Carrier mediated delivery of NGF: Alterations in basal forebrain neurons in aged rats revealed using antibodies against low and high affinity NGF receptors

The distribution of low and high affinity nerve growth factor (NGF) receptors was investigated in the basal forebrain during aging and NGF treatment. A peripheral administration model for NGF was utilized. NGF was conjugated to a transferrin receptor antibody (OX-26-NGF), and this conjugate was injected into the tail vein of aged Fischer 344 male rats (24 months) twice weekly for 5 weeks (equivalent to 50 μg of NGF/injection). Controls were injected with a non-conjugated mixture of OX-26 and NGF. The aged rats treated with conjugate showed a significant increase in cell size of p75- and trkA-immunoreactive neurons in the medial septal nucleus and vertical limb of the diagonal band as compared to controls. A significant increase in cell size of trkA-immunoreactive neurons was also observed in the horizontal limb of the diagonal band in rats treated with conjugate. Rats treated with conjugate also showed a significant increase in overall staining density for p75 and trkA antibodies in the medial septal nucleus as compared to controls. A significant increase in staining density of p75-immunoreactive structures was also observed in the vertical and horizontal limbs of the diagonal band. Therefore, treatment with OX-26-NGF conjugate has regional effects on both the low and high affinity NGF receptors in terms of cell body size and staining density in the basal forebrain of aged rats. The current findings support the idea that this delivery system might be useful in therapeutic approaches involving the delivery of neurotrophic factors and other large molecules into the brain. J. Comp. Neurol. 387:1–11, 1997. © 1997 Wiley-Liss, Inc.     

Loss of forebrain cholinergic neurons and impairment in spatial learning and memory in LHX7-deficient mice

The identification of the genetic determinants specifying neuronal networks in the mammalian brain is crucial for the understanding of the molecular and cellular mechanisms that ultimately control cognitive functions. Here we have generated a targeted allele of the LIM-homeodomain-encoding gene Lhx7 by replacing exons 3-5 with a LacZ reporter. In heterozygous animals, which are healthy, fertile and have no apparent cellular deficit in the forebrain, b-galactosidase activity reproduces the pattern of expression of the wild-type Lhx7 locus. However, homozygous mutant mice show severe deficits in forebrain cholinergic neurons (FCNs), while other classes of forebrain neurons appear unaffected. Using the LacZ reporter as a marker, we show that in LHX7-deficient mice FCN progenitors survive but fail to generate cholinergic interneurons in the striatum and cholinergic projection neurons in the basal forebrain. Analysis of behaviour in a series of spatial and non-spatial learning and memory tasks revealed that FCN ablation in Lhx7 mutants is associated with severe deficits in spatial but only mild impairment of non-spatial learning and memory. In addition, we found no deficit in long-term potentiation in mutant animals, suggesting that FCNs modulate hippocampal function independently of its capacity to store information. Overall our experiments demonstrate that Lhx7 expression is required for the specification or differentiation of cholinergic forebrain neurons involved in the processing of spatial information.     

NGF gene transfer to intrinsic basal forebrain neurons increases cholinergic cell size and protects from age-related, spatial memory deficits in middle-aged rats

Administration of nerve growth factor (NGF) by intracerebroventricular infusion or transplantation of NGF-secreting cells to the basal forebrain improves spatial memory in aged animals. Using the adeno-associated virus (AAV) vector system, basal forebrain neurons were transduced to produce NGF ectopically for long intervals (at least 9 months). Rats received intraseptal injections of either the control vector, pTR-UF4, or the pTR-NGFmyc at 3 months of age, prior to testing their performance in the Morris water task. An age-related decrease in the acquisition of the hidden platform location was found at 12 months of age in the pTR-UF4 control group, but not in the pTR-NGFmyc group. Further, when compared to 3 month old untreated animals, the control group, but not the pTR-NGFmyc group, was impaired at 12 months of age. Concomitant to preventing age-related memory deficits, the NGF gene transfer increased cholinergic neuron size by 34% in the medial septum. This approach may therefore represent a viable therapy for age-related dementia involving dysfunction in cholinergic activity and memory, such as Alzheimer’s disease.     

