MOLECULAR AND DEVELOPMENTAL NEUROSCIENCE Brain-derived neurotrophic factor-mediated effects on mitochondrial respiratory coupling and neuroprotection share the same molecular signalling pathways Anthony Markham, 1 Ian Cameron, 1 Rasneer Bains, 1 Paul Franklin, 1 Janos P. Kiss, 2 Leslie Schwendimann, 3,4 Pierre Gressens 3,4,5 and Michael Spedding 6 1 Institute of Pharmacy, Chemistry and Biomedical Sciences, School of Health, Natural and Social Science, University of Sunderland, Sunderland, UK 2 Servier Research Institute for Medicinal Chemistry, Budapest, Hungary 3 Inserm, U676, Paris, France 4 Universite ´ Paris 7, Faculte ´ de Me ´ decine Denis Diderot, IFR02, Paris, France 5 AP HP, Ho ˆ pital Robert Debre ´ , Service de Neurologie Pe ´ diatrique, Paris, France 6 Les Laboratoires Servier, 50 Rue Carnot, Suresnes 92150, France Keywords: brain-derived neurotrophic factor, glial cell line-derived neurotrophic factor, mice, mitochondria, neurodegeneration, oxidative metabolism Abstract Intracerebral injection of ibotenate into mouse pups induced grey matter lesions and white matter cysts; co-administration of brain- derived neurotrophic factor (BDNF) produced a dose-dependent reduction in these lesions. In contrast, glial cell line-derived neurotrophic factor (GDNF) had no significant effect, whereas nerve growth factor (NGF) or interleukin-1b (IL-1b) resulted in dose- dependent exacerbation. The neuroprotective effects of BDNF were abolished by co-administration of anti-BDNF antibody or MEK inhibitors, or ABT-737, a BH3 mimetic and Bcl-2 antagonist. The actions of BDNF, GDNF and NGF were measured in a parallel in vitro study on the oxidative metabolism of mouse brain mitochondria. BDNF produced a concentration-dependent increase in the respiratory control index (RCI, a measure of respiratory coupling efficiency, ATP synthesis, and organelle integrity) when co- incubated with synaptosomes containing signal transduction pathways; but GDNF failed to modify RCI, and NGF had only weak effects. BDNF had no effect on pure mitochondria, and enhanced oxidation only when complex I substrates were used. The effect of BDNF was inhibited by anti-BDNF antibody, MEK inhibitors or ABT-737, and also by IL-1b, indicating that the mitochondrial effects are mediated via the same MEK–Bcl-2 pathway as the neuroprotection. The complex I inhibitor rotenone, a compound implicated in the aetiology of Parkinson’s disease, inhibited both the in vitro mitochondrial and in vivo neuroprotective effects of BDNF. The ability of BDNF to modify brain metabolism and the efficiency of oxygen utilization via a MEK–Bcl-2 pathway may be an important component of the neuroprotective action observed with this neurotrophin. Introduction Brain-derived neurotrophic factor (BDNF) is the main activity- dependent neurotrophin in the central nervous system, and couples neuronal activity to neurotrophic effects, balanced between excitatory gutamatergic and inhibitory GABAergic systems, via its tyrosine kinase receptor (TrkB) (Schinder & Poo, 2000; Kalb, 2005). However, it is normally essential for activity and metabolic coupling to run in parallel, and, in order to evaluate this, we have developed a novel synaptosome–mitochondrial preparation that allows the astro- cyte ⁄ nerve ending ⁄ mitochondrial unit to be evaluated in vitro; this preparation was fully validated for the rat (Markham et al., 2004), and is now extended to mouse brain mitochondria. In this preparation, BDNF, the activity-dependent neurotrophin, was shown to increase the respiratory coupling index (RCI) of rat brain (but not of liver or heart) mitochondria, through a MEK kinase mechanism via complex 1. The approach was specifically designed to mimic the synaptosomal nerve ending but with the oxygen electrode directly in contact with the mitochondria, which is the only way to measure oxygen use directly. On electron microscopy, the reconstituted sonicated synaptosome– mitochondria mixture looks similar to nerve endings (Markham et al., 2004). Brain-derived neurotrophic factor has also marked neuroprotective effects against excitoxic lesions induced via a-amino-3-hydroxy-5- methyl-4-isoxazole-propionic acid (AMPA) and N-methyl-d-aspartate excitotoxins via the TrkB–MEK–mitogen-activated protein kinase (MAPK) cascade in newborn mice, in a model that we have extensively characterized (Gressens et al., 1997; Husson et al., 2002; Dicou et al., 2003; Plaisant et al., 2003a). During the course Correspondence: Michael Spedding, as above. E-mail: michael.spedding@fr.netgrs.com Received 22 April 2010, accepted 17 November 2011 European Journal of Neuroscience, Vol. 35, pp. 366–374, 2012 doi:10.1111/j.1460-9568.2011.07965.x ª 2012 The Authors. European Journal of Neuroscience ª 2012 Federation of European Neuroscience Societies and Blackwell Publishing Ltd European Journal of Neuroscience