Study design: Descriptive histologic analysis of spinal cord gene therapy.
Objective: To maximize protein expression in rat spinal cord using recombinant adenoassociate virus viral vector.
Summary of background data: There are few reports of spinal cord genetic transfer. There have been no reports that compare techniques to increase protein expression through genetic alterations or have illustrated successful genetic transfer to spinal cord astrocytes.
Methods: Adenoassociate virus constructs were packaged using three separate plasmids: a cis plasmid with the expression cassettes (pAM/neuron-specific enolase/green fluorescent protein/woodchuck posttranscriptional regulatory element/simian virus 40/polyadenylase or pAM/glial fibrillary acid protein/green fluorescent protein/woodchuck posttranscriptional regulatory element/simian virus 40/polyadenylase), the Ad-adenoassociate virus helper trans plasmid, and the essential region from the adenovirus genome (pFDelta6). The adenoassociate virus 2/5 capsid gene replaces the adenoassociate virus 2 capsid region in the trans construct, resulting in a different cellular tropism. Thirty-two adult (300-375 g) male Sprague-Dawley rats underwent L1 laminectomies. A total volume of 6 microL was injected directly into the spinal cord parenchyma at a rate of 600 nL/min with adenoassociate virus 2/glial fibrillary acid protein/green fluorescent protein, adenoassociate virus 2/neuron-specific enolase/green fluorescent protein, adenoassociate virus 2/5/glial fibrillary acid protein/green fluorescent protein, or adenoassociate virus 2/5/neuron-specific enolase/green fluorescent protein and either a low- (4 x 10(8)) or high-titer (1 x 10(10)) viral solution.
Results: The gene expression (green fluorescent protein reporter) was present in the cell bodies and axonal processes of all adenoassociate virus/green fluorescent protein constructs. However, a greater spread of virus was observed in rats injected with adenoassociate virus 2/5 compared with adenoassociate virus 2. In addition, more neurons were transduced with adenoassociate virus 2/5 than adenoassociate virus 2, and green fluorescent protein expression in neurons transduced with adenoassociate virus 2/5 appeared more intense compared with adenoassociate virus 2 neurons. The difference observed between adenoassociate virus 2 and adenoassociate virus 2/5 at 4 x 10(8) genomic particles/mL was not as profound when the virus titer was raised to 1 x 10(10) genomic particles/mL. Green fluorescent protein expression was observed in astrocytes following injection of rat spinal cords with either adenoassociate virus 2 or adenoassociate virus 2/5 carrying the glial fibrillary acid protein/green fluorescent protein construct. However, unlike neuron-specific enolase-driven expression, there was less overall expression, but a substantial increase in green fluorescent protein expression was observed with adenoassociate virus 2/5 compared with adenoassociate virus 2 with high virus titers. Furthermore, unlike the neuron-specific enolase promoter, glial fibrillary acid protein-driven expression of green fluorescent protein was not restricted to astrocytes alone. The glial fibrillary acid protein construct was able to transfect glial cells and maintain glial expression.
Conclusion: Adenoassociate virus can readily transduce spinal cord neurons and is an efficient nonpathologic vector to deliver expression cassettes. Increased titers and the adenoassociate virus 2/5 serotype appeared to maximize expression.