Dynamic metabolic changes after permanent cerebral ischemia in rats with/without post-stroke exercise: a positron emission tomography (PET) study

Abstract

Objectives:

Recent studies have suggested that rehabilitation therapy can accelerate functional recovery after a stroke. Although often overlooked, the cortical hemisphere contralateral to an infarction plays an important role. This study investigates alterations in metabolism of both the damaged (‘ipsilateral’) as well as the undamaged (‘contralateral’) hemisphere using ¹⁸F-fluorodeoxyglucose (FDG)-micro-positron emission tomography (PET) in a rat permanent stroke model (with or without post-injury exercise) in order to elucidate the relative importance of either hemisphere to the recovery process following stroke.

Methods:

Thirty-six adult, male Sprague-Dawley rats were divided into four groups before subsequent surgery: sham controls with or without exercise, and ischemic (‘stroke’) groups with or without exercise. Fluorodeoxyglucose micro-PET imaging was performed at 7, 14, and 21 days after the designated procedure according to group assignment. The imaging data was analyzed by ANOVA using SPMratIHEP software.

Results:

Both exercise and ischemia have measurable effects on the motor cortex as well as on the striatum, the effects of which notably include the contralateral hemisphere. To that end, regions of the contralateral motor cortex and striatum have been found to be in a hypermetabolic state following exercise. We further observed that exercise reversed the hypometabolism caused by ischemia back to control levels from day 7 through day 21 on the ipsilateral side. Its effect on the contralateral hemisphere, notably, bolsters an already vigorous response observed after ischemic insult. Thus, the beneficial effect of exercise, as inferred by an increase in metabolic activity, is evident in both hemispheres.

Discussion:

These findings suggest that the contralateral hemisphere can compensate for the damaged cortex by remodeling neuronal activity. Thus, clinical treatments specifically targeted to the ‘intact’ hemisphere following stroke may provide a complimentary strategy for promoting recovery of functional deficits and for improving quality of life in stroke patients.     

Effects of Exercise Following Lateral Fluid Percussion Brain Injury in Rats

Previous studies have suggested that brain-derived neurotrophic factor (BDNF) is involved in memory and learning, and may be neuroprotective following various brain insults. Exercise has been found to increase BDNF mRNA levels in various brain regions, including specific subpopulations of hippocampal neurons. In the present study, we were interested in whether following traumatic brain injury, exercise could increase BDNF mRNA expression, attenuate neuropathology, and improve cognitive and neuromoter performance. We subjected adult male Sprague-Dawley rats to a fluid percussion brain injury, followed by either 18 days of treadmill exercise or handling. Spatial memory was evaluated in a Morris Water Maze (MWM) and motor function was evaluated with a battery of neuromotor tests. Neuropathology was evaluated by measuring the cortical lesion volume and the extent of neuronal loss in the hipocampus. Expression of BDNF mRNA in the hippocampus was assessed with in situ hybridization and densitometry. Hybridization signal for BDNF mRNA was significantly increased bilaterally in the exercise group in hippocampal regions CA1 and CA3 (p<0.05), but not in the granule cell layer of the dentate gyrus. No significant differences were observed between the groups in neuropathology, spatial memory, or motor performance. This study suggests that after traumatic brain injury, exercise elevates BDNF mRNA in specific regions of the hippocampus.     

Brain BDNF levels elevation induced by physical training is reduced after unilateral common carotid artery occlusion in rats

