THE JOURNAL OF COMPARATIVE NEUROLOGY 324:81-93 (1992)

Musculotopic Organization of the Hypoglossal Nucleus in the Cynomolgus Monkey, Macaca fascicularis ALAN J. SOKOLOFF AND TERRENCE W. DEACON Biological Anthropology, Harvard University, Cambridge, Massachusetts 02138

ABSTRACT The movements of the tongue in feeding and vocalization are enabled by a complex system of interdigitated muscle fibers in the tongue body. Because of this complexity, the detailed anatomical connections between individual intrinsic tongue muscles and corresponding motoneurons in the hypoglossal nucleus have not been described for any mammal. In this study we describe the distribution of retrogradely labeled neurons in the hypoglossal nucleus, following injections of wheat-germ agglutinin-horseradish peroxidase into different regions of the tongue of 21 cynomolgus monkeys. These experiments demonstrate a spatial organization of hypoglossal motoneurons that reflects the anatomical and functional organization of tongue body muscles: motoneurons innervating the transversus and verticalis muscles are located in medial hypoglossal nucleus regions, motoneurons innervating the genioglossus are located in intermediate hypoglossal nucleus regions, motoneurons innervating the hyoglossus and inferior longitudinalis are located in ventrolateral hypoglossal nucleus regions, and motoneuPons innervating the styloglossus and superior longitudinalis are located in dorsolateral hypoglossal nucleus regions. Motoneurons innervating the suprahyoid muscle, the geniohyoid, are situated in a cell column separated ventrally from the main body of the hypoglossal nucleus. Motoneurons innervating the palatoglossus are located in the nucleus ambiguus and, possibly, in dorsolateral hypoglossal nucleus regions. Motoneurons of the medial divisions of the hypoglossal nucleus innervate tongue muscles that are oriented in planes transverse to the long axis of the tongue whereas motoneurons of the lateral divisions innervate tongue muscles that are oriented parallel to this axis. These results suggest that the segregation of motoneurons corresponds to t h e functional distinction between tongue protrusion and retrusion. o 1992 Wiley-Liss, Inc. Key words: tongue, brainstem, WGA-HRP, motoneuron, somatotopy

The tongue of mammals is critical for normal mastication, deglutition, respiration, and vocalization. In humans the tongue takes on an additional role in speech production. The importance of the tongue in these behaviors is reflected in human clinical disorders which involve dysfunction of the neural system controlling tongue movement. Tongue thrusting (Lowe, '81) and sudden infant death syndrome (Laitman and Crelin, '80) have both been linked to developmental anomalies of the nervous system controlling tongue movement. Despite the functional importance of the tongue in oral motor behaviors, little is known about the organization and innervation of tongue body musculature. Studies of swallowing and chewing indicate that different tongue body regions may behave differently in a given tongue movement (Hiiemae and Crompton, '85); however, the anatomical and neural organization allowing regional tongue body control has not been elucidated. Detailed descriptions of tongue

o 1992 WILEY-LISS, INC.

anatomy have been reported for only a few mammals (for review, see Lowe, '81; Sokoloff, '89), principally the human and cat (Abd-El Malek, '38, '39). These studies generally identify eight tongue muscles. Four extrinsic muscles originate outside of the tongue body: the hyoglossus, styloglossus, palatoglossus, and genioglossus. Four intrinsic muscles originate and insert within the body of the tongue: the transversus, verticalis, inferior longitudinalis, and superior longitudinalis. The divergent terminations of extrinsic muscles and the local arrangement of intrinsic muscles provide an anatomical substrate for the observed regional specificity of tongue body movements. Because of the complex interdigitation of the muscles of the tongue body, investigations of the organization of the Accepted June 12,1992 Address correspondence to Alan J. Sokoloff, Division of Biology and Medicine, Brown University, Providence RI 02912.

AJ. SOKOLOFF AND T.W. DEACON

82

SUPERK LONGlTUDll MEDIAN SEPTUM

\

\A

Fig. 1. Schematic diagram of muscles of t h e tongue body of the cynomolgus monkey. Arrows refer to the extrinsic attachments of extrinsic tongue muscles. T h e geniohyoid, a suprahyoid muscle located ventral t o t h e genioglossus, courses from t h e mandible to t h e hyoid bone, and does not enter t h e tongue body.

hypoglossal nucleus (HGN) motoneurons that innervate tongue muscles have focused exclusively on the portions of extrinsic muscles that reside outside of the tongue body. These studies demonstrate a musculotopic innervation of hyoglossus, styloglossus, and genioglossus muscles in the rat, cat, dog, and macaque (Barnard, '40; Lowe, '81; Uemura-Sumi et al., '81). In these animals, motoneurons innervating tongue retrusor muscles (hyoglossus and styloglossus) are located in dorsal or dorsolateral HGN regions, whereas motoneurons innervating the main tongue protrusor muscle (genioglossus) are located in ventral, ventromedial, or intermediate regions of the nucleus (Krammer et al., '79; Uemura et al., '79; Uemura-Sumi et al., '81; Lowe, '81). The organization of the motoneurons that innervate intrinsic tongue muscles and the fibers of extrinsic muscles within the tongue body has not been investigated. In the present study we describe the location of HGN motoneurons that innervate individual muscles of the tongue body of the cynomolgus monkey (Macaca fascicuAbbreviations GG

GH HG HGN IL IV L Ld Lv M Mm Mi

PG SG SL T Ve V

vc

genioglossus muscle geniohyoid muscle hyoglossus muscle hypoglossal nucleus inferior longitudinalis muscle fourth ventricle lateral nucleus lateral nucleus pars doralis lateral nucleus pars ventralis medial nucleus medial nucleus pars medialis medial nucleus pars intermedialis palatoglossus muscle styloglossus muscle superior longitudinalis muscle transversus muscle verticalis muscle ventral nucleus ventrocaudal nucleus

Zuris). Experimental results are discussed in the light of recent research on tongue body function and comparative studies of HGN organization.

