In vitro and in vivo characterization of MPEP, an allosteric modulator of the metabotropic glutamate receptor subtype 5: review article.

There is a need to identify subtype-specific ligands for mGlu receptors to elucidate the potential of these receptors for the treatment of nervous system disorders. To date, most mGlu receptor antagonists are amino acid-like compounds acting as competitive antagonists at the glutamate binding site located in the large extracellular N-terminal domain. We have characterized novel subtype-selective mGlu(5) receptor antagonists which are structurally unrelated to competitive mGlu receptor ligands. Using a series of chimeric receptors and point mutations we demonstrate that these antagonists act as inverse agonists with a novel allosteric binding site in the seven-transmembrane domain. Recent studies in animal models implicate mGlu(5) receptors as a potentially important therapeutic target particularly for the treatment of pain and anxiety.

The metabotropic glutamate receptor antagonist 2-methyl-6-(phenylethynyl)-pyridine (MPEP) blocks fear conditioning in rats.

Glutamate receptors play an essential role in fear-related learning and memory. The present study was designed to assess the role of the group I metabotropic glutamate receptor (mGluR) subtype 5 in the acquisition and retrieval of conditioned fear in rats. The selective mGluR5 antagonist 2-methyl-6-(phenylethynyl)-pyridine (MPEP) was applied systemically (0.0, 0.3, 3.0, 30.0 mg/kg per os) 60 min before the acquisition training and before the expression of conditioned fear, respectively, in the fear-potentiated startle paradigm. MPEP dose-dependently blocked the acquisition of fear. This effect was not due to state-dependent learning. MPEP also prevented the expression of fear at a dose of 30.0 mg/kg. As a positive control for these effects, we showed that the benzodiazepine anxiolytic compound diazepam (1.25 mg/kg intraperitoneally) also blocked acquisition and expression of fear potentiated startle. MPEP did not affect the baseline startle magnitude, short-term habituation of startle, sensitisation of startle by footshocks or prepulse inhibition of startle. These data indicate a crucial role for mGluR5 in the regulation of fear conditioning. In the highest dose MPEP might exert anxiolytic properties.

Electrophysiological characterization of CGP68730A a N-methyl-D-aspartate antagonist acting at the strychnine-insensitive glycine site.

1. Electrophysiological experiments were performed in vitro and in vivo. Voltage clamp recordings were done in Xenopus oocytes. Extracellular recordings were done in vitro in the neocortical slice and in the CA1 region of the hippocampal slice and in vivo in the CA1 region of the hippocampus of the anaesthetized rat. 2. In oocytes expressing either the human NMDAR1A/2A or 1A/2B subunit combinations, CGP68730A [sodium (-)-9-bromo-2,3,6,7-tetrahydro-5,6-dioxo-5H-pyrazino[1,2,3-de]-1,4-benzo thiazine-3-acetic acid] antagonized L-glutamate / glycine induced currents with calculated IC50s of 20.5 and 81.6 nM, respectively. 3. In vitro, CGP68730A was tested on NMDA induced depolarizations in the neocortical slice preparation and on epileptiform activity in hippocampal slices bathed in Mg2+-free-medium, which is known to be NMDA mediated. In both in vitro models CGP68730A exhibited antagonistic effects on the NMDA receptor mediated responses. 4. In vivo CGP68730A was tested on NMDA induced excitations in the CA1 region. CGP68730A abolished NMDA induced excitations when applied microiontophoretically. However, only weak effects on NMDA induced excitation were observed after systemic administration at 100 mg/kg i.v.. These results indicate that CGP68730A has poor central nervous system bioavailability. 5. In oocytes, an increase in the glycine concentration from the EC80 to the EC95.99 shifted the inhibition curves for CGP68730A to the right. Furthermore, in neocortical slices and in anaesthetized rats CGP68730A inhibited NMDA mediated depolarizations, and this effect could be reversed by the addition of the glycine mimetic D-serine. This indicates that these effects of CGP68730A are mediated by an action on the strychnine-insensitive glycine site. 6. Selectivity tests in oocytes and in the neocortical slice preparation, using NMDA, kainate and AMPA showed that CGP68730A was selective in antagonizing NMDA receptor mediated responses. In oocytes, the compound was about 1000 times less potent on the rat GluR3 and the human GluR6 receptors than on the human NMDAR1A/2A subunit combination. In the neocortical slice preparationCGP68730A had no effects on AMPA or kainate induced depolarizations at concentrations of 3 and 10 microM. At 30 microM CGP68730A reduced the effects of each of the three agonists tested. 7. Thus, CGP68730A seems to be a selective antagonist at the strychnine-insensitive glycine coagonist site of the NMDA receptor. However, the compound showed no obvious central NMDA antagonistic effects following intravenous application.