We investigated the contribution of blood flow elevation in the cerebrovasculature to physical training-induced brain-derived neurotrophic factor (BDNF) levels elevation in the brain. Brain-derived neurotrophic factor protein levels were measured in the motor cortex 24 h after the last session of a forced treadmill walking (30 minutes a day, 18 m/minute for 7 consecutive days). Unilateral common carotid artery occlusion and modulation of exercise intensity (0 versus −10% inclination of the treadmill) were used as strategies to reduce the (normal) elevation of flow in the cerebrovasculature occurring during exercise. Administration of N-nitro-L-arginine methyl ester (L-NAME, 60 mg/kg before each exercise sessions) and genetic hypertension (spontaneously hypertensive rats) were used as approaches to reduce stimulation of nitric oxide production in response to shear stress elevation. Vascular occlusion totally and partially abolished the effect of physical training on BDNF levels in the hemisphere ipsilateral and contralateral to occlusion, respectively. BDNF levels were higher after high than low exercise intensity. In addition, both genetic hypertension and L-NAME treatment blunted the effects of physical training on BDNF. From these results, we propose that elevation of brain BDNF levels elicited by physical training involves changes in cerebral hemodynamics.     

Effects of exercise and muscle type on BDNF,NT‐4/5, and TrKB expression in skeletal muscle

Muscle‐derived neurotrophins are thought to contribute to the adaptation of skeletal muscle to exercise, but the effects of brief exercise interventions on BDNF, NT‐4/5, and trkB are not understood. RNA was extracted for RT‐PCR from soleus and medial gastrocnemius of Sprague‐Dawley rats exercised on a treadmill at speeds up to 20 m/min at 5% incline for 5 or 10 days. BDNF expression was elevated in soleus following 5 days (184%, P < 0.001) but not 10 days of exercise. NT‐4/5 and trkB were not affected at either time‐point. BDNF mRNA was significantly higher in soleus at rest when compared with medial gastrocnemius (193%, P < 0.05). No significant effects of muscle type were detected for NT‐4/5 and trkB. Our results indicate differential control of BDNF expression between soleus and medial gastrocnemius following 5 days of exercise. BDNF may be a protein with an uncharacterized contribution to the acute adaptation of skeletal muscle to exercise, whereas NT‐4/5 shows no response. Muscle Nerve, 2009     

Enlarged infarct volume and loss of BDNF mRNA induction following brain ischemia in mice lacking FGF-2

FGF-2, a potent multifunctional and neurotrophic growth factor, is widely expressed in the brain and upregulated in cerebral ischemia. Previous studies have shown that intraventricularly or systemically administered FGF-2 reduces the size of cerebral infarcts. Whether endogenous FGF-2 is beneficial for the outcome of cerebral ischemia has not been investigated. We have used mice with a null mutation of the fgf2 gene to explore the relevance of endogenous FGF-2 in brain ischemia. Focal cerebral ischemia was produced by occlusion of the middle cerebral artery (MCAO). We found a 75% increase in infarct volume in fgf2 knock-out mice versus wild type littermates (P < 0.05). This difference in the extent of ischemic damage was observed after 24 h, and correlated with decreased viability in fgf2 mutant mice following MCA occlusion. Increased infarct volume in fgf2 null mice was associated with a loss of induction in hippocampal BDNF and trkB mRNA expression. These findings indicate that signaling through trkB may contribute to ameliorating brain damage following ischemia and that bdnf and trkB may be target genes of FGF-2. Together, our data provide the first evidence that endogenous FGF-2 is important in coping with ischemic brain damage suggesting fgf2 as one crucial target gene for new therapeutic strategies in brain ischemia.     

Effect of treadmill exercise on 5-HT, 5-HT1A receptor and brain derived neurophic factor in rats after permanent middle cerebral artery occlusion

