Thyrotropin-releasing hormone selectively depresses glutamate excitation of cerebral cortical neurons.

The microiontophoretic application of thyrotropin-releasing hormone causes a selective reduction in neuronal excitation evoked by L-glutamate but not by acetylcholine in rat cerebral cortex. Thyrotropin-releasing hormone has no influence on the activity of acetylcholinesterase or on choline uptake and release from cerebral synaptosomes. This evidence for a selective interaction between a centrally acting peptide and an excitatory amino acid neurotransmitter may indicate a specific locus of thyrotropin-releasing hormone action at glutamate-activated receptor sites.

Beta amyloid gene duplication in Alzheimer's disease and karyotypically normal Down syndrome.

With the recently cloned complementary DNA probe, lambda Am4 for the chromosome 21 gene encoding brain amyloid polypeptide (beta amyloid protein) of Alzheimer's disease, leukocyte DNA from three patients with sporadic Alzheimer's disease and two patients with karyotypically normal Down syndrome was found to contain three copies of this gene. Because a small region of chromosome 21 containing the ets-2 gene is duplicated in patients with Alzheimer's disease, as well as in karyotypically normal Down syndrome, duplication of a subsection of the critical segment of chromosome 21 that is duplicated in Down syndrome may be the genetic defect in Alzheimer's disease.

Pharmacological modulation of cortisol secretion and dexamethasone suppression in Alzheimer's disease.

We have investigated the dexamethasone suppression of cortisol release in a group of 28 patients with senile dementia of the Alzheimer type (SDAT) after stimulation by physostigmine and clonidine, as compared with basal conditions. All patients but one had previously been evaluated with a depression symptom checklist and had submitted to a standard Dexamethasone Suppression Test (DST). SDAT patients showed normal baseline cortisol values measured at 4:00 PM. DST was reproducible, but nonsuppression did not appear to be a feature of the disease, nor of the dementia syndrome, although a majority of the most demented patients were found to be nonsuppressors. Physostigmine stimulated cortisol secretion in 20 of 24 cases, irrespective of the severity of dementia. Clonidine induced a secretion in 12 of 15 cases, but this was less than that observed after cholinergic stimulation. Physostigmine made cortisol release significantly less sensitive to the suppressive effect of dexamethasone than clonidine in SDAT. This double response should be tested as a possible predictor of a cholinergic therapeutic effect.

Tetrahydroaminoacridine and physostigmine increase cerebral glucose utilization in specific cortical and subcortical regions in the rat.

The effects of the anticholinesterases tetrahydroaminoacridine (THA) and physostigmine on local cerebral glucose utilization (LCGU) were studied in the conscious rat, using the autoradiographic [14C]deoxyglucose technique. THA (5 mg/kg i.p.) increased LCGU significantly in 8 of the 43 regions studied. A higher dose of THA (10 mg/kg) produced a metabolic activation in 19 of the 43 regions. LCGU increased in cortical areas (including parietal and temporal cortices), the septohippocampal system, the thalamus, the lateral habenula, the basolateral amygdala, the superior colliculus, and the substantia nigra. Scopolamine (4 mg/kg i.p.) reversed the THA-induced LCGU increase. Physostigmine (0.2 and 0.5 mg/kg) increased LCGU in 15 and 22 regions, respectively. The average magnitude of the change induced by 0.5 mg/kg of physostigmine was similar to that observed after THA at 10 mg/kg, but the topography of the effects was somewhat different. Physostigmine increased LCGU in the preoptic magnocellular area, the brainstem, and the cerebellum but not in the parietal cortex. The effects in the septohippocampal system were smaller than those induced by THA. The regional topography of the LCGU increase overlapped the distribution of the M2 muscarinic receptors and that of acetylcholinesterase activity. These data suggest that the major effects of THA and physostigmine on LCGU result from their anticholinesterase action.

Metabolic response to tacrine (THA) and physostigmine in the aged rat brain.

