Experimentally restricting transgene expression exclusively to astrocytes has proven difficult. Using adeno-associated-virus-mediated gene transfer, we assessed two commonly used glial fibrillary acidic protein promoters: the full-length version gfa2 (2,210-bp human glial fibrillary acidic protein [GFAP] promoter) and the truncated variant gfaABC1D (681-bp GFAP promoter). The capacity to drive efficient, but also cell-type specific, expression of the EGFP in astrocytes was tested both in vitro in rat primary cortical cultures as well as in vivo in the rat striatum. We observed an efficient, but not entirely astrocyte-specific, gfa2-driven reporter expression. gfaABC1D exhibited a weaker activity, and most importantly, off-target, neuronal expression of the transgene occurred in a larger fraction of cells. Therefore, we explored the potential of a microRNA (miR)-specific target-sequence-based approach for abolishing off-target expression. When miR124 target sequences were incorporated into the 3' UTR, neuronal gene expression was effectively silenced. However, unexpectedly, the insertion of an additional sequence in the 3' UTR clearly diminished transgene expression. In conclusion, the gfaABC1D promoter on its own is not sufficient to specifically target transgene expression to astrocytes and is not well suited for AAV-based gene targeting, even if short promoter sequences are required. The combination with a miR de-targeting sequence represents a promising experimental strategy that eliminates off-target, neuronal expression.
Keywords: GFAP; adeno-associated virus (AAV); astrocyte; gene therapy; gfa2; gfaABC(1)D; microRNA; promoter; targeting; transgene.
Copyright © 2017 The Authors. Published by Elsevier Inc. All rights reserved.