Top of pageAbstract
Adeno-associated virus (AAV) is an attractive vector for gene delivery because of its low toxicity and ability to express transgenes for many months. Collectively, the different serotypes of AAV show broad tissue and cell tropism. However, AAV does not preferentially transduce endothelia when administered in vivo. In our approaches to develop gene therapy for the CNS manifestations of the lysosomal storage diseases (mucopolysaccharidoses; [e.g. Beta-glucuronidase deficiency]) and the ceroid lipofuscinoses (e.g. CLN2 or tripeptidyl protease I [TPP-1] deficiency) endothelia remain an attractive target. Endothelial expression of TPP-1 or beta-glucuronidase results in both apical and basolateral secretion of enzyme; in the brain, this would allow exposure of recombinant enzyme to the underlying brain parenchyma. We used phage display to identify peptides that mediate selective and efficient binding to brain endothelial cells in vivo. Selected peptides were genetically incorporated into the AAV2 capsid, and peptide modified viruses (PM-AAVs) tested in vitro and in vivo. AAV2-ITR genomes were efficiently packaged into PM-AAVs and viral titers were comparable to those obtained with native AAV2 (5-10^11 vector genomes/ml). PM- and native-AAV2 vectors were subsequently injected intravenously via tail vein and vector genomes in various tissues quantified 28 days later. Approximately 10^4 more PM-AAV was present in brain relative to native AAV2. Moreover, while AAV2 genomes were predominantly found in liver, PM-AAV was highest in brain. Our primary data indicate that PM-AAV2 can be retargeted to brain by use of in vivo phage display-derived peptides. This strategy may prove valuable for correction of the CNS deficiencies of several fatal metabolic diseases.