COL7A1 Editing via RNA Trans-Splicing in RDEB-Derived Skin Equivalents

Abstract

Mutations in the COL7A1 gene lead to malfunction, reduction or complete absence of type VII collagen (C7) in the skin's basement membrane zone (BMZ), impairing skin integrity. In epidermolysis bullosa (EB), more than 800 mutations in COL7A1 have been reported, leading to the dystrophic form of EB (DEB), a severe and rare skin blistering disease associated with a high risk of developing an aggressive form of squamous cell carcinoma. Here, we leveraged a previously described 3'-RTMS6m repair molecule to develop a non-viral, non-invasive and efficient RNA therapy to correct mutations within COL7A1 via spliceosome-mediated RNA trans-splicing (SMaRT). RTM-S6m, cloned into a non-viral minicircle-GFP vector, is capable of correcting all mutations occurring between exon 65 and exon 118 of COL7A1 via SMaRT. Transfection of the RTM into recessive dystrophic EB (RDEB) keratinocytes resulted in a trans-splicing efficiency of ~1.5% in keratinocytes and ~0.6% in fibroblasts, as confirmed on mRNA level via next-generation sequencing (NGS). Full-length C7 protein expression was primarily confirmed in vitro via immunofluorescence (IF) staining and Western blot analysis of transfected cells. Additionally, we complexed 3'-RTMS6m with a DDC642 liposomal carrier to deliver the RTM topically onto RDEB skin equivalents and were subsequently able to detect an accumulation of restored C7 within the basement membrane zone (BMZ). In summary, we transiently corrected COL7A1 mutations in vitro in RDEB keratinocytes and skin equivalents derived from RDEB keratinocytes and fibroblasts using a non-viral 3'-RTMS6m repair molecule.

Keywords: COL7A1; RNA therapy; RNA trans-splicing; dystrophic epidermolysis bullosa.

Figure 1

Schematic overview of the endogenous splicing processes. After the introduction of the RTM into the cell’s nucleus, the binding domain (BD) specifically targets the COL7A1 pre-mRNA by binding the intron 64/exon 65 junction. By providing an alternative splice acceptor site, the RTM facilitates the replacement of the mutated gene region within the COL7A1 pre-mRNA with the wild-type cDNA sequence (exon 65–exon 118) from the RTM, leading to a correction of the mutation. To enable an easy distinction between the cis-spliced and the trans-spliced pre-mRNA on the RNA level, the RTM carries five silent mutations within exon 65, which can be easily targeted with a polymorphism-specific reverse primer as recently described [21]. Created with BioRender.com (accessed on 1 February 2023).

Figure 2

Endogenous trans-splicing in RDEB keratinocytes and fibroblasts. (A) Correctly trans-spliced mRNA products were detected after transfection with the RTMS6m minicircle vector via sqRT-PCR, using a COL7A1-specific forward primer and an RTM-specific reverse primer, targeting all five silent mutations (black arrows) introduced at the beginning of the RTM sequence. The trans-spliced mRNA product (blue boxes) has a size of 209 bp and was detected in RDEB keratinocytes and RDEB fibroblasts transfected with the MC-RTMS6m plasmid. Healthy keratinocytes (hKc) and untransfected RDEB keratinocytes and fibroblasts were used as negative controls. (B) Sanger sequencing of the PCR product showed all silent mutations introduced by the RTMS6m and hence confirmed the accuracy of the trans-splicing reaction. (C) Additionally, NGS analysis showed trans-splicing in both RDEB keratinocytes and fibroblasts, with an efficiency of ~1.5% and ~0.6%, using the same exon 61/exon 62 junction forward primer in combination with an exon 68 specific reverse primer. The wild-type COL7A1 sequence was used as reference. Analysis of NGS results was performed with CRISPResso2 [25].

Figure 3

C7 expression in RDEB keratinocytes and fibroblasts. (A) Immunofluorescence staining of RDEB patient cell lines treated with the MC RTMS6m plasmid (right). RTM-treated patient keratinocytes, as well as fibroblasts, showed a restoration of C7 in individual cells (white arrows). Patient cell lines treated with the parental MC vector were used as negative controls (middle). Keratinocytes and fibroblasts from healthy donors served as a positive control (left). The cell’s nuclei were stained with 4′,6-Diamidino-2-phenylindol (DAPI, blue). Scale bar = 20 µm. (B) Mean fluorescence intensity was analysed with Fiji, and calculations were done using GraphPad Prism 9 (n = 3; mean ± SEM) [27,28]. (C) Expression of C7 after RTMS6m treatment of keratinocytes and fibroblasts was confirmed by Western blot analysis. Signal intensity was quantified using densitometric analysis with the Image Lab™ software from Biorad and plotted using GraphPad Prism 9.

Figure 4

Correction of RDEB skin equivalents. (A) Immunofluorescence staining performed on cryosections showed enhanced C7 expression (green) in RDEB fibroblast/RDEB keratinocyte SEs topically treated with the RTMS6m (right) as compared to untreated RDEB/RDEB SE (middle). SEs from wild-type fibroblasts and healthy keratinocytes were used as a positive control (left). Cell nuclei were stained with 4′,6-Diamidino-2-phenylindol (DAPI, blue). Scale bars: 20 µm. E= Epidermis; D = Dermis (B) Formation and organisation of the SE were confirmed by H&E staining. Scale bars: 100 µm. (C) The Mean Fluorescence Intensity (MFI) of the C7 expression was analysed with Fiji [27,28]. Final MFI calculations were done by subtracting the background MFI from the basal layer fluorescence. Treatment with RTMS6m led to increased C7 expression in four independent experiments (left). Calculation of the mean MFI in reference to the hKc SE C7 expression showed a significant (*) 4-fold increase in treated vs. untreated RDEB SE (right). Significance was determined via an unpaired Student’s t-test (n = 4; p = 0.0272; mean ± SEM).