Pharmacodynamic Biomarker of Myotonic Dystrophy - ABSTRACT. Myotonic dystrophy type 1 (DM1) and type 2 (DM2) are dominantly-inherited degenerative
disorders that cause muscle weakness and multi-systemic symptoms. Both disorders involve RNA toxicity, a
novel gain-of-function mechanism, triggered by expression of RNAs that have thousands of CUG- or CCUG-
repeats (DM1 or DM2, respectively). Key proteins that regulate RNA alternative splicing, that have high affinity
for CUG- and CCUG-repeats, become trapped on repetitive RNA, causing mis-regulated splicing of transcripts
encoding critical components of skeletal muscle. In mouse models we have used a variety of approaches,
including oligonucleotides, gene therapy vectors, and small molecules, to show that reduction of RNA toxicity
causes normalization of RNA splicing and muscle function, suggesting that DM is at least partly reversible. We
also reported that targeted RNAseq analysis of misregulated splice events is reliable and sensitive for
monitoring therapeutic impact in mice, providing a near-real-time assessment of target engagement. These
developments, and prospects for conducting biomarker-driven trials, have spurred many drug development
programs for DM, including some that are now advancing to IND-enabling studies or early-phase trials. To
prepare for these studies we developed human targeted RNAseq similar to that employed in mouse studies.
Building on comprehensive discovery by deep RNAseq of more than 90 muscle samples, we identified
transcripts that show splicing dysregulation, and then developed targeted RNAseq to evaluate 22 splice events
that show large effects in DM1. We then determined a composite splicing index for each sample. The splicing
index stratifies the population in a large cohort of DM1 patients, correlates with the extent of muscle weakness,
shows acceptable test-retest agreement of splicing defects over 2-3 months, and provides a serviceable
pharmacodynamic biomarker. However, as the field advances and we gain more experience, important
limitations have emerged, particularly in the method for obtaining muscle samples and the selection of splice
events for targeted RNAseq – two aspects that are closely linked. Here we seek to optimize and show feasibility
for a new sampling method, called myoaspiration, that enables multi-sampling. i.e. sampling of more muscles
on more occasions. In parallel, we seek to optimize and validate a new targeted RNAseq assay, applicable to
both DM1 and DM2, designed to accommodate myoaspirate samples, and providing greater precision and
stronger inferences about target engagement in muscle fibers. Performance of the assay will be initially
evaluated using previously collected samples in our BioBank, and then proceed to prospective evaluation of
newly-collected myoaspirate samples from DM1 and DM2 patients, focusing on the muscles that are selectively
affected or relatively spared. Upon completion, the project is expected to provide a simple yet effective
pharmacodynamic biomarker for therapeutic trials leading to drug approval, and for post-approval studies to
compare the effectiveness of different agents, regimens, and combinations.