Extreme 13C,15N and 17O isotopic enrichment in the young planetary nebula K4-47.

Carbon, nitrogen and oxygen are the three most abundant elements in the Galaxy after hydrogen and helium. Whereas hydrogen and helium were created in the Big Bang, carbon, nitrogen and oxygen arise from nucleosynthesis in stars. Of particular interest1,2 are the isotopic ratios 12C/13C, 14N/15N and 16O/17O because they are effective tracers of nucleosynthesis and help to benchmark the chemical processes that occurred in primitive interstellar material as it evolved into our Solar System3. However, the origins of the rare isotopes 15N and 17O remain uncertain, although novae and very massive stars that explode as supernovae are postulated4-6 to be the main sources of 15N. Here we report millimetre-wavelength observations of the young bipolar planetary nebula K4-47 that indicate another possible source for these isotopes. We identify various carbon-bearing molecules in K4-47 that show that this object is carbon-rich, and find unusually high enrichment in rare carbon (13C), oxygen (17O) and nitrogen (15N) isotopes: 12C/13C = 2.2 ± 0.8, 16O/17O = 21.4 ± 10.3 and 14N/15N = 13.6 ± 6.5 (uncertainties are three standard deviations); for comparison, the corresponding solar ratios7 are 89.4 ± 0.2, 2,632 ± 7 and 435 ± 57. One possible interpretation of these results is that K4-47 arose from a J-type asymptotic giant branch star that underwent a helium-shell flash (an explosive nucleosynthetic event that converts large quantities of helium to carbon and other elements), enriching the resulting planetary nebula in 15N and 17O and creating its bipolar geometry. Other possible explanations are that K4-47 is a binary system or that it resulted from a white dwarf merger, as has been suggested for object CK Vul8. These results suggest that nucleosynthesis of carbon, nitrogen and oxygen is not well understood and that the classification of certain stardust grains must be reconsidered.

Prebiotic chemical evolution in the astrophysical context.

An ever increasing amount of molecular material is being discovered in the interstellar medium, associated with the birth and death of stars and planetary systems. Radio and millimeter-wave astronomical observations, made possible by high-resolution laboratory spectroscopy, uniquely trace the history of gas-phase molecules with biogenic elements. Using a combination of both disciplines, the full extent of the cycling of molecular matter, from circumstellar ejecta of dying stars - objects which expel large amounts of carbon - to nascent solar systems, has been investigated. Such stellar ejecta have been found to exhibit a rich and varied chemical content. Observations demonstrate that this molecular material is passed onto planetary nebulae, the final phase of stellar evolution. Here the star sheds almost its entire original mass, becoming an ultraviolet-emitting white dwarf. Molecules such as H2CO, HCN, HCO(+), and CCH are present in significant concentrations across the entire age span of such nebulae. These data suggest that gas-phase polyatomic, carbon-containing molecules survive the planetary nebula phase and subsequently are transported into the interstellar medium, seeding the chemistry of diffuse and then dense clouds. The extent of the chemical complexity in dense clouds is unknown, hindered by the high spectral line density. Organic species such as acetamide and methyl amine are present in such objects, and NH2CHO has a wide Galactic distribution. However, organophosphorus compounds have not yet been detected in dense clouds. Based on carbon and nitrogen isotope ratios, molecular material from the ISM appears to become incorporated into solar system planetesimals. It is therefore likely that interstellar synthesis influences prebiotic chemistry on planet surfaces.

Chemical complexity in the winds of the oxygen-rich supergiant star VY Canis Majoris.

