coextensive with salivatory neurons that labeled after HRP injections ir~ the mandibular and sublingual salivary glands. A degree of somatotopic orgenization within the lingual preganglionic group was indicated by the rescOts of regional injections of enzyme and was confirmed by performing unilateral chorda tympani and glossopharyngeal neurectomies prior to extensive biliateral injections of HRP.
I Present adclrus: Department of Anatomy, School of Vvterinary Medicine, Tuskegee Institute, Ala. 36088, U.S.A.
118
INTRODUCTION
Previous anatomic and physiologic studies on the motor innervation of the tongue have concentrated on the somatic efferent neurons to the lingual muscles [1,12,14,2.t]. The v~:eral motor innervation of the glands and blood vessels of the tongue h ~ received less attention. Histologlc studies of serial sections of the canine brogue have revealed an abundance of lingual ganglia associated with the glossopharyngeal and lingual nerves [4,17,18]. Many of the ganglia along the branches of the glossopharyngeal nerve in the caudal one-third of the tongue are associated closely with the lingual and gustatory glands. This proximity has suggested that the postganglionic fibers subserve v secretomotor fupcti,)n. Glands are absent in the rostral two-thirds of the canine tongue, yet gangiia are encountered frequently here along the ramifications of the lingual nerve. The studies of Brown [3] and Pritchard and Daniel [16] on the dog's tongue have rev¢~!c ~ v a abundance of arterio~enous anastomoses. The an~stomotic segment.~ : , e richly innervated [3] and it may be that the fibers from the more rostral ganglia serve largely in their vasomotor control. The categorization of the autonomic ganglia of the tongue has b,~e.n debated. Despite their peripheral location, some investigators have suggested that they contain sympathetic neurons that have migrated from the cervical ..~ympathetic trunk [18,22]; others have maintained that they are accumula.'.ions of parasympathetic cell bodies [ 6,17 ]. The origin and course of the preganglionic fibers to the lingual ganglia have not been studied experimentally. It has been stated that preganglionic parasympathetic fibers to the tip of the tongue travel via the chorda tympani, and those to the caudal cne-third, via the glossopharyngeal nerve [5]. The visceral efferent fibers in these nerves are said to originate from the superior and inferior salivator,,, nuclei. However, the location of the salivatory neurons and the existence of two discrete nuclei have been the subjects of conside-'able debate [ 8--11,19--21,23,25,26]. The present experimental investigation was undertaken in order to define the autonomic innervation of the tongue. More specifically, the objectives were to determine: (1) the nature of the lingual ganglia (i.e. whether they consis~d of parasympathetic and/or sympathetic neurons); (2) the pre.~ise origin of the preganglionic fibels that project to the lingual ganglia; and (3) the route of these preganglionic axons to the tongue. MATEI:~IALS A N D M E T H O D S
A total of 26 neonatal dogs (2--4 weeks of age) were available for this study. Surgical anesthesia was effected by i.p. injection of fanl~Lnyl citratedroperidol (lnno~,ar Vet Pitman-Moore, Washington Crossing, N.J. 08560, U.S.A.) (0.2--0.4 ml/kg) followed by i.v. administration of pentobarbital sodium (4--3 mg/kg). A 30% aqueous solution of horseradish peroxida:;e (Sigma type IV) was prepared jt st prior to administration. The enzyme w~s
119
delivered by means of a 10/~I I-fatal}tonsyrinse. In the first group of 4 pups 80--160 p l of H R P solution were injected deep to the mucosa at multiple sims along the left half of the tongue. In each of 3 other pups, injections of H R P (80 pl) were confined to the left rostral, middle or caudal one-third of the tongue. The left lingual mucosa was scm'ified with a glass knife prior to topical application of H R P in 6 animals. In 2 animals, the left dorsal aspect of the tongue was scarified before 100 .alof H R P were applied; in the remaining animals, the sc~ication and enzyme applicati.)n (50 #I) were limited to the