Effects of acidic fibroblast growth factor on cholinergic neurons of nucleus basalis magnocellularis and in a spatial memory task following cortical devascularization

Focal Photothrombotic Lesion of the Rat Motor Cortex Increases BDNF Levels in Motor-Sensory Cortical Areas Not Accompanied by Recovery of Forelimb Motor Skills

Brain infarct triggers neurodegeneration that often shades spontaneous plasticity, occurring in the areas related anatomically and functionally to the infarcted structures. Neurotrophins which promote neuronal survival and plasticity, may protect neurons and enhance remodeling of the remaining circuits, leading to restoration of function. In particular, the crucial role of brain-derived neurotrophic factor (BDNF) in cortical function is well documented. Since BDNF was implicated in the mechanism of postinfarct recovery, we investigated whether focal photothrombosis in the motor cortex of adult rats modifies cortical BDNF protein levels in a time- and region-dependent fashion. In parallel, we aimed to establish, which cortical cells respond with altered BDNF expression and whether these alterations are reflected by forelimb motor skill impairment and recovery, evaluated up to 1 month postinfarct. The distribution of BDNF protein was visualized immunohistochemically and BDNF tissue levels were evaluated with enzyme-linked immunosorbent assay (ELISA). Ipsilateral to the infarct, an increase in BDNF levels occurred both in injured and neighboring regions already 24 h after photothrombosis. This increase was sustained up to postlesion day 7 in the motor cortex and reduced at 28 days. No BDNF changes were detected in homotopic regions of the contralateral cortex. The time-course of enhanced neurotrophic expression was paralleled by bilateral deficits in skilled reaching, which was the only clear and measurable motor impairment observed in the study. We conclude that the spontaneous increase of BDNF is not sufficient to protect neurons from degeneration in the lesion proximity whereas plasticity reported in the adjacent regions may be attributable to enhanced BDNF-related stimuli, which do not counteract the impairment of skilled reaching but might be, at least in part, responsible for the absence of deficits in other functional/behavioral tests.

Apoptosis in the rat basal forebrain during development and following lesions of connections

Evidence suggests that neurotrophins are essential for the survival and phenotypic maintenance of cholinergic basal forebrain (BF) neurons. We evaluated the pattern of programmed cell death in the BF of the rat during development and after ablations of the cerebral cortex, a major target area and source of neurotrophins for BF neurons. We identified dying cells using the TUNEL (terminal deoxynucleotidyl-transferase-mediated dUTP-biotin nick end labelling) method and confirmed their apoptotic morphology with electron microscopy. Moreover, we demonstrated the expression of the apoptotic marker active caspase-3 in cells with features of apoptosis. TUNEL(+) cells were present in the developing BF during the first two postnatal weeks. Their frequency peaked at postnatal day (P)1 and at P5. TUNEL used in conjunction with immunofluorescence for neuronal nuclear protein (NeuN) showed that, at both peak stages, the majority of apoptotic cells were neurons. Extensive lesions of the cerebral cortex at different ages (P0, P7 and P14) did not induce significant changes in the frequency of apoptotic BF neurons. However, they resulted in alterations in the morphological phenotype of choline acetyltransferase (ChAT)-immunoreactive neurons in the BF, and a reduction in their number which was inversely proportional to the age at which the lesions were performed. We suggest that: (i) apoptosis is temporally coordinated with the morphological and neurochemical differentiation of BF neurons and the establishment of connections with their target areas; and (ii) cortical ablations do not affect the survival of BF neurons, but they influence the phenotype of cholinergic BF neurons.