Transgenic activation of Ras in neurons promotes hypertrophy and protects from lesion-induced degeneration.

Ras is a universal eukaryotic intracellular protein integrating extracellular signals from multiple receptor types. To investigate its role in the adult central nervous system, constitutively activated V12-Ha-Ras was expressed selectively in neurons of transgenic mice via a synapsin promoter. Ras-transgene protein expression increased postnatally, reaching a four- to fivefold elevation at day 40 and persisting at this level, thereafter. Neuronal Ras was constitutively active and a corresponding activating phosphorylation of mitogen-activated kinase was observed, but there were no changes in the activity of phosphoinositide 3-kinase, the phosphorylation of its target kinase Akt/PKB, or expression of the anti-apoptotic proteins Bcl-2 or Bcl-X(L). Neuronal Ras activation did not alter the total number of neurons, but induced cell soma hypertrophy, which resulted in a 14.5% increase of total brain volume. Choline acetyltransferase and tyrosine hydroxylase activities were increased, as well as neuropeptide Y expression. Degeneration of motorneurons was completely prevented after facial nerve lesion in Ras-transgenic mice. Furthermore, neurotoxin-induced degeneration of dopaminergic substantia nigra neurons and their striatal projections was greatly attenuated. Thus, the Ras signaling pathway mimics neurotrophic effects and triggers neuroprotective mechanisms in adult mice. Neuronal Ras activation might become a tool to stabilize donor neurons for neural transplantation and to protect neuronal populations in neurodegenerative diseases.

2-Methyl-6-(phenylethynyl)-pyridine (MPEP), a potent, selective and systemically active mGlu5 receptor antagonist.

In the present paper we describe 2-methyl-6-(phenylethynyl)-pyridine (MPEP) as a potent, selective and systemically active antagonist for the metabotropic glutamate receptor subtype 5 (mGlu5). At the human mGlu5a receptor expressed in recombinant cells, MPEP completely inhibited quisqualate-stimulated phosphoinositide (PI) hydrolysis with an IC50 value of 36 nM while having no agonist or antagonist activities at cells expressing the human mGlu1b receptor at concentrations up to 30 microM. When tested at group II and III receptors, MPEP did not show agonist or antagonist activity at 100 microM on human mGlu2, -3, -4a, -7b, and -8a receptors nor at 10 microM on the human mGlu6 receptor. Electrophysiological recordings in Xenopus laevis oocytes demonstrated no significant effect at 100 microM on human NMDA (NMDA1A/2A), rat AMPA (Glu3-(flop)) and human kainate (Glu6-(IYQ)) receptor subtypes nor at 10 microM on the human NMDA1A/2B receptor. In rat neonatal brain slices, MPEP inhibited DHPG-stimulated PI hydrolysis with a potency and selectivity similar to that observed on human mGlu receptors. Furthermore, in extracellular recordings in the CA1 area of the hippocampus in anesthetized rats, the microiontophoretic application of DHPG induced neuronal firing that was blocked when MPEP was administered by iontophoretic or intravenous routes. Excitations induced by microiontophoretic application of AMPA were not affected.

Reevaluation of ACEA 1021 as an antagonist at the strychnine-insensitive glycine site of the N-methyl-D-aspartate receptor.

Electrophysiological experiments were performed in vitro and in vivo to characterize the inhibitory effects of 6,7-dichloro-5-nitro-1,4-dihydro-2,3-quinoxalinedione (ACEA 1021; licostinel) on rat brain glutamate receptors. In vitro, ACEA 1021 was tested on N-methyl-D-aspartate (NMDA)-induced depolarizations in the neocortical slice preparation and on epileptiform activity in Mg2+-free hippocampal slices, which is known to be NMDA receptor mediated. In both in vitro models, ACEA 1021 exhibited antagonistic effects on the NMDA receptor-mediated responses. Selectivity tests in the neocortical slice preparation, using NMDA, kainate and alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) showed that 10 microM ACEA 1021 reduced NMDA and kainate responses to 27.9 and 79.9% of the control value, respectively, whereas responses to AMPA were increased by 2.4% above the control value, thus showing that at this concentration ACEA 1021 acts preferentially at NMDA receptors. However, at 30 microM, all the NMDA-, AMPA- and kainate-induced responses were reduced. In vivo, ACEA 1021 was tested on NMDA-induced excitation in the CA1 region. After systemic administration of ACEA 1021, central effects were observed at 10 mg/kg i.v. in the CA1 region. These results indicate that ACEA 1021 is centrally active and inhibits NMDA receptor-mediated responses. Interestingly, selectivity tests in the CA1 region did not show clear differences in the action of ACEA 1021 on NMDA- and AMPA-induced excitations. Furthermore, ACh-induced excitations were also reduced. Thus, at low concentrations, ACEA 1021 seems to be a selective antagonist at the strychnine-insensitive glycine site of the NMDA receptor. However, at 30 microM in vitro and at 10 mg/kg in vivo, non-NMDA receptor-mediated actions of ACEA 1021 are observed. Our results suggest that these additional effects of ACEA 1021 may contribute to its anticonvulsive properties in mice as well as to its neuroprotective properties in animal models of cerebral ischemia.

