Spinal cord regeneration is limited by multiple barriers, for which gene delivery provides a versatile approach to target the various cellular processes. However, the ability to influence cellular responses is limited by the delivery efficiency. Biomaterial scaffolds have been used to deliver DNA locally in order to maintain an elevated concentration in the microenvironment. An approach successfully used in vitro is surface-mediated delivery, in which DNA is immobilized to the biomaterial surface. Previous studies with encapsulation and release of 800 µg plasmid into a spinal cord lateral hemisection demonstrated localized transgene expression for two weeks, with decreasing levels further away from the injury site. Relative to controlled release strategies, the surface delivery mechanism generally employs DNA complexed with a transfection reagent, which both facilitates immobilization and cellular trafficking. Additionally, this approach has the potential to reduce the quantity of DNA required.

In this report, we investigate delivery of plasmid and lipoplexes in the injured spinal cord by immobilization to a multiple channel bridge. Initial studies characterized binding, release, and transfection in vitro. The bridges were made of poly (lactide co glycolide) (PLG), coated with different extracellular matrix (ECM) components, and DNA was adsorbed to the surface comparing three different methods: i) incubation with the bridge, ii) drying DNA onto ECM coated bridges, and iii) DNA mixed with ECM and subsequent drying onto the bridge. DNA was complexed with the lipid Transfast in a 0.5:1 weight ratio. In vitro experiments indicated that fibronectin coated PLG surfaces enhanced DNA adsorption (80% efficiency) compared with collagen I, laminin, and no ECM. The two drying methods using laminin resulted in similar adsorption efficiencies as fibronectin. Importantly, transgene expression was greatest on fibronectin for all methods, with the incubation technique producing greater expression than the two drying methods. The incubation method also produced a homogeneous distribution of rhodamine-labeled DNA on the surface and had the greatest number of transfected cells. Increasing the amount of DNA incubated over 3 ug did not enhance the levels of transgene expression. Fibronectin coated bridges incubated with DNA resulted in a slightly slower release for adsorbed lipoplexes relative to plasmid (98.7% versus 74.5% after 1 day, and 99.4% and 88.4 % after 21 days).

For in vivo studies, bridges were incubated with 16 µg of DNA for 16-20 hrs and implanted in a rat spinal cord lateral hemisection model. Interestingly, adsorption of lipoplexes resulted in 4 times higher levels of transgene expression at the injury site compared with plasmid adsorption after 1 week. Expression levels 0.5 cm caudal of the injury site were increased 650-fold, whereas segments 0.5 cm rostral and 1 cm caudal had expression increased 170 fold. These results suggest that lipoplexes are more efficient in the spinal cord than plasmid, allowing for transfection of cells adjacent to the implant site. Compared with the encapsulation method of 800 µg plasmid, surface-mediated delivery of lipoplexes enhanced total transfection in the spinal cord 11.5 times. Longer timepoints are currently under investigation to assess whether the lipoplexes allow for a longer duration of transgene expression, exceeding 2 weeks. A system for efficient gene delivery into the spinal cord can be used to identify therapeutic targets and may ultimately enable regenerative therapies.