Sequential Triggered Delivery of Adeno-Associated Virus from a Collagen-Gag Scaffold

Introduction: Chronic wounds are a major issue facing healthcare providers and patients. These wounds do not progress through the wound healing cascade correctly, leaving them chronically inflamed. Increasingly, tissue engineering scaffolds are being developed to help promote wound healing. Separately, adeno-associated virus (AAV) are becoming ever more popular as a means of genetic editing. However, to date, no one has created a system capable of timed release of multiple AAV that can drive coordination in wounds. We set out to develop a system capable of triggerable AAV therapeutic release, that also allowed diffusion of a different AAV therapeutic. To provide a suitable scaffold for tissue growth, hydrogel loaded with AAV was placed within pockets in a collagen-gag scaffold. These pockets can then be individually triggered by ultrasound. Using a reporter AAV we characterized our system, demonstrating the ability to achieve both high and low level diffusion. Then we showed ultrasound can be used to trigger release from the low-level diffusion pocket, while leaving the other unaffected giving our system precisely timed control. Lastly, we showed a clinically relevant example of our system in use by releasing two pro-angiogenic AAVs sequentially. This system can be readily adapted with other viral vectors for application in diverse biological pathways.

Materials and

Methods: Hydrogels were prepared by ionically crosslinking alginate solution with a mixture containing calcium sulfate and loading it with adeno-associated virus. Collagen-glycosaminoglycan (collagen-GAG) scaffolds were fabricated with pockets by using an inverted mold design and crosslinked both dehydrothermally and chemically. Gels were loaded with 1x10^9 to 1x10^10 genomic copies/gel that were designed to express enhanced green fluorescent protein (eGFP), mCherry, vascular endothelial growth factor (VEGF), or platelet derived growth factor (PDGF). 50µL of these gels were injected into collagen-glycosaminoglycan (collagen-GAG) pockets. Scaffolds for release studies were stored in a 12 well plate (for diffusion/ultrasound experiments) over a period of 10 days and gels allowed to diffuse or the gels were ultrasound treated. Ultrasonication was applied using a Sonics Vibracell VCX130 was used and set to 30 seconds at 30% amplitude. Conditioned media was collected and added to HEK293t cells, and the percentage of cells expressing genes of interest was confirmed via flow cytometry or ELISA.

Results, Conclusions, and Discussions: We confirmed alginate will stay in the collagen-GAG pocket when injected (See Fig 1A). AAV successfully released from the gels, with diffusion being sustained for up to 9 days (See Fig 1B). However, the viral load affects how much release/transfection occurs with the low-dose loaded gels yielding 6% and the high 100%. Past studies have looked at virus release from alginate for local delivery but there are fewer explorations on dosage effects. While most studies see a high level of viral release they don’t explore triggering release long term after system implantation. By starting with a low level of baseline diffusion, we then successfully triggered viral release with ultrasonication generating a burst of protein expression at a user-defined time days after starting the experiment (See Fig 1C). By combining both the high level diffusion and low level diffusion with an ultrasonic trigger we achieved a sequential release of AAV. This was demonstrated by releasing a constant level of AAV-VEGF-eGFP and then triggering a AAV-PDGF-mCherry release (See Fig 1D). On the day of ultrasonic triggering the test group expressed significantly increased PDGF as compared to non-triggered control (901 pg/ml vs 260 pg/ml). This shows that our system allows for truly sequential release, yet given the pocketed design approach, we can easily modify the device for a variety of uses by simply changing out the AAV loaded into the gels and ultrasound regime. This system allows sequential release of adeno-associated virus from alginate pockets in a collagen-GAG scaffold. To date AAV release systems have either been diffusion driven or triggered, whereas our system is capable of both diffusion and triggered release. By further incorporating this system within a proven regenerative scaffold, it provides both the correct extracellular matrix environment, and the right signaling to allow users to better mimic natural signal pathways when designing treatments for a wide range of tissue engineering and other pathologies.

Acknowledgements (Optional):