Therapeutic Approaches to Neuromuscular Repeat Disorders: Facioscapulohumeral Muscular Dystrophy and Myotonic Dystrophy Type 1

Facioscapulohumeral muscular dystrophy (FSHD) and myotonic dystrophy type 1 (DM1) represent the most common adult-onset muscular dystrophies, affecting more than 50,000 people in the United States. Although their underlying pathophysiology differs, both diseases are caused by dominant, gain-of-function mutations amenable to therapeutic approaches that target RNA. The relatively high prevalence of FSHD and DM1 as monogenetic disorders and clear treatment strategies make both diseases ideal targets for therapeutic development. In this article, we review FSHD and DM1 pathophysiology and detail current clinical trials and late preclinical development of disease-modifying agents.

Facioscapulohumeral Muscular Dystrophy

FSHD is the second most common adult-onset myopathy after myotonic muscular dystrophy. FSHD is recognized clinically by a pattern of facial, scapular, and upper (humeral) weakness, although weakness usually is not limited to these muscles. Clinical presentation is most commonly seen in those in their late teens to early 30s, although childhood and older adult presentations can be seen. Progression of weakness typically is slow. Twenty percent of people with FSHD are wheelchair-dependent by age 50,^1,2 and a small minority develop significant respiratory muscle weakness. The social and economic burdens caused by FSHD are large.^1,2 Delays in the diagnosis of FSHD likely are related to the expectation that inherited myopathies present in childhood, a tendency toward asymmetric weakness, signs of inflammation commonly found on FSHD muscle biopsies, and complex genetic testing, which excludes FSHD from most myopathy gene panels.

The genetics underlying FSHD are complex but present unique therapeutic opportunities (Figure 1). FSHD is a toxic gain-of-function disease caused by loss of epigenetic silencing of the D4Z4 repeat region on the short arm of chromosome 4, either through contraction of the repeat (FSHD1; approximately 95% of cases) or direct loss of methylation without contraction of the D4Z4 region (FSHD2).³ Both FSHD1 and FSHD2 result in relaxation of chromatin, which, in the correct genetic context, results in inappropriate expression of the double homeobox protein 4 (DUX4) gene. Myocytes appear particularly vulnerable to the negative effects of DUX4 because of promotion of full-length DUX4 transcripts (DUX4fl), which is a toxic isoform. DUX4fl induces multiple injurious pathways, including inflammatory, oxidative stress, and apoptosis.⁴ Even in myocytes poised to express DUX4, expression is rare and intermittent, making measurement of DUX4 transcripts difficult.

DUX4 has a role in early embryonic development but has no defined role in muscle beyond this point, making it an ideal target for strategies aimed at reducing or eliminating postembryonic expression (Table 1). Difficulties with this approach include the rare expression of target transcripts and lack of a natural animal model.³ In addition to DUX4 reduction, multiple nontargeted or downstream therapies have been attempted, which have not been fruitful and are not discussed here.

Small Molecules

Most compounds we typically consider “drugs” are small molecules. Small molecules tend to be easily absorbed by the body, have broad biodistribution, and are taken into cells to bind to a specific target. In FSHD, small molecules may aim to reduce DUX4 expression, although other upstream and downstream targets also have been considered. The furthest along of these agents is losmapimod (Fulcrum Therapeutics, Cambridge, MA), which theoretically inhibits the DUX4 promotor through the P38 mitogen-activated protein kinase (MAPK) pathway. Whereas losmapimod reduced DUX4 target genes by 80% in rodents, a recently completed phase II trial did not demonstrate an effect on DUX4, although there were some modest functional gains (Efficacy and Safety of Losmapimod in Subjects With FSHD, NCT04003974).⁵ A phase III trial is ongoing (Efficacy and Safety of Losmapimod in Treating Participants With FSHD [REACH], NCT05397470). It should be noted that there is some concern about the long-term use of losmapimod because the P38 pathway is not unique to DUX4 and is important both in myogenesis and in the function of other cell types.

Oligonucleotides

In FSHD, oligonucleotides aim to reduce DUX4 expression by harnessing preexisting cellular mechanisms for regulation of RNA turnover. As the name suggests, oligonucleotides are structures made up of small numbers of joined nucleic acids. These include antisense oligonucleotides (ASOs), which are modified DNA sequences designed to pair with specific DUX4 mRNA sequences and target them for degradation through RNAse H enzyme recruitment.⁶ RNA interference (RNAi) uses small double-stranded RNA molecules, which are processed and loaded into the RNA-induced silencing complex. These complexes then engage with the target mRNA and induce turnover. ASOs and RNAi require regular dosing as their effects wane over time, although virally delivered RNAi may provide persistent DUX4 suppression through perpetual production of its silencing product.⁷ To exert their therapeutic effects, these oligonucleotides must survive extracellular travel, avoid excretion by the kidney, and enter the target cell to gain access to their target RNA. They also must be designed carefully to limit off-target effects and induction of an immune response.

