vestibular nucleus projections to the edinger–westphal and anteromedian nuclei of rabbits

Brain Research 963 (2003) 121–131www.elsevier.com/ locate/brainres

Research report

V estibular nucleus projections to the Edinger–Westphal andanteromedian nuclei of rabbits

*Carey D. BalabanDepartments of Otolaryngology, Neurobiology and Communication Disorders, University of Pittsburgh, Eye and Ear Institute, 203 Lothrop Street,

Pittsburgh, PA 15213,USA

Accepted 8 November 2002

Abstract

The Edinger–Westphal nucleus (EW), anteromedian nucleus (AM) and adjacent neurons in the ventral tegmental area (VTA) aresources of preganglionic parasympathetic innervation of intraocular smooth muscle, including blood vessels, pupillary muscle and theciliary body in mammals. They also have central connections that are believed to affect parasympathetic outflow indirectly. This studyutilized anterograde transport of biotinylated dextran amine andPhaseolus vulgaris leucoagglutinin to demonstrate direct projections fromthe vestibular nuclei to the Edinger–Westphal and anteromedian nuclei in rabbits. The rabbit AM and adjacent VTA contain moderate tointensely choline acetyltransferase (ChAT)-immunopositive neurons. The rabbit EW, by contrast, is nearly devoid of ChAT-immuno-positive neurons. Vestibular nucleus projections to these regions originate from all levels of the superior, medial and lateral vestibularnuclei, but do show topographic organization. The densest terminations were observed in AM and the ventral and central aspects of EW.The projections to AM terminate in both ipsilaterally and in a narrow paramedian region. Predominantly ipsilateral terminations wereobserved in VTA. Terminations on ChAT-positive cells in AM and VTA were verified in three rabbits. It is suggested that projections tosome intensely ChAT-positive AM and VTA neurons may be a substrate for vestibular influences on lens accommodation, pupillaryconstriction and regulation of intraocular circulation during changes in posture and gravitoinertial challenges. The projections toChAT-negative (and weakly immunoreactive) cells in AM,VTA and EW, on the other hand, are likely to contribute vestibular signals to avariety of motor responses via descending pathways. 2002 Elsevier Science B.V. All rights reserved.

Theme: Motor systems and sensorimotor integration

Topic: Vestibular system

Keywords: Vestibular nucleus; Parasympathetic nervous system; Intraocular innervation; Edinger–Westphal nucleus; Anteromedian nucleus

1 . Introduction [7,9,40,44,53,54]. The caudal medial vestibular nucleusand the inferior vestibular nucleus can influence postcra-

Recent anatomic and physiologic studies have demon- nial parasympathetic and sympathetic outflow, either di-strated direct connections between the vestibular nuclei rectly or indirectly, via descending projections to theand brain stem regions that influence sympathetic and nucleus of the solitary tract, dorsal motor vagal nucleus,parasympathetic outflow (for a review, see Ref. [11]). nucleus ambiguus and the rostral ventrolateral medullaryThese so-called ‘vestibulo-autonomic pathways’ originate reticular formation. An ascending pathway also originatesfrom an extensive region within the vestibular nuclei, from the dorsal aspect of the superior vestibular nucleus,which includes the dorsal aspect of the superior vestibular pars alpha (or caudoventral aspect) of the lateral vestibularnucleus, pars alpha (or caudoventral aspect) of the lateral nucleus, and the caudal half of the medial vestibularvestibular nucleus, and the caudal half of the medial nucleus and the inferior vestibular nucleus. This ascendingvestibular nucleus and the inferior vestibular nucleus projection terminates densely in the parabrachial nucleus,

which has both ascending projections to the amygdala,hypothalamus and prefrontal cortex, and descending con-*Tel.: 11-412-647-2298; fax:11-412-647-0108.

E-mail address: [email protected](C.D. Balaban). nections to medullary regions that control sympathetic and

0006-8993/02/$ – see front matter 2002 Elsevier Science B.V. All rights reserved.PI I : S0006-8993( 02 )03955-0

mailto:[email protected]

122 C.D. Balaban / Brain Research 963 (2003) 121–131

parasympathetic outflow. These pathways have been sug- kg body weight) were premedicated with atropinegested as substrates for vestibular contributions to phenom- methylnitrate (0.06–0.2 mg, s.c.) and anesthetized withena as diverse as cardiovascular control, respiratory pat- sodium pentobarbital (40 mg/kg, 4 ml /kg total volume,terns, gastrointestinal function and anxiety disorders i.v.). Surgical procedures were performed under aseptic[6,10,11]. conditions. The head was fixed with zygoma clamps in a

