Just as each tree — oak, elm or birch — has its own distinct shape and pattern of branches and twigs, so each neuron develops a unique dendritic arbour that allows it to receive and process inputs from other neurons. How neuronal activity influences this process has now become clearer thanks to a study by Wayman et al.
Although dendritic development is partly controlled by intrinsic factors, external influences — in particular, neuronal activity caused by afferent inputs — are crucial if a neuron is to produce an appropriate and functional dendritic tree. One of the molecular pathways that mediates activity-dependent dendritic development involves the glycoprotein WNT, which works through various signalling cascades to modulate the actin cytoskeleton and thereby influence the growth of dendrites. But how does neuronal activity cause the production of WNT?
Wayman et al. carried out a careful analysis of a previously unknown calcium signalling pathway that links neuronal activity to the synthesis of WNT2, and thereby to neuronal morphology. First, using small inhibitory RNAs (siRNAs), the authors provided evidence that activity-dependent dendritic development depends on the activation of the 
- and 
-isoforms of calmodulin-dependent protein kinase I (CaMKI) by CaMK kinase (CaMKK). As previously shown in other systems by the Soderling laboratory, this leads to activation of the ERK signalling pathway and subsequently of the transcription factor CREB.
When the authors screened for CREB target genes, they discovered that WNT2 was bound and activated by CREB in response to neuronal activation and calcium influx. In this study, the authors used a stimulation protocol that activated N-methyl-D-aspartate receptors (NMDARs), pointing towards a pathway for dendritic development that goes from NMDAR activation and calcium influx, through CaMKK and CaMKI to CREB and finally to WNT2.
However, other studies have implicated calcium influx through L-type voltage-gated calcium channels, rather than NMDARs, in dendritic modelling. It is possible that both are involved, depending on the type of neuron and stimulus. Further work should clarify this, as well as help to pinpoint how different cues work together — or oppose each other — in the development of neuronal dendrites.
