Cholinergic Elements in the Zebraﬁsh Central Nervous …...tem in the vertebrate central nervous...

Cholinergic Elements in the ZebrafishCentral Nervous System: Histochemical

and Immunohistochemical Analysis

DIEGO CLEMENTE, ANGEL PORTEROS, EDUARDO WERUAGA,

JOSE RAMON ALONSO, FRANCISCO JAVIER ARENZANA, JOSE AIJON,

AND ROSARIO AREVALO*

Departamento de Biologıa Celular y Patologıa, Instituto de Neurociencias de Castilla yLeon, Universidad de Salamanca, E-37007 Salamanca, Spain

ABSTRACTRecently, the zebrafish has been extensively used for studying the development of the central

nervous system (CNS). However, the zebrafish CNS has been poorly analyzed in the adult. Thecholinergic/cholinoceptive system of the zebrafish CNS was analyzed by using choline acetyl-transferase (ChAT) immunohistochemistry and acetylcholinesterase (AChE) histochemistry inthe brain, retina, and spinal cord. AChE labeling was more abundant and more widely distrib-uted than ChAT immunoreactivity. In the telencephalon, ChAT-immunoreactive (ChAT-ir) cellswere absent, whereas AChE-positive neurons were observed in both the olfactory bulb and thetelencephalic hemispheres. The diencephalon was the region with the lowest density of AChE-positive cells, mainly located in the pretectum, whereas ChAT-ir cells were exclusively located inthe preoptic region. ChAT-ir cells were restricted to the periventricular stratum of the optictectum, but AChE-positive neurons were observed throughout the whole extension of the lami-nation except in the marginal stratum. Although ChAT immunoreactivity was restricted to therostral tegmental, oculomotor, and trochlear nuclei within the mesencephalic tegmentum, awidespread distribution of AChE reactivity was observed in this region. The isthmic regionshowed abundant AChE-positive and ChAT-ir cells in the isthmic, secondary gustatory andsuperior reticular nucleus and in the nucleus lateralis valvulae. ChAT immunoreactivity wasabsent in the cerebellum, although AChE staining was observed in Purkinje and granule cells.The medulla oblongata showed a widespread distribution of AChE-positive cells in all mainsubdivisions, including the octavolateral area, reticular formation, and motor nuclei of the cranialnerves. ChAT-ir elements in this area were restricted to the descending octaval nucleus, theoctaval efferent nucleus and the motor nuclei of the cranial nerves. Additionally, spinal cordmotoneurons appeared positive to both markers. Substantial differences in the ChAT and AChEdistribution between zebrafish and other fish species were observed, which could be importantbecause zebrafish is widely used as a genetic or developmental animal model. J. Comp. Neurol.474:75–107, 2004. © 2004 Wiley-Liss, Inc.

Indexing terms: acetylcholine; cholinergic; cholinoceptive; teleost

The neurotransmitter acetylcholine (ACh) is one-step-synthesized in the cytoplasm of cholinergic neurons by theenzyme choline acetyltransferase (ChAT) and is degradedat the synaptic cleft by the enzyme acetylcholinesterase(AChE). To investigate the cholinergic–cholinoceptive sys-tem in the vertebrate central nervous system (CNS), dif-ferent antibodies against ACh (Geffard et al., 1985), cho-linergic receptors (Andre et al., 1984; Wenthold et al.,1990), and vesicular ACh transporter (Andre et al., 1984;Ichikawa et al., 1997) have been developed, even thoughChAT antibodies are regarded as the most reliable mark-ers for the identification of cholinergic elements. Addition-ally, comparison of the distribution of ChAT with results

obtained after AChE histochemistry provides evidencesabout the cholinoceptive nature of some AChE-positive

Grant sponsor: Spanish Fondo de Investigaciones Sanitarias; Grantnumber: PI021730; Grant sponsor: Junta de Castilla y Leon.

*Correspondence to: Rosario Arevalo, Dpto. de Biologıa Celular y Pato-logıa, Universidad de Salamanca, Facultad de Medicina, Campus Miguelde Unamuno, c/Alfonso X el Sabio, 1, E-37007 Salamanca, Spain. E-mail:[email protected]

Received 6 August 2003; Revised 25 November 2003; Accepted 26 Jan-uary 2004

DOI 10.1002/cne.20111Published online in Wiley InterScience (www.interscience.wiley.com).

THE JOURNAL OF COMPARATIVE NEUROLOGY 474:75–107 (2004)

© 2004 WILEY-LISS, INC.

neurons, as previously described (Nickell and Shipley,1988; Molist et al., 1993; Le Jeune and Jourdan, 1994;Crespo et al., 1999).

ACh plays an integral role in basic cerebral functions. Itmodulates the neuronal responsiveness to sensory stimuli(McKenna et al., 1989; Murphy and Sillito, 1991) and

Abbreviations

A anterior thalamic nucleusAChE acetylcholinesteraseAON anterior octaval nucleusCad anterior commissure, dorsal partCav anterior commissure, ventral partCC crista cerebellarisCCe corpus cerebelliCG central grayCGS central gray stratumCh horizontal commissureChAT choline acetyltransferaseCIL central nucleus of the hypothalamic inferior lobeCM corpus mamillareCON caudal octavolateral nucleusCPN central pretectal nucleusCpos posterior commissureCtec tectal commissureCWS white central stratumDAO dorsal accessory optic nucleusDc central nucleus of the dorsal telencephalic areaDd dorsal nucleus of the dorsal telencephalic areaDH dorsal hornDIL diffuse nucleus of the hypothalamic inferior lobeDl lateral nucleus of the dorsal telencephalic areaDm medial nucleus of the dorsal telencephalic areaDON descending octaval nucleusdot dorsomedial optic tractDp posterior nucleus of the dorsal telencephalic areaDTN dorsal tegmental nucleusECL external cellular layerEG eminentia granularisENV entopeduncular nucleus, ventral partEW Edinger–Westphal nucleusFd dorsal funiculusFld dorsal part of the lateral funiculusFlv ventral part of the lateral funiculusfr retroflexus fascicleFv ventral funiculusGCL ganglion cell layerGL glomerular layerH habenulaHc caudal zone of the hypothalamic periventricular nucleusHd dorsal zone of the hypothalamic periventricular nucleusHv ventral zone of the hypothalamic periventricular nucleusHyp hypothalamusICL internal cellular layerIII oculomotor nerveImRN intermediate reticular nucleusIN isthmic nucleusINL inner nuclear layerIO inferior oliveIPL inner plexiform layerIR inferior rapheIRN inferior reticular nucleusLC locus coeruleusLCa lobus caudalis cerebelliLIX glossopharyngeal lobeLLF lateral longitudinal fascicleLRN lateral reticular nucleusLVII facial lobeLX vagal lobeMA Mauthner cellMaON magnocellular octaval nucleusMFN medial funicular nucleusMLF medial longitudinal fascicleMNV mesencephalic nucleus of the trigeminal nerveMON medial octavolateral nucleusMS marginal stratumNEd dorsal part of the entopeduncular nucleusNEv ventral part of the entopeduncular nucleusNIII oculomotor nucleus

NIn interpeduncular nucleusNIX-Xm glossopharyngeal and vagal motor nucleiNLL nucleus of the lateral lemniscusNLV nucleus lateralis valvulaeNMLF nucleus of the medial longitudinal fascicleNSO nucleus of the spino-occipital nervesNT nucleus taeniaeNVIc caudal part of the abducens nucleusNVIIm facial motor nucleusNVIr rostral part of the abducens nucleusNVmd dorsal part of the trigeminal motor nucleusNVmv ventral part of the trigeminal motor nucleusNXm vagal motor nucleusoc optic chiasmOEN octaval efferent nucleusOS optic stratumOT optic tectumot optic tractP posterior thalamic nucleusPGc central preglomerular nucleusPGl lateral preglomerular nucleusPGm medial preglomerular nucleusPit pituitary � hypophysisPL perilemniscal nucleusPM magnocellular preoptic nucleusPON posterior octaval nucleusPPa anterior part of the parvocellular preoptic nucleusPPp posterior part of the parvocellular preoptic nucleusPPv ventral part of the periventricular pretectal nucleusPSm magnocellular superficial pretectal nucleusPSp parvocellular superficial pretectal nucleusPTN posterior tuberal nucleusPTp periventricular nucleus of the posterior tuberculumPVO paraventricular organPVS periventricular stratumRN red nucleus of SchnitzleinRTN rostral tegmental nucleusSC suprachiasmatic nucleusSCN subcommissural nucleusSE subeminential nucleusSFGS superficial fibrous and gray stratumSG subglomerular nucleusSGN secondary gustatory nucleussgt secondary gustatory tractSO secondary octaval populationSR superior rapheSRN superior reticular nucleusSV subvalvular nucleusT tangential nucleusTel telencephalonTGN tertiary gustatory nucleusTL torus longitudinalisTLa torus lateralisTS torus semicircularisTSc central nucleus of the torus semicircularis,TSvl ventrolateral nucleus of the torus semicircularisVal valvula cerebelli, lateral partVam valvula cerebelli, medial partVd dorsal nucleus of the ventral telencephalic areaVH ventral hornVI abducens nerveVIII octaval nerveVL ventrolateral thalamic nucleusVM ventromedial thalamic nucleusvot ventrolateral optic tractVs supracommissural nucleus of the telencephalic ventral

areaVv ventral nucleus of the telencephalic ventral areaX vagal nerve

76 D. CLEMENTE ET AL.

participates in the neural circuits of control of sleep, learn-ing, and memory (Deutsch, 1971; Sitaram et al., 1978;Hagan and Morris, 1988; Hasselmo and Bower, 1993;Jones, 1993; Levin and Simon, 1998; Perry et al., 1999;Van der Zee and Luiten, 1999). In fish, ACh takes part inthe alteration of the visual responsiveness of optic circuit-ries (Fite and Wang, 1986; King and Schmidt, 1991), inthe processing of gustatory information during feeding(Molist et al., 1993), in the modulation of the telencephaliccircuitry (Perez et al., 2000), or in motor information pro-cessing (Molist et al., 1993).

The distribution of ChAT-containing cells in fish hasbeen described in either specific regions (telencephalon:Villani et al., 1994; diencephalon: Tumosa et al., 1984;Ekstrom and Korf, 1986; Wullimann and Roth, 1992; Vil-lani et al., 1994; mesencephalon: Rhodes et al., 1986;Zottoli et al., 1987, 1988; Molist et al., 1993; myelenceph-alon: Danielson et al., 1988) or in the whole brain (Scyli-orhinus canicula: Anadon et al., 2000; Acipenser baeri:Adrio et al., 2000; Phoxinus phoxinus: Ekstrom, 1987;Porichthys notatus: Brantley and Bass, 1988; Salmo truttafario and Oncorhynchus mykiss: Perez et al., 2000). Incontrast, reports about the distribution of AChE activityin fish brain are only partial (habenulo-interpeduncularsystem: Villani et al., 1987; mesencephalon: Contestabileand Zannoni, 1975; Molist et al., 1993; cerebellum: Con-testabile and Zannoni, 1975; gustatory lobes: Contest-abile, 1975). Nevertheless, extensive parts of the brainsuch as the telencephalon and most of the rhombenceph-alon have not yet been described.

Studies in zebrafish (Danio rerio) have contributed no-tably to our understanding about early nervous systemdevelopment and its genetic regulation in three areas: (1)the differentiation of the first neurons and their associ-ated tracts and commissures, (2) neuromeres and expres-sion of regulatory genes, and (3) generation of mutants bymeans of saturation mutagenesis (Wullimann et al.,1996). Nonetheless, data about the distribution of activesubstances in the zebrafish adult brain are scarce andmainly focused on the catecholaminergic (Ma, 1994a,b,1997) and nitrergic (Holmqvist et al., 2000) systems. Thedevelopment of the zebrafish cholinergic system has beenwidely analyzed by using AChE histochemistry (Hanne-man et al., 1988; Hanneman and Westerfield, 1989; Wil-son et al., 1990; Ross et al., 1992). Notwithstanding, thereare no comprehensive studies on the organization of thecholinergic system in the adult zebrafish brain as the finalstep of the developmental process. In the present report,we describe the distribution of ChAT immunoreactivity inthe adult zebrafish brain, comparing it with AChE activityto shed more light on the neurochemical organization ofthe zebrafish adult brain. We also compare our data withprevious reports on other fish and other vertebrates toprovide additional information on the cholinergic systemfrom a phylogenetic point of view.

MATERIALS AND METHODS

Subjects

Ten 1-year-old adult zebrafish Danio rerio (Hamilton-Buchanan, 1882) of both sexes were obtained from a localsupplier and used for this study. The animals were main-tained in a 12 hours light:12 hours darkness cycle in tanksat a constant temperature of 28°C before killing. All pro-

cedures were in accordance with the guidelines of theEuropean Communities Directive (86/609/EEC) and cur-rent Spanish legislation for the use and care of animals(BOE 67/8509-12, 1998) and conform to NIH guidelines.

Histological processing

The animals were deeply anesthetized with MS-222(Sigma, St. Louis, MO) and decapitated. The heads werefixed in 4% paraformaldehyde and 0.2% picric acid in 0.1M phosphate buffer, pH 7.4 (PB), for 4 hours at 4°C. Afterseveral washes in PB, tissue was cryoprotected in 30%(w/v) sucrose in PB. Coronal and sagittal 20-�m-thicksections of the whole head were obtained with a cryostat(Leica, Nussloch, Germany) and thaw-mounted ongelatin-coated slides in three series: one for Nissl staining,one for ChAT immunohistochemistry, and the last one forAChE histochemistry.

ChAT immunohistochemistry

A pretreatment to inactivate the endogenous peroxi-dases was performed in a solution containing 2% H2O2 in0.1 M phosphate-buffered saline, pH 7.4 (PBS), for 20minutes. Sections were then preincubated for 1 hour in 5%donkey normal serum (Santa Cruz Biotechnology, SantaCruz, CA) and 0.2% Triton X-100 (Probus, Badalona,Spain) diluted in PBS. Immunohistochemistry was per-formed by incubating the sections sequentially in (1) pri-mary goat polyclonal antibody against ChAT (ChemiconInternational, Temecula, CA), 1:250 in PBS containing 5%normal donkey serum and 0.2% Triton X-100 for 3 days at4°C; (2) biotinylated donkey anti-goat IgG (Santa CruzBiotechnology), 1:250 in PBS containing 5% normal don-key serum and 0.2% Triton X-100 for 1 hour at roomtemperature; (3) Vectastain Elite ABC reagent (VectorLabs., Burlingame, CA), 1:250 in PBS containing 0.2%Triton X-100 for 1.5 hours at room temperature. The re-action product was visualized by using 0.025% 3,3�-diaminobenzidine (DAB) and 0.0033% H2O2 in 0.2 M Tris-HCl, pH 7.6. The course of the reaction was monitoredunder the microscope, and the reaction was stopped byrinsing the slides with 0.2 M Tris-HCl, pH 7.6. Controltissue was incubated with normal donkey serum (1:100)instead of the primary antibody, the secondary antibody,or the ABC reagent. No staining was observed in any ofthe controls.

The primary antibody used in this study has been usedbefore in a wide range of vertebrate species, including fish(Shiromani et al., 1987; Medina and Reiner, 1994; Gros-man et al., 1995; Adrio et al., 2000; Anadon et al., 2000;Perez et al., 2000; Pombal et al., 2001). The specificity ofthe antibody has been tested previously by Western blotanalysis of brain protein extracts from rat, dogfish, stur-geon, and trout (Anadon et al., 2000).

AChE histochemistry

AChE histochemistry was carried out following a vari-ation (Hedreen et al., 1985) of the Koelle method(Geneser-Jensen and Blackstad, 1971). After washing in0.1 M acetate buffer, pH 6.0, sections were incubated atroom temperature for 30 minutes in an incubating me-dium made up of 65 mM acetate buffer, pH 6.0, 1.7 mMacetylthiocholine iodide (Sigma), 5 mM sodium citrate(Merck, Darmstadt, Germany), 3 mM copper sulfate (Pro-labo, Fontenay-St. Bois, France), 0.2 mM potassium ferri-

77ZEBRAFISH CHOLINERGIC SYSTEM

cyanide (Panreac, Madrid, Spain), and 0.2 mM ethoprop-azine (Sigma) as an inhibitor of nonspecific esterases.

AChE activity was visualized enhancing the intensity ofthe precipitate with 0.0125% DAB (Sigma) and 0.0033%H2O2 in 0.2 M Tris-HCl buffer, pH 7.6. The reaction wasmonitored under the microscope and stopped by washingthe sections in the same Tris-HCl buffer. The sectionswere dehydrated in an ethanol series, cleared with xyleneand cover-slipped with Entellan (Merck). Omission of thesubstrate acetylthiocholine iodide was carried out as con-trol for the AChE histochemistry, and no residual activitywas observed.

Analysis

Sections were analyzed with a Leica microscopeequipped with brightfield condensers. Brightfield digitalimages were obtained with an Apogee KX digital camera(Apogee Instruments, Inc., Tucson, AZ) coupled to anOlympus Provis AX70 photomicroscope. The original im-ages were processed digitally with Adobe Photoshop 7.0(San Jose, CA) software. The capture software was con-nected to a trichromatic sequential filter (Cambridge Re-search & Instrumentation, Inc., Boston, MA). After con-version into black and white, the sharpness, contrast, andbrightness were adjusted to reflect the appearance seenthrough the microscope. Schematic drawings at differentrepresentative coronal section levels were made by using acamera lucida and CANVAS version 7.0.2 software(Deneba Software, Miami, FL). A different number of pos-itive cells belonging to each positive nuclei were selected,and their maximum diameters were measured by using adigitizer tablet connected to a semiautomatic image anal-ysis system (MOP-Videoplan, Kontron, Munich). Whenthe labeled cells were scarce, all stained cells were mea-sured.

