Cervical Primary Afferent Input to Vestibulospinal ...

THE JOURNAL OF COMPARATIVE NEUROLOGY 353:529-538 (1995)

RB ER mlation T Pression of ig n ox iproje •ocJ.uComp .sti the on Spinal easing and extko o k matomi of mmunocyco 7 l 4 toting the nt 2 double3 .

Cervical Primary Afferent Input to Vestibulospinal Neurons Projecting to the Cervical Dorsal Horn: An Anterograde and Retrograde Tracing Study in the Cat S. BANKOUL, T. GOTO, B. YATES, AND V.J. WILSON Laboratory of Neurophysiology, The Rockefeller University, New York, New York 10021-6399

ABSTRACT Vestibulospinal neurons in the caudal half of the medial and descending vestibular nuclei terminate in the cervical spinal cord, not only in the ventral horn and intermediate zone but also in the dorsal horn. The purpose of the present study was to examine whether the areas containing these vestibulospinal neurons are reached by cervical primary afferent& I n one group of experiments, wheat germ agglutinin-horseradish peroxidase conjugate and horseradish percaidase were pressure injected into spinal ganglia C labeled fibers and boutons in the caudal part (caudal to the dorsal cochlear nucleus) of the 2 ipsilateral medial and descending vestibular nuclei. This projection was verified in experiments -inCwhich wheat germ agglutinin-horseradish peroxidase conjugate was microiontophoretically 8 a ninto d ther caudal e v half e a ofl either e d the medial or the descending vestibular nuclei and injected revealed a n t retrogradely e r o g labeled r a cells d eonly l iny ipsilateral spinal ganglia C cells in C3. In another group of experiments, after microiontophoretic injections of Phaseolus 2 leucoagglutinin or Biocytin into either the medial or the descending vestibular nuclei, -oulgaris C anterogradely • labeled fibers and boutons were present i n the cervical spinal cord, mainly 7 bilaterally in the dorsal horn (laminae 1-VI) but also, to a lesser extent, in the ventral horn and primary afferent information to ,intermediate w i t h zone. The a existence m ofaa loop x that i mrelays u cervical m vestibulospinal neurons projecting to the cervical spinal cord, in particular the dorsal horn, may o f have implications for vestibular control over local information processing in the cervical dorsal horn.

© 1995 Wiley-Liss, Inc.

Indexing terms: v e s t i b u l a r nuclei, neck afferent, spinal ganglia, tracing study, spinal cord

Vestibulospinal projections have been the subject o f anatomical and physiological studies for many years. I n mammals, three vestibulospinal projections have been established: a lateral vestibulospinal tract, a medial vestibulospinal tract, and a caudal vestibulospinal tract (for review, see Wilson and Melvill Jones, 1979; Sjolung and Bjorldund, 1982; Hata and Watanabe, 1990). The lateral vestibulospinal tract originates in the lateral vestibular nucleus and courses in the ventral part of the anterior funiculus as far as lumbar levels (Carpenter, 1960; Petras, 1967; Brodal, 1974). The medial vestibulospinal tract originates chiefly in the medial vestibular nucleus, but also from the lateral and descending nuclei, and sends its fibers within the medial longitudinal fasciculus as far caudally as the midthoracic levels (Nyberg-Hansen, 1964; Wilson and Melvill Jones, 1 Originate in the lateral vestibular nucleus, parts o f the 9 7 o 9 1995 WILEY-LISS, INC. ) . T h e

medial vestibular nucleus, t h e descending vestibular nucleus, and the cell group f of Brodal and Pompeiano (1957; see also Peterson and Coulter, 1977; Peterson et al., 1978; Leong et al., 1984). For all three vestibulospinal projections, the termination area was originally considered to be the ventral horn and the intermediate zone. Subsequent physiological evidence of vestibulospinal influence outside these areas (Erulkar et al., 1966; Uchino and Hirai, Accepted September 16, 1994. This paper is dedicated to Prof. W. Zenker on the occasion of his 70th birthday. S. Bankoul's present address is Institute o f Anatomy, University o f Fribourg, Rue Gockel, 1, 1700 Fribourg, Switzerland. T. Goto's present address is Department of Orthopedic Surgery, Omiya Red Cross Hospital, 903 Kami-Ochiai, Yono City, Saitama 338, Japan. Address reprint requests to S. Bankoul, Institute of Anatomy, Rue Gockel 1, 1700 Fribourg, Switzerland.

