Using muscle homing peptide CyPep10 to deliver phosphorodiamidate morpholino oligomers in the mdx mouse

Abstract

The severe muscle wasting disorder Duchenne muscular dystrophy (DMD) is characterized by the absence of dystrophin, a protein that is essential for muscle stability. Restoring this protein has therapeutic potential. Antisense oligonucleotides (ASOs), designed to target and skip exons, can restore the reading frame that is disrupted in these patients, enabling the production of partially functional dystrophin. Achieving optimal dystrophin restoration remains challenging due to limited delivery and cellular uptake. Muscle homing peptides conjugated to ASOs are a way to achieve this. Previously, CyPep10 (CP10) has been used to significantly increase exon skipping efficiency for the 2′-O-methyl phosphorothioate chemistry in the mdx mouse model for DMD. Here, we explore the effect of using peptide CP10 as a conjugate to phosphorodiamidate morpholino oligomers (PMOs) ASOs to improve muscle delivery, thereby hoping to achieve increased treatment efficiency. Overall, we confirmed the homing ability of CP10 and observed significantly increased muscle tissue concentration levels of PMO when CP10 was conjugated. This did not lead to increased levels of exon skipping or dystrophin restoration. Conjugating both a cell-penetrating peptide (CPP) and CP10 to a PMO showed that increased exon skipping efficiency can be achieved to a slightly greater extent than with CPP-PMO treatment.

Graphical abstract

This study by Schneider and colleagues found that conjugation of muscle homing peptide CyPep10 to a phosphorodiamidate morpholino oligomer (PMO) increased muscle tissue concentration of PMOs and that double conjugation with a cell-penetrating peptide with CyPep10 improved exon skipping efficiency as well as muscle pathology hallmarks.

Introduction

Duchenne muscular dystrophy (DMD) is one of the most common and severe forms of muscular dystrophy.¹ This disease is caused by mutations in the DMD gene that lead to disruptions of the open reading frame, resulting in the loss of dystrophin, a protein that is normally translated from this gene.² Dystrophin connects the actin cytoskeleton within muscle fibers to the extracellular matrix surrounding these fibers and is needed for muscle stability and integrity.³ Lack of this protein leads to muscle damage upon contraction, which in turn results in chronic inflammation and deposition of fibrosis and constant cycles of de- and regeneration of muscle fibers, until regeneration fails and muscle is replaced by adipose tissue.⁴ Thus, people without functional dystrophin experience progressive muscle loss, which will eventually lead to premature death from respiratory or cardiac failure, around the age of 30.⁵ Current standards of care include the use of corticosteroids to slow down disease progression by decreasing inflammation and by providing symptomatic support, but there remains an unmet medical need, and patients would benefit from therapy that addresses the cause of the disease, i.e., the lack of dystrophin.^6,7

Exon skipping therapy makes use of specifically designed antisense oligonucleotides (ASOs) to recognize and hide target exons from pre-mRNA transcripts during splicing. Restoration of the reading frame enables the synthesis of shorter, yet partially functional, dystrophin proteins, similar to those observed in Becker muscular dystrophy patients.^8,9 Effective design of exon skipping ASOs depends on the location and size of the mutation. In DMD, most mutations are clustered around a hotspot between exons 42 and 55.¹⁰ Targeted skipping of specific exons in this region could be beneficial for a substantial portion of the patient group. For DMD, several mutation-specific ASOs have been evaluated in clinical trials and have been approved: eteplirsen (exon 51), casimersen (exon 45), and golodirsen and viltolarsen (both exon 53).

The approved ASOs for DMD are all phosphorodiamidate morpholino oligomers (PMOs).¹¹ Treatment is administered via weekly intravenous infusion and uptake by muscles is evidenced by dystrophin restoration measured in biopsies and a milder disease phenotype in treated patients. However, there remains a need for further improvement and increased levels of dystrophin restoration are expected to lead to greater clinical benefit.^12,13 Moreover, it is thought that the greatest improvement to efficiency would be via increased delivery to skeletal muscles.

Improving the uptake of 2′-O-methyl phosphorothioate (2OMePS) ASOs by muscle in an animal model system for DMD, the mdx mouse, has previously been achieved using a muscle homing peptide. This peptide, CyPep10 (CP10), along with other peptide candidates, was identified to be muscle homing through in vivo phage display biopanning.¹⁴ For a detailed overview of the use phage display biopanning to identify these muscle homing peptides, we refer the reader to this 2018 paper by Jirka and colleagues.¹⁴ The mdx mouse model carries a point mutation in exon 23 of the Dmd gene leading to a premature stop codon and a dystrophic pathology.¹⁵ Among all the peptides identified in this paper, CP10 was the only conjugate that achieved significant improvement in exon skipping in skeletal muscle tissues and heart compared with an unconjugated ASO of the same chemistry.¹⁴ The CP10 peptide conjugate is a 7-mer cyclized peptide with the following sequence: QLFPLFR. This peptide has only one positively charged amino acid, in contrast with the more commonly used class of arginine-rich cell-penetrating peptides (CPPs). These short cationic peptides are known to increase translocation across the cell membrane in a tissue-aspecific manner.¹⁶ CPPs have also been shown to achieve significantly improved uptake of ASOs in the muscles and heart of the mdx mouse model.¹⁷ While initially promising, it was later found that using CPPs led to renal toxicity in non-human primates and patients at the doses needed to achieve a therapeutic effect.^18,19,20 In the MOMENTUM trial investigating a CPP-PMO (SRP-5051), adverse events like hypomagnesemia and a decline in kidney function were observed in a subset of patients, and it was decided to stop the clinical development of this CPP-PMO. Thus, alternative peptides to improve delivery to muscle and heart are required.

In this study, we tested whether CP10 could improve the delivery of a PMO targeting exon 23 in mdx mouse muscle tissues and consequently increased skipping and dystrophin restoration. We compared this CP10-PMO conjugate with naked PMO (also known as AVI-4225¹⁷) and a CPP-PMO (also referred to as RC-1001¹⁷), where the CPP was an acetyl-R6Gly peptide (peptide oligonucleotide conjugates, US patent 9161948). Combined conjugation of CPP and CP10 to the same PMO was also evaluated to determine if there was a synergistic effect of combining these conjugates. Our work did not demonstrate improved therapeutic effects with CP10 conjugation to PMO ASOs, as was found for 2OMePS ASOs. However, we did observe improved tissue concentration with PMOs that were conjugated to CP10.