Exploration of P-type Ca2+ channels as drug targets for the treatment of epilepsy or ischemic stroke.

We investigated the neuroprotective efficacy of the P-type Ca2+ channel antagonist daurisoline against electroshock-induced convulsions in rats and mice, hypoxic/hypoglycemic-induced damage in rat hippocampal slices and brain damage induced by occlusion of the middle cerebral artery (MCA) in rats. Daurisoline applied intravenously (i.v.) (bolus of 1-60 mg/kg) reduced the spontaneous activity of rat cerebellar Purkinje cells in a dose-dependent manner, a result demonstrating activity in the brain with systemic administration of the compound. While this effect reversed rapidly in about 10-20 min following bolus-application of the drug at doses of up to 30 mg/kg, a dose of 60 mg/kg consistently induced a depression of respiration followed by death of the animals. Daurisoline administered at 10-30 mg/kg did not prevent electroshock-induced convulsions in mice or rats, nor did it reduce the neuronal damage in hippocampal slices induced by a hypoxic/hypoglycemic insult in vitro by MCA occlusion in vivo. These observations do not support the hypothesis that P-type Ca2+ channels are promising drug targets for the acute treatment of epileptic convulsions and/or ischemic stroke.

Excitatory acoustic responses in the inferior colliculus of the rat are increased by GABAB receptor blockade.

This study sought to investigate the influence of GABAB receptor activation on acoustically induced excitation within the rat inferior colliculus. To this end, the GABAB receptor antagonist, CGP 35348, was applied systemically and iontophoretically. Single and multibarrel electrodes were used for extracellular recordings within the central nucleus of the inferior colliculus. The experimental model, a paired-pulse stimulus paradigm, applied two identical acoustic stimuli, 200 msec apart, evoking corresponding responses characterized by the second being consistently weaker than the first. Abolishment of the acoustically evoked response, following iontophoretic application of the GABAB receptor agonist, L-baclofen, verified the existence of GABAB receptors in all inferior colliculus cells tested. Intravenous application of CGP 35348 (200 mg/kg) evoked a 24% overall increase in stimulus responses. Likewise, a 13% increase in total evoked excitation was observed, following iontophoretic application. There was no significant reduction of inhibition on the second evoked response in the paired-pulse model, following either systemic or iontophoretic application of CGP 35348. This result implies that the decreased magnitude of the second response, with an interpulse interval of 200 msec, is not influenced by GABAB receptor mediated inhibition. These findings do indicate, however, that GABAB receptors play a small, but significant role during the processing of acoustic information, within the inferior colliculus.

Giant neurons in the rat reticular formation: a sensorimotor interface in the elementary acoustic startle circuit?

