Centrifugal Fibers Research Articles

Survival of isthmo-optic neurons after early removal of one eye

In a series of normal 12- or 13-day-old chick embryos in which the neuroanatomical tracer wheat germ agglutinin conjugated to horseradish peroxidase (WGA-HRP) was injected into one eye, about 25,000 neurons were found to be retrogradely labeled in the isthmic region of the brain. Of these, about 100 were found on the side of the injection, whereas about 22,000 were found in the contralateral isthmo-optic nucleus (ION, the nucleus of origin of centrifugal fibers to the retina) and a further 3000 ectopic cells were present on this side but lay outside the borders of the ION. After comparable injections on the 18th day of incubation or later, no more than 40 retrogradely labeled neurons were seen on the side of the injection, and on the contralateral side the numbers of neurons in the ION and in the ectopic ION population were reduced by about 60% to approximately 9500 and 1500 cells, respectively. In the present study we have re-examined the effects of removing one optic vesicle or optic cup on the survival and distribution of isthmo-optic neurons in two further series of chick embryos: in the first series the animals were injected intraocularly with WGA-HRP on the 12th day of incubation; in the other the eye injections were done on day 18, 19 or 20. In animals in which the eye rudiment was removed on or before stage 13 of the Hamburger and Hamilton series 23, essentially all of the centrifugally projecting neurons on both sides of the brain could be labeled from the surviving eye at day 12; however, in chicks operated upon after stage 13 the distribution of the labeled ION neurons was indistinguishable from that in a normal 12 day embryo. In monocular chicks that were allowed to survive beyond the period of naturally occurring cell death in the ION (i.e. beyond day 17) a variety of labeling patterns was seen in the ION and the surrounding region. In nearly all of the animals in which the eye was removed after stage 13, the number of labeled neurons on the side contralateral to the surviving eye was comparable to that seen in a normal, unoperated embryo, whereas virtually no cells were labeled in the isthmic region of the opposite (ipsilateral) side. In animals in which the eye removal was done on or before stage 13 the ION and ectopic ION populations on both sides were frequently labeled, but in no case did the total number of labeled neurons on the two sides exceed the number found on one side in a normal chick. In other animals operated on at these stages, very few labeled cells were found on the side of the surviving eye, but the number on the contralateral side was again only about half that found in a normal animal. Analysis of the anterogradely labeled retinofugal projection in these brains suggests that there is a close correlation between the number of surviving isthmo-optic neurons on each side and the magnitude of the crossed and uncrossed retinal projections to the optic tectum. Taken together the present findings provide further evidence in support of the notion that during development the axons of the centrifugal neurons compete with each other within the retina, for a trophic agent that is available in only limited supply. In addition, they indicate that the availability of the trophic agent is not regulated by the magnitude of the centrifugal input to the retina, since even in animals in which the retina was innervated by twice the normal number of centrifugal fibers, the surviving eye was never able to support more than the normal number of isthmo-optic neurons. Several possible factors that may be involved in determining the laterality of the centrifugal projection in monocular chicks are discussed.

Octopamine release from centrifugal fibers of the Limulus peripheral visual system.

Octopamine, a biogenic amine, is synthesized and stored within centrifugal (efferent) fibers that project from the brain to the lateral and ventral eyes of the horseshoe crab, Limulus polyphemus. The experiments described here show that depolarization of Limulus lateral and ventral eyes, produced by elevating the concentration of extracellular K+, causes the selective release of newly synthesized octopamine from centrifugal fibers in a manner that requires the influx of extracellular Ca2+. Conjugates of octopamine and tyramine that are also stored within centrifugal fibers are not released in response to K+-induced depolarization. These findings add further support to the hypothesis that octopamine is a neurotransmitter synthesized by and released from centrifugal fibers in Limulus eyes. This amine may be responsible for many of the alterations in lateral eye structure and function that are mediated by centrifugal innervation.

