Neural circuits in the mammalian brain comprise hundreds to millions of neurons, connected with each other in complex and highly specific ways. Studies over the past few decades have identified a handful of genes that are critical for wiring up a handful of circuits. It is clear, however, that we have just scratched the surface. A key problem is that the number of plausible candidates is large, while methods for assessing their roles in vivo are laborious, expensive and time-consuming. Large scale genetic tests and screens of the type that have been transformative in invertebrates (flies and worms) are infeasible in the most-used mammalian model system, the mouse for many practical reasons such as long generation time and small litter size.
In their recent collaborative study with the Zipursky lab at UCLA, Sarin et al. harnessed recent advances in molecular biology to design an alternative strategy. First, they coupled RNA sequencing (RNAseq) technology with methods for efficient cell-type purification to identify genes expressed by the main neural types in a simple circuit at three developmental stages. Then, they adapted CRISPR-based genetic engineering to mutate individual genes in cells of the circuit in otherwise wild-type animals. The system they chose was the outer retina, in which rod and cone photoreceptors form synapses on two classes of interneurons called bipolar and horizontal cells. Synapse formation by photoreceptors is specific in three different ways: cellular (rods and cones synapse on rod bipolar cells and cone bipolar cells, respectively); subcellular (rods and cones synapse on horizontal cell axons and dendrites, respectively); and laminar (rod and cone terminals are confined to outer and inner strata, respectively, within a thin neuropil).
Based on their RNAseq data, Sarin et al. chose 30 genes encoding recognition and signaling molecules to mutate in developing rods and bipolar cells; they then analyzed cells and synapses in “patchs” of mutated cells. Because the vectors included fluorescent proteins, they were able to rapidly identify the patches and subject them to histological analysis using confocal microscopy. Several gave interesting phenotypes, and of these they chose Wnt5, a morphogen and axon guidance molecule, to analyze in detail. They showed that it is produced by rod bipolar cells and acts on rods. Taking advantage of the ability to mutate many genes fairly rapidly, they were then able to test eight potential receptors and intracellular mediators, thereby characterizing the Wnt signal transduction pathway. It would have been prohibitive to test so many candidate intercellular signals and receptors using conventional mouse genetics. However, to validate the method they compared results to those obtained by germ-line mutants of two genes, and found satisfying agreement. They concluded that Wnt5 signaling regulates a previously unappreciated step in formation of the outer retinal circuit, in which processes of pre- and postsynaptic cell are confined to a thin neuropil layer, thereby ensuring that synaptic partners can find each other efficiently.
Together, the results of Sarin et al. not only provide new information about how the outer retina wires up during development, but also provides a general approach can be applied to other regions of the brain.