area of the foliate papillae (2 dog~) and the w~llate papillae (2 dogs). Extrinsic lingual muscles were exposed surgically on the left and injected with 80--130 #I of H R P in a total of 7 animals. Multiple injections of enzyme were restricted to single muscles in 6 of these animals (styloglossus in 2 dogs, hyoglossus in 2 dogs, genioglossus in 2 dogs). All 3 of these muscles were injected in the remaining animal. In two animals, H R P (80 /~I) was injected into the caudal one-third of the tongue (i.e. the region of the lingual glands) on the right after the mandibular and sublingual salivary glands had been exposed and injected with H R P (100 #I) at multiple sites on tke left. In 4 animals, bilateral intralingual injections of H R P (130/~I) were preceded by unilateral cranial nerve transections. In two cases, both the left chorda tympani and glossopharyngeal nerve were transected prior to enzyme injections. In the third animal, the bilateral injections were preceded by transection of the left chorda tympani, and in the fourth by transection of the left glossophalTngeal nerve. After post-injection ~m~rival periods of 24--48 h, the animals were euthanatized by barbiturate overdose and perfused via the left cardiac ventricle with a chilled Karnovsky-type fixative contmning 0.5% paraformaldehyde, 2.5% glutaraldehyde, and 2.5% sucrose in 0.1 M phosphate buffer (pH 7.4). In all cases, the brains were removed promptly, cut transversely into blocks, and stc,red in chilled fixative overnight. The spinal cords and cranial cervical ganglia were also removed from the first group of 4 pups. The brain stem blocks from all animals, the cranial cervical ganglia from the first 4 animals, and cranial thoracic spinal segments from 2 of the~e 4 animals were transferred to 0.1 M phosphate buffer (pH 7.4) prepsrei in a 3 0 % sucrose solution. After these tissues had settled in the sucrose solution, they were removed, frozen, and sectioned serially at 60 /~m. Sections were collected se:ially and the H R P reaction product was demonstrated by treatment with hydrogen peroxide and 3',3'.diaminobenzidine according to the method of Nauta et al. [15]. The sections were mounted and counterstained with cresyl violet. The slides were examined with bright- and dark-field optics. The intralingual dispersion of injected H R P was studied on 60/~m frozen sections of the tongue. These sections were reacted and stained in the same way as the nervous tissues.
120
I(ESULTS
The nature oi ti~e lingual ganglia and the origin of their preganglionic fibers Multiple injections of HRP were m a d e into the left half of the t o n g u e m 4 dogs in an effort to instill e n z y m e t h r o u g h o u t the interior of the t o n g u e . ~ubsequent inspecti(m of serial frozen sections of the tongues revealed diffuse (leposiL~ of e n z y m e reaction p r o d u c t e n c o m p a s s i n g most of the lingual ganglia and glands and mucil of the lingual muscle mass. These injections resulted in consistent retrograde labeling of cell bodies in the ipsilateral cranial cerw, cal ganglion (Fig. 1). The majority of the labeled cell bodies were located in the caudal half of the ganglion. Serial Iransverse frozen secti¢,ns were prepared from the first 4 thoracic spinal .,egm,,nt.q in 2 of the 4 t,,gs. r~:¢aml:,atmn "-" ' " of these spinal sections revealed a ¢omFlete lack of neur(~:,,,l l~bt,,~.g. The absence of label in the zona intermedia of the thoracic segments and the presence of substantial label in the cranial cervical ganglion indicated t h a t the lingual ganglia were n o t in receipt of proganglionic s y m p a t h e t i c fibers, but rather that the s y m p a t h e t i c projections to t h , tongue consist of postganglionic fibers. The preceding observations, t)y elimination, suggested that the lingual ganglia were c o m p o s e d largely, if not exclusively, of p a r a s y m p a t h e t i c neur(ms, l hus, the brain stems of these pups were sectioned serially and
Fig. 1. Dark-field rnierograph showing labeled neuronal cell bodies in the cranial cervical ganglion following multiple HI(P inwetions into tile tongue. 430"~.