NGF receptor immunoreactivity in aged rat brain

The cellular distribution of nerve growth factor (NGF) receptor (NGFR) immunoreactivity in 3 cholinergic nuclei (medial septal nucleus, nucleus of the diagonal band and nucleus basalis magnocellularis) of the aged rat brain was compared to that of young-adult animals. In young-adult rats, NGFR immunoreactivity was strong in the neuronal body and in the whole dendritic tree. In aged animals, NGFR immunoreactivity was weak in both cell body and dendrites and was practically absent in the dendrite's distal portion. The loss of dendritic NGFR may play a critical role in the decline of neuronal function in the aging brain.     

Acute application of NGF increases the firing rate of aged rat basal forebrain neurons.

Nerve growth factor (NGF) has been widely used in animal models to ameliorate age-related neurodegeneration, but it cannot cross the blood-brain barrier (BBB). NGF conjugated to an antibody against the transferrin receptor (OX-26) crosses the BBB and affects the biochemistry and morphology of NGF-deprived basal forebrain neurons. The rapid actions of NGF, including electrophysiological effects on these neurons, are not well understood. In the present study, two model systems in which basal forebrain neurons either respond dysfunctionally to NGF (aged rats) or do not have access to target-derived NGF (intraocular transplants of forebrain neurons) were tested. One group of transplanted and one group of aged animals received unconjugated OX-26 and NGF comixture as a control, while other groups received replacement NGF in the form of OX-26-NGF conjugate during the 3 months preceding the electrophysiological recording session. Neurons from animals in both the transplanted and aged control groups showed a significant increase in firing rate in response to acute NGF application, while none of the conjugate-treated groups or young intact rats showed any response. After the recordings, forebrain transplants and aged brains were immunocytochemically stained for the low-affinity NGF receptor. All conjugate treatment groups showed significantly greater staining intensity compared to controls. These data from both transplants and aged rats in situ indicate that NGF-deprived basal forebrain neurons respond to acute NGF with an increased firing rate. This novel finding may have importance even for long-term biological effects of this trophic factor in the basal forebrain.     

Hypertrophy of basal forebrain neurons and enhanced visuospatial memory in perinatally choline-supplemented rats

The effects of choline supplementation during two time-frames of early development on radial-arm maze performance and the morphology of basal forebrain neurons immunoreactive for the P75 neurotrophin receptor (NTR) in male and female Sprague–Dawley rats were examined. In the first experiment, rats were supplemented with choline chloride from conception until weaning. At 80 days of age, subjects were trained once a day on a 12-arm radial maze for 30 days. Compared to control littermates, supplemented rats made fewer working and reference memory errors; however, the memory enhancing effects of choline supplementation were greater in males than females. A morphometric analysis of NTR-immunoreactive cell bodies at three levels through the medial septum/diagonal band (MS/DBv) of these rats revealed that perinatal choline supplementation caused the somata of cells in the MS/DBv to be larger by 8–15%. In a second experiment, choline supplementation was restricted to embryonic days 12–17. A developmental profile of NTR immunoreactive cell bodies in the MS/DBv of 0-, 8-, 16-, 30- and 90-day old male and female rats again revealed that cell bodies were larger in choline-supplemented rats than controls. As in the behavioral studies, the effect of choline supplementation was greater in male than female rats. These data are consistent with the hypothesis that supplementation with choline chloride during early development leads to an increase in the size of cell bodies of NTR-immunoreactive cells in the basal forebrain and that this change may contribute to long-term improvement in spatial memory.     

Loss of basal forebrain P75(NTR) immunoreactivity in subjects with mild cognitive impairment and Alzheimer's disease.

The long-held belief that degeneration of the cholinergic basal forebrain was central to Alzheimer's disease (AD) pathogenesis and occurred early in the disease process has been questioned recently. In this regard, changes in some cholinergic basal forebrain (CBF) markers (e.g. the high affinity trkA receptor) but not others (e.g., cortical choline acetyltransferase [ChAT] activity, the number of ChAT and vesicular acetylcholine transporter-immunoreactive neurons) suggest specific phenotypic changes, but not frank neuronal degeneration, early in the disease process. The present study examined the expression of the low affinity p75 neurotrophin receptor (p75(NTR)), an excellent marker of CBF neurons, in postmortem tissue derived from clinically well-characterized individuals who have been classified as having no cognitive impairment (NCI), mild cognitive impairment (MCI), and mild AD. Relative to NCI individuals, a significant and similar reduction in the number of nucleus basalis p75(NTR)-immunoreactive neurons was seen in individuals with MCI (38%) and mild AD (43%). The number of p75(NTR)-immunoreactive nucleus basalis neurons was significantly correlated with performance on the Mini-Mental State Exam, a Global Cognitive Test score, as well as some individual tests of working memory and attention. These data, together with previous reports, support the concept that phenotypic changes, but not frank neuronal degeneration, occur early in cognitive decline. Although there was no difference in p75(NTR) CBF cell reduction between MCI and AD, it remains to be determined whether these findings lend support to the hypothesis that MCI is a prodromal stage of AD.     

