Rescue of cardiomyopathy through U7snRNA-mediated exon skipping in Mybpc3-targeted knock-in mice

Exon skipping mediated by antisense oligoribonucleotides (AON) is a promising therapeutic approach for genetic disorders, but has not yet been evaluated for cardiac diseases. We investigated the feasibility and efficacy of viral-mediated AON transfer in a Mybpc3-targeted knock-in (KI) mouse model of hypertrophic cardiomyopathy (HCM). KI mice carry a homozygous G>A transition in exon 6, which results in three different aberrant mRNAs. We identified an alternative variant (Var-4) deleted of exons 5–6 in wild-type and KI mice. To enhance its expression and suppress aberrant mRNAs we designed AON-5 and AON-6 that mask splicing enhancer motifs in exons 5 and 6. AONs were inserted into modified U7 small nuclear RNA and packaged in adeno-associated virus (AAV-U7-AON-5+6). Transduction of cardiac myocytes or systemic administration of AAV-U7-AON-5+6 increased Var-4 mRNA/protein levels and reduced aberrant mRNAs. Injection of newborn KI mice abolished cardiac dysfunction and prevented left ventricular hypertrophy. Although the therapeutic effect was transient and therefore requires optimization to be maintained over an extended period, this proof-of-concept study paves the way towards a causal therapy of HCM.

• Figure S1. Evidence for an alternative Mybpc3 variant mRNA in wild-type mouse cardiac myocytes. • Figure S2. Evaluation of Var-4 Mybpc3 mRNA during mouse development. • Figure S3. Effect of 2OMePS-modified AONs on Mybpc3 mRNAs and cMyBP-C proteins in WT and KI cardiac myocytes. • Figure S4. Evaluation of the Mybpc3 mRNA species in wildtype cardiac myocytes.     Figure S2. Evaluation of Var-4 Mybpc3 mRNA during mouse development. A) RT-PCR from RNA extracted from cardiac myocytes isolated from neonatal C57BL/6J or Black swiss wild-type mice after the first (1.) and second (2.) round of PCR using primers located in exons 4 and 9 of Mybpc3 and in Gapdh. B) RT-PCR from RNA extracted from ventricular tissue of wild-type Black swiss mice at different postnatal development after the first (1.) and second (2.) round of PCR using primers located in exons 4 and 9 of Mybpc3 or in Gapdh. C) RT-PCR from RNA extracted from cardiac tissue isolated from wild-type C57BL/6J mice at different times of embryonic and postnatal development after the first (1.) and second (2.) round of PCR using primers located in exons 4 and 9 of Mybpc3 and in Gapdh. D) RT-PCR (1 round of PCR only) from RNA extracted from cardiac tissue of Mybpc3-targted knock-in (KI) mice at different postnatal windows using primers located in MYBPC3 exons 4 and 9.
The expected fragments are indicated by arrowheads. Abbreviation: M, 100-bp molecular weight marker.
These data showed that Var-4 is detected at a low level during the entire development in both C57BL/6J and Black swiss wild-type mice. These data support the view that it is a natural splicing variant of Mybpc3. Furthermore, Var-4 mRNA was also detected in young and old KI mice with no major different intensity.  Cardiac myocytes were isolated from wild-type neonatal mice and transduced for 72 h with adeno-associated virus serotype 6 (AAV6-U7snRNA-AON-5 or AAV6-U7snRNA-AON-5+6). At the end of the experiment, cells were treated (+) or not (-) with 300 µg/ml of the translation inhibitor emetine to prevent degradation of the nonsense mRNAs. The nonsense Mybpc3 mRNAs deleted of either exon 5 (Ex5) or exon 6 (Ex6) were stabilized after emetine treatment, suggesting that they are normally degraded by the nonsense-mediated mRNA decay. On the other hand, Var-4 mRNA deleted of both exons 5+6 was stable and detected in baseline condition 72 h after transduction with AAV6-U7-AON-5+6 in a wild-type sample. The expected amplicons are indicated by the arrowheads. Abbreviation: M, 100-bp molecular weight marker.
These data showed that the exon skipping strategy mediated by AAV-based gene transfer of AONs works in wild-type neonatal mouse cardiac myocytes, and that the skipping of the single exon 5 or the single exon 6 results in nonsense mRNAs, which are normally degraded by the nonsense-mediated mRNA decay (NMD). When the NMD is blocked by the translation inhibitor emetine, the nonsense mRNAs were detected.

