5'RNA Trans-Splicing Repair of COL7A1 Mutant Transcripts in Epidermolysis Bullosa

Abstract

Mutations within the COL7A1 gene underlie the inherited recessive subtype of the blistering skin disease dystrophic epidermolysis bullosa (RDEB). Although gene replacement approaches for genodermatoses are clinically advanced, their implementation for RDEB is challenging and requires endogenous regulation of transgene expression. Thus, we are using spliceosome-mediated RNA trans-splicing (SMaRT) to repair mutations in COL7A1 at the mRNA level. Here, we demonstrate the capability of a COL7A1-specific RNA trans-splicing molecule (RTM), initially selected using a fluorescence-based screening procedure, to accurately replace COL7A1 exons 1 to 64 in an endogenous setting. Retroviral RTM transduction into patient-derived, immortalized keratinocytes resulted in an increase in wild-type transcript and protein levels, respectively. Furthermore, we revealed accurate deposition of recovered type VII collagen protein within the basement membrane zone of expanded skin equivalents using immunofluorescence staining. In summary, we showed for the first time the potential of endogenous 5' trans-splicing to correct pathogenic mutations within the COL7A1 gene. Therefore, we consider 5' RNA trans-splicing a suitable tool to beneficially modulate the RDEB-phenotype, thus targeting an urgent need of this patient population.

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

Figure 1

Fluorescence-reporter based screening for most functional RTMs. (A) Trans-splicing was evaluated either in the presence of a mini-gene target for screening of highly functional BDs, or in keratinocytes, where successful trans-splicing to the endogenous COL7A1 target pre-mRNA was investigated. In the presence of a functional BD, specific trans-splicing results in a single mRNA transcript encoding either DsRed (transfection control) and full-length AcGFP in the mini-gene reporter setting, or a 5′AcGFP/COL7A1 exons 65–118 hybrid when splicing to the endogenous transcript. (B,C) HEK293FT cells co-transfected with the mini-gene target and individual RTMrs (i,ii) and positive control (iii). Hybrid DsRed and AcGFP expression is shown by fluorescence microscopy and flow cytometry. AcGFP/DsRed ratios were calculated from the amount of GFP-positive cells (sector B2 + B4)/total amount of transfected cells (sector B1 + B2 + B4), and is given in percent (%). RTMrA showed an AcGFP/DsRed ratio of 97.65%; RTM B, 97.03%. 100% of cells transfected with the positive control expressed both DsRed and acGFP. (D,E) In keratinocytes, correct trans-splicing to COL7A1 was confirmed by RT-PCR using a GFP forward and a COL7A1 exon 67 reverse primer. The 335 bp amplification product was verified to be the GFP-COL7A1 fusion by Sanger sequencing.

Figure 2

Endogenous trans-splicing: Gene and protein expression in RDEB-KC. (A) Schematic of the endogenous trans-splicing process. The complete transcript region from exon 1 to 64, which harbors the mutation of interest, is replaced by its wild-type copy provided by RTMe. (B) SqRT-PCR performed on an RDEB patient cell line retrovirally transduced with RTMeA and RTMeB. Means ± SD of fold changes of COL7A1 expression over wild-type keratinocytes (KC) of three independent experiments are given. Unpaired, two-sided Student’s t-test was performed and p-values ≤ 0.05 were considered significant (*). (C) Western blot analysis showed re-expression and secretion of C7.

Figure 3

C7 immunostaining of RDEB-KC and skin equivalents. (A) C7-immunostaining (green) on RTM-transduced RDEB patient cells showed re-expression of C7. Cell nuclei were counterstained using DAPI (blue). (B) Correct localization of C7 at the BMZ was confirmed in cryosections of skin equivalents generated from RTMeA-transduced and mock-transduced (empty plasmid) patient KCs and in human wild-type (wt) skin. E = epidermis; D = dermis; arrows mark the dermal-epidermal junction. Scale bars: 50 μm.