Canadian biologists showed that neurons afflicted with Alzheimer's disease develop disruptive G-quadruplex structures in their DNA — and the researchers found a gene that prevents that.

A paper published March 23 in Nature Communications details the newly revealed function of the BMI1 gene, which is known to inhibit signs of aging in the brain and is now an even more promising target for a potential treatment of the prevalent neurodegenerative disease.

The researchers also discovered where G-quadruplex structures form along DNA and how they disrupt DNA transcription, which could have broader implications for other brain diseases that are also accompanied by the structures.

Despite significant research into Alzheimer's disease, no genetic mutations have been found to directly cause the condition, according to Gilbert Bernier, a research director at Montreal's Hôpital Maisonneuve-Rosemont and the paper's lead author. For more than a decade, his lab has instead investigated the connection between Alzheimer's and the expression of genes, particularly those associated with aging.

In 2009, the lab found that when the BMI1 gene was silenced in neurons of mice, their brains and eyes showed signs of rapid aging. The researchers concluded that the BMI1 protein encoded by the gene stops neurons from exhibiting symptoms of aging or prematurely dying. Their follow-up research revealed low counts of BMI1 in the brains of nonfamilial Alzheimer's disease patients and showed that lacking one BMI1 allele led to DNA damage in repetitive genetic sequences in mice.

The latest study expands the known protective role of BMI1, which regulates the gene silencing and the packaging of DNA into chromatin structures. It focused on G-quadruplex structures, which are composed mostly of repetitive sequences of guanine — one of the four bases that spell out DNA, in addition to the "A," "C" and "T" bases — and are a departure from DNA's usual double-stranded helix shape. Also called G4 structures, they are known to create genomic instability and disrupt DNA replication and transcription.

Analyses of brains and stem cell-derived neurons from Alzheimer's patients showed numerous G4 structures, which had few BMI1 proteins near them. In contrast, healthy neurons with regular BMI1 expression had significantly fewer G4 structures, implying that the gene has a preventive effect when expressed normally.

"In the context of Alzheimer's disease, it's completely new," said Bernier, who is also a neuroscience professor at the University of Montreal. "It has never been described before, that in Alzheimer's disease patient brain, you have these G4 popping up all over in neurons."

The creation of G4 structures appears to be the result of decompression of heterochromatin, condensed and often repetitive DNA that limits DNA expression and is regulated, in part, by BMI1. The researchers also found that G4 structures alter gene expression and can be drastically reduced in Alzheimer's disease neurons by more tightly compacting heterochromatin.

Bernier and his co-authors also discovered that G4 structures essentially form only in actively transcribed LINE1s, a category of repetitive genetic sequences that is largely noncoding or "junk" DNA and make up 17% of the human genome. Also known as L1s, these sequences are mostly inactive, while the few active segments can disrupt DNA elsewhere and are known to cause some diseases.

The researchers learned that G4 structures are created only when active L1s are read by the DNA-transcribing protein RNA polymerase II. This was demonstrated by inhibiting the transcription protein, which reversed G4 structure production.

"That's a very fundamental discovery," Bernier said. "It's going to be in a textbook!"

In total, G4 structures seem to be forming in Alzheimer's disease neurons because the cells insufficiently express BMI1, which can prevent the disruptive transcription of L1s by keeping them tightly packed in heterochromatin, according to Bernier.

He said his team's findings raise the question of how BMI1 expression is reduced on the molecular level and whether it occurs as individuals age. The answer could eventually lead to a way to reactivate expression of the gene, he said, and that could be a treatment for Alzheimer's disease if reduced BMI1 expression turns out to be the origin of the disease.

"You find a way to increase the gene expression back to normal, in principle, well, you just cure Alzheimer's disease," Bernier said.

The study, "G-quadruplexes originating from evolutionary conserved L1 elements interfere with neuronal gene expression in Alzheimer's disease," published March 23 in Nature Communications, was authored by Roy Hanna and Andrea Barabino, Hôpital Maisonneuve-Rosemont; Anthony Flamier, Whitehead Institute of Biomedical Research; and Gilbert Bernier, Hôpital Maisonneuve-Rosemont and University of Montreal.

Correction: A previous version of this story incorrectly described researchers' discovery of how G-quadruplex structures disrupt DNA. The error has been corrected.