Septohippocampal pathways contribute to system consolidation of a spatial memory: sequential implication of GABAergic and cholinergic neurons.

Studies of the neuropharmacological substrates of spatial memory formation have focused on the contribution of septohippocampal pathways. Although these pathways include, among others, cholinergic and GABAergic fibers innervating the hippocampus, research has essentially been oriented towards the role of their cholinergic component. Recently, a few studies investigated the role of GABAergic septohippocampal projections. These only focused on almost immediate or recent memory and yielded discrepant results. GABAergic lesions impaired learning or had no effects. Given the role of the hippocampus in memory consolidation and the potential modulatory influence of the septum on hippocampal function, it is relevant to study the role of the septohippocampal interface in memory stabilization. We performed investigations with relatively selective lesions of GABAergic (using oxerin-saporin) or/and cholinergic (using 192 IgG-saporin) medial septum/vertical limb of the diagonal band of Broca (MS/vDBB) neurons in rats, and assessed acquisition of a spatial memory and its subsequent recall in the water maze. Following a 6-day training phase during which all groups improved performance to comparable levels, retention was tested 1, 5, or 25 days later. At the 1-day delay, all groups performed above chance and did not differ significantly among each other. At the 5-day delay, only rats with GABAergic or combined lesions exhibited a retention deficit. At the 25-day delay, all three lesion groups performed at chance level; in these groups, performance was significantly lower than that found in sham-operated rats. Immunochemical and histochemical verifications of the lesion extent/selectivity showed extensive GABAergic damage after intraseptal orexin-saporin infusions or cholinergic damage after 192 IgG-saporin infusions, with relatively limited damage to the other neurotransmitter system. Our data show that GABAergic and cholinergic septohippocampal neurons both contribute to memory stabilization, and could do so in a sequential way: GABAergic processes could be engaged at an earlier stage than cholinergic ones during system consolidation of a spatial memory.

MRI study of the cavum septum pellucidum in obsessive-compulsive disorder.

The cavum septum pellucidum (CSP), a putative marker of neurodevelopmental anomaly, has been associated with an increased risk of several psychiatric disorders. The purpose of this study was to evaluate the CSP in patients with obsessive-compulsive disorder (OCD) compared with healthy control subjects. Seventy-one patients with OCD and 71 healthy volunteers matched for age and sex were evaluated with magnetic resonance imaging. We evaluated the CSP using criteria employed in previous studies: presence of the CSP, length of the CSP, and overall size of the CSP, measured in five grades, ranging from grades 0 (no CSP) to 4 (severe CSP). We evaluated OCD symptom severity using the Yale-Brown Obsessive-Compulsive Scale (Y-BOCS). The CSP presence was significantly greater in the OCD group (60.6%) than in control subjects (29.6%), and CSP size grade was significantly larger in the OCD group (chi(2) = 15.609, P = 0.004). CSP length showed no significant group difference. Among patients with OCD, those with a CSP had higher scores on the obsession subscale of the Y-BOCS than those without a CSP (Z = -2.358, P = 0.018), while they did not show significant difference from those without a CSP in the compulsion subscale of the Y-BOCS, age, duration of illness, or age at onset. These results indicate that neurodevelopmental alterations in midline structures might contribute to the pathogenesis of OCD.

In vitro activation of the medial septum-diagonal band complex generates atropine-sensitive and atropine-resistant hippocampal theta rhythm: an investigation using a complete septohippocampal preparation.

