, 2010). Although many other Cre lines targeting dopamine receptor-expressing neurons exist, other lines tend to have

sparser label in striatum and may only represent a restricted subset of D1R- or D2R-expressing projection neurons. Direct-pathway MSNs in the dorsal striatum directly project to SNr, with major projections to the EP learn more and a smaller fraction of projections to the GP. As expected, when monosynaptic rabies virus was injected into AAV-infected D1R-Cre mice, dense projections associated with the direct pathway were labeled, terminating in SNr and EP, with some projections to GP (Figure 2A). Fluorescent label in GP in D1R-Cre mice is a combination of fibers traversing to EP/SNr, direct projections from D1R-expressing MSNs, and projections from monosynaptically connected D2R-expressing MSNs (D2R MSNs are known to frequently form connections onto D1R MSNs [Planert et al., 2010 and Taverna et al., 2008]). When the striatum was examined at higher power, a stark border between striatum and globus pallidus is detectable, emphasizing the specificity of infection to striatal neurons (Figure 2B). In contrast, when monosynaptic rabies virus was injected into AAV-infected D2R-Cre animals, AZD6244 ic50 projections associated with the indirect pathway were obvious (Figure 2C), heavily innervating GP but sparing EP and SNr. Few, if any, direct-pathway

MSN axons are visible because D1R→D2R MSN connectivity is extremely low (Planert et al., 2010 and Taverna et al., 2008). At higher power, the sharp border between striatum and globus pallidus was again detectable, but heavy labeling of axon terminals in GP again emphasizes the specificity of virus targeting to indirect-pathway MSNs. The sites of primary injection

within the striatum were constrained to the same area of striatum, as diagrammed in (Figures 2E and 2F) and charted in (Figure 2G). As noted above, even though the injection sites were somewhat near the border of GP, genetic restriction of primary infection to either D1R- or D2R-expressing MSNs provided nearly complete restriction of primary infection to dorsal striatum (Figures 2B and 2D). A few cells in the GP were sometimes labeled (Figure 2A), indicating that these cells likely provide direct input to neurons in the dorsal striatum (Bevan et al., 1998). However, due to their proximity to the nearly injection site, these inputs were not analyzed further. Rabies virus infection was rarely detectable at the injection site in wild-type animals (Figure S1); in two animals injected with AAV9-FLEX-hGTB, some rabies label was detectable near the injection site, likely due to tiny amounts of leak TVA expression. Four animals injected with AAV9-pEF1α-FLEX-GTB had no detectable rabies virus label anywhere in the brain. Wild-type mice never had any label outside of striatum, indicating that rabies glycoprotein is not expressed at high enough levels in the absence of Cre to allow for transsynaptic spread of rabies virus.