The interstellar medium is enriched primarily by matter ejected from old, evolved stars. The outflows from these stars create spherical envelopes, which foster gas-phase chemistry. The chemical complexity in circumstellar shells was originally thought to be dominated by the elemental carbon to oxygen ratio. Observations have suggested that envelopes with more carbon than oxygen have a significantly greater abundance of molecules than their oxygen-rich analogues. Here we report observations of molecules in the oxygen-rich shell of the red supergiant star VY Canis Majoris (VY CMa). A variety of unexpected chemical compounds have been identified, including NaCl, PN, HNC and HCO+. From the spectral line profiles, the molecules can be distinguished as arising from three distinct kinematic regions: a spherical outflow, a tightly collimated, blue-shifted expansion, and a directed, red-shifted flow. Certain species (SiO, PN and NaCl) exclusively trace the spherical flow, whereas HNC and sulphur-bearing molecules (amongst others) are selectively created in the two expansions, perhaps arising from shock waves. CO, HCN, CS and HCO+ exist in all three components. Despite the oxygen-rich environment, HCN seems to be as abundant as CO. These results suggest that oxygen-rich shells may be as chemically diverse as their carbon counterparts.

Elevation of nerve growth factor and antisense knockdown of TrkA receptor during contextual memory consolidation.

We report here a series of experiments establishing a role for nerve growth factor and its high-affinity receptor TrkA in contextual memory consolidation. In all experiments, we trained rats in a novel chamber using tone and shock. Our first experiment revealed that endogenous nerve growth factor (NGF) increases in the hippocampus at a critical time during consolidation that occurs 1 week after training. NGF levels at other intervals (24 hr and 2 and 4 weeks after training) did not differ from those of naive control animals. In our second experiment, we blocked effects that NGF has at 1 week after training by infusing antisense TrkA phosphorothioate DNA oligonucleotide. Reduction of septohippocampal TrkA receptor expression selectively impaired memory consolidation for context but not for tone. Animals with antisense TrkA oligonucleotide infused into the medial septal area or CA1 of the hippocampus froze less when placed in the training chamber than did animals infused with inactive randomized oligonucleotide. At 4 weeks after training, antisense TrkA oligonucleotide had no effect on freezing. Third, we correlated levels of freezing with choline acetyltransferase (ChAT) and vesicular acetylcholine transporter (VAChT) immunohistochemistry. Antisense TrkA infused into CA1 of the hippocampus reduced cell body cross-sectional area for cholinergic cells in the medial septal area and decreased the density of hippocampal terminals labeled for ChAT and VAChT proteins. Cholinergic cell body measurements were significantly correlated with freezing. Taken together, these results indicate a role for nerve growth factor acting via the TrkA receptor on ChAT and VAChT proteins in contextual memory consolidation.

Dendritic encoding: an alternative to temporal synaptic coding of conscious experience.

In this commentary, arguments are made for a dendritic code being preferable to a temporal synaptic code as a model of conscious experience. A temporal firing pattern is a product of an ongoing neural computation; hence, it is based on a neural algorithm and an algorithm may not provide the most suitable model for conscious experience. Reiteration of a temporal firing code as suggested in a preceding article (Helekar, 1999) does not necessarily improve the situation. The alternative model presented here is that certain synaptic activity patterns, possibly those possessing universal features as suggested by Helekar, can become encoded in the dendritic structure. Following dendritic encoding, quantum phenomena in those specific dendrite sets could illuminate the static image of that encoded synaptic activity. It is the activation of the static image that would be equivalent to conscious experience; thus, conscious awareness would not be directly affiliated with synaptic activity. This dendrite encoding model may go farther than other models to explain the gestalt nature of consciousness, insofar as quantum entanglement could produce an interconnectedness between specific sets of dendrites-an interconnectedness that need not be based on neural computation or neural connections.

Hippocampal microtubule-associated protein-2 alterations with contextual memory.