121
Fig. 2. Dark-field micrograph s h o w i n g labeled fusiform neurons in the medullary reticular formation after HRP injection into the tongue. 450x.
e x a m i n e d in an e f f o r t to i d e n t i f y the. origin of preganglionic p a r a s y m p a t h e t ' c fibers to t h e t o n g u e . After unilateral injections o f H R P into t h e t o n g u e , labeled s o m a t a a p p e a r e d c o n s i s t e n t l y , as e x p e c t e d , in t h e hypoglo~qal nucleus. A n o t h e r g r o u p of labeled cells, h o w e v e r , was a,-ranged d i f f u s e l y in t h e lateral reticular f o r m a t i o n . T h e ~ , f u s i f o r m or m u l t i p o l a r n e u r o n s (Fig. 2) c o n t a i n e d finer Nissi flak(.s a n d were n o t a b l y smaller than the hypoglossal s o m a t a (Fig. 3). N e u r o n s in these s a m e t w o lo~,ations also labeled w h e n d e e p unilateral inj,.,ctions of e n z y m e were restricted ~(~ the caudal o n e - t h i r d , m i d d l e o n e - t h i r d or rostra~ o n e - t h i r d of the t o n g u e . Superficial applit:ations of H R P to scarified expan.ses ()f t h e lingual m u c o s a in 6 dogs failed to lal)(,I n e u r o n s w i t h i n t h e brain s ~ m . It was c o n t : l u d e d , t h e r e f o r e , t h a t t h e medullary n e u r o n s supplied somali(: or visceral m o t o r fibers to t h e s u b m u c o s : ' l areas of t h e t o n g u e . Ad(litional e x p e r i m e n t s wt,re pt,rforme(l in an e f f o r t to distinguish t h e
122
Fig. 3. Dark-field micrograph of labeled neurons in the hypogl _o$~__1 nucleus following linRual HRP injectior~s. 440x.
somatic motor n,:urons to lingual muscles from the visceral neurons projecting to th ~ ling aal ganglia. In one series of experiments, enzyme injections were confii: ~1 to the extrinsic lingual muscles. These muscles, in contrast to the intrinsic lingual muscles, lack proximate autonomic ganglia, and it was anticipated ~.hat enzyme injections into these muscles would label only somatic motor neurons in the medulla. After multiple injections of enzyme into the individual extrinsic muscles (i.e. the left styloglossus, the left hyoglossus or the left genioglossus), labeled neurons appeared in aggregates in the hypc,gloss~l nucleus. The enzyme injections into the extrinsic muscles differed from those into the tongue in that they failed to produce retrograde labelir~g within the lateral reticular formation of the medulla. These findings suggested that the somatic efferent fibers to lingual mus,~les, both intrinsic and extrinsic, arose from the hypoglossal nucleus, while the lingual ganglia received preganglionic fiber projections from the smaller neurons in the lateral reticular formation. The reticular neurons that labeled after unilateral intralingual injections were situated dorsal to the rostral four-fifths of the ipsilateral facial nucteus and extended over the caudal one-third of the superior olivary nucleus. These small fu~iform or multipolar cells were diffusely arranged over an area that was b o u r d e d laterally by the spinal nucleus of the trigeminal nerve, medially by the nucleus reticularis gigantocellularis and dorsally by the vestibular nucl .'i.