Figure S5. Tissue distribution of GFP after systemic administration of AAV9 in a Mybpc3-targeted knock-in mouse.
A) A 4-wk-old KI mouse received adeno-associated virus (AAV) serotype 9 encoding GFP (7.6x10 10 vg) by systemic administration into the tail vein. RT-PCR analysis was performed on RNA extracted from different tissues (heart, lung, liver and kidney) 4 weeks after AAV injection. For PCR amplification primers located in the GFP sequence were used. The expected amplicon is indicated by the arrowhead. Abbreviation: M, 100-bp molecular weight marker. B) 1-day-old KI mouse received AAV9-CMV-GFP (2x10 11 vg) by systemic administration into the temporal vein for 7 days. 8-day-old control mouse (CTRL) did not receive anything. Hearts were extracted, rinsed in high K+, embedded in tissue Tek, and 10-µm-thick sections were performed. Sections were stained with primary antibodies directed against GFP (Santa Cruz, 1:200) and -actinin (SIGMA, 1:800) and with the secondary antibodies anti-rabbit Alexa Fluor-488 (1:600) and anti-mouse Alexa Fluor-546 (1:600). Nuclei were stained with DRAQ5 TM (1:800). Slides were embedded in Mowiol and immunofluorescence analysis was performed by confocal microscopy using a Zeiss Axiovert microscope with a 40x-oil objective. Confocal images were recorded with a Zeiss LSM 710 system. These data showed that i) GFP is more expressed in the heart than in the other organs, supporting the cardiotropism of the AAV serotype 9; ii) transduction efficiency of AAV9-CMV-GFP at a dose of 2x10 11 gave a complete transduction of the heart 7 days after systemic administration.   1-day-old KI mice received either AAV9-U7-AON-5+6 (2x10 11 vg) or PBS by systemic administration into the temporal vein for 7 days. Hearts were extracted, rinsed in high K+ and embedded in tissue Tek, and 10-µm-thick sections were performed. Sections were stained with primary antibodies directed against the MyBP-C motif of cardiac myosin-binding protein C (cMyBP-C, 1:200) and -actinin (1:800, Sigma) and with the secondary antibodies anti-rabbit Alexa Fluor-488 (1:600, Molecular Probes) and anti-mouse Alexa Fluor-546 (1:600, Molecular Probes). Nuclei were stained with DRAQ5TM (1:800, Molecular Probes). Control (CTRL) was stained only with the secondary antibodies. Slides were embedded in Mowiol and immunofluorescence analysis was performed by confocal microscopy using a Zeiss Axiovert microscope with a 100x-oil objective. Confocal images were recorded with a Zeiss LSM 710 system.
These data showed that cMyBP-C is organized in doublets in the A-band of the sarcomere and in alternation withactinin located at the Z-disk of the sarcomere. No major difference in the sarcomeric pattern was observed between the PBS-and AAV9-U7-AON-5+6 injected mice. This suggests that the stochiometry of the sarcomere is well preserved. These data showed that the expression of Var-4 prevents LVH and accumulation of Acta1 on the one hand, and on the other hand restores cardiac function and normal levels of Nppa and Nppb mRNA in the KI hearts. These data suggest that Var-4 is not toxic in vivo and supports the exon skipping approach using AON-5+6 in vivo to rescue the phenotype of KI mice.  Neonatal KI mice received 2x10 11 vg of adeno-associated virus (AAV9-U7-AON5+6) via the temporal vein, which corresponds to a mean dose of 1.44x10 14 vg/kg of body weight. Ventricular tissues were removed 55 days or 7 days after AAV9 administration, and genomic DNA was extracted using Trizol or the Extract-N-AmpTM Tissue PCR kit (SIGMA) according to manufacturer's instructions. Multiplex PCR was performed on 20 ng DNA using primer pairs for AAV9 and for Gapdh (for AAV9: 5'-AGT GGC CAA CTC CAT CAC TA-3' and 5'-CAC AGA TGC TCA GAC CAC TTT TGC GGA AGT-3', and for Gapdh: 5'-ATT CAA CGG CAC AGT CAA G-3' and 5'-TGG CTC CAC CCT TCA AGT-3'). On the left panel is shown the results of the PCR on 1% agarose gel and on the right panel the quantification of the AAV9 DNA particles normalized to Gapdh DNA. **P<0.01 vs 7 days, Student's t-test.
These data showed that AAV9 particles were detected in all mice that received AAV9 and were 4-fold lower 55 days than 7 days after AAV9 administration. No AAV9 particles were detected in PBS-treated mice.