Effect of local 6-OHDA and 5,6-DHT injections into the rat olfactory bulb on neophobia and learned aversion to a novel food

This study investigated the role of the noradrenergic (NA) and the serotoninergic (5-HT) centrifugal fibers on the olfactory bulb (OB). For this purpose, behavioral responses largely depending on olfactory information were measured in rats after discrete bilateral injections into the OB with either 6-hydroxydopamine (6-OHDA) or 5,6-dihydroxytryptamine (5,6-DHT) solutions. The evaluation of the level of endogenous 5-HT in the OB was measured by high performance liquid chromatography. The NA fibers appeared lesioned after the 6-OHDA treatment only, whereas the 5-HT level was significantly reduced in 5,6-DHT treated bulbs only (34% of the control value). The animals were fed on usual stock-diet (S) and had the choice between S and a novel food (V) on neophobia test, and again on the following day after aversive conditioning (0.9 mEq/kg LiCl i.p.). The behavioral results showed that the drug administrations were followed by a significant effect in one case only: the intake of V was significantly enhanced in 6-OHDA treated animals during neophobia test. Thus, neophobia was reduced by 6-OHDA. The data are interpreted as follows: (1) the lesion of the bulbar NA or 5-HT fibers does not induce total anosmia; and (2) the NA innervation of the OB could play a role in the modulation of behavioral responses to biological odorants.

Hypothalamo-retinal centrifugal projection in the dog

Horseradish peroxidase (HRP)-filled neurons and their processes were consistently detected in the ventral portion of the dog hypothalamus after intraocular injection of HRP. The number of HRP-filled neurons decreased in parallel with extent of the resection of the optic nerve. HRP-filled neurons were never detected in specimens with a complete resection of the optic nerve. These findings strongly indicate that these HRP-filled neurons in the ventral hypothalamus are the source of centrifugal fibers to the retina.

Olfactory bulb excitability selectively modified in behaving rats after local 6-hydroxydopamine treatment

Previous studies have shown that the olfactory bulb (OB) responsiveness was selectively enhanced towards food odor in free-moving hungry rats. This modulation of the olfactory input depended on the action of centrifugal fibers. The present study investigated the possible involvement of the noradrenergic (NA) projections from the locus coeruleus to the OB in this adaptative control of the OB excitability. After plugging of the olfactory lumen, 17 rats received discrete injections of 6-hydroxydopamine solution (10 μg of the salt; 40 nmol free base) into one OB (6-OHDA bulb) and the vehicle into the contralateral side (control bulb); 10 sham operated rats received intrabulbar isotonic saline solution injected bilaterally (sham bulbs). In the 12 6-OHDA bulbs with the noradrenaline content significantly reduced, the mean dopamine and noradrenaline levels were, respectively, 90% and 34% of the control values. One week after treatment, the average irritability score was significantly higher in the 6-OHDA treated animals than in the sham-operated ones. The main electrophysiological effect of the treatment was to amplify the selective enhancement of OB responsiveness to food odor normally occurring in hungry animals; this effect was restricted to the 6-OHDA bulbs. The results are discussed considering the possible involvement of NA fibers in selective attention to odors at the OB level.

The topographic organization of associational fibers of the olfactory system in the rat, including centrifugal fibers to the olfactory bulb.

This study analyzed the topographic organization of the associational fibers within the olfactory cortex of the rat, by using the autoradiographic method. Small injections of 3H-leucine were placed in all of the subdivisions of the olfactory cortex, to label selectively the fibers arising in each area. Intracortical fibers were identified from all of the olfactory cortical areas except the olfactory tubercle and were classified into two major systems (the layer Ib system and the layer II-deep Ib system) on the basis of their laminar pattern of termination (see Luskin and Price, '83). The layer Ib fiber system arises in the anterior olfactory nucleus, piriform cortex, and lateral entorhinal area, and is broadly organized in relation to the lateral olfactory tract. Cortical areas deep to or near the lateral olfactory tract are preferentially interconnected with areas near the tract, while parts of the cortex lateral and caudal to the lateral olfactory tract are most heavily interconnected with areas lateral, caudal, and medial to the tract. Commissural projections from the anterior olfactory nucleus and the anterior piriform cortex match some (but not all) components of the ipsilateral layer Ib fiber system. The layer II-deep Ib fiber system arises in three small areas--the ventral tenia tecta, the dorsal peduncular cortex, and the periamygdaloid cortex. The fibers from the ventral tenia tecta terminate in layer II of the anterior olfactory nucleus and are topographically organized. The fibers from the dorsal peduncular cortex and the periamygdaloid cortex are more widely distributed, especially in the lateral and caudal parts of the cortex. Two other intracortical projections do not fit into either of these fiber systems. The nucleus of the lateral olfactory tract projects bilaterally to the islands of Calleja and the medial edge of the anterior piriform cortex. The anterior cortical nucleus projects to many parts of the olfactory cortex, but the fibers end in both superficial and deep parts of layer I (layer Ia and Ib). There are projections from several of the olfactory cortical areas to the cortical areas surrounding the olfactory cortex. Virtually all of the olfactory areas also project to the ventral and dorsal endopiriform nuclei deep to the piriform cortex and/or to the polymorph zone deep to the olfactory tubercle. In addition, projections have been demonstrated to the deep amygdaloid nuclei, especially from the more ventromedial and caudal parts of the olfactory cortex.