530 S. BANKOUL E l At 1984) has been supported from the anatomical experiments . Injections into the caudal medial and descending v of Donevan et al. (1990, 1992) in the cat and of Bankoul and e g nuclei (cMV1V, cD1W). A f t e r a skin incision, th Neuhuber (1992) in the rat, showing vestibulospinal fibers tibular and terminals in the dorsal horn. neck muscles were removed from the occipital bone, th e . membrane, and the rostra' three cervi In a recent study, Bankoul and Neuhuber (1992) demon- atlantooccipital e vertebrae. After partial trepanation of the occipital bone strated that primary afferents from upper cervical spinal ca gentle retraction of the cerebellum, the floor of the ganglia in the rat project to the area from which antero- and l gradely labeled fibers to the dorsal horn of the cervical fourth ventricle was exposed on one side. The micropipette spinal cord can be traced. was prepositioned using coordinates calculated according t the atlas of Berman (1968). After visual identification of the It was the aim of this study to investigate in the cat, using o modern anterograde and retrograde tracer techniques, the cMVN and the cDVN and final positioning of the micropi. following two questions. First, which vestibular nuclei, and pette, iontophoretic injections of Biocytin (Sigma; two cats), which part o f these nuclei, do cervical primary afferent Phaseolus vulgaris leucoagglutinin (PHA-L; Vector; two cats), and WGA—HRP (Sigma; three cats) were made. fibers reach directly? Second, is there a vestibulospinal projection to the dorsal horn o f the cervical spinal cord Biocytin was applied as a 10% solution in 0.05 M Tris (pH 7.6) over a period of 10-15 minutes using glass pipettes originating in the area o f termination of direct primary with tip diameters of 25-40 pm and a 4 pA positive current. alTerents? PHA-L was used as a 2.5% solution i n 0.1 M sodium phosphate-buffered saline (pH 7.4). The tracer was injected over a period of 20-30 minutes using glass pipettes with tip MATERIALS AND METHODS diameters o f 20-40 g m and a 5 p A positive current Experiments were performed on 19 adult cats, 7 females delivered in pulses (10 seconds on, 10 seconds off). WGAand 12 males, weighing 2 7-5.8 kg. A l l animals were HRP was injected as a 2% solution in 0.9% NaCl over a initially anesthetized with Nembutal (sodium pentobarbi- period o f 20-25 minutes using glass pipettes with tip tal; Abbott; 30-35 mg/kg i.p.); supplementary doses (5 diameters o f 20-35 p.m and a 4 p A positive current mg/kg) were given to suppress withdrawal reflexes. For delivered in pulses (10 seconds on, 10 seconds off). After all injections of tracers, cats were fixed in a rigid head holder. injections, the glass pipettes were left in place for several All the surgery was performed under sterile conditions, and minutes before withdrawal. The survival time for animals the Principles of Laboratory Animal Care (NIH publication in Biocytin experiments was 4 and 6 days; survival ranged No. 85-23, revised 1985) were followed. from 14 to 20 days in PHA-L experiments and from 3 to 4 days in WGA-HRP experiments. Injection procedures In three control experiments, PHA-L and WGA-HRP Injections into cervical spinal ganglia. A f t e r a skin were injected into areas adjacent to the caudal vestibular incision, the dorsal neck muscles were removed on one side nuclei, including the nucleus of the solitary tract and parts of the cervical spine, and laminae of the vertebrae were of the dorsal motor nucleus of the vagus. After all injections, the wound was closed layer by layer, exposed. After a hemilaminectomy, spinal ganglia were and antibiotics were administered. For recovery, the cats exposed and wheat germ agglutinin-horseradish peroxidase placed in the Laboratory Animal Research Center (WGA-HRP; Sigma; 2% solution in 0.9% saline; six cats) were under veterinary supervision. was injected by pressure at the following levels: C2 (two cats) and C3, C4, C6, and C8 (one cat each). In one additional Tracer demonstration ganglion ( C Mannheim; 30% aqueous solution) was used as tracer. A After appropriate survival times, cats were deeply anes3 total of 2-4 p,I of tracer was injected in each experiment by thetized with Nembutal (45-50 mg/kg, i.p.) and perfused ), pressure over a period of 10-15 minutes using glass pi- through the ascending aorta with a warm prewash of 2,000 h o r with s e tip diameters ranging from 40 to 50 p.m. The ml of 0.9% saline containing 5,000 IU heparin. The animals pettes pipettes left in place for 5 minutes before withdrawal. were then perfused, depending on the tracer applied, with r a d i were s After survival times o f 2-4 days, animals were deeply the following solutions: 3,500 ml of a cold mixture of 4% h anesthetized and perfused as described further below. paraformaldehyde and 0.5% glutaraldehyde i n 0,05 M p e r phosphate buffer, pH 7.4 (for Biocytin experiments), or o x i 3,500 ml of a cold mixture of 2.5% glutaraldehyde and 0.5% paraformaldehyde in 0.05 M phosphate buffer, pH 7.4 (for d a s PHA-L experiments), or 3,500 ml of a cold mixture of 1% Abbreviations e paraformaldehyde and 1.25% glutaraldehyde i n 0.1 M CN cuneate nucleus ( phosphate buffer, p H 7.4 (for WGA-HRP experiments). DCN dorsal cochlear nucleus H Brainstems, spinal cords, and spinal ganglia were dissected; DMV dorsal motor nucleus of the \mgt.'s DVN postfixed for 2-4 hours i n the same solution used for descending vestibular nucleus R ECN external cuneate nucleus perfusion; and then stored overnight i n cold phosphate INT P nucleus intercalatus buffer containing 20% sucrose. IO inferior olive ; LVN Serial horizontal o r coronal cryostat sections o f the lateral vestibular nucleus MLF brainstem, the spinal cord, or the spinal ganglia were cut at medial longitudinal fasciculus B o MVN medial vestibular nucleus 40 p.m and stored in 0.05 M phosphate buffer. For Bi0Cytill e h NTS nucleus of the solitary tract experiments, the sections were then preincubated freely PH nucleus praepositus hypoglossi r i RB floating for 1.5-2 hours a t room temperature i n Trisrestiform body 5ST n g spinal trigeminal tract buffered saline (TBS; 0.05 M, pH 7.4) containing 1% Triton SVN superior vestibular nucleus X-100. After five rinses in TBS, the sections were incubated e r hypoglossal nucleus for 1 hour at room temperature in a freshly mixed solution

