(H-319) An Endosomal Escape Trojan Horse Platform to Improve Cytosolic Delivery of Nucleic Acids

Endocytosis has long served as a major bottleneck toward nucleic acid delivery as DNA drugs remain trapped within endosomes. Current trends to overcome endosomal entrapment provide varied success; however, active delivery agents such as endosomal escape peptides (EEP) have emerged as a prominent strategy to improve cytosolic delivery. Yet, these membrane active agents have poor selectivity for endosomal membranes, leading to toxicity in clinical trials. A hallmark of the endosome is the acidic environment, which aids in degradation of foreign materials. Here, I develop a pH-triggered spherical nucleic acid (SNA) that provides a smart ASO release with endosomal acidification and selective membrane disruptive activity. We anchor i-Motif DNA to a gold nanoparticle core (AuNP), where the complement strand contains both an antisense oligonucleotide (ASO) sequence and a functionalized EEP. By orienting the EEP toward the AuNP core, the EEP is inactive until released in response to endosomal acidification. This mimics a Trojan Horse! In this study, we characterize a small library of i-Motif duplexes to develop the endosomal trigger. We evaluate antisense efficacy using HIF-1a, a hypoxic indicator upregulated in many cancers, and demonstrate dose dependent activity through RT-qPCR. Further, modular inclusion of the AuNP and EEP leads to significant improvements in ASO efficacy. Finally, we show that the Trojan Horse SNA benefits synergistically from nuclease- and pH-driven release strategies, with increased ASO endosomal escape efficiency. Overall, this study develops a modular platform that improves cytosolic delivery of nucleic acid therapeutics and offers key insight for overcoming intracellular barriers.

In this study, the pH-response of the i-Motif DNA trigger is characterized through UV-Vis absorbance and fluorescence spectroscopy measurements. I compare i-Motif (structured C-rich DNA) to a scrambled and non-C-rich strand to show sequence and composition specificity. Absorbance measurements are used to characterize single stranded pH-response whereas fluorescence measurements characterize duplex dissociation. Through FRET, a Cy3B dye (pH-independent dye) is quenched by an AuNP or a blackhole quencher when bound, and an increase in fluorescence is used to demonstrate pH release. To study endosomal escape peptide (EEP) activity to enhance DNA cytosolic delivery, I screened a small EEP library that were conjugated to DNA through copper click chemistry and purified using reverse-phase HPLC. I compared cellular activity of DNA-EEP conjugates using flow cytometry and measured mean fluorescence intensity in a high throughput manner within HeLa cells. After selecting an ideal i-Motif and EEP, I measured therapeutic efficacy within HeLa cells using HIF-1a as the mRNA of interest. I tested many aspects of the modular platform to evaluate activity such as with/without AuNP core, with/without EEP, with/without transfection agent, with/without pH-responsive DNA, and with/without nuclease resistant modifications at varying concentrations. All groups were normalized to a scrambled antisense sequence within the respective delivery group, an untreated control, as well as housekeeping gene of interest (18S). Lastly, confocal and fluorescence lifetime imaging microscopy was used to visualize experiments and evaluate endosomal escape. Experiments were measured in triplicate at physiological conditions with biological replicates to evaluate statistical significance.

I quantify the tunability of i-Motif DNA and determine that the transition pH increases with increasing C-repeat density. Additionally, a scrambled i-Motif with the same composition, but different sequence has diminished pH-response whereas a strand with low C-base composition does not display detectable pH transition (pH 5 to 8). pH transition characterization using Boltzmann sigmoidal distributions (full width at half maximum) also shows that the profile narrows with increased C-composition. These findings imply that i-Motif pH response is sequence dependent and tunable.

A major significance of this study is the application of i-Motif DNA within physiological conditions as it is not well characterized in therapeutics, and it has especially not been characterized as a duplex trigger mechanism to drive the release of its complement. To maintain physiological significance, the melting temperature must be >50^oC to be stable for extended incubation periods which directly hinders i-Motif dissociation due to increased activation energy. A screen of nine i-Motif duplexes reveals little to mild trends from both increasing C-density as well as i-Motif overhang length; however, the screen did reveal a pair that has a functional duplex release at pH 5.5, which is responsive in pH ranges found in late endosomes. Although an ideal design would trigger closer to neutral pH (pH 6.5 to 6.0) to avoid nucleases found in late endosomes/lysosomes, pH 5.5 is reasonable for the scope of this study. Further, EEP and AuNP functionalization does not significantly hinder pH response and increases cellular uptake and antisense efficacy in a modular manner.

The study concludes by evaluating the cellular release profile and antisense efficacy to determine the influence of both pH- and nuclease-driven release mechanisms. We find that combining these, using a non-backbone modified i-Motif anchor, enhanced overall antisense activity. Groups containing pH-responsive anchors (both PS- and PO-backbones) had increased endosomal escape efficiency compared to non-pH responsive groups. As this strategy maintains activity in a dose-dependent manner, we conclude by developing a platform technology named DNA EndosomaL Escape Vehicle Response (DELVR) that is modular in design and improves endosomal escape efficiency for antisense oligonucleotides.