It has been well documented that exercise promotes neurological rehabilitation in patients with cerebral ischemia. However, the exact mechanisms have not been fully elucidated. This study aimed to discuss the effect of treadmill exercise on expression levels of 5-HT, 5-HT1A receptor (5-HT1AR) and brain derived neurophic factor (BDNF) in rat brains after permanent middle cerebral artery occlusion (pMCAO). A total of 55 rats were randomly divided into 3 groups: pMCAO group, pMCAO and treadmill exercise (pMCAO + Ex) group, and sham-operated group. Rats in pMCAO + Ex group underwent treadmill exercise for 16 days. Neurological function was evaluated by modified Neurological Severity Scores (mNSS). High-performance liquid chromatography-electrochemical detection system was used to determine the content of 5-HT in cortex tissues. The protein levels of 5-HT1AR, BDNF and synaptophysin were measured by Western blot. The mNSS in pMCAO + Ex group was lower than that in pMCAO group on day 19 post-MCAO (p < 0.001). The content of 5-HT dropped to 3.81 ± 1.86 ng/ml in pMCAO group (43.84 ± 2.05 ng/ml in sham-operated group), but increased in pMCAO + Ex group (10.06 ± 1.80 ng/ml). The protein expressions levels of synaptophysin, 5-HT1AR and BDNF were downregulated after cerebral ischemia (p < 0.05), and upregulated after treadmill exercise (p < 0.05). These results indicate that treadmill exercise improves neurologic function, enhances neuronal plasticity and upregulates the levels of 5-HT, 5-HT1AR and BDNF in rats with pMCAO.     

Endurance exercise facilitates relearning of forelimb motor skill after focal ischemia

Endurance exercise (i.e. running), by up-regulating brain-derived neurotrophic factor (BDNF) and other modulators of synaptic plasticity, improves attention and learning, both critical components of stroke rehabilitation. We hypothesized that, following middle cerebral artery occlusion in male Sprague-Dawley rats, endurance exercise would act synergistically with a challenging skilled forelimb task to facilitate motor recovery. Animals were randomly assigned to one of four rehabilitation conditions: no rehabilitation, running only, reach training only, and reach training preceded by running (run/reach training) for 5 weeks beginning 5 days after stroke. The behavioral outcome, morphological change and mRNA expression of proteins implicated in neuroplasticity (BDNF, synapsin I and microtubule-associated protein 2) were compared. Endurance exercise on a motorized running wheel, prior to reach training, enhanced recovery of skilled reaching ability but did not transfer to gross motor skills such as postural support (forelimb asymmetry test) and gait (ladder rung walking test). Microtubule-associated protein 2 staining density in the run/reach group was slightly enhanced in the contralateral motor cortex compared with the contralateral sensory and ipsilateral cingulate cortices, suggesting that running preceding reach training may have resulted in more dendritic branching within the motor cortex in this group. No significant differences in mRNA levels were detected among the training paradigms; however, there was a trend toward greater BDNF and synapsin I mRNA in the reaching groups. These findings suggest that exercise facilitates learning of subsequent challenging reaching tasks after stroke, which has the potential to optimize outcomes in patients with stroke.     

Transgenic mice overexpressing truncated trkB neurotrophin receptors in neurons show increased susceptibility to cortical injury after focal cerebral ischemia

It has been suggested that the increased production of endogenous BDNF after brain insults supports the survival of injured neurons and limits the spread of the damage. In order to test this hypothesis experimentally, we have produced transgenic mouse lines that overexpress the dominant-negative truncated splice variant of BDNF receptor trkB (trkB.T1) in postnatal cortical and hippocampal neurons. When these mice were exposed to transient focal cerebral ischemia by occluding the middle cerebral artery for 45 min and the damage was assessed 24 h later, transgenic mice had a significantly larger damage than wild-type littermates in the cerebral cortex (204 +/- 32% of wild-type, P = 0.02), but not in striatum, where the transgene is not expressed. Our results support the notion that endogenously expressed BDNF is neuroprotective and that BDNF signaling may have an important role in preventing brain damage after transient ischemia.     