Cholinergic neurons in the basal forebrain of aged female mice

Aging is associated with at least down-regulation of several cellular functions and diminished responsiveness to internal and external signals, and possibly with direct cell death. Consequently, pharmacological manipulations may be less effective in aged than in young organisms. In the present study, we investigated whether the basal forebrain cholinergic neurons and the estrogen receptor alpha (ERalpha) which they contain respond to changes in estrogen availability in aged female mice. The mice were sham-operated, ovariectomized, or ovariectomized and treated with 17beta-estradiol at the age of 18 months. Three months later, the mice were perfused and brain sections were double immunostained for choline acetyltransferase (ChAT) and ERalpha. Cell counting with a stereological method revealed that changes in the estrogen level have no effect on the total number of ChAT-immunoreactive (ir) neurons in the basal forebrain. However, the percentage of ChAT-ir neurons containing ERalpha-ir was higher in the ovariectomized mice than in the sham-operated or estrogen-treated mice. This was specific for the medial septum and vertical diagonal band of Broca. The findings indicate that even at old age the ERalphas in cholinergic cells are able to respond to changes in estrogen levels, though in a region-specific manner. This is naturally important for studies aiming to develop therapies for the elderly.     

Circuit-specific alterations in hippocampal synaptophysin immunoreactivity predict spatial learning impairment in aged rats.

The present study examined the long-standing concept that changes in hippocampal circuitry contribute to age-related learning impairment. Individual differences in spatial learning were documented in young and aged Long-Evans rats by using a hippocampal-dependent version of the Morris water maze. Postmortem analysis used a confocal laser-scanning microscopy method to quantify changes in immunofluorescence staining for the presynaptic vesicle glycoprotein, synaptophysin (SYN), in the principal relays of hippocampal circuitry. Comparisons based on chronological age alone failed to reveal a reliable difference in the intensity of SYN staining in any region that was examined. In contrast, aged subjects with spatial learning deficits displayed significant reductions in SYN immunoreactivity in CA3 lacunosum-moleculare (LM) relative to either young controls or age-matched rats with preserved learning. SYN intensity values for the latter groups were indistinguishable. In addition, individual differences in spatial learning capacity among the aged rats correlated with levels of SYN staining selectively in three regions: outer and middle portions of the dentate gyrus molecular layer and CA3-LM. The cross-sectional area of SYN labeling, by comparison, was not reliably affected in relation cognitive status. These findings are the first to demonstrate that a circuit-specific pattern of variability in the connectional organization of the hippocampus is coupled to individual differences in the cognitive outcome of normal aging. The regional specificity of these effects suggests that a decline in the fidelity of input to the hippocampus from the entorhinal cortex may play a critical role.     

GABAergic control of basal forebrain cholinergic neurons and memory

The involvement of the GABAergic innervation of basal forebrain neurons in the rats' conditional visual discrimination performance was examined. Performance in such a task is based on the subjects's ability to retrieve information about response rules, and previous experiments have demonstrated that basal forebrain lesions interfere with this ability. Following the acquisition of the task, chronic guide cannulae were stereotaxically implanted into the substantia innominata of both hemispheres, and the animals were retrained. Administration of the GABAA-agonist muscimol into the substantia innominata (0, 25, 50 ng/0.5 microliters/hemisphere) dose-dependently decreased the number of correct responses, increased the number of errors of omission, increased response latency, but did not affect side bias. Systemic co-administration of the cholinesterase inhibitor physostigmine (0, 0.1, 0.2 mg/kg; i.p.) exclusively interacted with the effects of muscimol on correct responding. Specifically, physostigmine dose-dependently intensified and attenuated the muscimol-induced reduction in correct responding. Although it cannot be excluded that alternative neuronal mechanisms were involved in the mediation of the effects of muscimol and their interaction with physostigmine, these findings support previous evidence indicating that the activity of basal forebrain cholinergic neurons is controlled by a GABAergic input, and that this neuronal link is involved in mnemonic processing.     