Localization of Fibroblast Growth Factor-1 in Cholinergic Neurons Innervating the Rat Larynx

Cholinergic neurons in the dorsal motor nucleus of the vagus (DMNV) are particularly vulnerable to laryngeal nerve damage, possibly because they lack fibroblast growth factor-1 (FGF1). To test this hypothesis, we investigated the localization of FGF1 in cholinergic neurons innervating the rat larynx by immunohistochemistry using central-type antibodies to choline acetyltransferase (cChAT) and peripheral type (pChAT) antibodies, as well as tracer experiments. In the DMNV, only 9% of cChAT-positive neurons contained FGF1, and 71% of FGF1-positive neurons colocalized with cChAT. In the nucleus ambiguus, 100% of cChAT-positive neurons were FGF1 positive. In the intralaryngeal ganglia, all ganglionic neurons contained both pChAT and FGF1. In the nodose ganglia, 66% of pChAT-positive neurons were also positive for FGF1, and 90% of FGF1-positive ganglionic cells displayed pChAT immunoreactivity. Neuronal tracing using cholera toxin B subunit (CTb) demonstrated that cholinergic neurons sending their axons from the DMNV and nucleus ambiguus to the superior laryngeal nerve were FGF1 negative and FGF1 positive, respectively. In the nodose ganglia, some FGF1-positive cells were labeled with CTb. The results indicate that for innervation of the rat larynx, FGF1 is localized to motor neurons, postganglionic parasympathetic neurons, and sensory neurons, but expression is very low in preganglionic parasympathetic cholinergic neurons.

Comparison of FGF1 (aFGF) expression between the dorsal motor nucleus of vagus and the hypoglossal nucleus of rat

Neurons in the dorsal motor nucleus of the vagus (DMNV) are more severely affected by axonal injury than most other nerves, such as those of the hypoglossal nucleus. However, the mechanism underlying such a response remains unclear. In this study, we compared the expression of fibroblast growth factor 1 (FGF1), a neurotrophic factor, between the DMNV and the hypoglossal nucleus by RT-PCR and immunohistochemical analyses. RT-PCR showed that the level of FGF1 mRNA expression in the DMNV was lower than that in the hypoglossal nucleus (P<0.01). Immunohistochemistry revealed that FGF1 was localized to neurons. FGF1-positive neurons in large numbers were evenly distributed in the hypoglossal nucleus, whereas FGF1-positive neurons were located in the lateral part of the DMNV. Double immunostaining for FGF1 and choline acetyltransferase demonstrated that 22.7% and 78% of cholinergic neurons were positive for FGF1 in the DMNV and hypoglossal nucleus, respectively. A tracing study with cholera toxin B subunit (CTb) demonstrated that cholinergic neurons sending their axons from the DMNV to the superior laryngeal nerve were FGF1-negative. The results suggest that the low expression of FGF1 in the DMNV is due to severe damage of neurons in the DMNV.

Microsphere embolism-induced cortical cholinergic deafferentation and impairments in attentional performance

Ischemic events have been hypothesized to play a critical role on the pathogenesis of dementia and the acceleration of cognitive impairments. This experiment was designed to determine the consequences of microvascular ischemia on the cortical cholinergic input system and associated attention capacities. Injections of microspheres ( approximately 50 microm diameter; approximately 5000 microspheres/100 microL) into the right common carotid artery of rats served as a model of microvascular ischemia and resulted in decreases in the density of cholinergic fibers in the ipsilateral medial prefrontal cortex and frontoparietal areas. Furthermore, dense astrogliosis, indicated by glial fibrillary acidic protein (GFAP) immunohistochemistry, was observed in the globus pallidus, including the areas of origin of cholinergic projections to the cortex. Fluoro-Jade B staining indicated that loss of neurons in the cortex was restricted to areas of microsphere-induced infarcts. Attentional performance was assessed using an operant sustained attention task; performance in this task was previously demonstrated to reflect the integrity and activity of the cortical cholinergic input system. Embolized animals' performance was characterized by a decrease in the animals' ability to detect signals. Their performance in non-signal trials remained unaffected. The residual density of cholinergic axons in prefrontal and frontoparietal areas correlated with the animals' performance. The present data support the hypothesis that microvascular ischemia results in loss of cortical cholinergic inputs and impairs associated attentional performance. Microsphere embolism represents a useful animal model for studying the role of interactions between microvascular disorder and impaired forebrain cholinergic neurotransmission in the manifestation of cognitive impairments.