Using immunohistochemistry and immunoblots, we show that alterations in hippocampal microtubule-associated protein-2 appear to be highly correlated with contextual memory as measured by significantly heightened fear responses. Compared to naive controls, rats trained in a novel context showed significantly increased immunostaining for the high molecular weight microtubule-associated protein-2a/b. This increase was observed 2 weeks after training and it was selective for hippocampal CA1 and CA2 pyramidal cells. Pre-exposure to the training context one month before training altered the hippocampal microtubule-associated protein-2 response; in these animals only the dentate gyrus showed significantly increased microtubule-associated protein-2a/b. Training-related increases in immunohistochemical staining for microtubule-associated protein-2 suggested that there was an increase in overall intact protein, an increase in immunoreactive breakdown products, or changes in protein compartmentalization. Immunoblots of hippocampal homogenates reacted with monoclonal antibodies to microtubule-associated protein-2a/b showed an increased presence of breakdown products in trained animals compared to untrained controls. Additional immunoblot studies demonstrated statistically significant decreases in the levels and/or phosphorylation state of the low molecular weight microtubule-associated protein-2c in the hippocampus of trained animals as compared to that of controls. These alterations in microtubule-associated protein-2 may reflect dendritic remodeling related to contextual memory storage.

A structural basis for memory storage in mammals.

It is proposed that altered dendrite length and de novo formation of new dendrite branches in cholinoceptive cells are responsible for long-term memory storage, a process enabled by the degradation of microtubule-associated protein-2. These memories are encoded as modality-specific associable representations. Accordingly, associable representations are confined to cytoarchitectonic modules of the cerebral cortex, hippocampus, and amygdala. The proposed sequence of events leading to long-term storage in cholinoceptive dendrites begins with changes in neuronal activity, then in neurotrophin release, followed by enhanced acetylcholine release, muscarinic response, calcium influx, degradation of microtubule-associated protein-2, and finally new dendrite structure. Hypothetically, each associable representation consists of altered dendrite segments from approximately 5000-15,000 cholinoceptive cells contained within one or a few module(s). Simultaneous restructuring during consolidation of long-term memory is hypothesized to result in a similar infrastructure among dendrite sets, facilitating co-activation of those dendrite sets by neurotransmitters such as acetylcholine, and conceivably enabling high energy interactions between those dendrites by phenomena such as quantum optical coherence. Based on the specific architecture proposed, it is estimated that the human telecephalon contains enough dendrites to encode 50 million associable representations in a lifetime, or put another way, to encode one new associable representation each minute. The implications that this proposal has regarding treatments for Alzheimer's disease are also discussed.

A possible role for cholinergic neurons of the basal forebrain and pontomesencephalon in consciousness.

Excitation at widely dispersed loci in the cerebral cortex may represent a neural correlate of consciousness. Accordingly, each unique combination of excited neurons would determine the content of a conscious moment. This conceptualization would be strengthened if we could identify what orchestrates the various combinations of excited neurons. In the present paper, cholinergic afferents to the cerebral cortex are hypothesized to enhance activity at specific cortical circuits and determine the content of a conscious moment by activating certain combinations of postsynaptic sites in select cortical modules. It is proposed that these selections are enabled by learning-related restructuring that simultaneously adjusts the cytoskeletal matrix at specific constellations of postsynaptic sites giving all a similar geometry. The underlying mechanism of conscious awareness hypothetically involves cholinergic mediation of linkages between microtubules and microtubule-associated protein-2 (MAP-2). The first reason for proposing this mechanism is that previous studies indicate cognitive-related changes in MAP-2 occur in cholinoceptive cells within discrete cortical modules. These cortical modules are found throughout the cerebral cortex, measure 1-2 mm2, and contain approximately 10(3)-10(4) cholinoceptive cells that are enriched with MAP-2. The subsectors of the hippocampus may function similarly to cortical modules. The second reason for proposing the current mechanism is that the MAP-2 rich cells throughout the cerebral cortex correspond almost exactly with the cortical cells containing muscarinic receptors. Many of these cholinoceptive, MAP-2 rich cells are large pyramidal cell types, but some are also small pyramidal cells and nonpyramidal types. The third reason for proposing the current mechanism is that cholinergic afferents are module-specific; cholinergic axons terminate wholly within individual cortical modules. The cholinergic afferents may be unique in this regard. Finally, the tapering apical dendrites of pyramidal cells are proposed as primary sites for cholinergic mediation of linkages between MAP-2 and microtubules because especially high amounts of MAP-2 are found here. Also, the possibility is raised that muscarinic actions on MAP-2 could modulate microtubular coherence and self-collapse, phenomena that have been suggested to underlie consciousness.

