A major hurdle for exploring RNA interference (RNAi) in a therapeutic setting is still the issue of in vivo delivery of small RNA molecules (siRNAs). The chemical modification of polyethylenimines (PEIs) offers a particularly attractive avenue towards the development of more efficient non-viral delivery systems. Here, we explore tyrosine-modified polyethylenimines with low or very low molecular weight (P2Y, P5Y, P10Y) for siRNA delivery. In comparison to their respective parent PEI, they reveal considerably increased knockdown efficacies and very low cytotoxicity upon tyrosine modification, as determined in different reporter and wildtype cell lines. The delivery of siRNAs targeting the anti-apoptotic oncogene survivin or the serine/threonine-protein kinase PLK1 (polo-like kinase 1; PLK-1) oncogene reveals strong inhibitory effects in vitro. In a therapeutic in vivo setting, profound anti-tumor effects in a prostate carcinoma xenograft mouse model are observed upon systemic application of complexes for survivin or PLK1 knockdown, in the absence of in vivo toxicity. We thus demonstrate the tyrosine-modification of (very) low molecular weight PEIs for generating efficient nanocarriers for siRNA delivery in vitro and in vivo, present data on their physicochemical and biological properties, and show their efficacy as siRNA therapeutic in vivo, in the absence of adverse effects.

The delivery of small interfering RNAs (siRNA) is an efficient method for gene silencing through the induction of RNA interference (RNAi). It critically relies, however, on efficient vehicles for siRNA formulation, for transfection in vitro as well as for their potential use in vivo. While polyethylenimines (PEIs) are among the most studied cationic polymers for nucleic acid delivery including small RNA molecules, polypropylenimines (PPIs) have been explored to a lesser extent. Previous studies have shown the benefit of the modification of small PEIs by tyrosine grafting which are featured in this paper. Additionally, we have now extended this approach towards PPIs, presenting tyrosine-modified PPIs (named PPI-Y) for the first time. In this study, we describe the marked improvement of PPI upon its tyrosine modification, leading to enhanced siRNA complexation, complex stability, siRNA delivery, knockdown efficacy and biocompatibility. Results of PPI-Y/siRNA complexes are also compared with data based on tyrosine-modified linear or branched PEIs (LPxY or PxY). Taken together, this establishes tyrosine-modified PPIs or PEIs as particularly promising polymeric systems for siRNA formulation and delivery.

Among non-viral vectors, cationic polymers, such as poly(propylene imine) (PPI), play a prominent role in nucleic acid delivery. However, limitations of polycationic polymer-based DNA delivery systems are (i) insufficient target specificity, (ii) unsatisfactory transgene expression, and (iii) undesired transfer of therapeutic DNA into non-target cells. We developed single-chain antibody fragment (scFv)-directed hybrid polyplexes for targeted gene therapy of prostate stem cell antigen (PSCA)-positive tumors. Besides mono-biotinylated PSCA-specific single-chain antibodies (scFv(AM1-P-BAP)) conjugated to neutravidin, the hybrid polyplexes comprise β-cyclodextrin-modified PPI as well as biotin/maltose-modified PPI as carriers for minicircle DNAs encoding for Sleeping Beauty transposase and a transposon encoding the gene of interest. The PSCA-specific hybrid polyplexes efficiently delivered a GFP gene in PSCA-positive tumor cells, whereas control hybrid polyplexes showed low gene transfer efficiency. In an experimental gene therapy approach, targeted transposition of a codon-optimized p53 into p53-deficient HCT116^{p53−/−/PSCA} cells demonstrated decreased clonogenic survival when compared to mock controls. Noteworthily, p53 transposition in PTEN-deficient H4^PSCA glioma cells caused nearly complete loss of clonogenic survival. These results demonstrate the feasibility of combining tumor-targeting hybrid polyplexes and Sleeping Beauty gene transposition, which, due to the modular design, can be extended to other target genes and tumor entities.