The 5HT1A receptor agonist, 8-OH-DPAT, protects neurons and reduces astroglial reaction after ischemic damage caused by cortical devascularization

Serotonin 1A (5HT1A) receptor agonists have shown neuroprotective properties in different models of central nervous system injury. Activation of neuronal 5HT1A receptors appears to be involved in the neuroprotective effects. It remains to be elucidated if astroglial cells are responsive to the 5HT1A neuroprotective effects. The participation of astroglial S100B trophic factor has been proposed since 5HT1A activation leads to S100B release and nanomolar concentration level of this molecule showed pro-survival activity in neuronal cultures. Using the cortical devascularization model (CD; unilateral pial disruption), a procedure that results in localized ischemia without producing direct physical damage to brain tissue, we tested the effects of a full 5HT1A agonist, 8-OH-DPAT, or the antagonist WAY-100635 on cortical neuronal survival, astroglial cell response and S100B expression. Wistar rats were subjected to CD lesion which consisted of a craniotomy followed by physical damage to the underlying pial blood vessels. Two and twenty-four hours after the CD lesion, animals received intraperitoneally 8-OH-DPAT (1 mg/kg), WAY-100635 (1 mg/kg) or vehicle (sterile saline). At 3, 7 or 14 days post-lesion, animals were sacrificed and their brains processed for immunohistochemistry to detect GFAP, vimentin, MAP-2, S100B and nuclear Hoechst staining. S100B level in the brain cortex and serum was quantified by an ELISA assay. Serum S100B was considered an index of S100B release. 8-OH-DPAT treatment reduced neuronal death, dendrite loss, astroglial hypertrophy and hyperplasia. In contrast, WAY-100635 treatment increased these parameters of damage. S100B intracellular immunoreactivity in astrocytes and total S100B level showed long-lasting changes after the CD lesion and subsequent treatments depending on the 5HT1A activity. The level of serum S100B was increased in 8-OH-DPAT-treated animals. Increased damage observed in WAY-100635-treated animals supports the hypothesis that the protective 8-OH-DPAT action may be mediated by specific 5HT1A receptors. The reduction in astroglial hypertrophy and hyperplasia as well as long-term changes in S100B immunoreactivity and increased S100B release that we observed allows us to hypothesize that astroglial cells may play an important role in 5HT1A-mediated neuroprotection.

Degeneration of the basalocortical pathway from the cortex induces a functional increase in galaninergic markers in the nucleus basalis magnocellularis of the rat

The present work aimed 1) to evaluate whether an increase in galanin or galanin receptors could be induced in the nucleus basalis magnocellularis (nbm) by degeneration of the basalocortical neurons from the cortex and 2) to analyze the consequences of such an increase on cortical activity. First, a mild ischemic insult to the frontoparietal cortex was performed to induce the degeneration of the basalocortical system; galanin immunoreactivity, galanin binding sites, and cholinergic muscarinic receptors were quantified through immunocytochemistry and autoradiography. Second, galanin infusions in the nbm were undertaken to mimic a local increase of the galaninergic innervation; cortical acetylcholine release, cerebral glucose use, and cerebral blood flow were then measured as indices of cortical activity. As a result of the cortical ischemic lesion, the postsynaptic M1 and presynaptic M2 muscarinic receptors were found to be reduced in the altered cortex. In contrast, galaninergic binding capacity and fiber density were found to be increased in the ipsilateral nbm in parallel with a local decrease in the cholinergic markers such as the muscarinic M1 receptor density. Galanin infusion into the nbm inhibited the cortical acetylcholine release and cerebral blood flow increases elicited by the activation of the cholinergic basalocortical system but failed to affect acetylcholine release, cerebral blood flow, and cerebral glucose use when injected alone in the nbm. These results demonstrate that degeneration of the basalocortical system from the cortex induces an increase in galaninergic markers in the nbm, a result that might suggest that the galaninergic overexpression described in the basal forebrain of patients with Alzheimer's disease can result from a degeneration of the cholinergic basalocortical system from the cortex. Because galanin was found to reduce the activity of the basalocortical cholinergic system only when this one is activated, galanin might exert its role rather during activation deficits than under resting conditions such as the resting cortical hypometabolism, which is characteristic of Alzheimer's disease.