MATERIALS AND METHODS Experimental animals A total of 27 adult cynomolgus monkeys (M. fusciculuris) were used in this study. Six of these were obtained as fixed specimens from other studies and used for purposes of tongue dissection and brainstem histology. Twenty-one were used in tracer studies.

Muscle architecture The tongues and adjacent infrahyoid and suprahyoid regions of six monkeys were dissected under a dissecting microscope to identify the origins and courses of tongue muscles (Fig. 1).Selected sections from the tongues of eight cynomolgus monkeys used in other parts of this study were stained by the Milligan trichrome method (Humason, '79). Four of these tongues were cut in serial, 75 pm sections on a freezing microtome in the transverse plane, three in the sagittal plane, and one in the horizontal plane. The tissue was analyzed and photographed in brightfield or polarized light with a Zeiss Axioplan microscope to determine the arrangement of individual muscle fibers. Reconstruction of tongue body musculature allowed the analysis of injection sites in the tracer experiments.

Organization of the hypoglossal nucleus Cytoarchitecture. The brainstems of six cynomolgus monkeys were cut in transverse 75 km sections on a freezing microtome and alternating sections were mounted directly onto gelatin-coated slides and air-dried. Sections were stained in neutral red or cresyl violet and analyzed under brightfield microscopy to determine HGN cytoarchitecture to assist, in analysis of tracer experiments. To

MONKEY HYPOGLOSSAL NUCLEUS

83

2

26

D

n

U

23

Fig. 2. Composite diagrams illustrating injection-site location of experimental cases. a, b, and c each illustrate four transverse sections of the tongue and d illustrates a sagittal section near the midline of the tongue. The extent of injected label is represented in black. Numbers correspond to cases listed in Tables 1 and 2. A, anterior; D, dorsal; P, posterior; V, ventral.

facilitate comparison of the rostrocaudal position of labeled neurons in different cases, the HGN was divided into five levels, each level representing one fifth of the rostrocaudal extent of the HGN. These levels are designated, from the caudal pole of the HGN to the rostral pole, level 1(caudal), level 2 (obex),level 3 (middle), level 4 (midrostral),and level 5 (rostral), as depicted in Figures 3 and 4. Bucer injections. Tongue muscles in 21 cynomolgus monkeys were injected with 0.3-3.0 p1 of a 5% solution of wheat-germ agglutinin-conjugated horseradish peroxidase (WGA-HRP, Sigma Chemical Co., St. Louis, MO) using a 1 pl, 5 pl, or 10 p1 Hamilton syringe. Muscle injections were done under either halothane or ketamine anaesthesia. Eighteen to seventy-two hours following WGA-HRP injection, the animals were reanaesthetized with an overdose of pentobarbital sodium and perfused through the left ventricle with 1-2 liters of 0.9% saline, followed first by 1-2 liters of a fixative containing 0.5% paraformaldehyde, 2.0% glutaraldehyde, and 2.0% sucrose in a 0.1 M phosphate buffer, and second by 0.5-1.0 liters of 10% sucrose in 0.1 M phosphate. The brainstem, upper cervical spinal cord, tongue, and adjacent muscles were removed and stored for

1 to 3 days in a 20% sucrose solution of 0.1 M phosphate at 4°C. Histologg. The brainstem and upper cervical spinal cord were cut serially in 75 pm transverse sections on a freezing microtome. Frozen 75-100 pm sections of the tongue body and adjacent extrinsic tongue and mouth floor muscles were also cut in the transverse, sagittal, and horizontal planes. Brainstem and tongue tissue sections were collected in a solution containing 30% sucrose and 30% ethylene glycol in 0.1 M phosphate buffer and stored for up to 1week at 0°C. Brainstem and muscle tissue from all experiments was processed with tetramethylbenzidine (TMB) following Mesulam ('78) and Deacon et al. ('83). Every second or fourth section of brainstem tissue and every fourth or sixth section of tongue body muscle was reacted with TMB for 2-4 hours, mounted on gelatin-coated slides, air dried, and counterstained in neutral red. Brainstem and tongue slides were examined in brightfield, darkfield, or polarized light microscopy to determine the extent of the injection site and resulting brainstem label. Brainstem and tongue sections with WGA-HRP label were photographed and traced.

