Exon-Skipping Oligonucleotides Restore Functional Collagen VI by Correcting a Common COL6A1 Mutation in Ullrich CMD

Abstract

Collagen VI-related congenital muscular dystrophies (COL6-CMDs) are the second most common form of congenital muscular dystrophy. Currently, there is no effective treatment available. COL6-CMDs are caused by recessive or dominant mutations in one of the three genes encoding for the α chains of collagen type VI (COL6A1, COL6A2, and COL6A3). One of the most common mutations in COL6-CMD patients is a de novo deep intronic c.930+189C > T mutation in COL6A1 gene. This mutation creates a cryptic donor splice site and induces incorporation of a novel in-frame pseudo-exon in the mature transcripts. In this study, we systematically evaluated the splice switching approach using antisense oligonucleotides (ASOs) to correct this mutation. Fifteen ASOs were designed using the RNA-tiling approach to target the misspliced pseudo-exon and its flanking sequences. The efficiency of ASOs was evaluated at RNA, protein, and structural levels in skin fibroblasts established from four patients carrying the c.930+189C > T mutation. We identified two additional lead ASO candidates that efficiently induce pseudo-exon exclusion from the mature transcripts, thus allowing for the restoration of a functional collagen VI microfibrillar matrix. Our findings provide further evidence for ASO exon skipping as a therapeutic approach for COL6-CMD patients carrying this common intronic mutation.

Keywords: Collagen VI; Ullrich muscular dystrophy; antisense oligonucleotide; congenital muscular dystrophy; deep intronic mutation; exon skipping; extra cellular matrix.

Graphical abstract

Figure 1

Exon-Skipping Strategy Using Antisense Oligonucleotide (A) The deep intronic mutation c.930+189C > T in COL6A1 creates a cryptic splice donor site responsible for the insertion of an in-frame pseudo-exon (PE) between exons 11 and 12 in the mature transcripts, as confirmed by PCR and Sanger sequencing. The exon-skipping strategy designed to target the c.930+189C > T mutation to skip the PE is illustrated. (B) Schematic representation of PE and its flanking sequences. The top panel shows the potential splicing enhancer and splicing silencer motifs predicted by the Human Splicing Finder (http://www.umd.be/HSF/). The bottom panel illustrates the 15 ASOs designed to anneal to the PE and its flanking 5′ and 3′ splice sites. (C) M-fold of PE and its flanking 5′ and 3′ sites using RNAfold web server (http://rna.tbi.univie.ac.at//cgi-bin/RNAWebSuite/RNAfold.cgi). The PE is highlighted in yellow, and the region recognized by ASOs is indicated in blue.

Figure 2

Evaluation of ASOs Specificity and Efficiency in Pseudo-Exon Skipping at the RNA Level (A) Gel electrophoresis of PCR products amplified in RNA samples isolated from UCMD fibroblasts treated with ASOs at 20 nM for 24 h with Lipofectamine transfection. (B) Schematic representation of quantitative real-time PCR assay. Two sets of primers with specific probes were used in the analysis. One set of primers with a specific probe complementary to the pseudo-exon (PE) to exclusively amplify the mutant transcripts. Another set of primers to amplify the total COL6A1 transcripts, including both wild-type and mutant transcripts. (C) Quantitative real-time PCR was performed in RNA samples collected from UCMD fibroblasts treated with 20 nM ASOs after 24 h of transfection, using specific primers and probes. The two regions, where ASOs are capable of efficiently skipping the PE from the mutant transcripts, are underlined by a dashed line in red. Data were normalized to untreated samples and analyzed by one-way ANOVA and post-Bonferroni test. Data are presented as mean ± SD (∗p ≤ 0.05; ∗∗p ≤ 0.01; ∗∗∗p ≤ 0.001).

Figure 3

The Efficiency of ASOs on 2′-OMe Backbone in Restoring ECM Collagen VI Protein (A) Patient skin fibroblasts were treated with ASOs at 20 nM using Lipofectamine as transfection reagent. After 24 h, the transfection medium was replaced with growth medium containing l-ascorbic acid (50 μg/mL) for 48 h before being processed for collagen VI immunostaining. (B) Collagen VI protein in the ECM (in green) and nuclei (in blue) were displayed by immunofluorescence staining. Pictures were captured under fluorescence microscopy at 10× (upper panel) and 40× (lower panel) original magnification. Scale bars: 100 μm (upper panel) and 50 μm (lower panel), respectively. (C) Mean intensity and the area covered by collagen VI were quantified in fibroblasts treated with a single transfection of ASO-scr, ASO-3, ASO-4, ASO-5, and ASO-6 at 20 nM for 24 h followed by l-ascorbic acid (50 μg/mL) treatment for 48 h. Data represent mean ± SD from analysis of six individual field images acquired at 40× original magnification under fluorescence microscopy. Data were analyzed by one-way ANOVA and post-Bonferroni test (∗p ≤ 0.05; ∗∗p ≤ 0.01; ∗∗∗p ≤ 0.001).