Concepts of CNS Plasticity in the Context of Brain Damage and Repair

BACKGROUND

How effectively the brain can respond to injury and undergo structural repair has become one of the most exciting areas of contemporary basic and translational neuroscience research. Although there are no clinical treatments yet available to enhance repair of the damaged brain, there are a number of potential therapies being investigated. New drugs are designed to provide some degree of neuroprotection by preventing injured or vulnerable nerve cells from dying, or they are given in the hope of stimulating regenerative processes that could lead to the restoration or the formation of new connections that were lost because of the injury.

MAIN OUTCOME MEASURES

The developments in pharmacology are based primarily upon understanding the molecular mechanisms of drug actions at the level of the genome or with respect to cellular metabolism. Although there is a substantial interest in the pharmacology of brain repair, there seems to be less concern with the various theories of central nervous system plasticity, organization, and reorganization after an injury.

CONCLUSIONS

This review discusses some of the older and current ideas and theories that have been presented over the years to explain recovery of function. We then provide an overview of what is being done in the laboratory to develop new and safe drugs for the treatment of traumatic brain injuries.

Effects of an acidic fibroblast growth factor fragment analog on learning and memory and on medial septum cholinergic neurons in senescence-accelerated mice

We examined the effects of repeated subcutaneous injections of an acidic fibroblast growth factor fragment analog, [Ala16] acidic fibroblast growth factor (1-29), on learning and memory and on the choline acetyltransferase immunoreactivity of forebrain neurons in senescence-accelerated mice. One group of accelerated senescence-prone mice (accelerated senescence-prone-8) received [Ala16] acidic fibroblast growth factor (1-29), whereas the other group of accelerated senescence-prone-8 mice and a group of accelerated senescence-resistant mice (control) received vehicle solution. Injections began at three weeks after birth and were given weekly for 10 months. In a passive avoidance test, the mean retention latency at three, six and nine months of age was significantly longer in controls (vehicle-treated accelerated senescence-resistant-1) and acidic fibroblast growth factor fragment-treated accelerated senescence-prone-8 than in vehicle-treated accelerated senescence-prone-8 mice, and the latency in acidic fibroblast growth factor fragment-treated accelerated senescence-prone-8 mice was significantly shorter than that in controls only at nine months of age. In the Morris water maze task, the mean latency to climb onto the platform was significantly longer in acidic fibroblast growth factor fragment- and vehicle-treated accelerated senescence-prone-8 mice than in controls. However, the mean latency in the third and fourth trial blocks was significantly shorter for acidic fibroblast growth factor fragment-treated accelerated senescence-prone-8 than for vehicle-treated accelerated senescence-prone-8 mice. In the probe trials, controls and acidic fibroblast growth factor fragment-treated accelerated senescence-prone-8 mice spent significantly more time in the quadrant in which the platform had previously been located than in the other three quadrants. In acidic fibroblast growth factor fragment-treated accelerated senescence-prone-8 mice, the density of medial septum neurons intensely stained for choline acetyltransferase was significantly greater than that in vehicle-treated accelerated senescence-prone-8 mice, but significantly less than that in controls. The results indicate that the beneficial effect of [Ala16] acidic fibroblast growth factor (1-29) on learning and memory function in accelerated senescence-prone-8 mice may be related to a preservation of function in medial septum cholinergic neurons.

Differentiation and angiogenesis of central nervous system stem cells implanted with mesenchyme into ischemic rat brain