Results

CP10 is efficiently taken up by cultured muscle and heart cells in vitro

The uptake of CP10 in muscle cells was evaluated in vitro using human myotubes (control and patient derived) and cardiomyocytes with fluorescein isothiocyanate (FITC)-labeled CP10 (Figure 1). By evaluating FITC fluorescent signal after incubation of this peptide in the media of the cells, we confirmed myotube uptake similar to that of the previous work that describes the identification of this muscle homing peptide.¹⁴

Figure 1.

Confirmation of muscle uptake of CP10 in vitro

Human control and patient myotubes (left and middle) and cardiomyocytes incubated with FITC labeled CP10 (green). Nuclei in blue (DAPI). Cells were imaged at an original magnification of 20×.

CP10-PMO has a slightly greater efficiency than naked PMO but not CPP-PMO in vitro

To assess in vitro efficacy, C2C12 myotubes were treated with various doses of naked PMO, CPP-PMO, and CP10-PMO (Figures S30A and S30B) for 24 h to determine exon skipping efficiency by RT-PCR. At all doses we observed the lowest levels of exon skipping with naked PMO ranging from 0% exon skipping at 0.2 μM to ∼48% exon skipping at 30 μM (Figure 2). CP10-PMO treatment led to higher levels of exon skipping compared with naked PMO treatment for each of the tested doses, except for the lowest (Figure 2). CPP-PMO, however, induced the highest levels of exon skipping, with 95.5% exon skipping at 30 μM compared with 72% exon skipping with CP10-PMO treatment (Figure 2).

Figure 2.

In vitro testing of peptide PMO conjugates

C2C12 myotubes were incubated with different doses of naked PMO, CPP-PMO, and CP10-PMO for 24 h, whereafter exon skipping was evaluated. Error bars display standard deviation of the mean.

CP10-PMO conjugate does not lead to improved exon skipping and dystrophin restoration in mdx mice

Therefore, we proceeded to use the CP10 peptide for conjugation to a PMO targeting mouse exon 23 (CP10-PMO) and assess safety and efficiency in the mdx mouse model. As references, we used unconjugated PMO (naked PMO) and the CPP-PMO conjugate. Four weekly doses of 30 mg/kg PMO or the molar equivalent were injected intravenously for 4 weeks. One week after the last treatment, tissues were harvested for analysis. RNA was isolated from the gastrocnemius, triceps, and heart muscles and the diaphragm. RT-PCR analysis revealed that, at the RNA level, we did not observe a beneficial effect for CP10-PMO compared with naked PMO in skeletal muscles (gastrocnemius and triceps) or in diaphragm and heart (Figure 3A). In the skeletal muscles (gastrocnemius and triceps), naked PMO treatment resulted in 30% and 23% exon skipping whereas CP10-PMO led to 26% and 29% exon skipping, respectively. In the diaphragm, a lower level of skipping was observed across all treatment groups, with naked PMO and CP10-PMO both resulting in 7% exon skipping. Negligible exon skipping was observed in the heart for naked PMO (0.2%) and CP10-PMO (0.1%). In contrast, the CPP-PMO, as expected, significantly increased exon skipping in all tissues with 51%, 49%, 25%, and 12% exon skipping in the gastrocnemius, triceps, diaphragm, and heart, respectively (Figure 3A).

Figure 3.

Exon skipping and dystrophin restoration (in %) of single PMO conjugates and naked PMO treatment in vivo

(A) Exon skipping levels in triceps, gastrocnemius, diaphragm, and heart. (B) Dystrophin restoration levels for the same tissues. Error bars display standard deviation of the mean.

∗p < 0.05, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001.

After exon skipping analysis, we investigated whether our treatments had an effect on total dystrophin restoration in selected tissues using the Jess automated western blot system (Figure 3B). In line with our exon skipping results, no significant change in dystrophin restoration levels was observed for the CP10-PMO compound, compared with naked PMO. In triceps, 6% dystrophin restoration was observed for CP10-PMO, which was at the same level as naked PMO. For gastrocnemius, 4% dystrophin restoration was measured for CP10-PMO, compared with 3% for naked PMO. For the diaphragm, 0.9% dystrophin restoration was found for CP10-PMO and 0.3% for mice treated with naked PMO. In the heart, treatment with CP10-PMO resulted in lower dystrophin restoration compared with naked PMO, as 0.1% dystrophin restoration was observed, compared with 0.2% in the naked-PMO treated group. As expected, the significantly improved exon skipping levels of CPP-PMO led to significantly greater dystrophin restoration in the gastrocnemius (17%), triceps (17%), and diaphragm (2%). This was not the case for heart tissue, where only 0.1% of dystrophin restoration was observed. Since levels of dystrophin restoration below 2.5% were based on predicted values, it was decided to not perform any statistical analysis on this outcome measure for the diaphragm and the heart.

Multiple conjugate PMOs show improved exon skipping levels compared with CP10 and PMO in linear but not in branched form in vivo

We next tested whether conjugating both CP10 and a CPP to the PMO could improve efficiency. CP10 and CPP were conjugated to the PMO in either a linear (L-) or branched (B-) fashion (Figures S30C and S30D). C2C12 mouse myotubes were used to evaluate exon skipping levels of these compounds compared with the single CPP-PMO conjugate at different concentrations. In vitro, we found exon skipping levels of up to 90% in C2C12 cells when treated with the highest dose (30 μM) of CPP-PMO. These levels were also reached after cells were treated with an intermediate dose (15 μM) of the double-conjugated PMOs. We further saw that treatment with a low dose of 5 μM CPP-PMO led to ∼20% of exon skipping, whereas levels of 40% or even up to 80% were observed for L-CP10-CPP-PMO or B-CP10-CPP-PMO, respectively, at the same concentration (Figure 4).

Figure 4.

Exon 23 skipping of L-CP10-CPP-PMO vs. B-CP10-CPP-PMO conjugates compared with CPP-PMO treatment in C2C12 cells

C2C12 myotubes were treated with different concentrations of each compound and exon skipping levels were calculated. Double-conjugated PMOs resulted in exon skipping levels similar to those compared with CPP-PMO at lower dosing. Error bars display standard deviation of the mean.