123
Another set of experiments was performed in an effort to define the anatomic relationship between the preganglionic neurons to the lingual gland areas and those supplying the major salivary glands. A prior HRP study in the dog [4] had revealed that the preganglionic parasympathetic fibers to the mandibular (submaxillary) and sublingual salivary glands originated from the ipsilateral lateral reticular formation nf the medulla. In order to achieve a precise topographic comparison between these salivatory neurons and those projecting to the lingual glands, HRP was injected into the gland area of the caudal one-third of the tongue on the right and ~lso into the left mandibular and sublingual salivary glands in two dogs. Left--right comparison of the labcled medullary reticular neurons revealed that the preganglionic neurons to the glands of the tongue, for the most part, were co-extensive with those to the two major salivary glands (Fig. 4). The majority of labeled cells on both sides of the brain stem were located in the nucleus reticularis parvicellularis. Both labeled populations consisted of small fusiform or multipolar neurons (Figs. 2 and 5), yet those labeling as a result of intralingual injections were smaller on the average of 100 cell measurements (29.4 × 21.7 pm) than those labeling after injections into the major salivary glands (38.2 × 21.3 ~m). Some medium sized neurons (55 × 20 /am) also labeled in the area immediately ventral to the medial vestibular nucleus as a result of injections into the nlandibular and sublingual glands. Similar medium-sized neurons were not labeled on the side of the intralingual injections. Although more cell bodies labeled as a result of injections into the major salivary glands on the left (Figs. 2, 4 and 5), the labeled preganglionic neurons on both sides were defined by similar boundaries. Both labeled cell populations shared the same caudal limit (i.e. a point just rostral to the caudal end of the facial nucleus). However, the preganglionic neurons to the major salivary glands extended beyond the rostral end of the facial nucleus and disappeared over the superior olivary nucleus at a point just caudal to the descending root of the facial nerve (Fig. 4). Thus, the preganglionic neurons to the glandular region of the tongue fell caudally within the slightly larger salivatory center containing the preganglionic neurons to the mandibular and sublingual glands, i.e. lhe so-called superior or rostral salivatory nucleus.
Somatotopic organization of the lingual parasympathetic the cranial nerves conveying their preganglionic fibers
neuror~s and
In 3 dogs, deep unilateral injections of enzyme were restricted to the cranial, the middle, or the caudal one-third of the tongue. The labeling of lingual preganglionic neurons in the nucleus reticularis parvicellularis was most abundant as a result of injections into the caudal or glandular one:;hird of the tongue. The labeled neurons in this case, however, did not exter:d to the rostrai end of the facial nucleus as they had after injections that were either limited to or included the rostral one-third of the tongue. In l hese latter instances, the labeled neurons were found rostral to the facial nu(leus and reached the rostral end of the salivatory center. These results suggested a
124
I1
I
t
t
VI
Fig. 4. D!agrammatic illustrations of brain s t e m sections to show the comparative distribution of labeled cell bodies (solid dots) in the medullary reticular f o r m a t i o n following caudal intralingual HRP injections on the right (R) and injections o f H R P into the man-
125
Fig. 5. Dark-field micrograph showir.~ labeled fusiform neuronal cell bodies in the medullary reticular formation following rr.andibular-sublingual HRP injections. degree of rostrocaudal somatotopic organization of the lingual preganglionic neurons. Additional experiments were performed to explore f,~ther the suggested s o m a t o t o p y and also to determine the routes of the preganglionic fibers to the tongue. In these experiments, bilaterally symmetrical intralingual injections of HRP were preceded by selective unilateral neurectomies. '~ ,~; extent of preganglionic neuronal labeling on the operated and unoperated si0~s was then contrasted on serial sections of the medulla. In one dog, bilateral deep injections of HRP along the entire length of the tongue were preceded by transection of the left chorda tympani. In this case, the labeled preganglionic neurons on the operated side were n o t found rostral to the facial nucleus, whereas those on the intact side e x t e n d e d over the caudal one-third of the superior olivary nucleus. The caudal extent of the labeled l~eurons was similar on both sides of the medulla. When left glossopharyngeal n e u r e c t o m y preceded bilateral intralingual injections dibular and sublingual salivary glands on the left (L). The brain stem sections are arranged in a caudal (A) to rostral (J) .sequence. Abbreviations: NVC, caudal vestibular nucleus; NVM, medial vestibular nucleus; NCL, lateral vestibular nucleus; RVIII, vestibulo-c(~chlear nerve; NRV, nucleus and spinal tract of trigeminal nerve; NVII, facial nucleus; PY, pyramid; NVIII, cochlear nucleus; SO, superior olivary nucleus; G, genu of facial ne'..~e; NVI, abducens nucleus; RVII, descending root of facial nerve; TB, trapezoid t;ody.
