Comparing lipid nanoparticle mediated delivery of Cas9 mRNA and protein for cystic fibrosis gene editing

Introduction: Cystic fibrosis is an autosomal recessive disorder that is estimated to affect around 70,000 people in the world, and is caused by mutations in the CFTR gene. This disease results in non-functional CFTR proteins, which prevents the transport of chloride ions from the interior of epithelial cells to the airway surface liquid in the lungs, resulting in the accumulation of thick mucus that reduces breath capacity and gas exchange. Multiple clinical trials have been approved to treat CF with small drug molecules to alleviate symptoms, but fail to accommodate many mutations or address the gene itself. Gene editing platforms have been developed to address the source of the mutations in the genome, but this requires machinery to deliver to the site of action in a cell, which is difficult for pre-existing methodologies. Nucleic acid-containing lipid nanoparticles offer a promising new platform due to their previous approval, ability to encapsulate various cargoes, and high modularity in targeting specific organs, like the lungs. LNPs have been advantageous in gene editing platforms through their ability to deliver CRISPR-Cas9 components either through Cas9 mRNA or Cas9 protein. LNPs can be used to perform editing in the genome to correct various sequences as well as target the CFTR gene itself. Further, though the use of cationic lipids has been used for systemic delivery of LNPs to the lung via the bloodstream, LNPs can be further optimized to target cell subtypes associated with CF such as ionocytes and basal epithelial cells.

Materials and

Methods: Lipid-Nanoparticle Formulation
Five ionizable lipids were synthesized and purified. Lipid nanoparticles were formulated with DOTAP, ionizable lipid, phospholipid, cholesterol, and lipid PEG combined at previously optimized ratios in an ethanol phase and Cas9 mRNA or protein, donor DNA, and guide RNA in an aqueous phase. The phases were mixed microfluidically and dialyzed in PBS.
Biodistribution
Dialyzed LNPs containing Cas9 mRNA or protein, non-targeting donor DNA, and scrambled guide RNA were combined with DiR and dialyzed in PBS. LNPs were intravenously injected into C57BL6 black mice, sacrificed after 6 hr, and organs were imaged for fluorescence at ex. 750 em. 780 nm for the DiR stain to determine tropism of the LNP.
Editing analysis in vitro
H1299 lung cells were plated at 10,000 cells/well and treated with LNPs 24 hrs after. 72 hrs after treating, the cells were trypsinized and genomic DNA was extracted. The purified genomic DNA was amplified with PCR and purified using agarose gel electrophoresis. Sanger sequencing was performed on the DNA and indels were calculated with TIDER.
LoxP editing analysis in vivo
LNPs were formulated as described with guide RNAs for the LoxP cassette, scrambled DNA, and either Cas9 protein or mRNA. LNPs were injected into Ai9 mice (n = 4) in three doses of 0.15mg/kg sgRNA over three days. Mice were sacrificed 8 days after the final injection. Organs were isolated and imaged with IVIS for tdTomato fluorescence at ex. 554 em. 581, and then processed for flow cytometry.

Results, Conclusions, and Discussions: To identify lead candidates for editing, a library of unique ionizable lipid nanoparticles encapsulating Cas9 (either protein or 5-methoxyuridine modified mRNA), guide RNA (sgRNA), and donor DNA for blue fluorescent protein (BFP) were delivered to H1299 GFP+ cells in vitro and then processed for flow cytometry three days after treatment to evaluate BFP knock-in to the genome through BFP+ fluorescence.

From this initial study, top performers of both types of formulations were used in a biodistribution study to determine the tropism of each LNP. DiR-stained LNPs were injected intravenously to follow lipid-nanoparticle trafficking. All formulations demonstrated lung tropism after analyzing the heart, lung, liver, kidney, and spleen, consistent with their prior optimization for this target. The top performer is contrasted with a standard base lipid in Figure 2A, showing Lipid 1 exhibits significantly higher proportions of lung tropism as compared to the base lipid, with both formulations.

To evaluate editing in vivo, Lipid 1 was used in protein and mRNA formulations and injected into Ai9 mice to the formulations. Controls with scrambled sgRNAs were used for both formulations. As seen in Figure 2C, tdTomato fluorescence indicates successful editing. After organs were isolated, tdTomato mean fluorescence counts were evaluated and similar high lung tropism was seen as the earlier biodistribution experiment in Figure 2B. However, this study led to the intriguing contrast in mRNA formulation editing efficacy and protein formulations, where the protein formulation exhibited over two times more tdTomato fluorescence than the mRNA formulations when normalized to background fluorescence. The enhanced editing capabilities of the protein formulation as compared to the mRNA formulation demonstrates the potential advantage of Cas9 protein in lung-tropic gene editing compared to Cas9 mRNA due to the kinetic optimization of delivering Cas9 protein along with its sgRNAs to the gene of interest, while Cas9 mRNA must be translated into Cas9 protein and trafficked to the genome with the delivered sgRNAs, which poses a more difficult mechanism. Further investigation into the editing efficacy of Cas9 protein versus Cas9 mRNA will be used for corrective editing of the CFTR gene itself, addressing the common mutations f508del and G542X.

Acknowledgements (Optional):