To evaluate effective means for delivering exogenous neurotrophins to neuron populations in the brain, we compared the distribution and transport of brain-derived neurotrophic factor (BDNF), nerve growth factor (NGF), and neurotrophin-3 (NT-3) following intracerebral delivery. Rats received an injection of radioiodinated or unlabeled neurotrophin into the lateral ventricle and were killed humanely after 1.5-24 hours. Other rats received continuous infusion of unlabeled neurotrophin into the lateral ventricle, the striatum, or the hippocampus for 3-14 days. The neurotrophins were detected by autoradiography or immunohistochemistry. There were striking differences between BDNF, NGF, and NT-3 in their penetration through brain tissue. These differences occurred regardless of the site or method of delivery, but were most pronounced following a bolus intracerebroventricular (ICV) injection. After ICV injection, NGF was widely distributed in tissues around the ventricles and at the surface of the brain, whereas the penetration of BDNF into brain tissue was distinctly less than that of NGF, and the penetration of NT-3 was intermediate. An ICV injection of NGF produced prominent but transient labeling of cells that contain the low-affinity NGF receptor, whereas an injection of BDNF prominently labeled the ventricular ependyma. During ICV infusion (12 micrograms/day), the distribution of BDNF was no greater than that observed after a 12-micrograms bolus injection. A sixfold increase in the amount of BDNF infused (72 micrograms/day) produced a more widespread distribution in the brain and doubled the depth of penetration into periventricular tissues near the cannula. Corresponding to their differences in penetration, NGF was retrogradely transported by basal forebrain cholinergic neurons after ICV or intrastriatal delivery, whereas NT-3 was transported by a few basal forebrain neurons after ICV delivery, and BDNF was rarely detected in neurons after ICV delivery. Delivery of BDNF directly to the striatum or the hippocampus labeled numerous neurons in nuclei afferent to these structures. In situ hybridization studies confirmed that the high-affinity BDNF receptor (TrkB) was much more widely expressed in neurons than was the high-affinity NGF receptor (TrkA). Moreover, mRNA for truncated forms of TrkB was expressed at high levels in the ependyma, the choroid epithelium, and the gray matter. It is likely that binding of BDNF to TrkB, which appears to be more abundant and ubiquitous than TrkA, restricts the diffusion of BDNF relative to that of NGF.

Alterations in neostriatal dopamine metabolism, release, and biosynthesis were determined 3, 5, or 18 days following partial, unilateral destruction of the rat nigrostriatal dopamine projection. Concentrations of dopamine and each of its metabolites, 3,4-dihydroxyphenylacetic acid (DOPAC), homovanillic acid (HVA), and 3-methoxytyramine (3-MT) were markedly decreased in the lesioned striata at 3, 5, or 18 days postoperation. The decline in striatal high-affinity [3H]dopamine uptake closely matched the depletion of dopamine at 3 and 18 days postoperation. However, neither DOPAC, HVA, nor 3-MT concentrations were decreased to as great an extent as dopamine at any time following lesions that depleted the dopamine innervation of the striatum by greater than 80%. In these more severely lesioned animals, dopamine metabolism, estimated from the ratio of DOPAC or HVA to dopamine, was increased two- to four-fold in the injured hemisphere compared with the intact hemisphere. Dopamine release, estimated by the ratio of 3-MT to dopamine, was more increased, by five- to sixfold. Importantly, the HVA/dopamine, DOPAC/dopamine, and 3-MT/dopamine ratios did not differ between 3 and 18 days postlesioning. The rate of in vivo dopamine biosynthesis, as estimated by striatal DOPA accumulation following 3,4-dihydroxyphenylalanine (DOPA) decarboxylase inhibition with NSD 1015, was increased by 2.6- to 2.7-fold in the surviving dopamine terminals but again equally at 3 and 18 days postoperation. Thus, maximal increases in dopamine metabolism, release, and biosynthesis occur rapidly within neostriatal terminals that survive a lesion. This mobilization of dopaminergic function could contribute to the recovery from the behavioral deficits of partial denervation by increasing the availability of dopamine to neostriatal dopamine receptors. However, these presynaptic compensations are not sufficient to account for the protracted (at least 3-week) time course of sensorimotor recovery that has been observed following partial nigrostriatal lesion.