126 of the enzyme, the lingual neurons at the rostral end of the salivatory center were labeled bilaterally; however, the labeling on the operated side did n o t extend as far caudally over the facial nucleus as it did on the intact side. When both the left chorda tympani and glossopharyngeal nerve were transected prior to bilateral HRP injections, lingual preganglionic neurons appeared to label in usual abundance in the salivatory center on the right, but were completely absent on the left. From these experiments it was concluded that the preganglionic neurons t o the tongue project from the nucleus reticularis parvicellularis to the tongue via the chorda tympani and glossopharyngeal nerves. Although tbe cell bodies of origin of the chorda tympani and glossopharyngeal fibers were overlapped throughout much of this r, ucleus, the most rostral of the lingual preganglionic neurons projected exclusively ~Ja the chorda tympani and the most caudal ~xclusively via the glossopharyngeal nerve.
DISCUSSION In the present study, retrograde transport of H R P from sites throughout the ganglionated areas of the tongue afforded the means to clarify the nature of the lingual ganglia by defining the origin of their preganglionic fibers. The classification of the lingual ganglia along the ramifications ~f the lingual and glossopharyngeai nerves has been subject to vario,Js interpretations. The ganglia accompanying the bl~mches of the glossopha.'yngeal nerve in the caudal one-third of the tongue have been perceived commonly as a source of secretory motor fibers to the lingual glands [18]. This view was supported by Simonetta's extirpatlon of seemingly homologous ganglia in the sheep and by the proximity of these ganglia to tlte lingual glands [18]. Most previous investigations of the lingual ganglia, including Sarbach's study on the distribution of the ganglia in the tongue of the dog, cat, guinea pig, rat and rabbit, have been descriptive rather than experimental. As such, these studies lacked a means of determining whether the ganglia along lingual and glossopharyngeal nerves in the tongue were composed of sympathetic [18~2] or parasympathetic [6,17] cell bodies. Since intralingual injections of enzyme labeled postganglionic sympathetic neurons in the crania} cervical ganglion, but failed to label any preganglionic neurons in the thoracic spinal cord, it seemed unlikely that the lingualganglia were sympathetic in nature. If, as the suggested alternative, the ganglia were parasympathetic, the same enzyme injections would be expected to label preganglionic neurons in the brain stem. These injections, in fact, resulted in consistent ipsilateral neuronal labeling in the hypoglossal nucleus, and the parvicellular reticular nucleus. Since the injected enzyme had diffused widely within the tongue, its transport from somatic motor mid preganglionic terminals would result in "etrograde labeling of both somatic and visc~ra} neuronal groups in the brain stem. Retrograde labeling of hypoglossal neurons as a result of intralinguaJ injections of H R P had been demonstrated
127 previously [12,13]. In an effort to distinguish clearly the origins of the somatic and visceral motor projections to the tongue, enzyme was injected into the extrinsic lingual muscles. Since the extrinsic muscles lacked associated autonomic ganglia, itwas assumed that these injections would label only the somatic motor neurons in the medulla. The injections of the extrinsic muscles resembled the intralingual injections in that they also produced labeling of many hypoglossal neurons. However, they differed it, that they failed to label any neurons within the nucleus re~icularis parvicellularis. On comparing and contrasting the~e results, it was concluded that somatic efferent projections to the lingual muscle fibers originated from the hypoglossal nucleus, and that the lingual ganglia receive preganglionic parasympathetic projections from the nucleus reticularis parvicellularis. Textbooks state that parasympathetic preganglionic fibers ploiect via the chorda tympani to the rostral two-thirds of the tongue and via the glossopharyngeal nerve to the caudal one-third, and these lingual preganglionic fibers have been assumed to