RESULTS

Our description of the zebrafish CNS is based on thetopographical atlas for the zebrafish brain of Wullimannet al. (1996). ChAT immunoreactivity and AChE activitywere only observed in neuronal elements and no glial cellswere positive. After ChAT immunohistochemistry andAChE histochemistry, positive neurons showed differentstaining intensities. Strongly stained neuropile was de-tected in specific nerve bundles and nuclei. AChE-positiveneurons were distributed throughout all brain subdivi-sions, although they were especially abundant in the tel-encephalon, mesencephalon, and medulla oblongata. Incontrast, the distribution of ChAT immunoreactive(ChAT-ir) cells was restricted to specific nuclei or regionsof the brain: preoptic region, optic tectum, tegmentum,and diverse nuclei within the rhombencephalon (Fig. 1).Our findings are described from the rostralmost to thecaudalmost regions of the CNS, beginning with the retinaand ending with the spinal cord.

Retina

The patterns of staining of ChAT and AChE in both theinner nuclear layer and the inner plexiform layer of theretina were similar. Scattered, round, small neurons(7.59 � 0.17 �m; n � 69) were observed positive to bothAChE and ChAT in the vitrealmost part of the innernuclear layer (Fig. 2a,b), and two intensely positive neu-ropile bands were also observed positive to both markers

in the inner plexiform layer, one close to the inner nuclearlayer and another in the boundary with the ganglion celllayer. In addition, some round stained neurons (7.68 �0.54 �m; n � 69) in the ganglion cell layer were observedto be ChAT-ir but they were AChE-negative (Fig. 2a,b).These cells were less numerous than those in the innernuclear layer.

Telencephalon

The zebrafish telencephalon is composed of paired olfac-tory bulbs and the adjacent telencephalic hemispheres. Inboth regions, AChE-positive cells were detected (Fig. 1a–d). No ChAT-ir cells were observed in the telencephalon,but ChAT-ir fibers could be observed arriving at the tel-encephalic hemispheres.

No ChAT immunoreactivity was observed in the ze-brafish olfactory bulb. AChE-positive small neurons(9.57 � 0.25 �m; n � 75), usually aggregated in groups oftwo or three cells, appeared within the external cellularlayer of the olfactory bulb (Fig. 3a). On some occasions, theinitial portion of one dendrite was AChE-stained. Theirlocation, distribution, and morphology led us to identifythem as mitral cells.

Although ChAT-ir cell somata were absent in the ze-brafish telencephalic hemispheres, some ChAT-ir plex-uses were observed in the caudalmost part of the dorsalarea within the medial, dorsal, and central nuclei (Fig.3b). These fibers ran toward the dorsal telencephalon,passing through the dorsal and ventral nucleus of theventral area of the telencephalic hemispheres where theydid not branch. Scattered ChAT-ir fibers were observedarriving at the entopeduncular nucleus, both in the dorsaland in the ventral parts. In all the regions where ChAT-irfibers were present, some AChE-positive neurons weredetected. AChE-positive small neurons were observed inboth the dorsal area and the ventral area of the telence-phalic hemispheres. The density of cells in the dorsal areawas higher than in the ventral area (Fig. 1b–d). In theformer, AChE-positive cells were present in the dorsal,medial, central, posterior and lateral nuclei (Fig. 3c).These neurons were always small in size (8.20 � 0.15 �m;n � 90; 9.62 � 0.20 �m; n � 40 for the dorsocentralnucleus), with round or ovoid somata (Fig. 3d). The high-est density of positive neurons was observed in the dorsaland lateral nuclei. Some of the stained neurons within thecentral nucleus showed the proximal portion of one pro-cess, oriented to the lateral part of the section, as AChE-positive (Fig. 3c).

In the ventral area, AChE-positive neurons were simi-lar in shape and size to those of the dorsal area. They werelocated in the central nucleus and in the dorsal and ven-tral parts of the entopeduncular nucleus (Fig. 3e).

Diencephalon

In a dorsoventral arrangement, the zebrafish dienceph-alon is composed of the epithalamus, dorsal thalamus,ventral thalamus, posterior tuberculum, and hypothala-mus. The preoptic region, the synencephalon and the pre-tectum, regions that are closely related with the dien-cephalons, will also be described in this section. ChAT-and AChE-positive cells were scarce in this region (Fig.1c–j; Table 1). Whereas ChAT-ir neurons were exclusivelydetected within the preoptic region, AChE-positive oneswere also observed in the ventral thalamus, posterior tu-berculum, pretectum, and synencephalon. However, in


some of these regions a strong ChAT-ir innervation couldbe detected.

In the preoptic region, two nuclei showed ChAT-ir cells.The first was the anterior part of the parvocellular preop-tic nucleus. This nucleus showed some isolated, lightlystained cells (7.19 � 0.75 �m; n � 42) close to the ventriclealong the dorsoventral extension of the nucleus. Some ofthem presented one branched process oriented toward themedial and dorsal parts of the brain. The second, themagnocellular preoptic nucleus (Fig. 4a), showed someChAT-ir neurons (12.32 � 0.33; n � 45), whose axonsinnervated the neurohypophysis. Furthermore, in the pre-optic region, a profuse innervation in the ventralmost partof the parvocellular preoptic nucleus was observed. AChE-positive cells, with the same characteristics as theChAT-ir cells, were detected in the magnocellular preopticnucleus and in the anterior part of the parvocellular pre-optic nucleus. The latter showed another AChE-positiveneuronal population. This population was formed by onegroup located in the ventralmost part of the nucleus pre-senting intensely stained somata (8.52 � 0.19 �m; n � 50)and the beginning of one long, thin, dorsally orientedprocess (Fig. 4b). In addition, in the posterior part of theparvocellular preoptic nucleus and in the suprachiasmaticnucleus, a few small (8.22 � 0.75 �m; n � 46), round cellswith a moderate staining intensity were detected.

The thalamus and the posterior tuberculum did notpresent any ChAT-ir neurons. Nevertheless, in the ventralthalamus, AChE-positive cells were observed in the ven-tromedial and ventrolateral nuclei. They showed mediumsomata (10.53 � 0.24 �m; n � 40) and were located in theventrolateral part of both nuclei without displayingstained processes (Fig. 4c). Within the posterior tubercu-lum, the periventricular nucleus showed two subpopula-tions of strongly AChE-positive neurons. One of them wasformed by small (8.43 � 0.35 �m; n � 65), round, abun-dant cells, whereas the second was composed of medium(11.52 � 0.65 �m; n � 30), pyriform, scarcely stainedneurons. Cells of the latter group exhibited the initialportion of one laterally oriented process emerging fromthe apical part of the soma (Fig. 4d).

Some ChAT-ir fibers were observed arriving at the cau-dalmost part of the nucleus of the medial longitudinalfascicle within the synencephalon, where they impingedChAT-immunonegative cells (Fig. 4e). In this nucleus,large AChE-positive neurons (16.67 � 0.52 �m; n � 40)with medium labeling intensity in the somata and in oneor two dendrites were detected (Fig. 4f). In some of theseneurons, both the soma and the processes were impingedby AChE-positive preterminal puncta on their surfaces(Fig. 6f). The location and morphology of the cells withChAT-ir buttons were similar to those of the AChE-positive neurons with stained preterminal puncta on theirsurfaces.

A high number of ChAT-ir fibers were detected aroundthe parvocellular superficial, magnocellular superficial,central, posterior, and accessory pretectal nuclei and alsoin the parvocellular superficial pretectal nucleus (Fig. 5a).In all these nuclei, intense AChE-positive neuropile wasobserved (Fig. 5b). This neuropile appeared more stronglypositive in both parts of the superficial pretectal nucleusthan in the rest of the aforementioned nuclei, where itshowed medium staining intensity. AChE-positive me-dium cells (11.42 � 0.29 �m; n � 82) with round somatawere observed in the pretectum within the magnocellular

superficial, central, posterior, and accessory pretectal nu-clei and in the dorsal accessory optic nucleus. The cellslocated in the dorsal part of the magnocellular superficialpretectal nucleus (Fig. 5b) showed strong labeling,whereas the rest of the positive cells within the pretectumappeared weakly positive.

In the hypothalamus, a dense network of stronglystained ChAT-ir fibers were detected (Fig. 5c) in somenuclei, such as in the dorsal and ventral zones of theperiventricular hypothalamus, the anterior tuberal nu-cleus, the lateral hypothalamic nucleus, and in the centraland the diffuse nuclei of the inferior lobe. ChAT-ir fibersinnervated the neurohypophysis. In contrast, AChE-positive neurons and fibers were not present in the afore-mentioned hypothalamic nuclei (Fig. 5d).

Mesencephalon

The teleostean mesencephalon includes the optic tec-tum, a cupuliform structure with a layered organizationdorsally, and the torus semicircularis and the mesence-phalic tegmentum ventrally. ChAT-ir and AChE-positivecells were observed within the optic tectum and the teg-mentum, whereas the torus semicircularis presentedAChE-positive cell bodies but no ChAT-ir ones (Fig. 1e–l).

Optic tectum. Within the optic tectum, ChAT-ir cellbodies appeared exclusively in the periventricular stra-tum (Fig. 6a). The positive cells were located in the centralportion of the stratum and showed small (6.43 � 0.27 �m;n � 36), round, or pyriform somata and one apical den-drite oriented perpendicular to the lamination. From thisdendrite, the axon that ramified mainly in the boundarybetween the superficial fibrous and gray and central graystrata emerged. This thick dendrite ramified in the fourdense neuropile bands within the optic, superficial fibrousand gray, central gray, and central white strata, respec-tively. All of them were narrow and intensely stainedexcept for that in the superficial fibrous and gray stratum,which was thicker and more weakly positive. In addition,some ChAT-ir fibers were observed within the tectobulbartract.

AChE-positive elements showed a more widespread dis-tribution in the optic tectum. Thus, some scattered andisolated AChE-positive cells were observed in the optic,superficial fibrous and gray, central gray, and centralwhite strata (Fig. 6b). These cells were medium in size(11.09 � 0.15 �m; n � 87) and weakly stained. The shapeof their somata was ovoid with a predominant verticalorientation, although some horizontal cells were observedin the central white stratum. In the periventricular stra-tum, a subpopulation of small (6.58 � 0.34 �m; n � 36),round cells located in the central portion of the stratumwas AChE-positive. In addition, four darkly positivebands of neuropile were detected in the optic tectum. Eachband was located within one of the following strata: optic,superficial fibrous and gray, central gray, and centralwhite. The patterns of location, thickness and stainingintensity of AChE-positive neuropile bands were similarto those observed for the ChAT-immunoreactivity.

Torus semicircularis. Although no ChAT-ir cell bod-ies were detected within the torus semicircularis, somepositive fibers were observed within the ventrolateral nu-cleus (Fig. 1h). These fibers branched profusely withinthis area before they ran to the optic tectum. On the otherhand, scattered, small (8.87 � 0.54 �m; n � 40), roundcells with strong AChE labeling were observed in the


ventrolateral nucleus of the torus semicircularis, notshowing any specific disposition or location within thenucleus.

Tegmentum. In the tegmentum, ChAT-ir cells weredistributed in the dorsal part of the rostral tegmentalnucleus (Fig. 6c). These medium neurons (10.46 � 0.43

�m; n � 40) showed the positive axon that arrives at thenucleus of the medial longitudinal fascicle impinging onsome ChAT-immunonegative cell bodies. Another sub-population in this nucleus displayed the axon ventrallyoriented, branching into the hypothalamus. ChAT-ir cellswere also observed in the oculomotor nucleus (Fig. 7a) and

Fig. 1. a–q: Schematic drawings of transverse sections throughthe brain of an adult zebrafish. The complete drawings show, on theleft side, the cytoarchitecture of the zebrafish brain and, on the right,the distribution of AChE-positive (solid circles) neurons. In the halfdrawing, ChAT-immunoreactive cells (stars) and processes are repre-

sented. The section level is shown in the schematic representation ofthe zebrafish brain in the upper right corner of each plate. For abbre-viations, see list. Scale bars � 500 �m in c (applies to a–c), f (appliesto d–f), i (applies to g–i), l (applies to j–l), n (applies to m,n), q (appliesto o–q).


Figure 1 (Continued)

in the trochlear nucleus. The immunostained neurons ofthe oculomotor nucleus showed round or ovoid somata(14.42 � 0.36 �m; n � 45) and possessed one intenselystained axon that left the brain ventrally, joining theoculomotor nerve. The stained axons of the round neurons(14.65 � 0.54 �m; n � 40) of the trochlear nucleus wereoriented dorsally, coursed toward the trochlear decussa-tion, and left the brain dorsally through the trochlearnerve. ChAT-ir fibers were detected in some regions where

ChAT-ir somata were not observed. Thus, ChAT-ir fiberswere observed in the Edinger–Westphal nucleus, the dor-sal tegmental nucleus, the red nucleus of Schnitzlein, andaround the lateral longitudinal fascicle, within the nu-cleus of the lateral lemniscus and within the perilemniscalnucleus (PL).

In the zebrafish tegmentum, the density of AChE-stained cells was very high (Table 1), and their neuronsshowed different sizes (from 8 to 14 �m). In the dorsal and



rostral tegmental nuclei, some AChE-positive cells(10.78 � 0.51 �m; n � 30) were observed which were veryabundant in the latter (Fig. 6d). These cells were detectedin the whole extension of these nuclei, displaying fusiformor round somata and the beginning of one or two positiveprocesses. The red nucleus of Schnitzlein presented onegroup of intensely AChE-stained neurons. These neuronsshowed pyriform somata (13.49 � 0.81 �m; n � 20) withone or two thick stained processes ventrally oriented (Fig.6d). The mesencephalic nucleus of the trigeminal nervedisplayed AChE-positive cells (11.21 � 0.78 �m; n � 30)with medium to weak labeling intensity (Fig. 6e). Thesecells were located forming one group separated from theventricle and close to the posterior commissure, dorsolat-erally to the subcommissural nucleus. AChE staining also

occurred in the Edinger–Westphal nucleus, where positivecells showed strong labeling intensity. These cells hadmedium somata (11.34 � 0.65 �m; n � 30;Fig. 6f) andrarely showed the beginning of one process oriented to-ward the ventral part of the section. The nucleus of thelateral lemniscus and the PL showed intensely AChE-stained cells with small (8.44 � 0.78 �m; n � 56), roundsomata surrounding the lateral longitudinal fascicle.Thus, neurons from the nucleus of the lateral lemniscusappeared forming one group dorsally to the fascicle andneurons belonging to the PL were located laterally to itwithout showing any specific arrangement.

AChE-positive cells were observed in the whole rostro-caudal extension of the oculomotor nucleus (Fig. 7b).These cells showed intense AChE-labeling and medium,



round or ovoid somata (14.56 � 0.54 �m; n � 45). Caudalto this nucleus, AChE-positive cells were located in thetrochlear nucleus, displaying round or pyriform morphol-ogies (14.76 � 0.76 �m; n � 40), and one ventrally ori-ented process. Finally, one subpopulation of AChE-positive cells was observed in the dorsalmost part of theinterpeduncular nucleus. These cells exhibited mediumlabeling intensity and small, round somata (7.56 � 0.13�m; n � 54).

Rhombencephalon

The rhombencephalon was the area with the highestdensity of AChE activity (Table 1). AChE-positive cellswere abundant throughout the isthmus, octavolateralarea, reticular nuclei, and motor nuclei of the cranialnerves (Fig. 1g–p). ChAT immunoreactivity was restrictedto the isthmic region, the motor nuclei of the cranialnerves, the octaval efferent nucleus, the descending octa-

val nucleus, and the nucleus of the spino-occipital nerves(Figs. 1j–p, 8a).

Isthmus. There are some nuclei in this region thatdisplayed AChE and ChAT reactivity. Some ChAT-irsmall round cells (7.57 � 0.54 �m; n � 70) were detectedin the secondary gustatory nucleus (Fig. 8b). ChAT-iraxons emerged from neurons of this nucleus and arrivedat the hypothalamus where they branched profusely.Other processes were observed forming a thin fascicleoriented to the middle of the section. ChAT-ir mediumneurons (11.34 � 0.98 �m; n � 35) and positive neuro-pile were observed in the lateral and ventral boundariesof the isthmic nucleus. Some medium neurons (12.33 �0.26 �m; n � 50) appeared ChAT-ir in the rostral part ofthe superior reticular nucleus, close to the isthmic nu-cleus and the secondary gustatory nucleus. These cellsshowed ovoid or pyriform somata and two thick stainedprocesses ventrally oriented toward caudal areas of thebrain. Furthermore, some granule cells presented ChATimmunoreactivity in the nucleus lateralis valvulae(7.43 � 0.39 �m; n � 70).

All the isthmic nuclei that displayed ChAT-ir neuronsalso showed AChE-positive cells. In all cases, the cellsshowed the same location and morphology as thoseChAT-ir, but they were more abundant (Fig. 8c). Theisthmic nucleus presented a strongly AChE-positiveneuropile and round medium neurons (11.68 � 0.87 �m;n � 35) located in the external boundaries of the nu-cleus, mainly in the ventral and lateral parts (Fig. 8d).AChE-positive small round cells were observed in thesecondary gustatory nucleus (7.31 � 0.56 �m; n � 70)and in the nucleus lateralis valvulae (7.55 � 0.92 �m;n � 70). Stained neurons of the superior reticular nu-cleus (12.56 � 0.76 �m; n � 50) were intensely AChE-positive in the somata and in the proximal portions oftwo or three dendritic trunks (Fig. 8e).

