Although the nature and locations of the repeat expansions are different for DM1 and DM2, the underlying molecular mechanism, sequestration of MBNL and consequent missplicing of a wide range of transcripts, is ostensibly the same. It then begs the question, why are the phenotypes of DM1 and DM2 different?

Understanding Molecular Mechanisms in DM1 and DM2

One “logical” interpretation for the milder phenotype in DM2 is that CCUG_exp repeats may be less toxic. Perhaps MBNL is not sequestered as efficiently by DM2 repeats (hence DM2 patients typically have much longer repeats) and thus not depleted to the same extent as in DM1. But there are many potential alternative hypotheses, including differences due to the genomic locations of the expanded repeats or the presence of as yet unknown modifiers that act upon the expanded repeat transcript or other potential targets upstream of the ensuing splicing anomalies.

Drs. Beatriz Llamusi (University of Valencia) and Nicolas Charlet-Berguerand (Université de Strasbourg) and colleagues have identified an RNA binding protein that acts to reduce MBNL depletion in DM2, but not in DM1. Their data suggests a molecular explanation that, at least in part, reconciles the phenotypic differences between DM1 and DM2.

rbFOX1 binds CCUG_exp, but not CUG_exp RNA

To assess whether alternative RNA binding proteins may be more efficient at binding to CCUG_exp versus CUG_exp, and thereby mitigate MBNL sequestration in DM2, the research team conducted an unbiased assay for C2C12 cell nuclear extract interactions with the two expanded repeat sequences. They identified several proteins that bound to the expaned repeat sequences, but, among the hits, only rbFOX family RNA binding proteins interacted with CCUG_exp, but not CUG_exp. They further showed that rbFOX1 localizes to nuclear foci only in cells containing the DM2 repeats.

The research team them utilized gel-shift and UV-crosslinking experiments to demonstrate the mechanism by which rbFOX1 is retained in nuclear foci containing CCUG_exp sequences—these studies confirmed a direct interaction between rbFOX1 and CCUG_exp.

rbFOX1 is Bound to CCUG_exp in DM2 Patient Cells

In additional experiments, the research team established that rbFOX1 co-localized with CCUG_exp–containing foci in primary muscle cultures derived from DM2 patients, as well as in sections from DM2 patient skeletal muscle. Yet, additional studies showed that rbFOX1 was not immobilized in CCUG_exp RNA foci in the same manner as MBNL. rbFOX1 was not detected in similar nuclear analysis of DM1 patient cells or tissues.

Excess rbFOX1 Displaces MBNL from CCUGexp and its Overexpression Corrects Splicing and Muscle Atrophy

Although rbFOX1 and MBNL both bind CCUG_exp, their interactions may or may not be competitive. But, expression of an excess of either rbFOX1 or MBNL was shown to compete the other away from binding to the DM2 expanded repeat; depletion of rbFOX1 increased the localization of MBNL to CCUG_exp. These data are consistent with the hypothesis that rbFOX1 and MBNL compete for the same binding site.

The next step was to test whether an overabundance of rbFOX1 could mitigate the downstream effects of MBNL depletion. Overexpression of rbFOX1 in C2C12 cells carrying CCUG_exp resulted in partial correction of splicing of Clcn1 and Tnnt2 mini-genes, while reduction of rbFOX1 by siRNA increased the splicing defect. Finally, the research team overexpressed rbFOX1 in Drosophila models of DM1 and DM2 and showed rescue of skeletal muscle atrophy only in the DM2 model.

Taken together, the investigators here provide a molecular model to explain, at least in part, the phenotypic differences between DM1 and DM2. Reduced sequestration of MBNL by CCUG_exp versus CUG_exp repeats by rbFOX1 would mitigate the splicing defects and thereby potentially contribute towards the milder pathology in DM2.

Reference:

rbFOX1/MBNL1 competition for CCUG RNA repeats binding contributes to myotonic dystrophytype 1/type 2 differences.
Sellier C, Cerro-Herreros E, Blatter M, Freyermuth F, Gaucherot A, Ruffenach F, Sarkar P, Puymirat J, Udd B, Day JW, Meola G, Bassez G, Fujimura H, Takahashi MP, Schoser B, Furling D, Artero R, Allain FHT, Llamusi B, Charlet-Berguerand N.
Nat Commun. 2018 May 22;9(1):2009. doi: 10.1038/s41467-018-04370-x.

Frontiers of Neurology Publishes a Series of DM Reviews

A new set of six review articles is being published on various aspects of DM. These articles comprise a special issue of Frontiers in Neurology, entitled: ‘Beyond Borders: Myotonic Dystrophies—A European Perception', co-edited by Profs. Giovanni Meola and Benedikt Schoser. Together, these articles represent an important new reference source for DM stake-holders.

Coverage for A Broad Range of DM Topics

Wenninger et al. (2018) note the difficulties caused in interpretation of the highly variable and multi-systemic phenotypes exhibited in DM1 and DM2. They have focused this review on what they define as the ‘core clinical features’ of DM, with the intent that this perspective can facilitate early diagnosis and treatment.