Central neural influences on the human retina during selective attention

Two controversial issues were investigated. One concerned the question of whether centrifugal fibers project to the human retina, and the other, the question of whether precortical filters which differentially transmit relevant and irrelevant information during selective attention exist. The issues were addressed jointly by studying the effects of selective attention on evoked responses obtained from the retina and over the occipital cortex. Selective attention was manipulated by requiring subjects to respond to stimuli presented at various locations in the visual field while attempting to ignore stimuli presented at other locations. Retinal responses to stimuli presented at attended to locations were found to be larger than those at unattended to locations. A similar result was obtained for responses recorded over the occipital cortex. The results are interpreted as follows: (1) They support the hypothesis that sustained voluntary attention to specific locations in space is mediated, at least partly, by centrally controlled centrifugal pathways which differentially modify synaptic transmission at a precortical level; (2) such differential alteration can occur as far peripherally as the retina; and (3) such alteration at the retinal level necessarily requires the existence of centrifugal optic nerve fibers. The results are discussed within the context of existing evidence bearing on the issues.

The distribution of centrifugal terminals in the pigeon retina.

The terminals of the centrifugal fibres of the pigeon retina have been labelled by anterograde axonal transport or diffusion of horseradish peroxidase, and their distribution studied in whole amounts of the retina. Centrifugal terminal arborisations are concentrated in a band near the projection of the horizontal meridian on the retina. Within this band there are high density areas next to the area centralis and in the mid-temporal retina. The arborisations are largely absent from the red field and are nearly randomly arrayed. Their distribution resembles that of the displaced ganglion cells but not of other retinal neurons. We estimate that approximately 7,000 centrifugal terminal arborisations are present in each retina. Approximately two-thirds of these are convergent, with branches ending in large terminals clustered around a single cell in the amacrine sublayer; often penetrating up to 10 microns into the inner nuclear layer. The remainder are divergent endings which innervate a larger area of retina with widely spaced small terminals about 1 micron in diameter. Possible synaptic contacts between centrifugal terminals and displaced ganglion cells are occasionally seen; these contacts and the similarity in distribution suggest a link between the centrifugal and accessory optic systems.

Field potential and intracellular potential studies of the olfactory bulb in the carp: evidence for a functional separation of the olfactory bulb into lateral and medial subdivisions

1. Field potentials in the olfactory bulb and intracellular potentials from mitral cells were analyzed in the carp. Electrical shocks were applied to a part of the input (the lateral or medial bundle of the olfactory nerve: l-ON or m-ON respectively) or output pathways (lateral or medial olfactory tract: LOT or MOT respectively) of the olfactory bulb in order to activate the olfactory bulb partially. 2. When shocks were applied to the regions described above, the distributions of the C2-wave component (which reflects the synaptic depolarization of the peripheral dendrites of granule cells) of field potentials were different between the lateral and medial parts of the olfactory bulb. It was suggested that the patterns of excitatory synaptic inputs to peripheral dendrites of the granule-cell population are different between the two parts of the olfactory bulb. 3. The distributions of the C3- and C4-wave components (which reflect the synaptic depolarization of somata and deep dendrites of granule cells by volleys in centrifugal nerve fibers to the olfactory bulb) of field potentials were similar between the two parts of the olfactory bulb. It was suggested that the patterns of excitatory synaptic inputs from the centrifugal fibers to somata and deep dendrites of the granule-cell population are similar between the two parts of the bulb. 4. Mitral cells activated antidromically by LOT shocks or synaptically by l-ON shocks were located mainly in the lateral part of the olfactory bulb. On the other hand, mitral cells activated antidromically by MOT shocks or synaptically by m-ON shocks were located mainly in the medial part of the bulb. 5. Mitral cells, showing IPSPs in response to LOT, MOT and l-ON shocks, but not to m-ON shocks, were located mainly in the lateral part of the olfactory bulb. On the other hand, mitral cells, showing IPSPs to MOT and m-ON shocks, but not to LOT and l-ON shocks, were located mainly in the medial part of the bulb. Such a spatial distribution of mitral cells showing IPSPs was in accordance with that expected from the spatial distribution of the field potentials. 6. These results suggest that the olfactory bulb of the carp can be functionally separated into two subdivisions (the lateral and medial parts); the activities of neurons in the one part exert little influence on neurons in the other part. 7. From these results we suppose that the olfactory system of the carp is composed of two separate systems; in the lateral olfactory system, the lateral part of the olfactory bulb receives inputs mainly from the lateral bundle of the olfactory nerve and sends outputs to the LOT, while in the medial olfactory system, the medial part of the olfactory bulb receives inputs mainly from the medial bundle of the olfactory nerve and sends outputs to the MOT.