Figure 4

The Efficiency of ASOs in 2′-MOE Backbone in Pseudo-Exon Skipping at RNA Level (A) Representative picture of PCR products from fibroblasts treated with ASO-5, ASO-6, ASO-5/6, or ASO-9 in 2′-MOE backbone for 24 h with Lipofectamine transfection. (B) Quantitative real-time PCRs of total COL6A1 and mutant COL6A1 transcripts were performed in RNA samples collected from four patient skin fibroblasts treated with ASO-5, ASO-6, ASO-5/6, or ASO-9 at concentrations ranging from 2.5 to 20 nM. Data are presented as mean ± SD. Data were analyzed by one-way ANOVA and post-Bonferroni test (∗p ≤ 0.05; ∗∗p ≤ 0.01; ∗∗∗∗p ≤ 0.0001).

Figure 5

The Efficiency of ASOs on 2′-MOE Backbone in Restoring ECM Collagen VI Protein (A) Representative images of immunofluorescence staining of collagen VI protein (in green) and nuclei (in blue) in patient skin fibroblasts treated with 2′-MOE ASOs. Pictures were captured under fluorescence microscopy at 10× (top panel) and 40× (lower panel) original magnification. Scale bars: 100 μm for 10× magnification and 50 μm for 40× original magnification. (B) The structure of ECM collagen VI expression in fibroblasts from healthy control and UCMD patients treated with 20 nM ASOs. Images were captured under confocal microscope. In healthy condition, collagen VI forms the liner microfibrils (white arrow) in ECM, while in UCMD patients the linear structure is replaced with discontinuous and speckled microfibrils (yellow asterisk). Treatment of ASO-5, ASO-6, ASO-5/6, or ASO-9 restored the collagen VI deposition pattern to linear microfibrils. Scale bar: 10 μm. (C) Mean intensity and the area covered by collagen VI were quantified in fibroblasts treated with a single transfection of ASO-scr, ASO-5, ASO-6, ASO-5/6, and ASO-9 at 20 nM. Data represent mean ± SD. Data were analyzed by one-way ANOVA followed by post-Bonferroni test (∗p ≤ 0.05; ∗∗p ≤ 0.01; ∗∗∗p ≤ 0.001). (D) The beaded microfibril network of collagen VI protein was visualized by scanning electron microscopy in UCMD fibroblasts treated with 20 nM ASO-5 or ASO-5/6. Scale bar: 100 nm.

Figure 6

The Efficiency of ASOs in Pseudo-Exon Skipping by Gymnotic Delivery (A) Representative images of PCR products from fibroblasts treated with ASO-5, ASO-6, ASO-5/6, or ASO-9 in 2′-MOE backbone for 4 days without the use of any transfection reagent. (B) Quantitative real-time PCRs of total COL6A1 and mutant COL6A1 transcripts were performed in RNA samples collected from four patient skin fibroblasts treated with ASO-5, ASO-6, or ASO-5/6 at concentrations ranging from 50 to 800 nM. Data are presented as mean ± SD. Data were analyzed by one-way ANOVA and post-Bonferroni test (∗p ≤ 0.05; ∗∗p ≤ 0.01; ∗∗∗p ≤ 0.001). (C) Representative images of ECM collagen VI protein expression in patient skin fibroblasts treated with ASO-5, ASO-6, ASO-5/6, and ASO-9 at 800 nM for 4 days with gymnotic delivery. Collagen VI protein (in green) and nuclei (in blue) are displayed by immunofluorescence staining. The linear collagen VI microfibrils are indicated by white arrows and the discontinuous collagen VI microfibrils by yellow asterisks. Scale bar: 50 μm. (D) Mean intensity and the area covered by collagen VI were quantified in fibroblasts after the gymnotic delivery of ASO-5, ASO-6, ASO-5/6, and ASO-9 at 800 nM. Data represent mean ± SD. Data were analyzed by one-way ANOVA and post-Bonferroni test (∗p ≤ 0.05; ∗∗p ≤ 0.01; ∗∗∗p ≤ 0.001).