For the in vivo assessment of the safety and efficiency of the CP10-CPP-PMO compounds compared with naked PMO and CPP-PMO, we used a regimen of 4 weekly intravenous doses of 30 mg/kg molar equivalent. The bodyweight of all mice within each treatment group were recorded weekly. We observed a progressive increase in weight over time, with no differences observed between the groups (Figure 5). After sacrifice, serum was collected to assess the safety profile of the tested compounds with clinical chemistry biomarkers for liver and kidney function and tissue and muscle breakdown. We found that, after CP10-PMO treatment, there was no significant difference in any of the assessed markers when compared with the naked PMO group (Figure 6). CPP-PMO treatment showed a similar result when looking at general liver and kidney function (alkaline phosphatase, total bilirubin, albumin, total protein, globulin, and blood urea nitrogen); the levels did not significantly differ from naked PMO treatment. CPP-PMO treatment did, however, significantly decrease levels of aspartate transaminase (AST), alanine transaminase (ALT), and creatine kinase (CK) and lactate dehydrogenase (LDH) compared with naked PMO treatment. It is known that AST and ALT are elevated in patients with DMD and the mdx mouse due to muscle damage, making these outcome measures additional indicators of tissue damage along with CK and LDH.²¹ For the double-conjugated PMOs, we again observed no significant differences in the levels for general kidney and liver function compared with naked PMO treatment; however the biomarkers for muscle damage (AST, ALT, CK, and LDH) were lower as observed with CPP-PMO treatment, albeit not always significantly different from naked ASO treatment (Figure 6).

Figure 5.

Bodyweight (grams) of individual mice in all treatment groups per week (n = 4–5)

Body weights were measured each week and increased with time in a similar fashion over time for each of the groups. No statistical differences were found between the treatment groups.

Figure 6.

Overview of measured serum markers per treatment group

General markers measured from serum for liver and kidney function and biomarkers for muscle damage are shown. Error bars display standard deviation of the mean. ∗p < 0.05; ∗∗p < 0.005; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001.

We observed that the conjugation of both CP10 and CPP to the PMO led to higher exon skipping levels as compared with naked PMO (Figure 7A) when evaluated in the mdx mouse model. B-CP10-CPP-PMO treatment led to 37%, 25%, and 15% exon skipping in gastrocnemius, diaphragm, and heart, respectively. L-CP10-CPP-PMO treatment led to even higher exon skip levels (55%, 50%. and 20% for gastrocnemius, diaphragm. and heart, respectively). Comparing the levels of these double-conjugated compounds with the naked PMO treatment, which led to 20%, 6%, and 0% exon skipping in gastrocnemius, diaphragm, and heart, respectively, we found that the increased levels of skipping were significantly higher in all measured tissues for the L-CP10-CPP-PMO-treated mice and only significantly higher for heart tissue for the B-CP10-CPP-PMO treatment group. When comparing these double-conjugated PMOs with their single-conjugated counterparts (CPP-PMO or CP10-PMO), L-CP10-CPP-PMO induced exon skipping as high (gastrocnemius) or higher (heart and diaphragm) than CPP-PMO and had higher exon skipping levels in all tissues when compared with CP10-PMO. B-CP10-CPP-PMO also resulted in higher exon skipping levels when compared with CP10-PMO; however, lower exon skipping levels were observed in diaphragm and gastrocnemius tissue when compared with CPP-PMO. This compound induced higher exon skipping levels than CPP-PMO in heart, albeit lower than L-CP10-CPP-PMO. However, exon skipping levels between either of the double-conjugated PMOs (B- or L-CP10-CPP-PMO) were never significantly different from their single-conjugated counterparts (CPP-PMO and CP10-PMO) in any of the measured tissues.

Figure 7.

Exon skipping (in %), dystrophin restoration, and pathological hallmark assessment after treatment with multiple PMO conjugates in vivo

(A) Exon skipping levels in gastrocnemius, diaphragm, and heart are shown. (B) Dystrophin restoration for the same tissues is depicted. (C1) Representative images of gastrocnemius tissue of immunofluorescence analysis of dystrophin- and MYH3-positive fibers for each of the treatment groups. (C2) Representative images of H&E analysis in gastrocnemius tissues of each of the treated groups. (D) Mean number of dystrophin positive fibers per square millimeter in gastrocnemius tissues in each treatment group. (E) H&E analysis for gastrocnemius tissues to reveal pathology levels in the different treatment groups. (F) Mean number of regenerating fibers (MYH3 positive) in gastrocnemius tissues in the different treatment groups. Error bars display standard deviation of the mean. ∗p < 0.05; ∗∗p < 0.005; ∗∗∗p < 0.001.

Our Jess analysis showed that L-CP10-CPP-PMO induced significantly higher levels of dystrophin restoration in all analyzed tissues when compared with naked PMO (Figure 7B). In gastrocnemius, L-CP10-CPP-PMO vs. naked PMO resulted in 26% vs. 3% restoration, respectively. For diaphragm, 10% vs. 0.7% dystrophin restoration was observed for L-CP10-CPP-PMO and naked PMO, respectively. In heart, L-CP10-CPP-PMO led to 3% dystrophin restoration, whereas naked PMO induced 0.2% restoration. In this second round of in vivo tests, we observed that CPP-PMO treatment resulted in 22%, 6%, and 1% dystrophin restoration and that CP10-PMO induced 5%, 1%, and 0.2% dystrophin restoration in gastrocnemius, diaphragm, and heart, respectively. The levels of dystrophin restoration for L-CP10-CPP-PMO were thus higher than their single-conjugated counterparts (CPP-PMO and CP10-PMO); however, these differences were not significant for any of the tissues. B-CP10-CPP-PMO treatment also showed higher dystrophin restoration levels in all analyzed tissues when compared with the naked-PMO treatment group; however, the difference in restoration was only significantly different for heart tissue and not in the gastrocnemius or diaphragm. For this compound 17%, 7%, and 3% dystrophin restoration was observed in the gastrocnemius, diaphragm, and heart, respectively. For the heart and the diaphragm, the levels of dystrophin restoration induced by B-CP10-CPP-PMO were higher than their single-conjugated counterparts, whereas in the gastrocnemius muscle, the single CPP-PMO induced greater dystrophin restoration than B-CP10-CPP-PMO. However, the differences between the levels of dystrophin restoration induced by B-CP10-CPP-PMO and their single-conjugated counter parts were never significant in any of the analyzed tissues.