IfL ET At. nding ICISIOn, t h e bone, th

•ee cervi ipital ca b .o n e l icropipett 0 ..cording t e 0 of the ttion o re micropi;otwo cats), actor; tw f are ot made. I Tris shpipettes -e vl e sodium is injected c !s with tip ucurrent 3 r WGACI r Over a with tip e current •nAfter all irt several •.animals Si ranged un 3 to 4 GA-HRP estibular :nd parts by layer, the cats : Center

ly anes)erfused of 2,000 animals A, with e of 4% 0.05 M nts), or id 0.5% 7.4 (for a of 1% 0.1 M ments). isected; md for Psphate

of the cut at iocytin freely •TrisTriton ibated dution

p l O a fhorseradish peroxidase; Vectastain ABC Kit; Vector) i n M After a last rinse (5 x ) i n TBS, the sections were TBS. r A processed according to the diaminobenzidine (DAB) method e (Graham and Karnovsky, 1966) for 2-4 minutes. RIn PHA-L experiments, the sections were preincubated a Y freely floating for 1 hour at room temperature in a blocking g A buffer (BB; 10% fetal calf serum and 0.5% Triton X-100 in e 0.05 F M TBS, pH 7.2), followed by an incubation in antin Phaseolus antibody (goat, diluted 1:500 in BB) for 48 hours F tat 4°C. After five rinses in TBS, the sections were incubated E 1 hour a t room temperature i n TBS containing a for A biotinylated rabbit anti-goat IgG (Vectastain ABC Kit; R (Vector). After five rinses in TBS, sections were processed E for 1 hour in a freshly mixed solution of reagent A (avidin A N vPH) and reagent B (biotinylated horseradish peroxidase) in T After a last rinse in TBS, sections were processed TBS. iaccording to the DAB method for 3-4 minutes. I d In WGA-HRP experiments (anterograde and retrograde), N HRP was detected on free-floating sections using the tetrai P methylbenzidine (TMB) method (Mesulam, 1978). Every n U second section through the brainstem was processed accordD ing T to the DAB method (Graham and Karnovsky, 1966). H Sections processed with TMB were counterstained with T )neutral red. In the different experiments, labeled neurons O examined under brightfield illumination or under were a darkfield polarization optics ailing and Wassle, 1979). V n E d S RESULTS r T e WGA-HRP injections into different cervical I spinal ganglia a B g After pressure injection into spinal ganglia C2-C8, anteroU egradely labeled fibers were revealed in the ipsilateral cuneL nate and external cuneate nuclei. I n addition, fibers and O boutons could be detected in the medial (MVN) and the tdescending (DVN) vestibular nuclei. The fibers penetrated S B the two nuclei either from the lateral border of the DVN or P (from the ventral border of the DVN and MVN (Fig. 1A). IThe area i n which the fibers could be seen extended b N rostrocaudally from the caudal tip of the DVN and MVN to ithe level of the dorsal cochlear nucleus (DON; Fig. 2). The A odensity of anterogradely labeled fibers in this region varied L taccording to the level of the injected spinal ganglion. It was N ihighest after C2 and C3 injections and diminished as Einjections were made more caudally in C4 and C6. Injections ninto spinal ganglion C8 revealed only a very few scattered U yfibers. In all experiments, the density of fibers and boutons R l in the brainstem decreased in a caudal-rostral gradient Ofrom the caudal border of the MVN and DVN to the level of athe dorsal cochlear nucleus. In the coronal plane, fibers and N t en passant and terminal boutons were distributed over the S eDVN and MVN (Fig. 1B,C). d After HRP pressure injection into spinal ganglion C3, there was no difference in the labeling pattern from that seen after WGA-HRP injection into the ganglion, although the labeling seemed to be less intense.

Biocytin, PHA-L, and WGA-HRP injections into the caudal medial and descending vestibular nuclei (cMVN, c D . Retrogradely labeled cells i n cervical spinal ganglia. After WGA-HRP V N ) iontophoresis into the cMVN and cDVN,

531 DAB processing of the sections through the brainstem revealed injection sites confined to the boundaries of these nuclei. No diffusion to surrounding structures such as the nucleus of the solitary tract, the praepositus hypoglossi nucleus, the external cuneate nucleus, o r the reticular formation was seen. I n some experiments, sections processed with TMB revealed a slight diffusion to the nucleus of the solitary tract (Fig. 4). However, the ventrodorsal and mediolateral extension of different injection sites was reconstructed from serial coronal sections incubated with DAB, because it is known that injection sites stained with TMB appear to be larger than the actual zone of tracer uptake (Mesulam and Rosene, 1979; Lipp and Schwegler, 1980). Histochemical processing of the sections through spinal ganglia C eral spinal ganglia C 2 detectable in spinal ganglia Cg, T2, or T4 or in any contralat2 — eralT ganglia. The number of labeled cells in all animals was — C in C7 (between three and seven cells) and highest in 4 lowest (v Fe i g32. and 40 cells). After different survival times, C3e(between r7 the distribution was similar i n all animals T h e largest 5 ) d. a l e diameters of labeled cells ranged from 40 to 60 p.m. rN Anterogradely e to labeled fibers in the cervical spinal cord. rlAfter oa PHA-L g b eandl Biocytin injections into the cMVN and the recDVN, a d dimmunohistochemical processing o f the sections cthrough lbrainstem revealed rather small injection sites e l ye the confined within the borders of the two nuclei (Figs. 3, 4). No lldiffusion as to surrounding structures was detectable. WGAw e b HRP injection sites were described in the previous parargraph. e l Ael l injections produced labeling o f fibers i n the cervical spinal cord. T h e fibers coursed i n the dorsal ecolumns and dorsolateral funiculi on the ipsilateral side but dwere also found to a lesser degree in the ventrolateral and cventromedial funiculi (Fig. 6). Most of the fibers and en epassant and terminal boutons labeled in the gray matter were located i n the ipsilateral and, t o a lesser extent, l contralateral dorsal horn (Fig. 6, 7A,B). However, scattered l fibers were detectable in the intermediate zone and the sventral horn (Fig. 6, 7C). Fibers could be found as far i caudally as C8, but the density decreased below C5. Segments as far caudal as T6 were also removed and examined, nbut, even though the survival times were appropriate to i allow the tracer (PHA-L and WGA-HRP) to travel this far, ponly a few labeled fibers, mainly in the ventral horn, could be found. s In the dorsal horn, the majority of the fibers and boutons i spread over laminae II—V (Rexed, 1954) and, to a lesser l extent, laminae I and VI (Fig. 6). The density of fibers was agreater in the lateral and medial parts than in the intermediate part of the dorsal horn. t After WGA-HRP injections into the cMVN and cDVN, - some retrogradely labeled cells were also present in the cervical spinal cord. They were located bilaterally in the region of the central cervical nucleus, the ventral horn, and the dorsal horn (laminae III—V). These cells could be seen as far caudally as the C8 segment.