Cerebellum. The zebrafish cerebellum has threeparts: the valvula cerebelli, which is divided into medialand lateral subdivisions, the corpus cerebelli, and thelobus vestibulolateralis, which is formed by the lobus cau-dalis and the eminentia granularis. No ChAT-ir elementsappeared in the zebrafish cerebellum (Fig. 9a). Neverthe-less, AChE-positive neurons were detected in the threesubdivisions. All Purkinje cells were positive to AChEhistochemistry in the dorsal and ventral portions of thevalvula cerebelli, medial part; in the corpus cerebelli andin the lobus caudalis (Fig. 1i–n). These cells presentedovoid or pyriform somata (10.03 � 0.15 �m; n � 57),although stained processes were not observed (Fig. 9b,c).Nonetheless, in each region the distribution of positivePurkinje cells was not uniform. Thus, stained Purkinjecells were more abundant in the ventral than in the dorsalpart of the valvula cerebelli. In the corpus cerebelli, thedensity of AChE-positive Purkinje cells was higher in therostral and caudal parts and in the lateral regions than inthe medial and dorsal ones, respectively. With regard tothe lobus caudalis, AChE-positive cells were more abun-dant in the lateral and dorsal parts than in the centralportion. In addition to the Purkinje cells, a subpopulationof granule cells appeared AChE-positive in the three cer-ebellum subdivisions.

Medulla oblongata. ChAT immunoreactivity was ex-clusively observed in the motor nuclei of the cranialnerves, the octaval efferent nucleus, the descending octa-val nucleus, and the nucleus of the spino-occipital nerves.

Fig. 2. ChAT and AChE distribution in the zebrafish retina (coro-nal sections). a: The zebrafish retina presents ChAT-immunoreactivecells (arrowheads) in the inner nuclear layer (INL) and ganglion celllayer (GCL). Two bands of ChAT-immunoreactive neuropile formedby the amacrine cells are observed in the inner plexiform layer (IPL).b: Neurons located in the vitrealmost part of the INL (arrowheads)and two intense bands in the IPL of the retina are positive afterAChE-histochemistry. For other abbreviations, see list. Scale bar �25 �m in b (applies to a,b).


All regions that displayed ChAT-ir neurons also showedAChE-positive cell bodies. Nevertheless, there are othernuclei that only showed AChE-positive neurons: the restof the nuclei within the octavolateral area, raphe andreticular nuclei, Mauthner cells.

Trigeminal, abducens, and facial motor nuclei. Largeand intensely ChAT-stained neurons appeared in the mo-tor nuclei of the cranial nerves. Both dorsal (located dor-sally to the medial longitudinal fascicle) and ventral (ven-trolaterally situated to this fascicle) parts of thetrigeminal motor nucleus presented ChAT-ir large pyri-form cells (16.52 � 1.05 �m; n � 30) with thick processes(Fig. 10a). The abducens nucleus showed ChAT-ir largeneurons (16.76 � 0.88 �m; n � 30) in both the rostral andcaudal parts (Fig. 10c), presenting some stained processesthat ramified in the superior and intermediate reticularnuclei. The ChAT-ir axon of the motoneurons was ob-served leaving the brain along the abducens nerve (Fig.10c). The facial motor nucleus showed some ChAT-ir pyri-form neurons (16.56 � 0.77 �m; n � 30) with one ventro-laterally oriented dendrite that ramified in the region ofthe intermediate reticular nucleus.

The rostralmost portions of both the dorsal and ventralparts of the trigeminal motor nucleus showed AChE-positive neurons (Fig. 11b). These AChE-stained neurons

were large (16.66 � 1.0 �m; n � 30), with round or pyri-form somata and some of them had stained processes.Some neurons of the rostral and caudal parts of the abdu-cens nucleus (Fig. 10d) and of the facial motor nucleuswere positive to AChE histochemistry. These cells showedlarge (abducens: 16.22 � 0.81 �m, n � 30; facial: 16.71 �0.87, n � 30) round or pyriform somata (Fig. 10d) and thebeginning of one stained process ventrally and ventrolat-erally oriented, respectively.

Reticular formation. The reticular formation did notpresent any stained neuron in either the reticular nuclei,the raphe nuclei, or the Mauthner cells. In contrast,AChE-positive cells were present. AChE-positive neuronswhich appeared in the superior raphe, close to the midlineof the brain, were round or ovoid, medium in size (11.76 �0.84 �m; n � 30), and intensely positive (Fig. 8c). Theovoid cells in this nucleus showed a predominant verticalorientation. The two rhombencephalic subdivisions (inter-mediate and inferior) of the reticular nucleus presentedAChE-positive neurons (Fig. 10d). In both nuclei, AChE-positive cells showed two sizes: medium (11.33 � 0.26 �m;n � 100) and very large (20.33 � 1.07 �m; n � 30). Thecells within these two nuclei were less intensely stainedthan those observed in the superior part of the reticularnucleus. AChE-positive neurons of the former two nuclei

Fig. 3. ChAT immunoreactivityand AChE activity in the telencepha-lon (coronal sections). a: Section of theolfactory bulb showing AChE-positivemitral cells in the external cellularlayer (ECL). b: ChAT-immunoreactivefibers arriving at the medial and cen-tral nuclei of the dorsal telencephalicarea. c,d: AChE-positive elements inthe telencephalic hemispheres. AChE-positive neurons are mainly locatedin the dorsal telencephalic area.Cells of the central nucleus of thedorsal area (inset area) showed onestained process ventrolaterally ori-ented. e: AChE-positive neurons inthe dorsal (NEd) and ventral (NEv)parts of the entopeduncular nucleus.For other abbreviations, see list. Scalebar � 25 �m in e (applies to b, d, e); 50�m for a; 100 �m for C; 25 �m forinset.


showed positive terminal puncta over the somata and thestained processes. In the lateral reticular nucleus, somemedium neurons (11.66 � 0.65 �m; n � 40) with round orovoid somata showed intense AChE staining (Fig. 11f).

The somata and the initial portion of both the lateraland ventral dendrites of the Mauthner cells were weaklystained after AChE histochemistry (Fig. 11c,d). AChE-positive terminals impinging on them were clearly ob-served (Fig. 11c).

Octavolateral area. ChAT-immunopositive cell bodies(10.55 � 0.76 �m; n � 30) were observed in the caudal-most part of the descending octaval nucleus, ventrally tothe caudal octavolateral nucleus of the octavolateral area.A dense ChAT-ir innervation was detected in the wholeextension of the medial and caudal octavolateral nuclei, aswell as in the descending and magnocellular octaval nu-clei. However, the most conspicuous ChAT-ir neuronswithin the octavolateral area were observed in the octavalefferent nucleus showing large pyriform somata (16.42 �0.62 �m; n � 42) with their stained dendrites ventrolat-erally directed (Fig. 11a).

The octavolateral area showed one of the highest den-sities of AChE-positive neurons within the zebrafish CNS(Table 1). The octaval efferent nucleus showed numerousand intensely AChE-stained pyriform neurons (16.76 �0.32; n � 42). These cells showed one ventrally or ventro-laterally oriented long dendrite (Fig. 10b). Numerous,weakly AChE-positive neurons of small size (9.49 � 0.51�m; n � 75) were observed in the medial octavolateralnucleus (Fig. 11d,e). They did not present positive pro-cesses, and their distribution was very homogeneous inthe whole rostrocaudal and mediolateral extension of thenucleus. In the secondary octaval population, some AChE-stained neurons (10.53 � 0.24 �m; n � 48) were detected

close to the fourth ventricle, preferentially located in therostral and medial parts of this region. A group of AChE-positive cells (9.36 � 0.37 �m; n � 90) were observed inthe central portion of the anterior octaval nucleus (Fig.11c,d). Some of these cells showed one afferent stainedprocess forming part of the octaval nerve. The root of theoctaval nerve divided the magnocellular octaval nucleusinto two regions. Large positive neurons (15.79 � 0.86 �m;n � 30) were observed in both parts (Fig. 11e). Some of thepositive neurons of the dorsal group presented AChE-positive buttons impinging on them. The cells of the ven-tral part were smaller (10.76 � 0.65 �m; n � 50) and lessintensely stained than those of the dorsal region. Someneurons from both groups presented one or two stainedprocesses, some of them parallel to the axons of the octavalnerve. The descending octaval nucleus showed an exten-sive labeling throughout its whole rostrocaudal extension.Intensely stained cells (10.11 � 0.76 �m; n � 90) wereespecially abundant in the lateralmost part of the nucleusand close to the ventricle. In the posterior octaval nucleus,scarce AChE-positive rounded and medium neurons(11.86 � 0.84 �m; n � 75) were detected around the rootof the vagal nerve (Fig. 11f). The caudal octavolateralnucleus showed abundant and intensely AChE-positiveneurons. The distribution of the AChE-positive neuronswas homogeneous in the whole rostrocaudal and medio-lateral extension of this nucleus. These AChE-positiveneurons displayed round or ovoid medium (10.36 � 0.36�m; n � 90) somata, and they did not display stainedprocesses.

Caudal rhombencephalon. Caudal to the octaval effer-ent nucleus, motoneurons belonging to the visceromotorcolumn (16.99 � 0.38 �m; n � 50), composed of the glos-sopharyngeal and vagal motor nuclei, appeared ChAT im-munoreactive (Fig. 12a). These cells had two stained pro-cesses. The laterally oriented axon could be followedleaving the brain throughout the IX and X nerves. Inaddition, one ventrally oriented dendrite ramified withinthe intermediate reticular nucleus, the inferior reticularnucleus or the nucleus of the spino-occipital nerves (Fig.12a).

In the medial caudal part of the medulla oblongata,numerous strongly ChAT-ir cells were observed within thenucleus of the spino-occipital nerves (Fig. 11a). This groupof cells partially overlaps with the caudal part of the vagalmotor nucleus and extends down to the rostral portion ofthe spinal cord (Fig. 12a). These cells (14.34 � 0.34 �m;n � 45) showed round or pyriform somata and two pro-cesses with different orientations. The dorsal one coursedtoward more medial or dorsal areas, whereas the ventralone branched profusely in the ventral part of the caudalmedulla oblongata.

Some AChE-positive neurons appeared in the viscero-motor column. The AChE-positive neurons were large(17.10 � 0.32 �m; n � 50), strongly stained, and presentedone stained process ventrolaterally oriented toward theroot of the vagal nerve. They were mainly located in themedial part of the column, close to the fourth ventricle(Fig. 12b). Motoneurons belonging to the nucleus of thespino-occipital nerves appeared AChE-positive. Theseneurons showed round or pyriform somata (14.67 � 0.13�m; n � 45), but their processes did not present AChEreactivity.

TABLE 1. Distribution of Positive Cells1

CNS region AChE ChAT

Retina � ��Telencephalon

Olfactory bulb � �Dorsal telencephalic area �� Ventral telencephalic area � �

DiencephalonPreoptic area � �Ventral thalamus �� Posterior tuberculus � �Synencephalon (NMLF) �� Pretectum ��

MesencephalonOptic tectum �� Torus semicircularis �� Tegmentum ��

RhombencephalonCerebellum (Purkinje cells)Valvula cerebelli �� Corpus cerebelli �� Lobus vestibulolateralis �� Medulla oblongataOctavolateral area �� Reticular formation

Nucleus raphes superior � �Nuclei reticularis

Superior �� Intermediateis �� Inferior �� Lateralis ��

Spinal cordVentral horn ��

1Relative density of AChE-positive and ChAT-immunoreactive cells in the main sub-divisions of the zebrafish CNS. Density has arbitrarily been assigned to values asfollows: � � low abundance; �� moderate abundance; �� high abundance;�� very high abundance; � � absence. CNS, central nervous system. For otherabbreviations, see list.


Spinal cord

The spinal cord presented both ChAT-ir and AChE-positive neurons (Fig. 1q). Large ChAT-ir neurons (20.04 �1.34 �m; n � 30) were observed in the ventral horn of thespinal cord (Fig. 12c). These cells showed weak labeling inthe somata and in the initial portion of thick dendritictrunks emerging from the cell body. Moreover, another pop-ulation of ChAT-ir cells was located close to the midline ofthe spinal cord, ventral to the central canal and presentedstronger intensity of labeling and smaller cell bodies

(14.64 � 0.52 �m; n � 30) than cells from the former popu-lation. These cells showed one or two stained and brancheddendrites ventrolaterally oriented. In the rostralmost part ofthe spinal cord, the ChAT-ir population overlapped with themotoneurons of the nucleus of the spino-occipital nerves. TheAChE-positive motoneurons were located in the same posi-tion and showed the same morphology as the large ChAT-ircells described above (Fig. 12d). Nevertheless, they showed astrong intensity of labeling in their large somata (19.89 �0.89 �m; n � 30).

Fig. 4. ChAT- and AChE-positive ele-ments in the zebrafish diencephalon (trans-verse sections). a: ChAT-immunoreactiveneurons in the magnocellular preoptic nu-cleus. b: Strongly AChE-positive neurons inthe anterior part of the parvocellular preopticnucleus (PPa), showing stained processes ori-ented toward the top of the section (arrows).c: AChE-positive cells in the lateralmost partof the ventromedial (VM) and ventrolateral(VL) thalamic nuclei. Medial is to the right.d: Section at the level of the posterior tuber-culum showing some AChE-stained cells inthe periventricular nucleus of the posteriortuberculum (PTp). In this nucleus, in addi-tion to the small round stained neurons, largepyriform neurons can be observed with thebeginning of one stained process oriented lat-erally. e: The nucleus of the medial longitu-dinal fascicle presented some neurons im-pinged on by ChAT-immunoreactiveterminals (arrows). f: Photomicrographshowing a section of the AChE-stained neu-rons within the nucleus of the medial longi-tudinal fascicle. For other abbreviations, seelist. Scale bar � 25 �m in f (applies to a,e,f);50 �m for b,d; 40 �m for c.


DISCUSSION

Methodological considerations

In this study, we carried out a variation (Hedreen et al.,1985) of the Koelle method (Geneser-Jensen and Blacks-tad, 1971) for the histochemical demonstration of AChE.The specificity of AChE histochemistry in the zebrafishbrain is supported by two facts: (1) no staining was ob-served in control sections where the substrate was omittedin the histochemical procedure, and (2) there is a gradienton the intensity of AChE labeling among nuclei from lightto intense. This last point suggests that the method issensitive to varying levels of enzyme activity within pop-ulations of cells in the same animal. The difference in thelabeling was not as evident in the case of ChAT immuno-histochemistry. All ChAT-containing cells showed strongintensity of staining. It was more difficult to assign a

greater or lesser quantity of enzyme among ChAT-ir cells,with the exception of the spinal cord, where stained cellsshowed two different morphologies with different intensi-ties of ChAT immunoreactivity.

Because of its high sensitivity, AChE histochemistry isa valuable marker to locate cholinergic as well as cholino-ceptive neurons in the CNS (Butcher, 1983; Bradford,1986), because AChE occurs in cholinergic and also incholinoceptive neurons (Silver, 1974). The ganglion celllayer of the retina was the only region where ChAT-ir cellbodies were observed in the absence of AChE stainedneurons. However, there were areas lacking ChAT-ir neu-rons that presented AChE-positive somata (e.g., telen-cephalon, some strata of the optic tectum, some nucleiwithin the mesencephalic tegmentum, the octavolateralarea, and the cerebellum). Most of these areas showed

Fig. 5. Transverse sections of the ze-brafish pretectum and hypothalamus show-ing the ChAT/AChE distribution. a: The su-p e r fi c i a l p r e t e c t u m s h o w s C h A T -i m m u n o r e a c t i v e fi b e r s w i t h i n t h eparvocellular superficial pretectal nucleus(PSp) and also around the rest of the nuclei ofthis area. Lateral is to the left. b: AlthoughAChE-positive neurons were exclusively ob-served around the magnocellular superficialpretectal nucleus (PSm) and the dorsal acces-sory optic nucleus (DAO), a strong neuropilewas detected in all the nuclei within thisregion. Lateral is to the left. c,d: The hypo-thalamus showed a profuse ChAT-immunoreactive innervation (c) in the dorsalzone of the periventricular nucleus (Hd) andin the diffuse nucleus of the inferior lobe(DIL), whereas AChE-positive elements wereabsent from this diencephalic area (d). Forother abbreviations, see list. Scale bar � 100�m in d (applies to c,d); 60 �m for a,b.


intense ChAT-ir innervation. In contrast, in the cerebel-lum and olfactory bulb, neither somata nor fibers wereChAT immunoreactive, although some AChE-positivecells were detected. Moreover, there were regions, such asthe hypothalamus, with a rich cholinergic innervationthat did not display AChE-positive elements. A previousstudy of the distribution of AChE and ChAT in the CNS ofthe cyprinid Phoxinus demonstrated that (1) AChE-positive cells were more numerous than ChAT-immunopositive cells; (2) some of the regions that showedChAT-immunopositive neurons also presented AChE-positive cells; (3) there were some regions that displayedChAT-ir cells and did not show AChE-positive elements(Ekstrom, 1987). The three possibilities were observed inthe zebrafish brain. To compare our data with previousreports on other fish and other vertebrates regarding thecholinergic system, we follow the same scheme as in theresults, beginning with the retina and ending with thespinal cord.

Retina

AChE-positive cells were observed in the vitrealmostpart of the inner nuclear layer and in two bands of positiveneuropile in the inner plexiform layer of the zebrafish.ChAT immunoreactivity was detected in the same locationas AChE reactivity and also in cells of the ganglion celllayer. Our observations in the zebrafish retina are inagreement with previous data about the distribution ofChAT-ir cells in the goldfish retina (Tumosa et al., 1984).Posterior studies have demonstrated the noncholinergicnature of the ganglion cells (Zottoli et al., 1985; Ross andGodfrey, 1986; Tumosa and Stell, 1986). Moreover, we didnot observe ChAT-ir fibers along the optic nerve. There-fore, the innervation observed in the retinorecipient strataof the optic tectum should have a different source (seeabove). Both types of ChAT-ir cells in the retina could beamacrine cells with stained processes along the innerplexiform layer. The distribution of ChAT-ir cells in theretina is conserved among vertebrates (Schmidt et al.,1987; Mitrofanis and Stone, 1988; Rodieck and Marshak,1992; Deng et al., 2001; Zhang and Wu, 2001), but thereare differences in the number, size, and shape of the cells.ChAT-ir amacrine cells (so-called “starburst” cells) ofmammals might be involved in the origin of the directionalselectivity and optokinetic eye movements (Yoshida et al.,2001). The presence and specific role of “starburst” cells infish have not been established and clarified. Strikingly, wedetected only one subpopulation of AChE-containing cellslocated in the inner nuclear layer, but two different neu-ropile bands were observed in the inner plexiform layer.Thus, dendrites belonging to AChE-positive amacrinecells of the inner nuclear layer might branch into bothsublaminae, which is in contrast to the descriptions for“starburst” amacrine cells in the mammalian retina(Famiglietti, 1983, 1992), where an individual “starburst”amacrine cell contributes to only one sublamina. A betterexplanation of this phenomenon might be that the AChEstaining of the vitrealmost sublamina within the innerplexiform layer might be the result of the branching ofChAT-positive “starburst” amacrine cells located in theganglion cell layer as it was described in mammals(Famiglietti, 1983, 1992).