We then proceeded to evaluate whether the levels of dystrophin restoration resulted in any changes in the number and or distribution of fibers expressing dystrophin and whether treatment had any effect on measurable pathology levels (inflammation and fibrosis and regenerating fibers). The three gastrocnemius muscles with the highest level of exon skipping per treatment group were evaluated.

For the compounds that resulted in the significantly highest levels of exon skipping (CPP-PMO and L-CP10-CPP-PMO), we also observed significantly more dystrophin positive fibers (Figures 7C1, 7D, and S31). Naked PMO-treated mice displayed a mean of 54 fibers per square millimeter of gastrocnemius tissue, while L-CP10-CPP-PMO resulted in 291 dystrophin positive fibers per square millimeter of tissue. L-CP10-CPP-PMO treatment resulted in a higher number of dystrophin-positive fibers when compared with their single conjugation counterparts CPP-PMO and CP10-PMO; however, this result was not significantly different. For CPP-PMO, 209 dystrophin positive fibers per square millimeter were observed, and CP10-PMO treatment led to 67 fibers/mm² that were positive for dystrophin. Like its linear counterpart, B-CP10-CPP-PMO treatment also resulted in significantly more dystrophin positive fibers when compared with naked PMO treatment, where 216 dystrophin positive fibers per square millimeter were observed. The most significant difference for the level of dystrophin-positive fibers was observed between naked PMO and L-CP10-CPP-PMO (Figure 7D). We also observed that, in tissues with a lower number of positive fibers (naked PMO and CP10-PMO), clusters of dystrophin restoration throughout the tissue were present, whereas a high number of dystrophin-expressing fibers led to a uniform distribution of positive fibers across the tissue (CPP-PMO, B-CP10-CPP-PMO, and L-CP10-CPP-PMO) (Figure 7C1).

When assessing muscle histology with hematoxylin and eosin (H&E) analysis, the greatest amount of fibrosis and inflammation was observed for the gastrocnemius tissues of mice treated with naked PMO (10.8%; Figures 7C2, 7E, and S32). Lower levels of pathological features were present for all other treatment groups; however, a significant decrease in pathology was only present in the gastrocnemius tissues of mice treated with L-CP10-CPP-PMO, where 1.7% of pathology was measured (Figure 7E).

To examine the level of regenerating fibers, we used immunofluorescence and assessed MYH3-positive fibers (Figures 7C1, 7F, and S31). We observed the lowest level of regeneration in the L-CP10-CPP-PMO-treated group (0.05 fibers/mm²); however, we found that none of the treatments led to any significant differences between any of the treatment groups (Figure 7F).

CP10 conjugate leads to increased PMO concentrations in muscle

To determine if improved muscle delivery was achieved we assessed the concentration of PMO using an enzyme-linked oligonucleotide hybridization assay (ELOHA) for lysates of quadriceps and gastrocnemius skeletal muscle tissue and heart tissue for our different treatment groups. We found that significantly increased concentrations of PMO could be detected for skeletal muscles treated with L-CP10-CPP-PMO, as well as B-CP10-CPP-PMO and CP10-PMO, compared with naked PMO (Figures 8A and 8B). In the heart, L-CP10-CPP-PMO and B-CP10-CPP-PMO treatment resulted in significantly higher levels of PMO concentrations compared with naked PMO (Figure 8C).

Figure 8.

Concentration of PMO in quadriceps, gastrocnemius, and heart tissue

The concentration of PMO in quadriceps, gastrocnemius, and heart tissue lysates for each treatment group was determined using an ELOHA. ∗p < 0.05; ∗∗p < 0.005; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001.

Discussion

In our studies we explored whether the muscle homing peptide CP10, which previously improved uptake of 2OMePS ASOs, could increase muscle delivery for PMOs. We found that, even though increased muscle homing was observed for CP10 conjugated PMO, no effective improvement on exon skipping was observed.

The results of exon skipping, dystrophin restoration, and pathology measures in our experiments correlate with each other. When higher levels of exon skipping and dystrophin restoration were observed, lower levels of muscle breakdown biomarkers (LDH and CK), inflammation, fibrosis, and regenerating fibers were present, confirming that the cycle of de- and regeneration, one of the hallmarks of DMD pathology, could be slowed down with in vivo treatment. From these correlations it was clear that CP10 conjugation to the PMO did not lead to the desired therapeutic effects of improved exon skipping, dystrophin restoration, or histological features, even though increased PMO concentrations were measured in muscle tissue compared with naked PMO. PMO concentration measurements were performed on whole muscle tissue lysates and do not indicate where in the tissue or cells the PMO is located. From our results, it is apparent that CP10 has muscle-homing capacity; however, CP10-PMO likely does not reach the nucleus to a great enough extent to lead to increased levels of exon skipping and subsequent dystrophin restoration. An evident sign for the presence of differences in nuclear uptake can be observed when looking at CPP-PMO and naked PMO treatments. CPP-PMO treatment induced significantly more exon skipping in skeletal muscles compared with naked PMO; however, the muscle tissue concentration levels are the same or only slightly higher. These results show that a greater proportion of the total PMO available reaches the nucleus when conjugated to the CPP, compared with when naked PMO is used. Evaluating distinct cellular compartments could help us to identify the current bottleneck for PMO conjugated with CP10 and accelerate efficacy improvements.

We speculate that the discrepancy between the effect of conjugation of CP10 to a 2OMePS vs. PMO in part lies in the different pharmacokinetic properties of these ASO chemistries.²² In comparison with PMOs, the phosphorothioate backbone of 2OMePS prolongs time in circulation by serum protein binding, which prevents clearance. Binding to mostly albumin, 2OMePS are cleared less rapidly by glomerular filtration compared with PMOs, which are rapidly cleared.²³ Our findings suggest that the conjugation of CP10 to PMO likely does not prevent renal clearance to a high enough extent and, therefore, does not allow the CP10-PMO to be in circulation long enough to benefit from the muscle homing capacity of the CP10 peptide. We thus proceeded to test whether combining CPP and CP10 conjugation onto a single PMO would improve the efficiency of PMO treatment. From our results, it was clear that a slight synergistic effect could be observed when CPP and CP10 were conjugated in a linear manner onto the PMO in comparison with either CPP-PMO, CP10-PMO, or naked PMO when looking at exon skipping and dystrophin restoration. However, the increase compared with CPP-PMO was limited. Thus, while the uncharged backbone makes PMO easily accessible for conjugation and an ASO with a large safety profile, the poor bioavailability seems to overrule a large part of the positive effect of a short muscle homing peptide. This is in line with previous reports in the mdx mouse model, where small muscle homing peptides could not improve the efficiency of PMOs. Yin and colleagues (2009 and 2010)^24,25 identified a muscle homing peptide, which resulted in a beneficial effect on exon skipping levels and functional outcomes in mdx mice only when combined with a CPP.