André et al. (2018) review the evidence that DM1 and DM2 have an impact throughout the skeletal muscle life cycle, influencing its development, growth, regenerative capacity, and premature aging. They suggest that invoking the well-recognized involvement of the expanded repeat and RNA dysfunction alone may be too simplistic an explanation for the pathogenesis of skeletal muscle in DM.

Matloka et al. (2018) take on the important issue of using cell-based systems to understand pathogenic mechanisms and to screen candidate therapeutics for DM. Their article reviews the history of utilization of cellular models in DM1 and catalogs models that are available for basic science and translational research.

López-Morató et al. (2018) recognize the extensive efforts to therapeutically target the depletion of MBNL by expanded repeat DMPK transcripts using large molecules specifically designed for the target. In their review, they look beyond those efforts at other targets that have been engaged by small molecule drug discovery and development programs for DM1.

Two other articles remain to be published—see below.

Publication Status

The complete articles by Chakraborty et al. (2018) and Mahyera et al. (2018) are not yet available, although the abstracts are on the Frontiers of Neurology website. The work by Dr. Chakraborty and colleagues discusses use of fly models to understand the molecular and physiologic pathogenesis of the heart in DM1 and DM2, and the small molecules that have been evaluated for cardiac dysfunction in these models. The work by Dr. Mahyera and colleagues discusses the population distribution and structure of expanded repeat tracts in DM2 patients in Germany.

Four of the article pdfs are available, open-access, at the Frontiers in Neurology website. The other two articles will be posted to this site when published.

References:

Abnormalities in Skeletal Muscle Myogenesis, Growth, and Regeneration in Myotonic Dystrophy
André LM, Ausems CRM, Wansink DG, Wieringa B.
Front Neurol. 2018 May 28;9:368. doi: 10.3389/fneur.2018.00368. eCollection 2018. Review.
[This article is available as a pdf on the journal website but was not yet listed in PubMed at the time of this writing, so the link to the PubMed abstract was not yet available.]

Modelling of myotonic dystrophy cardiac phenotypes in Drosophila.
Chakraborty M, Llamusi B, Artero R.
Front Neurol. doi: 10.3389/fneur.2018.00473. eCollection 2018. Review.
[This article was not listed in PubMed, and it is listed as ‘provisionally accepted’ on the Frontiers in Neurology website. Thus, the citation is incomplete and neither the PubMed link nor the pdf were available at the time of this writing.]

Small Molecules Which Improve Pathogenesis of Myotonic Dystrophy Type 1.
López-Morató M, Brook JD, Wojciechowska M.
Front Neurol. 2018 May 18;9:349. doi: 10.3389/fneur.2018.00349. eCollection 2018. Review.

Distribution and Structure of DM2 repeat tract alleles in the German Population.
Mahyera AS, Schneider T, Halliger-Keller B, Schrooten K, Hörner E-M, Rost S, Kress W.
Front Neurol. doi: 10.3389/fneur.2018.00473. eCollection 2018. Review.
[This article was not listed in PubMed, and it is listed as ‘provisionally accepted’ on the Frontiers in Neurology website. Thus, the citation is incomplete and neither the PubMed link nor the pdf were available at the time of this writing.]

Cells of Matter-In Vitro Models for Myotonic Dystrophy.
Matloka M, Klein AF, Rau F, Furling D.
Front Neurol. 2018 May 23;9:361. doi: 10.3389/fneur.2018.00361. eCollection 2018. Review.

Core Clinical Phenotypes in Myotonic Dystrophies.
Wenninger S, Montagnese F, Schoser B.
Front Neurol. 2018 May 2;9:303. doi: 10.3389/fneur.2018.00303. eCollection 2018. Review.

Microsatellite expansions triggering a variety of neurological disorders occur in both coding and non-coding regions. Trinucleotide CNG expansions (e.g., DM1) predominate in exonic and UTR regions, while intronic expansions can vary from tri- to hexanucleotide repeats and are associated with eight diseases that include DM2, C9orf72 ALS/FTD, and Fuchs endothelial corneal dystrophy. How the sequence and differential localization of microsatellite expansions impacts disease mechanisms is not fully known.

Dr. Maury Swanson (University of Florida), Łukasz Sznjader, a 2016-2017 MDF Fellow, and researchers from the University of Rochester, Houston Methodist Hospital, and Adam Mickiewicz University in Poland, have recently evaluated the concept that intronic microsatellite expansions themselves act as a trigger to host gene intron retention and thereby to the pathogenesis of disease. In this study, they assessed a variety of neurological diseases linked to GC- and A/AT-rich intronic expansions.