Further studies on the development of the isthmo-optic nucleus with special reference to the occurrence and fate of ectopic and ipsilaterally projecting neurons.

AbstractUsing wheat germ agglutinin‐conjugated horseradish peroxidase (WGA‐HRP) and the fluorescent dyes true blue and nuclear yellow, we have reex‐amined the time of arrival at the retina of the centrifugal fibers from the contralateral isthmo‐optic nucleus (ION), and have followed, quantitatively, the appearance and fate of other neurons that can be retrogradely labeled from the ipsilateral and contralateral eyes.The earliest age at which ION neurons can be retrogradely labeled is on the ninth day of incubation; the cells labeled at this stage are located in the ventrolateral part of the nucleus on the contralateral side. Over the course of the next 48 hours of development more and more cells can be labeled in a distinct ventrolateral to dorsomedial progression within the nucleus. Since this labeling sequence parallels the time course of generation of the ION neurons it is suggested that the axons of the first ION cells to be generated are the first to reach the contralateral retina, and that in this system there is a close relationship between the time that neurons withdraw from the cell cycle and the time that their axons reach their target area.In addition to the neurons in the ION, about 1,500 cells outside of the nucleus can be retrogradely labeled by WGA‐HRP injected into the contra‐lateral eye in posthatched chicks. This is slightly less than half the number of “ectopic isthmo‐optic neurons” that can be similarly labeled on the 13th day of incubation (when the ION is numerically complete). The reduction in the number of ectopic ION neurons occurs over the same period as the phase of naturally occurring cell death in the ION itself–between the 13th and 17th days of incubation. Long‐term labeling experiments with true blue indicate that the disappearance of about 53% of the ectopic ION cells is due to their death rather than the elimination of axon collaterals. In their morphology the ectopic neurons resemble the cells in the ION at early stages in their development, and there is indirect evidence that the ectopic ION neurons which survive also receive an input from the optic tectum through the‐tecto‐isthmal tract. On these and other grounds it is suggested that the ectopic neurons are indeed “misplaced” ION cells.Between the tenth and 13th days of incubation a small and rather variable number of neurons (58–178) was retrogradely labeled from the ipsilateral eye. Of these neurons with aberrant, ipsilaterally projecting axons, an average of just under 30 lay within the interior of the ION, about 33 were along its borders, and 38 were observed scattered among the ectopic ION cells. Double labeling experiments indicate that early in their development some of these neurons have axon collaterals which project to the contralateral eye. In a large number of animals injected with WGA‐HRP after the 17th day of incubation, only a single neuron was seen in the interior of the ipsilateral ION, but, on average, about nine labeled neurons were found along its borders and about 20 were ectopically placed.The numbers of ectopic ION neurons and ION cells with ipsilaterally projecting axons that we have observed are substantially higher than those reported in a previous study (Clarke and Cowan, '76) in which it was also suggested the both classes of cells were effectively eliminated between the 13th and 17th days of incubation. The observed differences are attributable to the greater sensitivity of the WGA‐HRP method when used with the chromogen tetramethyl benzidine. It is now clear that although cells in the ION with aberrantly projecting axons (or axon collaterals) may be selected against during the later stages of development, the fate of the ectopic ION neurons is not significantly different from that of the cells in the ION itself, but since they lie outside the nucleus their dendritic processes are not subject to the same morphogenetic influences.

Substance P-like immunoreactivity in the parietal eye visual system of the lizard Uta stansburiana.