The increased tissue concentration of CP10-PMO and the PMOs conjugated with CP10 in combination with CPP compared with naked PMO and CPP-PMO, encourages us to believe that improved delivery to muscle tissue is achieved by using CP10. This indicates that a greater proportion of the administered PMO is in principle available to reach the nucleus, which brings us one step closer to efficient improvement of uptake for PMO treatment. In future studies, we aim to identify and further develop muscle homing peptides that will enable PMOs to reach their final destination.

Our results show that treating mdx mice with 30 mg/kg weekly for 4 weeks with L-CP10-CPP-PMO induced significantly higher muscle tissue delivery and exon skipping. We also observed lower levels of muscle damage serum markers compared with naked PMO treatment with no differences in liver and kidney safety markers. Improved distribution of dystrophin positive fibers seemed to lead to improved therapeutic effects, since almost complete recovery of pathology was observed. A mean of only 1.7% of inflammation and fibrosis was present and negligible amounts of regeneration were found in the muscles of L-CP10-CPP-PMO-treated mice. It is known that single CPP conjugation leads to improved effects.¹⁷ Here we provide evidence that a combination with muscle homing peptide CP10 improved effects to a somewhat greater extent and resulted in a significantly higher level of compound availability in the muscle tissue. We currently know the use of only CPPs conjugated to a PMO in patients can lead to hypomagnesemia.^18,19,20 Our results could indicate that using a muscle homing peptide in combination with a CPP could help to reduce the dosage to a level that could subsequently reduce adverse side effects of the CPP. Currently, this proof-of-concept study includes a single time point and one dose. Future studies will be aimed at identifying doses at which improved effects can still be observed for L-CP10-CPP-PMO by performing dose escalation experiments, to assess whether the CP10-CPP conjugation can reduce the therapeutic dose compared with CPP alone, thus reducing toxicity. Furthermore, longer-term experiments, including functional analyses, and more extensive histological studies are warranted. Alternatively, research in the future will focus on studying and improving CP10 conjugation so that enhanced PMO activity in the nucleus could increase the therapeutic potential of PMO treatment, which could even possibly eliminate the need for CPPs altogether.

A notable result that we found is the effect of conjugation strategy on efficiency. We found that both L-CP10-CPP-PMO and B-CP10-CPP-PMO had different effects in skeletal muscle and diaphragm. As described previously, L-CP10-CPP-PMO significantly improves exon skipping and dystrophin restoration in these tissues, whereas B-CP10-CPP-PMO did not. This compound results in lower exon skipping compared with CPP-PMO in both tissues, leading to marginally improved dystrophin restoration in the diaphragm and performed worse than CPP-PMO in the gastrocnemius. In the heart, the method of conjugation did not affect exon skipping and dystrophin restoration. These results highlight that the conjugation strategy matters, which suggests that testing several conjugation strategies is warranted in these types of experiments. Currently, it is not known why our linear and branched compounds achieved different results. We speculate that the difference between branched and linear conjugation could be due to differences in their rigidity. It is possible that receptor binding spots were more accessible for linearly conjugated PMOs compared with their branched counterparts, since there would have been less steric hinderance of each of the conjugated peptides to each other (see Figures S30C and S30D for organic chemistry structure diagrams of both compounds). Future studies will be needed to find out more about the mechanism of action for the different strategies of conjugation of these peptides.

This study gave clear insights into the importance of considering different properties of ASO chemistries and linking methods. Our study also offers new insights into strategies for improving the uptake of PMO ASOs, which we believe are crucial for addressing the long-standing challenge of suboptimal delivery and activity of ASOs.

Materials and methods

In vitro testing of CP10

Dissolving CP10

The FITC-labeled CP10 peptide was dissolved in ultrapure Milli-Q water with the addition of 10% DMSO to ensure proper homogeneous dissolving and reduce aggregate formation. To determine the concentration of the peptide solution, the nanodrop was used, using the labels setting to measure absorbance. To ensure that the peptide concentration was measured under the right pH conditions, the stock solution was diluted 1:40 in Tris-HCl (pH 7.5) and measurements were considered correct when the peak was observed at 495 nm. Final concentration measurements were used to calculate the amount needed to get to a final concentration of 2.25 μM in 1 mL of the medium of the cultured cells.

Cell culture

Patient-derived (8036), control human skeletal muscle cells (KM155), and human cardiomyocytes (SV40) were cultured in a six-well plates on either 0.5% gelatin (8036 and KM155)- or collagen (SV40)-coated glass coverslips at 37°C and 5% CO₂. Myoblasts were seeded and cultured until 80%–90% confluence in proliferation medium, which consisted of skeletal muscle cell growth basal medium (PromoCell, C-23060) supplemented with growth medium supplementMix (Gibco C-39365), 15% heat-inactivated (HI) fetal bovine serum (FBS) (Gibco), and 50 μg/mL gentamicin (Sigma, G1272). When the desired confluency was reached, cells were switched to differentiation medium, which consisted of DMEM (1×)+GlutaMax-I (Gibco, 61965-026) supplemented with 2% HI horse serum (Gibco) and 1% penicillin/streptomycin (Gibco, 15140-122). After 72 h of differentiation, clusters of nuclei in myotube-like cell shapes could be observed, confirming proper differentiation. Cardiomyocytes were cultured in Prigrow I medium (applied biological materials TM001) containing 10% HI FBS and 1% penicillin/streptomycin. When full confluency was reached, the uptake experiment was initiated. For mouse C2C12 cells expressing EGFP (IVS2-654),²⁶ cells were expanded in collagen-coated flasks in DMEM (Gibco, 10-569-044) supplemented with 10% FBS (Cytiva, SH30071.03). At 24 h after seeding, media were changed to 100 μL of low serum media, which comprised DMEM media (Gibco, 10-569-044) supplemented with 2% horse serum (Gibco, 16050122) and 1× I-T-S- supplement (Sigma, I1884-1Vl). Cells were allowed to differentiate for 4 days, with low serum media changes every other day.

