gene silencing by siRNA delivery to the corneal epithelium in a keratin-12- bioluminescence mouse model

Abstract

Purpose: The barrier properties of the corneal epithelium as well the large size and negative charge of siRNAs (short interfering RNAs) make topical delivery of these macromolecules challenging. We have generated a bioluminescence mouse model that expresses firefly luciferase (luc2) in the anterior corneal epithelium. Using live animal imaging, we have developed a rapid in vivo assay of gene silencing therapies aimed at diseases of the corneal epithelium.

Methods: The codon-optimized luciferase gene luc2 was knocked into the murine keratin K12 (Krt12) locus by gene targeting in embryonic stem cells to generate C57BL/6J mice expressing luciferase in the anterior epithelium. The 3’ untranslated region of luc2 contained a multiple targeting cassette (MTC) - a non-coding synthetic construct consisting of several human corneal dystrophy mutations and their flanking sequences. This one mouse model can be used to assess allele-specific siRNAs against any one of several human mutations. The resultant reporter gene animals were treated with Accell “self-delivery” siRNA (non-targeting control versus luc2 siRNA) by intrastromal injection (n=3 mice per treatment). Then, animals were treated with siRNA targeting the human keratin 12 mutation Leu132Pro present in the MTC; left eye (native- or Accell-control siRNA) versus right eye (native- or Accell- Leu132Pro siRNA; n=5 mice per treatment). Bioluminescence was quantified using a Xenogen IVIS Lumina imager with LivingImage 3.2 software. Statistics were performed using SPSS v.21.

Results: All targeting siRNA measurements were normalized to non-targeting control measurements. “Self-delivery” Accell siRNAs produced potent and sustained in vivo gene silencing. At days 2-7, >50% silencing of reporter gene expression was obtained with Accell siRNA. An Accell version of the allele-specific Leu132Pro siRNA worked as well as Accell siRNA against the luc2 coding sequence. In comparison, native siRNA only resulted in significant knockdown at day 1 (p<0.05).

Conclusions: We have developed and validated a robust bioluminescence mouse model for use in assessing siRNA therapeutics in vivo. Importantly, our ability to successfully monitor siRNA activity in corneal epithelial cells in real-time allows us to develop and assess topical formulations to achieve siRNA topical delivery for the ocular surface, bringing these therapeutics closer to clinical use.