Targeted delivery and endosomal cellular uptake of DARPin-siRNA bioconjugates: Influence of linker stability on gene silencing

Abstract

Specific cell targeting and efficient intracellular delivery are major hurdles for the widespread therapeutic use of nucleic acid technologies, particularly siRNA mediated gene silencing. To enable receptor-mediated cell-specific targeting, we designed a synthesis scheme that can be generically used to engineer Designed Ankyrin Repeat Protein (DARPin)-siRNA bioconjugates. Different linkers, including labile disulfide-, and more stable thiol-maleimide- and triazole- (click chemistry) tethers were employed. Crosslinkers were first attached to a 3′-terminal aminohexyl chain on the siRNA sense strands. On the protein side thiols of a C-terminal cysteine were used as anchoring sites for disulfide- and thiol-maleimide conjugate formations, while strain-promoted azido-alkyne cycloadditions were carried out at a metabolically introduced N-terminal azidohomoalanine. After establishing efficient purification methods, highly pure products were obtained. Bioconjugates of EpCAM-targeted DARPins with siRNA directed at the luciferase gene were evaluated for cell-specific binding, uptake and gene silencing. As shown by flow cytometry and fluorescence microscopy, all constructs retained the highly specific and high-affinity antigen recognition properties of the native DARPin. As expected, internalization was observed only in EpCAM-positive cell lines, and predominantly endolysosomal localization was detected. Disulfide linked conjugates showed lower serum stability against cleavage at the linker and thus lower internalization into endosomes compared to thiol-maleimide- and triazole-linked conjugates, yet induced more pronounced gene silencing. This indicates that the siRNA payload needs to be liberated from the protein in the endosome. Our data confirm the promise of DARPin-siRNA bioconjugates for tumor targeting, but also identified endosomal retention and limited cytosolic escape of the siRNA as the rate-limiting step for more efficient gene silencing.