Everything’s good for something. Including, it turns out, 5’ untranslated trinucleotide repeats.
In the Feb. 17, 2020, issue of Nature Neuroscience, researchers have demonstrated a role for such repeats in controlling protein levels of fragile X mental retardation protein (FMRP).
In the biomedical community, such repeats are best known for causing dozens of diseases once they repeat too many times. Huntington’s disease; C9ORF72, the most common genetic form of amyotrophic lateral sclerosis (ALS); and myotonic dystrophy are all united, along with dozens of others, by the presence of repeat expansions.
Even healthy individuals, however, have trinucleotide repeats – they just have fewer repeats. Huntington’s disease, for example, results from more than 35 repeats of the trinucleotide CAG. But normal individuals, too, have at least half a dozen CAGs at the same spot.
Now, a team led by University of Michigan researchers has demonstrated that for FMRP, those normal-length CAG repeats played a role in regulating protein levels locally at synapses.
Fragile X syndrome, which was the first repeat expansion disease to be discovered, is the most common monogenic cause of autism.
But the consequences of expansion differ depending on the number of repeats in the FMR1 gene, which codes for FMRP. Most individuals have about 30 repeats of the trinucleotide CGG in their FMR1 gene. Persons with fragile X syndrome have 200 or more repeats, while those with 55 to 200 CGG repeats suffer from fragile X-associated tremor/ataxia syndrome (FXTAS).
Previous work has shown that “repetitive elements in the genome can trigger protein translation in the absence of [an AUG] start codon,” corresponding author Peter Todd told BioWorld.
Todd and his team were looking at that form of translation, called RAN (for “repeat-associated non-AUG initiated”) translation in cell culture. As one of their control groups, they shortened the CGG repeats in FXTAS neurons.
“We thought that when we made [the repeats] the ‘normal’ size, this funny translation would go away,” Todd said.
Their failed control group prompted Todd, who is associate professor of neurology at the University of Michigan Medical School, and his colleagues to look at possible functions of RAN-associated translation.
They showed that under normal circumstances, RAN-associated translation resulted in lower FMRP levels, most likely by keeping ribosomes busy with translation in the upstream region rather than FMRP itself.
Activation of the metabotropic glutamate receptor mGluR5 suppressed RAN translation and led to rapid increases of FMRP levels in synapses.
“This ability to rapidly make that protein out in the synapse is, we think, important to synaptic function,” Todd said. MGluR5 plays a role in synaptic plasticity, which is the cellular underpinning of learning and information storage.
Additionally, when they blocked RAN translation with an antisense oligonucleotide (ASO), FXTAS neurons survived longer than their untreated brethren.
Todd said that the work is at the “tissue culture and cells in a dish” stage. The team is hoping to expand into animal models to see whether targeting RAN translation could ultimately be a therapeutic approach to fragile X-associated disorders.
Several co-authors on the paper are from Ionis Pharmaceuticals Inc., which collaborated on the ASO aspect of the work now published in Nature Neuroscience.
Ionis is in clinical trials with ASOs targeting several nucleotide repeat expansion (NRE) diseases, including Huntington’s disease and ALS. In those diseases, however, getting rid of the offending protein is a promising strategy, while getting rid of FMRP leads to fragile X syndrome. The ASO “effectively fixed two problems simultaneously,” Todd said.
It prevented translation of the expanded repeats, which themselves are toxic and can kill neurons.
But FMRP is an RNA-binding protein, and its loss in fragile X syndrome and FXTAS also leads, essentially, to supply chain problems as its binding partners can no longer be transported.
By targeting RAN translation rather than the protein itself, FMRP synthesis in synapses was increased.
The work could lead to new strategies for the treatment of NRE disorders, which are mostly neurological, though they are often complex, with mixed neurological and non-neurological symptoms.
Expansions, in that case under the name microsatellite instability (MSI), also occur in a fraction of cancers across multiple anatomical locations – Keytruda (pembrolizumab, Merck & Co. Inc.) received the first tumor-agnostic approval for MSI-hi tumors. But the underlying cause of MSI-hi tumors is a DNA repair deficiency, and the expansions themselves are biomarkers for that deficiency, rather than disease drivers.
It is likely, however, that there are many more NRE-driven diseases waiting to be discovered.
“We keep discovering new repeat diseases all the time,” Todd said.
The work also suggests that much remains to be learned about the functions of short repeating sequences in the absence of disease.
Because of the way genomes are sequenced in relatively short stretches, and then assembled via overlapping the ends of those shorter pieces, “we don’t see them in whole genome data,” Todd said. “We just ignore that piece of the genome in our analysis, because it’s too hard to look at.
“Three percent of our genomes,” he said, “is made up of exactly this kind of repeat, and we basically know zero about what they do.”