The presence of microsatellites in introns alone conveys very low risk of expansion and triggering of inherited diseases—of 80,000 intronic microsatellites with expansion potential, only 0.01% are actually known to expand and cause disease. The research team mapped the intronic localization of both pathogenic expanded microsatellites and unexpanded repeats and demonstrated a bias of disease-causing microsatellites toward splice sites and thereby represent potential triggers for splicing alterations and intron inclusion. CNBP intron 1 inclusion as determined from a DM muscle RNA-seq database and patient biopsy samples was substantially elevated in contrast to intron inclusion across a broad range of neuromuscular disease controls. Notably, only CNBP intron 1 (the site of the microsatellite expansion) was spliced in, while other downstream CNBP introns were not. Further studies showed that intron 1 retention was not a developmentally regulated event.

Dr. Swanson and colleagues then tested the hypothesis that CNBP intron 1 inclusion in DM2 patient peripheral blood could be used as a disease biomarker. These studies confirmed intron inclusion in DM2, but not in controls, in a repeat length-dependent fashion. At this point it was as yet unclear whether the expanded CCUG repeat itself was mechanistic in altering CNBP splicing to include intron 1. A mouse reporter gene model was used to confirm that intron inclusion was driven by CCUGexp in a repeat length-dependent manner. Finally, the research team showed that intronic retention occurred only for GC- (e.g., DM2), but not A/AT-rich (e.g., Friedrich’s ataxia), microsatellite expansion diseases.

Taken together, this report identifies the molecular mechanisms underlying intronic retention in DM2 and emphasizes its value as a putative biomarker, particularly since intronic retention is so easily assessed in peripheral blood using routine RT-PCR assays. Such an approach to disease detection is much more efficient than conventional genetic strategies as a means of mapping heritable microsatellite expansion disorders and can be broadly used in discovery of the genetic basis of currently undiagnosed diseases.

Reference:
Intron retention induced by microsatellite expansions as a disease biomarker.
Sznajder ŁJ, Thomas JD, Carrell EM, Reid T, McFarland KN, Cleary JD, Oliveira R, Nutter CA, Bhatt K, Sobczak K, Ashizawa T, Thornton CA, Ranum LPW, Swanson MS.
Proc Natl Acad Sci U S A. 2018 Apr 2. pii: 201716617. doi: 10.1073/pnas.1716617115. [Epub ahead of print]

Robust, reliable, and reproducible endpoints are essential for decision making in early stage drug discovery and development programs. Given the high level of understanding of molecular events in the pathogenesis of DM, assays that focus on careful quantification of disease-causing molecules are essential for go/no-go decision making in drug screening programs. Thus far, methodology for reliable quantification of mutant RNA copy number, although a critically important tool for drug developers, has been a question mark. A recent study utilized a battery of techniques (Northern blotting, reverse transcriptase-quantitative polymerase chain reaction, RNA-sequencing and fluorescent in situ hybridization analyses) to demonstrate a very low abundance—one to a few dozen molecules per cell—Dmpk/DMPK mRNA in mouse model and patient samples (Gudde et al., 2016).

Dr. David Brook (University of Nottingham) and colleagues have reported out on studies of methodology to precisely quantify (a) the absolute numbers of mutant RNA transcripts and (b) alternative splicing in DM1 and DM1 patient cell samples (Wojciechowska et al., 2018). In these studies, they evaluated the potential value of medium throughput technologies—Multiplex Ligation-Dependent Probe Amplification (MLPA) and droplet digital PCR (ddPCR)—that could be feasibly applied in drug development programs.

To quantitatively assess alternative splicing, the research team designed assay methodology that would evaluate eight key splicing events in DM1 and DM2 patient skeletal muscle biopsy samples. MLPA used two probes for each exon (exon-On and exon-Off). Likewise, ddPCR utilized dual TaqMan probes for each splicing event, specific to exon-On and exon-Off. Transcripts evaluated met specific selection criteria and represented known changes at different disease stages. Both MLPA and ddPCR met rigorous criteria for splicing quantification (sensitivity, reliability) and results showed strong correlation across the two methods.

To achieve absolution quantification of DMPK mRNA copy number, the group optimized ddPRC assays and used them to characterize DM1 skeletal muscle biopsies and cultured fibroblasts. Mutant and wild type DMPK transcripts were differentiated on the basis of the rs527221 G>C SNP in exon 10. Results suggested that highly sensitive, reproducible, and accurate quantification was achievable even with the low mutant mRNA copy number (15-20 molecules/cell) that characterize DM.

In addition to assessing the appropriateness of MLPA and ddPCR as biomarker assays for myotonic dystrophy, the research team demonstrated that the absolute abundance of mutant DMPK transcripts in the nuclear foci of DM1 patient fibroblasts was very low. Thus, sensitive and reliable assays that can avoid floor effects are important for the field.

Finally in situ fluorescence hybridization was used to estimate mutant DMPK mRNA in individual nuclear foci. The median number of mutant CUG transcripts per nuclear foci was between two and four, with between four and six total foci detected per DM1 patient fibroblast.

