Introduction: RNA-based therapeutics are innovative and emerging approaches to target human diseases. Nonetheless, specific agents for cardiac diseases are still missing. One of the main reasons for the lack of breakthroughs in cardiac research is a lack of delivery methods in ex vivo models, and that single-cell cultures are often poorly bridging to in vivo settings. Herein, we present a delivery strategy for ncRNAs in living myocardial slices (LMS). This well-established ex vivo model involves ultrathin myocardial slices from the left ventricle of various species and allows testing of pharmacological agents and environmental changes and the effects on the multicellular myocardium.

Purpose: This study aims at establishing a reproducible highly translational model for pre-clinical testing of RNA-therapeutics.

Methods: LMS were prepared from the left ventricle of human, pig and rat tissue, mechanically stretched with 3D-printed rings and cultured for 24 to 48 h. All transfections were initially performed with Cy3-labeled siRNAs and miRNAs, as well as an eGFP modRNA, to observe intracellular uptake. Several transfection reagents (Lipofectamine RNAiMAX, Protamine Sulfate) and cell type-specific delivery systems (Glyko-PEIs) were used. The fluorescence emissions were captured via Cytation (Agilent Technologies, Inc.) every two hours, then cryosections were generated to evaluate the penetration depth. Finally, targeted reduction of gene expression was investigated using miRNAs targeting different genes (e.g. HMOX2, FGF1, SPP1). Both siRNAs and miRNAs were first validated for functionality in a luciferase-reporter assay and cell culture (e.g. H9C2 cells and neonatal rat cardiomyocytes (NRCM)). LMS were then treated with the identified ncRNA candidates, and gene expression was examined using RNA-Isolation and RT-qPCR.

Results: Cytation live imaging demonstrated a significant and dose-dependent intracellular uptake for both siRNAs and miRNAs using Lipofectamine RNAiMax. The addition of protamine sulfate, a positively charged peptide, resulted in an even faster and more substantial increase in cellular uptake. On the contrary, these effects were not observed with modRNA. In addition, we investigated the efficiency of polyethyleneimines chemically bound to different sugar molecules as transfection agents. These so-called Glyko-PEIs thus target individual cardiac cell types with different Sugar-Receptor patterns. GlykoOne-PEI proved to be the most efficient transporter and, even at lower concentrations, outperformed the other PEIs bound to sugar molecules. Similarly to RNAiMax, the addition of protamine sulfate showed a significant increase in the uptake rate of the RNA-Glyko-PEI complexes. However, LMS cryosections showed that siRNA could penetrate only in the first 2-3 cell layers of both sides of approximately 15 cell layers. The gene knockdown capabilities of siRNAs and in silico predicted miRNAs on the target genes were confirmed with luciferase reporter assays and RT-qPCR from NRCMs transfected with the respective candidates. Nevertheless, subsequent RT-qPCRs of RNA from LMS treated with up to 200nM HMOX2-siRNA and 500nM miRNA failed to achieve functional knockdown of HMOX2.

Conclusions: We herein report a highly translational cell type-specific delivery method for siRNAs and miRNAs in LMS. This platform could be possibly applied to pre-clinical testing of ncRNA candidates or to discover novel therapeutic strategies for cardiac diseases.