Enabling mRNA Therapy for Fragile X Syndrome

Introduction: Safe delivery of mRNA to the brain will revolutionize the treatment of neurological diseases. Thanks to the COVID-19 mRNA vaccines and numerous therapy-based clinical trials, it is now clear that lipid nanoparticles (LNPs) are a clinically viable means to deliver RNA therapeutics. However, LNP-mediated mRNA delivery to the brain remains elusive. Over the past decade, numerous studies have shown that cells in our body communicate with each other via extracellular vesicles (a.k.a. exosomes), which are around 100 nm in size and consists of lipid bilayer membrane similar to synthetic lipid-based nanocarriers. We hypothesized that rationally designed LNPs based on extracellular vesicle (EV) mimicry would enable efficient delivery of RNA therapeutics to brain cells without undue toxicity. As a paradigm application, we develop an mRNA therapy for Fragile X Syndrome (FXS) using EV-inspired LNPs.

Materials and

Methods: LNPs consisted of four components similar to the formulation used in the mRNA COVID-19 vaccines (Moderna and Pfizer): ionizable lipid, cholesterol, helper lipid and polyethylene glycol (PEG)-lipid. For the in vitro screen, we tested ten classes of helper lipids based on their abundance in extracellular vesicle membranes, commercial availability, and large-scale production feasibility while keeping rest of the LNP components unchanged. The transfection kinetics of GFP mRNA encapsulated in LNPs and doped with 16 mol% of helper lipids was tested using SIM-A9 cells (murine microglia cell line). The successful LNP candidates (enableing >80% transfection efficacy) were then tested in vivo to deliver luciferase mRNA to the mice brain via intrathecal injection. LNPs were then tested to deliver Cre recombinase mRNA in Ai9 mouse brain to identify the cells being targetted in vivo using flow cytometry and histology. The best performing LNP formulations were assessed for delivering Cas9 mRNA and CD81 sgRNA (model protein) in C57 mouse brain to study brain targeted gene editing. The best LNP formulation was then assessed to deliver human and mouse fragile mental retardation protein (FMRP) mRNA in adult FMR1 KO mouse to assess FXS phenotype rescue.

Results, Conclusions, and Discussions: LNP-mediated transfection of GFP mRNA resulted in stable GFP expression for 5 days in SIM-A9 cells in vitro. The luciferase expression in vivo confirmed successful delivery of mRNA throughout the brain (Figure 1A). Cre recombinase experiment (Figure 1B) showed that mRNA was successfully delivered to a range of different cells in the mouse brain including microglia, neurons, neural stem cells, oligodendrocytes and astrocytes. Sanger sequencing showed that CRISPR-Cas9 editing was successful in ~40% of cells in the mouse brain. Both mouse and human FMRP expression attenuated social behavior deficits and stereotyped and repetitive behavior in FMR1 KO mice without affecting WT mouse behaviour.
In conclusion, we have developed a library of LNPs that can transfect in vivo brain cells with high efficacy. RNA therapeutics have a broad range of applications for neurological diseases where one or more protein is dysregulated (up-/down-regulated) leading to the disease manifestation. The technology developed here can be used for potentially treating such diseases by delivering single or multiple mRNAs to different brain cells, and holds great potential as a tool to study the brain function.

Acknowledgements (Optional):