We examined the parietal eye visual system of the iguanid lizard Uta stansburiana for the presence of substance P-like immunoreactivity by use of both immunofluorescence and peroxidase-antiperoxidase techniques. In the parietal eye no substance P-containing somata were found; however, its plexiform layer contained small (ca. 1 micron diam) immunoreactive fibers. These fibers apparently originate outside the parietal eye. Immunoreactive fibers also were found in the parietal nerve, the dorsal sac, and the leptomeninx of the pineal gland. No labeled somata were observed in any of these regions in either normal or colchicine treated animals. Previously we demonstrated that a system of centrifugal fibers to the parietal eye originates from neurons in the dorsal sac (Engbretson et al. 1981). The apparent absence of substance P-containing neurons in the dorsal sac suggests that the substance P-containing fibers in the parietal eye are not the previously observed centrifugal fibers. The source of the substance P-containing fibers in the parietal eye is unknown. The pars dorsolateralis of the left medial habenular nucleus receives a dense substance P-positive projection. No such projection was seen in the right habenula. Simultaneous visualization of the terminals of ganglion cells of the parietal eye (labeled with orthograde intraaxonally transported horseradish peroxidase) and substance P-like immunofluorescence showed that the locus of habenular immunoreactivity is distinct from the projection field of the parietal eye. Thus the substance P-positive terminals in the habenula do not originate in the parietal eye. Transection of the parietal nerve confirmed this conclusion.

The origin of centrifugal fibers to the inner ear in Caiman crocodilus. A horseradish peroxidase study

The origin of acoustic and vestibular efferent fibers was investigated in Caiman crocodilus. After injection of horseradish peroxidase (HRP) into the cochlear duct or into the ampullae of the horizontal and anterior semi-circular canals, cells in the medulla oblongata exhibited retrogradely transported HRP reaction product. Efferent vestibular fibers were found in the medial reticular nucleus, at the level of the oliva superior. There were more labeled neurons ipsilateral to the injection site. Efferent acoustic neurons were found close to or inside the rostroventral division of the oliva superior. They spread out into the medial reticular nucleus bilaterally. More labeled efferent acoustic neurons occurred contralateral to the injection site.

A retinopetal nucleus in the preoptic area in a teleost, Navodon modestus

Following horseradish peroxidase (HRP) applications to the optic nerve of a teleost ( Navodon modestus), a retinopetal nucleus was identified in the contralateral preoptic area. The nucleus was composed of small (7–10 μm) round cells. Centrifugal fibers from the nucleus were traced to the inner nuclear layer of the ipsilateral retina by both orthograde HRP and Fink-Heimer methods. The cells in the nucleus showed no neurosecretion. The retinopetal nucleus or neurons could not be found in Carassius carassius. No retinopetal neurons were found in the optic tectum in both species.

Origin of centrifugal fibers to the labyrinth in the frog ( Rana esculenta). A study with the fluorescent retrograde neuronal tracer ‘Fast blue’

After injecting a solution of a fluorescent retrograde neuronal tracer (Fast blue, Diamidino compound 253/50) into the perilymphatic space of the frog labyrinth ( Rana esculenta), labeled cells were found in the ventral and dorsal nuclei of the VIIIth nerve and in the nucleus reticularis medius. We consider these labeled cells to be the origin of the afferent innervation of the frog labyrinth. No evidence was found for the existence of a direct cerebello-labyrinthine connection.

Habenular asymmetry and the central connections of the parietal eye of the lizard.

Histochemical and autoradiographic analyses of the axonal transport of horseradish peroxidase and tritiated amino acids were employed to study the central connectivity of the lizard parietal eye. Somata and processes of centrifugal fibers to the parietal eye were localized in tissue of the dorsal sac and in the leptomeningeal sheath of the pineal gland. Analyses of series of transverse sections of the brain showed the left medial habenular nucleus to be subdivided into pars dorsolateralis and pars ventromedialis, and the right medial habenular nucleus not to be so subdivided. Centripetal fibers of parietal eye ganglion cells project to only the pars dorsolateralis of the left medial habenular nucleus and terminate there in two distinct fields. The asymmetry of the lizard habenula may be a specialization associated with the unilateral projection from the parietal eye.