A.J. SOKOLOFF AND T.W. DEACON

84 Injection sites. Tracer injections were placed into the tongue body in 17 animals, into the extra-lingual portion of the genioglossus muscle in two animals and in the palatoglossus and geniohyoid muscle in two others (see Figure 2). Although the palatoglossus and geniohyoid muscles are not commonly classified with tongue body muscles (Williams et al., '89),previous experimental evidence suggests that these muscles are innervated by hypoglossal motoneurons in some mammals (Holstege et al., '83; Kitamura et al., '83, '85, '86; Uemura et al., '79; Uemura-Sumi et al., '81). Data analysis. Because of the complex interdigitation of muscles within the tongue body it was difficult to confine injections to fibers of a single muscle. As a result it was often not feasible to determine the innervation of a single tongue muscle from examination of a single tracer experiment. In cases where the injection involved two or more muscles, the innervation of an individual tongue muscle was determined by comparing different pairs of cases that satisfied three criteria: 1) the injection in one case differs from that in the other in the involvement of a particular muscle, 2) all other muscles involved in the injection in one case are also involved in the other, and 3) the injections in both cases are located in a similar region of the tongue body. When the patterns of HGN label for two such cases are compared, HGN label present in one case but not the other can be said to be due to the unique inclusion of the particular muscle in the injection site of the one case. Most WGA-HRP injections involved one side of the tongue body. In all such cases label was confined to the ipsilateral HGN, confirming previous reports of an exclusively ipsilateral tongue innervation (Uemura, '79; UemuraSumi et al., '81).This fact was used to help further analyze tongue injection sites. In a few cases, spread of WGA-HRP to both sides of the tongue resulted in a bilateral injection with an ipsilateral predominance. In these cases motoneuron label in the ipsilateral HGN was attributed to muscles involved in the ipsilateral injection site whereas motoneuron label in the contralateral HGN was attributed to muscles involved in the contralateral injection site. Comparison of both sides in a case with asymmetric tracer uptake site could provide information on subtle topographic differences in motor innervation. A list of all injections and the muscles involved in each case is included in Table 1.

RESULTS Cytoarchitecture Cells in the HGN are organized into two main nuclei, a medial nucleus (extending from level 1 to 5) and a lateral nucleus (extending from level 2 to 51, which are separated by a band of white matter (Figs. 3, 4). These nuclei are further divisible on cytoarchitectonic criteria into separate subnuclei, the medial nucleus into a pars medialis (Mm, level 2 to 4) and a pars intermedialis (Mi, level 2 to 41, and the lateral nucleus into apars ventralis (Lv, level 2 to 4)and a pars dorsalis (Ld, level 2 to 4). Together the medial and lateral nuclei form the major portion or "main body" of the HGN. A compact cell group, the ventral nucleus, lies ventral to the main body of the HGN (Figs. 3, 4). In addition, a column of neurons located ventrolateral to the caudal HGN was found to provide motor innervation to the tongue body. This cell column is called the ventrocaudal nucleus (Fig. 4).

TABLE 1. Case by Case Involvement of Tongue Muscles' T

Ve

1

T

2

T T

Ve Ve Ve Ve

3 4 7 8 10

T

11

T

12 13

T

19 20 22 26 22 23 24 Z5

SL SL SL SL

Ve

SL

T

Ve

SL

'7

Ve Ve Ve Ve

IL

GG

HG

IL IL IL

GG GG GG GO GG GG GG GG GG GG

HG

IL 1L

HG HG

SG

PG

SG SG

PG PG

GH

GH

HG

PG

T T T

IL

GG

HC

IL IL

GG GG

IL IL

GG GG

HG

SL T T

HG HG

SG

PG

-

'Tongue body injections listed in numerical order. GG, genioglossus; GH, geniohyoid, HG, hyoglossus; IL, inferior longitudinalis; PG, palatoglossus. SG, styloglossus; SL, superior longitudinalis; T, transversus; Ve. verticalis Symbols in italic indicate bilateral involvement.

Motorpool organization Genioglossus. Sixteen tracer injections involve fibers of the genioglossus. Following exclusive injection of the genioglossus (cases 8 and 10; contralateral spread case 25; Fig. 2, Table l ) ,labeled neurons are located in Mi (level 2 to 4) and in the rostral portion (level 5 ) of the medial nucleus (Fig. 5a; Table 2). Most tongue body injections involved genioglossus in tandem with transversus and verticalis fibers (Table 1). In these cases, labeled neurons are consistently present in Mi; in some of these cases labeled neurons are also present in level 5 of the medial nucleus (Table 2). Injections that do not involve genioglossus fibers (cases 20, 14, 22, and 23) lack labeled neurons in either Mi or level 5 of the medial nucleus. Transversus and uwticalis. Fourteen tracer injections involve fibers of the transversus and verticalis muscles. Most cases involving transversus and verticalis also involved fibers of several other tongue muscles (Figs. 2, 5b); however, in a few cases uptake was limited to transversus, verticalis, and genioglossus (Fig. 2; Table 1).In these cases (ipsilateral uptake in case 22 and contralateral uptake in cases 1, 3, and 19) HGN label is confined to the medial nucleus (Fig. 5c). In the one case (case 20) involving transversus and verticalis fibers but not genioglossus fibers, label in the medial nucleus is confined to level 2 of Mm. In concert with data from genioglossus injections, these results suggest that motoneurons innervating transversus and verticalis are located in level 1through level 4 of the medial nucleus confined to Mm. Injections involving transversus but not verticalis fibers (ipsilateral injection in cases Z4,Z5, and 12; contralateral in cases 1and 19) show

Fig. 3. Composite photomicrograph of Nissl-stained brainstem sections, identifying the rostrocaudal extent of cytoarchitectonically defined subdivisions. Representative sections from the caudal (level 1). obex (level 2), middle (level 3), midrostral (level 4), and rostral (level 5) levels of the HGN are shown. Arrows identify the separation of the medial nucleus from the lateral nucleus. IV, fourth ventricle; L, lateral nucleus; Ld, lateral nucleus, pars dorsalis; Lv, lateral nucleus, pars ventralis; M, medial nucleus; Mi, medial nucleus, pars intermedialis; Mm, medial nucleus, pars medialis; VC, ventrocaudal nucleus; V, ventral nucleus. Calibration bar, 100 pm.

