Single neuron labeling and genetic manipulation

Those who have observed brain sections stained by the Golgi method would agree with Ramon y Cajal1: “What an unexpected sight!...everything is simple, clear and unconfused. There is no longer any question of interpretation.” The Golgi method labels a very small population of random neurons in their entirety in an otherwise unstained brain, allowing visualization of dendritic trees of individual neurons and tracing of long distance axonal projections1. It is difficult to overestimate the enormous contribution this method has brought to neuroscience. Now imagine that one can use genetic manipulation to create, at will, singlylabeled neurons in intact brain tissue or in vivo, and moreover, knock out endogenous genes in only these labeled neurons. This will help us to assess the functions of genes in single clearly labeled neurons, increasing the power of phenotypic detection; avoid pleiotropic effects of genes by focusing on the tissue and developmental stages of interest; and determine cellautonomy of gene action. The cellular and molecular mechanisms that ensure the elaborate connection and function of the nervous system can then be dissected with single neuron resolution. How can one achieve this purpose? Genetically mosaic animals, in which a subset of somatic tissues have different genotypes compared to the rest of the organisms, have long been used to attack biological problems in Caenorhabditis elegans, Drosophila and mice. Traditionally, genetic mosaic animals were generated via spontaneous or X-ray-induced mitotic recombination, resulting in progeny homozygous mutant for a candidate gene of interest in the heterozygous (and therefore phenotypically normal in most cases) background. In mice, chimaeras can also be generated by mixing embryonic cells of different genotypes. With the introduction of sequence-specific recombination systems such as FLP/FRT or Cre/LoxP, not only can one dramatically Single neuron labeling and genetic manipulation