originate from the superior and inferior salivatory nuclei [2,5]. Yet attempts to locate these ~alivatory nuclei by means of induced retrograde chromatolysis have }ed to widely varying results. Kohnstamm's early investigations in the dog indicated that there were superior and inferior salivatory nuclei that formed mainly contralateral projections from the brain stem [10,11]. The superior nucleus consisted of large motor neurons that extended as far as the rostral pole of the motor nucleus of the tTigeminal nerve; the inferior nucleus consisted of smaller neurons at the level of the caudal olivary nucleus. Solomowicz, however, found that the preganglionic fibers to the mandibular salivary gland in the dog originated bilaterally from large neurons in the region of the lateral vestibular nucleus [21]. Yagita and Hsyama identified the superior salivatory nucleus as a loose group of small neurons situated in the medullary reticular formation at the level of the facial nucleus [26]. Yagita later described the inferior ~iivatory nucleus as a direct caudal continuation of the superior nucleus [25]. Torvik failed to find two discrete salivary nuclei in the kitten [ 23]. He concluded that the preganglionic facial and glossopharyngeal fibers arose ipsilaterally from a single loose aggregate of neurons in the nucleus reticularis parvicellularis. Using a modified Koelle technique, Schute and Lewis disputed the findings based on retrograde cell change and identified salivatory nuclei medial and lateral to the genu of the facial nerve root [20]. Recent HRP investigations on the innervation of the mandibular and sublingual gland~ in the rat [8], cat [19], and dog [4] confirmed Torvik's location of t~ e salivatory center in the nucleus reticularis parvicellularis. The present study reaffirmed the parvicellular reticular nucleus as a salivary center and demonstrated that the origin of the preganglionic fibers to the lingual ganglia also lies within this nucleus. These lingual preganglionic neurons, although slightly smal'er on the average and fewer, were co,extensive with the preganglionic neurons to the mandibular and sublingual glands. A degree of somatotopic organization of the lingual preganglionic neurons was detected through experiments in which unilateral chorda tympani
128 or glossopharyngeal nerve transection preceded bilateral intraling, ml injections of HRP. The results indicated that the most rostral lingual preganglionic neurons (i.e. those rostral to the facial nucleus) projected to the tongue via the chorda tympani and the most caudal neurons (i.e. those closest to the caudal pole of the facial nucleus) projected via the glossopharyngeal nerve. Between these extremes, however, the lingual preganglionic neurons were not clearly separable into glossopharyngeal and chorda tympani groups. The preceding results also seemed pertinent to the question of two discrete salivatory nuclei. Torvik, in contrast to earlier workers, had observed that the preganglionic projections through the chorda tympani and glossopharyngeal nerves arose from a diffuse but confluent group of cells in the nucleus reticularis parvicellularis [23]. The present study supported this observation in specific regard to the lingual preganglionic neurons. These formed a short, diffuse, but uninterrupted column in which the chorda tympani and glossopharyngeal units were overlapped except at the rostral and caudal extremes. While the present study indicated that the lingual ganglia contain parasympathetic neurons which receive preganglionic projections from the nucleus reticularis parvicellularis via the chorda tympani and glossopharyngeal nerves, the functions served by th~se autonomic neurons have not been resolved. As noted previously, the proximity of the caudal (glossopharyngeal) ganglia to the lingual glands has suggested a secretory motor function. The effectors se~wed by the ganglia in the non-glandular, rostral two.thirds of the tongue are a matter of greater speculation. In the absence of glands, vessels appear as the predominant, autonomically innervated effectors. Prior studies have called attention to the abundance of richly innervated arteriovenous anastomoses in the canine tongue and their possible role in thermoregulation [3,16]. It seems, therefore, reasonable to speculate that the arteriovenous anastomotic segments receive substantial postganglionic projections from the lingual ganglia. This speculation is in accord with Hellekant's [7] recent demonstration of lingual vasodilator fibers within the chorda tympani of the rat. ACKNOWLEDGEMENTS