Control experiments A small PHA-L injection into the nucleus of the solitary tract, adjacent to the caudal tip of the DVN and encompassing the area where a little spread of reaction product was seen i n TMB-processed slides, did not reveal fibers o r boutons in the ipsi- or contralateral dorsal horn of the cervical spinal cord. Two WGA-HRP injections, centered into the nucleus of the solitary tract adjacent to the caudal

532

S. BANKOUL ET AU

C2 - C8

A

Fig. 1. A : Summarizing diagram of labeled primary afferent fibers P h o t o m i c r o g r a p h o f a fiber i n t h e M V N , forming a "basket like" in the caudal MVN and DVN, after wheat germ agglutinin-horseradish s t r u c t u r e (arrow) around a cell. C: Labeled fibers in the DVN showing peroxidase (WGA-HRP) injections into cervical spinal ganglia C 2 —C s • B : s o m e s w e l l i n g s ( a r r o w ) . S c a l e b a r s

E T A L .

FIZIMARY AFFERENT INPUT TO VESTIBULOSPINAL NEURONS tip of the DVN, but with a slight diffusion to the dorsal motor nucleus of the vagus, also did not result in labeling in the cervical dorsal horn.

MEMIN,

obex midline

10 0 0g

Fig. 2. T h e shaded area represents, i n a horizontal plane, t h e location of labeled fibers in the MVN and DVN after different injections into the spinal ganglia (C2—C drawn as a reference for the rostrocaudal extent. 8 ). T h e d o r s a l c o c h l e a r n u c l e u s ( D C N ) i s

like" awing

5

3

3

DISCUSSION Primary afferent projection A primary afferent projection from the neck to the MVN, DVN, and cell group X in mammals was reported in older (Corbin et al., 1937; Escolar, 1948) as well as in more recent studies (Rustioni and Macchi, 1968; Keller and Hand, 1970; Pfister and Zenker, 1984; Bakker et al., 1985; Edney and Porter, 1986; Pfaller and Arvidsson, 1988; Arvidsson and Pfaller, 1990; Prihoda et al., 1991). A cervical afferent input to the cell group X has been shown electrophysiologically (Wilson et al., 1976). Mergner et al. (1982) demonstrated activation of neurons in the MVN of the cat after natural stimulation of cervical afferent& One possible morphological basis for such an activation was described for the rat, in which anterograde (Pfaller and Arvidsson, 1988; Neuhuber and Zenker, 1989; Arvidsson and Pfaller, 1990) as well as retrograde (Bankoul and Neuhuber, 1990) tracer studies demonstrated a cervical primary afferent projection to the MVN, in particular to its caudal region. In most of these studies, the labeled fibers in vestibular nuclei were obtained after tracer applications to whole cervical spinal ganglia of different levels. Because such injections label all functional sets o f cervical primary afferents, a distinction between proprioceptive and exteroceptive qualities of these fibers would be impossible. With the rat, however, Neuhuber and Zenker (1989) were able to demonstrate, by the absence of labeled fibers after tracer application to the greater occipital nerve, that projections to the vestibular nuclei originate from proprioceptive neurons. I t is also o f interest that observations made in monkey (Fitz-Kitson, 1985) suggest that neck proprioceptors have direct, monosynaptic access to the descending and medial vestibular nuclei in primates and that, presumably, such a projection may also exist in humans. The present study provides evidence for a cervical primary afferent projection to vestibulospinal neurons in both the DVN and the MVN of the cat. Because WGA-HRP was used for injections into cervical ganglia, the labeled fibers could also be the result of trans-

Fig. 3. Photomicrograph of a Biocytin injection into the MVN (Exp. 1) at low magnification (A), showing the center of the injection (arrow) in relation to the MVN in a frontal plane (+ marks the fourth ventricle), and a higher magnification of a labeled cell (B) within the injection site. Scale bars = 125 p,m in A, 60 p-111 in B.

534 .

S

.

BANKOUL ET AL

obex midline

obex midline Fig. 4. S u m m a r i z i n g diagram showing reconstructions o f Biocytin (A; 1, 2), PHA-L (A; 5, 6) and WGA-HRP (B; 3, 4, 7) injections, as revealed with tetramethylbenzidine (TMB), which is known t o exaggerate the injection site, as it was mentioned earlier in Results. Injections are shown in the horizontal and corresponding transverse planes.