Cooperative effects of Sonic Hedgehog and NGF on basal forebrain cholinergic neurons

In ventromedial cells of the developing CNS, Sonic hedgehog (Shh) has been shown to affect precursor proliferation, phenotype determination, and survival. Here we show that Shh and its receptor, Ptc-1, are expressed in the adult rat basal forebrain, and that Ptc-1 is expressed specifically by cholinergic neurons. In basal forebrain cultures, Shh was added alone and in combination with nerve growth factor (NGF), and the number of cholinergic neurons was determined by choline acetyltransferase (ChAT) immunocytochemistry. By 8 days in vitro, Shh and NGF show a synergistic effect: the number of ChAT-positive cells after treatment with both factors is increased over untreated cultures or cultures treated with either factor alone. While Shh increases the overall basal level of proliferation, double-labeling of dividing neuronal precursors with [(3)H]thymidine followed by ChAT immunocytochemistry after they mature, demonstrates that the specific increase in cholinergic neurons is not due to this proliferation enhancement. These experiments imply a role for Shh in the development of postmitotic cholinergic neurons and suggest a therapeutic value for Shh in neurodegenerative disease.     

Nerve growth factor receptor immunoreactive profiles in the normal, aged human basal forebrain: colocalization with cholinergic neurons

A monoclonal antibody raised against the receptor for nerve growth factor (NGF) has been used to map the distribution of NGF receptor-containing profiles within the human basal forebrain of four male and three female elderly patients without neurologic or psychiatric illness. Immunohistochemically processed tissue reveals a continuum of NGF receptor-positive neurons located within the medial septum, vertical and horizontal limb nuclei of the diagonal band, and nucleus basalis. NGF receptor-containing neurons are also found within the bed nucleus of the stria terminalis, the anterior commissure, the internal capsule, and the internal and external medullary laminae of the globus pallidus. Virtually all (greater than 95%) NGF receptor-containing neurons colocalize with the specific cholinergic marker choline acetyltransferase (ChAT) or the nonspecific marker acetylcholinesterase (AChE). Conversely, a few cholinergic perikarya are found which are not NGF receptor positive (and vice versa). These findings demonstrate that human basal forebrain neurons on which NGF receptor immunoreactivity is detected are primarily cholinergic and analogous to the nonhuman primate Ch1-Ch4 subgroups of Mesulam et al. (J. Comp. Neurol., 214:170-197, '83). NGF receptor-containing fiber tracts are observed emanating from the medial septum and vertical limb nucleus of the diagonal band coursing medially within the fornix. Another fascicle originating mainly from the nucleus basalis and travelling within the external capsule enroute to the cortex is observed innervating all cortical layers. Comparison of NGF receptor- and ChAT-containing neurons reveals cholinergic perikarya within the striatal complex, whereas virtually no NGF receptor-containing neurons are found in these structures. An occasional displaced NGF receptor-containing neurons is seen in the ventrolateral portion of the putamen and the white matter underlying the nucleus accumbens. These data are discussed in terms of the relationship of NGF receptor- and ChAT-containing neurons within the basal forebrain and in terms of the possible functional significance of NGF in normal and diseased brain.     

Characteristics of memory impairment following lesioning of the basal forebrain and medial septal nucleus in rats

Memory impairment in rats with lesions of the basal forebrain (BF) and medial septal nucleus (MS) including cell bodies of the cortical and septohippocampal cholinergic systems, respectively, were compared in order to evaluate the functional contribution of the two cholinergic systems to memory. Biochemical assay revealed that lesioning of the BF and MS resulted in marked and selective decreases in both choline acetyltransferase and acetylcholinesterase activities in the cerebral cortex and hippocampus, respectively. Rats with BF lesions exhibited a severe deficit in a passive avoidance task; acquisition of passive avoidance by repeated training was sluggish, and the acquired response was rapidly eliminated in a subsequent extinction test. However, only slight impairment of passive avoidance was observed in rats with MS lesions. Memory impairment in rats with BF or MS lesions was also investigated using two spatial localization tasks, the Morris water task and the 8-arm radial maze task. Both BF and MS lesions elicited a significant impairment in the Morris water task that required reference memory, as demonstrated by an apparent increase in the latency to escape onto a hidden platform in a large water tank. The impairment was much more obvious in the BF-lesioned rats. In contrast, in the radial maze task primarily requiring working memory, rats with lesions of the MS showed severe disruption, exhibiting a marked increase in total errors, a decrease in the number of initial correct responses, and an apparent change in the strategy pattern. However, corresponding changes in the rats with BF lesions were slight. These results suggest that BF lesions may lead to substantial long-term memory impairment while MS lesions may primarily produce short-term or working memory impairment, indicating a qualitatively different contribution of the two cholinergic systems to memory. It is also suggested that these two experimental animal models may be useful for evaluation of therapeutic drugs for senile dementia of the Alzheimer type.     