The medial septum and diagonal band complex (MS-DB) is believed to play a key role in generating theta oscillations in the hippocampus, a phenomenon critical for learning and memory. Although the importance of the MS-DB in hippocampal theta rhythm generation is generally accepted, it remains to be determined whether the MS-DB alone can generate hippocampal oscillations or is only a transducer of rhythmic activity from other brain areas. Secondly, it is known that hippocampal theta rhythm can be separated into an atropine-sensitive and insensitive component. However, it remains to be established if the MS-DB can generate both types of rhythm. To answer these questions, we used a new in vitro rat septohippocampal preparation placed in a hermetically separated two side recording chamber. We showed that carbachol activation of the MS-DB generated large theta oscillations in the CA1 and CA3 regions of the hippocampus. These oscillations were blocked by applying either the GABA(A) receptor antagonist bicuculline or the AMPA/kainate antagonist DNQX to the hippocampus. Interestingly, the application of the muscarinic receptor antagonist atropine produced only a partial decrease in the amplitude, without modification of the frequency, of theta. These results show for the first time, that upon optimal excitation, the MS-DB alone is able to generate hippocampal oscillations in the theta frequency band. Moreover, these MS-DB generated theta oscillations are mediated by muscarinic and nonmuscarinic receptors and have a pharmacological profile similar to theta rhythm observed in awake animals.

Seizures and electroencephalography findings in 61 patients with fetal alcohol spectrum disorders.

Fetal alcohol spectrum disorders (FASD) cause neurodevelopmental abnormalities. However, publications about epilepsy and electroencephalographic features are scarce. In this study, we prospectively performed electroencephalography (EEG) and brain magnetic resonance (MR) imaging in 61 patients with diagnosis of FASD. One patient had multiple febrile seizures with normal EEGs. Fourteen children showed EEG anomalies, including slow background activity and interictal epileptiform discharges, focal and/or generalized, and 3 of them had epilepsy. In one patient, seizures were first detected during the EEG recording and one case had an encephalopathy with electrical status epilepticus during slow sleep (ESES). Focal interictal discharges in our patients did not imply the presence of underlying visible focal brain lesions in the neuroimaging studies, such as cortical dysplasia or polymicrogyria. However, they had nonspecific brain MR abnormalities, including corpus callosum hypoplasia, vermis hypoplasia or cavum septum pellucidum. The latter was significantly more frequent in the group with EEG abnormal findings (p < 0.01).

Low intracerebroventricular doses of cholinotoxin AF64A do not affect the morphology of gonadotropin hormone-releasing hormone (GnRH)-immunoreactive fibers in the rat septum.

Ethylcholine aziridinium (AF64A) induces cholinergic lesion in animal models of AD. Although higher concentrations of AF64A are known to induce nonspecific, cholinergic, and non-cholinergic lesions, low concentrations are believed to be selectively cholinotoxic. However, morphological evidence of this phenomenon has not been demonstrated yet. The present study demonstrates that while AF64A damaged septal cholinergic fibers, periventricular GnRH-immunoreactive fibers remained intact, confirming the highly selective cholinotoxicity of AF64A at appropriate concentrations.

Restoration of peripheral V2 receptor vasopressin signaling fails to correct behavioral changes in Brattleboro rats.

Beside its hormonal function in salt and water homeostasis, vasopressin released into distinct brain areas plays a crucial role in stress-related behavior resulting in the enhancement of an anxious/depressive-like state. We aimed to investigate whether correction of the peripheral symptoms of congenital absence of AVP also corrects the behavioral alterations in AVP-deficient Brattleboro rats. Wild type (WT) and vasopressin-deficient (KO) male Brattleboro rats were tested. Half of the KO animals were treated by desmopressin (V2-receptor agonist) via osmotic minipump (subcutaneous) to eliminate the peripheral symptoms of vasopressin-deficiency. Anxiety was studied by elevated plus maze (EPM), defensive withdrawal (DW) and marble burying (MB) tests, while depressive-like changes were monitored in forced swimming (FS) and anhedonia by sucrose preference test. Cell activity was examined in septum and amygdala by c-Fos immunohistochemistry after 10 min FS. KO rats spent more time in the open arm of the EPM, spent less time at the periphery of DW and showed less burying behavior in MB suggesting a reduced anxiety state. KO animals showed less floating behavior during FS revealing a less depressive phenotype. Desmopressin treatment compensated the peripheral effects of vasopressin-deficiency without a significant influence on the behavior. The FS-induced c-Fos immunoreactivity in the medial amygdala was different in WT and KO rats, with almost identical levels in KO and desmopressin treated animals. There were no differences in central and basolateral amygdala as well as in lateral septum. Our data confirmed the role of vasopressin in the development of affective disorders through central mechanisms. The involvement of the medial amygdala in the behavioral alterations of vasopressin deficient animals deserves further attention.