Telencephalon

In the olfactory bulb, AChE activity was observed inmitral cells, as previously described in sharks (Kusunokiet al., 1973). Furthermore, in Scyliorhinus,this cell typeshowed ChAT immunoreactivity (Anadon et al., 2000),which is in contrast to other studies of vertebrates, wherecholinergic neurons were not detected in this telence-phalic region: teleosts (cyprinids: Ekstrom, 1987; batrach-oids: Brantley and Bass, 1988; salmonids: Perez et al.,2000), amphibians (Marın et al., 1997), reptiles (Medina etal., 1993), birds (Medina and Reiner, 1994), and mammals(Ichikawa et al., 1997).

Within the dorsal area of the telencephalic hemi-spheres, the dorsal, medial, and central nuclei showedAChE-positive cells and ChAT-ir axons which could indi-cate a cholinoceptive nature of the AChE-positive neuronswithin these telencephalic nuclei in the zebrafish. InPorichthys, there were some dense ChAT-ir terminalfields in the same nuclei as in zebrafish within the dorsaltelencephalic hemispheres (Brantley and Bass, 1988). Thesame affirmation could be valid for the AChE-positivecells observed in both parts of the entopeduncular nu-cleus, because of the presence of ChAT-ir axons surround-ing them.

The absence of cholinergic cells in the dorsal area of thetelencephalon has been considered as a primitive charac-teristic of vertebrates. No cholinergic cells have been re-ported in this area in either cyprinids (Ekstrom, 1987;present results) or other teleostean families (Brantley andBass, 1988; Perez et al., 2000), nor in the pallium (homol-ogous to the telencephalic dorsal area) of amphibians(Ciani et al., 1988; Marın et al., 1997) or most reptilesstudied so far (turtles: Mufson et al., 1984; Powers andReiner, 1993; crocodile: Brauth et al., 1985; the lizardGekko: Hoogland and Vermeulen-Van der Zee, 1990), withthe exception of the lizard Gallotia (Medina et al., 1993).Moreover, pallial cholinergic neurons are absent almostentirely from cortical regions in birds (Medina and Reiner,1994). In mammals, this characteristic is not a conservedfeature. Cholinergic cells are present in the cortex of therat (Eckenstein and Thoenen, 1983; Houser et al., 1983;Levey et al., 1984; Parnavelas et al., 1986; Blaker et al.,1988; Reiner, 1991), mouse (Mufson and Cunningham,1988), but not in the guinea pig (Maley et al., 1988), or catsor dogs (Kimura et al., 1981; Vincent and Reiner, 1987;St-Jacques et al., 1996). Furthermore, ChAT-ir neuronshave been identified in fetal monkey cerebral cortex (Hen-dry et al., 1987) but not in the cortex of adult primates,including humans (Mesulam et al., 1984; Satoh andFibiger, 1985; Mesulam and Geula, 1988; Geula et al.,1993; Alonso and Amaral, 1995). Thus, the presence ofcortical cholinergic neurons seems to be a secondary fea-ture acquired relatively late during the evolution of ver-tebrates, but it is not a feature certainly shared by am-niotes.

Cholinergic cells have been reported previously in sub-pallial areas in mammals (Kimura et al., 1984; Mesulamet al., 1984; Satoh and Fibiger, 1985; Vincent and Reiner,1987; Maley et al., 1988; Mufson and Cunningham, 1988),birds (Medina and Reiner, 1994), reptiles (Brauth et al.,1985; Hoogland and Vermeulen-Van der Zee, 1990; Me-dina et al., 1993), amphibians (Marın et al., 1997), andboth cyprinid (Ekstrom, 1987) and other teleosts analyzed(Brantley and Bass, 1988, Perez et al., 2000). Nonetheless,


Figure 6


the absence of ChAT-containing cells in the basal telen-cephalon of lamprey (Pombal et al., 2001), dogfish (Ana-don et al., 2000), and sturgeon (Adrio et al., 2000) suggeststhat this feature is not a primitive condition of verte-brates, acquired by teleosts. Among all species of teleostsstudied, cholinergic cells are absent only in the ventral

part of the telencephalon in the zebrafish (present re-sults). This finding could indicate that the presence ofcholinergic cells in this area is a conserved feature in thisgroup but has been secondarily lost in zebrafish.

Diencephalon

Preoptic region. The distribution of cholinergic neu-rons in the preoptic region is very heterogeneous amongdifferent species of teleosts. Thus, the low density of cho-linergic cells that we found in the preoptic region of thezebrafish is in agreement with the results from the cyp-rinid Phoxinus (Ekstrom, 1987) and the noncyprinidPorichthys (Brantley and Bass, 1988), but different fromthe salmonids Oncorhynchus and Salmo (Perez et al.,2000). In the sturgeon, cholinergic neurons have beenidentified in the anterior part of the parvocellular preopticnucleus (Adrio et al., 2000) and among teleosts in the trout(Perez et al., 2000) and in the zebrafish (present results)but not in the cyprinid European minnow (Ekstrom, 1987)or in the noncyprinid midshipman (Brantley and Bass,1988).

The AChE-positive cells observed in the ventralmostregion of the anterior part of the parvocellular preopticnucleus of the zebrafish showed the same location as thecholinergic cells within the preoptic region assigned to theorganum vasculosum laminae terminalis in dogfish andtrout (Anadon et al., 2000; Perez et al., 2000). However, inthe zebrafish, these cells were ChAT-immunonegative;moreover, they received a strong cholinergic input, whichsuggests a possible cholinoceptive nature.

The magnocellular preoptic nucleus projects massivelyto the neurohypophysis in cyprinids (Anglade et al., 1993).

Fig. 7. Transverse sections of the zebrafish brain showing ChAT-and AChE-positive motoneurons of the oculomotor nucleus. a: ChAT-immunostained processes of the ChAT-immunoreactive motoneuronswithin this nucleus aggregated to form the root of the oculomotornerve (arrows). b: AChE-positive neurons in the oculomotor nucleus

(NIII; thick arrows). AChE-positive Purkinje (thin arrows) andgranule cells (arrowheads) can be observed in the ventralmost part ofthe valvula cerebelli, medial part (Vam). For other abbreviations, seelist. Scale bar � 50 �m in b (applies to a,b).

Fig. 6. Distribution of ChAT-immunoreactive (-ir) and ChAT-positive neurons in the zebrafish mesencephalon (coronal sections). a:Photomicrograph of the optic tectum showing ChAT-positive cells inthe periventricular stratum (PVS; arrow). The apical dendrites ramify(arrowheads) forming part of the four bands of neuropile (asterisks) inthe optic (OS), superficial fibrous and gray (SFGS), central gray (CGS)and central white strata (CWS). b: AChE-positive neurons (arrows)were distributed in the SFGS, CGS, CWS, and PVS. Four bands ofAChE-positive neuropile were observed in the same strata as thoseobserved ChAT-immunopositive (asterisks): OS, SFGS, CGS, andCWS. c: ChAT-ir neurons of the rostral tegmental nucleus (RTN) withprocesses coursing toward the nucleus of the medial longitudinalfascicle (NMLF). Cells of the NMLF (inset area) are surrounded byChAT-ir terminals. d: In addition to the high number of AChE-stainedneurons that can be observed in the RTN, some large and intenselystained neurons with the beginning of a stained process orientedtoward the ventral part of the section can be detected in the rednucleus of Schnitzlein (RN). Medial is to the right. e: General over-view of the diencephalon/mesencephalon transition area showingAChE-positive neurons in the mesencephalic nucleus of the trigemi-nal nerve (MNV). Some small AChE-positive neurons can be observedin the ventral part of the torus semicircularis (TS) and in the periven-tricular nucleus of the posterior tuberculum (PTp). f: AChE-positivecells in the NMLF and in the Edinger–Westphal nucleus (EW). Cellsof the NMLF are surrounded by AChE-positive terminals (enlargedarea). For other abbreviations, see list. Scale bar � 50 �m in f (appliesto c,d,f); 25 �m for inset f; 25 �m for a,b; 100 �m for e.


Fig. 8. Distribution of ChAT- and AChE-positive elements in therhombencephalon (a,b are sagittal sections and c–e are transversesections). a: General overview of the zebrafish brain showing theexpression of ChAT immunoreactivity in the optic tectum (OT), sec-ondary gustatory nucleus (SGN), rostral (NVIr) and caudal (NVIc)parts of the abducens nucleus, facial motor nucleus (NVIIm), glosso-pharyngeal and vagal motor nuclei (NIX-Xm), and nucleus of thespino-occipital nerves. b–e: ChAT/AChE distribution in the isthmicregion. b: ChAT-immunoreactive fibers from the SGN run toward thedorsal zone of the hypothalamic periventricular nucleus. Large neu-

rons in the superior reticular nucleus (SRN) and motoneurons in bothparts of the trigeminal motor nucleus (NVmd and NVmv) can also beobserved. Caudal is to the right. c: General overview of the rostralrhombencephalon showing AChE-positive elements in the superiorraphe (SR), isthmic nucleus (IN), nucleus lateralis valvulae (NLV)and SRN. d: High magnification of the AChE-positive staining in theIN. Lateral is to the right. e: Neurons of the SRN, which presentAChE-positive somata and the initial portions of two or three pro-cesses. Lateral is to the right. For other abbreviations, see list. Scalebar � 25 �m in e (applies to d,e); 250 �m for a; 100 �m for b,c.


Moreover, a cholinergic projection from this nucleus to thehypophysis has been described in the Siberian sturgeonand in the trout (Adrio et al., 2000; Perez et al., 2000). Theorigin of the cholinergic innervation observed in the neu-

rohypophysis in the zebrafish could be the magnocellularpreoptic nucleus. Furthermore, the parvocellular preopticnucleus projects to the telencephalon in teleosts (Stried-ter, 1990) and could be the origin of the cholinergic inner-vation of the zebrafish telencephalon; however, no dataare available about this connection in cyprinids.

Epithalamus, thalamus, and posterior tuberculum.

The epithalamus consists of the dorsal and ventral nucleiof the habenula and the light-sensitive pineal organ(epiphysis). The presence of cholinergic neurons in theepiphysis has been reported in lampreys (Pombal et al.,1999; Yanez et al., 1999), elasmobranchs (Anadon et al.,2000), cyprinids (Phoxinus: Ekstrom and Korf, 1986), andsalmonids (Salmo and Oncorhynchus: Perez et al., 2000),but they are absent in the epiphysis in both cyprinid(Danio: present results) and noncyprinid teleosts (Porich-thys: Brantley and Bass, 1988) and in chondrosteans (Ad-rio et al., 2000).

The habenula presents intense ChAT immunoreactivityin amphibians (Marın et al., 1997), reptiles (Medina et al.,1993), birds (Sorenson et al., 1989; Medina and Reiner,1994), and mammals (Houser et al., 1983; Mesulam et al.,1984; Contestabile et al., 1987; Vincent and Reiner, 1987;Mufson and Cunningham, 1988; Ichikawa et al., 1997). Inthe first fish analyzed, cholinergic cells were absent fromthis diencephalic region (cyprinid, European minnow: Ek-strom, 1987; batrachoid, midshipman: Brantley and Bass,1988; anguilid, eel: Molist et al., 1993). With these data,Marın et al (1997) suggested that the existence of cholin-ergic cells within the habenular complex is exclusive totetrapods and appeared initially in early amphibians,which could be supported by our results in zebrafish. How-ever, ChAT-containing cells have been described in thehabenula in other fish, some of them belonging to primi-tive groups (dogfish: Anadon et al., 2000; Siberian stur-geon: Adrio et al., 2000; trout: Perez et al., 2000). Thisfinding demonstrates that the existence of cholinergiccells within the habenular complex may be considered asa common feature of the cholinergic system in vertebrates.Moreover, this could be a primitive (plesiomorphic) char-acteristic of fish that has been modified in some teleostsand has been conserved in tetrapods.

In the thalamus of all fish species hitherto studied,cholinergic cells are absent or scarce and are exclusivelylocated in the dorsal thalamus (Ekstrom, 1987; Brantleyand Bass, 1988; Adrio et al., 2000; Anadon et al., 2000;Perez et al., 2000). In contrast, in our study, AChE-positive neurons appeared exclusively in the ventromedialand ventrolateral nuclei of the ventral thalamus andChAT-ir cells were not observed in any thalamic nuclei.Moreover, the presence of cholinergic neurons in the thal-amus has been described in birds (pigeon: Medina andReiner, 1994) and in mammals (monkey: Rico and Cavada,1998). All these data favor the hypothesis that the pres-ence of cholinergic cells in the thalamus appeared severaltimes during evolution, so they are homoplastic (parallelevolution).

AChE-positive cells are observed in the periventricularnucleus of the posterior tuberculum in the zebrafish. Thisnucleus presented AChE/ChAT reactivity in the eel(Molist et al., 1993). In Phoxinus, Salmo, and Acipenser,there was a subpopulation of cholinergic cells in the para-ventricular organ (Ekstrom, 1987; Adrio et al., 2000; Perezet al., 2000). No AChE/ChAT-positive cells were present inthis nucleus in the zebrafish posterior tuberculum.

Fig. 9. ChAT and AChE distribution in the cerebellum (transversesections). a,b: General overview of the corpus cerebelli (CCe) andlobus caudalis (LCa), showing no ChAT-immunoreactive elements (a)but numerous AChE-positive Purkinje and granule cells (b). c: Highermagnification of positive Purkinje cells in the ventral part of thevalvula cerebelli, part medialis. Scale bar � 30 �m in c; 120 �m fora,b.


Hypothalamus. The zebrafish hypothalamus is con-spicuously the largest part of the diencephalon. No cellsomata positive to AChE histochemistry or ChAT im-munohistochemistry were observed in the zebrafish hy-pothalamus. This finding is in agreement with a previ-ous study of the distribution of cholinergic cells in thebrain of a noncyprinid teleost Porichthys notatus(Brantley and Bass, 1988). The presence of cholinergicor cholinoceptive cell groups is poorly represented inmost teleosts studied to date. Thus, in the cyprinidPhoxinus phoxinus, ChAT-positive cells were detectedonly in the caudal zone of the periventricular hypothal-amus (Ekstrom, 1987). In the trout, a salmonid, twogroups of putative cholinergic cells were described ex-clusively in the anterior tuberal nucleus and in a zonelateral to the paraventricular organ (Perez et al., 2000).In nonteleostean fish, such as the dogfish, some cholin-ergic cells appeared in the boundary between the pos-terior recess organ with the hypophysis and saccus vas-

culosus and in the lateral tuberal nucleus (Anadon etal., 2000). In chondrosteans, cholinergic cells were de-scribed in the lateral recess wall, diffuse nucleus of theinferior lobe, anterior tuberal nucleus, lateral tuberalnucleus, and dorsolateral wall of the posterior recess(Adrio et al., 2000). In contrast, cholinergic cells in thehypothalamus of the rest of the vertebrate groups aremore widely distributed. Cholinergic cells have beendescribed in the periventricular hypothalamus of am-phibians (Marın et al., 1997) and in different hypotha-lamic regions of mammals (Tago et al., 1987, 1989;Woolf, 1991; Ichikawa et al., 1997), reptiles (Medina etal., 1993; Powers and Reiner, 1993) and birds (Medinaand Reiner, 1994). Therefore, this poor representationor absence of the cholinergic/cholinoceptive system inthe teleostean hypothalamus seems to be a derivedcharacteristic acquired during the evolutive radiation ofteleosts. In other words, the presence of cholinergicpopulations in the hypothalamus of vertebrates might

Fig. 10. Transverse sections of the zebrafish brain showing ChAT-and AChE-positive motoneurons of the rhombencephalic motor nu-cleus of the cranial nerves. a: Both dorsal (NVmd) and ventral (NVmv)parts of the trigeminal motor nucleus showed ChAT-immunoreactiveneurons. The ventral motor root of the trigeminal nerve is also stained(arrowheads). b: Both NVmd and NVmv are AChE-positive. Lateral isto the left. c: Photomicrograph of the caudal part of the abducens

nucleus (NVIc). One ChAT-immunoreactive axon (arrowheads) rantoward the root of the abducens nerve (VI). Lateral is to the right.d: The NVIc shows one small group of AChE-stained neurons locatedjust below the intermediate part of the reticular nucleus (ImRN).AChE-positive giant cells of the ImRN show two thick processes(higher magnification). For other abbreviations, see list. Scale bar �60 �m in d (applies to a–d); 25 �m for inset d.


be a primitive feature that was completely lost or highlyreduced during the evolutive radiation of teleosts.