The critical role of cholinergic basal forebrain neurons in morphological change and memory encoding: a hypothesis.

It has been known for a long time that cholinergic basal forebrain neurons which project to the cerebral cortex play a role in learning and memory. Behavioral studies following lesions, for example, repeatedly have suggested multiple learning-related roles for these neurons. Apart from behavioral studies, cholinergic neurons have been shown to possess extraordinarily plastic axons. This plasticity has not been related comprehensively to mnemonic devises, even though morphological changes in the CNS are prime candidates for the neural engram. In this paper, I propose a hypothesis that relates these two characteristics of cholinergic neurons. This hypothesis is that plastic cholinergic axon terminals induce structural reorganization in their targets during memory storage. Possible intracellular mechanisms are examined, whereby acetylcholine release in the cerebral cortex could cause postsynaptic structural changes. Finally, the characteristics of the overall cholinergic-cholinoceptive cell "engram" are elaborated with particular attention paid to the encoding of the stimulus properties along with the context and meaning of the stimulus.

Global and serial neurons form A hierarchically arranged interface proposed to underlie memory and cognition.

It is hypothesized that the cholinergic and monoaminergic neurons of the brain from a global network. What is meant by a global network is that these neurons operate as a unified whole, generating widespread patterns of activity in concert with particular electroencephalographic states, moods and cognitive gestalts. Apart from cholinergic and monoaminergic global systems, most other mammalian neurons relay sensory information about the external and internal milieu to serially ordered loci. These "serial" neurons are neurochemically distinct from global neurons and commonly use small molecule amino acid neurotransmitters such as glutamate or aspartate. Viewing the circuitry of the mammalian brain within the global-serial dichotomy leads to a number of novel interpretations and predictions. Global systems seem to be capable of transforming incoming sensory data into cognitive-related activity patterns. A comparative examination of global and serial systems anatomy, development and physiology reveals how global systems might turn sensation into mentation. An important step in this process is the permanent encoding of memory. Global neurons are particularly plastic, as are the neurons receiving global inputs. Global afferents appear to be capable of reorganizing synapses on recipient serial cells, thus leading to enhanced responding to a signal, in a particular context and state of arousal.

Choline acetyltransferase mRNA plasticity with Pavlovian conditioning.

Choline acetyltransferase mRNA and somal area increased selectively in the ventral nucleus basalis of rats trained that a tone signals immediate shock (i.e., predicts danger). Retrograde tracing showed the affected cells correspond to those that project to the auditory cortex. Behavior responses and mRNA increased significantly above those of control groups trained with the tone not signaling immediate shock. In one of those control groups, animals learned that the same tone signaled a shock-free period before shock. These animals showed a visibly decreased riboprobe and a trend toward smaller somal areas. These results implicate transcriptional regulation of choline acetyltransferase in long-term memory storage. Selective attention and inattention to the tone are possible components of memory encoded by the molecular changes reported here.

Trophic-factor modulation of cortical acetylcholinesterase reappearance following transection of the medial cholinergic pathway in the adult rat.