There is also physiologic evidence that vestibular in- stereotaxic apparatus (Narishige Instruments, Tokyo,formation influences parasympathetic outflow to the eye. Japan) with the head tilted 458 nose-down. LidocaineMarkham et al. [34] demonstrated monocular accommoda- (1–2%, s.c.) was injected along the incision line and thetive (lens-thickening) changes during slow (18 /s) roll aponeurosis of the superficial cervicoauricularis musclerotation, which peaked when the ipsilateral ear was down was divided. The underlying muscle layers were retractedand were absent when the ipsilateral ear was up. Further- to expose the occipital bone, atlas and atlanto-occipitalmore, limited experimental evidence suggests that coordi- membrane. The medulla was exposed by removing thenated actions of smooth and striated muscle of the eye can atlanto-occipital membrane and enlarging the foramenbe elicited by vestibular stimulation. For example, me- magnum dorsally with rongeurs. After completion of thechanical (‘air-puff’) stimulation of the utricle in cats and injections, the craniotomy was packed with Gelfoam andmonkey has been reported to produce conjugate nystag- the soft tissues were sutured in layers. Penicillin (80,000–mus, with pupillary constriction during fast phases and 100,000 U/day, i.m.) was administered during the survivaldilation during slow phases [19]. Since these responses period. Consistent with NIH requirements, the procedureswere unaffected by removing sympathetic input to the eye, in these studies have been reviewed and approved by thethey were presumed to be strictly parasympathetic. Electri- University of Pittsburgh Institutional Animal Care andcal stimulation of the utricle in cats also elicited lens Utilization Committee.accommodation [35] via parasympathetic mechanisms. The rabbits were given 13% mannitol (2.3–3.0 ml /kg,However, with the exception of Furuya and Markham’s i.v.) to increase the exposure of the floor of the fourth[20] report that some collaterals of vestibulo-oculomotor ventricle by osmotically shrinking the cerebellum andprojections project to the Edinger–Westphal nucleus in brain stem. Using the obex and the floor of the fourthcats, there is no anatomical evidence of a direct pathway ventricle as landmarks, rabbits were given an iontophoreticfrom the vestibular nuclei to the potential sources of injection ofPhaseolus vulgaris leucoagglutinin [PHA-L,preganglionic parasympathetic fibers to the ciliary gang- Vector Laboratories, 2.5% solution in sodium phosphate-lion. buffered saline (PBS), pH 8.0] and/or biotinylated dextran

Recent studied indicate that preganglionic parasympa- amine (BDA, Molecular Probes, 7–10% solution in PBS,thetic innervation of the eye originates primarily from pH 7.0) into the vestibular nuclei and/or nucleus pre-somata within the anteromedian nucleus and adjacent positus hypoglossi (10–15mm tip diameter, 4mA positiveventral tegmental area, with a lesser contribution arising current, 10 min). Some rabbits received a pressure in-from the Edinger–Westphal nucleus [2,14–16,46]. Evi- jection of 50–100 nl (100mg/ml) recombinant tetanusdence from cats indicates that these preganglionic para- toxic C fragment (rTTC, Boeringer-Mannheim), Fluoro-sympathetic neurons include intensely ChAT-immuno- Gold or tetramethylrhodamine-dextran-amine (Molecularreactive neurons in the anteromedian nucleus and ventral Probes) into the contralateral vestibular nuclei, ventrolater-tegmental area, while weakly ChAT-immunopositive and al medulla or cerebellar nodulus. This communication isChAT-negative neurons tend to project centrally. This restricted to data from sites confined to the vestibularpaper demonstrates a more extensive direct projection from nuclei and nucleus prepositus hypoglossi, with no evidencethe vestibular nuclei to both choline acetyltransferase of spread to the cerebellum, nucleus tractus solitarius orChAT-positive and ChAT-negative neurons in the Eding- the dorsal medullary reticular formation.er–Westphal nucleus and anteromedian nucleus in rabbits.Hence, it is suggested that vestibular information may 2 .2. Immunohistochemical and histochemical proceduresinfluence distinct groups of central and peripheral projec-tion neurons. After survival times ranging from 4 to 10 days, the

rabbits were euthanized with a pentobarbital overdose andperfused transcardially with phosphate-buffered saline

2 . Materials and methods (PBS) followed by the paraformaldehyde–lysine–sodiummetaperiodate (PLP) fixative of McLean and Nakane [37].

2 .1. Surgical procedures The brains were post-fixed for 18–24 h at 48C in asolution of 4% paraformaldehyde–30% sucrose in 50 mM

In addition to a series of animals produced for this phosphate buffer and cryoprotected in a 30% sucrose–50study, this report includes data from animals utilized in mM phosphate buffer solution for 2–3 days. Frozenprevious studies of the organization of vestibular nuclear sections (40mm, transverse plane) were cut on a slidingoutput pathways [7,9]. New Zealand white rabbits (2.7–4.4 microtome and sets of every fourth to sixth section were

C.D. Balaban / Brain Research 963 (2003) 121–131 123

Fig. 1. The distribution of choline acetyltransferase (ChAT)-immunopositive neurons is shown in representative coronal sections through (A) theanteromedian nucleus (am) and (B)–(D) the oculomotor (III) and Edinger–Westphal (ew) nuclei of a rabbit. The approximate borders of theEdinger–Westphal nucleus are outlined. (A) The anteromedian nucleus consisted predominantly of ChAT-immunopositive somata. Two examples ofChAT-positive somata in the adjacent ventral tegmental area (near the exiting oculomotor nerve fibers) are indicated by arrows. (B, C) In contrast to theChAT-immunopositivity of virtually all cells in the oculomotor nucleus (n.III), ChAT-immunopositive neurons were very rare in the Edinger–Westphalnucleus (ew). The moderately ChAT-immunopositive ew neuron in panel (B) (arrow) is shown in higher magnification below in panel (D).

placed in 50 mM phosphate buffer (pH 7.2–7.4). For cleared in xylene and coverslipped with either Permount orlonger term storage, sections were maintained at220 8C in non-fluorescent DPX (Fluka).a solution of 30% sucrose–30% ethylene glycol solution in Sections from both control rabbits and from three BDA-50 mM phosphate buffer. injected animals were also stained immunohistochemically

