Purpose: As key mediators of intercellular communication, extracellular vesicles (EVs) have emerged as a new therapeutic modality, specifically via a microRNA (miRNA) transfer mechanism. This makes them ideal gene delivery vehicles in a number of applications that can benefit from effective gene delivery, including spinal cord injury (SCI) and sepsis. However, endogenous EV RNA levels are low, potentially limiting therapeutic potency. As loading via electroporation remains controversial in the field, there is a need to explore alternate methods of loading and evaluate them in in vivo models.
Methods: In this study, Human Embryonic Kidney Cell (HEK293T) derived EVs were loaded with miRNA using two novel loading methods we have developed: sonication and pH-gradient. To demonstrate the in vivo efficacy of sonication-loaded EVs in SCI, they were co-loaded with a miR-155 inhibitor and miRNA-124, both known to modulate microglial activation post-SCI. Loaded HEK EVs were injected intravenously in SCI mice (100ug per mouse) beginning at 3 weeks post-injury and once daily for 2 days. Functional outcomes were measured via Bas Mouse Scale (BMS) for locomotion and Y-maze to asses spatial working memory both before and after injection. To evaluate EV therapeutic effect for sepsis and inflammatory responses, bone-marrow derived macrophages were treated with EVs containing miR-146. Cytokine level and protein downregulation were assessed via ELISA and Western blot, respectively.
Results: While the BMS scores remained similar between the loaded and unloaded EV groups, miRNA loaded EVs resulted in transient improved performance in Y-maze tests, indicating improved cognitive function. Though this study indicated promise in using sonication, we demonstrated that this method can degrade RNA and thus have developed a novel method of loading using a pH-gradient to circumvent this. Loading via the pH-gradient method increased the incorporation of miRNA after passive incubation, without inducing and aggregation or RNA degradation. Compared to a no-treatment control, there was a 54% knockdown of GAPDH mRNA using protonated EVs loaded with GAPDH siRNA, compared to only 4% using unmodified EVs. Loaded EVs were then evaluated in their ability to deliver miRNA in a sepsis model. Bone-marrow derived macrophages (BMDMs) were treated with loaded EVs containing miR-146, which induced a dose-dependent increase in MIP-2 secretion as well as downregulation of the Irak1 (interleukin-1 receptor-associated kinase 1), comparable to the response achieved using a commercial product to mediate miRNA transfection. In vivo studies are underway to evaluate the efficacy of this EV formulation in a murine air pouch model.
Conclusion: These studies demonstrate that the pH-gradient method is a promising new mode of EV loading and can enhance the therapeutic potential of EVs and ongoing studies involve developing these techniques to produce efficient and scalable methods of EV loading for gene delivery.
Sheng Wang– University of Maryland School of Medicine
Junfang Wu– University of Maryland School of Medicine
Wei Chao– University of Maryland School of Medicine
Steven Jay– University of Maryland