Mild experimental brain injury differentially alters the expression of neurotrophin and neurotrophin receptor mRNAs in the hippocampus

The molecular events responsible for impairments in cognition following mild traumatic brain injury are poorly understood. Neurotrophins, such as brain-derived neurotrophic factor (BDNF), have been identified as having a role in learning and memory. We have previously demonstrated that following experimental brain trauma of moderate severity (2.0-2.1 atm), mRNA levels of BDNF and its high-affinity receptor, trkB, are increased bilaterally in the hippocampus for several hours, whereas NT-3 mRNA expression is decreased. In the present study, we used in situ hybridization to compare BDNF, trkB, NT-3, and trkC mRNA expression in rat hippocampus at 3 or 6 h after a lateral fluid percussion brain injury (FPI) of mild severity (1.0 atm) to sham-injured controls at equivalent time points. Mild FPI induced significant increases in hybridization levels for BDNF and trkB mRNAs, and a decrease in NT-3 mRNA in the hippocampus. However, in contrast to the bilateral effects of moderate experimental brain injury, the present changes with mild injury were restricted to the injured side. These findings demonstrate that even a mild traumatic brain injury differentially alters neurotrophin and neurotrophin receptor levels in the hippocampus. Such alterations may have important implications for neural plasticity and recovery of function in people who sustain a mild head injury.     

Electrical stimulation accelerates and increases expression of BDNF and trkB mRNA in regenerating rat femoral motoneurons

Electrical stimulation promotes the speed and accuracy of motor axonal regeneration. The positive effects of stimulation are mediated at the cell body. Here we characterize the effect of electrical stimulation on motoneuronal expression of BDNF and its receptor, trkB, two genes whose expression levels in motoneurons correlate with regeneration and are regulated by electrical activity in a variety of neurons. We used semiquantitative in situ hybridization to measure expression of mRNA encoding BDNF and the full-length trkB receptor at intervals of 8 h, 2 days and 7 days after unilateral femoral nerve cut, suture, and stimulation. Expression in regenerating motoneurons was compared to that of contralateral intact motoneurons. BDNF and trkB signals were not significantly upregulated 8 h and 2 days after femoral nerve suture and sham stimulation. By 7 days, there was a 2-fold increase in both BDNF and trkB mRNA expression. In contrast, stimulation of cut and repaired nerves for only 1 h led to rapid upregulation of BDNF and trkB mRNA by 3-fold and 2-fold, respectively, within the first 8 h. The stimulation effect peaked at 2 days with 6-fold and 4-fold increases in the signals, respectively. Thereafter, the levels of BDNF and trkB mRNA expression declined to equal the 2-fold increase seen at 7 days after nerve repair and sham-stimulation. We conclude that brief electrical stimulation stimulates BDNF and trkB expression in regenerating motoneurons. Because electrical stimulation is known to accelerate axonal regeneration, we suggest that changes in the expression of BDNF and trkB correlate with acceleration of axonal regeneration.     

Quantitative analysis of contralateral hemisphere hypertrophy and sensorimotor performance in adult rats following unilateral neonatal ischemic-hypoxic brain injury

The immature nervous system is capable of considerable compensatory reorganization following injury. This has been studied extensively following many different types of injury in humans and laboratory animals. One common risk factor associated with perinatal brain injury that has been associated with such reorganization is an ischemic-hypoxic event. Using the established Levine model of neonatal ischemic-hypoxia (IH) to create unilateral striatal, cortical and hippocampal damage, we investigated anatomical changes in the undamaged hemisphere contralateral to the injury. Specifically, we measured cross-sectional area (mm²) of brain sections at the level of + 1.20 and - 2.12 mm from bregma. In addition, we examined sensorimotor deficits in these animals during development and as adults by measuring the amount of time that the animals were able to remain on a rotating treadmill. Our results show that some animals exhibited hypertrophy in the hemisphere contralateral to the lesion as compared to measurements taken from normal control animals. Additionally, we have demonstrated that, following IH, animals that showed significant contralateral whole-hemisphere hypertrophy were able to remain on the Rota-Rod treadmill significantly longer than the animals that did not exhibit this hypertrophy. We conclude that there are compensatory reorganizational changes that occur in the undamaged hemisphere contralateral to injury in some animals following neonatal ischemic-hypoxic brain injury. Furthermore, our data suggest that this plasticity in the contralateral hemisphere may be functionally advantageous.     