Axonally transported PHA-L was visualized immuno- for choline acetyltransferase (ChAT). The free-floatinghistochemically by standard published methods (e.g., Ref. sections from the injected animals were stained initially for[7]). For visualizing BDA transport, free-floating 40mm BDA as described above, using a nickel-enhanced DABfrozen sections were rinsed successively in distilled water chromogen, then rinsed in three changes of PBS. Sections(3310 min), 0.9% H O and distilled water to suppress from the injected and the control animals were placed2 2

endogenous peroxidase activity, followed by a preincuba- overnight at 48C in a block solution of PBS containingtion for 2 h at room temperature in 0.5% Triton X-100 in 0.3% Triton X-100, 2.5% normal horse serum (Sigma) andPBS. After a rinse in PBS, the sections were incubated for 1% bovine serum albumin (BSA, Sigma). The sections1 h in ABC reagent, rinsed in buffer and reacted for were then transferred to a 1:100 dilution of goat anti-ChATvisualizing sites of peroxidase activity with either a nickel- (Chemicon, A144P) for 2 h at room temperature, followedenhanced DAB (for BDA) or a standard DAB (for PHA-L, by 48 h at 48C. The tissue was then incubated in a 1:1002 mg DAB, 8.3ml H O in 10 ml 500 mM sodium acetate dilution of biotinylated rabbit anti-goat IgG (Vector Lab-2 2

buffer, pH 6.0) chromogen. Sections were mounted on oratories) in PBS containing 2% BSA for 1 h at roomsubbed slides, dehydrated through a graded alcohol series, temperature. After three rinses in PBS, ChAT immuno-


reactivity was visualized with standard ABC Vectastain tial nucleus of Cajal, and ventrally and ventrolaterally by(Vector Laboratories) methods and a standard DAB (2 mg the oculomotor nucleus. In marked contrast to the subja-DAB, 8.3 ml H O in 10 ml 500 mM sodium acetate cent oculomotor nucleus, the Edinger–Westphal nucleus at2 2

buffer, pH 6.0) chromogen. the level of the oculomotor nucleus lacked ChAT-immuno-positive neurons (Fig. 1). The ventral aspect of theEdinger–Westphal nucleus extends rostral to the

3 . Results oculomotor nucleus. This extension corresponds to the areadesignated the anteromedian nucleus in cats and monkeys,

The classically defined rabbit Edinger–Westphal nucleus which also projects to the ciliary ganglion [15,48,51].occupies the rostral half of the oculomotor nuclear com- Unlike the dorsal /caudal aspect of the Edinger–Westphalplex [1]. It is bordered dorsally by the periaqueductal grey, nucleus, the vast majority of neurons in the ventral aspectlaterally by the nucleus of Darkschewitch and the intersti- of the anteromedian nucleus displayed moderate-to-intense

Fig. 2. Photomicrographs of ChAT-immunopositive neurons in the rabbit oculomotor nucleus (n. III), dorsal (AMd) and ventral (AMv) anteromedialnucleus and the adjacent ventral tegmental area (VTA). (A) Intensely ChAT-immunoreactive oculomotor nucleus motoneurons have large multipolarsomata and long, rectilinear dendrites. (B) The ventral aspect of the anteromedian nucleus contains a relatively dense packing of ChAT-immunoreactiveneurons with fusiform to stellate somata in the ventral aspect. Individual neurons varied from intense to moderate ChAT immunoreactivity; relatively fewcells were ChAT-negative. (C) The dorsal aspect of the anteromedian nucleus contains a sparse distribution of ChAT-immunoreactive neurons with stellatesomata. (D–F) Intensely ChAT-immunopositive neurons were scattered near the oculomotor nerve rootlets in the ventral tegmental area. The somataranged in shape from stellate (D–E) to fusiform (F).


ChAT-immunopositivity in the rabbit (Fig. 1). Two re-gions can be distinguished in the rabbit anteromediannucleus, a ventral portion containing densely packed,stellate to fusiform shaped, ChAT-positive neuronalsomata (Fig. 2B) and a dorsal portion containing a sparserdistribution of ChAT-positive (Fig. 2C) and ChAT-nega-tive neurons. The shapes of these neurons contrastedsharply in appearance from the intensely immunopositivemultipolar somata and rectilinear proximal dendrites of theoculomotor nucleus motoneurons (Fig. 2A). The ChAT-positive anteromedian nucleus neurons were also contigu-ous spatially with intensely ChAT-positive stellate tofusiform neurons in the ventral tegmental area, which werescattered near intensely ChAT-positive axons in theoculomotor nerve root (Fig. 2D–F). These ventral tegmen-tal area neurons had the same range of somatic shapes andChAT staining intensities as the anteromedian nucleusneurons. Immunopositive axons of some ChAT-positiveanteromedian nucleus and ventral tegmental area neuronscould be traced into the adjacent ChAT-immunopositivefascicles of the oculomotor nerve root. Because theseChAT-positive cells are both located in sites that give riseto parasympathetic innervation of the ciliary ganglion inother species [2,15,16,33,46,51], these cell bodies arelikely to include the source of preganglionic sympatheticinnervation of the eye [46].