Sites of behavioral and neurochemical action of ACTH-like peptides and neurohypophyseal hormones.

In order to localize the site of action of neuropeptides in relation to their effects on behavior and memory various approaches have been used. As a result of studies using rats bearing lesions in different areas of the limbic system as well as of studies in which neuropeptides were locally applied into various areas of the brain it appeared that the limbic system (amygdala, hippocampus, septum and some thalamic areas) plays an essential role in the effect of vasopressin and ACTH and their derivatives on behavior and memory. Neurochemical studies generally indicate that changes occur in catecholamine utilization in these various limbic regions upon administration of these neuropeptides. It can be concluded that the effects of vasopressin in the terminal regions of the coeruleo-telencephalic noradrenalin system correlate with its effects on consolidation of memory. It is likely that the effects of vasopressin on other transmitter systems (e.g. dopamine in the amygdala and serotonin in the hippocampus) correspond with the effect of this neuropeptide on retrieval processes. In addition, regional differences in biotransformation of the neurohypophyseal hormones suggest that different patterns of behaviorally active fragments of these peptides may be present locally in the brain.

Gonadal steroids target AMPA glutamate receptor-containing neurons in the rat hypothalamus, septum and amygdala: a morphological and biochemical study.

Interactions between glutamate and gonadal steroids are involved in the regulation of limbic and hypothalamic functions. We hypothesized that hormonal signals affect excitatory neurotransmission by regulating the expression of glutamate receptors (GluR) in limbic and hypothalamic regions. To test this hypothesis, first, the coexpression of dl-alpha-amino-3-hydroxy-5-methyl-4-isoxazone-propionate (AMPA) GluR1, GluR2/3, and androgen receptors or estrogen receptors was revealed in the same cells of septal, amygdaloid, and hypothalamic areas by double immunocytochemistry. The highest incidence of co-localization was detected in hypothalamic regions. To demonstrate a regulatory role of testosterone or estradiol on AMPA receptor expression, the hormonal milieu of male and female rats was manipulated by gonadectomy and hormonal treatment. GluR1 and GluR2/3 expression was assessed by Western blots. Statistical analysis demonstrated that testosterone and estradiol have a stimulatory influence on the expression of AMPA receptors in the hypothalamus. The regulatory effect of estradiol on AMPA receptors was found to be site and gender specific: after estradiol treatment, samples taken from the hypothalamus contained increased levels of GluR1 and GluR2/3, whereas in the septum, bed nucleus and amygdala, no changes could be detected. Furthermore, the increase in hypothalamic GluR 2/3 levels was two times higher in females, compared with males, whereas the changes in hypothalamic GluR 1 levels showed no sex differences. Our results support the hypothesis that the interaction between gonadal steroids and glutamate involves hormone regulation of GluR. This mechanism seems to be gender and site specific, suggesting that excitatory neurotransmission and related physiological mechanisms also may be distinctly different in males and females.

Novel characteristics of glutamate-induced cell death in primary septohippocampal cultures: relationship to calpain and caspase-3 protease activation.

Studies examined the phenotypic characteristics of glutamate-induced cell death and their relationship to calpain and caspase-3 activation. Cell viability was assessed by fluorescein diacetate and propidium iodide staining and lactate dehydrogenase release. Calpain and caspase-3 activity was inferred from signature proteolytic fragmentation of alpha-spectrin. Characterization of cell death phenotypes was assessed by Hoechst 33258 and DNA fragmentation assays. Exposure of septohippocampal cultures to 1.0, 2.0, and 4.0 mmol/L glutamate induced a dose-dependent cell death with an LD50 of 2.0 mmol/L glutamate after 24 hours of incubation. Glutamate treatment induced cell death in neurons and astroglia and produced morphological alterations that differed from necrotic or apoptotic changes observed after maitotoxin or staurosporine exposure, respectively. After glutamate treatment, cell nuclei were enlarged and eccentrically shaped, and aggregated chromatin appeared in a diffusely speckled pattern. Furthermore, no dose of glutamate produced evidence of internucleosomal DNA fragmentation. Incubation with varying doses of glutamate produced calpain and caspase-3 activation. Calpain inhibitor II (N-acetyl-Leu-Leu-methionyl) provided protection only with a narrow dose range, whereas carbobenzoxy-Asp-CH2-OC(O)-2,6-dichlorobenzene (Z-D-DCB; pan-caspase inhibitor) and MK-801 (N-methyl-D-aspartate receptor antagonist) were potently effective across a wider dose range. Cycloheximide did not reduce cell death or protease activation.