Strikingly, a profuse ChAT-ir innervation was observedin the hypothalamus, but no AChE-positive cells wereobserved in this region. It has been reported previously, inthe striatum, that AChE is not always necessarily present

in the target cell of a cholinergic synapse (Henderson,1981; Bolam et al., 1984). Therefore, the hydrolysis of AChin the zebrafish hypothalamus should be performed byother cholinesterases different from AChE, which shouldbe inactivated during the histochemical procedure for thedetection of AChE. There are different reports that show

Fig. 11. ChAT and AChE distribution in the zebrafish octavolat-eral area (transverse sections). a,b: ChAT-immunoreactive (a) andAChE-positive (b) neurons in the octavolateral efferent nucleus show-ing the beginning of stained dendrites ventrolaterally oriented (ar-rows). c: The somata and the beginning of the dendrites of the Mau-thner cells (MA) are positive to AChE histochemistry. In the inset,some AChE-positive terminals impinging on the cell body of the MAcan be observed. d: Weakly AChE-positive cells in the medial octavol-ateral nucleus (MON) and strongly positive cells in the anterior octa-

val nucleus (AON). The lateral dendrite of an MA can be observed.e: AChE-positive neurons in the magnocellular octaval nucleus(MaON) adjacent to the root of the octaval nerve (VIII). f: In theposterior octaval nucleus (PON), AChE-positive neurons can be ob-served dorsal and ventral to the root of the vagal nerve (X). AChE-positive cells are also observed in the lateral reticular nucleus (LRN).For other abbreviations, see list. Scale bar � 60 �m in f (applies toa,b,d–f); 120 �m for c; 30 �m for c inset.


afferents to the inferior hypothalamic lobe: the nucleusglomerulosus or its homologue, the posterior pretectal nu-cleus (Sakamoto and Ito, 1982; Wullimann and Meyer,1990); the suprachiasmatic nucleus and the tegmentum(Wullimann et al., 1991); the telencephalon (Airhart et al.,1988); and the secondary gustatory nucleus (Morita et al.,1980). However, only that from the gustatory center hasbeen described as cholinergic in fish (eel: Molist et al.,1993; trout: Perez et al., 2000), which is in agreement withour observations of a strong cholinergic projection fromthe secondary gustatory nucleus to the hypothalamic in-ferior lobe.

Synencephalon and pretectum. In the zebrafish sy-nencephalon, AChE-positive neurons were observed ex-clusively in the nucleus of the medial longitudinal fasciclefrom early stages of zebrafish development (Hannemanand Westerfield, 1989). This nucleus also showed AChElabeling in Anguilla (Molist et al., 1993), but ChAT-immunoreactive neurons were not observed in this area inany studied vertebrate except in the dogfish (Anadon etal., 2000). This finding could mean that AChE-positive

cells of the nucleus of the medial longitudinal fasciclecould be cholinoceptive, which is supported by the pres-ence of AChE-positive and ChAT-ir terminals impingingon neurons of this nucleus.

The data about the cholinergic nature of the fish pretec-tum are controversial, even among the different te-leostean groups. Thus, whereas ChAT-ir neurons were notpresent in chondrosteans (Siberian sturgeon: Adrio et al.,2000) and some teleosts, including both cyprinids (ze-brafish: present results) and fish belonging to other te-leostean groups, such as the midshipman (Brantley andBass, 1988), ChAT-ir cells were described in elasmo-branchs (dogfish: Anadon et al., 2000), and both cyprinids(European minnow: Ekstrom, 1987) and noncyprinids(trout: Perez et al., 2000). The presence of ChAT-immunopositive elements has also been described in thepretectum of birds (Sorenson et al., 1989; Medina andReiner, 1994). In contrast with fish and birds, cholinergiccell bodies are not present in the pretectum of amphibians(Marın et al., 1997), reptiles (Medina et al., 1993; Powersand Reiner, 1993), or mammals (Houser et al., 1983; Me-

Fig. 12. ChAT/AChE elements in the caudal rhombencephalonand rostral spinal cord (coronal sections). a: The caudalmost part ofthe vagal motor nucleus (NXm) and some neurons of the nucleus ofthe spino-occipital nerve (NSO) display ChAT immunoreactivity.Whereas the stained cells of the NXm (arrowheads) and some neuronsof the NSO (small arrow) show a strong staining, other neurons of theNSO are bigger and more weakly positive (large arrow). Some mo-toneurons of the rostralmost spinal cord can be observed (open ar-rows). b: Strongly AChE-positive neurons of the vagal motor nucleus(NXm) with a stained process oriented toward the root of the vagal

nerve. Vagal lobe (LX) does not show stained neurons. c: ChAT-immunoreactive neurons in the ventral horn (VH) of the spinal cordshowing two morphologies. Large, weakly stained primary motoneu-rons (arrow) are intermingled with smaller and more intenselystained secondary motoneurons (arrowheads). d: General overview ofthe rostral spinal cord with AChE-positive motoneurons within theventral horn (VH). A detail of stained motoneurons can be observed inthe inset. For other abbreviations, see list. Scale bar � 80 �m in d; 30�m for d inset; 60 �m for a,b; 100 �m in c.


sulam et al., 1984; Satoh and Fibiger, 1985; Kasa, 1986;Vincent and Reiner, 1987; Maley et al., 1988; Woolf, 1991).All these data suggest a homoplastic evolution of the cho-linergic pretectal system in fish and birds (Medina andReiner, 1994).

AChE-positive neurons were observed in the magnocel-lular superficial, central, posterior, and accessory pretec-tal nuclei and in the dorsal accessory optic nucleus. Wul-limann and Meyer (1990) detected AChE-positive cells inCarassius in the magnocellular superficial pretectal nu-cleus and positive neuropile in the parvocellular part ofthe same nucleus and in the posterior and accessory pre-tectal nuclei. Nevertheless, in another cyprinid, Phoxinus,the magnocellular superficial pretectal nucleus is the onlynucleus that shows AChE-positive elements (Ekstrom,1987). With regard to other noncyprinid teleosts, AChE-positive elements have been described in Osteoglossum(Wullimann and Meyer, 1990) in the parvocellular andmagnocellular superficial pretectal nuclei, in the corticalnucleus and in the posterior pretectal nucleus and inHemichromis (Wullimann and Meyer, 1990), in the mag-nocellular, parvocellular and intermediate superficial pre-tectal nuclei and in the cortical nucleus. These abundantspecies-specific differences in the AChE distribution ob-served among the fish analyzed may be related to the highstructural variability in the pretectum of teleosts, a reflec-tion of the different connectivity that this region showsamong different fish groups analyzed (Meek and Nieuwen-huys, 1998).

Although ChAT-ir somata were absent in the zebrafishpretectum, a rich cholinergic innervation was detected inthis region. This could indicate the cholinoceptive natureof the AChE-positive pretectal neurons in the zebrafish.Moreover, the parvocellular superficial pretectal nucleusshowed a rich innervation, not only around it, but also inits inner part, in agreement with previous results in othercyprinids, such as the European minnow (Ekstrom, 1987),and in fish belonging to other teleostean families, such asthe trout (Perez et al., 2000). This region is strongly in-nervated by the isthmic nucleus and/or by the optic tectumin Navodon (Murakami et al., 1986), Carassius (Zottoli etal., 1988), Salmo and Oncorhynchus (Perez et al., 2000),and Cyprinus (Xue et al., 2001), both areas displayingChAT-ir cell somata in the zebrafish (present results).

Mesencephalon

Optic tectum. AChE-positive cells were more numer-ous than ChAT-positive ones in the optic tectum of thezebrafish. AChE-positive cells were distributed in the op-tic, superficial fibrous and gray, central gray, centralwhite, and periventricular strata, in agreement with pre-vious studies carried out in another cyprinid, the goldfish(Villani et al., 1979), and in other noncyprinid teleosts: thecatfish and the gar Lepisosteus (Villani et al., 1979; North-cutt and Butler, 1993). In contrast, other reports aboutAChE distribution in this part of the fish brain onlyshowed stained type XIV cells of Meek and Schellart(1978) within the periventricular stratum in other cyprin-ids (Phoxinus: Ekstrom, 1987; Carassius: Wullimann andMeyer, 1990) as well as in fish belonging to other families(Ossteoglossum: Wullimann and Meyer, 1990; Anguilla:Molist et al., 1993). ChAT immunohistochemistry re-vealed that these type XIV cells are the only cholinergicneural somata in the optic tectum of different species ofteleosts (Porichthys: Brantley and Bass, 1988; Anguilla:

Molist et al., 1993; Salmo and Oncorhynchus: Perez et al.,2000), including all cyprinids so far analyzed (Phoxinus:Ekstrom, 1987; Carassius: Tumosa and Stell, 1986; Tu-mosa et al., 1986; Zottoli et al., 1988; Danio: presentresults) and are a well-conserved feature among all groupsof teleosts. Therefore, ChAT- and AChE-positive cells inthis stratum might be cholinergic in the zebrafish. Theremaining AChE-positive but ChAT-immunonegativecells could be cholinoceptive and would be contacted by therich innervation observed forming ChAT-ir neuropilebands within the optic, superficial fibrous and gray, cen-tral gray, and central white strata. The bands were simi-lar in location and thickness to those described with AChEhistochemistry, indicating that there is no substan-tial ChAT-negative/AChE-positive innervation. TheseChAT-ir bands are originated, at least in part, by theapical dendrite of type XIV cells, but also by the axons ofthese neurons and other fibers from different sources.Thus, a cholinergic input to the optic tectum has beendescribed from the isthmic nucleus in several fish species(Ekstrom, 1987; Brantley and Bass, 1988; Zottoli et al.,1988; Molist et al., 1993; Perez et al., 2000) and from thesuperior reticular nucleus (Ekstrom, 1987; Zottoli et al.,1988). AChE-positive bands of neuropile are not exclusiveto the zebrafish optic tectum, being observed in other fish(Contestabile, 1975; Contestabile and Zannoni, 1975; Os-wald et al., 1980; Zottoli et al., 1987). This AChE-positiveneuropile is located in retinorecipient strata (Vanegas andIto, 1983), but it has been demonstrated that the retino-tectal fibers are not cholinergic, but they receive a cholin-ergic input within the optic tectum (Henley et al., 1986)and retinal ganglion cells are not cholinergic (Tumosa andStell, 1986). The origin of this cholinergic innervation ofthe optic tectum might be the isthmic nucleus and thesuperior reticular nucleus in the goldfish (Zottoli et al.,1988), which possessed AChE- and ChAT-positive neuronsin the zebrafish.

ChAT-immunopositive cells have been described in theoptic tectum of birds (Medina and Reiner, 1994) but not inlampreys (Pombal et al., 2001), elasmobranchs (Anadon etal., 2000), chondrosteans (Adrio et al., 2000), amphibians(Marın et al., 1997), or reptiles (Brauth et al., 1985; Me-dina et al., 1993; Powers and Reiner, 1993). Mammalsshow cholinergic neurons in the superior colliculus, ho-mologous to the optic tectum of the rest of vertebrates(Tago et al., 1987; Vincent and Reiner, 1987). This findingcould indicate that the presence of cholinergic cells in theoptic tectum is a secondary feature acquired separately byteleosts, birds, and mammals during evolution (ho-moplasy).

Torus semicircularis. We detected AChE-positivecells and a profuse ChAT-ir innervation in the ventrolat-eral nucleus of the torus semicircularis of the zebrafish.The torus semicircularis is a sensory integrative mesen-cephalic area acting as a target for the ascending octavol-ateralis system (Wullimann et al., 1996). The ventrolat-eral nucleus of the torus semicircularis receives afferentsfrom the medial and caudal octaval nucleus (both ofthem containing abundant AChE-positive but ChAT-immunonegative neurons in zebrafish) included in themechanoreceptive ascending pathway. Cholinergic axonsfrom the isthmic nucleus project to the optic tectumthroughout the torus semicircularis (Zottoli et al., 1988).Some of these axons could make synapses with AChE-positive neurons of the torus semicircularis, suggesting a


cholinoceptive nature of AChE-containing cells in thismesencephalic area. AChE staining in the torus semicir-cularis in fish has only been reported in the eel (Molist etal., 1993), but without displaying ChAT-ir cells, whichwere observed in the zebrafish.

Tegmentum. The cholinergic system is widely distrib-uted in the zebrafish tegmentum. As in other brain re-gions, AChE-positive elements were more abundant thanChAT-ir ones. Thus, we detected AChE reactivity in themesencephalic nucleus of the trigeminal nerve, rostraland dorsal tegmental nuclei, Edinger–Westphal nucleus,nucleus of the lateral lemniscus, PL, oculomotor andtrochlear nuclei, and interpeduncular nucleus. ChAT-irsomata were observed exclusively in the rostral tegmen-tal, oculomotor, and trochlear nuclei.

There are no data about the cholinergic nature of therostral tegmental nucleus of Grover and Sharma (1981).Our results are the first reported that show a connectionfrom the rostral tegmental nucleus to the nucleus of themedial longitudinal fascicle and to the hypothalamus,since it has been reported that this nucleus projects to theoptic tectum in cyprinid (Grover and Sharma, 1981) andnoncyprinid fish (Northcutt, 1982). In all fish species an-alyzed, the oculomotor nucleus, as well as the trochlearnucleus, showed AChE activity and/or ChAT immunore-activity (Rhodes et al., 1986; Ekstrom, 1987; Brantley andBass, 1988; Molist et al., 1993; Anadon et al., 2000; Adrioet al., 2000; Perez et al., 2000). Our results are in agree-ment with the observed conservative cholinergic nature ofthese two cranial nerve motor nuclei throughout fish evo-lution.

A population of AChE-positive cells in the Edinger–Westphal nucleus was observed. Cholinergic neurons inthis nucleus have been described in amphibians (Marın etal., 1997), reptiles (Medina et al., 1993; Powers andReiner, 1993), birds (Sorenson et al., 1989; Medina andReiner, 1994), and mammals (Mesulam et al., 1984; Tagoet al., 1989), but this nucleus has been described as ChAT-immunonegative in all teleosts studied so far (Ekstrom,1987; Brantley and Bass, 1988; Molist et al., 1993; Perezet al., 2000; present results). Moreover, this nucleusshowed ChAT-ir terminals, suggesting the cholinoceptivenature of the AChE-positive neurons.

In the red nucleus of Schnitzlein, AChE-positive butChAT-immunonegative neurons have been described inthe eel (Molist et al., 1993) and in the zebrafish (presentresults). The dorsalmost part of the interpeduncular nu-cleus showed scattered AChE-positive and ChAT-immunonegative neurons in the zebrafish. ChAT-irperikarya have not been described in the interpeduncularnucleus in the dogfish (Anadon et al., 2000), sturgeon(Adrio et al., 2000), goldfish (Villani et al., 1994), or trout(Perez et al., 2000). In these species, the main origin ofcholinergic innervation to this nucleus arises from thehabenula (Villani et al., 1987, 1994; Anadon et al., 2000;Adrio et al., 2000; Perez et al., 2000), which is AChE- andChAT-negative in the zebrafish and in other teleosts(Phoxinus phoxinus: Ekstrom, 1987; Porichthys notatus:Brantley and Bass, 1988). These results demonstrate agreat species-specific variation in the habenulo-interpeduncular system among fish, even among speciesfrom the same family as is the case of cyprinids (Ekstrom,1987; Villani et al., 1994; present results). We observedscarce AChE-positive neurons in the dorsalmost portion ofthe interpeduncular nucleus. This part of the nucleus

receives afferents from the dorsal nucleus of the telence-phalic ventral area in goldfish (Villani et al., 1996), whichdid not present either AChE-positive or ChAT-ir cells inzebrafish. Moreover, we did not detect cholinergic inner-vation from any other nuclei to the interpeduncular nu-cleus. This phenomenon could be the same previouslydescribed in neurons of the nigrostriatal system, whichhave large amounts of AChE without an identifiable cho-linergic afferent pathway (Butcher and Woolf, 1982). Inthese neurons, AChE might have another function differ-ent from ACh hydrolysis (Hefti and Lichtensteiger, 1978;Chubb et al., 1980; Llinas et al., 1984).

Rhombencephalon

Isthmus. The presence of cholinergic cells in the isth-mic nucleus has been detected in all vertebrate groupsanalyzed. Thus, the isthmic nucleus is cholinergic in lam-preys (Pombal et al., 2001), elasmobranchs (Anadon et al.,2000), chondrosteans (Adrio et al., 2000), teleosts (Ek-strom, 1987; Brantley and Bass, 1988; Zottoli et al., 1988;Perez et al., 2000; present results), amphibians (Marın etal., 1997), reptiles (Medina et al., 1993; Powers andReiner, 1993), and birds (Medina and Reiner, 1994). Thehomologue of the isthmic nucleus in mammals, the para-bigeminal nucleus, is also cholinergic (Mufson et al., 1986;Tago et al., 1989; Woolf, 1991). A cholinergic connectionfrom the isthmic/parabigeminalis nucleus to the optictectum/superior colliculus has been observed in mammals(Beninato and Spencer, 1986; Mufson et al., 1986), birds(Sorenson et al., 1989; Bagnoli et al., 1992; Medina andReiner, 1994), reptiles (Medina et al., 1993; Powers andReiner, 1993), amphibians (Ricciuti and Gruberg, 1985;Desan et al., 1987; Marın et al., 1997; Marın and Gonza-lez, 1999), and teleosts (Ekstrom, 1987; Brantley andBass, 1988; Zottoli et al., 1988; Molist et al., 1993; Perez etal., 2000). Thus, the cholinergic nature of the isthmic/parabigeminal nucleus and this connection of the visualpathway seem to be a well-conserved feature throughoutphylogeny. The isthmic nucleus also projects to the par-vocellular and/or magnocellular superficial pretectal nu-cleus (Murakami et al., 1986; Striedter and Northcutt,1989; Perez et al., 2000; Xue et al., 2001). Therefore, itcould be the source of the profuse cholinergic innervationobserved in the parvocellular superficial pretectal nucleusin zebrafish.