Will central nervous systems in the adult mammal regenerate after bypassing a lesion? A study in the mouse and chick visual systems

We determined whether or not optic axons in the adult mammalian retina would regenerate well if allowed to bypass a lesion scar. In one series we produced microlesions with fine needles in the ganglion cell fiber layer of adult mouse retinas and later examined the retinas as silver-stained flat mounts to observe the behavior of optic axons that bypassed the lesion. Such axons continued to grow abortively, i.e., grew randomly and for only short distances, whether growing within a sector of Wallerian (anterograde) degeneration or in the neighboring zone of intact optic axon bundles. In a second series we produced optic nerve crushes in adult mice and observed the behavior of optic axons growing retrograde into the retina. These fibers similarly grew abortively whether in a zone of intact fiber bundles or when a retinal lesion was produced with the crush, in the sector of Wallerian degeneration. Retinal lesions in newly hatched chicks produced a comparable picture of abortive (short-distance) growth, but the optic and centrifugal fibers had a greater tendency to remain oriented within the ganglion cell fiber layer than did mouse axons. This improved orientation may be the consequence of the greater number of optic fibers in the chick retina and hence the greater opportunity for nonspecific contact guidance. The results indicate that blockage (by a lesion scar or myelin debris) and hypoxia are not the key causes of regenerative failure, as regeneration failed even when those factors were minimal.

Retinal efferents from the pretectal area in the rat

Horseradish peroxidase (HRP) labeled neuronal somas were consistently found in the medial pretectal area after intraocular injection of HRP. The labeled cells represent the source of centrifugal fibers to the rat retina.

Effect of intervening lesions on amino acid distributions in rat olfactory cortex and olfactory bulb.

Levels of the proposed neurotransmitter amino acids glutamate, aspartate, gamma-aminobutyric acid (GABA), and glycine were measured within the layered structures of the olfactory bulb and olfactory cortex following unilateral transections of the lateral olfactory tract or of virtually all fiber tracts of the olfactory peduncle. Distributions of the amino acids on both lesion and control sides were examined and compared by means of a mapping procedure. The results suggest: 1) Glutamate and aspartate are specifically associated with mitral (and presumably also tufted) cell axons and terminals in the piriform cortex. The distribution of aspartate in the olfactory bulb is further suggestive of a specific association of aspartate with mitral cell dendrites and somata. 2) Glutamate might be specifically associated with some centrifugal fibers traveling to the olfactory bulb in or near the anterior commissure. 3) GABA might be specifically related to some certrifugal fibers to the olfactory bulb in addition to its prominent association with granule cells of the bulb. 4) Glycine is unlikely to play a prominent neurotransmitter role in either the olfactory bulb or olfactory cortex.

Central projections of the frontal organ of Rana pipiens, as demonstrated by the anterograde transport of horseradish peroxidase.

The central projections of the frontal organ of Rana pipiens are more widespread, and more similar to those of the epiphysis, than previously realized. Fibers labeled with horseradish peroxidase were traced to the amygdala, the epiphysis, the pretectal region, and several nuclear areas of the mesencephalic and diencephalic central gray. When peroxidase reaction product was carefully distinguished from neuromelanin by means of polarization microscopy, no unequivocally labeled cell bodies off centrifugal fibers could be found.

Localization of acetylcholinesterase in the main and accessory olfactory bulbs of the mouse by light and electron microscopic histochemistry.

Experiments were conducted to examine by light and electron microscopy the localization of acetylcholinesterase (AChE) in the main (MOB) and accessory (AOB) olfactory bulbs of the normal mouse. Evidence from the literature for cholinergic innervation of the mammalian olfactory bulb was then assessed in light of possible correlation between reported sites of termination of centrifugal fibers to the olfactory bulb and the localization of AChE. AChE-positive nerve fibers were concentrated in the periglomerular region and internal plexiform layer of the MOB. Stained fibers were also present in the granule cell, mitral cell, and external plexiform layers as well as within glomeruli. A few neurons in all layers of the MOB contained AChE reaction product. Unlike the MOB, AChE-positive fibers were not present in the glomerular layer of the AOB. AChE-positive fibers were concentrated in the inner plexiform layer, whereas fewer stained fibers were observed in the external plexiform and mitral cell layer and granule cell layer. Lightly stained neurons were found in the deeper portions of the external plexiform and mitral cell layer and granule cell layer. Ultrastructurally, AChE reaction product in the MOB and AOB was predominantly associated with small unmyelinated axons. Reaction product was also observed adjacent to axon terminals and dendrites. Occasionally within the MOB, AChE activity was found within periglomerular, tufted, short-axon, mitral, and granule cells. In the AOB, however, intracellular AChE activity was observed within some mitral/tufted cells and only a few granule cells. In conclusion, the AChE reaction product was mainly associated with axons in regions of the MOB where centrifugal fibers have been reported. Accessory olfactory bulb AChE localization was different from that of the MOB, suggesting a different pattern of cholinergic input to the AOB. The small amounts and sites of intraneuronal AChE reaction product in cells of the olfactory bulb indicate cholinoceptive rather than cholinergic function.