Individual differences in spatial memory among aged rats are related to hippocampal PKCgamma immunoreactivity

We reported previously that the extent of spatial memory impairment among aged rats was correlated positively with levels of protein kinase Cgamma in hippocampal homogenates measured by quantitative Western blotting (Colombo et al., 1997). In the current study, immunocytochemistry was used to test whether the relationship between elevated PKC-gamma and memory impairment among aged rats could be localized further within regions of the hippocampus. Six- and 24-month-old male Long-Evans rats were first trained in the water maze on a standard place-learning task and then trained 2 weeks later on a transfer task designed for rapid acquisition. In comparison with young rats, aged rats with impaired spatial memory had increased PKCgamma-immunoreactivity (PKCgamma-ir) in CA1 of the hippocampus, but not the dentate gyrus. In addition, PKCgamma-ir in CA1 was correlated positively with spatial memory impairment among aged rats on the standard place-learning and the transfer training tasks. The current results are consistent with our previous report of PKCgamma in hippocampal homogenates, and show further that the relationships between PKCgamma-ir and memory impairments among aged rats are most evident in area CA1. Thus age-related impairments of spatial memory, as well as deficits in the flexible use of previously acquired information, may result from dysregulation of PKCgamma.     

Immunohistochemical detection of a monoclonal antibody directed against the NGF receptor in basal forebrain neurons following intraventricular injection.

It has been shown by autoradiography that, following intraventricular administration, a monoclonal antibody directed against the rat nerve growth factor (NGF) receptor is specifically accumulated bilaterally by numerous cholinergic neurons of the basal forebrain. This is consistent with the evidence that cholinergic basal forebrain neurons have NGF receptors and respond to NGF under a variety of experimental conditions. The present study demonstrates that the immunohistochemical detection of unmodified monoclonal antibody in cholinergic forebrain neurons following transport from CSF is feasible, although injection of larger amounts of the antibody is required to obtain an image equivalent to the one obtained with the autoradiographic method. The location of the immunohistochemical product clearly indicates that the antibody has been internalized, probably in an endosomal compartment.     

Specificity of 192 IgG-saporin for NGF receptor-positive cholinergic basal forebrain neurons in the rat

A monoclonal antibody to the rat nerve growth factor (NGF) receptor, 192 IgG, accumulates bilaterally and specifically in cholinergic basal forebrain (CBF) cells following intraventricular injection. An immunotoxin composed of 192 IgG linked to saporin (192 IgG-saporin) has been shown to destroy cholinergic neurons in the basal forebrain. We sought to determine if intraventricular 192 IgG-saporin affected choline acetyltransferase (ChAT) enzyme activity in the CBF terminal projection fields. ChAT assays from 192 IgG-saporin-treated animals showed significant time-dependent decreases in ChAT activity in the neocortex, olfactory bulb and hippocampus, compared to PBS- or OKT1-saporin-injected controls. ChAT and tyrosine hydroxylase activity in the striatum was always unchanged by 192 IgG-saporin. ChAT immunohistochemistry was confirmative of major cell loss in the CBF, while other cholinergic nuclei appeared unremarkable. The data provide further evidence of the selectivity of 192 IgG-saporin in abolishing cholinergic, NGF receptor-positive CNS neurons.     