Intraventricular nerve growth factor administration prevents lesion-induced loss of septal cholinergic neurons in aging rats.

In young adult rats transection of the fimbria results in loss of cholinergic cell bodies in the septum and this lesion-induced loss is prevented by intraventricular administration of NGF. The present study examined whether NGF administration is equally effective in aging animals. Eighteen-month-old rats received fimbrial transections and were given intraventricular injections of NGF during four weeks. Septal cholinergic neurons were then visualized using NGF receptor immunohistochemistry, which represents a reliable marker for cholinergic neurons in the septal area. The fimbrial transections reduced the number of septal NGF receptor-positive cells to a similar extent as in young animals. NGF treatment of aging rats protected these cells as effectively as in young adult rats.

Nerve growth factor induces galanin gene expression in the rat basal forebrain: implications for the treatment of cholinergic dysfunction.

Nerve growth factor (NGF) is a potential treatment for cholinergic dysfunction associated with Alzheimer's disease (AD). In rats, NGF activates gene expression of the acetylcholine synthetic enzyme choline acetyltransferase (ChAT) and prevents age- and lesion-induced degeneration of basal forebrain (BF) cholinergic neurons. Cholinergic neurons in the BF coexpress galanin (GAL), a neuropeptide that has been shown to impair performance on memory tasks possibly through the inhibition of cholinergic memory pathways. NGF up-regulates both ChAT and GAL gene expression in cultured pheochromocytoma cells; however, the effect of chronic in vivo NGF administration on GAL gene expression within the BF has not been studied. We used in situ hybridization and quantitative autoradiography to assess GAL and ChAT gene expression within the BF of adult male rats following chronic intracerebroventricular infusion of NGF or cytochrome c. We now report that, in addition to stimulating ChAT gene expression, NGF strongly up-regulated the GAL gene in the rat cholinergic BF. NGF had no effect on GAL gene expression in other noncholinergic forebrain regions. NGF induction of GAL gene expression in the BF was specific, because gene expression for another neuropeptide, neurotensin, present within noncholinergic BF neurons was unchanged. Our data provide the first evidence that in vivo NGF administration up-regulates GAL gene expression in the cholinergic BF. These results suggest that the concurrent induction of GAL in the BF could limit the ameliorating actions of NGF on cholinergic dysfunction.

Immunohistochemical localization of caffeine-induced c-Fos protein expression in the rat brain.

Although caffeine is the most widely used central nervous system stimulant, the neuronal populations and pathways mediating its stimulant effects are not well understood. Using c-Fos protein as a marker for neuronal activation, the present study investigated the pattern of c-Fos induction at 2 hours after low locomotor-stimulant doses (1, 5, 10, and 30 mg/kg, i.p.) of caffeine and compared them with those after a higher dose (75 mg/kg, i.p.) or saline injection in adult male rats. Fos-immunoreactive neurons were counted in selected nuclei across the entire brain. Caffeine induced an increase in locomotor activity in a dose-dependent manner up to doses of 30 mg/kg and a decline at 75 mg/kg. Quantitative analysis of Fos-immunoreactive neurons indicated that no structures showed significant Fos expression at doses below 75 mg/kg or a biphasic pattern of Fos expression, as in locomotion. In contrast, caffeine at 75 mg/kg induced a significant increase compared with the saline condition in the number of Fos-immunoreactive neurons in the majority of structures examined. The structures included the striatum, nucleus accumbens, globus pallidus, and substantia nigra pars reticulata and autonomic and limbic structures including the basolateral and central nuclei of the amygdala, paraventricular and supraoptic hypothalamic nuclei, periventricular hypothalamus, paraventricular thalamic nuclei, parabrachial nuclei, locus coeruleus, and nucleus of the solitary tract. The locomotor-enhancing effects of low doses of caffeine did not appear to be associated with significant Fos expression in the rat brain.

