Over the past few years, it has become evident that local synthesis of proteins in axons and dendrites has a vital role in neural development and synaptic plasticity. We still know relatively little about how local protein synthesis is regulated, although there is some evidence that extracellular signals are involved. In a study reported in The Journal of Neuroscience, Schratt et al. present compelling new findings that indicate that brain-derived neurotrophic factor (BDNF) regulates the local translation of a defined subset of mRNAs at the synapse.

BDNF is known to stimulate the translation of several mRNAs in neurons, but it is unclear whether this represents a local or a global translational response. To investigate this phenomenon further, Schratt et al. carried out a genome-wide screen to identify mRNAs that became associated with the translational machinery in response to BDNF.

The authors isolated cortical neurons from the brains of embryonic day 18 rats, and cultured the neurons for 14 days to allow them to form synapses before exposing them to BDNF. They extracted mRNA from the cells, and centrifuged this preparation on a sucrose density gradient to separate out the fraction that was presumed to be undergoing active translation (the polysomal fraction). The authors then synthesized complementary DNA from this fraction, and hybridized it to an Affymetrix rat microarray. Out of about 3,500 mRNAs that were expressed by the neurons, as many as 143 entered the polysomal fraction in response to BDNF.

The authors further narrowed down their search to mRNAs that are regulated by BDNF at the level of transcription initiation. Specifically, they looked for cases in which BDNF-induced polysome association was prevented by rapamycin, an inhibitor of the mTOR–PI3K (mammalian target of rapamycin–phosphatidylinositol 3-kinase) signalling pathway. Forty-eight of the 143 mRNAs fell into this category.

To investigate whether BDNF was regulating local protein synthesis at the synapse, the authors focused on one of these mRNAs, Homer2, which encodes a postsynaptic adaptor protein. They found several lines of evidence to indicate that Homer2 translation was initiated in the dendritic compartment in response to BDNF. First, Homer2 mRNA was enriched in synaptoneurosomes (purified preparations of intact synapses), and the authors observed that its translation could be induced in synaptoneurosomes by the addition of BDNF. Second, in situ hybridization experiments showed that Homer2 mRNA was localized in the dendritic compartment of cultured hippocampal neurons. Last, sequences in the 3′ untranslated region of the Homer2 mRNA were sufficient to target a green fluorescent protein-coding mRNA to the dendrites of hippocampal neurons in culture.

These findings strongly support the idea that BDNF stimulates the translation of a select group of mRNAs at the synapse, through a mechanism that includes the mTOR–PI3K pathway. The genomic approach that was developed for this study could be used to build up a profile of mRNAs that are translated in response to extracellular signals at various stages of development or synaptic plasticity. It should also provide new material with which to probe the mechanisms that enable certain mRNAs to be singled out for translation at specific sites.