Laminar patterns of cortical acetylcholinesterase (AChE) activity are reestablished in the adult, pharmacologically unmanipulated rat following axotomy of the medial cholinergic pathway. The extent to which trophic and/or growth promoting or inhibiting agents modulate AChE fiber reappearance is not fully understood. Such studies, however, would further clarify possible roles for these agents in neuronal plasticity in response to injury, as well as in plastic processes associated with normative functions. In the present experiments, we explored trophic modulation by intracortically infusing nerve growth factor (NGF) or somatostatin into cingulate cortex at a site distal to transection of the medial cholinergic pathway. Comparisons were made with sham-operated or noninfused transected controls, as well as with transected animals infused with renin or antibodies against NGF. Administration began 2 days after axotomy and continued at successive 3-day intervals for 4 weeks. It was found that, proximal to the lesion site, NGF increased and somatostatin decreased optical density of AChE; the number of AChE-containing fibers was unaltered compared to controls. Distal to the knife cut, both NGF and somatostatin increased number of AChE fibers but did not alter overall AChE optical density. Nonetheless, NGF produced an increase in the number of intensely staining puncta both proximal and distal to the cut. Neither renin nor anti-NGF antibodies produced statistically significant effects on optical density or number of fibers at any cortical locus studied. We conclude that NGF and somatostatin have opposite effects on the expression of AChE: whereas NGF increases AChE levels, somatostatin inhibits AChE accumulation in proximal fibers, perhaps by actions on synthesis or transport. Fiber proliferation, which only occurred distally, was affected positively by both NGF and somatostatin, indicating that neurite-promoting effects produced by both agents are confined to tissue regions where neurite extension is stimulated by axotomy. Increases in AChE-positive puncta produced by NGF, however, were not confined to regions of fiber proliferation.

Pavlovian conditioning alters cortical microtubule-associated protein-2.

Three experiments indicate that Pavlovian conditioning to tone alters microtubule-associated protein-2 (MAP-2) in the temporal cortex. First, increased MAP-2 immunohistochemistry was evident in temporal cortex following tone-shock pairings but not light-shock pairings. In the second experiment, animals given tone paired with shock (compared with animals trained with tone unpaired with shock or given tone only) showed MAP-2 immunohistochemical changes in the temporal cortex, as well as in the frontal and cingulate cortex, the hippocampus and amygdala. In experiment 3, quantitative immunoblots showed decreased intact MAP-2 and increased breakdown products selectively in temporal cortex following fear conditioning to tone. Conditioning to tone also increased sizes of MAP-2 rich pyramidal somata and apical dendrites in temporal and frontal cortex.

Cholinoceptive cells in rat cerebral cortex: somatodendritic immunoreactivity for muscarinic receptor and cytoskeletal proteins.

Adult rat telecephalon was surveyed for cells demonstrating immunopositivity for muscarinic receptor (M35 antibody), microtubule-associated proteins, neurofilaments, and brain-spectrin. Neurons immunostained for muscarinic receptor were found in frontal, parietal, temporal, and occipital isocortex where they accounted for approximately 15-16% of all neurons. This labeling involved a large proportion of layer V pyramidal cells, some layer III pyramidal cells and a small proportion of non-pyramidal cells in layers II-VI. In the hippocampus, pyramidal cells, non-pyramidal cells and granular cells were immunoreactive, as were many pyramidal cells in subicular and entorhinal cortices. In every cortical region examined, cells demonstrating muscarinic receptor were morphologically identical to cells stained lightly to moderately for acetylcholinesterase following pretreatment with diisopropylfluorophosphate, and they were found in similar numbers and in a similar laminar distribution. These characteristics further corresponded to those of cells whose somatodendritic compartments were intensely immunostained by antibodies to microtubule-associated proteins (MAP): MAP-1, MAP-2, MAP-5; neurofilament proteins (NF): NF-68kD, NF-160kD, NF-200kD; and brain-spectrin. Double immunostaining using a fluorescence method followed by an avidin-biotin staining procedure revealed that cortical cells which possessed immunoreactivity for muscarinic receptor demonstrated an 80-85% overlap with cells that were immunoreactive for MAP-2 (and tau) or NF-200kD. Following unilateral ibotenic acid lesions of the nucleus basalis, MAP-2 immunostaining was reduced in the ipsilateral isocortex. This significant reduction was most evident in the parietal cortex, exactly where maximal loss of acetylcholinesterase-containing fibers occurred. The same lesion produced no significant difference in immunodensity of muscarinic receptor, MAP-1, MAP-5 NF-68kD, NF-160kD and NF-200kD. Thus, cortical cholinoceptive cells are enriched with cytoskeletal components and cholinergic afferents modulate cortical MAP-2.