Some AChE/ChAT-positive cells were observed in thesuperior reticular nucleus. This finding is in agreementwith the observations described in other teleosts (Ek-strom, 1987; Brantley and Bass, 1988; Zottoli et al., 1988;Molist et al., 1993; Perez et al., 2000). This nucleusprojects to the superficial pretectal nucleus, contralateralpreoptic region, and optic tectum in other cyprinids (Ek-strom, 1987; Zottoli et al., 1988), regions where the pres-ence of AChE-positive neurons and, in the case of thepretectal nucleus and optic tectum, AChE-positive neuro-pile were observed in the zebrafish.

The secondary gustatory nucleus showed ChAT-ir cellsin the zebrafish, which is in agreement with previousobservations made in the cyprinid Carassius (Zottoli et al.,1988), and in other noncyprinid fish: Anguilla (Molist etal., 1993) and Salmo and Oncorhynchus (Perez et al.,2000). However, in another cyprinid (Phoxinus), Ekstrom(1987) described a similar group of ChAT-ir neurons asnucleus “a” of Nieuwenhuys and Pouwels (1983). The sec-ondary gustatory nucleus projects to the hypothalamus in


teleosts but also to another diencephalic nucleus that re-ceives different names but that hodologically are similar:preglomerular tertiary gustatory nucleus in Lepomis(Wullimann, 1988), Oreochromis (Yoshimoto et al., 1998),and in the cyprinid Carassius auratus (Rink and Wulli-mann, 1996), lateral thalamic nucleus in some catfish(Kanwal et al., 1988; Lamb and Caprio, 1993), and nucleusglomerulosus in Carassius carassius (Morita et al., 1980).Therefore, in all teleosts, there is a connection from thesecondary gustatory nucleus to a preglomerular tertiarygustatory nucleus. However, in zebrafish, as well as introut (Perez et al., 2000), the cholinergic projection fromthe secondary gustatory nucleus arrived exclusively at thehypothalamus. There are two possibilities to explain thisobservation. The first possibility is that different regionsof the secondary gustatory nucleus possess different tar-gets in the diencephalon. Although different subnucleihave been described within the secondary gustatory nu-cleus in one catfish, all of them project to the same nucleiwithin the diencephalon (Lamb and Finger, 1996). Fur-ther hodological investigations should be carried out toverify this possibility in zebrafish. A more plausible expla-nation could be that ChAT-ir neurons within the second-ary gustatory nucleus do not represent the total number ofneurons within this nucleus. At least one other subpopu-lation should exist that, by using a different neurotrans-mitter, projects to the preglomerular tertiary gustatorynucleus. However, no data are available about the natureof the neurotransmitters used by the neurons of the sec-ondary gustatory nucleus, apart from ACh.

In catfish and in goldfish, a secondary visceral nucleushas been described that receives projections from the gen-eral viscerosensory region of the medulla, not from gusta-tory centers (Finger and Kanwal, 1992). This nucleusshowed a similar topography compared with that de-scribed for the cholinergic secondary gustatory nucleus ingoldfish (Zottoli et al., 1988), which suggests that, at leastin this species, this cholinergic cell group might be therelated to the viscerosensory system. However, specificstudies of the connections of the secondary gustatory nu-cleus with the primary sensory areas are necessary toresolve this question in the zebrafish.

Cerebellum. In the three subdivisions of the zebrafishcerebellum, AChE-positive/ChAT-immunonegative Pur-kinje and granule cells are the only stained neuronaltypes. The presence of AChE staining and/or the absenceof ChAT immunoreactivity in Purkinje cells are sharedfeatures in different fish species (Salmo and Poecilia: Con-testabile and Zannoni, 1975; Carassius: Contestabile andZannoni, 1975; Contestabile et al., 1977; Ictalurus: Con-testabile et al., 1977; Phoxinus: Ekstrom, 1987; Porichtys:Brantley and Bass, 1988; Acipenser: Adrio et al., 2000;Scyliorhinus: Anadon et al., 2000; Salmo and Oncorhyn-chus: Perez et al., 2000; Danio: present results). TransientChAT expression in Purkinje cells has been described inthe early postnatal development of the rat (Gould andButcher, 1987). ChAT-positive neurons in adult animalshave been previously identified as Golgi cells in dogfishand cat cerebellum (Ikeda et al., 1991; Anadon et al.,2000). Nonetheless, cholinergic cells were not observed inthe cerebellum of teleosts, amphibians, reptiles, birds, orother mammals. Furthermore, no cholinergic innervationhas been observed onto Purkinje cells.

AChE-positive granule cells were observed in severalfish species (Contestabile and Zannoni, 1975; Contestabile

et al., 1977) and in other vertebrates (Contestabile andTabanelli, 1977; Villani et al., 1977; Kusunoki et al., 1987;Siripholvat et al., 1988; Robertson and Roman, 1989).However, there are differences in the distribution of gran-ule cells. Thus, although in the goldfish all granule cellsappeared positive to AChE histochemistry (Contestabileet al., 1977), in the catfish (Contestabile et al., 1977) andin the zebrafish (present results) there is only one sub-population that shows AChE reactivity. On the otherhand, granule cells are not described as cholinergic in anyof the fish analyzed (Ekstrom, 1987; Brantley and Bass,1988; Anadon et al., 2000; Adrio et al., 2000; Perez et al.,2000; present results).

The cerebellum has been described as one region whereAChE exists beyond the requirements or in the absence ofcholinergic transmission in mammals (Appleyard andJahnsen, 1992). In guinea pig, AChE enhances the re-sponse of Purkinje cells to excitatory amino acids releasedby granule cells (Appleyard and Jahnsen, 1992). In thezebrafish Purkinje cells, AChE may have the same func-tion, because it has been described that fish granule cellsuse the excitatory neurotransmitter glutamate in theirsynapses with Purkinje cells (Somogyi et al., 1990).

Medulla oblongata and spinal cord.

Trigeminal, abducens, and facial motor nuclei. Mo-toneurons of the trigeminal, abducens, and facial motornuclei are cholinergic in lampreys (Pombal et al., 2001),elasmobranchs (Anadon et al., 2000), chondrosteans (Ad-rio et al., 2000), teleosts (Ekstrom, 1987; Brantley andBass, 1988; Zottoli et al., 1988; Perez et al., 2000; presentresults), amphibians (Marın et al., 1997), reptiles (Medinaet al., 1993; Powers and Reiner, 1993), birds (Medina andReiner, 1994), and mammals (Armstrong et al., 1983; Tagoet al., 1989). The results obtained in zebrafish support thehypothesis that motoneurons of cranial nerves are cholin-ergic throughout vertebrate phylogeny.

Reticular formation. The intermediate and inferior re-ticular nuclei displayed AChE-positive neurons, but noChAT-ir cells were detected. Neurons from these two nu-clei showed AChE-positive preterminal puncta on them.Due to the morphological characteristics of reticular nu-clei, there are not many data about their afferents. Thereare data in cyprinids of afferents from the optic tectum(Grover and Sharma, 1981; Luiten, 1981) and the cerebel-lum (Wullimann and Northcutt, 1988). However, in thezebrafish, the only cholinergic neuronal type within theoptic tectum is the periventricular type of XIV cells, whichdo not project out of this structure, and no cholinergic cellswere present in the cerebellum. Therefore, in the ze-brafish, the cholinergic input to the reticular nuclei shouldoriginate in other regions different from the optic tectumand the cerebellum. On the other hand, although no cho-linergic cells were present in the zebrafish intermediateand inferior reticular nuclei, they have been reported inlampreys (Pombal et al., 2001), elasmobranchs (Anadon etal., 2000), teleosts (Ekstrom, 1987; Brantley and Bass,1988; Perez et al., 2000), amphibians (Marın et al., 1997),reptiles (Medina et al., 1993; Powers and Reiner, 1993),birds (Medina and Reiner, 1994), and mammals (Kimuraet al., 1981; Hendry et al., 1987; Vincent and Reiner, 1987;Maley et al., 1988; Mufson and Cunningham, 1988; Rug-giero et al., 1990). Additionally, AChE-positive neuronswere detected in the lateral reticular nucleus of zebrafish,where ChAT-ir cells were not present. This lateral retic-ular nucleus is cholinergic in reptiles (Brauth et al., 1985;


Medina et al., 1993; Powers and Reiner, 1993) and birds(Medina and Reiner, 1994).

Mauthner cells are located in the rostral octavolateralregion. They mediate fast-escape motor responses, impor-tant in predator avoidance after reception of unexpectedvibrational and/or visual stimuli (Eaton and Bombardieri,1978). In the present study, both the somata and proximaldendrites of Mauthner cells were positive to AChE histo-chemistry, but they did not show ChAT immunoreactivity,similarly to what has been reported in the goldfish(Rhodes et al., 1986). Previous studies demonstrated, withindirect evidence, that Mauthner cells use ACh as a neu-rotransmitter (Spira et al., 1970; Model et al., 1972; Day etal., 1983). Different reasons may account for the lack ofChAT staining in the zebrafish Mauthner cell: (1) TheChAT molecule may be synthesized in the soma and trans-ported in an inactive immunonegative form to the termi-nals, where it is latter activated. However, the consistentfinding of ChAT-ir cell bodies, dendrites, and axons inmany other cell groups in the zebrafish argues againstsuch an interpretation. (2) In a neuron as large as theMauthner cell, the ChAT molecule may be distributeddiffusely throughout the cytoplasm. It might be possiblethat AChE in the Mauthner cells plays a different role inthe degradation of other transmitters besides ACh (Chubbet al., 1980). It is also possible that the cell is receivingcholinergic input in the cell body region. We observedpreterminal puncta positive to AChE histochemistry onthe surface of the Mauthner cell somata and proximaldendrites as has been observed in the rainbow trout (Pfis-ter et al., 1973), perhaps indicating that Mauthner cells,at least in these two species, receive cholinergic input.However, no ChAT-ir fibers were observed impinging onthem.

Octavolateral area. The presence of a cholinergic effer-ent nucleus within the octaval region is a conserved fea-ture in most fish hitherto analyzed. Thus, it is present inPorichthys (Brantley and Bass, 1988), Carassius (Daniel-son et al., 1988), Anguilla (Roberts et al., 1994), but not inthe cyprinid Phoxinus (Ekstrom, 1987). There are alsoreports about a cholinergic eight-nerve-efferent nucleus inamphibians (Marın et al., 1997), reptiles (Medina et al.,1993; Powers and Reiner, 1993), birds (Sorenson et al.,1989; Medina and Reiner, 1994), and mammals (Tago etal., 1989). Therefore, the presence of cholinergic efferentsthat modulate the stato-acoustic receptors of the inner earis a well-conserved characteristic among vertebrates. Inthe teleosts studied to the present, cholinergic cells in theoctavolateral area are absent or poorly represented (Ek-strom, 1987; Brantley and Bass, 1988; Perez et al., 2000;present results). Nonetheless, the octavolateral area con-tains abundant cholinergic cells in dogfish (Anadon et al.,2000). In other vertebrate groups, ChAT-containing cellsappear in very concrete regions (Kimura et al., 1984; Car-penter et al., 1987, 1990; Jones and Beaudet, 1987; Tagoet al., 1989; Barmack et al., 1992; Powers and Reiner,1993; Medina and Reiner, 1994; Marın et al., 1997). Theseresults may indicate that the presence of cholinergic cellsin the octaval region may be a primitive feature of verte-brates. A reduction of these populations is observed intetrapods, whereas teleosts may have lost these popula-tions secondarily.

On the other hand, AChE activity was displayedthroughout all the rostrocaudal octavolateral area. Thus,AChE-positive cells were detected in the medial and cau-

dal octavolateral nuclei, secondary octaval population,and anterior, magnocellular, descending and posterior oc-taval nuclei. Nuclei within the octavolateral area receive aprofuse ChAT-ir innervation, which could indicate thecholinoceptive nature of the AChE-positive neuronswithin the aforementioned nuclei. The available dataabout the innervation of the octavolateral area in cyprin-ids are very scarce. There are reports about afferents tothis area from the acoustic and vestibular organs(Echteler, 1984; McCormick and Braford, 1994) and adescending projection from the octaval population to thedescending octaval nucleus (Echteler, 1984; McCormickand Hernandez, 1996). In the zebrafish, the octaval pop-ulation did not show cholinergic somata. Therefore, a dif-ferent source of the cholinergic fibers present in the octa-val area must be investigated. Some of this innervationcould be originated in the descending octaval nucleus,which showed ChAT-ir cells in the zebrafish. Neverthe-less, the observed efferents of this nucleus in cyprinids arerestricted to the trochlear and oculomotor nuclei in gold-fish (Torres et al., 1992, 1995) and to the torus semicircu-laris in carp (Echteler, 1984).

Caudal rhombencephalon. The presence of cholinergiccells in the visceromotor column is common in all verte-brates studied to date, which is supported by our data inzebrafish (Armstrong et al., 1983; Ekstrom, 1987; Brant-ley and Bass, 1988; Zottoli et al., 1988; Tago et al., 1989;Medina et al., 1993; Powers and Reiner, 1993; Medina andReiner, 1994; Marın et al., 1997; Adrio et al., 2000; Ana-don et al., 2000; Perez et al., 2000; Pombal et al., 2001).

ChAT-ir motoneurons of the nucleus of the spino-occipital nerves overlap rostrally with the caudalmost por-tion of the ChAT-ir vagal motor nucleus and ventrallywith the rostralmost ChAT-ir neurons of the spinal cord.This nucleus shows cholinergic neurons in other fish(Brantley and Bass, 1988; Adrio et al., 2000; Anadon et al.,2000; Perez et al., 2000; Pombal et al., 2001). The nucleusof the spino-occipital nerves is homologous to the hypo-glossal motor nucleus of tetrapods, which is cholinergic inall vertebrate groups analyzed (Woolf, 1991; Medina et al.,1993; Medina and Reiner, 1994; Marın et al., 1997).

Spinal cord. The size and location of the large AChE-and ChAT-positive neurons of the rostral spinal cord issimilar to that described for primary motoneurons in aprevious study in zebrafish (Westerfield et al., 1986). Inaddition to large motoneurons, we observed a subpopula-tion of ChAT-ir smaller cells in the spinal cord. Differenttypes of ChAT-ir neurons have been described in the spi-nal cord of other fish in addition to primary motoneurons.Cholinergic cells were observed in the dorsal horn, themarginal nucleus and the ventral funiculus in dogfish(Anadon et al., 2000). In the Siberian sturgeon, somepositive cerebrospinal fluid-contacting cells were de-scribed (Adrio et al., 2000), and, in the trout, some inter-stitial cells within the ventral funiculus were also ChAT-ir(Perez et al., 2000). The small ChAT-ir cells observed inthe zebrafish spinal cord did not correspond to any of thelocations or morphologies observed in the aforementionedfish. The morphology and branching pattern of ChAT-irsmall cells in zebrafish indicate that they correspond tosecondary motoneurons, similar to those previously de-scribed in this species (Westerfield et al., 1986) and ingoldfish (Fetcho, 1991). The presence of AChE reactivityin motoneurons has been described from early stages ofzebrafish development (Hanneman and Westerfield,


1989). Cholinergic motoneurons have been described indogfish (Anadon et al., 2000), Siberian sturgeon (Adrio etal., 2000), goldfish (Rhodes et al., 1986), midshipman(Brantley and Bass, 1988), brown and rainbow trout(Perez et al., 2000), and zebrafish (present results). More-over, motoneurons have been described as cholinergic inother vertebrate classes (mammals: Kimura et al., 1981;Satoh and Fibiger, 1985; Vincent and Reiner, 1987; Maleyet al., 1988; birds: Medina and Reiner, 1994; reptiles:Medina et al., 1993; Powers and Reiner, 1993; and am-phibians: Ciani et al., 1988; Marın et al., 1997).

Functional implications

As a general rule for the CNS of the zebrafish, all nucleithat displayed ChAT-ir cells showed AChE-positive neu-rons (except the ganglion cell layer of the retina) and mostof the regions with AChE-positive cells, but withoutChAT-ir ones, presented a more or less ChAT-ir rich in-nervation, in accordance with the cholinergic or cholino-ceptive nature of the AChE-positive cells.

Several concrete areas did not show these parallelismsin AChE/ChAT expression patterns. The hypothalamushad a profuse ChAT-ir innervation, whereas AChE activ-ity was not present, which is in contrast to the classicarrangement in ACh neurotransmission. However, it hasbeen reported that AChE is not necessarily present in thetarget cell of a cholinergic synapse. As an example, over90% of the cells in the rat striatum do not contain AChE,but they receive a strong cholinergic input from the sub-stantia nigra (Henderson, 1981; Bolam et al., 1984).

Finally, AChE-containing neurons were present in re-gions that did not display any cholinergic innervation (i.e.,cerebellum, olfactory bulb, interpeduncular nucleus). Ithas been reported that neurons containing large amountsof AChE can be cholinoceptive, although identifiable cho-linergic afferents are not present in the vicinity (Butcherand Woolf, 1982). The possibility that AChE could haveanother function, distinct from the hydrolysis of ACh,cannot be ruled out. Thus, a carboxy- and amino-peptidaseactivity toward substance P and enkephalins (Chubb etal., 1980), or a neuromodulatory activity similar to thatexerted by dopamine (Hefti and Lichtensteiger, 1978; Lli-nas et al., 1984), have been attributed to AChE in differentregions of the CNS.

CONCLUSIONS

The distribution of cholinergic cells in the zebrafishCNS shows some similarities and differences in compari-son to other teleosts and other vertebrates. In some re-gions, these variations are present even among fish be-longing to the cyprinid family. Thus, we described fourgroups of regions. The first group, composed of cranialmotor nuclei, isthmic area, and spinal cord, is cholinergicin all vertebrates studied to date, including zebrafish. Asecond group is comprised by regions, such as the retina,the olfactory bulb, the dorsal telencephalon, the cerebel-lum, and the tegmentum, where the pattern of distribu-tion of cholinergic cells is the same as in most vertebrategroups. There are other regions such as the optic tectum,the hypothalamus, and the stato-acoustic system wheretheir cholinergic characteristics are well conserved amongfish but are completely divergent from most groups ofvertebrates. Finally, specific diencephalic regions, such asthe pretectum, the thalamus, the epithalamus, and the

posterior tuberculum, show a highly divergent cholinergicorganization among all vertebrate species. These findingssuggest that, together with many features that are com-mon to most vertebrates, and even to most fish, the ze-brafish cholinergic system has evolved several character-istics that are unique to this species.