JL ET AL .

PRIMARY AFFERENT INPUT TO VESTIBULOSPINAL NEURONS

5

3

5

Fig. 6. C a m e r a lucida drawing o f transverse section through the spinal cord at C3 level showing, in summary, labeled axons and boutons after different tracer injections into the caudal DVN and MVN. The dots represent boutons, and the curve dashes represent axons. The arrow indicates the side ipsilateral to the injection.

A Fig. 5. Retrogradely labeled cells in the spinal ganglion C2 (A) and C3 (B) after a WGA-HRP injection into the DVN (Fig. 4B; 4). Scale bar = 100 um.

neuronal transport of this tracer after longer survival times, as reported previously (Trojanowski, 1983; Robertson and Grant, 1985; Pfaller and Arvidsson, 1988). The following facts, however, argue against such a possibility. First, comparison of the labeling pattern after shorter and longer survival times showed no substantial difference. Second, after injection o f free HRP, which i s considered not t o be transported transneuronally in adult mammals (Trojanowski, 1983; Janjua and Leong, 1983; Neal and Carey, 1986) except after intracellular application (Hongo et al., 1981; Triller and Korn, 1981), the areas with labeled terminals and fibers were the same as after WGA-HRP injections. The lower labeling intensity most probably was due to the lower sensitivity of the HRP technique. The anterogradely demonstrated primary afferent projection was corroborated by retrograde labeling of cells in the cervical spinal ganglia after WGA-HRP injections into the termination area for cervical primary afferents within the DVN and MVN. The pattern of anterograde and retrograde labeling (see Results) indicates that in the cat the direct primary afferent projection to the caudal vestibular nuclei originates mainly from upper cervical spinal ganglia. This caudal region of the two nuclei, reached by primary afferent fibers, overlaps wIth the rostrocaudal distribution of vestibulospinal neurons projecting to the dorsal horn of the spinal cord, as discussed further below.

Vestibulospinal projection In recent morphological studies in the cat (Donevan et al., 1990, 1992; Shinoda et al., 1992), it was demonstrated that

vestibulospinal neurons project not only to the ventral horn and the intermediate zone but also to the dorsal horn of the upper cervical spinal cord (C1--C4). After PHA-L injections into different parts o f the vestibular nuclear complex, mainly the rostral MVN and adjacent parts of the DVN and LVN, scattered fibers in the cervical dorsal horn could be found bilaterally as far dorsal as lamina II. For the rat, Bankoul and Neuhuber (1992) reported a vestibulospinal projection originating in the caudal part of the MVN and projecting mainly to the dorsal horn of the upper cervical spinal cord. An interesting aspect of this vestibulospinal projection is revealed in a recent study in the rat (Bankoul and Neuhuber, 1992), where cells of origin of the projection to the cervical dorsal horn are found in a restricted area of the caudal MVN, which receives a primary afferent input from the upper cervical ganglia. Double labeling experiments in the rat (Bankoul, 1994) reveal primary afferent baskets around vestibulospinal neurons in the MVN, retrogradely labeled after injections confined to the dorsal horn of different levels of the upper cervical spinal cord. Previously reported injection sites into the vestibular nuclei of the cat, giving rise to vestibulospinal collaterals to the dorsal horn, were not located in the area of the vestibular nuclei that receives cervical primary afferent input (Donevan e t al., 1990, 1992). Therefore, o u r attention was focused on a possible vestibulospinal projection to the dorsal horn from within the "primary afferent termination area." Our anterograde tracing results corroborate certain aspects o f the results obtained from studies in the rat (Bankoul and Neuhuber, 1992) and cat (Donevan et al., 1990, 1992), but there are certain differences as well. In the rat, cells projecting to the cervical dorsal horn are located only in the caudal part of the MVN (Bankoul and Neuhuber, 1992). Neither the rostral part of this nucleus nor the DVN contribute to this projection. I n the cat, however, after more rostral injections (rostra! to an imaginary line drawn through the middle of the rostrocaudal

536

S. BANKOUL ET

Fig. 7. Photomicrographs of labeled fibers with boutons after PHA-L injections (Exp. 5, 6) into the DVN and MVN. Fibers are located in the dorsal horn laminae II and III (A) and V and VI (B) and in the ventral horn (C). Scale bar = 50 p,m.

extent of the MVN) into the MVN and DVN, a scattered projection to the upper dorsal horn was detectable (Donevan et al., 1990, 1992). Our experiments were focused on the caudal part of the MVN and DVN and showed that this area contains cells projecting mainly bilaterally t o the cervical dorsal horn (laminae Furthermore, the rostrocaudal extent of boutons and fibers in the cervical spinal cord obtained in our experiments varies from that described in the rat (Bankoul and Neuhuber, 1992), where C6 was the most caudal segment with detectable fibers and boutons. In our experiments, the density of labeled fibers and boutons decreased below C5, but some scattered fibers and boutons could be seen as far caudal as segment T6. The distribution pattern changed below C5, so that i n the upper thoracic cord the few remaining fibers were located in the ventral part of the gray matter. It is interesting to note that the highest density of vestibular fibers and boutons i s i n the upper cervical segments, whose dorsal root ganglia produce the highest density of primary afferent fibers detectable in the MVN and DVN. In the white matter of the rat, most of the labeled axons coursed in the ipsilateral ventral funiculus, especially in the medial longitudinal fasciculus and, to a lesser extent, in the dorsal and lateral funiculi. Some fibers, however, were also detectable in the same funiculi on the contralateral side. In the cat, after more rostral injections (Donevan et al., 1990, 1992), the fibers coursed bilaterally in all funiculi, with a predominance in the ventromedial, ventrolateral, and lateral funiculi. In our experiments, fibers were also present throughout all funiculi except the lateral one, but a majority of the fibers was located in the dorsal and dorsolateral funiculi.