The effects of the centrally acting anti-cholinesterases tacrine (tetrahydroaminoacridine, THA) and physostigmine (PHY), on local cerebral glucose utilization (LCGU) have been studied in 27-month-old rats, using the autoradiographic [14C]deoxyglucose technique. THA (10 mg/kg i.p.) increased LCGU significantly in 13 of the 54 regions studied (24%) including insular, parietal, temporal, and retrosplenial cortices, septohippocampal system, thalamus, lateral habenula, and superior colliculus. In these regions, the average THA-induced increase in LCGU was 24% above control. The whole brain mean LCGU was not significantly increased. PHY (0.5 mg/kg) increased LCGU in 18% of the regions (average elevation, 23%). The whole brain mean LCGU increased by 7% (p < 0.05). The regional distributions of THA- and PHY-induced increases in LCGU were extremely similar and overlapped the distribution of the M2 muscarinic receptors and that of acetylcholinesterase activity, suggesting that the major effects of THA and PHY on LCGU result from their anticholinesterase action. As compared to those of 3-month-old rats, both the number of regions affected and the amplitude of the metabolic activation were significantly less in aged rats. However, the drugs were still active in old rats and compensated for the age-related hypometabolism in some brain areas.

Effects of metrifonate, a cholinesterase inhibitor, on local cerebral glucose utilization in young and aged rats.

The effects of the centrally acting anti-cholinesterase metrifonate (MFT) and its metabolite dichlorvos (2,2-dichlorovinyl dimethyl phosphate; DDVP) on local cerebral glucose utilization (LCGU) have been studied in 3- and 27-month-old rats, using the autoradiographic [14C]deoxyglucose technique. In 3-month-old rats, MFT (80 mg/kg i.p.) increased LCGU significantly in 17 of the 54 regions studied, including insular, cingulate, and temporal cortices, ventral hippocampus, thalamus, lateral habenula, substantia nigra, and superior colliculus. In these regions, the average MFT-induced increase in LCGU was 23% above control. The average hemispheric LCGU increased by 10% (p < 0.01). DDVP (5 mg/kg) increased LCGU in 19 regions (average increase 26%). The average hemispheric LCGU increased by 9% (p < 0.01). Regional distributions of MFT- and DDVP-induced increases in LCGU were similar and overlapped the distribution of the acetylcholinesterase activity. In 27-month-old rats, MFT was active in 18 regions (average increase 25%). The whole-brain mean LCGU increased by 10% (p < 0.01). MFT compensated for the age-related hypometabolism in some brain areas including insular, temporal, and retrosplenial cortices, substantia nigra, and superior colliculus. The effects of MFT on LCGU were preserved in old rats, at variance with other anticholinesterases (tacrine, physostigmine). Which are less active in the aged rat brain.

Age-related alterations in the properties of hippocampal pyramidal neurons among rat strains.

We compared age-related alterations in the electrophysiological and pharmacological properties of CA1 hippocampal pyramidal neurons in three strains of rats (Sprague-Dawley, Fisher 344, and Wistar) at 3-4 and 25-32 months of age, using the in vitro slice preparation. The most consistent age-related alterations in the properties of rat hippocampal neurons were: a decrease in membrane excitability, a decrease in the amplitude and duration of inhibitory postsynaptic potentials and a decreased sensitivity to the effect of the cholinergic agonist carbachol. In contrast, no consistent alterations in calcium-dependent events were observed in these strains of rats. The age-related changes in the duration of the afterhyperpolarizations (AHPs) were different (and even opposite) depending on the strain studied. Our results show that age-related changes observed in a given strain are not necessarily present in all strains of the same species.

Age-related changes in parvalbumin- and GABA-immunoreactive cells in the rat septum.

The calcium binding protein parvalbumin is present in GABAergic neurons of the medial septum-diagonal band of Broca (MS-DBB) region that project to the hippocampal formation. We examined the distribution pattern, the number, and the morphological features of the parvalbumin-containing cells (parv+) in the MS-DBB region of 2- to 3-, 8- to 9-, 15- to 16-, and 26- to 27-month-old Sprague-Dawley rats. A significant reduction in the number of parv+ cells was observed as a function of age. The mean somal area of the parv+ cells was significantly reduced in the 26- to 27-month-old rats. A significant reduction in the number of parv+ cells was also observed in the 26- to 27-month-old rats in the cingulate cortex, but not in the striatum or the hippocampal formation. No significant age-related changes were observed in the number of the GABA-immunoreactive cells in the MS-DBB region nor in the cingulate cortex. In conclusion, there is an age-related decrease in the number of parv+ cells, with no change in the number of GABA-immunoreactive cells in the MS-DBB region of the rat. Because GABA and parvalbumin are colocalized in the MS-DBB neurons, the results suggest that the level of parvalbumin is decreased, but that the cells are not lost.

Age-related changes in dopaminergic and serotonergic indices in the rat forebrain.

