A neuroscientist argues for a broader approach to brain mapping.

Efforts to map all of the connections between the brain's neurons — known as synapses — are gathering momentum. Neural 'wiring diagrams' have even garnered a label: the 'connectome'. But I worry that the connectome will be a waste of time and money, and that we'll eventually need to redo the whole thing.

Why am I so negative? Although the connectome is ambitious, it's not ambitious enough. As currently envisioned, it ignores most brain cells as well as possible functional connections between those cells.

Although you wouldn't know it from all the attention they receive, neurons are a relatively minor type of brain cell, making up less than 10% of the human brain. And synapses between neurons make up only a small subset of all possible functional connections in the brain. Most brain cells are actually glia, which have long been neglected by neuroscientists owing to their lack of electrical signalling. But glia are increasingly being recognized as having important roles in brain function.

For example, consider the recent study of adenosine and sleep led by Philip Haydon and Marcos Frank at the University of Pennsylvania in Philadelphia (M. M. Halassa et al. Neuron 61, 213–219; 2009). Adenosine binds to receptors on neurons, thereby regulating neuronal signalling. Interestingly, adenosine seems to represent 'sleepiness': it accumulates during wakefulness and dissipates during sleep. Where does it come from? It is generated from adenosine triphosphate (ATP), which is secreted by astrocytes — a major type of glia.

Therefore, if we want to map the functional brain connections controlling sleep, we need to include glia and the extracellular space between glia and neurons. If we're going to understand brain function by mapping the brain, we need to include most of the brain in our map.

Discuss this paper at http://blogs.nature.com/nature/journalclub