Functional considerations It is of interest to note that vestibulospinal neurons, projecting predominantly to the dorsal horn of the cervical spinal cord, are situated in a zone of the DVN and MVN

where cervical primary afferents end. In the rat, it has been reported in a morphological study that primary afferents that contact a small, well-localized cell group in the caudal MVN (Bankoul and Neuhuber, 1992), which in turn projects to the superficial laminae of the upper cervical dorsal horn, are mostly proprioceptive (Neuhuber and Zenker, 1989). In an earlier electrophysiological study in the cat, Fredrickson et al. (1965) were able to demonstrate that neck afferents, having access to the vestibular nuclei, were mostly proprioceptive and not extero- or nociceptive. The question arises, what influence do they exert on processing in the upper cervical dorsal horn? A clue as to the function o f the vestibulospinal projection to the dorsal horn may be provided by the termination pattern of the fibers (laminae IT-V and, to a lesser extent, I and VI), which coincides remarkably well with the termination field of thin-caliber afferent fibers (A8/ C) from muscles and joints (laminae IV, V; Mense, 1990) as well as of many cutaneous afferents, which respond to extreme positions of joints, muscular hypwda, and changes caused by inflammation (Mense, 1986; Proske et al. 1988). Vestibulospinal neurons, projecting to the cervical dorsal horn and receiving primary afferent information from neck muscles, could modulate this input to the cervical dorsal horn during head movements. This modulation may, in part, be inhibitory because there are findings suggesting that -y-aminobutyric acid (GABA) may be the transmitter of some of these vestibulospinal fibers, at least in the rabbit (Blessing et al., 1987) and rat (Nomura et al., 1984; Nagai et al., 1985; Jones et al., 1991). Numerous studies (e.g., Lindsay et al., 1976; Thoden and Schmidt, 1979; Mergner et al., 1983; Ezure and Wilson, 1984; Dutia and Hunter, 1985; Wilson, 1991) have described the important role that afferent information from neck proprioceptors plays in the reflex control of posture. Our findings in the cat may represent the morphological basis for a cervicovestibulocervical loop, whose main function could be the maintenance o f an appropriate balance between primary afferent information from neck muscles a• different secondary afferents reaching the vestibular nuclei. 1111

:OUL ET AL J I V I V I A RY ACKNOWLEDGMENTS A F F E The authors thank Mrs. S. Wong for skillful technical R E N T a I N grs• s M. Kaczorowski and Mr. C. Weber for expert photoP U g s r a and DC 02187. 02619 T p i hi T cs LITERATURE CITED O w t, V E S o ka ganglia in the rat, studied by anterograde transport of WGA-HR.P. J. T I B Comp. Neurol. 292:349-362. rrn Bakker, D.A., F.J.R Richmond, V.C. Abtahams, and J. Courvifle (1985) U L O vk c Patterns of primary afferent termination in the external cuneate nucleus i S P I . from cervical axial muscles in the cat. J. Comp. Neurol. 241:467-479. de B N A L sT a, neurons projecting to the dorsal horn: A double labeling study in the rat. N s Experientia 50:A70. oh M o Bankoul, S., and W.L. Neuhuber (1990) A cervical primary afferent input to E ki nr vestibular nuclei, as demonstrated by retrograde transport of wheat os U , germ agglutinin-horseradish peroxidase i n the rat. Exp. Brain Res. us 79:405-411. R Js lBankoul, . S., and W.L. Neuhuber (1992) A direct projection from the medial . O ,t vestibular nucleus t o the cervical spinal dorsal horn o f the rat, as ,L S N u demonstrated by anterograde and retrograde tracing. Anat. Embryo!. a .. 185:77-85. S n Berman, A l . (1968) The brainstem of the cat, a cytoarchitectonic atlas with (d H d 5 1y stereotaxic coordinates. Madison: The University of Wisconsin Press. o K W.W., S.C. Hedger, and W.H. Oertel (1987) Vestibulospinal 9w 3 it has been Blessing, at, . in rabbit includes GABA-synthesizing neurons. Neurosci. Lett. 9l pathway 7 afferents lady P 80:158-162. 4a s in the caudal Brodal, A. (1974) Anatomy of the vestibular nuclei and their connections. In f )s H.H. Kornhuber led): Handbook o f Sensory Physiology, Vol. VI/1. at turn projects C Berlin: Springer, pp. 239-351. ls d dorsal horn, ee A., and O. Pompeian° (1957) The vestibular nuclei in the cat. J. lu Anat. Lker, 1989). In Brodal, 91:438-454. ri e t, Fredrickson Carpenter, vp M.B. (1960) Fiber projections from the descending and lateral n r Leck afferents, ip vestibular nuclei in the cat. Am. J. Anat. 107:1-22. ( f tostly proprio- Corbin, c KB., W.T. Lhamon, and D.W. Petit (1937) Peripheral and central 1 connections o f the upper cervical dorsal root ganglia in the rhesus iestion arises, ao o 9 monkey. J. Comp. Neurol. 66:405-414. lr in the upper 9r A I L , M . Neuber-Hess, and P.K. Rose (1990) Multiplicity of nction of the Donevan, pt 0 vestibulospinal projections to the upper cervical spinal cord of the cat: A rg smay be pro) study with the anterograde tracer Phaseolus vulgar ie (laminae II-V eNeurol. 302:1-14. C Comp. m cides remark- Donevan, d u c oJ.A. a gMacDonald, g l u t i n i P.A. n . Brennan, and P.K. Rose (1992) en is l eA.H., ab Morphology of single vestibulospinal collaterals in the upper cervical diber afferent . ne J r spinal cord of the cat. II. Collaterals originating from axons outside the minae IV, V; t ventral funiculi. J. Comp. Neurol. 322:343-359. yy r 'erents, which r Dutia, aN M i l , and M.J. Hunter (1985) The sagittal vestibulocollic reflex and ular hypoxia, aa its interaction with neck proprioceptive afferents in the decerebrate cat. 1986; Proske flIl J. Physiol. (London) 359: i7-29. f ecting to the Edney, p eH a D.P., and J.D. Porter (1986) Neck muscle afferent projection to the 'rent informar brainstem of the monkey: Implication for the neural control of gaze. J. rg Comp. Neurol. 250:389-398. os sinput to the er S.D., J.M. Sprague, B.L. Whitsel, S. Dogan, and P.J. Jannetta This module- Crulkar, j ns (1966) Organization of the vestibular projection to the spinal cord of the e e are findings a ti cat. J. Neurophysiol. 29:626-664. c ) may be the Escolar, in J. (1948) The afferent connections of the 1st, 2nd, and 3rd cervical ts ibers, at least nt nerves in the cat. A n analysis by Marchi and Rasdolsky methods. J. it Comp. Neurol. 89:79-91. iomura et al., p o Ezure, us a K. and V.J. Wilson (1984) Interaction of tonic neck and vestibular in the forelimb of the decerebrated cat. Exp. Brain Res. 54: i; Thoden and tnNreflexes sn 289-292. sand Wilson, tS D. (1985) The direct connections of the C2 dorsal root ganglia in o save described Fitz-Ritson, oc f the Macaca irus monkey: Relevance to the chiropractic profession. J. m from neck ve Manipulative Physiol. Then 8:147-156. C e posture. Our Fredrickson, J.M., D. Schwarz, and H.H. Kornhuber (1965) Convergence 4, s and interaction of vestibular and deep somatic afferents upon neurons in Lological basis -a -lain function , tt h e vestibular nuclei of the cat. Acts Otolaryngol. 61:168-188. C RC., and M.J. Karnovsky (1966) The early stages of absorption of in e balance be- Graham, 8 injected horseradish peroxidase in the proximal tubules of mouse kidney: b muscles and d d cytochemistry by a new technique. J. Histochem. Cytou Ultrastructural ;ibular o chem. 14:291-302. l r o s s a p l i r n o a o l t

