Dendrites come in different shapes and sizes, much like most things in this world. In Drosophila, the homeodomain protein Cut is crucial for the development of dendrites. In a recent paper in Cell, Grueber and colleagues asked how the morphology of a subset of Drosophila neurons, the dendritic arborization sensory neurons (DASNs), develop their characteristic branched dendrites.

There are four classes of DASNs (I–IV), which differ in their dendrite morphologies. Class I neurons have the simplest morphology; class II and IV neurons have more complex branching patterns and greater dendritic fields than class I neurons; and class III neurons have a distinctively intricate 'spiked' organization — numerous short terminal branches that extend from the main dendritic trunk.

Drawing from previous studies of DASNs, the authors hypothesized that levels of Cut might correlate with dendritic morphology. This idea was confirmed by their observations. Highly spiked neurons had high levels of Cut, whereas neurons with simple dendrite morphology had low or undetectable levels.

To discern the function of Cut, the authors performed loss-of-function experiments and predicted that, as class III neurons show the highest level of Cut expression, a reduction in Cut levels would lead to a marked change in phenotype. Indeed, class III cut− neurons had no higher-order dendritic branching and a reduced number of spikes on the dendrite trunk.

In class IV cut− neurons, higher-order branches were lost and overall dendrite length reduced. Loss of Cut function in some class II neurons led to the inhibition of dendritic growth and branching. Interestingly, class I neurons, which showed little or no Cut expression in the wild type, remained unaffected in the cut− clones.

If high levels of Cut are responsible for class-specific morphology, then expressing Cut in neurons with low wild-type Cut levels (as in class I neurons) might lead to the development of higher-order dendritic branches. Indeed, the authors found that dendrites of class I neurons that were forced to express Cut (or a mammalian Cut homologue) showed altered morphology that resembled that of class III dendrites.

So, Cut has a role in class-specific dendrite patterning, but how does it perform this function? Does it act in an ON/OFF fashion, switching on the machinery that is required for specific branching? Or, does its function change according to the level at which it is expressed? By increasing levels of Cut in class II and class IV neurons, it was possible to alter the morphologies of their dendrites to become more like those of class III neurons. This ability of neurons to 'jump classes' indicates that Cut functions in a level-dependent manner — the differing levels among neurons being important for their different morphologies.

So, Cut functions to develop the finer morphological features of the specific dendrites. The report indicates that dendrite morphology is not fixed at the time of neuronal birth but rather, it is moulded throughout early development.