Age-related changes in the content of dopamine (DA), homovanillic acid (HVA), dihydroxyphenylacetic acid (DOPAC), 3-methoxytyramine (3-MT), serotonin (5-HT) and 5-hydroxyindoleacetic acid (5-HIAA) in anterior cerebral cortex, hippocampus and striatum of the rat have been investigated using HPLC with electrochemical detection. A significant decrease in HVA was observed in the striatum and hippocampus of the aged (27 months) animals, as compared to the controls (2.4 to 2.6 months). A significant decrease in DA levels was also observed in the hippocampus but not in the striatum. In contrast, the level of DA in the cerebral cortex was markedly increased in the aged animals. A concomitant increase in 3-MT level was observed. Finally the level of 5-HIAA was significantly increased in striatum and hippocampus.

NMDA receptor activation in the aged rat: electrophysiological investigations in the CA1 area of the hippocampal slice ex vivo.

The effects of aging on activation of N-methyl-D-aspartate (NMDA) receptors were studied in the CA1 field of hippocampal slices from young (2-4 months old) and aged (25-32 months old) Sprague-Dawley rats with the use of ex vivo extra- and intracellular electrophysiological recording techniques. No significant age-related changes of the unitary NMDA-receptor mediated excitatory postsynaptic potentials (EPSPs), recorded from the pyramidal cells after stimulation of the stratum radiatum in a magnesium-free medium and isolated in the presence of the non-NMDA receptor antagonist 6-cyano-7-nitroquinoxaline-2,3-dione, were found. Simultaneously, the magnitude of synaptic plasticity which involved NMDA receptor activation was not altered. No significant age-related modifications in the mechanisms controlling glutamate release and of postsynaptic NMDA receptor responsiveness were revealed. Considering the 30-40% decrease in NMDA binding sites in the aged hippocampus, our results suggest the occurrence of compensatory mechanisms which are discussed.

Septo-hippocampal neurons in the aged rat: relation between their electrophysiological and pharmacological properties and behavioral performances.

The performances of aged (24-26 months) rats in two behavioral tasks (passive avoidance and spontaneous alternation) have been studied. Subpopulations of old animals were found to be impaired in these tasks. Most of the impaired animals, however, were not impaired in both tasks. The properties of the septo-hippocampal neurons (SHNs) were subsequently studied in the same group of experimental animals, anesthetized with urethane, using electrophysiological techniques. The spontaneous activity of SHNs displaying a rhythmically bursting activity (RBA) was significantly higher in animals impaired in the spontaneous alternation task. The proportion of SHNs with RBA was significantly lower and the frequency of the RBA was higher in animals impaired in the passive avoidance task. The pharmacological properties of the SHNs were not significantly different in the various groups. The significance of these complex correlations with regard to the age-related alterations of SHNs properties is discussed.

Protein SP40,40-like Immunoreactivity in the Rat Brain: Progressive Increase With Age.

The pattern of distribution of SP40,40-like immunoreactive structures has been studied in the rat brain using a well-characterized polyclonal antibody raised against the SP40,40 protein. Protein SP40,40 is the human counterpart of the rat sulphated glycoprotein 2, whose mRNA shows widespread expression in the developing and mature brain. In young adult rats few immunoreactive structures were observed. Some immunoreactive neurons were found in the cingulate cortex, the arcuate and perifornical hypothalamic nuclei, as well as glial labelling in the hypothalamus. A striking increase in the number of immunoreactive cells was observed as a function of age. In 20 - 22-month-old rats, numerous immunoreactive cells were observed in the cingulate cortex, several thalamic and hypothalamic nuclei, the red nucleus, olivary nuclei, superior colliculus, and many cranial nerve nuclei. Whereas the immunoreactivity was restricted to a diffuse labelling of the cell bodies and processes in young rats, other forms of labelling were observed in aged rats: punctate cytoplasmic labelling and intensely stained granules with no visible cell membrane. A further increase in the density of the immunoreactive material was observed in 30 - 31-month-old rats. Double labelling experiments demonstrated that the SP40,40 immunoreactivity was almost exclusively located in neurons and not in glial cells (with the exception noted above). The distribution of SP40,40 immunoreactivity in aged rats did not coincide with the distribution of the microtubule-associated tau protein, OX42 or lipofuscin.

Loss of Calbindin-28K Immunoreactivity in Hippocampal Slices from Aged Rats: a Role for Calcium?

