Jiang and J.-Y. Kim, unpublished data). Past studies have used AAV-GFP virus for in vivo imaging following stereotaxic injection into mice and monkeys (Stettler et al., 2006; Lowery et al., 2009). Local injection has the benefit of eliminating background fluorescence from distant projection neurons, but at the cost of having less control over the density of labeled cells due to a sharp gradient in transduction from the site of injection. Neonatal transduction provides improved

consistency see more in the expression pattern, and offers a serviceable alternative to Thy1-XFP lines (Feng et al., 2000), particularly when working with models that already require multiple transgenes or modified alleles. Viral transgenesis also Sunitinib provides access to neurons not labeled in the Thy1-XFP mice, notably Purkinje cells of the cerebellum, which in the past have required acute injection of synthetic dyes for morphological study in vivo (Gobel & Helmchen, 2007). Given the high plasticity of cerebellar circuitry and the progressive but poorly understood degeneration

of Purkinje neurons in many inherited ataxias (Boyden et al., 2004; Carlson et al., 2009), chronic in vivo imaging of these arbors during motor learning and disease will likely grant new insight into cerebellar function and dysfunction. Combined with the potential to genetically manipulate the labeled neurons, neonatal viral transduction opens the possibility for experiments probing the relationship between targeted proteins, dendritic morphology, and neuronal function within single cells of the intact brain (O’Connor et al., 2009). Although this technique has many advantages over past methods, several limitations should be noted. First, as mentioned above, the small packaging size of AAV limits the length and number of transgenes that can be co-expressed. In some situations this can be overcome by trans-splicing of co-injected viruses, but this may not

be possible in every setting (Lai et al., 2005; Ghosh & Duan, 2007). Second, widespread transduction may not be ideal when more restricted expression is needed. Where available, spatial or cell-type specificity could be attained using Cre-dependent flex-signal viruses (Atasoy et al., 2008) with Cre-expressing transgenic CHIR-99021 mouse lines (e.g. nagy.mshri.on.ca/). In other cases, selectivity might be achieved using an intersectional strategy of complementary elements introduced on co-injected viruses (Dymecki et al., 2010; Haubensak et al., 2010; Fujimoto et al., 2011). Third, the level of viral gene expression varies between cells due to differences in the multiplicity of infection inherent in viral transgenesis. This fluctuation may complicate some studies of neuronal function, but may be lessened at extremes of high and low titers where infection can be maximised or dilution-limited to a single particle.