Anterogradely labeled axons could be traced to theEdinger–Westphal nucleus and anteromedian nucleus frominjections of BDA and PHA-L confined to the vestibularnuclei. Examples of two BDA injection sites and onePHA-L injection site are shown in photomicrographs inFig. 3. The injection sites consisted of a dense ‘core’where the tracer filled the somata and neuropil, surroundedby a ‘halo’ region containing heavily labeled neuronalsomata. The extent of the core and halo regions of BDA(Fig. 4) and PHA-L (Fig. 5) sites that produced labeling inthe Edinger–Westphal and anteromedian nuclei includedthe superior, medial and lateral vestibular nuclei. Twopreviously described PHA-L injection sites in the inferiorvestibular nucleus [not shown; 91020 (Ref. [9], Fig. 2) and91027 (Ref. [8], Fig. 8)] and injections involving primarilythe nucleus prepositus hypoglossi failed to produce antero-grade labeling in either the Edinger–Westphal or theanteromedian nucleus.Fig. 3. Photomicrographs of representative injection sites in the vestibu-

Anterogradely labeled axons from these injection siteslar nuclei that produced anterogradely labeled axons in the Edinger–Westphal and anteromedian nuclei. (A) This BDA injection site was entered the ipsilateral medial longitudinal fasciculus andcentered in the ventrolateral aspect of the superior vestibular nucleus (S).ascended to the level of the oculomotor nucleus. ManyNote the dense reaction at the center of the site and the presence ofaxons entered the trochlear and oculomotor nuclei andlabeled cell bodies within the ‘halo’ region. 4v, fourth ventricle. (B) This

formed relatively dense terminal arbors aroundBDA injection was centered near the caudal pole of the superiormotoneurons (Fig. 6B). Axons entered the rostroventralvestibular nucleus (S), without encroachment into adjacent regions of

either the medial vestibular nucleus (M) or pars beta of the lateral aspect of the Edinger–Westphal nucleus, primarily alongvestibular nucleus (L ). (C) This PHA-L injection involved the lateralb the midline, and formed varicose terminal arbors nearaspect of the medial vestibular nucleus (M) and the medial aspect of the neuronal somata (Fig. 7D) in both the median and ipsila-pars alpha of the lateral vestibular nucleus (L ), without involvement ofa teral paramedian aspects of the region of the nucleus (Fig.either the medial aspect of the medial vestibular nucleus or nucleus

6B) which contains few ChAT-positive neurons. However,prepositus hypoglossi (PH). The calibration bar represents 1 mm inpanels (A) and (C) and 2 mm in panel (B). as reported previously by Furuya and Markham [20] in the


Fig. 4. Chartings of BDA injection sites in the vestibular nuclei that produced anterograde transport to the anteromedian and Edinger–Westphal nuclei. Thesites are defined as both the dense core and halo region. Eight injection sites are charted on a standard series of camera lucida drawings of transversesections from a paraffin embedded reference brain (sectioned at 10mm). The nomenclature for the vestibular nuclei distinguishes the superior vestibularnucleus (S), rostral medial vestibular nucleus (M ), caudal medial vestibular nucleus (M ), lateral vestibular nucleus parsa (L ), lateral vestibular nucleusr c a

parsb (L ), lateral vestibular nucleus pars gamma (or Deiters nucleus, L ) and inferior vestibular nucleus (I). Nucleus prepositus hypoglossi (PH), nucleusb g

tractus solitarius (NTS), the dorsal motor nucleus of the vagus nerve (X), abducens nucleus (6), group y (y), inferior cerebellar peduncle (ICP) and thefastigial (F), interposed (Ip) and dentate (De) cerebellar nuclei are also indicated. The sections are arranged in series from caudal (lower right) to rostral(upper left).

cat, an occasional axon extended from the oculomotor 4 . Discussionnucleus to form varicosities en passage in the ventralaspect of the Edinger–Westphal nucleus. Isolated axons The current literature indicates that there are twobearing varicosities en passage were also observed. The different distribution patterns of ChAT-immunoreactivemost elaborate terminal branches tended to be in the neurons in the mammalian Edinger–Westphal nucleus,medial (or median) aspect of the ventral two-thirds of the anteromedian nucleus and ventral tegmental area. TheseEdinger–Westphal nucleus (Fig. 6B); less dense termina- patterns also appear to correspond to the origin of oculartions were present near the dorsal and lateral margins of parasympathetic innervation in different species. Inthe nucleus (ipsilateral to the injection site). These termi- macaques, ChAT-immunopositive neurons are prominentnals were also found in both the dorsal and ventral aspects in the Edinger–Westphal nucleus [23,38], but appear to beof the ipsilateral anteromedian nucleus (Fig. 6A). The absent in the ventral tegmental area among the oculomotorvaricose axons in the anteromedian nucleus were associ- nerve rootlets. Correspondingly, the ocular preganglionicated with both ChAT-positive (Fig. 7C) and ChAT-nega- parasympathetic neurons in primates appear to be confinedtive (Fig. 7B) somata or dendrites in the double-stained to the Edinger–Westphal complex and the anteromediansections. In addition, axons formed varicosities with nucleus [2,14–16]. By contrast, rabbits (this communica-intensely ChAT-positive neurons ipsilaterally in the adja- tion), cats [25,46] and rats [3] have many ChAT-immuno-cent ventral tegmental area (Fig. 7A), which are a source positive neurons in the anteromedian nucleus and ventralof parasympathetic innervation of the ciliary ganglion [46]. tegmental area, but relatively few ChAT-positive neuronsThe location and distribution of the terminations in the are located in the Edinger–Westphal nucleus. The ChAT-Edinger–Westphal nucleus, anteromedian nucleus and ven- immunopositive neurons in the rabbit anteromedian nu-tral tegmental area did not vary as a function of the cleus and ventral tegmental area have a similar range oflocation of injection sites within the vestibular nuclei. staining intensities and fusiform to stellate somatic shapes