ACKNOWLEDGMENTS

The authors thank Miss M.T. Sanchez for her technicalassistance and Mr. G.H. Jenkins for his help with theEnglish version of the article.

REFERENCES

Adrio F, Anadon R, Rodrıguez-Moldes I. 2000. Distribution of cholineacetyltransferase (ChAT) immunoreactivity in the central nervous sys-tem of a chondrostean, the Siberian sturgeon (Acipenser baeri). J CompNeurol 426:602–621.

Airhart MJ, Shirt JO, Kriebel RM. 1988. Telencephalic projections to thegoldfish hypothalamus: an anterograde degeneration study. Brain ResBull 20:503–514.

Alonso JR, Amaral DG. 1995. Cholinergic innervation of the primate hip-pocampal formation. I. Distribution of choline acetyltransferase immu-noreactivity in the Macaca fascicularis and Macaca mulatta monkeys.J Comp Neurol 355:135–170.

Anadon R, Molist P, Rodrıguez-Moldes I, Lopez JM, Quintela I, CervinoMC, Barja P, Gonzalez A. 2000. Distribution of choline acetyltrans-ferase immunoreactivity in the brain of an elasmobranch, the lesserspotted dogfish (Scyliorhinus canicula). J Comp Neurol 420:139 –170.

Andre C, Guillet JG, De-Backer JP, Vanderheyden P, Hoebeke J, StrosbergAD. 1984. Monoclonal antibodies against the native or denatured formsof muscarinic acetylcholine receptors. EMBO J 3:17–21.

Anglade I, Zandbergen T, Kah O. 1993. Origin of the pituitary innervationin the goldfish. Cell Tissue Res 273:345–355.

Appleyard M, Jahnsen H. 1992. Actions of acetylcholinesterase in theguinea-pig cerebellar cortex in vitro. Neuroscience 47:291–301.

Armstrong DM, Saper CB, Levey AI, Wainer BH, Terry RD. 1983. Distri-bution of cholinergic neurons in rat brain: demonstrated by the immu-nocytochemical localization of choline acetyltransferase. J Comp Neu-rol 216:53–68.

Bagnoli P, Fontanesi G, Alesci R, Erichsen JT. 1992. Distribution of neu-ropeptide Y, substance P, and choline acetyltransferase in the devel-oping visual system of the pigeon and effects of unilateral retinalremoval. J Comp Neurol 318:392–414.

Barmack NH, Baughman RW, Eckenstein FP. 1992. Cholinergic innerva-tion of the cerebellum of rat, rabbit, cat, and monkey as revealed bycholine acetyltransferase activity and immunohistochemistry. J CompNeurol 317:233–249.

Beninato M, Spencer RF. 1986. A cholinergic projection to the rat superiorcolliculus demonstrated by retrograde transport of horseradish perox-idase and choline acetyltransferase immunohistochemistry. J CompNeurol 253:525–538.

Blaker SN, Armstrong DM, Gage FH. 1988. Cholinergic neurons within therat hippocampus: response to fimbria-fornix transection. J Comp Neu-rol 272:127–138.

Bolam JP, Ingham CA, Smith AD. 1984. The section-Golgi-impregnationprocedure-3. Combination of Golgi-impregnation with enzyme histo-chemistry and electron microscopy to characterize acetylcho-linesterase-containing neurons in the rat neostriatum. Neuroscience12:687–709.

Bradford HF. 1986. Chemical neurobiology. New York: Freeman.Brantley RK, Bass AH. 1988. Cholinergic neurons in the brain of a teleost

fish (Porichthys notatus) located with a monoclonal antibody to cholineacetyltransferase. J Comp Neurol 275:87–105.

Brauth SE, Kitt CA, Price DL, Wainer BH. 1985. Cholinergic neurons in thetelencephalon of the reptile Caiman crocodilus. Neurosci Lett 58:235–240.

Butcher LL. 1983. Acetylcholinesterase histochemistry. In: Bjorklund A,Hokfelt T, editors. Methods in chemical neuroanatomy. Handbook forchemical neuroanatomy. Vol. 1. Amsterdam: Elsevier. p 1–49.

Butcher LL, Woolf NJ. 1982. Monoaminergic-cholinergic relationships and


the chemical communication matrix of the substantia nigra and neos-triatum. Brain Res Bull 9:475–492.

Carpenter MB, Chang L, Pereira AB, Hersh LB. 1987. Comparisons of theimmunocytochemical location of choline acetyltransferase in the ves-tibular nuclei of the monkey and rat. Brain Res 418:403–408.

Carpenter MB, Huang Y, Pereira AB, Hersh LB. 1990. Immunocytochem-ical features of the vestibular nuclei in the monkey and cat. J Hirnfor-sch 31:585–599.

Chubb IW, Hodgson AJ, White GH. 1980. Acetylcholinesterase hydrolyzessubstance P. Neuroscience 5:2065–2072.

Ciani F, Franceschini V, Del Grande P. 1988. Histochemical and biochem-ical study of the acetylcholinesterase and choline actyltransferase inthe brain and spinal cord of the frog, Rana esculenta. J Hirnforsch29:157–163.

Contestabile A. 1975. Histochemical study on the distribution of someenzyme activities in the vagal and facial lobes of the goldfish, Carassiusauratus. Histochemistry 44:123–132.

Contestabile A, Tabanelli G. 1977. Cholinesterase patterns in the cerebel-lum of reptiles. Histochemistry 51:195–200.

Contestabile A, Zannoni N. 1975. Histochemical location of acetylcholines-terase in the cerebellum and optic tectum of four freshwater teleosts.Histochemistry 45:279–288.

Contestabile A, Villani L, Ciani F. 1977. Ultrastructural analysis on ace-tylcholinesterase localization in the cerebellar cortex of teleosts. AnatEmbryol (Berl) 152:15–27.

Contestabile A, Villani L, Fasolo A, Franzoni MF, Gribaudo L, OktedalenO, Fonnum F. 1987. Topography of cholinergic and substance P path-ways in the habenulo-interpeduncular system of the rat. An immuno-cytochemical and microchemical approach. Neuroscience 21:253–270.

Crespo C, Brinon JG, Porteros A, Arevalo R, Rico B, Aijon J, Alonso JR.1999. Distribution of acetylcholinesterase and choline acetyltrans-ferase in the main and accessory olfactory bulbs of the hedgehog (Eri-naceus europaeus). J Comp Neurol 403:53–67.

Danielson PD, Zottoli SJ, Corrodi JG, Rhodes KJ, Mufson EJ. 1988. Loca-tion of choline acetyltransferase to somata of posterior lateral lineefferents in the goldfish. Brain Res 448:158–161.

Day JW, Hall DH, Hall LM, Bennett MV. 1983. alpha-Bungarotoxin label-ing and acetylcholinesterase location at the Mauthner fiber giant syn-apse in the hatchetfish. J Neurosci 3:272–279.

Deng P, Cuenca N, Doerr T, Pow DV, Miller R, Kolb H. 2001. Location ofneurotransmitters and calcium binding proteins to neurons ofsalamander and mudpuppy retinas. Vision Res 41:1771–1783.

Desan PH, Gruberg ER, Grewell KM, Eckenstein F. 1987. Cholinergicinnervation of the optic tectum in the frog Rana pipiens. Brain Res413:344–349.

Deutsch JA. 1971. The cholinergic synapse and the site of memory. Science174:788–794.

Eaton RC, Bombardieri RA. 1978. Behavioral functions of the Mauthnerneuron. In: Faber DS, Triller A, editors. Neurobiology of the Mauthnercells. New York: Raven Press. p 221–244.

Echteler SM. 1984. Connections of the auditory midbrain in a teleost fish,Cyprinus carpio. J Comp Neurol 230:536–551.

Eckenstein FP, Thoenen H. 1983. Cholinergic neurons in the rat cerebralcortex demonstrated by immunohistochemical location of cholineacetyltransferase. Neurosci Lett 36:211–215.

Ekstrom P. 1987. Distribution of choline acetyltransferase-immuno-reactive neurons in the brain of a cyprinid teleost (Phoxinus phoxinusL.). J Comp Neurol 256:494–515.

Ekstrom P, Korf HW. 1986. Putative cholinergic elements in the photosen-sory pineal organ and retina of a teleost, Phoxinus phoxinus L. (Cyp-rinidae). Distribution of choline acetyltransferase immunoreactivity,acetylcholinesterase-positive elements and pinealofugally projectingneurons. Cell Tissue Res 246:321–329.

Famiglietti EV. 1983. “Starburst” amacrine cells and cholinergic neurons:mirror-symmetric on and off amacrine cells of rabbit retina. Brain Res261:138–144.

Famiglietti EV. 1992. Dendritic co-stratification of ON and ON-OFF direc-tionally selective ganglion cells with starburst amacrine cells in rabbitretina. J Comp Neurol 324:322–335.

Fetcho J. 1991. Spinal network of the Mauthner cell. Brain Behav Evol37:298–316.

Finger TE, Kanwal JS. 1992. Ascending general visceral pathways withinthe brainstem of two teleost fishes: Ictalurus punctatus and Carassiusauratus. J Comp Neurol 320:509–520.

Fite KV, Wang SR. 1986. Microiontophoresis and single-unit analysis ofcholinergic drugs in the optic tectum of frog. Brain Behav Evol 28:198–206.

Geffard M, McRae-Degueurce A, Souan ML. 1985. Immunocytochemicaldetection of acetylcholine in the rat central nervous system. Science229:77–79.

Geneser-Jensen FA, Blackstad TW. 1971. Distribution of acetyl cholines-terase in the hippocampal region of the guinea pig. I. Entorhinal area,parasubiculum, and presubiculum. Z Zellforsch Mikrosk Anat 114:460–481.

Geula C, Schatz CR, Mesulam M-M. 1993. Differential localization ofNADPH-diaphorase and calbidin-D28k within the cholinergic neuronsof the basal forebrain, striatum, and brainstem in the rat, monkey,baboon and human. Neuroscience 54:461–476.

Gould E, Butcher LL. 1987. Transient expression of cholineacetyltransferase-like immunoreactivity in Purkinje cells of the devel-oping rat cerebellum. Brain Res 431:303–306.

Grosman DC, Lorenzi MV, Trinidad AC, Strauss WL. 1995. The humancholine acetyltransferase gene encodes two proteins. J Neurochem65:484–491.

Grover BG, Sharma SC. 1981. Organization of extrinsic tectal connectionsin the goldfish (Carassius auratus). J Comp Neurol 196:471–488.

Hagan JJ, Morris RGM. 1988. The cholinergic hypothesis of memory: areview of animal experiments. In: Iversen LL, Iversen SD, editors.Handbook of psychopharmacology. New York: Plenum Press. p 237–323.

Hanneman E, Westerfield M. 1989. Early expression of acetylcholinester-ase activity in functionally distinct neurons of the zebrafish. J CompNeurol 284:350–361.

Hanneman E, Trevarrow B, Metcalfe WK, Kimmel CB, Westerfield M.1988. Segmental pattern of development of the hindbrain and spinalcord of the zebrafish embryo. Development 103:49–58.

Hasselmo ME, Bower JM. 1993. Acetylcholine and memory. Trends Neu-rosci 16:218–222.

Hedreen JC, Bacon SJ, Price DL. 1985. A modified histochemical techniqueto visualize acetylcholinesterase-containing axons. J Histochem Cyto-chem 33:134–140.

Hefti F, Lichtensteiger W. 1978. Subcellular distribution of dopamine insubstantia nigra of the rat brain: effects of gamma-butyrolactone anddestruction of noradrenergic afferents suggest formation of particlesfrom dendrites. J Neurochem 30:1217–1230.

Henderson Z. 1981. Ultrastructure and acetylcholinesterase content ofneurones forming connections between the striatum and substantianigra of rat. J Comp Neurol 197:185–196.

Hendry SH, Jones EG, Killackey HP, Chalupa LM. 1987. Cholineacetyltransferase-immunoreactive neurons in fetal monkey cerebralcortex. Brain Res 465:313–317.

Henley JM, Lindstrom JM, Oswald RE. 1986. Acetylcholine receptor syn-thesis in retina and transport to optic tectum in goldfish. Science232:1627–1629.

Holmqvist B, Ellingsen B, Alm P, Forsell J, Øyan A-M, Goksøyr A, Fjose A,Seo H-C. 2000. Identification and distribution of nitric oxide synthasein the brain of adult zebrafish. Neurosci Lett 292:119–122.

Hoogland PV, Vermeulen-Van der Zee E. 1990. Distribution of cholineacetyltransferase immunoreactivity in the telencephalon of the lizardGekko gekko. Brain Behav Evol 36:378–390.

Houser CR, Crawford GD, Barber RP, Salvaterra PM, Vaughn JE. 1983.Organization and morphological characteristics of cholinergic neurons:an immunocytochemical study with a monoclonal antibody to cholineacetyltransferase. Brain Res 266:97–119.

Ichikawa T, Ajiki K, Matsuura J, Misawa H. 1997. Location of two cholin-ergic markers, choline acetyltransferase and vesicular acetylcholinetransporter in the central nervous system of the rat: in situ hybridiza-tion histochemistry and immunohistochemistry. J Chem Neuroanat13:23–39.

Ikeda M, Houtani T, Ueyama T, Sugimoto T. 1991. Choline acetyltrans-ferase immunoreactivity in the cat cerebellum. Neuroscience 45:671–690.

Jones BE. 1993. The organization of central cholinergic systems and theirfunctional importance in sleep-waking states. Prog Brain Res 98:61–71.

Jones BE, Beaudet A. 1987. Distribution of acetylcholine and catechol-amine neurons in the cat brainstem: a choline acetyltransferase andtyrosine hydroxylase immunohistochemical study. J Comp Neurol 261:15–32.


Kanwal JS, Finger TE, Carpio J. 1988. Forebrain connections of the gus-tatory system in ictalurid catfishes. J Comp Neurol 278:353–376.

Kasa P. 1986. The cholinergic systems in brain and spinal cord. ProgNeurobiol 26:211–272.

Kimura H, McGeer PL, Peng J-H, McGeer EG. 1981. The central cholin-ergic system studied by choline acetyltransferase immunohistochemis-try in the cat. J Comp Neurol 200:151–201.

Kimura H, McGeer PL, Peng J-H, McGeer EG. 1984. Cholineacetyltransferase-containing neurons in the rat brain. In: Bjorklund A,Hokfelt T, Kuhar MJ, editors. Handbook of chemical neuroanatomy.Amsterdam: Elsevier. p 51–67.

King WM, Schmidt JT. 1991. The long latency component of retinotectaltransmission: enhancement by stimulation of isthmic nucleus or tecto-bulbar tract and block by nicotinic cholinergic antagonists. Neuro-science 40:701–712.

Kusunoki T, Tsuda Y, Takashima F. 1973. The chemoarchitectonics of theshark brain. J Hirnforsch 14:13–26.

Kusunoki T, Kishida R, Kadota T, Goris RC. 1987. Chemoarchitectonics ofthe brainstem in infrared sensitive and nonsensitive snakes. J Hirn-forsch 28:27–43.

Lamb CF, Caprio J. 1993. Diencephalic gustatory connections in the chan-nel catfish. J Comp Neurol 337:400–418.

Lamb CF, Finger TE. 1996. Axonal projections pattern of neurons in thesecondary gustatory nucleus of channel fish. J Comp Neurol 365:585–593.

Le Jeune H, Jourdan F. 1994. Acetylcholinesterase-containing intrinsicneurons in the rat main olfactory bulb: cytological and neurochemicalfeatures. Eur J Neurosci 6:1432–1444.

Levey AI, Wainer BH, Rye DB, Mufson EJ, Mesulam M-M. 1984. Cholineacetyltransferase-immunoreactive neurons intrinsic to rodent cortexand distinction from acetylcholinesterase-positive neurons. Neuro-science 13:341–353.

Levin ED, Simon BB. 1998. Nicotinic acetylcholine involvement in cogni-tive function in animals. Psychopharmacology (Berl) 138:217–230.

Llinas R, Greenfield SA, Jahnsen H. 1984. Electrophysiology of partcom-pacta cells in the in vitro substantia nigra-a possible mechanism fordendritic release. Brain Res 294:127–132.

Luiten PGM. 1981. Afferent and efferent connections of the optic tectum inthe carp (Cyprinus carpio L.). Brain Res 220:51–65.

Ma PM. 1994a. Catecholaminergic systems in the zebrafish. I. Number,morphology, and histochemical characteristics of neurons in the locuscoeruleus. J Comp Neurol 344:242–255.

Ma PM. 1994b. Catecholaminergic systems in the zebrafish. II. Projectionpathways and pattern of termination of the locus coeruleus. J CompNeurol 344:256–269.

Ma PM. 1997. Catecholaminergic systems in the zebrafish. III. Organiza-tion and projection pattern of medullary dopaminergic and noradren-ergic neurons. J Comp Neurol 381:411–427.

Maley BE, Frick ML, Levey AI, Wainer BH, Elde RP. 1988. Immunohis-tochemistry of choline acetyltransferase in the guinea pig brain. Neu-rosci Lett 84:137–142.

Marın O, Gonzalez A. 1999. Origin of tectal cholinergic projections inamphibians: a combined study of choline acetyltransferase immunohis-tochemistry and retrograde transport of dextran amines. Vis Neurosci16:271–283.

Marın O, Smeets WJAJ, Gonzalez A. 1997. Distribution of choline acetyl-transferase immunoreactivity in the brain of anuran (Rana perezi,Xenopus laevis) and urodele (Pleurodeles waltl) amphibians. J CompNeurol 382:499–534.