Taken together, both MLPA and ddPCR technologieswere shown to produce highly reproducible and simultaneous assessment of genes aberrantly spliced in DM, and ddPCR was an effective measure of absolute mutant DMPK transcript copy numbers per cell. Validation of this methodology addresses an important gap in tools available for the therapy development pipeline for DM.

References:

A low absolute number of expanded transcripts is involved in myotonic dystrophy type 1 manifestation in muscle.
Gudde AE, González-Barriga A, van den Broek WJ, Wieringa B, Wansink DG.
Hum Mol Genet. 2016 Apr 15;25(8):1648-62. doi: 10.1093/hmg/ddw042. Epub 2016 Feb 16.

Quantitative Methods to Monitor RNA Biomarkers in Myotonic Dystrophy.
Wojciechowska M, Sobczak K, Kozlowski P, Sedehizadeh S, Wojtkowiak-Szlachcic A, Czubak K, Markus R, Lusakowska A, Kaminska A, Brook JD.
Sci Rep. 2018 Apr 12;8(1):5885. doi: 10.1038/s41598-018-24156-x.

Cardiac Biomarkers and DM

Cardiac rhythm disturbances represent a cardinal feature and a leading cause of death in DM. To elevate the level of patient cardiac care, consensus-based care recommendations are now available for DM1, children with CDM or DM1, and DM2. The molecular basis of cardiac dysfunction in DM, however, has been difficult to discern. It has been suggested that serum levels of high-sensitive cardiac troponin T and N-terminal pro B-type natriuretic peptide may be predictive of cardiac risk and potentially useful for stratification (Valaperta et al., 2016 and 2017), but these have not yet shown promise as efficacy biomarkers for interventional clinical trials. Similarly, mis-splicing of SCN5A has been implicated in DM cardiac conduction defects (Freyermuth et al., 2016; Pang et al., 2018), but its value as a clinical trial tool has not been determined. Thus, specific biomarkers for cardiac involvement in DM have not yet been established and their absence represents an important gap in the ability to assess candidate therapeutics for a key phenotype in this patient group.

Assessing Mis-Splicing of Cardiac-Relevant Transcripts

Drs. Rosanna Cardani and Giovanni Meola (IRCCS-Policlinico San Donato and University of Milan) and colleagues initiated a study of alternative splicing of several genes that could be used to follow the cardiac phenotype in DM1 or DM2 patients. Analyses were performed in patients with and without cardiac involvement; molecular analyses focused on TNNT2 expression, in part because it is mis-spliced in both skeletal and cardiac muscle in DM.  Dr. Laura Valentina Renna, a former MDF fellow, contributed to this work.

The research team evaluated skeletal muscle biopsies in 24 DM1, 9 DM2 patients, and 10 age-matched controls; subjects also underwent muscle strength evaluations (MRC scale), staging of DM1 (MIRS), ECG, and Holter tests. Their study also included a range of histologic and immunocytochemical markers to establish correlations between observed splicopathies and skeletal muscle status.

TNNT2 encodes cardiac troponin T (cTnT). The research team showed that TNNT2 mis-splicing was more evident in skeletal muscle biopsies from DM1 subjects (where the fetal isoform was > 50% of total transcript) versus DM2. The authors suggest that greater mis-splicing in DM1 may relate not only to the more severe myopathy, but to the general disease severity, including cardiac involvement, in DM1. Finally, the level of TNNT2 mis-splicing strongly correlated with QRS duration abnormalities in DM1 (but not DM2), suggesting that alternative splicing of TNNT2 may have value as a cardiac biomarker.

TNNT2 as a Biomarker of DM1 Severity?

Overall, the authors cannot conclude that TNNT2 mis-splicing is a specific biomarker of cardiac involvement in DM, only that data are suggestive that it may be when further data are acquired. They do argue that TNNT2 mis-splicing may function as a biomarker of disease severity in DM1, the potential of which should be explored in natural history studies and interventional clinical trials.

References:

High-sensitive cardiac troponin T (hs-cTnT) assay as serum biomarker to predict cardiac risk in myotonic dystrophy: A case-control study.
Valaperta R, Gaeta M, Cardani R, Lombardi F, Rampoldi B, De Siena C, Mori F, Fossati B, Gaia P, Ferraro OE, Villani S, Iachettini S, Piccoli M, Cirillo F, Pusineri E, Meola G, Costa E.
Clin Chim Acta. 2016 Dec 1;463:122-128. doi: 10.1016/j.cca.2016.10.026. Epub 2016 Oct 22.

Cardiac involvement in myotonic dystrophy: The role of troponins and N-terminal pro B-type natriuretic peptide.
Valaperta R, De Siena C, Cardani R, Lombardia F, Cenko E, Rampoldi B, Fossati B, Brigonzi E, Rigolini R, Gaia P, Meola G, Costa E, Bugiardini R.
Atherosclerosis. 2017 Dec;267:110-115. doi: 10.1016/j.atherosclerosis.2017.10.020. Epub 2017 Oct 21.

