Abstract
The Cre-lox system is an indispensable tool in neuroscience research for targeting gene deletions to specific cellular populations. Here we assess the utility of several transgenic Cre lines, along with a viral approach, for targeting cerebellar Purkinje cells. Using a combination of a fluorescent reporter line (Ai14) to indicate Cre-mediated recombination and a floxed Dystroglycan line (Dag1(flox)) we show that reporter expression does not always align precisely with loss of protein. The commonly used Pcp2(Cre) line exhibits a gradual mosaic pattern of Cre recombination in Purkinje cells from P7-P14, while loss of Dag1 protein is not complete until P30. Ptf1a(Cre) drives recombination in precursor cells that give rise to GABAergic neurons in the embryonic cerebellum, including Purkinje cells and molecular layer interneurons. However, due to its transient expression in precursors, Ptf1a(Cre) results in stochastic loss of Dag1 protein in these neurons. Nestin(Cre), which is often described as a "pan-neuronal" Cre line for the central nervous system, does not drive Cre-mediated recombination in Purkinje cells. We identify a Calb1(Cre) line that drives efficient and complete recombination in embryonic Purkinje cells, resulting in loss of Dag1 protein before the period of synaptogenesis. AAV8-mediated delivery of Cre at P0 results in gradual transduction of Purkinje cells during the second postnatal week, with loss of Dag1 protein not reaching appreciable levels until P35. These results characterize several tools for targeting conditional deletions in cerebellar Purkinje cells at different developmental stages and illustrate the importance of validating the loss of protein following recombination.