MONKEY HYPOGLOSSAL NUCLEUS

85

‘\J \

I

/

/

3

I

2

‘I ‘I ‘ I

1 Figure 3

AJ. SOKOLOFF AND T.W. DEACON

86

level 5

- - - (rostral) --

level 4 (midrostral)- -

level 3 (middle)

level 2 (obex)

' -

- --

level I (caudal)

--

-,

-----

_ _ - -- -.

Fig, 4. A three-dimensional reconstruction of the HGN is illustrated on the left, and representative cross-sections are illustrated on the right. The 3-D reconstruction is divided to correspond to each of the five rostrocaudal levels distinguished in the text. The HGN is composed of two major divisions, the medial nucleus (MI and the lateral nucleus (L). These are further subdivided into a medial nucleus pars medialis

(Mm)and medial nucleus pars intermedialis (Mi),and a lateral nucleus pars ventralis (Lv) and lateral nucleus pars dorsalis (Ld). Two separate cell groups, the ventral nucleus (V), and the ventrocaudal nucleus (VC) are separated from the main portion of the HGN. Dotted lines indicate correspondences between the reconstruction and each of the five sections depicted.

the same pattern of labeling in Mm as injections that involved both transversus and verticalis fibers. This suggests overlap in the locations of neurons innervating these two muscles. Labeled neurons in Mm are comparatively small, and some exhibit dendrites that extend across the midline into the contralateral Mm. In several cases involving deep midline tongue injection (cases 2 , 3 , and 111, labeled neurons are observed ventrolatera1 and caudal to the well-defined limits of the HGN (Fig. 6a), identified above as the ventrocaudal nucleus. Although many muscles were injected in these cases, the deep central location of injections that labeled the ventrocaudal nucleus suggests that it may innervate transversus fibers. Hyoglossus and inferiorlongitudinalis. Thirteen tracer injections involve fibers of the hyoglossus and/or inferior longitudinalis (see Table 1).Owing to the complex interdigitation of different tongue muscles along the inferior border of the tongue body, all cases which involve hyoglossus and inferior longitudinalis also involve injection of several other tongue muscles (Fig. 2). To aid in the analysis of tracer

injections involving hyoglossus and inferior longitudinalis, label located in the medial nucleus is attributed to the injection of transversus, verticalis, or genioglossus (see above). Injection results indicate that neurons which innervate hyoglossus and inferior longitudinalis are located in Lv and its most rostral extent in L. In every case in which either hyoglossus or inferior longitudinalis fibers are injected, retrogradely labeled cells are present in Lv and level 5 L. In most cases which lack injection of either hyoglossus or inferior longitudinalis fibers, labeled cells are not observed in Lv or level 5 L. (Two exceptions involve label in Lv following injections into superior longitudinalis; for discussion see below.) Comparisons of cases which differ in the injection of hyoglossus or inferior longitudinalis fibers (e.g., for hyoglossus cases 24 and 25; for inferior lonstudinalis cases 2 and 13,25 and 22) verify a Lv and rostral L location of motoneurons which innervate these muscles (Fig. 7a,b). Figure 7a shows a case (24) in which hyoglossus injection labels cells in the most rostral division of L. Figure 7b

MONKEY HYPOGLOSSAL NUCLEUS

87

TABLE 2. Rostrocaudal Location of Neurons Labeled in Individual Cases Level 1 Number

M

Mm

1 2

M M M M

Mm Mm Mm Mm V

3 4 7 8 10

11 12 13

Level 2 Mi Lv MI MI MI Mi

Lv Lv Lv Lv

Ld

Mm Mm

Ld

Mm

Mi

Mm

MI

V

18 19

20 22 26 22 23 24

25

M

Mm NA Mm Mm Mm Mm

Mm Lv

Lv

Mm Mm

Mi

Mi

Mm

Mi

Mi

Mm Mm

Mi Mi

Mi

Ld Ld Ld

Mm

MI

Mm

Mm Mm

MI MI

MI

Lv

Ld

Lv

Ld Ld Ld Ld

Lv Lv

Mi

Mi

Ld

Mm

Ld Ld

Mm

Mi Mi MI

Level 5 M L L

M

L L

M M M

Mi

Lv

Mi

Mi

MI

Mi Mi

Ld Mm Mm

Mm

Mi

14

16

Lv Lv Lv LV

Level 4 Ld

MI

Mm Mm

MI MI

Mm

Mi M M

Level 3 Mi Lv

Lv Lv

Ld M

L

Ld Lv

Mm

L

Lv L

Lv Lv Lv

Ld Ld

Lv

L L L

Case numbers are in the far left column. Cases listed in boldface include labeled cells also within the ventrocaudal nucleus. L, lateral nucleus; Ld, lateral nucleus, pars dorsalis; Lv, lateral nucleus, pars ventralis; M, medial nucleus; Mi, medial nucleus, pars intermedialis; Mm, medial nucleus, pars medialis; NA, nucleus ambiguus; V, ventral nucleus. Symbols in italic indicate bilateral label.