Splicing misregulation of SCN5A contributes to cardiac-conduction delay and heart arrhythmia in myotonic dystrophy.
Freyermuth F, Rau F, Kokunai Y, Linke T, Sellier C, Nakamori M, Kino Y, Arandel L, Jollet A, Thibault C, Philipps M, Vicaire S, Jost B, Udd B, Day JW, Duboc D, Wahbi K, Matsumura T, Fujimura H, Mochizuki H, Deryckere F, Kimura T, Nukina N, Ishiura S, Lacroix V, Campan-Fournier A, Navratil V, Chautard E, Auboeuf D, Horie M, Imoto K, Lee KY, Swanson MS, de Munain AL, Inada S, Itoh H, Nakazawa K, Ashihara T, Wang E, Zimmer T, Furling D, Takahashi MP, Charlet-Berguerand N.
Nat Commun. 2016 Apr 11;7:11067. doi: 10.1038/ncomms11067.

CRISPR -Mediated Expression of the Fetal Scn5a Isoform in Adult Mice Causes Conduction Defects and Arrhythmias.
Pang PD, Alsina KM, Cao S, Koushik AB, Wehrens XHT, Cooper TA.
J Am Heart Assoc. 2018 Oct 2;7(19):e010393. doi: 10.1161/JAHA.118.010393.

TNNT2 Missplicing in Skeletal Muscle as a Cardiac Biomarker in Myotonic Dystrophy Type 1 but Not in Myotonic Dystrophy Type 2.
Bosè F, Renna LV, Fossati B, Arpa G, Labate V, Milani V, Botta A, Micaglio E, Meola G, Cardani R.
Front Neurol. 2019 Sep 27;10:992. doi: 10.3389/fneur.2019.00992. eCollection 2019.

Cutting Through the Hype

As if we needed actual data to understand that CRISPR technology is a hot topic—a PubMed search on “CRISPR” yields over 7,000 hits, while the NIH RePORTER database search returns 2,316 active research grants with reference to the topic. Potentially transformative therapeutic strategies have always been hyped, but the explosion in the mass media coverage makes it difficult to believe that anyone remains who has not heard the term CRISPR and the unfortunate implication that marketed therapies are “imminent.”

Our task in the scientific, pharma/biotech and advocacy communities is to counter the therapeutic misconception that CRISPR will be a panacea for any inherited disease, while taking pragmatic steps to address the not inconsequential barriers of rigorous efficacy studies, scale-up production, delivery/exposure, efficiency, and even ethics to optimize and evaluate the potential of this important new tool.

Finding the Optimal CRISPR Targets

Initial efforts to apply CRISPR/Cas9 approaches to DM focused on reducing/removing expanded CTG tracts from the DMPK gene. An additional challenge for gene editing in expanded repeat disorders is the need to remove large DNA tracts. Indeed, Dr. Bé Wieringa and colleagues found that removal of expanded CTG repeats was feasible, but required cutting from both sides of the repeat to avoid increasing the genomic instability that drives DM (see http://www.myotonic.org/gene-editing-dm). DNA targeting may ultimately prove to be the optimal approach for DM, but more developmental and mechanistic work is required to facilitate in situ removal of large expansions safely and efficiently.

Within the CRISPR field, the approach of targeting RNA has attracted considerable interest. Two recent papers from Dr. Feng Cheng’s lab (Broad Institute of MIT and Harvard) have demonstrated RNA targeting using CRISPR/Cas13.  In the first (Cox et al., 2017), they evaluated programmable editing of RNA transcripts to alter coding and thereby correct mutations at the level of the transcript (a system they designate as RNA Editing for Programmable A to I Replacement or REPAIR). This system is designed to address single base mutations and thus may have limited to no applicability in repeat expansion disorders, although continued attention to the associated technological advances is warranted. The second publication (Abudayyeh et al., 2017) demonstrates that CRISPR/Cas13a has the potential for targeted knockdown of RNA, with similar efficiency to and better specificity than RNAi. Such an approach is comparable to knockdown of DMPK transcripts with antisense oligonucleotides that recruit RNaseH mechanisms.

To Target RNA or DNA in DM?

Previously, Dr. Gene Yeo’s group showed that RNA-targeted Cas9 in an in vitro model degrades toxic DMPK transcripts, increases in the MBNL protein, and reduces or eliminates the gene splicing defects that characterize DM (see http://www.myotonic.org/modifying-gene-editing-technology-dm).

In a recent paper in Molecular Cell (Pinto et al., 2017), Dr. Eric Wang and colleagues (University of Florida) reported that deactivated Cas9 (dCas9) specifically binds expanded repeat tracts at the DMPK locus and impedes their transcription. They evaluated multiple CTG repeat lengths and showed that dCas9 coupled to a guide RNA achieved knockdown in a repeat length-dependent manner, with binding efficiency and potency of knockdown increasing with the number of repeats. Moreover, the group showed that the dCas9-gRNA strategy reduces nuclear foci, restores normative splicing, and blocks RAN translation in both HeLa cell models and DM1 patient myoblasts and in skeletal muscles of systemically injected HSA-LR mice.