Reestablishment of laminar patterns of cortical acetylcholinesterase activity following axotomy of the medial cholinergic pathway in the adult rat.

Little is known about injury-induced restructuring of cholinergic neurons projecting from the basal forebrain to nonhippocampal loci in the mature mammalian brain. In an attempt to address this issue further, we made unilateral knife cuts in the cingulate cortex of adult rats and assessed the rates and extents of fiber regeneration. Among the fibers interdicted by these transections were axons of the medial cholinergic pathway, which originates prominently in the diagonal band nuclei and projects to medial frontal, cingulate, and occipital cortices. Immediately following the lesion, acetylcholinesterase (AChE)-containing fibers proximal to the transection demonstrated increased enzyme activity that remained elevated for at least 12 weeks before returning to control values within 9 months. Loss of distal cholinesterase staining was seen from 1-2 weeks postaxotomy, with partial restoration of enzyme activity within 4 weeks. By 12 weeks, virtually complete restitution of laminar patterns of AChE innervation had occurred in the cortex posterior to the knife cut. At least part of this reinnervation was attributable to fibers growing or elongating across the glial scar, with the degree of regrowth being correlated positively with the rapidity and magnitude of wound closure. Sprouting from the cut ends of axons, as well as collateral growth of those and uncut adjacent fibers, are likely sources of fibers contributing to the restoration of normal laminar profiles of AChE activity.

Organization of central cholinergic neurons revealed by combined in situ hybridization histochemistry and choline-O-acetyltransferase immunocytochemistry.

Digoxigenin-labeled riboprobes and in situ hybridization of choline-O-acetyltransferase mRNA, both alone and in combination with immunohistochemical procedures for the synthetic enzyme of acetylcholine, were used to map the topography of putative cholinergic neurons in the rat central nervous system. Only the anti-sense riboprobe yielded specific labeling, which was absent in brain sections processed with sense riboprobe. Telencephalic neurons demonstrating the mRNA for choline-O-acetyltransferase and choline-O-acetyltransferase-like immunoreactivity were found in the caudate-putamen nucleus, nucleus accumbens, olfactory tubercule, Islands of Calleja complex, medial septal nucleus, vertical and horizontal limbs of the diagonal band, substantia innominata, nucleus basalis, and nucleus of the ansa lenticularis, as well as occasionally in the amygdala. Neurons in the cerebral cortex, hippocampus, and primary olfactory structures did not demonstrate hybridization signal, even though some cells in those areas were observed to exhibit choline-O-acetyltransferase-like immunopositivity. Thalamic cells were devoid of hybrido- and immunoreactivity, with the exception of several neurons located primarily in the ventral two-thirds of the medial habenula. A few cell bodies labeled with riboprobe and co-localizing choline-O-acetyltransferase-like immunopositivity were found in the lateral hypothalamus, caudal extension of the internal capsule, and zona incerta. Neurons in the pedunculopontine and laterodorsal tegmental nuclei evinced moderate hybridization signal, whereas cells of the parabigeminal nucleus were very weakly reactive. In contrast, motor neurons of the cranial nerve nuclei demonstrated high levels of choline-O-acetyltransferase mRNA and choline-O-acetyltransferase-like immunoreactivity. Putative cholinergic somata in the ventral horns and intermediolateral cell columns of the spinal cord and around the central canal were also labeled with riboprobe. It is concluded that hybridocytochemistry with digoxigenin-labeled riboprobes confirms the existence of cholinergic neurons in most of the neural regions believed to contain them on the basis of acetylcholinesterase pharmacohistochemistry and choline-O-acetyltransferase immunocytochemistry, with the prominent exceptions of the cerebral cortex, hippocampus, olfactory bulb, anterior olfactory nucleus, and caudal raphe nuclei, which apparently do not possess neurons expressing detectable levels of the mRNA for the synthetic enzyme of acetylcholine.

