Splicing correction by peptide-conjugated morpholinos as a novel treatment for late-onset Pompe disease

Abstract

Late-onset Pompe disease (LOPD) is overwhelmingly caused by a single mutation that disrupts splicing of acid-alpha glucosidase (GAA) and results in the accumulation of lysosomal glycogen in muscle cells leading to progressive muscle weakness in patients. Current therapeutics for LOPD do not meet the needs of patients and have largely been developed in mutant animal models lacking Gaa expression, which more closely mimic the less common infantile form of the disease. Here we design and evaluate peptide-conjugated phosphorodiamidate morpholino oligomers (PPMOs) to target the causative mutation in GAA and correct pathogenic splicing in muscle tissue. We show PPMO compounds correct LOPD splicing in both patient induced pluripotent stem cell-derived muscle cells and in skeletal muscle tissue after intravenous dosing in a newly developed humanized LOPD animal model that recapitulates patient LOPD splicing.

Keywords: MT: Oligonucleotides: Therapies and Applications; RNA medicines; late-onset Pompe disease; lysosomal storage disease; oligonucleotide therapeutics; splice switching antisense oligonucleotide.

Graphical abstract

Figure 1

PPMO increases GAA enzyme activity in patient-derived fibroblasts (A) GAA mis-splicing caused by the common IVS1 mutation (red) compared with canonical GAA splicing (black). Splicing events were detected by endpoint RT-PCR. (B) Schematic depicting the location of PPMO compounds in microwalk screen. (C) Members of the microwalk screen were assayed for increases in GAA enzyme activity in LOPD patient fibroblasts. NTC = nontargeting control. One-way ANOVA with Dunnett’s multiple comparison test. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗<0.0001. (D) Optimization of hit compound sequence RC-3001 using an abasic subunit to address high GC content. (E) Abasic-containing sequence (RC-3003) restores GAA enzyme activity in patient fibroblasts. Results are expressed as mean with error bars (SD).

Figure 2

LOPD patient iPSC-derived myotubes recapitulate splicing deficits observed in LOPD patients (A) LOPD patient fibroblasts were reprogrammed into iPSC lines and pluripotency was confirmed by immunostaining with OCT3/4. LOPD iPSCs were terminally differentiated into LOPD patient derived iPSC myoblasts and myotubes that express MYOD and MHC by immunostaining, respectively. PC, phase contrast. Scale bars: fibroblasts (500 μm), iPSC and myoblasts (100 μm), myotubes (200μm) (B) LOPD patient iPSC-derived myotubes faithfully recapitulate aberrant splicing reported in LOPD patients, including skipping of exon 2. C) GAA TV expression in LOPD patient iPSC-derived myotubes. TV1, NM_000152.5; TV2, NM_001079803.3; TV3, NM_001079804.3. Results are expressed as mean with error bars (SD).

Figure 3

PPMO correct GAA splicing in LOPD patient iPSC-derived myotubes (A) PPMO treatment decreases pathogenic splicing and increases full-length GAA as measured by endpoint RT-PCR. Marker in basepairs. Full-length GAA, 1100 bp; exon 2 skipping, 450 bp, 550 bp. (B) PPMO compounds increase levels of GAA TV1 and TV2 upon treatment. (C) PPMO treatment precisely restores the LOPD-critical GAA exon 1–2 junction with no detectable undesired splicing variants as measured by RNA amplicon sequencing. Results are expressed as mean with error bars (SD).

Figure 4

PPMO correct pathogenic GAA levels at transcript, protein, and lysosomal enzyme activity levels in LOPD patient iPSC-derived myotubes (A) PPMO treatment increases GAA TV1 and TV2 at the LOPD critical exon 1-2 junction and also increases. (B) Total GAA (all TVs) transcript levels as measured by qPCR. (C and D) PPMO treatment similarly increases GAA protein (one-way ANOVA with Dunnett’s multiple comparison test; ∗p < 0.05; ∗∗p < 0.01). Arrows indicate size of GAA primary translation (76 kDa) and post-translationally modified GAA (76 kDa). (E) Lysosomal GAA enzyme activity levels in patient iPSC-derived myotubes were increased after treatment with PPMO in a dose-dependent fashion. Results are expressed as mean with error bars (SD).

Figure 5

Generation and characterization of a relevant murine model of LOPD (A) Human GAA harboring the IVS1 mutation was inserted into the Rosa26 safe harbor locus. (B) Single copy integration was confirmed by genomic qPCR. (C) GAA TV expression in skeletal muscle was confirmed by qPCR. (D) GAA expression was confirmed in various tissues by qPCR analysis. (E) Endpoint RT-PCR analysis demonstrated that the mutant human GAA transgene was faithfully mis-spliced in the mouse and produced the same GAA splice variants found in LOPD patients. (F) GAA mis-splicing in mouse quadricep muscle was confirmed by RNA amplicon sequencing. Results are expressed as mean with error bars (SD).

Figure 6

Systemic PPMO treatment corrects GAA splicing in vivo (A) Endpoint RT-PCR qualitatively showed an increase in full-length GAA after single IV dose of RC-3002. (B) Amplicon sequencing confirmed an increase in correctly spliced exon 1–2 junction upon treatment with RC-3002. (C) qPCR analysis of RNA extracted from quadricep muscle of animals treated with a single dose of RC-3003 demonstrated a dose-dependent increase in GAA transcript levels that was found to correlate with (D) compound tissue concentrations. (E) treatment was well tolerated. One-way ANOVA with Dunnett’s multiple comparison test. ∗∗p < 0.01. Results are expressed as mean with error bars (SD).