Delivering anything therapeutic to the brain has long been a challenge, largely due to the blood-brain barrier, a layer of cells that separates the vessels that supply the brain with blood from the brain itself. Now, in a study published August 12 in Nature Biotechnology, researchers have found that double-stranded RNA-DNA duplexes with attached cholesterol can enter the brains of both mice and rats and change the levels of targeted proteins. The results suggest a possible route to developing drugs that could target the genes implicated in disorders such as muscular dystrophy and amyotrophic lateral sclerosis (ALS).
“It’s really exciting to have a study that’s focused on delivery to the central nervous system” with antisense oligonucleotides given systemically, says Michelle Hastings, who investigates genetic disease at the Rosalind Franklin University of Medicine and Science in Chicago and was not involved in the study. The authors “showed that it works for multiple targets, some clinically relevant.”
In 2015, Takanori Yokota of Tokyo Medical and Dental University and colleagues published a study showing that a so-called heteroduplex oligonucleotide (HDO)—consisting of a short chain of both DNA and an oligonucleotide with modified bases paired with complementary RNA bound to a lipid on one end—was successful at decreasing target mRNA expression in the liver. Yokota’s team later joined forces with researchers at Ionis Pharmaceuticals to determine whether HDOs could cross the blood-brain barrier and target mRNA in the central nervous system.
In the new study, the research team showed that an HDO designed to target a tumor-associated long noncoding RNA (Malat1) achieved a higher level of knockdown in the brain and spinal cord when tagged with cholesterol and injected intravenously than when the HDO was tagged with the lipid tocopherol and injecting subcutaneously. In both rats and mice, the effects of the knockdown were tied to the dose of the HDO and four doses—given a week apart—were most effective. An intravenously injected single-stranded oligonucleotide tied to cholesterol didn’t knock down Malat1.
“Duplexed oligos actually seem to be delivered much better than single-stranded oligos,” coauthor Frank Rigo, vice president of functional genomics and drug discovery at Ionis Pharmaceuticals, tells The Scientist. “The data seems to suggest that the single-stranded oligos may get trapped . . . and may not actually cross the vasculature as efficiently as the double strand,” he explains.
After their initial results with Malat1, the researchers generated HDOs for three clinically relevant genes: DMPK, which when mutated is linked to a type of muscular dystrophy; glial fibrillary acidic protein (Gfap), which can cause Alexander disease; and human superoxide dismutase 1 (SOD1), which, when mutated and engineered into mice, models amyotrophic lateral sclerosis. They found knockdowns of between about 20 and 60 percent of DMPK mRNA, depending on which central nervous system tissue they analyzed. DMPK protein levels decreased by about half in both the brain and muscles of mice that received the HDO intravenously. The targets of the other two HDOs demonstrated more modest knockdowns.
“The data seem strong,” Rudy Juliano, an emeritus professor at the University of North Carolina School of Medicine who did not participate in the study, writes in an email to The Scientist. Open questions, he adds, include whether most of the effect observed in the whole brain actually took place in neurons and not in supporting tissues, if the high doses the authors used—50 milligrams of HDO per kilogram of body weight—can be toxic, and why “this simple modification of oligo structure can overcome the extremely robust blood-brain barrier that limits the access of much smaller and much less polar molecules to the brain.”
“The first thing that struck me was the enormous dose that they are using. Fifty milligrams per kilogram is very high,” agrees David Male, a cell biologist at The Open University in the UK who was not involved in the work. It’s important to consider the dosing for clinical applications, he adds. “You might want to knock down inflammation in the brain in say, multiple sclerosis. You’re probably going to have to be giving quite sizable doses on quite a long-term basis if you’re just giving an oligonucleotide, which doesn’t effectively regenerate itself.”
In the study, the authors acknowledge the limitations of the current work, and Rigo says plans for optimization are already in place. “Theoretically, you would want to identify how those lipids are getting into the brain. What receptors and what systems are they using?” he asks. “We would want to expand the work and, if it’s using one or two receptors, find out which ones they are, and then harness those in a more specific way.”