Exogenous brain-derived neurotrophic factor prevents postischemic downregulation of [3H]muscimol binding to GABA(A) receptors in the cortical penumbra

We have previously shown that exogenous application of brain-derived neurotrophic factor (BDNF) reduces infarct volume in the cortical ischemic penumbra after experimental focal ischemia [Stroke 31 (2000) 2212-2217]. Since BDNF is known to modulate the expression and function of various neurotransmitter receptors, we addressed the question whether BDNF may act via modification of postischemic ligand binding to excitatory NMDA and AMPA and/or inhibitory GABA(A) receptors, respectively. Transient focal cerebral ischemia was induced in male Wistar rats for 2 h using the suture occlusion technique. A period of 30 min after occlusion of the middle cerebral artery, BDNF (300 microg/kg per hour in vehicle; n=5) or vehicle alone (n=5) was continuously infused intravenously for 3 h. Using quantitative receptor autoradiography, postischemic ligand binding of [(3)H]MK-801, [(3)H]AMPA and [(3)H]muscimol was analyzed in the ischemic core, the ischemic cortical penumbra and corresponding regions of the contralateral hemisphere. Transient focal ischemia caused a significant reduction of [(3)H]muscimol binding to GABA(A) receptors within the ischemic cortical penumbra of placebo-treated rats. This was largely prevented by exogenous application of BDNF. [(3)H]MK-801 and [(3)H]AMPA binding values were also reduced in the cortical penumbra and the corresponding area of the contralateral hemisphere. Our data suggest that the neuroprotective effect of BDNF against ischemic damage in the cortical penumbra may be mediated in part by maintained activity of the inhibitory GABAergic system which likely counteracts glutamate induced excitotoxicity.     

BDNF binding to truncated trkB.T1 does not affect gene expression

Truncated trkB.T1 is a splice variant of the neurotrophin receptor trkB. In spite of its abundance, and ability to bind and internalize BDNF, it is not clear whether it can transmit BDNF signaling. We tested this hypothesis by searching for proteins binding the evolutionarily conserved cyto-domain of trkB.T1, and by studying BDNF-induced changes of gene expression through DNA microarrays. Cells bearing trkB.T1 receptors presented morphological changes. However, no cytoplasmic interactors of trkB.T1 were found. In addition, BDNF-dependent modulation of gene expression was detected in cells bearing trkB.TK but not trkB.T1 receptors. These results suggest that the main function of trkB.T1 is to regulate local availability of neurotrophins and that it is unable to sense changes in BDNF availability.     

Target-derived BDNF (brain-derived neurotrophic factor) is essential for the survival of developing neurons in the isthmo-optic nucleus

Neurons in the peripheral nervous system depend on single neurotrophic factors, whereas those in the brain are thought to utilize many different trophic factors. This study examined whether some neurons in the brain critically depend on a single trophic factor during development. Neurons in the isthmo-optic nucleus (ION) of chick embryos respond to exogenous brain-derived neurotrophic factor (BDNF). Relatively high concentrations of endogenous BDNF were present in the ION of 14-18-day-old chick embryos. ION target cells in the retina were immunolabeled for BDNF but showed surprisingly low levels of BDNF mRNA. These data suggest that ION target cells derive some BDNF from other retinal sources. No BDNF mRNA was detected in the ION itself. ION neurons had a very efficient retrograde transport system for BDNF and exogenous BDNF arrived in the ION intact. When the ION was deprived of endogenous trkB ligands by injection of trkB fusion proteins in the eye, cell death of ION neurons was enhanced, and this effect was mimicked by BDNF-specific blocking antibodies in the eye. TrkB fusion proteins in the retina induced cell death of ION neurons prior to visible effects on ION target cells in the retina. Immunolabel for endogenous BDNF was sparse in pyknotic ION neurons, suggesting that ION neurons with low BDNF content were eliminated by apoptosis. These data show that BDNF is an essential target-derived trophic factor for developing ION neurons and thereby validate the neurotrophic hypothesis for at least one neuronal population in the brain.     