Fig. 5. Chartings of PHA-L injection sites in the vestibular nuclei that produced anterograde transport to the anteromedian and Edinger–Westphal nuclei.The sites are defined as both the dense core and halo region. Eight injection sites are charted on a standard series of camera lucida drawings of transversesections from a paraffin embedded reference brain (sectioned at 10mm). The nomenclature for the vestibular nuclei distinguishes the superior vestibularnucleus (S), rostral medial vestibular nucleus (M ), caudal medial vestibular nucleus (M ), lateral vestibular nucleus parsa (L ), lateral vestibular nucleusr c a

parsb (L ), lateral vestibular nucleus parsg (or Deiters nucleus, L ) and inferior vestibular nucleus (I). Nucleus prepositus hypoglossi (PH), nucleusb g

tractus solitarius (NTS), the dorsal motor nucleus of the vagus nerve (X), abducens nucleus (6), group y (y), inferior cerebellar peduncle (ICP) and thefastigial (F), interposed (Ip) and dentate (De) cerebellar nuclei are also indicated. The sections are arranged in series from caudal (lower right) to rostral(upper left). Note that projections to the Edinger–Westphal nucleus originated from sites involving all levels of the medial, lateral and superior vestibularnuclei.

as reported in cats [46]. Although most preganglionic groups of ChAT-positive neurons are likely to include bothparasympathetic neurons in cats were reported to be preganglionic parasympathetic cells [46] and centrallyintensely ChAT-immunopositive, some preganglionic neu- projecting neurons [26,42,46]. No topographic organiza-rons were ChAT-negative [46]. Similar data on ocular tion of the vestibular nucleus projections was apparent andpreganglionic parasympathetic neurons are unavailable for the terminations do not appear to be distributed pref-other species. As a result, it is not possible to state with erentially within the anteromedian nucleus, the Edinger–certainty that a particular neuron in these regions projects Westphal nucleus or the ventral tegmental area. Theperipherally or centrally on the basis of the intensity of widespread origin of these connections contrasts with theChAT immunoreactivity. However, it seems likely that more restricted origin of inferior and medial vestibularChAT-positive neurons in the anteromedian nucleus and nucleus projections to medullary preganglionic parasympa-adjacent ventral tegmental area include a population of thetic regions [5,7,9,40]. Thus, the Edinger–Westphal andpreganglionic parasympathetic neurons. anteromedian nuclei may be influenced by vestibular input

This study presents anatomic evidence of a direct from a different combination of sources than the descend-projection from all levels of the medial, lateral and ing vestibular projections to lower brain stem parasympa-superior vestibular nuclei to the Edinger–Westphal nucleus thetic and sympathetic regions [2,13–16,33,48,51].and anteromedian nucleus of rabbits. The terminals appear The terminations of vestibular nucleus afferents in theto be associated with both intensely to moderately ChAT- Edinger–Westphal nucleus, anteromedian nucleus and ven-positive somata in the anteromedian nucleus and adjacent tral tegmental area have the potential to affect parasympa-ventral tegmental area, as well as ChAT-negative neurons thetic outflow to the eye both directly (via preganglionicin anteromedian and Edinger–Westphal nuclei. Based upon parasympathetic neurons) and indirectly (via centrallythe limited direct evidence in the literature, the former projecting neurons). Recent experiments have shown that


Fig. 6. Camera lucida drawings of BDA axonal labeling in the anteromedian nucleus, ventral tegmental area and Edinger–Westphal nucleus from aninjection site into the superior vestibular nucleus. (A) This section is taken through the rostral aspect of the anteromedian nucleus near the level of Fig. 1A.Note the paramedian distribution of axons in the dorsal aspect of the anteriomedian nucleus (AMd), and the strictly ipsilateral distribution in the ventralaspect of the anteromedian nucleus (AMv) and among the rostralmost oculomotor nerve rootlets (III) in the adjacent ventral tegmental area VTA. (B) Thisdrawing shows a section through the rostral aspect of the oculomotor nucleus (n.III) of the same animal. Axon terminals in the oculomotor nucleus werelocated ipsilaterally in a dorsomedial region (neurons that project to the contralateral superior rectus muscle) and in a ventral and lateral regioncontainingipsilateral inferior rectus motoneurons. Varicose axons were observed in the medial and lateral aspects of the Edinger–Westphal nucleus (EW) and theVTA. Axons in passage were present in the medial longitudinal fasciculus (mlf). The cerebral aqueduct (Aq) is also illustrated.

peripherally directed axons from rat Edinger–Westphal cells give rise to spinal versus peripheral projections [16]nucleus neurons terminate predominantly on ciliary gang- and that separate groups of neurons project to the faciallion cells, smooth muscle cells of ciliary ganglion ar- nucleus and the inferior olive [26]. It has been suggestedterioles and near ciliary ganglion veinules [26]. Hence, that the pathway to the facial nucleus and spinal trigeminalthese efferents may subserve both preganglionic para- nucleus may be components of circuitry for a blinksympathetic and local (intraganglionic) vasoregulatory response to intense illumination [26], while projections toroles. The preganglionic parasympathetic fibers likely the rostral ventrolateral medulla may coordinate sympa-contribute to the control of lens accommodation, pupillary thetic and parasympathetic outflow to the eye. Because thediameter, choroidal blood flow, uveal blood flow, and cerebellar nuclei also project to the Edinger–Westphal andmaintenance of the integrity of the blood–aqueous humor anteromedian nuclei [36,49], projections to the lateralbarrier [12,18,21,52]. reticular nucleus, inferior olive, cerebellar cortex and