The mammalian acoustic startle response (ASR) is a relatively simple motor response that can be elicited by sudden and loud acoustic stimuli. The ASR shows several forms of plasticity, such as habituation, sensitization, and prepulse inhibition, thereby making it an interesting model for studying the underlying neuronal mechanisms. Among the neurons that compose the elementary startle circuit are giant neurons in the caudal pontine reticular nucleus (PnC), which may be good candidates for analyzing the neuronal basis of mammalian behavior. In a first step of this study, we employed retrograde and anterograde tracing techniques to identify the possible sources of input and the efferent targets of these neurons. In a second step, we performed intracellular recordings in vivo, followed by subsequent injections of HRP for morphological identification, thereby investigating whether characteristic features of the ASR are reflected by physiological properties of giant PnC neurons. Our observations demonstrate convergent, bilateral input from several auditory brainstem nuclei to the PnC, predominantly originating from neurons in the cochlear nuclear complex and the superior olivary complex. Almost no input neurons were found in the nuclei of the lateral lemniscus. As the relatively long neuronal response latencies in several of these auditory nuclei appear to be incompatible with the primary ASR, we conclude that neurons in the cochlear root nuclei most likely provide the auditory input to PnC neurons that is required to elicit the ASR. The giant PnC neurons have a remarkable number of physiological features supporting the hypothesis that they may be a neural correlate of the ASR: (1) they receive short-latency auditory input, (2) they have high firing thresholds and broad frequency tuning, (3) they are sensitive to changes in stimulus rise time and to paired-pulse stimulation, (4) repetitive acoustic stimulation results in habituation of their response, and (5) amygdaloid activity enhances their response to acoustic stimuli. Anterograde tracing showed that most giant PnC neurons are reticulospinal cells. Axon collaterals and terminal arbors were found in the reticular formation as well as in cranial and spinal motoneuron pools. The results of this study indicate that giant PnC neurons form a sensorimotor interface between the cochlear nuclear complex and cranial and spinal motoneurons. This neuronal pathway implies that the elementary acoustic startle circuit is composed of only three central relay stations and thus appears to be organized more simply than assumed in the past.

Multiphasic responses of cerebellar Purkinje cells to 1S,3R-ACPD: an in vivo study.

The effects of iontophoretically applied (1S,3R)-1-aminocyclopentane-1,3-dicarboxylic acid (1S,3R-ACPD), an agonist of metabotropic glutamate receptors, were examined in rat cerebellar Purkinje cells in vivo. Multibarrel electrodes were used for extracellular recordings of spontaneous single unit discharges and iontophoretic ejection of 1S,3R-ACPD. The effect of 1S,3R-ACPD depended on both the strength and the duration of the iontophoretic current. Application of the agonist with ejection currents at or slightly above the response threshold for up to 60 s resulted in an increased rate of action potential firing. With larger ejection currents of the same duration the initial increase in activity was followed by a depression and eventually a cessation of activity. In the transition phase between low frequency firing and firing arrest, Purkinje cells generated almost exclusively complex spikes. When the drug application was continued for longer durations (1-10 min) the initial response was followed by a characteristic cyclic firing pattern. These cycles consisted of alternating phases of mainly simple spike activity, predominantly complex spike activity and silent intervals. At the end of drug applications using large ejection currents, a prolonged period (on average 66 s) with almost no spiking activity was observed. This period ended with an abrupt onset of simple spike firing. These findings point to an important function of cerebellar metabotropic glutamate receptors in the regulation of Purkinje cell activity.

Phenylglycine derivatives antagonize the excitatory response to Purkinje cells to 1S,3R-ACPD: an in vivo and in vitro study.

The interactions of the phenylglycine derivatives (S)-4-carboxyphenylglycine (S-4CPG) and (R,S)-alpha-methyl-4-carboxyphenylglycine (MCPG) with responses of rat cerebellar Purkinje cells to (1S,3R)-1-aminocyclopentane-1,3-dicarboxylic acid (1S,3R-ACPD) were examined by intracellular recordings in acute cerebellar slices and extracellular recordings in vivo, using multibarrel electrodes. In vitro, both S-4CPG (100 microM to 1 mM) and MCPG (250 microM to 1 mM) reversibly and dose-dependently reduced an inward current induced by bath-applied 1S,3R-ACPD, an agonist at metabotropic glutamate receptors (mGluRs), in Purkinje cells voltage-clamped at -60 to -65 mV. S-4CPG applied at a concentration of 1 mM reduced the 1S,3R-ACPD induced current to 17% of control values but when applied alone also produced an inward current amounting to 26.8% of that induced by 1S,3R-ACPD. MCPG bath-applied at 250 microM, 500 microM, or 1 mM reduced the 1S,3R-ACPD-induced current to 85%, 56% or 3% of control values, respectively, and did not cause any current when applied alone even at a concentration of 1 mM. In vivo, iontophoretic application of 1S,3R-ACPD induced a transient increase followed by a decrease in the firing rate of Purkinje cells. The excitatory response of Purkinje cells to 1S,3R-ACPD was suppressed during ejection of either one of the phenylglycine derivatives, while the mechanism resulting in the decreased firing rate was not affected. Our observations demonstrate that both S-4CPG and MCPG antagonized the excitatory response of cerebellar Purkinje cells to 1S,3R-ACPD.(ABSTRACT TRUNCATED AT 250 WORDS)

Principal cells of the rat medial nucleus of the trapezoid body: an intracellular in vivo study of their physiology and morphology.