Hata, Y., and I . Watanabe (1990) Distribution o f lateral and medial vestibulospinal neurons i n the vestibular nucleus i n the cat. The Sixteenth Barany Society Meeting, Tokyo, Japan, abstract 57, p. 89, published by Japan Society for Equilibrium Research. Hongo, T., N. Kudo, M. Yamashita, N. Ishizuka, and H. Mannen (1981) Transneuronal passage of intraaxonally injected horseradish peroxidase (HRP) from group lb and II fibers into secondary neurons in the dorsal horn of the cat spinal cord. Biomed. Res. 2:722-727. thing, R.B., and H. Wassle (1979) Visualization of the HRP reaction product using the polarization microscope. Neurosci. Lett. 13:7-11. Janjua, M.Z., and S.K. Leong (1983) Cervical cord neurons labeled by horseradish peroxidase application t o the sciatic nerve i n the rat. Neurosci. Lett. 38:115-118. Jones, RE., C.J. Holmes, E. Rodriguez-Veiga, and L. Mainville (1991) GABA-synthesizing neurons in the medulla: Their relationship to serotonin-containing and spinally projecting neurons i n the rat. J. Comp. Neurol. 313:349-367. Keller, J.H., and P.J. Hand (1970) Dorsal root projections t o nucleus cuneatus of the cat. Brain Res. 20:1-17. Leong, S.K., J.Y. Shieh, and W.C. Wong (1984) Localizing spinal-cordprojecting neurons in adult albino rats. J. Comp. Neurol. 228:17. Lindsay, KW., T.D.M. Roberts, and J.R. Rosenberg (1976) Asymmetric tonic labyrinth reflexes and their interaction w i t h neck reflexes i n the decerebrated cat. J. Physiol. (London) 261:583-601. Lipp, H T. , and H. Schwegler (1980) Improved transport of horseradish peroxidase after injection with a non-ionic detergent (nonidet P-40) into mouse cortex and observations on the relationship between spread at the injection site and amount of transported label. Neurosci. Lett. 20:49-54. Lovick, TA., and J.P. Robinson (1983) Bulbar raphe neurons with projections to the trigeminal nucleus caudalis and the lumbar cord in the rat: A fluorescence double-labeling study. Exp. Brain Res. 50:299-308. Mense, S. (1986) Slowly conducting afferent fibers from deep tissues, neurological properties, and central nervous actions. Progr. Sens. Physiol. 6:139-219. Mense, S. (1990) Relationship between functional and morphological properties in single primary afferent neurons. In W. Zenker and W.L. Neuhuber (eds): The Primary Afferent Neuron. New York: Plenum Press, pp. 201-211. Mergener, T., D. Anastopulos, and W. Becker (1982) Neuronal responses to horizontal neck deflection in the group x region of the cat's medullary brainstem. Exp. Brain Res. 45:196-206. Mergner, T., L. Deecke, and W. Becker (1983) Vestibular-proprioceptive interactions: Neurophysiology and psychophysics. I n E. Horn led): Multimodai Convergences in Sensory Systems. Fortschritte der Zoologie, Bd. 28. Stuttgart: Fischer, pp. 241-252. Mesularn, M.M. (1978) Tetramethyl benzidine for horseradish peroxidase neurohistochemistry: A noncarcinogenic blue reaction product with superior sensitivity for visualizing neural afferents and efferent& J. Histochem. Cytochem. 26:106-117. Mesulam, MAL, and D.L. Rosene (1979) Sensitivity in horseradish peroxidase neurohistochemistry: A comparative and quantitative analysis of the nine methods. J. Histochem. Cytochem. 27:763-773. Nagai, T. , T. Maeda, H . Imai, P.L. McGeer, and E.G. McGeer (1985) Distribution o f GABA-T-intensive neurons i n the rat hindbrain. J. Comp. Neurol. 231:260-269. Neal, T.L., and R.G. Carey (1986) Modification of transport specificity of horseradish peroxidase: Anterograde and transneuronal properties. Soc. Neurosci. Abstr. 12:1565. Neuhuber, W U , and W. Zenker (1989) Central distribution o f cervical primary afferents in the rat, with emphasis on proprioceptive projection to vestibular, perihypoglossal, and upper thoracic spinal nuclei. J. Comp. Neurol. 280:231-253. Nomura, I., E. Senba, T. Kubo, T. Shiraishi, T. Matsunaga, M. Tohyama, Y. Shiotani, and J.Y. Wu (1984) Neuropeptides and -y-aminobutyric acid in the vestibular nuclei of the rat: An immunohistochemical analysis. Distribution. Brain Res. 311:109-118. Nyberg-Hansen, R. (1964) Origin and termination o f fibres from the vestibular nuclei descending in medial longitudinal fasciculus. J. Comp. Neurol. 22:355-367. Peterson, BM., and J.D. Coulter (1977) A new long spinal projection from the vestibular nuclei in the cat. Brain Res. 122:351-356. Peterson, B.W., R.A. Maunz, and K. Fukushima (1978) Properties of a new vestibulospinal projection, the caudal vestibulospinal tract. Exp. Brain Res. 32:287-292.