The putative atypical antipsychotic drug amperozide preferentially increases c-fos expression in rat medial prefrontal cortex and lateral septum.

The effects of acute, systemic administration of the putative atypical antipsychotic drug amperozide on c-fos expression in the rat forebrain were studied by means of Fos immunohistochemistry. Amperozide significantly increased the number of Fos-immunoreactive nuclei in the medial prefrontal cortex and the lateral septum but not in the nucleus accumbens (shell or core), the striatum, or the amygdala. With the exception of the nucleus accumbens-shell, where amperozide failed to produce statistically significant increases, the regional distribution of Fos immunoreactivity following amperozide was similar to that induced by atypical, but not by typical, antipsychotic drugs. In addition, after amperozide the number of Fos-positive nuclei was higher in the nucleus accumbens than in the dorsolateral striatum, a characteristic that is common to all known atypical antipsychotic agents.

Comparative effects of the ACTH 4-9 analogue (ORG 2766), ACTH 4-10 and [D-Phe7] ACTH 4-10 on medial septal self-stimulation behaviour in rats.

Experiments were performed to examine the effects of various analogues of ACTH on electrical self-stimulation behaviour elicited from the medial septal area using an ascending or descending sequence of stimulus intensities within a session. When an ascending sequence of threshold multiples was used ACTH 4-10 and the ACTH 4-9 analogue (ORG 2766) enhanced level pressing for low intensity stimulation but attenuated self-stimulation at greater current intensities. The analogue ORG 2766 appeared to be a thousand times more potent than ACTH 4-10; [D-Phe7] ACTH 4-10 inhibited the response rate at threshold level but was inactive at greater current intensities. The same effect was found following administration of ORG 2766 in a dose which was 20 times greater (1 microgram/rat) than that used in the first experiments. Lever pressing was not affected by treatment with ACTH 4-10 or ORG 2766 when a descending sequence of stimulus intensities was used within a session. Thus, ACTH-related peptides may affect motivational processes involved in self-stimulation rather than the reward of the stimulation per se. It is suggested that although ORG 2766 mimicked the action of ACTH 4-10 this synthetic peptide may have additional behavioural properties.

Serotonin mediates cardiovascular responses to acetylcholine, bradykinin, angiotensin II and norepinephrine in the lateral septal area of the rat brain.

The infusion of acetylcholine, bradykinin, angiotensin II, norepinephrine and serotonin into the lateral septal area produced a dose-dependent increase of arterial blood pressure and heart rate. A pattern of inhibition of these cardiovascular responses, produced by pretreatment of the lateral septal area with phentolamine, 6-hydroxydopamine, methysergide and 5,7-dihydroxytryptamine was disclosed. These results suggest that the effects of acetylcholine, bradykinin and partially of angiotensin II, depend on the release of norepinephrine and the actions of this neurotransmitter in turn depend on the integrity of the serotonergic system in the lateral septal area.

Thermoregulatory actions of centrally-administered vasopressin in the rat.

Dependent upon the route and/or site of administration, arginine vasopressin (AVP) evoked a number of thermoregulatory actions in the conscious rat. Infused into a lateral cerebral ventricle, arginine vasopressin produced short-lasting hypothermia of rapid onset. Injected into the preoptic area, arginine vasopressin caused long-lasting hyperthermia of rapid onset that was antagonized by the prior administration of a V1 receptor antagonist, [d(CH2)5 Tyr(Me)AVP]. Injections of arginine vasopressin into the nucleus accumbens, ventral septal area, substantia innominata and the dorsomedial hypothalamus were without effect on body temperature. Although the antipyretic action of arginine vasopressin within the ventral septal area has been well documented, these findings provide further evidence that this peptide exerts additional thermoregulatory actions that are both neuroanatomically and functionally specific.