The medial nucleus of the trapezoid body (MNTB) is one of several principal nuclei in the superior olivary complex (SOC) of mammals. It is classically thought to function as a relay station between the contralateral ventral cochlear nucleus and the lateral superior olive (LSO), playing a role among those brainstem nuclei that are involved in binaural hearing. In order to characterise the physiology and morphology at the cellular level of the major neuronal component of the MNTB, the principal cells, we have analysed these neurons in rats in vivo using intracellular recordings and horseradish peroxidase-labelling. Our data demonstrate that MNTB principal cells, when being stimulated acoustically via the contralateral ear, show a phasic-tonic response with an onset latency of 3.5 ms and a suppression of their spontaneous activity following stimulus offset. These neurons have an axonal morphology whose complexity has not yet been described. All cells (n = 10) projected exclusively ipsilaterally and had terminal axonal arbors in a variety of auditory brainstem nuclei. At least two and maximally seven auditory targets were innervated by an individual cell. Each cell projected into the LSO and the superior paraolivary nucleus (SPN). Additional projections that were intrinsic to the SOC were often observed in the lateral nucleus of the trapezoid body and in periolivary regions, with only one cell projecting into the medial superior olive. Most, if not all, MNTB principal cells also had projections that were extrinsic to the SOC, as their axons ascended into the lateral lemniscus. In two neurons the ascending axon formed terminal arbors in the ventral nucleus of the lateral lemniscus, and the dorsal nucleus of the lateral lemniscus could be identified as a target of one neuron. The location of the cell bodies of the MNTB principal cells correlated with the neurons' best frequencies, thereby demonstrating a tonotopic organisation of the MNTB, with high frequencies being represented medially and low frequencies laterally. The axonal projections into the LSO and the SPN were also tonotopically organised and the alignment of the tonotopically organised and the alignment of the tonotopic axes was similar to that in the MNTB. Our results confirm previous data from other species and suggest that MNTB principal cells have a great amount of physiological and morphological similarities across mammalian species. Furthermore, the complexity of the axonal projections indicates that these neurons play a role in auditory information processing which goes far beyond their previously described classical role.

Giant neurons in the caudal pontine reticular formation receive short latency acoustic input: an intracellular recording and HRP-study in the rat.

The reticular formation is composed of heterogeneous cell populations with multiple functions. Among these multiple functions is the processing of sensory information in the context of behavior. The purpose of the present study was to identify and characterize neurons in the reticular formation of the rat that receive auditory input. In order to do so, we combined intracellular electrophysiology in vivo with intracellular injection of horseradish peroxidase, enabling us to correlate electrophysiology unequivocally with anatomy at the single cell level. We found that many neurons in the caudal pontine reticular nucleus (PnC), which we analyzed intracellularly, responded to acoustic stimuli and were excited at short latency (mean EPSP latency: 2.6 ms; mean spike latency: 5.2 ms). This short latency suggests a direct input from the cochlear nucleus, the first central nucleus of the auditory pathway. The morphology revealed that the acoustically driven PnC neurons have very large somata (mean diameter: 44.0 microns). They can therefore be referred to as "giant PnC neurons." Complex dendritic arbors extended from these neurons into the reticular formation and thus formed a large membrane surface for the integration of multimodal inputs. Most of the giant PnC neurons sent their axons caudally into the medial longitudinal fasciculus and can therefore be regarded as reticulospinal neurons. The results demonstrate that the giant reticulospinal PnC neurons are in a position to transmit acoustic information very quickly to spinal cord neurons and to receive converging input from other parts of the brain. They are thus good candidates for participation in the mediation and modulation of acoustically elicited behaviors, such as the short latency acoustic startle response.

Loss of the acoustic startle response following neurotoxic lesions of the caudal pontine reticular formation: possible role of giant neurons.