At least 2 companies—Avidity Biosciences and Dyne Therapeutics—are using antibodies or antibody fragments against the transferrin receptor linked to an oligonucleotide to enhance uptake in muscle cells. Avidity Biosciences has advanced to a phase 1/2 trial of AOC 1020 and plans to enroll approximately 70 people with FSHD at ascending doses of its drug (Phase 1/2 Study of AOC 1020 in Adults with FSHD, NCT05747924).

CRISPR

People often ask about gene-editing techniques, specifically CRISPR (clustered regularly interspaced short palindromic repeats). There are now 2 CRISPR approaches for the treatment of FSHD. CRISPRe is a CRISPR-Cas9 gene-editing technique initially found in bacteria, in which a specific pattern of DNA repeats is recognized by CRISPR and then cut by Cas9 (CRISPR-associated protein 9). Because the DUX4 sequence is found in multiple areas in the human genome, targeting this sequence for editing is not ideal. Some groups have targeted the polyadenylation sequence for removal, because this sequence is critical for DUX4 expression. Results have been mixed.^8,9 Despite the exciting possibilities associated with CRISPRe, there are significant safety concerns regarding off-target genetic editing and the irreversible nature of the treatment. There also are concerns over the immunogenicity of CRISPR components given reports of preexisting immunity to bacterial Cas proteins. There are no clinical trials planned using a CRISPRe approach. CRISPRi is a modified technique in which Cas9 no longer cuts DNA, but instead helps block expression of a targeted gene sequence. This approach avoids the genotoxicity associated with CRISPRe. Epic Bio recently announced a plan for a phase 1 trial of EPI-321, a CRISPRi product, to treat FSHD.

Myotonic Dystrophy

Myotonic dystrophy is a multisystem disease with primary manifestations in muscle, heart, brain, and other organs (eye, skin, endocrine, and others). It is the most common inherited muscle disease of adults, with a worldwide clinical incidence of 1 in 8000, although this likely is an underestimate.¹⁰ There are 3 main types of myotonic dystrophy: type 1 (DM1), type 2, and congenital. Except for tideglusib ([AMO-02] AMO Pharma, Durham, NC), therapeutic development has focused largely on DM1, which is the focus of this review. We present a broad overview of therapies in development for DM1; the topic has been reviewed extensively elsewhere.^11,12

Clinical presentation of classic DM1 is highly variable, with age at onset ranging from childhood to late in life. Weakness and atrophy typically involve distal limb, facial, and oropharyngeal muscles early in the disease, with subsequent involvement of proximal muscles. Other cardinal manifestations include early cataracts, myotonia, frontal balding, cardiac conduction disease, and endocrinopathies (eg, diabetes, hypertriglyceridemia, thyroid disease).¹⁰ Respiratory weakness and sudden cardiac death may lead to a shortened lifespan.¹³

The genetic cause of DM1 explains some of the clinical variability. DM1 is caused by expansion of a trinucleotide CTG repeat tract in the myotonic dystrophy protein kinase gene (DMPK),¹⁴ with lengths of >50 CTG repeats being pathogenic.¹⁰ Once expanded, the CTG repeat is unstable in the germline and there is a tendency toward further repeat expansion in subsequent generations. However, there are likely important modifiers of severity that have yet to be identified, as repeat length itself accounts for only a small percentage of DM1 clinical variability overall.¹⁵

A key concept of the mechanism driving myotonic dystrophy is that the pathogenic trinucleotide repeat does not encode an abnormal DMPK protein. Instead, DMPK RNA transcripts containing the repeat are toxic through binding and sequestration of important RNA-processing proteins.¹⁶ Investigational therapies for DM1 have been designed around this mechanism (Table 2). They fall into 4 main categories: (1) lowering the amount of toxic RNA accumulation; (2) increasing the amount of RNA-processing proteins to overcome sequestration; (3) blocking the interaction between toxic RNA and RNA-processing proteins; and (4) targeting downstream pathways (Figure 2).

Lowering Toxic RNA Accumulation

Strategies to lower toxic repeat RNA have gained the most attention and have progressed to early phase clinical trials. The majority of these approaches involve using chemically modified oligonucleotides to co-opt the cell’s own machinery to induce turnover of toxic RNA. These oligonucleotide approaches include ASOs to engage the RNAse H enzyme or RNAi oligonucleotides that are loaded into and direct activity of the RNA-induced silencing complex.