Lesion of septal-hippocampal neurons with 192 IgG-saporin alters function of M1 muscarinic receptors.

Cholinergic neurons projecting from the medial septum to the hippocampus were lesioned with the selective neurotoxin 192 IgG-saporin. Injection of 300 ng of 192 IgG-saporin into the medial septum produced a 60% decrease in choline acetyltransferase activity. M1 muscarinic receptor function was examined by measuring enhancement of evoked release of norepinephrine from rat hippocampal slices by the M1 selective agonist McN-A-343. In hippocampal slices from rats which were lesioned with 192-saporin, the response to McN-A-343 was reduced compared to sham-operated controls. Pirenzepine binding demonstrated no change in M1 receptor number or affinity. However, the curve for displacement of pirenzepine by the muscarinic agonist oxotremorine-M was shifted to the right in hippocampal tissue from lesioned rats. This shift was identical to that produced by addition of the non-hydrolyzable GTP analogue GppNHp, which uncouples the M1 muscarinic receptor/G-protein complex. These results suggest that lesion of septal-hippocampal cholinergic inputs causes uncoupling of the M1 muscarinic receptor, decreasing responsiveness to stimulation. These findings are similar to reports of decreased M1 muscarinic receptor coupling to G-proteins and loss of function in Alzheimer's disease. The 192 IgG-saporin lesion may provide a viable animal model in which to study uncoupling of G-proteins and M1 muscarinic receptors.

Fimbria-fornix transection and excitotoxicity produce similar neurodegeneration in the septum.

Fimbria-fornix transection produces neuronal injury in the septum. This cellular pathology is characterized by somatodendritic vacuolar abnormalities in neurons. Because these cellular changes are reminiscent of some of the morphological abnormalities seen with glutamate receptor-mediated excitoxicity, we tested whether excitotoxic injury to the septal complex of adult rats mimics the degeneration observed within the dorsolateral septal nucleus and medial septal nucleus following fimbria-fornix transection. The septal complex was evaluated at various time-points (6 h to 14 days) by light and electron microscopy following unilateral injection of the N-methyl-D-aspartate receptor agonist quinolinate or the non-N-methyl-D-aspartate receptor agonist kainate, and the morphological changes observed were compared to those abnormalities in the medial septal nucleus and dorsolateral septal nucleus at three to 14 days after fimbria-fornix transection. The patterns of cytoplasmic abnormalities and vacuolar injury were morphologically similar in the somatodendritic compartment of neurons following excitotoxicity and axotomy paradigms. These similarities were most evident when comparing the persistently injured neurons in the penumbral regions of the excitotoxic lesions at one to 14 days recovery to neurons in the medial septal nucleus and dorsolateral septal nucleus at seven and 14 days after fimbria-fornix transection. Pretreatment with the N-methyl-D-aspartate receptor antagonist dizocilpine maleate prior to unilateral fimbria-fornix transection attenuated the somatodentritic vacuolar damage found within the ipsilateral dorsolateral and medial septal nuclei at 14 days recovery. Because glutamate is the principal transmitter of hippocampal efferents within the fimbria-fornix, we conclude that postsynaptic glutamate receptor activation participates in the evolution of septal neuron injury following fimbria-fornix transection. Thus, excitotoxicity is a possible mechanism for transneuronal degeneration following central nervous system axotomy.

Hippocampal theta activity following selective lesion of the septal cholinergic system.

