Neuroscientists are continuously breaking down barriers. Admittedly, some of the barriers we face have been of our own creation, but this fact only makes them more impervious to challenge. For example, the idea that glial cells are just supportive elements in the nervous system and have a negligible impact on neuronal communication has pervaded the literature for a long time. But as Phil Haydon discusses in this issue, this barrier between neurons and glia has fallen, as there is increasing evidence that astrocytes can significantly modulate synaptic transmission in vivo.

Other barriers look sturdier, as they seem to reflect the straightforward division of labour in the nervous system. For instance, the functional barrier between ion channels and protein kinases is apparently clear-cut. But the recent discovery of a channel protein with kinase activity, which is discussed in our Highlights section, indicates that even this boundary will probably have to be redrawn.

Certain barriers have already fallen in some fields but are only beginning to crumble in others. One example is the question of whether molecules that participate in neural development are also involved in adult plasticity. Although this idea is widely accepted for the neurotrophins, the evidence that supports a similar role for proteins such as Eph receptors and ephrins is just beginning to accrue. Robert Gerlai suggests in his Perspective that, similar to the neurotrophins, Eph receptors and ephrins might be as important in the adult brain as they are in the embryo.

But the study of Eph receptors and ephrins also illustrates that not every barrier must fall. David Wilkinson's account of the role of these molecules during embryonic development reminds us that boundary formation is a crucial step in the making of the nervous system. Some barriers are best left untouched.