shows a case in which hyoglossus is injected along with styloglossus fibers, and labeled cells are observed in Lv as well as Ld (see Styloglossus, below). Neurons in Lv labeled following hyoglossus injection are situated in the lateral border of Lv and send dendrites laterally or ventrolaterally into the adjacent reticular formation. The axons of these neurons course medially in the HGN before curving ventrally to exit through the hilus of the hypoglossal nerve. Styloglossus and superior longitudinalis. Eight tracer injections involve styloglossus and/or superior longitudinalis (see Table l). In every case in which styloglossus or superior longitudinalis are injected, there are retrogradely labeled cells in Ld. Only one case (case 1)demonstrated Ld label without apparent injection of either styloglossus or superior longitudinalis. Comparison of label in case 23 (involving styloglossus and hyoglossus) with case 12 (involving hyoglossus, inferior longitudinalis, and genioglossus, but not styloglossus) indicates that styloglossus involvement is associated with Ld label. Labeled neurons in case 2 3 are situated along the lateral edge of Ld. Most injections that involve styloglossus label Lv as well as Ld (Fig. 7c; the only exception is case 13). Lv label occurs without label in Ld when injections involve the hyoglossus and inferior longitudinalis, but not the styloglossus or superior longitudinalis. This suggests that Lv projects to the muscles running along the dorsal and ventral surfaces of the tongue, whereas Ld projections are confined to muscles running along the dorsal surface only. Two cases (22 and 14) involved the superior longitudinalis only (Fig. 2). In these cases neurons are most prominent in Ld and only a few labeled neurons are present in one level (3) of Lv. In these cases the lateral edge of Ld is predominantly label-free, and labeled neurons are located in the center and medial regions of Ld, medial to the location of Ld neurons which innervate styloglossus. Comparison of adjacent injections that differ in the involvement of superior longitudinalis fibers also suggests that Ld is the major source of superior longitudinalis innervation (e.g., cases 2 and 26, cases 13 and 16). The isolation of Ld label in these comparisons, the presence of contralateral Ld label in case

11, and the presence of Ld label in all cases with superior longitudinalis involvement (cases 2, 13,14,11,4,22,and 3) identifies Ld as the primary site of neurons innervating this muscle. Palatoglossus. Injection of WGA-HRP into the extralingual portion of palatoglossus was made in one case (case 18).The injection was placed adjacent to the tongue body in the region where the palatoglossus muscle leaves the cover of the palatoglossal arch to enter the tongue body. Following palatoglossus injection labeled neurons are present in nucleus ambiguus in a column extending from level 4 to levels just rostral to the rostral HGN pole (Fig. 8). Labeled neurons are not present in the HGN. Following injections into the posterior tongue body adjacent to the entry of the palatoglossus (cases 2 3 and 31, labeling is not observed in nucleus ambiguus but is observed in Ld. Owing to the involvement of styloglossus in these injections, however, Ld label cannot be definitively ascribed to palatoglossus uptake. No nucleus ambiguus labeling was observed following injections in any part of the tongue body, including regions adjacent to the entry of the palatoglossus. Geniohyoid. In one experiment (case 7; Fig. 2) an injection was centered in the geniohyoid muscle, but spread superiorly to involve a few of the immediately adjacent genioglossus fibers. Following this injection labeled neurons are located in the ventral nucleus at all ventral nucleus levels (level 2 and 31, in the most rostral regions of Mi (level 4) and in level 5 of the medial nucleus (Fig. 7b). The independent identification of Mi and level 5 of the medial nucleus as the site of genioglossus motoneurons suggests that only the ventral nucleus label is due to geniohyoid involvement. In the present study the ventral nucleus was only labeled in this single case of geniohyoid injection. A ventral location of motoneurons innervating the geniohyoid muscle has been reported in a number of mammals, including the macaque (Kitamura et al., '83, '85; Uemura et al., '79; Uemura-Sumi et al., '81). Neurons labeled in the ventral nucleus following geniohyoid injection are among the largest in the HGN and are clustered in a dense cell group which is separated from the

A.J. SOKOLOFF AND T.W. DEACON

88

Figure 5

MONKEY HYPOGLOSSAL NUCLEUS

89

case 3 1

a n

case 7 1

b

"u

Fig. 5 (opposite). The distribution of WGA-HRP label and representative photomicrographs of labeled neurons, following injections into different regions of the tongue body to demonstrate patterns of labeling in the medial nucleus: a: label in the medial nucleus pars intermedialis following injection into the genioglossus muscle, case 10; b: label throughout the main portion of the HGN following a large tongue body injection, case 3; c: label in the medial nucleus pars medialis and pars intermedialis following injection of transversus and genioglossus muscles, contralateral involvement in case 19. Arrows signify approximate rostrocaudal location of photomicrographs. calibration bars, 200 pm.