Calbindin-28K (CaBP) is a calcium-binding protein widely distributed in the brain. This protein appears to be involved in the sequestration and the translocation of intracellular free calcium. In this study, we have examined the distribution pattern of the structures immunoreactive for CaBP in the hippocampal formation from slices of young (4 months) and aged (24 - 27 months) rats previously submitted to electrophysiological measurements. We demonstrated a marked loss in the number of pyramidal cells immunoreactive for CaBP in aged rats as compared to young rats. A consistent decrease in the staining intensity was also revealed by optical density measurements. Some experiments have suggested that calcium homeostasis is modified in hippocampal neurons of aged rats. The loss of CaBP-like immunoreactivity (CaBP-LI) labelling could result from an increase in intracellular calcium concentrations. To support this hypothesis, we showed that in young rats (i) the CaBP-LI was enhanced in pyramidal neurons when the slice was preincubated in a calcium-free medium, and (ii) CaBP-LI was strongly decreased when the slice was preincubated in a high-calcium medium (5 mM) and when the entry of calcium into the cell was increased by a short application of an excitatory amino acid in the medium. Our results suggest that the loss of CaBP-LI in the hippocampus of aged rats could be due to an increase in intracellular calcium concentration. Preliminary observations of hippocampal slices at different times after induction of long-term potentiation (LTP) failed to show significant changes in CaBP immunoreactivity, suggesting that this calcium-binding protein is not directly involved in LTP processes.

Age-related changes in galanin-immunoreactive cells of the rat medial septal area.

Age-related changes in the cholinergic cells have been reported in the rat medial septal area. The neuropeptide galanin is colocalized with acetylcholine in the majority of the medial septal neurons. To assess possible age-related changes in the galanin-containing septal cells, we have examined, with immunohistochemical methods, the distribution pattern, density, and morphological features of galanin-containing cells in the rat medial septal nucleus (MS) and the nucleus of the diagonal band of Broca (DBB) in 1, 3-6, 9-12, 16-18, 24-27, and 28-30 month-old rats. A morphometric computerized analysis was also performed. In addition, the intensity of the immunolabelling was measured by densitometry. Galanin-like immunoreactivity (galanin-LI) was present in both the MS and the DBB. Our results clearly indicate a progressive age-related decrease in the number of galanin-positive cells throughout the MS-DBB complex. Our quantitative study revealed a significant loss of galanin-positive cells in the MS-DBB complex of 16-18 (50.4%), 24-27 (52.3%), and 28-30 (52.4%) month-old rats compared to 3-6 month-old animals. A non-significant reduction (28.6%) in galanin-LI cell number was observed in 3-6 month-old rats compared to 1 month-old animals. The morphometric analysis demonstrated a significant reduction (18%) in the surface of galanin-positive cells remaining in the 28-30 month-old group. Furthermore, a significant decrease in the immunolabelling intensity was consistently observed in animals of 16 month-old and older. To determine whether changes in galanin-positive cells were associated with cholinergic changes, the number of cells stained for acetylcholinesterase (AChE) was estimated in 3-6, 9-12, 16-18, and 24-27 month-old rats. There was a 43% decrease in the number of AChE-positive cells and a 71% loss of galanin-positive cells in 24-27 month-old rats compared to 3-6 month-old. The galanin-cell loss in the medial septal area was therefore associated with a parallel, although smaller, cholinergic septal cell loss.

Effects of neuropeptides on rat cortical neurons: laminar distribution and interaction with the effect of acetylcholine.

The effects of the microiontophoretic application of five different peptides (cholecystokinin octapeptide sulfated form, cholecystokinin octapeptide non-sulfated form, vasoactive intestinal polypeptide, angiotensin-II and substance P) on cortical neurons were studied in rats anaesthetized with urethane. Vertical electrode penetrations were made in the first somatic sensory cortex and the laminar position of the neurons determined by the reconstruction of the tracks based on extracellular dye deposits. The first type of effect observed was an excitation of some cortical neurons. These neurons were mostly found in infragranular layers, specially in layer Vb. Pyramidal tract neurons were more often excited by peptides than the cortical population taken as a whole. Substance P excited the largest percentage of neurons, followed by vasoactive intestinal polypeptide and cholecystokinin octapeptide sulfated form, whereas angiotensin II and cholecystokinin octapeptide non-sulfated form were the least potent in terms of frequency of neurons excited as well as of amplitude of the responses. The vast majority of the neurons excited by a peptide could also be excited by acetylcholine. A second and independent effect of peptides was observed: the neuronal excitation induced by acetylcholine could be depressed by the simultaneous application of peptide. This depressing effect was also the most frequently observed with substance P, followed by cholecystokinin and vasoactive intestinal polypeptide.