Previous tracing studies also demonstrated that the cerebellar nuclei have been regarded as potential feedbackEdinger–Westphal nucleus (and its rostral extension, the (or feedforward) signals to cerebellar circuits that modulateanteromedian nucleus) contains neurons that project to the preganglionic parasympathetic outflow to the eye [22].brain stem and spinal cord [16,26,32,33,41,51]. Antero- Thus, vestibular nucleus projections to ChAT-negativegrade and retrograde tracing methods have documented neurons in the anteromedian and Edinger–Westphal nu-projections to the spinal cord, inferior olive, spinal trigemi- cleus may provide an indirect influence on parasympatheticnal nucleus, facial nucleus, rostroventral reticular nucleus pathways to the eye.and lateral reticular nucleus [26,32,33,41,51]. Direct pro- Vestibular nuclear projections are merely one of ajections to the cerebellar cortex and nuclei have also been variety of inputs to the Edinger–Westphal and anterome-reported in cats and monkeys [36,42,45,47]. Limited dian nuclei. Previous studies have shown that these nucleidouble retrograde tracer evidence indicates that different also receive input from structures that include retinorecipi-


Fig. 7. Examples of anterogradely labeled axon terminals in the ventral tegmental area (A), anteromedian nucleus (B, C) and Edinger–Westphal nucleus(D). (A) A BDA-labeled varicose axon was applied to the dendrite of a ChAT-positive neuron (*) in the ventral tegmental area, immediately ventrolateralto the anteromedian nucleus. A portion of a terminal is also visible in the neuropil to the right of the ChAT-positive soma. (B) An elaborate BDA-labeledterminal is shown enveloping a ChAT-negative soma (star) in the anteromedian nucleus. (C) A coarse BDA-labeled terminal is shown in the neuropil of theanteromedian nucleus surrounded by ChAT-positive somata (*). (D) An elaborate, fine caliber BDA-labeled terminal is shown in a ChAT-negative regionwithin the Edinger–Westphal nucleus.

ent pretectal nuclei [4,28], interstitial nucleus of Cajal, activation of extraocular muscles, lens accommodation,nucleus of the posterior commissure, nucleus of and pupillary diameter regulation [31].Darkschewitsch, the midbrain periaqueductal gray [27], A brief consideration of responses to gravitoinertialcerebellar nuclei [36,49] and the spinal cord [55]. These stimulation suggests that vestibular information is germanediverse inputs may reflect the complexity involved in to the control of lens accommodation, pupillary constric-controlling intraocular smooth muscle and coordinating tion and regulation of intraocular circulation. When wetheir activity with skeletal musculature. Vergence eye fixate on an object approaching our nose, three motormovements are a prominent example of the intricacy of responses occur simultaneously: a convergent (disconju-this control problem: the eyes display a coordinated gate) eye movement mediated by the extraocular (striated)


[3] D.M. Armstrong, C.B. Saper, A.I. Levey, B.H. Wainer, R.D. Terry,muscles, lens accommodation [relaxation of the ciliaryDistribution of cholinergic neurons in rat brain: demonstrated by the(smooth) muscle leading to lens thickening] and pupillaryimmunocytochemical localization of choline acetyltransferase, J.

constriction (smooth muscle response). Although previous Comp. Neurol. 216 (1983) 53–68.reports have documented effects of otolith stimulation on ¨[4] J.A. Buttner-Ennever, B. Cohen, A.K.E. Horn, H. Reisine, Pretectallens accommodation [34,35] and pupillary diameter [19], projections to the oculomotor complex of the monkey and their role

in eye movements, J. Comp. Neurol. 366 (1996) 348–359.there is no published evidence regarding changes in either[5] C.D. Balaban, Projections from the parabrachial nucleus to thepupillary diameter or lens accommodation during linear

vestibular nuclei in rabbits: a visceral relay to vestibular circuits, in:vestibulo-ocular responses to movements along the naso-Abstracts, Midwinter Meeting of Association for Research in

occipital axis [39]. The existence of direct vestibular Otolaryngology, 1997, p. 69.projections to ChAT-positive neurons in the anteromedian [6] C.D. Balaban, Vestibular autonomic regulation, Curr. Opin. Neurol.

12 (1999) 29–33.nucleus and adjacent ventral tegmental area motivates[7] C.D. Balaban, Vestibular nucleus projections to the parabrachialinvestigation of whether linear vestibulo-ocular reflexes

nucleus in rabbits: implications for vestibular influences on au-include an accommodative component.tonomic function, Exp. Brain Res. 108 (1996) 367–381.

Finally, direct and indirect vestibular nucleus projections [8] C.D. Balaban, G. Beryozkin, Organization of vestibular nucleusto sources of parasympathetic innervation of the blood projections to the dorsal cap of Kooy in rabbits, Neuroscience 62

(1994) 1217–1236.vessels of the choroid, uvea, ciliary body and iris may[9] C.D. Balaban, G. Beryozkin, Vestibular nucleus projections tocontribute to regulation of intraocular circulation [12]

nucleus tractus solitarius and the dorsal motor nucleus of the vagusduring changes in posture. Postural alterations from stand-nerve: potential substrates for vestibulo-autonomic interactions, Exp.

