Each part of the mammalian brain holds a large number of neuronal types – at least a hundred and perhaps a thousand or more. These cells connect up in very specific ways as the nervous system forms and matures, assembling the complex circuits that underlie neural computations.
How does a neuron distinguish among many possible partners, choosing appropriate ones and rejecting others? A favorite model is that a “lock and key” mechanism exists, in which adhesive molecules on a presynaptic neuron (the “sending” side”) recognize binding partners on the proper postsynaptic partner.
But what are these molecules? Here, favorite candidates are members of large families of recognition molecules, which share general features and signaling capabilities but differ in their precise intercellular binding properties. Yet, this attractive idea has been hard to test.
In their new study, Duan and colleagues (Neuron, 2018 (PDF)) analyze the expression and roles of 20 members of such a family, the cadherins. These adhesion molecules have been implicated in many aspects of development, but tests to date have assayed only one or two at a time.
In a previous paper, Duan, Krishnaswamy et al. (Cell, 2014) showed that two different cadherins (cadherins 8 and 9) are essential for wiring up circuits in the mouse retina that calculate the direction in which an object is moving, and feeds this information to the brain. The circuit includes 8 types of neurons, with cadherin 8 and 9 each acting in one of them. Building on this work, they now show that at least 15 of the 20 cadherins are expressed by neurons in these circuits, and that at least 6 of them (cadherins 6-10 and 18) are required for circuit assembly.
To reach this conclusion, they generated single, double and triple mouse mutants that deleted one, two or three cadherins, and supplemented these reagents with other methods for decreasing or increasing cadherin expression in particular cell types. They then assayed the perturbations using structural and electrophysiological methods.
One striking result is that some of the cadherins work individually whereas others work in combination. Another is that each cadherin (or cadherin pair) is required for assembly of a specific synaptic type that serves a specific function. This striking correspondence between molecule, structure and function implies a modularity in circuit function.
The cadherins are expressed in subsets of neurons throughout the brain, and a recent collaborative study suggests that similar combinatorial coding may underlie circuit assembly in the hippocampus, a brain area implicated in memory (Basu, Duan, et al., Neuron, 2018).
Thus, as has happened frequently, lessons learned from the retina provide insights into brain development and structure generally. Moreover, the study on cadherins provides a strategy for querying other gene families that may be required for assembly of neural circuits.