Antisense therapy can reverse some aspects of myotonic dystrophy (DM): That's the conclusion from a new study in a mouse model, reported in the July 17 issue of Science.
Investigators reported that antisense molecules of DNA were injected directly into affected muscle, where they bound to the expanded RNA that is the cause of DM. That interaction prevented the RNA from trapping and inactivating another molecule, called muscleblind, a protein that has emerged as central to understanding the disease process of myotonic dystrophy.
Myotonic dystrophy is caused by an expansion of a nucleotide repeat in one of two genes, DMPK or ZNF9. The genes have little in common — DMPK is a protein kinase active in muscle, while ZNF9 is a transcription factor active in the nucleus. They are located on different chromosomes, and have few structural similarities. But each bears a repeated three- or four-nucleotide segment — CTG in DMPK, and CCTG in ZNF9. Expansion of that repeat, from a few dozen units in the normal gene to many hundreds or thousands of units in the mutated gene, causes disease — DM type 1 (from DMPK expansion) or DM type 2 (from ZNF9 expansion).
In both diseases, the expanded gene is transcribed to make an expanded messenger RNA. The expansion makes the mRNA “sticky,” and the mRNA molecules clump together in the cell nucleus, forming “foci” that trap RNA-binding proteins. One protein that is captured and bound with particularly high affinity is muscleblind, which regulates alternative splicing of RNA transcripts.
It has become clear that many features of DM might be explained by the dramatic reduction of active muscleblind — splicing of mRNA for the chloride channel, for instance, is regulated by muscleblind, and loss of muscleblind in otherwise normal mice recapitulates some aspects of DM, including loss of chloride channel expression and myotonia.
ANTISENSE INJECTION AMELIORATES PHENOTYPE
Because the interaction of muscleblind with the expanded RNA repeat contributes to DM, Charles A. Thornton, MD, wondered whether inhibiting their interaction could be therapeutic. Dr. Thornton, associate professor of neurology at the University of Rochester Medical Center in New York, has pioneered the investigation of the role of muscleblind in DM.
He turned to antisense morpholino molecules, nucleotide-like structures carrying the same four bases as RNA. The base sequence was designed to complement, and therefore bind to, the expanded RNA in a mouse model of DM1.
The transgenic mouse expresses an actin gene with a long CTG repeat, mimicking the mutation found in the human disease, though in a different gene. Like human cells in DM, mouse cells display RNA-rich clumps (“foci”) in the nucleus, and the mice develop myotonia.
Unlike DNA-or RNA-based antisense molecules, the chemical backbone of the morpholino does not trigger degradation by the cell, making it both longer lasting and safer in the event the molecule bound to non-expanded RNAs.
“The repeat sequence is present in other normal RNAs,” Dr. Thornton said, “and we didn't know the consequences of obliterating those transcripts. The morpholino is safer than other types of antisense molecules.”
Dr. Thornton's group injected the morpholino directly into muscle. They found that treatment led to a marked reduction in nuclear foci, and helped redistribute muscleblind away from the remaining foci and out into the rest of the nucleus. The splicing defects caused by muscleblind sequestration were normalized or nearly corrected by three weeks after a single injection. In particular, treatment rescued the improperly spliced chloride channel, restoring the channel to the muscle membrane and reducing myotonia. The effects remained strong at 14 weeks but had faded by eight months.
“Morpholino binding prevents muscleblind protein from binding to the expanded RNA,” Dr. Thornton said, allowing muscleblind to perform its job. There appeared to be a further benefit as well: “Once the proteins are stripped off or prevented from binding, the RNA begins to behave like RNA ought to behave,” he explained, moving out to the cytoplasm and coding for the DMPK protein. Loss of DMPK is thought to contribute to the cardiac symptoms in DM1.
The nuclear foci also trap other proteins, which the morpholino also presumably releases, and this may also contribute to the therapeutic effect. The role of these proteins in the DM phenotype are not as well characterized as that of muscleblind. “Muscleblind is certainly not the entire disease,” Dr. Thornton said.
“If this advances into the clinic, it would require periodic treatment. But we haven't gone in that direction with our thinking yet,” since so much more work is needed before then.
The next step, he said, is to develop morpholinos that can be distributed throughout the body by intravenous injection. “At that point, the treatment can be analyzed for safety and side effects.”
Those efforts are advancing in other diseases, including Duchenne muscular dystrophy, where systemic delivery in dogs has been reported. Regarding treatment of DM2, Dr. Thornton said: “We have no reason to suspect the same strategy would not work there.”
‘A LANDMARK DEVELOPMENT’
“This is a landmark development in showing the road forward in developing therapies for myotonic dystrophy,” said Stephen Tapscott, MD, PhD, professor of neurology at the University of Washington Fred Hutchinson Cancer Research Center in Seattle, who studies the role of normal repeats on genomic structure and function. “It's an important proof of principle paper, and based on excellent work by Dr. Thornton and others.”
“The next critical question would be therapeutic delivery,” he said. The work in the Duchenne dystrophic dog has shown it is feasible to deliver systemically [intravenously], “but whether it can be delivered in sufficient concentrations to be therapeutic is still unknown.”
One of the results from the paper was particularly interesting, Dr. Tapscott said. The movement of the expanded RNA out into the cytoplasm was tracked with a fluorescent protein that glowed brighter as more of the RNA was released from the nucleus. That suggests the system can be used for high-throughput screening for drug discovery. Drugs that trigger the same increase in fluorescence as the morpholino does would make good candidates for development of orally available drugs, obviating the need for morpholino injection, he said. “It's an open field with this result. It has overcome the barriers to rational drug discovery.”
• Wheeler TM, Sobczak K, Thornton CA, et al. Reversal of RNA dominance by displacement of protein sequestered on triplet repeat RNA Science 2009;325 (5938):336–339.