ing to supine (or inverted) are accompanied by increased Brain Res. 98 (1994) 200–212.intraocular pressure [17,29,30,50], with no change in [10] C.D. Balaban, J.F. Thayer, Neurological bases for balance–anxiety

links, J. Anxiety Disord. 15 (2001) 53–79.aqueous humor outflow [17]. Because increased intraocular[11] C.D. Balaban, B.J. Yates, Vestibulo-autonomic interactions: a tele-pressure passively reduces the vascular perfusion pressure

ologic perspective, in: S. Highstein, A. Popper (Editors), Springerof the eye [12], choroidal perfusion of intraocular struc-Handbook of Auditory Research: Vestibular Anatomy and Physi-

tures is maintained by both local autoregulatory and ology, Springer (in press).neuronal mechanisms [24,43]. Activation of parasympa- [12] A. Bill, Circulation in the eye, in: Microcirculation, Part 2,thetic innervation of the rabbit eye produces a prominent Handbook of Physiology, Section 2: The Cardiovascular System,

Vol. IV, American Physiological Society, Bethesda, MD, 1984, pp.vasoconstrictor effect in the anterior uvea and iris and a1001–1034.weak vasodilation of the choroid and the ciliary body in

[13] A. Brodal (Ed.), Neurological Anatomy in Relation to Clinicalrabbits [12]. The effects of cholinergic blockade suggest Medicine, Oxford University Press, New York, 1969, p. 431, 443.that the vasoconstrictor fibers have strong tonic activity in [14] R.M. Burde, Direct parasympathetic pathway to the eye: revisited,rabbits and the degree of vasoconstriction appears to be Brain Res. 463 (1988) 158–162.

[15] R.M. Burde, A.D. Loewy, Central origin of oculomotor parasympa-independent of pupillary constriction. Hence, it is sug-thetic neurons in the monkey, Brain Res. 198 (1980) 434–439.gested that vestibular nucleus influences on preganglionic

[16] R.M. Burde, J.J. Parelman, M. Luskin, Lack of unity of Edinger–parasympathetic neurons could contribute to vascular Westphal nucleus projections to the ciliary ganglion and spinal cord,regulation in the anterior iris and uvea during postural Brain Res. 249 (1982) 379–382.adjustments. [17] K.H. Carlson, J.W. McLaren, J.E. Topper, R.F. Brubaker, Effect of

body position on intraocular pressure and aqueous flow, Invest.Ophthalmol. Vis. Sci. 28 (1987) 1346–1352.

[18] H. Davson (Ed.), The Physiology of the Ocular and CerebrospinalA cknowledgements Fluids, Academic Press, New York, 1980, pp. 56–81.

[19] M. DeSantis, B.E. Gernandt, Effect of vestibular stimulation onThe author wishes to thank Maria Freilino, Gloria pupillary size, Exp. Neurol. 30 (1971) 66–77.

[20] N. Furuya, C.H. Markham, Arborization of axons in oculomotorLimetti and Jean Betsch for expert surgical and histologi-nucleus identified by vestibular stimulation and intracellular in-cal assistance. These studies were supported by R01jection of horseradish peroxidase, Exp. Brain Res. 43 (1981) 289–

DC00739 and P01 DC03417. A Core Grant for Vision 303.Research (EY08098) provided technical support for [21] P.D.R. Gamlin, Y. Zhang, R.A. Clendaniel, L.E. Mays, Behavior ofmaintenance of critical laboratory equipment. identified Edinger–Westphal neurons during ocular accommodation,

J. Neurophysiol. 72 (1994) 2368–2382.[22] H. Hultborn, K. Mori, N. Tsukahara, Cerebellar influence on

parasympathetic neurons innervating intra-ocular muscles, BrainR eferences Res. 159 (1978) 269–278.

[23] T. Ichikawa, T. Shimizu, Organization of choline acetyltransferase-[1] Y. Akagi, The localization of motor neurons innervating the containing structures in the cranial nerve motor nuclei and spinal

extraocular muscles in the oculomotor nuclei of the cat and rabbit, cord of the monkey, Brain Res. 779 (1998) 96–103.using horseradish peroxidase, J. Comp. Neurol. 181 (1978) 745– [24] J.W. Keil, Modulation of choroidal autoregulation in the rabbit, Exp.761. Eye Res. 69 (1999) 413–429.

[2] K. Akert, M.A. Glicksman, W. Lang, A. Huber, The Edinger– [25] H. Kimura, P.L. McGeer, J.H. Peng, E.G. McGeer, The centralWestphal nucleus in the monkey. A retrograde tracer study, Brain cholinergic system studied by choline acetyltransferase immuno-Res. 184 (1980) 491–498. histochemistry in the cat, J. Comp. Neurol. 200 (1981) 151–201.


[26] J. Klooster, A.J.M. Beckers, G.F.J.M. Vrensen, J.J.L. van der Want, and regions that mediate autonomic function in the rat, J. Vestib.The peripheral and central projections of the Edinger–Westphal Res. 7 (1997) 63–76.nucleus in the rat. A light and electron microscopic tracing study, [41] H. Qvist, E. Dietrichs, Afferents to the lateral reticular nucleus fromBrain Res. 632 (1993) 260–271. the oculomotor region, Anat. Embryol. 175 (1986) 261–269.