After cerebral infarction, necrosis in neural tissues is not usually repaired or reconstructed by the injured brain. We therefore examined the effects on postinfarction repair of implanting central nervous system (CNS) stem cells together with mesenchyme, because CNS stem cells can be expected to adapt and survive in the adult brain. Cerebral infarction was induced by the Koizumi-Longa method, using the adult male spontaneous hypertensive rat model. Reperfusion was performed an hour after middle cerebral artery occlusion. The rat mesencephalic neural plate at the early somite stage (embryonic day 10.5) together with the adjacent ventral mesenchymal tissues was dissected out under the microscope and immediately implanted into the ischemic rat striatum. One month later, the cognitive function was evaluated by the Morris water maze method. Histologic and immunohistochemical examinations of the graft were made with hematoxylin-eosin (H&E), neurofilament-200, and tyrosine hydroxylase (TH) stains. In the water maze study, mean latency times required to reach an escape platform in the implanted animals with surviving grafts were found to be shorter than in those without grafts, but longer than in normal animals. In the spatial probe trial, the number of animals seen to cross the area in the pool where the platform had been located was greater in the implanted rats with surviving grafts than in other groups. Multiple vascularization in the grafted area was observed histologically in H&E-stained tissues, and neurofilament-200-positive cells were recognized in the graft. TH staining revealed within the graft many immunoreactive neuron-like cell bodies with long dendrites. It was suggested that grafted CNS stem cells with mesenchyme may survive and differentiate into mature CNS tissue within the adult ischemic rat brain, constructing vessels in and around the grafts, and may therefore have the potential to be effective in the recovery of the cognitive function of the rat model.

NGF prevents further atrophy of cholinergic cells of the nucleus basalis due to cortical infarction in adult post-hypothyroid rats but does not restore cell size compared to euthyroid [correction of euthroid] rats

We have tested the hypotheses that nerve growth factor treatment in adult post-hypothyroid rats can: (1) restore cross-sectional area of cholinergic cells of the nucleus basalis and (2) prevent further atrophy of these neurons following cortical infarction. In addition, we assessed the expression of p75NGFR and p140trkA mRNAs in the nucleus basalis cells of post-hypothyroid rats. Rats were rendered hypothyroid by the addition of propylthiouracil to their diet beginning on embryonic day 19 until the age of 1 month. At this time both the pups and their dams continued to receive 0.05% propylthiouracil in their diet and the pups were thyroidectomized. At 60 days, propylthiouracil treatment was interrupted and thyroxine levels were restored to normal by daily subcutaneous administration of physiological levels of thyroxine. Morphometric analysis identified atrophied nucleus basalis magnocellularis cholinergic cells at two ages, days 75 and 105, identified by in situ hybridization for p75NGFR and p140trkA mRNAs in methylene blue stained cells (day 75) and choline acetyltransferase immunostaining (day 105). The mean number of silver grains (pixels) per microns2 (mean +/- S.E.M.) of cell body cross-sectional area for p75NGFR mRNA in the nucleus basalis magnocellularis of euthyroid rats was 3.43 +/- 0.89, which was not statistically different from post-hypothyroid animals (4.02 +/- 1.07). A similar finding was noted for p140trkA mRNA: mean number of grains in the euthyroid group was 5.54 +/- 0.96 and was not statistically different from the post-hypothyroid group (6.32 +/- 1.45). Nerve growth factor treatment in adulthood (between days 75 and 82) did not restore cross-sectional area from early thyroid deprivation. However, it prevented further atrophy of nucleus basalis magnocellularis neurons following cortical devascularization inflicted in adulthood (day 75).

Acidic fibroblast growth factor mRNA is expressed by basal forebrain and striatal cholinergic neurons

Evidence for the importance of the basal forebrain cholinergic system in the maintenance of cognitive function has stimulated efforts to identify trophic mechanisms that protect this cell population from atrophy and dysfunction associated with aging and disease. Acidic fibroblast growth factor (aFGF) has been reported to support cholinergic neuronal survival and has been localized in basal forebrain with the use of immunohistochemical techniques. Although these data indicate that aFGF is present in regions containing cholinergic cell bodies, the actual site of synthesis of this factor has yet to be determined. In the present study, in situ hybridization techniques were used to evaluate the distribution and possible colocalization of mRNAs for aFGF and the cholinergic neuron marker choline acetyltransferase (ChAT) in basal forebrain and striatum. In single-labeling preparations, aFGF mRNA-containing neurons were found to be codistributed with ChAT mRNA+ cells throughout all fields of basal forebrain, including the medial septum/diagonal band complex and striatum. By using a double-labeling (colormetric and isotopic) technique, high levels of colocalization (over 85%) of aFGF and ChAT mRNAs were observed in the medial septum, the diagonal bands of Broca, the magnocellular preoptic area, and the nucleus basalis of Meynert. The degree of colocalization was lower in the striatum, with 64% of the cholinergic cells in the caudate and 33% in the ventral striatum and olfactory tubercle labeled by the aFGF cRNA. These data demonstrate substantial regionally specific patterns of colocalization and support the hypothesis that, via an autocrine mechanism, aFGF provides local trophic support for cholinergic neurons in the basal forebrain and the striatum.