In vitro uptake experiment of the peptides

Using plain DMEM, cells were washed twice, and subsequently 1 mL of DMEM containing 2.25 μM of CP10 was added to the well, resulting in a level of DMSO that was below 1% to avoid cell toxicity. After 3 h of incubation, thorough washing was performed three times with plain DMEM. Then cells were fixed using ice-cold methanol for 5 min for myotubes and 10 min for cardiomyocytes, whereafter glass coverslips were mounted on microscopy slides using prolong gold containing DAPI. After drying, imaging was performed using a fluorescence microscope (BZ-X700, Keyence) to evaluate if green fluorescent signal could be observed with 20× magnification.

In vitro testing of peptide-PMO conjugates

To assess exon skipping, C2C12 cells were seeded at 10,000 cells/well in collagen coated 96-well plates (Corning, 356649). Differentiated myotubes were dosed for 24 h with compounds at the indicated concentrations in DMEM media with 10% FBS. RNA was extracted from treated cells using Zymo Quick-RNA 96 Kit (Catalog No. R1053). SuperScript III One-Step RT-PCR System with Platinum Taq DNA polymerase (Invitrogen, 12574-026) was used to synthesize cDNA. Primers used to assess exon skipping were: forward: 5′-CACATCTTTGATGGTGTGAGG-3′ and reverse 5′-CAACTTCAGCCATCCATTTCTG-3′. Exon skipping was quantified using LabChip GX Touch HT (Revvity), using the DNA 1K Reagent Kit (Catalog No. CLS760673).

In vivo testing of CP10 as a conjugate to PMO in the mdx mouse model

Compounds

Peptides (CP10, CPP, and L- and B-CP10-CPPs) were obtained from CPC Scientific and WuXi Apptec. The procedures for the preparation of the PMO and its peptide conjugates are described in supplemental information.

Animals

All experiments were approved by the Leiden University Medical Centre (LUMC) animal welfare body (PE.17.246.041, PE.17.246.025, and PE.17.246.054, performed under CCD number AVD1160020171407) and were carried out according to Dutch law. In this research, male mdx mice were recruited from local breeding pairs and were included at the age of 4 weeks (C57BL/10ScSn-Dmd^mdx/J). Using pathogen-free conditions, animals were housed with a 12-h light/dark cycle at 20.5°C. Ad libitum access to water and standard chow were available to all mice.

For our single-conjugate experiments, mice (n = 8 per treatment group) were intravenously injected via the tail vein with either naked PMO, CP10-conjugated PMO (CP10-PMO), or CPP-conjugated PMO (CPP-PMO) at 30 mg/kg PMO or the molar equivalent. Mice received four weekly injections and were sacrificed 1 week after the last injection.

For our double-conjugate experiment, mice (n = 5 per treatment group) were injected with either PMO, CP10-conjugated PMO (CP10-PMO), or CPP-conjugated PMO (CPP-PMO) or a combination of CP10 and CPP conjugated to the PMO (either conjugated in a branched or linear fashion) at 30 mg/kg PMO or the molar equivalent using the same regimen as for the single-conjugate experiment.

On the day of sacrifice, bodyweight was recorded and mice were put under a heating lamp whereafter blood was collected via a tail vein cut in capillary tubes (microvette CB300 CAT, Sarstedt). Tubes were left to stand for 30 min to 2 h and then spun down at 2,400 rpm for 15 min at 4°C whereafter the serum was retrieved and stored at −80°C until further use. After cervical dislocation, the gastrocnemius, tibialis anterior, quadriceps, triceps, diaphragm, heart, liver, and kidneys were collected.

Safety profile evaluation by serum marker screening

Collected serum samples were sent to IDEXX BioAnalytics for clinical chemistry evaluation. We used a custom panel that measured the following markers of liver and kidney function: alkaline phosphatase, ALT, AST, total bilirubin, blood urea nitrogen, globulin, total protein, and albumin. LDH and CK levels were also assessed to determine if tissue and muscle damage changed in response to treatment. Serum samples with an elevated hemolysis index were excluded from the dataset.

Exon skipping analysis

RNA was isolated by phase separation on ice. TRIzol isolation reagent (Invitrogen Thermo Fisher Scientific, REF21663501) was added to muscle tissues in tubes containing 1.4 mm zirconium beads and disrupted using a MagNaLyser (Roche) at 7,000 rpm for 20 s, which was repeated if needed. After full homogenization, at least one-half the volume of chloroform was added. Samples tubes were shaken vigorously by hand and then spun down at 13,000 rpm for 15 min at 4°C to separate the phases. The upper aqueous phase was retrieved and at least double the volume of isopropanol was added. Samples were spun down at 13,000 rpm for 10 min at 4°C to precipitate the RNA. Supernatant was removed and the RNA pellet was washed twice with 70% ethanol, whereafter the pellet was left to air dry. RNA was then resuspended in Milli-Q water, whereafter the concentration was determined using a nanodrop.

CDNA was synthesized using 1,000 ng RNA, with random hexamer primers (40 ng/μL) and dNTP mix (containing 10 mM of each nucleotide). This mix was incubated at 70°C for 5 min. Then, M-MLV reverse transcriptase (Promega, 200 U/μL) and rRNasin (Promega, 40 U/μL) were added to the reaction mix and incubated for 60 min at 42°C followed by 10 min of 70°C incubation to terminate the reaction. Then, PCR amplification was performed with primers at 10 pmol/μL: atccagcagtcagaaagcaaa (forward primer targeting mouse Dmd exon 22) and cagccatccatttctgtaagg (reverse primer targeting Dmd exon 24), Taq DNA polymerase (5 U/μL) and dNTPS (10 mM) in the following amplification protocol: 5 min at 94°C, followed by 30 cycles of 30 s at 94°C, 30 s at 60°C, and 30 s at 72°C, and a final extension time of 7 min at 72°C. Fragments were visualized by separation on a 1.5% agarose gel for 1 h at 130 V using the 100-bp GeneRuler DNA ladder of Thermo Fisher Scientific as a size standard. Exon skipping levels were semi-quantified using the bioanalyzer lab on a chip DNA1000 assay (Agilent 2100 bioanalyzer–DNA 1000 kit). The skip levels were determined by comparing the molarity of the skipped product with that of the total product (skipped and wild-type products).