Next Steps for DM

Efforts to target the expanded repeat DMPK RNA or DNA have shown potential as a therapeutic strategy, but it must be recognized that these efforts are at a preclinical proof of concept stage. Both approaches face questions of delivery, specificity of targeting and reagent efficiency once on target that are likely to become more complex when CRISPR as a candidate therapeutic transitions to clinical testing. For example, if AAV is used for CRISPR reagent delivery, the tissue targeting specificity inherent to many AAV serotypes must be overcome if the multi-systemic consequences of DM are to be addressed.

Hype is a poor substitute for pragmatic scientific progress. As we go forward, all parties are cautioned to avoid language that feeds therapeutic misconception—mouse studies are purely experiments in a so-so disease model—the mice are neither treated nor cured; clinical trials also are experiments, not treatments, until the agent receives regulatory approval. Finally, recruitment of substantive funding and scientific expertise to further optimize and test the RNA and DNA targeting strategies for DM is essential.

References:

RNA editing with CRISPR-Cas13.
Cox DBT, Gootenberg JS, Abudayyeh OO, Franklin B, Kellner MJ, Joung J, Zhang F.
Science. 2017 Oct 25. pii: eaaq0180. doi: 10.1126/science.aaq0180. [Epub ahead of print]

RNA targeting with CRISPR-Cas13.
Abudayyeh OO, Gootenberg JS, Essletzbichler P, Han S, Joung J, Belanto JJ, Verdine V, Cox DBT, Kellner MJ, Regev A, Lander ES, Voytas DF, Ting AY, Zhang F.
Nature. 2017 Oct 12;550(7675):280-284. doi: 10.1038/nature24049.

Impeding Transcription of Expanded Microsatellite Repeats by Deactivated Cas9.
Pinto BS, Saxena T, Oliveira R, Méndez-Gómez HR, Cleary JD, Denes LT, McConnell O, Arboleda J, Xia G, Swanson MS, Wang ET.
Mol Cell. 2017 Oct 18. pii: S1097-2765(17)30711-6. doi: 10.1016/j.molcel.2017.09.033. [Epub ahead of print]

Understanding Heterogeneity in DM

Evaluation of large cohorts in the DM1-Scope registry, established the impact of the variables of age and gender in DM1 (De Antonio et al., 2016; Dougan et al., 2016). Men are more likely to experience ‘traditional’ muscular system-related DM1 symptoms (myotonia, muscle weakness, cardiac and respiratory involvement), while other symptoms are more likely in women (cataracts, dysphagia, digestive tract dysfunction, incontinence, thyroid disorder and obesity). Likewise, age of onset has clear implications for DM1 symptom symptom occurrence and chronology. By contrast, the impact of the variables of age and gender is less well understood for DM2, but remains a critically important question due to the heterogeneity of the disease.

A Retrospective Study Informs the Natural History of DM2

Prof. Benedikt Schoser (Ludwig-Maximilians-University of Munich) and colleagues have explored variables in disease symptomatology in a retrospective study of a 307-patient cohort with genetically confirmed DM2 from the German Registry for Myotonic Dystrophy.

Study subjects showed mean onset at 42 years of age; most had a single symptom at onset, with proximal weakness, myalgia and myotonia the most frequent symptoms (in order of descending frequency). Weakness at disease onset was associated with older age. New neurological examinations (at an mean age of 56 years), showed 77% of the cohort with mild to moderate weakness of at least one muscle group. In decreasing order of frequency, multisystem involvement included cataract, hyperlipidemia, hypertension, thyroid disorders, affective disorders, and cardiac disease (implanted ICDs/pacemakers were frequent). A previously reported pattern of altered laboratory findings was confirmed in this DM2 cohort: hyper-CKaemia, dyslipidaemia, elevated GGT, low IgG, and elevated ALT and AST.

Age at onset of DM2 and gender were significantly and independently associated with specific patterns of symptoms. Advancing age was associated with increasing number of organ systems involved, but decreasing odds that a subject would develop either myotonia or myalgia. Women had a higher total number of organ systems involved.

A high percentage (87%) of study subjects had myotonia (evaluated by emg), although no gender differences were detected. In contrast to findings in DM1, muscle weakness was more common and more severe in women at time of diagnosis, while pain/myalgia was most common in men. As in DM1, cataract and thyroid and gallbladder diseases were observed more frequently in women. Women with DM2 also more frequently needed waling aids.

Patients experiencing early onset of DM2 were more likely to have broader organ system involvement.

Age and Gender Impact the DM2 Phenotype

Taken together, the research team observed that age and gender have a profound impact on the DM2 phenotype. Both female gender and age appear to contribute toward a greater burden of disease in DM2. Analysis of ancestry of patients in the registry suggested that a DM2 founder mutation may have arisen in the Upper/Lower Silesia region of Poland, Germany and the Czech Republic.
 