538 Petras, J.M. (1967) Cortical, tectal, and tegmental fiber connections in the spinal cord of the cat. Brain Res. 6:275-324. Pfaller, K., and J. Arvidsson (1988) Central distribution of trigeminal and upper cervical primary afferents i n the r a t studied by anterograde transport of horseradish peroxidase conjugated to wheat germ agglutinin. J. Comp. Neural. 268:91-108. Pfister, J., and W. Zenker (1984) The splenius capitis muscle of the rat, architecture and histochemistry, afferent and efferent innervation, as compared with that of the quadriceps muscle. Anat. Embryo'. 169.19-89. Prihoda, M., M.S. Hiller, and R. Mayr (1991) Central projections of cervical primary afferent fibers in the_guinea pig: An HRP and WGA-HRP tracer study. J. Comp. Neurol. 308:418-431. Proske, U., H.-G. Schaible, and R.F. Schmidt (1988) Joint receptors and kinesthesia. Exp. Brain Res. 72:219-224. Rexed, B. (1954) A cytoarchitectonic atlas of the spinal cord in the cat. J. Comp. Neurol. 100:297-379. Robertson, B., and G. Grant (1985) A comparison between wheat germ agglutinin- and choleragenoid-horseradish peroxidase as anterogradeiy transported markers in central branches of primary sensory neurons in the rat with some observations in the cat. Neuroscience 14:895-905. Rustioni, A., and G. Macchi (1968) Distribution of dorsal root fibers in the medulla oblongata of the cat. J. Comp. Neurol. 134:113-126.

S. BANKOUL ET AL, Shinoda, Y.T., Y. Ohgaki, Y. Sugiuchi, and T. Futami (1992) Morphology o f single medial vestibulospinal tract axons in the upper cervical spinal cord of the cat. J. Comp. Neurol. 316:151-172. SjOlung, B., and A. BjOrklund (1982) Brainstem Control of Spinal M e nisms. Amsterdam: Elsevier. Thoden, U., and P. Schmidt (1979) Vestibular-neck interaction in abducens d,. neurons. In R. Granit and O. Pompeian° (eds): Reflex Control of Posture and Movement. Amsterdam: Elsevier, pp. 561-566. Triller, A., and H . Korn (1981) Interneuronal transfer o f horseradish peroxidase associated with exo/endocytic activity on adjacent menbranes. Exp. Brain Res. 43:233-236. Trojanowski, J.Q. (1983) Native and derivatized lectins for in vivo studies of neuronal connectivity and neuronal cell biology. J. Neurosci. Methods 9: 185-204. Uchino, Y., and N. Hirai (1984) Axon collaterals of anterior semicircular canal—Activated vestibular neurons and their coactivation of extraocular and neck motoneurons in the cat. Neurosci. Res. 1:309-225. Wilson, V.J. (1991) Vestibulospinal and neck reflexes: Interaction in the vestibular nuclei. Arch. Ital. Biol. 129:43-52. Wilson, V.J., and G. Melvin Jones (1979) Mammalian Vestibular Physiology. New York: Plenum Press. Wilson, V.J., M. Maeda, J.L. Franck, and H. Shimazu (1976) Mossy fiber neck and second-order labyrinthine projections t o cat floccalus. J. Neurophysiol. 39:301-309.

1