Cat 'P300' and cholinergic septohippocampal neurons: depth recordings, lesions, and choline acetyltransferase immunohistochemistry.

The role of septohippocampal circuits in the generation of the P300 response in cats (n = 12) was explored in a series of depth recording, tract-tracing and lesion experiments. Systematic mapping of the hippocampus in 1-mm increments from rostral to caudal extent revealed large positive potentials, greater in amplitude to rare than to frequent stimuli, within the 200-500 ms range. Each map revealed maximal amplitude responses at diverse, widely distributed hippocampus loci. Furthermore, these electrical responses displayed polarity inversion within the hippocampus that was generally localized to the pyramidal cell layer; polarity inversion was also observed in the adjacent entorhinal cortex and amygdala. Injections of propidium iodide, a tract-tracing agent, into these inversion sites resulted in retrograde labeling of small clusters of choline acetyltransferase (ChAT)-positive neurons in the medial septal nucleus and vertical limb of the diagonal band. Aspiration lesions that bilaterally destroyed large amounts of caudal hippocampus from stereotaxic levels A4 to A1 resulted in a decreased number of cells expressing ChAT in the rostral basal nuclear complex. In only 2 cats was the preoperative presence of a significant vertex P300 absent postoperatively. In the majority of cases (5 of 8 animals), hippocampal aspiration produced an enhancement of the preoperative P300 potential. We conclude that cholinergic mechanisms are importantly, albeit not exclusively, involved in the mediation of P300 potentials in cats. Neurons mediating P300 responses appear to be organized in diverse clusters of septal and diagonal band cells. These septal cells may facilitate, and in turn be facilitated or inhibited as a function of hippocampal, or other, allocortical feedback loops.

Cholinergic neurons in the rat central nervous system demonstrated by in situ hybridization of choline acetyltransferase mRNA.

Digoxigenin-labeled RNA probes and in situ hybridization histochemistry were used to examine choline acetyltransferase gene expression in the rat central nervous system. Hybridization signal was present only in brain sections processed with the antisense riboprobe. The sense probe did not yield labeling, further validating the specificity of tissue reactivity. Telencephalic neurons containing the mRNA for the cholinergic synthetic enzyme were found in the caudate-putamen nucleus, nucleus accumbens, olfactory tubercule, islands of Calleja complex, medial septal nucleus, vertical and horizontal limbs of the diagonal band, substantia innominata, nucleus basalis, and nucleus of the ansa lenticularis. Some somata evincing hybridization signal were observed in the anterior amygdalar area, and an occasional such cell was seen in the basolateral and central amygdalar nuclei. Neurons in the cerebral cortex, hippocampus, and primary olfactory structures did not demonstrate hybridocytochemically detectable amounts of choline acetyltransferase mRNA. Thalamic cells were devoid of reactivity, with the exception of several neurons located primarily in the ventral two-thirds of the medial habenula. A few somata labeled with riboprobe were found in the lateral hypothalamus, caudal extension of the internal capsule, and zona incerta. Neurons in the pedunculopontine and laterodorsal tegmental nuclei were moderately reactive, whereas cells of the parabigeminal nucleus exhibited a very weak hybridization signal. No somata in the brainstem raphe nuclei, including raphe obscurus and raphe magnus, were observed to bind riboprobe. In contrast, motor neurons of the cranial nerve nuclei demonstrated relatively large amounts of choline acetyltransferase mRNA. Putative cholinergic somata in the ventral horns and intermediolateral cell columns of the spinal cord were also labeled with riboprobe, as were a few cells around the central canal. We conclude that hybridocytochemistry with digoxigenin-labeled riboprobes confirms the existence of cholinergic neurons (i.e. those that synthesize and use acetylcholine as a neurotransmitter) in most of the neural regions deduced to contain them on the basis of previous histochemical and immunocytochemical data. Notable exceptions are the cerebral cortex and hippocampus, which do not possess neurons expressing detectable levels of choline acetyltransferase mRNA.