Protein-energy malnutrition alters hippocampal plasticity-associated protein expression following global ischemia in the gerbil

Previously it has been demonstrated that protein-energy malnutrition (PEM) impairs habituation in the open field test following global ischemia. The present study examined the hypothesis that PEM exerts some of its deleterious effects on functional outcome by altering the post-ischemic expression of the plasticity-associated genes brain-derived neurotrophic factor (BDNF), its receptor tropomyosin-related kinase B (trkB), and growth-associated protein-43 (GAP-43). Male, Mongolian gerbils (11-12 wk) were randomized to either control diet (12.5% protein) or PEM (2% protein) for 4 wk, and then underwent 5 min bilateral common carotid artery occlusion or sham surgery. Tympanic temperature was maintained at 36.5 ± 0.5°C during surgery. Brains collected at 1, 3 and 7 d post-surgery were processed by in-situ hybridization or immunofluorescence. BDNF and trkB mRNA expression was increased in hippocampal CA1 neurons after ischemia at all time points and was not significantly influenced by diet. However, increased trkB protein expression after ischemia was exacerbated by PEM at 7 d in the CA1 region. Post-ischemic GAP-43 protein increased at 3 and 7 d in the CA1 region, and PEM intensified this response and extended it to the CA3 and hilar regions. PEM exerted these effects without exacerbating CA1 neuron loss caused by global ischemia. The findings suggest that PEM increases the stress response and/or hyper-excitability in the hippocampus after global ischemia. Nutritional care appears to have robust effects on plasticity mechanisms important to recovery after brain ischemia.     

Expression of trkB and trkC receptors and their ligands brain-derived neurotrophic factor and neurotrophin-3 in the murine amygdala

The neurotrophin brain-derived neurotrophic factor (BDNF) and neurotrophin-3 (NT-3) and their cognate receptors, trkB and trkC, have a variety of physiological brain functions, ranging from cell survival to mechanisms involved in learning and memory and long-term potentiation (LTP). LTP can be induced in the cortex and hippocampus, as well as within the amygdala. However, the role of neurotrophins in amygdalar LTP is largely unknown. Expression patterns of BDNF and NT-3 and their cognate receptors in the adult mouse amygdala have not been analyzed in detail. We have therefore examined the expression of trkB, trkC, BDNF, and NT-3 mRNA and protein in different amygdalar nuclei as well as in the hippocampal areas CA1-CA3 and the dentate gyrus. The distribution pattern of trkB, trkC, BDNF, and NT-3 mRNA in the murine hippocampus is comparable to that seen in rats. Within most amygdalar nuclei, a moderate BDNF mRNA expression was found; however, BDNF mRNA was virtually absent from the central nucleus. No expression of NT-3 mRNA was found within the amygdala, but trkC mRNA-expressing cells were widely distributed within this brain region. trkB mRNA was strongly expressed in the amygdala. Because trkB is expressed in a full-length and a truncated form (the latter form is also expressed by nonneuronal cells), we also investigated the distribution of full-length trkB mRNA-expressing cells and could demonstrate that this version of trkB receptors is also widely expressed in the amygdala. These results can serve as a basis for studies elucidating the physiological roles of these receptors in the amygdala.     

