It is easy to think of synaptic communication as one-way traffic, but in many cases, postsynaptic cells can also signal back to the presynaptic neuron to regulate synapse structure and function. The Drosophila neuromuscular junction (NMJ) illustrates how retrograde signalling contributes to synapse homeostasis, and two studies reported in Neuron identify new components of the signalling mechanism.

In the Drosophila embryo, motor neurons compensate for the growth of their target muscles by making extra synaptic boutons and by raising the number of neurotransmitter release sites at each synapse. These events rely on bone morphogenetic protein (BMP) signalling through the Wishful thinking (Wit) receptor on the presynaptic terminal. Until now, the Wit ligand that is released by the postsynaptic muscle cell was unknown. McCabe et al. knocked out the glass bottom boat (gbb) gene, which codes for a BMP homologue. This manipulation produced a phenocopy of the Wit-knockout phenotype, which is characterized by reductions in NMJ size and activity, and defects in synaptic ultrastructure. In vitro studies confirmed that Gbb acts as a ligand for Wit. These findings point towards a role for Gbb in controlling the growth of the NMJ.

In the second study, Haghighi et al. investigated the modulation of neurotransmission by retrograde signalling. They found that reducing Ca2+/calmodulin-dependent kinase II (CaMKII) activity in the postsynaptic muscle cell led to an increase in neurotransmitter release from the presynaptic neuron. Conversely, activating CaMKII constitutively in the muscle cell reduced the level of synaptic transmission. The authors propose that CaMKII senses the level of synaptic activity by monitoring Ca2+ levels. Synaptic activity causes an influx of Ca2+ into the muscle. This activates CaMKII, which in turn triggers a retrograde signal that inhibits neurotransmitter release from the presynaptic terminal. Once again, a BMP signal seems to be responsible, but it is not yet known whether Gbb is involved.

A third study by Pratt et al., which was also reported in Neuron, showed that postsynaptic CaMKII also mediates synapse remodelling in the mammalian brain. This indicates that at least some of the Drosophila retrograde signalling machinery has been conserved during evolution, and it seems likely that further examination of the Drosophila NMJ will bring to light more aspects of retrograde signalling that apply across the animal kingdom.