Evaluating dystrophin protein levels

Protein isolation was performed by adding protein isolation buffer containing 10% glycerol, 10% SDS, and 1.5 M Tris at a pH of 6.8 to zirconium beads filled tubes to which muscle tissue was added. Using the magnaylzer machine, tissues were homogenized as described previously and thereafter heated at 95°C for 10 min. The protein concentration was determined using the Pierce BCA Protein Assay Kit of Thermo Fisher Scientific. Samples were analyzed using the Jess Simple Western machine (ProteinSimple, Bio-Techne), using the 66–440 kDa Jess chemiluminescence separation module (SM-FL005-1) with 25 capillaries. All samples were set at a total protein concentration of 0.5 μg/μL using the protein isolation buffer mentioned, the 5× fluorescent master mix included in the EZ standard pack 3, and 1:100 sample buffer that is included in the separation module kit (Bio-Techne, SM-FL005). The separation module preparation protocol was done according to manufacturers’ recommendation with the deviation of loading 4 μL in the plate wells instead of 3 μL. The primary antibody solution that was used in our experiments contained antibodies for dystrophin and vinculin (mouse anti-rabbit, AB154168 and mouse anti-rabbit, REF70062, respectively), both diluted to 1:100 in antibody diluent 2. Anti-rabbit horseradish peroxidase antibody was used as a secondary antibody (Bio-Techne, #042-206). The area under the curve for both dystrophin and vinculin peaks were identified using the compass for simple western software (Protein-Simple, version 6.3.0). To determine the level of dystrophin restoration per individual sample, dystrophin values were normalized to vinculin. Percentages of restoration were determined based on a wild-type dilution curve with known percentages of dystrophin that was taken along on each Jess plate. Linear regression was used to determine the percentage of dystrophin restoration. In diaphragm and heart tissues, dystrophin restoration was found to be lower than 2.5%, which was found to be below the lower limit of quantification. To allow comparison between these samples, the percentage of restoration was predicted based on an overall WT curve that was generated with data from all wild-type curve measurements across different experiments.

Muscle histology and immune fluorescence analysis

Gastrocnemius tissues were snap frozen in liquid nitrogen cooled isopentane and stored at −80°C until further processing. Then serial cryosections (Leica CM3050 S) of the muscles of the three mice that showed highest exon skipping levels per group were made. Throughout the muscle, sections of 8 μm were made and put on superfrost PLUS slides (Menzel-Glazer, Fisher Emergo) as to have representative sections spanning the whole muscle. Slides were kept at −80°C until needed for staining.

H&E staining was performed to examine overall pathology using the following steps: slides containing the frozen tissue sections were defrosted at room temperature for up to 30 min and then fixed in ice-cold acetone (J.T. Baker, #808.2500) for 5 min, whereafter they were air dried for up to 30 min. Using Milli-Q water, the slides were rinsed. Then for 3 min hematoxylin (Mayer’s, Agilent #S330930-2) staining was performed with hematoxylin followed by another wash step using Milli-Q water followed by a more thorough washing step with running tap water for 5 min. To de-stain, slides were dipped in acid ethanol (70% ethanol with 1% glacial acetic acid, MilliporeSigma #1000562500) quickly 10 times and were then washed under running tap water twice for 1 min. One other rinse was performed by submerging slides in Milli-Q for 1 min, whereafter slides were put into an eosin bath (Eosin Y solution; Sigma-Aldrich, #HT110232) with the addition of 0.5 mL glacial acetic acid per 100 mL of eosin for 35 s. Thereafter, dehydration was performed using submersion into 80%, 90%, and 100% ethanol. Subsequently, slides were put in xylene baths (J.T. Baker) for 5 min, twice until they were left to air dry and mounted using coverslips (24 × 60 mm, Menzel Gläser, Fisher Emergo #360209) and pertex mounting medium (HistoLab, #00801) and left to dry overnight. The middle section of the tissue was used for analysis.

For immunofluorescent staining, slides were thawed and fixed the same as described in the H&E protocol, whereafter a hydrophobic pen was used to select a section from the middle of the tissue. These sections were rinsed using PBS twice for 2 min. Blocking was performed for 60 min using blocking reagent containing 0.05% Tween in PBS with 5% horse serum. The blocking reagent was removed and primary antibodies were added to the sections and left to incubate overnight at 4°C. Primary antibody solution consisted of blocking reagent containing the following antibodies: rabbit anti-dystrophin (ab152777 GR3443483-4) diluted 1:250, mouse anti-MYH3 (sc-53091, #J0233 IgG) diluted 1:20, and rat anti-laminin α-2 (sc-59854, G1217 IgG) diluted 1:50. Then, sections were washed twice for 2 min using PBS. After this, secondary antibodies were added and incubated in the dark for 60 min at room temperature. Secondary antibody mix consisted of fresh blocking reagent containing goat anti-rabbit Alexa Fluor 594 (A11037), diluted 1:1,000, mouse anti-MYH3 Alexa Fluor 647 (A21235) diluted 1:1,000, and goat anti-rat Alexa Fluor 488 (A11006) diluted 1:1,000. Hereafter, sections were washed using PBS, twice for 2 min and left to air dry. Then, slides were mounted using ProLong Gold Antifade containing DAPI (Thermo Fisher Scientific P36935).

H&E-stained sections were imaged at 20× magnification using a Pannoramic 250 Flash III scanner (3DHISTECH). Corrections, where background dirt and noise were removed, were performed using Adobe Photoshop (2022). For analysis, ImageJ software was used (version 2.3.0/1.53f51), where the color deconvolution plugin was used to split images into three channels (green, blue, and pink). The pink channel provides a clear contrast between pathological elements in the tissue and healthy muscle tissue. The threshold function was then used to distinguish healthy tissue from inflamed and fibrotic tissue. This measurement was compared with whole tissue measurements to determine the total percentage of measured pathological tissue. Two independent researchers performed quantification. The mean of these measurements is used for our results.

For immunofluorescent-stained tissues, imaging was performed using a Zeiss Axio Scan.Z1 Slide Scanner with a 20× objective. ZEN software (Carl Zeiss Microscopy) was used to export the images using the same settings for all images and Adobe Photoshop was used again to remove background noise. Dystrophin-positive and MYH3-positive fibers were counted manually using the point tool in ImageJ. The number of fibers was divided by the total area of the muscle to calculate the number of fibers per square millimeter.