Acidic FGF induces NGF and its mRNA in the injured neocortex of adult animals

Recently we reported that human recombinant acidic fibroblast growth factor (aFGF) is capable of preventing degeneration of nucleus basalis magnocellularis neurons in vivo and inducing growth of astrocytes in vitro. In the present study, the effects of aFGF on the concentration of nerve growth factor (NGF) and its messenger RNA were investigated in the rat cerebral cortex following unilateral cortical infarction. Lesioned animals exhibited a significant increase of NGF in the remaining cortex ipsilateral to the lesion. After combining cortical lesion with intracerebroventricular application of aFGF (12 micrograms/day for 7 days), we observed an 8-fold increase in the NGF concentration and a marked increase in the level of steady state NGF mRNA relative to controls ipsilaterally, and a less pronounced aFGF effect in the contralateral cerebral cortex. These results support the hypothesis that the neurotrophic effects previously shown for aFGF and basic FGF (bFGF) in neurotrophin-sensitive neurons is mediated by inducing increased production of NGF within the injured central nervous system (CNS) of adult animals.

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.

Differential expression of p140trk, p75NGFR and growth-associated phosphoprotein-43 genes in nucleus basalis magnocellularis, thalamus and adjacent cortex following neocortical infarction and nerve growth factor treatment

A loss of target-derived neurotrophic factors is hypothesized to be one of the major determinants of central nervous system neuronal degeneration. In order to obtain further insight into early neuronal responses to injury, lesion-induced alterations in the expression of high- and low-affinity nerve growth factor receptors, as well as growth-associated phosphoprotein-43 genes in nucleus basalis magnocellularis, thalamic and neocortical neurons were studied. For this purpose, unilateral cortical devascularization operations were conducted on adult rats. Animals received i.c.v. infusions of vehicle or nerve growth factor (12 micrograms/day) and were killed at one, three, seven and 15 days post-lesion. In situ hybridization studies using 35S-labelled oligonucleotide probes for p75NGFR, p140trk and growth-associated phosphoprotein-43 messenger RNAs reveals that these genes were differentially regulated following the lesion. In the nucleus basalis magnocellularis ipsilateral to the lesion, p140trk gene expression significantly decreased on days 3 and 7, while p75NGFR messenger RNA initially increased on day 3 and decreased on days 7 and 15 after lesion. GAP-43 messenger RNA levels were significantly increased in the nucleus basalis magnocellularis on post-lesion days 3 and 7. Moreover, in contrast to p75NGFR or 140trk, growth-associated phosphoprotein-43 messenger RNA levels were significantly increased in pyramidal neurons located in the remaining cortex adjacent to the cortical lesion at all time points. In the lateral and ventroposterior nuclei of the thalamus, growth-associated phosphoprotein-43 messenger RNA level was slightly increased on days 1 and 3 and was dramatically decreased, significantly below the levels in sham-operated controls, on post-lesion days 7 and 15. During nerve growth factor application, the level of p140trk messenger RNA in the lesioned nucleus basalis magnocellularis returned to values observed in the contralateral nucleus basalis magnocellularis while p75NGFR messenger RNA was increased above values noted in all animals not treated with nerve growth factor. Nerve growth factor treatment did not affect the expression of growth-associated phosphoprotein-43 messenger RNA in any of the areas studied. p140trk messenger RNA was not up-regulated during the time that nerve growth factor was applied, as observed for p75NGFR, but only eight days after interrupting nerve growth factor treatment. Three cell types, nucleus basalis magnocellularis, cortical pyramidal and thalamic neurons, were probably affected in different ways by the devascularization with respect to lesion extent. Consequently, the remaining number of synaptic contacts in each of these brain areas is most likely different which may lead to a differential regulation of growth-associated phosphoprotein-43 messenger RNA.(ABSTRACT TRUNCATED AT 400 WORDS)