Cortical projections of the nucleus of the diagonal band of Broca and of the substantia innominata in the rat: an anatomical study using the anterograde transport of a conjugate of wheat germ agglutinin and horseradish peroxidase.

The projections of the nucleus of the diagonal band of Broca and of the substantia innominata to the cerebral cortex were studied in the rat, using the anterograde transport of wheat germ agglutinin conjugated with horseradish peroxidase. Following diagonal band injections, fibers were observed ascending in the septum and reaching the cingulate cortex. They had a rostrocaudal, horizontal direction, mostly in layer VI and could be followed over long distances on sagittal sections. The fibers gave off collaterals which were seen ascending in the cerebral cortex and reaching more superficial layers. Following substantia innominata injections, fibers were observed to take two routes: the first one identical to the one described above and a second through the caudate-putamen reaching the temporo-insular cortex. Terminal fields had a more diffuse distribution following substantia innominata than following diagonal band injections. No clear laminar pattern of termination was observed. However the density of terminals was higher in layers IV, V and VI than in layers I, II and III. Since the conjugate used is not taken up by fibers of passage, this pattern of connection is believed to reflect the organization of the projection of the nucleus of the diagonal band and of the substantia innominata to the cerebral cortex.

Cerebral neocortical neurons in the aged rat: spontaneous activity, properties of pyramidal tract neurons and effect of acetylcholine and cholinergic drugs.

The properties of cortical cerebral neurons have been studied and compared in 2, 22 and 26 month-old Sprague-Dawley rats, using electrophysiological techniques. The mean spontaneous activity of the neurons in old animals (unidentified as well as pyramidal tract neurons) was not different from that of young adult rats. In contrast the mean latency of the antidromic response of pyramidal tract neurons to pyramidal tract stimulation was significantly longer in 26 month-old animals. No difference was observed in the effects of the excitatory amino acid glutamate applied by iontophoresis. The percentage of cortical neurons excited by the iontophoretic application of acetylcholine was similar in young and old animals. Neither the laminar distribution, nor the individual sensitivity of these neurons to acetylcholine were found to be modified. The pharmacological properties of the acetylcholine-induced excitations were unchanged, exhibiting muscarinic as well as nicotinic properties. These results are consistent with the suggestion that the impairment of the cholinergic system with aging is for a large part presynaptic. They also emphasize the fact that several physiological and pharmacological properties of the cerebral cortical neurons show little change with age in Sprague-Dawley rats.

Excitatory effect of acetylcholine on different types of neurons in the first somatosensory neocortex of the rat: laminar distribution and pharmacological characteristics.

In rats anaesthetized with either urethane, pentobarbital or fluothane the effects of acetylcholine, cholinergic agonists and antagonists (applied by iontophoresis) were studied on single cortical neurons of first somatosensory region. The laminar distribution of the neurons excited by acetylcholine was determined by the reconstruction of each electrode track based on a dye-deposit made at the last recording site. Neurons were identified using antidromic stimulation of the pyramidal tract, the ventrobasal thalamus and the corpus callosum. Neurons excited by acetylcholine could be segregated into two groups: one encompassing layer Vb and the upper part of layer VI, the other more deeply located at the limit between the cerebral cortex and the subjacent white matter. Neuronal responses to glutamate and nicotine, unlike those to actylcholine were evenly distributed in the cortex. Pyramidal tract neurons had corticothalamic neurons were frequently excited by acetylcholine and were shown to be located with the first group of acetylcholine sensitive neurons. Commissural neurons were rarely excited by acetylcholine and were not restricted to either group. The analysis of neuronal responses to acetylcholine and various agonists (carbachol, nicotine, acetyl-beta-methylcholine, carbamyl-beta-methylcholine, butyrylcholine) and antagonists (atropine, mecamylamine) revealed a prominent but not exclusive muscarine character. It is included (i) that cortical neurons of first somatosensory cortex which are excited by acetylcholine belong to two populations, one consisting, at least in part, of projection neurons (upper group) and the other of interneurons (lower group); (ii) that cortical acetylcholine receptors are of a 'mixed' type strongly weighted toward the muscarinic side.

Cholinergic and peptidergic projections from the medial septum and the nucleus of the diagonal band of Broca to dorsal hippocampus, cingulate cortex and olfactory bulb: a combined wheatgerm agglutinin-apohorseradish peroxidase-gold immunohistochemical study.