Downhill training upregulates mice hippocampal and striatal brain-derived neurotrophic factor levels

This study examined the effects of downhill treadmill exercise on brain-derived neurotrophic factor (BDNF) protein on the hippocampus and striatum of mice. Twenty-four adult mice were assigned to three groups: non-runners control, level or downhill (16 degrees decline) running exercise. The exercise schedule consisted of progressive treadmill running for 5 days week(-1) over 8 weeks. Blood lactate levels classified exercise intensity as moderate to high. Both training types increased citrate synthase activity of the soleus muscle when compared to untrained controls. While level running increased BDNF levels selectively in the hippocampus (68.5%), the eccentric running resulted in a pronounced BDNF increase in both the hippocampus (137.0%) and the striatum (49.9%). Further studies will specify whether the observed alterations in BDNF are due to downhill-induced upregulation or complex learning-induced mechanisms that influence BDNF levels in these brain regions.     

Upregulation of BDNF mRNA and trkB mRNA in the nigrostriatal system and in the lesion site following unilateral transection of the medial forebrain bundle

We have performed unilateral transection of the medial forebrain bundle (MFB) and studied BDNF mRNA and trkB mRNA levels at different postlesion times in the nigrostriatal system by means of in situ hybridization. BDNF mRNA levels were transiently induced in the substantia nigra pars compacta at 1 day postaxotomy. The disposition of BDNF mRNA expressing cells at this postlesion time in substantia nigra mimicked that of the dopaminergic neurons expressing the mRNA for the dopamine transporter. TrkB mRNA levels remained unaltered in the ventral mesencephalon at the different postlesion times examined-1 to 14 days. In contrast, trkB mRNA levels were significantly induced in the striatum at the longer postlesion time examined-14 days-when all neurodegenerative events are completed. It is becoming apparent that nigral BDNF mRNA levels are anterogradely transported to its target tissue in striatum. However, following axotomy, the lesion site represents a second potential target for BDNF action. Consequently, we also analyzed the pattern of mRNA expression for BDNF and trkB at the lesion site where dopaminergic axons are disconnected. There, we found notable inductions of both BDNF mRNA and trkB mRNA levels at 4 days postaxotomy. BDNF mRNA expressing cells were confined at the site of axotomy, which coincided precisely to that showing induction of trkB mRNA. Altogether, our results anticipate promising trophic roles of BNDF in the injured nigrostriatal system.     

Involvement of brain‐derived neurotrophic factor (BDNF) in the functional elimination of synaptic contacts at polyinnervated neuromuscular synapses during development

We use immunohistochemistry to describe the localization of brain‐derived neurotrophic factor (BDNF) and its receptors trkB and p75^NTR in the neuromuscular synapses of postnatal rats (P6–P7) during the synapse elimination period. The receptor protein p75^NTR is present in the nerve terminal, muscle cell and glial Schwann cell whereas BDNF and trkB proteins can be detected mainly in the pre‐ and postsynaptic elements. Exogenously applied BDNF (10 nM for 3 hr or 50 nM for 1 hr) increases ACh release from singly and dually innervated synapses. This effect may be specific for BDNF because the neurotrophin NT‐4 (2–8 nM) does not modulate release at P6–P7. Blocking the receptors trkB and p75^NTR (with K‐252a and anti‐p75‐192‐IgG, respectively) completely abolishes the potentiating effect of exogenous BDNF. In addition, exogenous BDNF transiently recruits functionally depressed silent terminals, and this effect seems to be mediated by trkB. Calcium ions, the L‐type voltage‐dependent calcium channels and protein kinase C are involved in BDNF‐mediated nerve ending recruitment. Blocking experiments suggest that endogenous BDNF could operate through p75^NTR receptors coupled to potentiate ACh release in all nerve terminals because the anti‐p75‐192‐IgG reduces release. However, blocking the trkB receptor (K‐252a) or neutralizing endogenous BDNF with the trkB‐IgG fusion protein reveals a trkB‐mediated release inhibition on almost mature strong endings in dual junctions. Taken together these results suggest that a BDNF‐induced p75^NTR‐mediated ACh release potentiating mechanism and a BDNF‐induced trkB‐mediated release inhibitory mechanism may contribute to developmental synapse disconnection. © 2009 Wiley‐Liss, Inc.