The effect of the excitotoxic N-methyl-D-aspartate agonist quinolinic acid in the caudal pontine reticular formation on the acoustic startle response was investigated in rats. Bilateral injections of 90 nmol of quinolinic acid led to large lesions in the reticular formation characterized by the loss of all neurons and a marked reduction or even abolition of the acoustic startle response; 18 nmol of quinolinic acid led to smaller lesions characterized by a selective loss of giant neurons within the caudal pontine reticular formation and a reduction of the startle amplitude. The partial correlation analysis revealed that the reduction of the amplitude of the acoustic startle response can be correlated with the loss of the giant neurons (r = 0.575; d.f. = 29; P less than 0.001) but not with the reduction of the number of all neurons (r = 0.207; d.f. = 29; P greater than 0.2) in the caudal pontine reticular formation. These findings were reconciled with electrophysiological and anatomical data indicating that the giant neurons in the caudal pontine reticular formation receive acoustic input and project to motoneurons of the spinal cord. It is concluded that the caudal pontine reticular formation is an important element of the startle pathway and that the giant reticulospinal neurons constitute an important part of the sensorimotor interface mediating this response.

Sensory neurons and motoneurons of the jaw-closing reflex pathway in rats: a combined morphological and physiological study using the intracellular horseradish peroxidase technique.

Motoneurons and muscle spindle afferents of the rat masseter muscle were physiologically and morphologically characterized. Their soma-dendritic morphology and axonal course were investigated using the intracellular horseradish peroxidase method. Following electrical stimulation of the masseter nerve, individual motoneurons were identified by antidromic all-or-none action potentials and individual sensory neurons by orthodromic action potentials. Using threshold separation an excitatory input from muscle spindles to a masseter motoneuron was demonstrated. The short latency difference of 0.34 ms between the mean orthodromic response in the sensory neurons and the beginning of the synaptic potential in the masseter motoneuron suggests a monosynaptic connection between the spindle afferents and the motoneurons. Following intrasomatic horseradish peroxidase injection large multipolar cell bodies of masseter motoneurons were found within the motor nucleus. Their positions corresponded to the topographic organization of the motor trigeminal nucleus as described in retrograde tracing studies. Dendrites of masseter motoneurons were complex and could be found far beyond the nuclear borders. Distal dendrites extended to the mesencephalic trigeminal nucleus, the supratrigeminal nucleus, the lateral lemniscus and the reticular formation. Within the reticular formation dendrites were seen in the intertrigeminal nucleus and the peritrigeminal zone. Unipolar cell bodies of muscle spindle afferents were found in the mesencephalic trigeminal nucleus after intra-axonal injection of horseradish peroxidase. For all reconstructed sensory neurons a similar axonal course was found. Axonal terminals were found ipsilateral in the motor trigeminal nucleus, indicating a direct connection between sensory neurons and motoneurons. Further collaterals were found ipsilateral in the supratrigeminal nucleus and caudal to the motor trigeminal nucleus in the parvocellular reticular nucleus alpha. Since the latter termination areas are important for bilateral control of jaw-movements, the muscle spindle afferents are likely to participate not only in a monosynaptic motor reflex, but also in more complex neuronal circuits involved in jaw-movements.

Morphological characterization of rat entorhinal neurons in vivo: soma-dendritic structure and axonal domains.

We used in vivo intracellular labeling with horseradish peroxidase in order to study the soma-dendritic morphology and axonal projections of rat entorhinal neurons. The cells responded to hippocampal stimulation with inhibitory postsynaptic potentials, and thus likely received direct or indirect hippocampal input. All cells (n = 24) showed extensive dendritic domains that extended in some cases for more than 1 mm. The dendrites of layer II neurons were largely restricted to layers I and II or layers I-III, while the dendrites of deeper cells could extend through all cortical layers. Computed 3D rotations showed that the basilar dendrites of deep pyramids extended roughly parallel to the cortical layering, and that they were mostly confined to the layer containing the soma and layers immediately adjacent. Total dendritic lengths averaged 9.8 mm +/- 3.8 (SD), and ranged from 5 mm to more than 18 mm. Axonal processes could be visualized in 21 cells. Most of these showed axonal branching within the entorhinal cortex, sometimes extensive. Efferent axonal domains were reconstructed in detail in 3 layer II stellate cells. All 3 projected axons across the subicular complex to the dentate gyrus. One of these cells showed an extensive net-like axonal domain that also projected to several other structures, including the hippocampus proper, subicular complex, and the amygdalo-piriform transition area. The axons of layer III and IV cells projected to the angular bundle, where they continued in a rostral direction. In contrast to the layer II, III and IV cells, no efferent axonal branches leaving the entorhinal cortex could be visualized in 5 layer V neurons. The data indicate that entorhinal neurons can integrate input from a considerable volume of entorhinal cortex by virtue of their extensive dendritic domains, and provide a further basis for specifying the layers in which cells receive synaptic input. The extensive axonal branching pattern seen in most of the cells would support divergent propagation of their activity.