Except within the thalamus, very few labeled cells co-stained for the inhibitory neuronal marker GAD67. Immunofluorescence staining in sections from mice injected at P3 demonstrated that the majority of cells transduced by AAV8 at this age were S100β-positive astrocytes (n = 3, Fig. 5K). These data indicate that the timing of intraventricular AAV8 injection can

strongly influence both the overall transduction efficiency as well as cell-type specificity. This unique property of AAV8 expands click here the potential repertoire for AAV targeting based on infection time, and provides a novel approach to astrocyte-specific transgene delivery. One advantage of viral-mediated gene transfer is that the viral titer can be GSK 3 inhibitor easily adjusted to alter the transduction efficiency. We tested whether viral dilution could be reliably harnessed to generate controllable transgene mosaicism, and at what dilutions different serotypes were effective. We prepared serial dilutions of AAV8-YFP and AAV1-YFP from ~1010 to ~108 particles/μL; 2 μL of each dilution was bilaterally injected into the lateral ventricles of

P0 pups (n = 5–8 for each condition). Dilution of both AAV8 and AAV1 reduced the transduction efficiency throughout the brain, with far less fluorescent protein expression at 109 particles/μL than at 1010 particles/μL (Fig. 6). Dilution of AAV8 resulted in progressively fewer neurons being transduced at 109 particles/μL than at 1010 particles/μL, and fewer still at 108 particles/μL than at 109 particles/μL. However, the spread of AAV8 infection was essentially identical among the different dilutions. In contrast, the spread of transduction with AAV1 declined sharply at the first 10-fold dilution to 109 particles/μL (Fig. 6A). To directly compare the transduction efficiency of AAV8 with AAV1, we co-injected the two

serotypes at the Thiamine-diphosphate kinase same titer (109 particles/μL, n = 4 per condition). As when injected alone at these titers, AAV8 transduced neurons throughout the brain, whereas transduction by AAV1 was largely restricted to the choroid plexus (Fig. 6B). The strong transduction of the ventricular epithelia suggests that high-affinity binding of AAV1 to these cells left little virus free to enter the rest of the brain. Next, we optimised the viral titers needed to attain reliable high- and low-density expression with each serotype based on serial dilution of each preparation. High-density neuronal transduction was consistently achieved by intraventricular injection of 4.0 × 109–2.0 × 1010 particles/hemisphere of AAV8 or 4.0 × 1010 particles/hemisphere of AAV1. Injection of virus at these concentrations left only a small population of wild-type cells surrounded by a field of transduced neighbors, ideal for studying the cell-extrinsic effects of a virally-delivered transgene. A complementary transduction pattern was attained by low-titer injections using 4.0 × 107 particles/hemisphere of AAV8 or 2.