Fig. 6 (above). The distribution of WGA-HRP label and representative photomicrographs of labeled neurons in the ventral and ventrocaudal nuclei: a: following a large tongue body injection involving deep tongue muscles, case 3, that labels cells in the ventrocaudal nucleus; and b: following injection involving primarily geniohyoid fibers, case 7, that labels cells in the ventral nucleus. Arrows signify approximate rostrocaudal location of photomicrographs. Calibration bars, 200 pm. (Squiggly black lines in b are blood vessel artifacts.)

main body of the HGN by the exiting axons of the hypoglossal nerve.

medial nucleus; motoneurons innervating genioglossus are located in Mi extending to level 5 of the medial nucleus, motoneurons innervating styloglossus and superior longitudinalis are located primarily in Ld and to a lesser extent Lv; and motoneurons innervating the hyoglossus and inferior longitudinalis are located in Lv and level 5 of the lateral nucleus (Fig. 9). These data identify a primary mediallateral division of the HGN that corresponds to the structural-functional organization of tongue musculature. From a functional perspective, motoneurons innervating protru-

DISCUSSION Our results demonstrate a musculotopic organization of HGN motoneurons in the cynomolgus monkey. This organization reflects topological and functional features of tongue body musculature. Motoneurons innervating transversus and verticalis are located in Mm extending to level 4 of the

A.J. SOKOLOFF AND T.W. DEACC

90 bt,.

a

lateral

"0 1

case 23

b

"0 1

case 4

c Fimirp 7

MONKEY HYPOGLOSSAL NUCLEUS

I1 .I

I'

91

'

\

I' II

Fig. 8. Neurons labeled in nucleus ambiguus following extra-lingual injection of palatoglossus demonstrating extra-hypoglossal labeling. No labeled cells are found in the HGN in this case. Calibration bar, 100 pm.

sor muscles are located in the medial nucleus and motoneurons innervating retrusor muscles are located in the lateral nucleus. These data also locate motoneurons innervating two extra-lingual muscles involved in the control of tongue position and movement. Motoneurons innervating geniohyoid are located in the ventral nucleus, separated ventrally from the main body of the HGN. Motoneurons innervating palatoglossus are located in nucleus ambiguus, and possibly in Ld (see below). In the single published study of the musculotopic organization of the macaque HGN, Uemura-Sumi et al. ('81) injected the extra-lingual portion of extrinsic tongue muscles and identified styloglossus and hyoglossus motoneurons in the lateral HGN, genioglossus motoneurons in the medial HGN, and geniohyoid motoneurons in a separate ventral cell group. Most HGN motoneurons, however, were not labeled following extra-lingual muscle injections (Uemura-Sumi et al., '81; their Fig. 1).Our findings concerning the location of motoneurons innervating the extrinsic muscles are in agreement with those of Uemura-Sumi et al. (%I), although we find that motoneurons innervating particular tongue muscles extend both more rostrally and more caudally than they report. For example, while UemuraSumi et al. ('81) localized motoneurons innervating genioglossus, styloglossus, and hyoglossus to middle HGN levels,

neurons labeled following injection of these muscles in the present study were found to extend from obex (level 2) to the rostra1 tip of the HGN (level 5 ) . In both studies ipsilateral innervation was observed for all tongue muscles. Studies in the cat (Uemura et al., '79; Yoshida et al., '811, dog (Kosaka and Yagita, '03; Barnard, '40; but see Chibuzo and Cummings, '821, and rat (Krammer et al., '79; Kitamura et al., '83, '85, '86) have demonstrated a musculotopic HGN organization similar to that reported for the macaque (Uemura-Sumi et al., '81; present study). The HGN in these animals is organized in two major nuclei, either ventral and dorsal nuclei, or ventromedial and dorsolateral nuclei. Retrograde tracer studies demonstrate that motoneurons innervating the protrusor muscle genioglossus are located in the ventral (rat) or the ventromedial (cat, dog) nucleus, while motoneurons innervating the retrusor muscles styloglossus and hyoglossus are located in the dorsal (rat)or dorsolateral (cat, dog) nucleus. Motoneurons innervating the geniohyoid muscle are located in a cell group separated ventrally from the main body of the HGN (Uemura et al., '79; Yoshida et al., '81; Miyazaki et al., '83; Kitamura et al., '83, '85, '86). This basic protrusor versus retrusor HGN organization and separate geniohyoid innervation therefore has been found in all mammals studied. The results of the present investigation in the cynomolgus monkey extend this basic protrusor versus retrusor HGN organization to include the motoneurons that innervate intrinsic, as well as extrinsic, tongue muscles. Fig. 7. The distribution of WGA-HRP label and representative photomicrographs of labeled neurons, following injections into ventral In an investigation of the innervation of the musculature and lateral regions of the tongue body: a: label in case 24 following a involved in swallowing in the cat, Holstege et al. ('83) tongue injection that involves transversus and genioglossus, as well as hyoglossus and inferior longitudinalis; b: label in case 23 following a observed labeled neurons in nucleus ambiguus and, in tiny injection that does not include any of the intrinsic muscles but does addition, observed a few labeled neurons in the lateral HGN following injection of 60 +1 of HRP into the soft palate. In include hyoglossus, styloglossus, and intra-lingual fibers of palatoglossus; c: label in case 4 following a n injection that includes the muscles the present study a localized 1 +1 injection of WGA-HRP injected in case 2 4 but additionallyinvolves the superior longitudinalis. into extra-lingual palatoglossus fibers labeled neurons in These cases show differences in distribution of labeled cells in the ventral (Lv) and dorsal (Ld) divisions of the lateral nucleus. Arrows nucleus ambiguus but not in the HGN. The position of signify approximate rostrocaudal location of photomicrographs.Calibra- labeled HGN neurons reported by Holstege et al. ('83)is similar to the position of neurons labeled following lateral tion bars, 200 pm.