This research was supported by NIH Grant 5-F34-FM06013-02. The authors thank Drs. A. de Lahunta, B.P. Halpern and J.M. Petras for their valuable directions during the experiment and for editing the manuscript. The authors are also grateful to Mr. W.P. Hamilton IV for the illustrations, to Dr. Isak Foss for his surgical assistance, to Mrs. Helen Lehman and Miss Barbara Loman for their technical help, and to Miss Alice CkLwan for secretarial support. REFERENCES
1 Barnard, J.W., The hypoglossalcomplex of v,~.rtebrates, J. comp. Neuro]., 72 (1940) 489--524,
129 2 Brodal, A., Neurological Anatomy, Oxford Universit) [:÷_-% London, 1969. 3 Brown, M.E., The occurrence of arterio-venous anastomoses in the tongue of the dog, Anat. Rec., 69 (1937) 287--292. 4 Chibuzo, G A., The Location of Cell Bodies of Primary Motor and Ser, sory Neurons that Innervate the Tongue of the Dog, Ph.D. Thesis, Cornell University, Ithaca, 1979. 5 Crosby, E.C., Humphrey, T. and Lauer, E.W., Correlative Anatomy o the Nervous System, MacMillan, New York, 1962. 6 Gomez, H., The innervation of lingual salivary glands, Anat. Rec., 139 (I 961 ) 69--76. 7 Hellekant, G., Vasodilator fibers to the tongue in the chorda tympani proper nerve, Acta physiol, scand., 99 (1977) 292--299. 8 Hiura, T., Sa]ivatory neurons innervate the submandibular and sublingual glands in the rat: horseradish peroxidase study, Brain Res., 137 (1977) 145--149. 9 Kohrustamm, O., Der Nucleus salivatorius Chorda tympani (Nervi intermedii), Anat. Anz., 21 (1902) 362--363. I0 Kohnstamm, O., Der Nucleus salivatorius und das cranioviscerale System (mit Demonstration), Arch. Psychiat. Nervenkr., 37 (1903) 660 G62. 11 K o h n s t ~ . O., Der Nucleus salivatorius inferior und das cranio-viscerale System, Neurol. Zbl., ~2 (1903) 699. 12 Kristensson, K and Olsson, Y., Uptake and retrograde axona] transport of protein tracers in hypo.,flossal neurons: fate of the tracer and reaction of the nerve cell bodies, Acta neuropatl~. (Berl.), 23 (1973) 43--47. 13 Kristensson, K., Olsson, Y. and Sj~strand, J., Axonal uptake and retrograde transport of exogenous proteins in the hypoglossal nerve, Brain Res., 32 (1971) 399--406. 14 Kristensson, K. and Sj6strand, J., Retrograde transport of protein tracer in the rabbit hypoglo~al nerve during regenermtir,,, Brain Res., 45 (197~.) 175--181. 15 Nauta, H.J.W., Pritz, M.B. and Llsek, R.J., .~tterents to the rat caudo-putamen studied with horseradish peroxidas..~. An evaluation of a retrograde neuroanatomical research method, Brain Res., 67 (1974) 219--238. 16 Pritchard, M.M.L. and Daniel. P.M., Arterio-venous anastomoses in the tongue of the dog, J. Anat. (Lond.), 87 (1953) 66--74. 17 Rakhawy, M.T., Phosphatases in the nervous tissue (the nature of the ganglionic nerve cells in the tongue), Acta Anat. (Basel), 83 (1972) 356--366. ~ Sarbach, K., De la r~partition de~ ganglions nervcux dans la langue de rongeurs et de carnivores, Schwelz. Mschr. Zahnheilk., 65 (1955) 1--17. !9 Satomi, H., Takahashi, K., Ise, H. and Yamamoto, T., Identification of the superior salivatory nucleus in the cat as studied by the HRP method, Neurosci. Lett. 14 (1979) 135--139. 20 Shute, C.C.D. and Lewis, P.R., The salivatory center in the rat, J. Anat. (Lond.), 94 (1960) 59--81. 21 Solomowicz, J., Vom Centrum der Submaxillardr/ise, Neurol. Zbl., 27 (1908) 724--727. 22 Stewart, F.W., The development of the cranial sympathetic ganglia in rat, J. comp. Neurol., 31 (1920) 163--218. 23 Torvik, A., Die Lokalisation des Speichelzertrum bei der Katze, Z. mikr.-anat. Forsch., 63 (1957) 317--326. 24 Wstkins, D.W, Hypoglossal nucleus: localizatio,~ of motoneuronal pools, Anat. Rec., 190 (1978) 5"6---577. 25 Ya~i,~, K., W.-~itere Untersuchungen fiber das So..~ichelzentrum, Anat. Ariz., 35 (1909) 70--'!5. 26 Ysbi'-a, K. ard ,ttayama, S., Uber des Speichelsekretioncentrum, Neurol. Zbl., 28 (1909) 738--~ 53.