The first-in-class trial for DM1 used naked ASOs not conjugated to a delivery molecule, and, despite good activity in preclinical studies, in a phase 1/2 clinical trial they failed to accumulate to an adequate level in skeletal muscle¹⁷ (A Safety and Tolerability Study of Multiple Doses of ISIS-DMPKRx in Adults With Myotonic Dystrophy Type 1, NCT02312011). Pursuant to this result, recent development has focused on boosting delivery of oligonucleotides by conjugating them to various ligands or packaging into gene therapy vectors (example, KT430, Astellas Pharma, Tokyo, Japan). Examples of oligonucleotide conjugates in preclinical and clinical trial include antibody or antibody fragments targeting the transferrin receptor (Avidity Biosciences, San Diego, CA, and Dyne Therapeutics, Waltham, MA), palmitate hexylamine phosphodiester fatty acid (Ionis Pharmaceuticals, Carlsbad, CA), or cell-penetrating peptides.¹⁸ The furthest along of these approaches is the phase 1/2 clinical trial by Avidity Biosciences (investigational AOC 1001, Avidity Biosciences, San Diego, CA), which reported interim results that showed a mean 45% reduction of DMPK mRNA and improvement of biomarkers that reflect increased activity of the RNA-processing proteins (Study of AOC 1001 in Adult Myotonic Dystrophy Type 1 Patients [MARINA], NCT05027269).

The next category of medicines identified through screening or rationally designed to reduce toxic RNA is small molecules. In general, these are molecules with intrinsic RNA-degradation properties that selectively bind the structure formed by the toxic RNA. All these approaches remain in preclinical status and have been reviewed in depth.^19,20

In addition to oligonucleotides and small molecules, engineered proteins that degrade specific RNAs (endonucleases) have been developed. This includes artificial site-specific endonucleases, and, more recently, CRISPR. Whereas the initial CRISPR-associated (Cas) proteins identified were capable of targeting DNA, the discovery of RNA-targeting Cas proteins broadened their application. Using adeno-associated virus (AAV) to deliver the CRISPR machinery to the cell, Cas-mediated RNA-targeting approaches are in preclinical development for DM1.^21,22 Primary concerns with this approach include specificity of knockdown, toxicity of AAV-mediated delivery, and inability to redose if expression of the protein wanes over time.

Another approach to reduce toxic RNA accumulation involves inhibiting transcription, or the initial production, of the expanded repeat. This includes genome modification by endonucleases (CRISPR or transcription activator-like effector nucleases) through removal of the repeat,²³ insertion of a sequence that stops transcription upstream of the repeat,²⁴ binding to the expanded repeat to block transcription,²⁵ or small molecules that (1) selectively bind to the repeat DNA to block the transcription machinery, (2) inhibit the transcription machinery directly, or (3) block somatic expansion of the repeat.¹⁹ Companies in preclinical development with such approaches include Genethon (Evry, France), Expansion Therapeutics (San Diego, CA), and Design Therapeutics (Carlsbad, CA).

Increasing the RNA-Processing Proteins to Overcome Sequestration

Preclinical studies have established proof of concept for this approach by delivering the protein directly, identifying small molecules that increase the protein,²⁶ or blocking cellular mechanisms that naturally reduce the amount of the RNA-processing proteins.²⁷ The latter approach involves delivering a cell-penetrating peptide-linked morpholino to block the cell’s own RNAi from reducing the RNA-processing protein. All 3 methods lead to increased levels of the RNA-processing proteins, allowing for recovery of function and correction of DM1 phenotypes in preclinical models.

Blocking the Interaction Between Toxic RNA and RNA-Processing Proteins

Initial work defining the mechanism of DM1 used morpholino oligonucleotides to bind and block the interactions of toxic RNA.²⁸ Morpholinos, which are synthetic ASOs, and peptide nucleic acids have been designed to bind to the toxic RNA repeat and block interactions with cellular proteins without causing knockdown. One drawback of a blocking approach is that it requires increased concentration of drug, compared with knockdown strategies, as knockdown oligonucleotides need to bind the target only once. Companies pursuing this strategy (NeuBase, Pepgen, Entrada-Vertex) have developed conjugates to enhance delivery to muscle tissue.

Small molecules that bind to the repeat RNA to block interactions with RNA-processing proteins have been identified,¹⁹ including the antibiotic erythromycin, for which there is a completed phase 2 clinical trial for DM1 (Safety and Efficacy Trial of MYD-0124 for Myotonic Dystrophy Type 1, jRCT2051190069).

Targeting Downstream Pathways

In addition to targeting the proximal molecular mechanism of DM1, small molecule drugs that interact with downstream signaling pathways (eg tideglusib) or were initially investigated to treat DM1 symptoms (eg mexiletine and metformin) have been found to have primary effects on the overall disease.^29-31 Because all 3 of these medications were already approved by the Food and Drug Administration before their application to DM1, they are furthest in clinical development, and are in phase 3 clinical trials for DM1 or congenital myotonic dystrophy (NCT05004129, NCT03692312, NCT04700046, NCT04624750, NCT05532813).