Reference:

Assessing the influence of age and gender on the phenotype of myotonic dystrophy type 2.
Montagnese F, Mondello S, Wenninger S, Kress W, Schoser B.
J Neurol. 2017 Oct 30. doi: 10.1007/s00415-017-8653-2. [Epub ahead of print]

Control of RNA Gain-of-Function Versus Toxic RAN Protein Mechanisms

Since the discovery of Repeat-Associated Non-AUG (RAN) translation, the pathogenesis of any disorder that is the consequence of microsatellite expansions must consider two distinct mechanisms—RNA gain-of-function (i.e., toxicity of expanded repeat RNA), toxicity of RAN proteins, or a combination of both mechanisms. For myotonic dystrophy (DM), the toxic RNA mechanism has predominated thus far, as there are only limited reports of the detection of RAN proteins in affected tissues. How these two potential pathogenic pathways in DM are regulated, and whether they may be in any way co-regulated, are currently open questions.

Muscleblind (MBNL) Plays a Key Regulatory Role Over RAN Translation

Dr. Laura Ranum (University of Florida) and colleagues have recently published studies that provide new insights into the regulation of molecular pathogenic pathways in DM2. Through assessments of expanded sense (CCTG) and antisense (CAGG) transcripts in DM2 autopsy brains and in vitro, they demonstrated the occurrence of bidirectional transcription. The research team also showed that transcripts containing a threshold repeat length produce the corresponding poly(LPAC) and poly(QAGR) RAN proteins, with protein production positively correlated with repeat length.  Cytoplasmic poly(LPAC) protein was localized to gray matter (neurons and glia), while poly(QAGR) was found in white matter, primarily in oligodendrocyte nuclei (but also in pathologic regions containing activated microglia). Using in vitro expression studies, the research team showed that both RAN proteins are toxic in neurons by mechanisms unrelated to RNA gain-of-function.

The production of poly(LPAC) protein was reduced or blocked by nuclear foci formation and retention of CCUG expansion RNA by MBNL1 binding. By contrast, CAGG transcripts do not bind MBNL1 and form nuclear foci, but rather are translocated to the cytoplasm, resulting in elevated poly(QAGR) RAN protein (CLIP experiments showed that poly(QAGR) protein does bind hnRNP A1).

A Model for Regulation of RAN Translation in DM2

Taken together, nuclear MBNL1 levels control the relative degree to which mis-splicing and RAN translation products contribute toward the pathogenesis of DM2. If free MBNL1 levels remain sufficient, expanded CCUG transcripts are retained in the nucleus, CCUG-mediated RAN translation is blocked, and disruption of splicing may not reach a level where functional consequences are observed. Depletion of MBNL results in mis-splicing, translocation of CCUG toxic RNA to the cytoplasm, and production of the poly(LPAC) RAN protein and its corresponding pathogenic contributions to DM2. By contrast, the poly(QAGR) RAN protein is not regulated by MBNL1 and it plays a role at least in the brain in DM2. The data supporting linkage of RNA gain-of-function and RAN protein-mediated pathology, and the differential localization of sense and antisense strand products within the brain, provide new insights into understanding the mechanisms underlying the neurologic components of DM2. These findings also support an RNA sequestration failure model as the mechanism for the action of toxic RAN proteins in the brain of patients with DM2.

Reference:

RAN Translation Regulated by Muscleblind Proteins in Myotonic Dystrophy Type 2.
Zu T, Cleary JD, Liu Y, Bañez-Coronel M, Bubenik JL, Ayhan F, Ashizawa T, Xia G, Clark HB, Yachnis AT, Swanson MS, Ranum LPW.
Neuron. 2017 Sep 13;95(6):1292-1305.e5. doi: 10.1016/j.neuron.2017.08.039.

Gene Editing for DM

Gene editing has garnered considerable publicity as the newest technology with potential for developing therapies for rare diseases. MDF previously published a primer, titled "Using Gene Editing to Correct DM," on the CRISPR/Cas9 technology that has been heavily promoted in the media.

Gene editing technology uses molecular mechanisms that were first developed in bacteria as a shield against invasion from viruses. This approach is rapidly moving into clinical trials for a select group of diseases—those where cells can be isolated from the body, edited, and then returned to patients as a viable treatment for the disease. These diseases are predominantly disorders of the blood and cancers, and several clinical trials are recruiting patients in China (HIV-infected subjects with hematological malignances; CD19+ refractory leukemia/lymphoma; esophageal cancer; metastatic non-small cell lung cancer; EBV-associated malignancies). At least one trial has been approved in the U.S. by the Food and Drug Administration (FDA) and is expected to start soon (this is also for a set of cancers).