A.J. SOKOLOFF AND T.W. DEACON

92

medial

lateral

Fig. 9. A schematic diagram illustrating the distinct groups of tongue body muscles that are innervated by each of the subnuclei of the HGN. Motoneurons from medial nuclear divisions (left) innervate muscles predominantly running transverse to the orientation of the tongue body (transversus, verticalis, and genioglossus), whereas motoneurons from lateral nuclear divisions (right) innervate muscles

predominantly running parallel to the long axis of the tongue (styloglossus, hyoglossus, superior longitudinalis, and inferior longitudinalis). Motoneurons in the ventral nucleus innervate the geniohyoid muscle. The three major subnuclear divisions of the HGN are shown diagrammatically separated to emphasize these structural-functional differences.

posterior tongue body injections which involved palatoglossus and styloglossus fibers in the present study (cases 23 and 3), suggesting the possibility that lateral HGN label in the experiments of Holstege et al. ('83) is due to tracer spread to other muscles, and perhaps to the intra-lingual portion of the palatoglossus or styloglossus. Alternately, the presence of lateral HGN label following extra-lingual palatoglossus injection in the study of Holstege et al. ('83) may reflect species differences in the sources of extralingual palatoglossus innervation. In the present study, posterior tongue body injections which involved the intra-lingual portion of palatoglossus fibers did not label motoneurons in nucleus ambiguus, nor did extra-lingual palatoglossus injections label hypoglossal motoneurons. This suggests either that axons that project from the nucleus ambiguus to the palatoglossus do not extend to contact the intra-lingual portion of the palatoglossus in the macaque, or that palatoglossus fibers receive innervation from two sources: intra-lingual fibers from the lateral HGN and extra-lingual fibers from nucleus ambiguus. It is interesting to note that Domenech-Ratto ('77) has proposed both a palatal and lingual developmental origin of the palatoglossus in the rat.

Cineradiographic studies of macaque feeding have demonstrated that different regions of the tongue body may move in different directions during feeding behaviors. During movement of food from the oral cavity to the oropharynx in swallowing, for example, the anterior and middle regions of the tongue dorsum are raised while the posterior region of the tongue dorsum is lowered (Franks et al., '84).How are differential movements of anterior versus posterior t,ongue regions produced? Are anterior versus posterior regional differences in tongue muscle activity reflected in a corresponding topographic segregation of hypoglossal motoneurons? The results of the present study do not reveal a discrete HGN innervation of anterior versus posterior fibers of any muscle. A subtle pattern of topographic innervation, however, is observed for genioglossus, transversus, and verticalis. Injections into anterior genioglossus fibers label motoneurons in the caudal HGN levels (levels 2 and 3 of Mi), whereas injections into deep and posterior genioglossus fibers additionally label motoneurons in more rostra1 HGN levels (levels 2 to 4 of Mi and level 5 of the medial nucleus; compare pattern of label in Fig. 5a,c; see also Table 2 ) . Similarly, injections into anterior transversus and vertica-

MONKEY HYPOGLOSSAL NUCLEUS lis fibers label motoneurons in levels 1to 3 of Mm, whereas injections into posterior transversus and verticalis fibers additionally label motoneurons in level 4 of Mm (compare Fig. 5b to 5c and 7a to 7c; Table 2). This topography may reflect a functional pattern of motor innervation which stresses a relatively discrete, independent control of the anterior tongue body. Double label and glycogen depletion studies are needed to discern if these topographic differences are associated with regional differences in the domains of individual motor units within the tongue body. In summary, the present data indicate that the major subdivisions of the hypoglossal nucleus correspond to structural and functional divisions of tongue musculature. The medial divisions, Mm and Mi, innervate tongue muscles that are oriented in planes transverse to the long axis of the tongue whereas the lateral divisions, Ld and Lv, innervate tongue muscles that are oriented in parallel with this axis. The overlap of motoneurons that innervate muscles with similar anatomical orientation may facilitate the simultaneous recruitment of these muscles in different tongue movements, for example, the hyoglossus and the inferior longitudinalis in tongue retrusion and the transversus and verticalis in tongue protrusion. In addition, overlap of the motoneuron pools that innervate similarly oriented intrinsic and extrinsic muscles may reflect shared developmental and evolutionary histories. Studies in non-mammal vertebrates that demonstrate a similar musculotopic organization of HGN motoneurons despite functional differences in tongue movement (e.g., in frogs: Sokoloff, '89; Sokoloff, '91) suggest that this musculotopic pattern of projections is remarkably conservative.

ACKNOWLEDGMENTS We thank A.W. Crompton for providing experimental facilities and support, M.D. Kramer for assistance in all stages of this research, and M.R. Szpir for comments on the manuscript. Financial support was provided in part by an NSF graduate fellowship (A.J.S.) and by the Milton Fund of Harvard University .