We have examined the distribution pattern and the density of various neuropeptide, neurotransmitter and enzyme containing neurons in the rat medial septum and the nucleus of the diagonal band of Broca to assess their possible involvement in the septohippocampal, septocortical and septobulbar pathways. Immunohistochemical methods were combined with the retrograde transport of a protein-gold complex injected in the hippocampus, the cingulate cortex or the olfactory bulb. Cholinergic neurons were the most numerous. Galanin-positive neurons were about two or three times less numerous than cholinergic cells. Both these cell types had a similar location though the choline acetyl transferase-like immunoreactive cells extended more caudally in the horizontal limb of the nucleus of the diagonal band of Broca. Immunoreactive cells for other neuroactive substances were few (calcitonin gene-related peptide, luteinizing hormone releasing hormone. [Met]enkephalin-arg-gly-leu) or occasional (dynorphin B, vasoactive intestinal polypeptide, somatostatin, neurotensin, cholecystokinin, neuropeptide Y and substance P). No immunoreactive cells for bombesin, alpha atrial natriuretic factor, corticotropin releasing factor, 5-hydroxytryptamine, melanocyte stimulating hormone, oxytocin, prolactin, tyrosine hydroxylase or arg-vasopressin were present. Choline acetyltransferase- and galanin-like immunoreactive cells densely participate to septal efferents. Cholinergic neurons constituted the bulk of septal efferent neurons. Galanin-positive cells were 22% of septohippocampal, 8% of septocortical, and 9% of septobulbar neurons. Galanin containing septohippocampal neurons were found in the medial septum and the nucleus of the diagonal band of Broca; galanin-positive septobulbar and septocortical cells were limited to the nucleus of the diagonal band of Broca. Occasional double-labellings were noticed with some peptides other than galanin. Luteinizing hormone-releasing hormone, calcitonin gene-related peptide and enkephalin were the most often observed; some other projecting cells stained for vasoactive intestinal polypeptide or dynorphin B. Luteinizing hormone-releasing hormone, calcitonin gene-related peptide and enkephalin were observed in septohippocampal neurons; luteinizing hormone-releasing hormone and vasoactive intestinal peptide were observed in septocortical neurons and calcitonin gene-related peptide, luteinizing hormone-releasing hormone and dynorphin B were observed in septo-bulbar cells. These results show that, in addition to acetylcholine, galanin is a major cellular neuroactive substance in septal projections to the hippocampus, the cingulate cortex and the olfactory bulb. The presence of septal projecting neurons immunoreactive for other peptides shows that a variety of distinct peptides may also participate, but in a smaller number, to septal efferent pathways.

Arousal-dependent properties of medial septal neurons in the unanesthetized rat.

We have performed a qualitative and quantitative analysis of the electrophysiological properties of medial septal neurons in the unanesthetized rat. The rat's head was held in a stereotaxic apparatus by a painless head-restrained system that was implanted seven days prior to the recording sessions. Extracellular recordings were made in a mixed population of antidromically identified septohippocampal neurons and unidentified medial septal neurons in different states of arousal and in response to peripheral and reticular stimulations. The spontaneous activity as well as the percentage of rhythmically bursting septal neurons varied significantly according to the state of arousal. Higher values were noted in paradoxical sleep (28 imp/s and 94% of bursting neurons) as compared with wakefulness with hippocampal theta rhythm (17.4 imp/s and 64.2% of bursting neurons) and slow wave sleep (12.3 imp/s and 8% of bursting neurons). The frequency of the bursts was significantly higher during paradoxical sleep. In individual medial septal neurons, arousing stimuli and paradoxical sleep could induce rhythmic bursting activity in previously non-bursting neurons provided that they were fast-firing neurons. No differences were noted in the functional characteristics of neurons in the medial septal nucleus as compared with the diagonal band of Broca. When the unanesthetized rats were compared with a group of urethane-anesthetized rats, the spontaneous activity was higher and more irregular in the absence of anesthesia. The percentage of the bursting neurons was significantly lower in the unanesthetized rats (32.3% vs 43.3%). However, the frequency of the bursts was higher (5.9 +/- 0.1 Hz vs 3.5 +/- 0.1 Hz). Since the patterns of activity of medial septal neurons fluctuate in different physiologically relevant states, previous classifications of these neurons made by ourselves and other authors, in urethane-anesthetized rats, may not be appropriate.