The characteristic electroencephalographic patterns within the hippocampus are theta and sharp waves. Septal neurons are believed to play an essential role in the rhythm generation of the theta pattern. The present study examined the physiological consequences of complete and selective damage of septohippocampal cholinergic neurons on hippocampal theta activity in rats. A selective immunotoxin against nerve growth factor receptor bearing cholinergic neurons (192 immunoglobulin G-saporin), [Wiley R. G. et al. (1991) Brain Res. 562, 149-153] was infused into the medial septal area (0.11-0.42 microgram). Hippocampal electrical activity was monitored during trained wheel running, drinking and the paradoxical phase of sleep, as well as following cholinomimetic treatment. A moderate dose of toxin (0.21 microgram) eliminated the septohippocampal cholinergic projection, as evidenced by a near total absence of choline acetyltransferase-immunoreactive neurons in the medial septum and the vertical limb of the diagonal band, and by the absence of acetylcholinesterase-positive fibers in the dorsal hippocampus. In the same rats, parvalbumin immunoreactivity, a reliable marker for septohippocampal GABAergic neurons, [Freund T. F. (1989) Brain Res. 478, 375-381], remained unaltered. In addition, retrograde transport of the tracer fluorogold demonstrated that the parvalbumin cell population preserved its axonal projection to the hippocampus. Following toxin treatment, the power of hippocampal theta, but not its frequency, decreased in a dose-dependent manner. Reduction of theta power occurred between three and seven days after the toxin treatment and remained unaltered thereafter up to eight weeks. A dose which eliminated all septohippocampal cholinergic neurons (0.21 microgram) left a small but significant theta peak in the power spectra during wheel running, paradoxical phase of sleep and intraseptal infusion of carbachol (5 micrograms). Peripheral administration of physostigmine (1 mg/kg) induced only slow (1.5-2.0 Hz) rhythmic waves. No changes were observed in the gamma (50-100 Hz) band. These findings indicate that the integrity of the septohippocampal GABAergic projection is sufficient to maintain some hippocampal theta activity. We hypothesize that cholinergic neurons serve to increase the population phase-locking of septal cells and thereby regulate the magnitude of hippocampal theta.

Effects of cholinergic-rich neural grafts on radial maze performance of rats after excitotoxic lesions of the forebrain cholinergic projection system--I. Amelioration of cognitive deficits by transplants into cortex and hippocampus but not into basal forebrain.

After ibotenate (10.0 mg/ml) lesions to the nucleus basalis and medial septal regions, at the source of the cortical and hippocampal branches of the forebrain cholinergic projection system, rats displayed long-lasting stable impairment in reference and working memory in both spatial (place) and associative (cue) radial maze tasks. Cell suspension transplants of cholinergic-rich fetal basal forebrain tissue dissected at embryonic day 15 substantially improved all aspects of radial maze performance to a comparable degree whether sited in cortex, hippocampus, or both regions of the host brain. No additive effects were obtained with grafts in both terminal regions, but total graft volume, assessed stereologically, showed a significant negative correlation with error scores. Rats with behaviourally effective grafts, like controls, were disrupted in the place task when tested in dim light which obscured extra-maze spatial cues. Lesioned rats were not affected by change in lighting. Grafts of cholinergic-poor fetal hippocampal tissue did not improve radial maze performance; neither did grafts of cholinergic-rich tissue placed within the host basal forebrain lesion sites. In rats with cholinergic-rich terminal grafts, cortical and hippocampal choline acetyltransferase activity was restored to control level, commensurate with site of transplant, whereas it was significantly reduced in lesioned animals and those with functionally ineffective grafts. The indiscriminate error pattern and insensitivity to changes in lighting shown by lesioned rats suggested that lesioning primarily disrupted attention rather than short- or long-term spatial or associative memory processes. Since rats with cholinergic-rich grafts showed both reduced errors and recovery of stimulus control, the data indicated that grafts affected information processing, rather than changes in motor or motivational processes. Changes in choline acetyltransferase activity and the behavioural efficacy of cholinergic-rich grafts are consistent with the involvement of acetylcholine in the behavioural deficits and recovery displayed by lesioned and grafted groups, but do not rule out contributions from other factors. The equipotency of grafts within each terminal region suggests also that there may be a considerable degree of functional cooperation between the two branches of the forebrain cholinergic projection system. Functional recovery may involve local, nonspecific synaptic or paracrine mechanisms within the target regions, since grafts were efficacious only when placed in the terminal areas, but not when sited homotopically in the basal forebrain, indicating that they did not achieve any functionally significant structural repair to the host brain at that site.