[27] J. Klooster, G.F.J.M.Vrensen, New indirect pathways subserving the [42] G.K. Røste, E. Dietrichs, Cerebellar cortical and nuclear afferentspupillary light reflex: projections of the accessory oculomotor nuclei from the Edinger–Westphal nucleus in the cat, Anat. Embryol. 178and the periaqueductal gray to the Edinger–Westphal nucleus, Anat. (1988) 59–65.Embryol. 198 (1998) 123–132. [43] C.E. Riva, P. Titze, M. Hero, B. Petrig, Effect of acute decreases of

¨[28] J. Klooster, G.F.J.M. Vrensen, L.J. Muller, J.J.L. van der Want, perfusion pressure on choroidal blood flow in humans, Invest.Efferent projections of the olivary pretectal nucleus in the albino rat Ophthalmol. Vis. Sci. 38 (1997) 1752–1760.subserving the pupillary light reflex and related reflexes. A light [44] D.A. Ruggiero, E.P. Mtui, K. Otake, M. Anwar, Vestibular afferentsmicroscopic tracing study, Brain Res. 688 (1995) 34–46. to the dorsal vagal complex: substrate for vestibulo-autonomic

[29] A.C. Kothe, The effect of posture on intraocular pressure and interactions in the rat, Brain Res. 743 (1996) 294–302.pulsatile ocular blood flow in normal and glaucomatous eyes, Surv. [45] H. Sekiya, K. Kawamura, S. Ishikawa, Projections from the Eding-Ophthalmol. 38 (Suppl.) (1994) S191–S197. er–Westphal complex of monkeys as studied by means of retrograde

[30] G.K. Krieglstein, M.E. Langham, Influence of body position on the axonal transport of horseradish peroxidase, Arch. Ital. Biol. 122intraocular pressure of normal and glaucomatous eyes, Ophthal- (1984) 311–319.mologica (Basel) 171 (1975) 132–145. [46] A. Strassman, P. Mason, F. Eckenstein, R. Baughman, R.

[31] R.J. Leigh, D.S. Zee, The Neurology of Eye Movements, Davis, Maciewicz, Choline acetyltransferase immunohistochemistry ofPhiladelphia, 1991. Edinger–Westphal and ciliary ganglion afferent neurons in the cat,

[32] A.D. Loewy, C.B. Saper, Edinger–Westphal nucleus: projections to Brain Res. 423 (1987) 293–304.the brain stem and spinal cord in the cat, Brain Res. 150 (1978) [47] T. Sugimoto, K. Itoh, N. Mizuno, Direct projections from the1–27. Edinger–Westphal nucleus to the cerebellum and spinal cord of the

[33] A.D. Loewy, C.B. Saper, N.D. Yamodis, Re-evaluation of the cat: an HRP study, Neurosci. Lett. 9 (1978) 17–22.efferent projections of the Edinger–Westphal nucleus in the cat, [48] T. Sugimoto, K. Itoh, N. Mizuno, Localization of neurons givingBrain Res. 141 (1978) 153–159. rise to the oculomotor parasympathetic outflow: a HRP study in cat,

[34] C.H. Markham, S.G. Diamond, N.E. Simpson, Ocular accommoda- Neurosci. Lett. 7 (1977) 301–305.tive changes in humans induced by positional changes with respect [49] T. Sugimoto, N. Mizuno, K. Uchida, Distribution of cerebellar fiberto gravity, Electroencephalogr. Clin. Neurophysiol. 42 (1977) 332– terminals in the midbrain visuomotor areas: an autoradiographic340. study in the cat, Brain Res. 238 (1982) 353–370.

[35] C.H. Markham, M.S. Estes, R.H.I. Blanks, Vestibular influences on [50] A. Tarkkanen, J. Leikola, Postural variations of the intraocularocular accommodation in cats, Int. J. Equilibrium Res. (1973) pressure as measured with the Mackay–Marg tonometer, Acta102–115. Ophthalmol. 45 (1967) 569–575.

[36] P.J. May, J.D. Porter, P.D.R. Gamlin, Interconnections between the [51] K. Toyoshima, E. Kawana, H. Sakai, On the neuronal origin ofprimate cerebellum and midbrain near-response regions, J. Comp. afferents to the ciliary ganglion in the cat, Brain Res. 185 (1980)Neurol. 315 (1992) 98–116. 67–76.

[37] I.W. McLean, P.K. Nakane, Periodate–lysine–paraformaldehyde [52] G. Westheimer, S.M. Blair, The parasympathetic pathways tofixative. A new fixative for immunoelectron microscopy, J. Histoch- internal eye muscles, Invest. Ophthalmol. 12 (1973) 193–197.em. Cytochem. 22 (1974) 1077–1083. [53] B.J. Yates, C.D. Balaban, A.D. Miller, K. Endo, Y. Yamaguchi,

[38] M.-M. Mesulam, E.J. Mufson, A.I. Levey, B.H. Wainer, Atlas of Vestibular inputs to the lateral tegmental field of the cat: potentialcholinergic neurons in the forebrain and upper brainstem of the role in autonomic control, Brain Res. 689 (1995) 197–206.macaque based on monoclonal choline acetyltransferase immuno- [54] B.J. Yates, L. Grelot, I.A. Kerman, C.D. Balaban, J. Jakus, A.D.histochemistry and acetylcholinesterase histochemistry, Neuro- Miller, The organization of vestibular inputs to nucleus tractusscience 12 (1984) 669–685. solitarius (NTS) and adjacent structures in the cat brainstem, Am. J.

[39] G.D. Paige, D.L. Tomko, Eye movement responses to linear head Physiol. 267 (1994) R974–R983.motion in the squirrel monkey. I. Basic characteristics, J. Neuro- [55] R.P. Yezierski, Spinomesencephalic tract: projections from thephysiol. 65 (1991) 1170–1182. lumbosacral spinal cord in the rat, cat, and monkey, J. Comp.

[40] J.D. Porter, C.D. Balaban, Connections between the vestibular nuclei Neurol. 267 (1988) 131–146.

vestibular nucleus projections to the edinger–westphal and anteromedian nuclei of rabbits

Documents