LITERATURE CITED Abd-El-Malek,S. (1938) A contribution to the study of the movements of the tongue in animals with special reference to the cat. J. Anat. 73:15-33. Abd-El-Malek, S. (1939) Observations on the morphology of the human tongue. J. Anat. 73:201-210. Barnard, J.W. (1940) The hypoglossal complex of vertebrates. J. Comp. Neurol. 72489-524. Chibuzo, G.A., and J.F. Cummings (1982) An enzyme tracer study of the organization of the somatic motor center for the innervation of different muscles of the tongue: Evidence for two sources. J. Comp. Neurol. 205:273-28 1.

93 Deacon, T.W., H. Eichenbaum, P. Rosenberg, and K. Eckmann (1983) Afferent connections of the perirhinal cortex in the rat. J. Comp. Neurol. 220:168-190. Domenech-Ratto, G. (1977) Development and peripheral innervation of the palatal muscles. Acta. Anat. 97t4-14. Franks, H.A., A.W. Crompton, and R.Z. German (1984) Mechanism of intraoral transport in macaques. Am. J. Phys. Anthropol. 65:275-282. Hiiemae, K.M., and A.W. Crompton (1985) Mastication, food transport and swallowing. I n M. Hildebrand, D.M. Bramble, K.F. Liem, and D.B. Wake (eds): Functional Vertebrate Morphology. Cambridge: Harvard University Press. Holstege, G., G. Graveland, C. Bijker-Bienond, and I. Schuddeboom (1983) Location of motoneurons innervating soft palate, pharynx, and upper esophagus: Anatomical evidence for a possible swallowing center in the pontine reticular formation. Brain Behav. Evol. 23:47-62. Humason, G. (1979) Animal Tissue Techniques (4th ed). San Francisco: Freeman and Lo. Kitamura, S., T. Nishiguchi, and A. Sakai (1983) Location of cell somata and the peripheral course of axom of the geniohyoid and thyrohyoid motoneurons: A horseradish peroxidase study in the rat. Exp. Neurol. 79:87-96. Kitamura, S., T. Nishiguchi, and A. Sakai (1985) A horseradish peroxidase study of rat lingual motoneurons with axons passing through the cervical nerve. Exp. Neurol. 87r20-34. Kitamura, S., T. Nishiguchi, J. Okubo, K. Chen, and A. Sakai (1986) An HRP study of the motoneurons supplying the rat hypobranchial muscles: Central localization, peripheral axon course and soma size. Anat. Rec. 21 6:73-81. Kosaka, K., and K. Yagita (1903) Experimentelle untersuchungen uber die ursprunge des nervus hypoglossus und seines absteigenden astes. Jahrb. F. Psychiat. U. Neur. 24:150-189. Krammer, E.B., T. Rath, and M.F. Lischka (1979) Somatotopic organization of the hypoglossal nucleus: A HRP study in the rat. Brain Res. 170:533-537. Laitman, J.T., and E.S. Crelin (1980) Developmental change in the upper respiratory system of human infants. Perinatol. Neonatol. 4:15-22. Lowe, A.A. (1981) The neural regulation of tongue movements. Prog. Neurobiol. 15r295-344. Mesulam, M-M. (1978) Tetramethyl benzidine for horseradish peroxidase neurohistochemistry: A noncarcinogenic blue reaction product with superior sensitivity for visualizing neural afferents and efferents. J. Histochem. Cytochem. 26: 106-1 17. Miyazaki, T., Y. Yoshida, M. Hirano, T. Shin, and T. Kanaseki (1983) Central location of the motoneurons supplying the thyrohyoid and the geniohyoid muscles as demonstrated by horseradish peroxidase method. Brain Res. 219:423427. Sokoloff, A.J. (1989) The Organization of the Hypoglossal Nucleus: An Experimental Neuroanatomical Investigation of Hypoglossal-lingual and Cortico-hypoglossal Projections in Mucaca fasciculuris and Other Vertebrate Species. Ph.D. thesis, Harvard University. Sokoloff,A.J. (1991) Musculotopic organization of the hypoglossal nucleus in the grass frog, Ranapipiens. 3. Comp. Neurol. 308:505-512. Uemura, M., K. Matsuda, M. Kume, Y. Takeuchi, R. Matsushima, and N. Mizuno (1979) Topographical arrangement of hypoglossal motoneurons: An HRP study in the cat. Neurosci. Lett. 13.99-104, Uemura-Sumi, M., N. Mizuno, S. Nomura, N. Iwahori, Y . Takeuchi, and R. Matsushima (1981) Topographical representation of the hypoglossal nerve branches and tongue muscles in the hypoglossal nucleus of macaque monkeys. Neurosci. Lett. 2231-35. Williams, P.L., R. Warwick, M. Dyson, and L. Bannister (1989) Gray's Anatomy, 37th British Edition. Edinburgh: Churchill Livingstone. Yoshida, Y., T. Miyazaki, M. Hirano, T. Shin, and T. Kanaseki (1981) Topographic arrangement of motoneurons innervating the suprahyoid and infrahyoid muscles: A horseradish peroxidase study in cats. Ann. Otol. Rhinol. Laryngol. 9.2259-266.