ELOHA to determine tissue concentration

Oligonucleotide capture probes were designed to hybridize to the 5′ end of the PMO sequence. These capture probes were covalently attached to Pierce Maleic Anhydride Activated Plates (Thermo Fisher Scientific 15110) by incubating coating solution (500-nM capture probe, 2.5% sodium bicarbonate solution) for 1 h at 37°C. Plates were then washed and blocked overnight with 10% milk/PBST at 4°C. Tissue samples were lysed in 20 μL of ELOHA lysis buffer (10 mM Tris-HCl pH 7.5, 0.5% IGEPAL, 100 mM NaCl, 5 mM EDTA) per milligram of protein before being digested with Proteinase K (Qiagen 19134) at a final concentration of 1 mg/mL for 60°C for 30 min; 30 μL of this digested lysate was then incubated with 120 μL Detection Probe Solution (333 nM biotinylated oligonucleotide detection probe targeting the 3′ sequence of the PMO, 500 mM guanidine thiocyanate, 0.04% lauryl sarcosine, 3 mM sodium citrate, and 1.25 mM DTT in PBST). These samples were then incubated overnight at 4°C on the blocked, capture probe-coated maleic anhydride plate. The next day, plates were washed with PBST before incubation with streptavidin-AP conjugate (Sigma-Aldrich 11093266910) for 1 h. Plates were washed with PBST and AttoPhos AP Fluorescent Substrate System (Promega S1000) was added for 20 min before the reaction was stopped with 10 μL of EDTA (0.5 M solution/pH 8.0) (Thermo Fisher Scientific 8BP2482100). Fluorescence was measured with a Spectramax i3 reader (Ex 440, Em555). Each sample was interpolated to a standard curve prepared for each compound with a sigmoidal, four-parameter logistic model to determine nanogram PMO per microliter of tissue lysate. The final tissue concentration value was reported as nanograms PMO per gram of tissue.

Statistical analysis

For the outcome measurements of our experiments, one-way ANOVA tests with a Dunnett’s multiple comparison test were used to determine if results between experimental groups vs. the control group, differed significantly from each other. When assumptions of equal variances (Brown-Forsythe test) or normal distribution (Shapiro-Wilk test) were not met the non-parametric alternative, a Kruskal-Wallis, was performed with a Dunn’s multiple comparisons test. All data are reported as mean values with SD. A result was deemed significantly different when p values of <0.05 were observed.

Exon skipping data of the CPP-PMO and CP10-PMO experimental groups compared with naked PMO for triceps and gastrocnemius were analyzed using one-way ANOVA. Exon skipping data for diaphragm and heart were analyzed using the non-parametric Kruskal-Wallis test. Dystrophin recovery data were analyzed using one-way ANOVA for triceps and the Kruskal-Wallis for the gastrocnemius. Exon skipping levels in the gastrocnemius, diaphragm, and heart for the double-conjugated compounds, CP10-PMO, and CPP-PMO compared with naked PMO were analyzed using the Kruskal-Wallis test. Dystrophin restoration was determined for the same tissues and analyzed by one-way ANOVA. The mean number of dystrophin-positive and MYH3-positive fibers per square millimeter in gastrocnemius tissues in each of these treatment groups was analyzed by one-way ANOVA and a Kruskal-Wallis test was performed to analyze H&E data of these groups. One-way ANOVA was used to analyze the differences in mean PMO concentration levels for each of the experimental groups compared with naked PMO in skeletal muscles and a Kruskal-Wallis test was used for heart tissue.

Data availability

The authors confirm that the data supporting the findings of this study are available within the article. Raw data were generated at Leiden University Medical Centre or Sarepta Therapeutics, and are available upon reasonable request.

Acknowledgments

We thank Dr. Vincent Mouly (Institute of Myology, Paris, France) for providing the human myogenic cell lines used in the study. We also thank Davy van de Vijver for his occasional help with tissue collection in our in vivo study. This work was carried out by the use of an unrestricted grant from Sarepta Therapeutics, Inc (SRP-LUMC-001). Xuyu Tan and Ryan A. Oliver were part of Sarepta Therapeutics at the time of the study. Kristin Ha passed away on May 19, 2024.

Author contributions

A.F.E.S., A.A.R., and S.M.G.J. were involved in conceptualization of this study. X.T., R.A.O., and V.G. synthesized and performed QC for the peptides and compounds used in this study. A.F.E.S. performed the main body of experimental work and analysis for this manuscript. C.L.T.d.W. performed all intravenous injections for this study and was continuously involved in tissue collection. M.L. performed experiments that involved the triceps muscle in this work. E.G.T., K.H., A.M., and S.G. were involved in tissue concentration experiments, serum marker screening experiments, and all C2C12 work for this study. A.F.E.S. wrote the original draft of this manuscript. V.G. provided the material and methods section for the used compounds in this study. E.G.T., A.M., V.G., and A.A.R. were involved in editing of this manuscript. A.A.R. and S.M.G.J. provided supervision throughout the duration of this study.

Declaration of interests

This work was funded by an unrestricted research grant from Sarepta Therapeutics, X.T., E.G.T., K.H., A.M., S.G., R.O., and V.G. were employees of Sarepta Therapeutics, at the time of this research. A.A.R. and S.J. are co-inventors on an LUMC owned patent on CP10. A.A.R. is a member of advisory committee of Sarepta.

For full transparency, the full disclosure statement from A.A.R.: A.A.R. discloses being employed by LUMC, which has patents on exon skipping technology, some of which have been licensed to BioMarin and subsequently sublicensed to Sarepta. As co-inventor of some of these patents, A.A.R. was entitled to a share of royalties. A.A.R. further discloses being an ad hoc consultant for PTC Therapeutics, Sarepta Therapeutics, Regenxbio, Dyne Therapeutics, Lilly, BioMarin Pharmaceuticals, Eisai, Entrada, Takeda, Splicesense, Galapagos, Sapreme, Italfarmaco, and Astra Zeneca. In the past 5 years, ad hoc consulting has occurred for Alpha Anomeric. A.A.R. also reports being a member of the scientific advisory boards of Eisai, Hybridize Therapeutics, Silence Therapeutics, Sarepta therapeutics, Sapreme, and Mitorx. SAB memberships in the past 5 years: ProQR. Remuneration for consulting and advising activities is paid to LUMC. In the past 5 years, LUMC also received speaker honoraria from PTC Therapeutics, Alnylam Netherlands, Italfarmaco, and Pfizer and funding for contract research from Sapreme, Eisai, Galapagos, Synaffix, and Alpha Anomeric. Project funding is received from Sarepta Therapeutics and Entrada via unrestricted grants.