McCormick CA, Braford MR Jr. 1994. Organization of inner ear endorganprojections in the goldfish, Carassius auratus. Brain Behav Evol 43:189–205.

McCormick CA, Hernandez DV. 1996. Connections of the octaval andlateral line nuclei of the medulla in the goldfish, including the cytoar-chitecture of the secondary octaval population in goldfish and catfish.Brain Behav Evol 47:113–137.

McKenna TM, Ashe JH, Weinberger NM. 1989. Cholinergic modulation offrequency receptive fields in auditory cortex: I. Frequency-specific ef-fects of muscarinic agonists. Synapse 4:30–43.

Medina L, Reiner A. 1994. Distribution of choline acetyltransferase immu-noreactivity in the pigeon brain. J Comp Neurol 342:497–537.

Medina L, Smeets WJAJ, Hoogland PV, Puelles L. 1993. Distribution ofcholine acetyltransferase immunoreactivity in the brain of the lizardGallotia galloti. J Comp Neurol 331:261–285.

Meek J, Nieuwenhuys R. 1998. Holosteans and teleosts. In: NieuwenhuysR, Ten Donkelaar HJ, Nicholson C, editors. The central nervous systemof vertebrates. Berlin: Springer. p 759–937.

Meek J, Schellart NA. 1978. A Golgi study of goldfish optic tectum. J CompNeurol 182:89–122.

Mesulam M-M, Geula C. 1988. Nucleus basalis (Ch4) and cortical cholin-ergic innervation in the human brain: observations based on the dis-tribution of acetylcholinesterase and choline acetyltransferase. J CompNeurol 275:216–240.

Mesulam M-M, Mufson EJ, Levey AI, Wainer BH. 1984. Atlas of cholin-ergic neurons in the forebrain and upper brainstem of the macaquebased on monoclonal choline acetyltransferase immunohistochemistryand acetylcholinesterase histochemistry. Neuroscience 12:669–686.

Mitrofanis J, Stone J. 1988. Distribution of cholinergic amacrine cells inthe retinas of normally pigmented and hypopigmented strains of ratand cat. Vis Neurosci 1:367–376.

Model PG, Spira ME, Bennett MV. 1972. Synaptic inputs to the cell bodiesof the giant fibers of the hatchetfish. Brain Res 45:288–295.

Molist P, Maslam S, Velzing E, Roberts BL. 1993. The organization ofcholinergic neurons in the mesencephalon of the eel, Anguilla anguilla,as determined by choline acetyltransferase immunohistochemistry andacetylcholinesterase enzyme histochemistry. Cell Tissue Res 271:555–566.

Morita Y, Ito H, Masai H. 1980. Central gustatory paths in the cruciancarp, Carassius carassius. J Comp Neurol 191:119–132.

Mufson EJ, Cunningham MG. 1988. Observations on choline acetyltrans-ferase containing structures in the CD-1 mouse brain. Neurosci Lett84:7–12.

Mufson EJ, Desan PH, Mesulam MM, Wainer BH, Levey AI. 1984. Cholineacetyltransferase-like immunoreactivity in the forebrain of the red-eared pond turtle (Pseudemys scripta elegans). Brain Res 323:103–108.

Mufson EJ, Martin TL, Mash DC, Wainer BH, Mesulam MM. 1986. Cho-linergic projections from the parabigeminal nucleus (Ch8) to the supe-rior colliculus in the mouse: a combined analysis of horseradish perox-idase transport and choline acetyltransferase immunohistochemistry.Brain Res 370:144–148.

Murakami T, Morita Y, Ito H. 1986. Cytoarchitecture and fiber connectionsof the superficial pretectum in a teleost, Navodon modestus. Brain Res373:213–221.

Murphy PC, Sillito AM. 1991. Cholinergic enhancement of direction selec-tivity in the visual cortex of the cat. Neuroscience 40:13–20.

Nickell WT, Shipley MT. 1988. Two anatomically specific classes of candi-date cholinoceptive neurons in the rat olfactory bulb. J Neurosci8:4482–4491.

Nieuwenhuys R, Pouwels E. 1983. The brainstem of actinopterygian fishes.In: Davis RE, Northcutt RG, editors. Fish neurobiology. Vol. 1. AnnArbor: University of Michigan Press. p 203–236.

Northcutt RG. 1982. Localization of neurons afferent to the optic tectum inthe long-nose gars. J Comp Neurol 204:325–335.

Northcutt RG, Butler AB. 1993. The diencephalon and optic tectum of thelong-nose gar, Lepisosteus osseus (L.): cytoarchitectonics and distribu-tion of acetylcholinesterase. Brain Behav Evol 41:57–81.

Oswald RE, Schmidt JT, Norden JJ, Freeman JA. 1980. Location of alpha-bungarotoxin binding sites to the goldfish retinotectal projection. BrainRes 187:113–127.

Parnavelas JG, Kelly W, Franke E, Eckenstein F. 1986. Cholinergic neu-rons and fibres in the rat visual cortex. J Neurocytol 15:329–336.

Perez SE, Yanez J, Marın O, Anadon R, Gonzalez A, Rodrıguez-Moldes I.2000. Distribution of choline acetyltransferase (ChAT) immunoreactiv-ity in the brain of the adult trout and tract-tracing observations on theconnections to the nuclei of the isthmus. J Comp Neurol 428:450–474.

Perry E, Walker M, Grace J, Perry R. 1999. Acetylcholine in mind: aneurotransmitter correlate of consciousness? Trends Neurosci 22:273–280.

Pfister C, Ritter J, Wenk H, Schwartz A, Meyer U. 1973. Enzyme histo-chemical studies of Mauthner cells of Salmo irideus (Gibbons, 1855) inbiomorphosis. Z Mikrosk Anat Forsch 87:128–144.

Pombal MA, Yanez J, Marın O, Gonzalez A, Anadon R. 1999. Cholinergicand GABAergic neuronal elements in the pineal organ of lampreys, andtract-tracing observations of differential connections of pinealofugalneurons. Cell Tissue Res 295:215–223.

Pombal MA, Marın O, Gonzalez A. 2001. Distribution of cholineacetyltransferase-immunoreactive structures in the lamprey brain.J Comp Neurol 431:105–126.


Powers AS, Reiner A. 1993. The distribution of cholinergic neurons in thecentral nervous system of turtles. Brain Behav Evol 41:326–345.

Reiner A. 1991. A comparison of neurotransmitter-specific andneuropeptide-specific neuronal cell types present in the dorsal cortex inturtles with those present in the isocortex in mammals: implicationsfor the evolution of isocortex. Brain Behav Evol 38:53–91.

Rhodes KJ, Zottoli SJ, Mufson EJ. 1986. Choline acetyltransferase immu-nohistochemical staining in the goldfish (Carassius auratus) brain:evidence that the Mauthner cell does not contain choline acetyltrans-ferase. Brain Res 381:215–224.

Ricciuti AJ, Gruberg ER. 1985. Isthmic nucleus provides most tectal cho-line acetyltransferase in the frog Rana pipiens. Brain Res 341:399–402.

Rico B, Cavada C. 1998. A population of cholinergic neurons is present inthe macaque monkey thalamus. Eur J Neurosci 10:2346–2352.

Rink E, Wullimann MF. 1996. Forebrain gustatory pathways in the gold-fish, Carassius auratus. In: Elsner N, Schnitzler H-U, editors. Gottin-gen neurobiology report 1996: proceedings of the 24th Gottingen Neu-robiology Conference 1996. Vol. II. New York: Georg Thieme VerlagStuttgar. p 287.

Roberts BL, Maslam S, Los I, Van der Jagt B. 1994. Coexistence of calci-tonin gene-related peptide and choline acetyltransferase in eel efferentneurons. Hear Res 74:231–237.

Robertson LT, Roman N. 1989. Distribution of acetylcholinesterase in thegranule layer of the cerebellum of the rhesus monkey (Macaca mu-latta). Brain Behav Evol 34:342–350.

Rodieck RW, Marshak DW. 1992. Spatial density and distribution of cho-line acetyltransferase immunoreactive cells in human, macaque, andbaboon retinas. J Comp Neurol 321:46–64.

Ross CD, Godfrey DA. 1986. Effect of enucleation on choline acetyltrans-ferase activity in layers of goldfish optic tectum. Brain Res 373:49–56.

Ross LS, Parrett T, Easter SS Jr. 1992. Axonogenesis and morphogenesisin the embryonic zebrafish brain. J Neurosci 12:467–482.

Ruggiero DA, Giuliano R, Anwar M, Stornetta R, Reis DJ. 1990. Anatom-ical substrates of cholinergic-autonomic regulation in the rat. J CompNeurol 292:1–53.

Sakamoto N, Ito H. 1982. Fiber connections of the corpus glomerulosus ina teleost, Navodon modestus. J Comp Neurol 205:291–298.

Satoh K, Fibiger HC. 1985. Distribution of central cholinergic neurons inthe baboon (Papio papio). I. General morphology. J Comp Neurol 236:197–214.

Schmidt M, Humphrey MF, Wassle H. 1987. Action and location of acetyl-choline in the cat retina. J Neurophysiol 58:997–1015.

Shiromani PJ, Armstrong DM, Bruce G, Hersh LB, Groves PM, Gillin JC.1987. Relation of pontine choline acetyltransferase immunoreactiveneurons with cells which increase discharge during REM sleep. BrainRes Bull 18:447–455.

Silver A. 1974. The biology of cholinesterases. Amsterdam: North Holland.p 355–388.

Siripholvat V, Watanabe T, Tomita T. 1988. Histochemical and cytochem-ical localization of acetylcholinesterase in the cerebellar cortex of thechicken. Jikken Dobutsu 37:437–446.

Sitaram N, Moore AM, Gillin JC. 1978. Experimental acceleration andslowing of REM sleep ultradian rhythm by cholinergic agonist andantagonist. Nature 274:490–492.

Somogyi P, Eshhar N, Teichberg VI, Roberts JDB. 1990. Subcellular local-ization of a putative kainate receptor in Bergmann glial cells using amonoclonal antibody in the chick and fish cerebellar cortex. Neuro-science 35:9–30.

Sorenson EM, Parkinson D, Dahl JL, Chiappinelli VA. 1989. Immunohis-tochemical location of choline acetyltransferase in the chicken mesen-cephalon. J Comp Neurol 281:641–657.

Spira M, Model PG, Bennett MVL. 1970. Cholinergic transmission at acentral synapse. J Cell Biol 47:199–200.

St-Jacques R, Gorczyca W, Mohr G, Schipper H. 1996. Mapping of the basalforebrain cholinergic system of the dog: a choline acetyltransferaseimmunohistochemical study. J Comp Neurol 366:717–725.

Striedter GF. 1990. The diencephalon of the channel catfish, Ictaluruspunctatus. II. Retinal, tectal, cerebellar and telencephalic connections.Brain Behav Evol 36:355–377.

Striedter GF, Northcutt RG. 1989. Two distinct visual pathways throughthe superficial pretectum in a percomorph teleost. J Comp Neurol283:342–354.

Tago H, McGeer PL, Bruce G, Hersh LB. 1987. Distribution of choline

acetyltransferase-containing neurons of the hypothalamus. Brain Res415:49–62.

Tago H, McGeer PL, McGeer EG, Akiyama H, Hersh LB. 1989. Distribu-tion of choline acetyltransferase immunopositive structures in the ratbrainstem. Brain Res 495:271–297.

Torres B, Pastor AM, Cabrera B, Salas C, Delgado-Garcıa JM. 1992.Afferents to the oculomotor nucleus in the goldfish (Carassius auratus)as revealed by retrograde labeling with horseradish peroxidase.J Comp Neurol 324:449–461.

Torres B, Fernandez S, Rodrıguez F, Salas C. 1995. Distribution of neuronsprojecting to the trochlear nucleus in goldfish (Carassius auratus).Brain Behav Evol 45:272–285.

Tumosa N, Stell WK. 1986. Choline acetyltransferase immunoreactivitysuggests that ganglion cells in the goldfish retina are not cholinergic.J Comp Neurol 244:267–275.

Tumosa N, Eckenstein F, Stell WK. 1984. Immunocytochemical location ofputative cholinergic neurons in the goldfish retina. Neurosci Lett 48:255–259.

Tumosa N, Stell WK, Johnson CD, Epstein ML. 1986. Putative cholinergicinterneurons in the optic tectum of goldfish. Brain Res 370:365–369.

Van der Zee EA, Luiten PG. 1999. Muscarinic acetylcholine receptors inthe hippocampus, neocortex and amygdala: a review of immunocyto-chemical location in relation to learning and memory. Prog Neurobiol58:409–471.

Vanegas H, Ito H. 1983. Morphological aspects of the teleostean visualsystem: a review. Brain Res 287:117–137.

Villani L, Ciani F, Contestabile A. 1977. Ultrastructural pattern of acetyl-cholinesterase distribution in the cerebellar cortex of the quail. AnatEmbryol (Berl) 152:29–41.

Villani L, Ciani F, Contestabile A. 1979. Electron microscope histochem-istry of acetylcholinesterase distribution in the optic tectum of teleosts.J Hirnforsch 20:539–551.

Villani L, Battistini S, Bissoli R, Contestabile A. 1987. Cholinergic projec-tions in the telencephalo-habenulo-interpeduncular system of the gold-fish. Neurosci Lett 76:263–268.

Villani L, Dipietrangelo L, Pallotti C, Pettazzoni P, Zironi I, Guarnieri T.1994. Ultrastructural and immunohistochemical study of thetelencephalo-habenulo-interpeduncular connections of the goldfish.Brain Res Bull 34:1–5.

Villani L, Zironi I, Guarnieri T. 1996. Telencephalo-habenulo-interpeduncular connections in the goldfish: a DiI study. Brain BehavEvol 48:205–212.

Vincent SR, Reiner PB. 1987. The immunohistochemical location of cholineacetyltransferase in the cat brain. Brain Res Bull 18:371–415.

Wenthold RJ, Altschuler RA, Hampson DR. 1990. Immunocytochemistry ofneurotransmitter receptors. J Electron Microsc Tech 15:81–96.

Westerfield M, McMurray JM, Eisen JS. 1986. Identified motoneurons andtheir innervation of axial muscles in the zebrafish. J Neurosci 6:2267–2277.

Wilson SW, Ross LS, Parrett T, Easter SS. 1990. The development of asimple scaffold of axon tracts in the brain of the embryonic zebrafish,Brachydanio rerio. Development 108:121–145.

Woolf NJ. 1991. Cholinergic systems in mammalian brain and spinal cord.Prog Neurobiol 37:475–524.

Wullimann MF. 1988. The tertiary gustatory center in sunfishes is notnucleus glomerulosus. Neurosci Lett 86:6–10.

Wullimann MF, Meyer DL. 1990. Phylogeny of putative cholinergic visualpathways through the pretectum to the hypothalamus in teleost fish.Brain Behav Evol 36:14–29.

Wullimann MF, Northcutt RG. 1988. Connections of the corpus cerebelli inthe green sunfish and the common goldfish: a comparison of perciformand cypriniform teleosts. Brain Behav Evol 32293–32316.

Wullimann MF, Roth G. 1992. Is the nucleus corticalis of teleosts a newcholinergic central nervous system for vertebrates? Neuroreport 3:33–35.

Wullimann MF, Meyer DL, Northcutt RG. 1991. The visually relatedposterior pretectal nucleus in the non-percomorph teleost Osteoglos-sum bicirrhosum projects to the hypothalamus: a DiI study. J CompNeurol 312:415–435.

Wullimann MF, Rupp B, Reichert H. 1996. Neuroanatomy of the zebrafishbrain. A topological atlas. Berlin: Birkhauser Verlag.

Xue HG, Yamamoto N, Yoshimoto M, Yang CY, Ito H. 2001. Fiber connec-tions of the isthmic nucleus in the carp (Cyprinus carpio) and tilapia(Oreochromis niloticus). Brain Behav Evol 58:185–204.


Yanez J, Pombal MA, Anadon R. 1999. Afferent and efferent connections ofthe parapineal organ in lamprey: a tract tracing and immunocytochem-ical study. J Comp Neurol 403:171–189.

Yoshida K, Watanabe D, Ishikane H, Tachibana M, Pastan I, Nakanishi S.2001. A key role of starburst amacrine cells in originating retinaldirectional selectivity and optokinetic eye movement. Neuron 30:771–780.

Yoshimoto M, Albert JS, Sawai N, Shimuzu M, Yamamoto N, Ito H. 1998.Telencephalic ascending gustatory system in a cychlid fish, Oreochro-mis (Tilapia) niloticus. J Comp Neurol 392:209–226.

Zhang J, Wu SM. 2001. Immunocytochemical analysis of cholinergic am-acrine cells in the tiger salamander retina. Neuroreport 12:1371–1375.

Zottoli SJ, Rhodes KJ, Mufson EJ. 1985. Choline acetyltransferase immu-nohistochemical evidence for cholinergic projections from the nucleiisthmi and reticularis mesencephali to the optic tectum in the goldfish,Carassius auratus. Biol Bull 169:558.

Zottoli SJ, Rhodes KJ, Mufson EJ. 1987. Comparison of acetylcholinester-ase and choline acetyltransferase staining patterns in the optic tectumof the goldfish Carassius auratus. A histochemical and immunocyto-chemical analysis. Brain Behav Evol 30:143–159.

Zottoli SJ, Rhodes KJ, Corrodi JG, Mufson EJ. 1988. Putative cholinergicprojections from the isthmic nucleus and the nucleus reticularis mes-encephali to the optic tectum in the goldfish (Carassius auratus).J Comp Neurol 273:385–398.


Cholinergic Elements in the Zebraﬁsh Central Nervous …...tem in the vertebrate central nervous...

Documents

Transcript of Cholinergic Elements in the Zebraﬁsh Central Nervous …...tem in the vertebrate central nervous...