For myotonic dystrophy (DM), multiple organ systems are affected and we cannot take the simple path of editing and returning cells to the body—treatment must address simply too much body tissue mass, including the brain, the heart, skeletal muscles, the gastrointestinal system, and other organs that are affected. Thus, for CRISPR/Cas9 to “work” in DM, the gene editing reagents will have to be efficiently delivered to virtually every cell in patients and effectively execute the deletion of CTG and CCTG repeat expansions from the DNA. The delivery of gene editing reagents into patients is an incredibly difficult undertaking and is likely years away from clinical trials in any disease.

Could a Modified CRISPR Technology be Effective in DM?

Investigators at the University of California San Diego, the University of Florida, and the National University of Singapore have recently reported early research that potentially ‘repurposes’ gene editing technology for a set of RNA disorders—myotonic dystrophy type 1 (DM1), myotonic dystrophy type 2 (DM2), a subset of Lou Gehrig’s disease (ALS) patients and Huntington’s disease. They have modified the Cas9 enzyme so it is targeted to toxic RNA, instead of the expanded DNA repeats in these diseases.

The researchers have optimized Cas9 so that it can specifically target and degrade expanded repeat RNA for DMPK and CNBP genes. In many ways, this is similar to the approach that Ionis Pharma is using to target CUG repeats RNA in DM1. 

Their development of an RNA-targeted Cas9 results in the degradation of toxic RNA, an increase in the MBNL protein, and reduction or elimination of the gene splicing defect that characterizes DM. The strategy uses gene therapy vectors to delivery the modified Cas9 enzyme. If this approach were to be effective, it’s likely that patients would only need a single intravenous injection to treat skeletal muscles, the heart, and the gastrointestinal system; because gene therapy does not cross the blood brain barrier, a second injection may be needed, into the fluid around the spinal cord, to treat the brain. To work toward clinical development, the researchers have formed a biotechnology company to raise funding and move the candidate therapy forward.

We Still Have a Considerable Way to Go Before this Novel Strategy is in the Clinic

While this approach shows promise, we should be cautioned that studies thus far have only tried the new experimental therapy in patient cells in tissue culture. Therapy development has to pass through preclinical testing in appropriate mouse models, preclinical safety testing and approval by the FDA before the first clinical trial can be launched. Importantly, this effort represents yet another shot on goal to develop a novel therapeutic for DM1 and DM2. MDF monitors all drug development efforts and will keep the community informed as to their progress.

Model organisms have yielded important insights into neuromuscular diseases. Findings from the relatively straightforward models now link unstable expansions of CTG and CTTG repeats to the phenotypes of myotonic dystrophy type 1 (DM1) and myotonic dystrophy type 2 (DM2) respectively. Yet, it would be a mistake to assume that we understand all therapeutically relevant pathogenic or disease modifying mechanisms in DM. A particularly vexing issue has been how DM1 and DM2 are mediated by MBNL sequestration, but yield phenotypes of differing severity. New fly models may provide some insights.

Novel Models for DM1 and DM2

To address divergent aspects of pathology in DM1 and DM2, Dr. Rubén Artero and colleagues (University of Valencia) generated and evaluated novel Drosophila models expressing the respective repeats (250 CTG or 1,100 CCTG) in skeletal and cardiac muscle. Flies expressing 20 CTG or CCTG repeats were also generated and used as controls.

Similar, Severe Phenotypes Seen in DM1 and DM2 Fly Models

The investigators showed that the established molecular features of DM—formation of nuclear aggregates, MBNL depletion, RNA splicing defects and upregulation of autophagy genes (Atg4, Atg7, Atg8a, Atg9 and Atg12)—occurred in their DM1 and DM2 models. They establish that expanded CCUG repeat RNA has similar potential in vivo toxicity as does CUG repeat RNA. Both models had severe skeletal (50% reduction in fiber cross-sectional area) and cardiac muscle phenotypes, and reduced survival. Cardiac dysfunction included altered systolic and diastolic intervals, deficits in contractility (percentage (%) of fractional shortening) and arrhythmias; some cardiac measures showed higher severity in the DM2 model fly. 

Do Unknown Factors Mitigate Cardiac Disease in DM2?

While understanding that no model organism can actually be said to “have DM,” fly and mouse models have informed understanding and treatment of DM. In the DM2 fly model, the cardiac phenotype is more severe than is seen in DM2 patients. The investigators suggest that while both CUG and CCUG expanded repeat RNA have the potential to cause severe striated muscle phenotypes, there may be mechanisms beyond the well-established toxic RNA pathway that reduce the toxicity they observed in the fly in human DM2. These findings and models may have relevance for identification of genetic modifiers or as validation screens for small molecule drug development.

Reference:

Expanded CCUG Repeat RNA Expression in Drosophila Heart and Muscle Trigger Myotonic Dystrophy Type 1-like Phenotypes and Activate Autophagocytosis Genes.
Cerro-Herreros E, Chakraborty M, Pérez-Alonso M, Artero R, Llamusí B.
Sci Rep. 2017 Jun 6;7(1):2843. doi: 10.1038/s41598-017-02829-3.