Sensitivity to cholinergic drug treatments of aged rats with variable degrees of spatial memory impairment

As a first step, the present experiment aimed at characterizing learning and memory capabilities, as well as some motor and sensorimotor faculties, in aged (24-26.5 months) Long-Evans female rats. As a second step, a psychopharmacological approach was undertaken in order to examine the sensitivity of aged rats to muscarinic blockade and to cholinomimetic treatments. Young adult (3-5.5 months) and aged rats were tested for beam-walking performance, locomotor activity in the home cage and an open field, and spatial learning/memory performance in a water maze and a radial maze. Spontaneous alternation rates were assessed in a T-maze. Statistical analysis discriminated between aged rats showing moderate impairment (AMI) and those showing severe impairment (ASI) in the water maze test. Beside their different degrees of impairment in the water maze, AMI and ASI rats were similarly (no significant difference) impaired in beam-walking capabilities, home cage activity and radial maze performance. In the spontaneous alternation task aged rats were not impaired and, in the open-field test, AMI rats were hypoactive, but not as much as ASI rats. Neither of the cognitive deficits was correlated with a locomotor or a sensorimotor variable, or with the body weight. When tested in the radial maze, a low dose of scopolamine (0.1 mg/kg i.p.) produced memory impairments which were significant in AMI and ASI rats, but not in young rats. Combined injections of scopolamine and physostigmine (0.05 and 0.1 mg/kg) or tacrine (THA, 3 mg/kg) showed physostigmine (0.1 mg/kg) to compensate for the scopolamine-induced impairments only in AMI rats. whereas THA was efficient in both AMI and ASI rats. The results indicate: (i) that rats with different degrees of spatial memory impairment in the water maze are similarly hypersensitive to muscarinic blockade when tested in a radial maze test; and (ii) that under the influence of a dose of scopolamine which is subamnesic in young rats, aged rats respond to anticholinesterase treatments according to the level of performance achieved in the water maze: moderately impaired rats are sensitive to both physostigmine and THA, whereas more severely impaired rats are sensitive only to THA.     

Alterations in dendritic morphology of frontal cortical neurons after basal forebrain lesions in adult and aged rats

The nucleus basalis magnocellularis (NBM) is the major cholinergic projection to neocortex in the rat and plays a role in the modulation of cortical activity. Lesions of the NBM decrease thickness of lamina II–III of frontal cortex and decrease soma size of lamina II–III neurons. Additionally, aging produces changes in neuron size and numbers in the basal forebrain and frontal cortex of rats. We assessed dendritic changes in neurons from lamina II–III of frontal cortex in adult, middle-aged, and aged rats three months after unilateral lesions of the NBM. While lesions did not affect dendritic morphology in young adult rats, they decreased total dendritic length in middle-aged and aged rats, with dendritic alterations most pronounced in middle-aged rats. In middle-aged rats, lesion-induced changes in basilar arbor were apparently due to decreased dendritic branching: lesions markedly decreased the number of first-, second-, and third-order branches, but did not affect higher-order branching. In aged rats, lesions resulted in a small decrease in dendritic material proximal to the soma and a pronounced decrease in dendritic material distal to the soma, apparently due to a decrease in the length of terminal branches. These results suggest that the plasticity of neocortical neurons in the basalocortical system changes with age, and that early in aging this system may be particularly vulnerable to neural damage.     

NGF mRNA is expressed by GABAergic but not cholinergic neurons in rat basal forebrain

Nerve growth factor (NGF) supports the survival and biosynthetic activities of basal forebrain cholinergic neurons and is expressed by neurons within lateral aspects of this system including the horizontal limb of the diagonal bands and magnocellular preoptic areas. In the present study, colormetric and isotopic in situ hybridization techniques were combined to identify the neurotransmitter phenotype of the NGF-producing cells in these two areas. Adult rat forebrain tissue was processed for the colocalization of mRNA for NGF with mRNA for either choline acetyltransferase, a cholinergic cell marker, or glutamic acid decarboxylase, a GABAergic cell marker. In both regions, many neurons were single-labeled for choline acetyltransferase mRNA, but cells containing both choline acetyltransferase and NGF mRNA were not detected. In these fields, virtually all NGF mRNA-positive neurons contained glutamic acid decarboxylase mRNA. The double-labeled cells comprised a subpopulation of GABAergic neurons; numerous cells labeled with glutamic acid decarboxylase cRNA alone were codistributed with the double-labeled neurons. These data demonstrate that in basal forebrain GABAergic neurons are the principal source of locally produced NGF. © 1995 Wiley-Liss, Inc.