SALT LAKE CITY—The “central dogma” of molecular biology says that DNA makes RNA makes protein, and in the classical picture of genetic diseases, gene mutations exert their effects by altering the amount of protein: either too little, as in the lysosomal storage disorders, or too much of the wrong type, as in Huntington’s disease. But new discoveries in myotonic dystrophy (DM) have brought to light an entirely new kind of genetic disease—one in which the problem is not protein at all, but rather the RNA itself. These findings not only explain the pathogenesis of DM but also suggest that curative treatment may be easier to develop than for protein-based genetic disorders. These developments were discussed by Charles A. Thornton, MD, who spoke at the 133rd Annual Meeting of the American Neurological Association.
The gene affected in DM type 1, called DMPK, normally carries a CTG trinucleotide repeat up to 35 units long. In patients with DM, however, the repeated section may be up to 5,000 units long. A similar mutation affects the gene for DM type 2, called ZNF9. The two proteins are entirely different in structure and function, which provided early evidence that the mutation was acting below the protein level.
The repeated section falls outside the coding region of the gene but is completely transcribed to form messenger RNA (mRNA) along with the protein-coding gene segments. But in DM, the repeated units (converted to CUGs during transcription) are longer than the entire rest of the transcript. Because of their great length and uniform sequence, the transcripts clump together, forming inclusions within the nucleus that can be seen under the microscope. “These are quite visible in each of the affected tissues,” said Dr. Thornton. He is Professor of Neurology at the University of Rochester Medical Center in New York.
Animal models of DM made by introducing the RNA repeat reproduced the inclusions, as well as some of the clinical features of the disease, including myotonia. This was compelling evidence that RNA was the true culprit, but it was not yet conclusive proof.
A “Breakthrough” Discovery
“A breakthrough in our understanding came when scientists isolated a set of nuclear proteins that bind to the repeated RNA with high affinity,” Dr. Thornton revealed. Called muscleblind proteins, they are the most abundant CUG-binding proteins in the nucleus. It appears that part of the normal function of muscleblind proteins is to link to RNA repeats, but in DM, they become trapped. The discovery of this sequestration of muscleblind proteins in the RNA inclusions has led to “a lot of excitement in the field,” he said. “We have a really excellent opportunity for therapeutic intervention.”
The muscleblind proteins control alternative splicing of RNA, the process in which certain segments (exons) of the transcript are either included or excluded in the final mRNA molecule. The result of normal alternative splicing is to control, in a tissue-specific pattern, the final set of proteins each cell produces. But in DM, inclusions reduce the amount of free muscleblind protein by 85% to 90%. The effect is to alter the expression of hundreds of other genes. Simply putting muscleblind back in to diseased cells can restore most of the changes in splicing. “There are clearly some transcripts not related to muscleblind function—it’s not the total story,” Dr. Thornton pointed out. But it is just as clearly the major player.
One gene whose splicing is controlled by muscleblind is the calcium reuptake pump found in muscle fibers. Alternative splicing creates two different forms that either include or exclude exon 22. In normal human muscle, exon 22 is included in almost all transcripts, but in DM muscle, it is partially or sometimes completely skipped.
In mice, simply knocking out the muscleblind protein mimics the effects of DM, with almost complete loss of exon 22 in the calcium pump. Restoring as little as half the normal amount of muscleblind normalizes splicing. Similarly, when the amount of high-repeat mRNA is reduced, more normal splicing can also occur. “The lesson is that the outcome of splicing is at least partly dependent on the mass of this toxic repeat RNA and the quantity of muscleblind,” Dr. Thornton said. “So long as the level of repeat RNA is low, there will be enough free muscleblind to execute splicing correctly. I would argue that this is the condition that prevails in myotonic dystrophy before the symptoms begin.”
Myotonia in diseases other than DM is most often caused by mutation of a specific chloride channel, called CLC1. In DM there is no mutation, but instead a large reduction in the number of channels per cell, due to improper splicing. “It seems reasonable to presume this is the cause of myotonia, but to test it formally, we needed to correct that defect, but not any other splicing defects.” Muscleblind normally excludes exon 7a in the muscle-specific channel; its inclusion causes transcription to abort before the protein can be completely made. So researchers used antisense RNA to block the exon in DM mice. The treatment “completely rescued the ion channel defect and nearly completely blocked myotonic discharges,” he said. “This was clear evidence that the splicing defect directly affects symptoms and led us to believe that the phenotype, even after it has fully developed in an animal model, could be completely reversed. But this is a piecemeal correction of a single splicing defect. Whether this would be worth pursuing in the clinic seems dubious. We should aim higher,” Dr. Thornton asserted.