Brain‐derived neurotrophic factor (BDNF) promotes the survival of dopamine (DA) neurons, enhances expression of DA neuron characteristics, and protects these cells from 6‐hydroxydopamine (6‐OHDA) toxicity in vitro. We tested the ability of BDNF or neurotrophin‐3 (NT‐3) to exert similar protective effects in vivo during chronic delivery of 6‐OHDA to the rat neostriatum. Chronic infusions of BDNF or NT‐3 (12 µg/day) above the substantia nigra were started 6 days before and continued during an 8‐day chronic intrastriatal infusion of 6‐OHDA. In control and neurotrophin‐treated animals, 6‐OHDA treatment selectively depleted 50–60% of nigrostriatal DA nerve terminals but produced little if any loss of pars compacta DA cell bodies. This partial DA lesion resulted in three rotations per minute toward the lesioned hemisphere after treatment with the DA release‐inducing drug d‐amphetamine. Compared with supranigral infusions of vehicle, BDNF and NT‐3 decreased the number of these ipsiversive rotations by 70 and 48% and increased by 20‐ and 10‐fold, respectively, the number of contraversive rotations observed after amphetamine injection. When challenged with the DA receptor agonist apomorphine, BDNF‐ and NT‐3‐treated animals also exhibited a seven‐ and 3.5‐fold increase in the number of contraversive rotations relative to the vehicle group, respectively. Compared with vehicle, BDNF increased striatal levels of homovanillic acid (HVA; 86%), 3,4‐dihydroxyphenylacetic acid (DOPAC; 42%), and 5‐hydroxyindoleacetic acid (5‐HIAA; 32%) and the HVA/DA (43%) and 5‐HIAA/serotonin (34%) ratios in the DA‐denervated striatum. NT‐3 augmented only striatal 5‐HIAA levels (24%). Neither factor altered the 6‐OHDA‐induced decrease in striatal DA levels or high‐affinity DA uptake and thus did not protect against the destruction of DA terminals and did not alter striatal D1 or D2 ligand binding. Choline, GABA, and glutamate uptake in the striatum were not altered by the lesion or neurotrophin treatment. Thus, BDNF and to a lesser extent NT‐3 reverse rotational behavioral deficits and augment striatal DA and 5‐HT metabolism in a partial DA lesion model.

The binding of biologically active and 125I-labeled neurotrophin-3 (NT-3) was studied with both dry film and emulsion autoradiography to compare with NGF binding and discover areas where NT-3 may function in vivo. The equilibrium binding of 300 pM 125I-NT-3 to rat brain sections was reversible and inhibited by unlabeled NT-3 (IC50, 420 pM). 125I-NT-3 bound in a saturable manner, with high affinity (Kd, 227-269 pM), and with a capacity (Bmax, 26 fmol/mg protein) that exceeded that of NGF by threefold. As with NGF, 125I-NT-3 also bound to a second population of sites with lower affinity (Kd, 2.8 nM) and higher capacity (Bmax, 170 fmol/mg protein). 125I-NT-3 binding was not blocked by NGF, or serum proteins, and brain-derived neurotrophic factor (BDNF) competed for it in a distinctly biphasic manner (IC50 values of 230 pM and 37 nM). Microdensitometry confirmed graphically and by Hill analysis the monophasic displacement of 125I-NT-3 and the biphasic displacement of 125I-NT-3 binding by BDNF in hippocampus, caudate-putamen, neocortex, and olfactory tubercle. In rat or cat, the topography of 125I-NT-3 binding differed from that reported for 125I-NGF binding or for the low-affinity NGF receptor. The highest binding densities were found in neocortical layers 1 and 2, the stratum oriens and radiatum of hippocampus, molecular layer of the dentate gyrus, nucleus of the lateral olfactory tract, entorhinal cortex, anterior olfactory nucleus, anteromedial thalamic nucleus, and amygdala. Moderate densities were found in neocortical layers 4-6, the neostriatum, amygdala, the dorsal root ganglia, and the central gray of spinal cord. Emulsion autoradiography also revealed binding in nerve terminal-rich regions of superficial neocortex and hippocampus but not on neural cell bodies. Binding was absent in many other brain regions, including cholinergic nuclei, and in all peripheral organs studied including liver, kidney, pancreas, heart, and skeletal muscle. 125I-NT-3 binding to sections of human basal ganglia resembled that seen in rat or cat, including high densities in the caudate, putamen, and superficial neocortex. The unique distribution and pharmacology of 125I-NT-3 binding to BDNF-sensitive and -insensitive sites in brain predict predominantly neuronal actions for these factors that are likely to be more widespread and distinct from those of NGF.

The capacity of D1 and D2 agonists and antagonists to regulate the in vivo release and metabolism of dopamine (DA) in mesolimbic and nigrostriatal DA neurons of the mouse was determined using gas chromatographic and mass fragmentographic (GC-MF) analysis. DA release was inferred from levels of 3-methoxytyramine (3-MT) and DA metabolism was inferred from levels of 3,4-dihydroxyphenylacetic acid (DOPAC) and homovanillic acid (HVA). DA release was increased by the D2 antagonists haloperidol and metoclopramide but not by the D1 antagonists SCH 23390 and SKF 83566. DA metabolism was increased by each of the four antagonists but to a greater extent with the D2 antagonists. The D2 agonists CGS 15855A and LY 171555 decreased DA release whereas the D1 agonist SKF 38393, at relatively high doses, only slightly affected DA release. Each of the three agonists decreased DA metabolism but again metabolism was more affected by the D2-selective drugs. The in vivo release of DA from mesolimbic and neostriatal DA neurons appears to be modulated by D2 but not by D1 receptors, whereas both receptor types can modulate DA metabolism.