RNA Sequence Restrains Fatal Encephalitis

Short Interfering RNA Protects Mice Against Japanese Encephalitis Virus and West Nile Virus

One short sequence of RNA can be used as a broad-spectrum antiviral agent for treating deadly brain inflammation caused by different, but related, viruses—at least in mice, a new study shows.

In findings that underscore the therapeutic potential of the fast-moving field of RNA interference, researchers in the CBR Institute for Biomedical Research have demonstrated in mouse models that a single treatment is all that is needed for protection against fatal encephalitis caused by the Japanese encephalitis virus and West Nile virus.

Photo by Graham Ramsay

In mouse models of fatal encephalitis caused by two different related viruses, Priti Kumar (right), Manjunath Swamy, and Premlata Shankar (not pictured) showed that a single short interfering RNA worked as a broad-spectrum antiviral agent.

The two viruses are among the major emerging and resurging mosquito-borne pathogens in the flavivirus family, which also includes its namesake viral cousin yellow fever (flavus is Latin for yellow), dengue, and St. Louis encephalitis.

Prevalent in Southeast Asia, Japanese encephalitis virus causes about 50,000 cases of encephalitis annually, killing about one third of the afflicted people and permanently damaging the brains of one half of the survivors. Once confined to Africa and the Middle East, West Nile virus came to the United States in 1999. From 66 cases in New York, the virus has spread to nearly 10,000 reported infections in 45 states, with more than 2,800 cases of brain inflammation. There are no effective drugs to treat these infections.

Clinical Trials
It has only been four years since scientists first showed that RNA interference—which protects plants, flies, and worms from viral infections—also works in mammalian cells. Three years ago, Judy Lieberman’s team at the CBR first showed in mice that short interfering RNA (siRNA) could be used therapeutically (see Focus, Feb. 21, 2003). Now, at least two experimental siRNA therapies already have advanced to phase I safety trials in people for age-related macular degeneration and respiratory syncytial virus (RSV).

“We wanted to find a single sequence that can protect against a wide range of viruses, so you don’t have to wait for a lab diagnosis,” said Priti Kumar, a postdoctoral fellow and first author of the paper in the April PLoS Medicine, which was published online Feb. 14. “It’s difficult to distinguish which virus is causing encephalitis during an epidemic. Tests for the virus take a long time to get results, especially in developing countries like India.”

Working under the supervision of CBR investigators Premlata Shankar and Manjunath Swamy, senior authors on the paper and HMS assistant professors of pediatrics at the CBR and Children’s Hospital Boston, Kumar took over the project from another postdoctoral fellow, Sang Kyung Lee, who had started the initial studies with the candidate siRNAs in cell lines. (He is now on the bioengineering faculty at Hanyang University in Seoul, Korea.)

Their search for an siRNA target had begun with a screen of the viral genomes. They looked for any conserved sequences the flaviviruses still shared despite the fast-evolving nature of the viruses. They concentrated on West Nile and Japanese encephalitis viruses, for which there are good mouse models. To further narrow the list of candidates, they used another set of simple rules that have emerged from RNA interference research.

Matchmaking
An ideal siRNA is 19 to 21 nucleotides long, Shankar said. Any longer and it can trigger a general interferon response in mammalian cells that shuts down protein synthesis. Except for some leeway on the end nucleotides, the sequence must be nearly an exact complementary match for the target RNA in the viral genome. It also must be distinct enough from all other sequences in the mammalian genome that the siRNA does not cause unintended consequences elsewhere.

Short interfering RNA starts as a double-stranded RNA. No more than half of the middle nucleotides of the short RNA can be guanine–cytosine pairs, which can make the short double-stranded RNA too sticky to separate into its sense and antisense strand. And the crucial antisense strand must start with adenines and thymidines, which make it easy to peel away, Swamy said.

The researchers found five candidates. The results were dramatic from the start with one promising siRNA that targeted a gene coding for the envelope protein. In cell lines, the siRNA made by lentiviruses provided complete protection against Japanese encephalitis virus.

By coincidence, other HMS researchers had discovered the function of this sequence through structural studies. “The peptide segment encoded by the RNA used in this work is the so-called ‘fusion loop,’ which inserts into the membrane of a target cell during the process of viral entry,” said Stephen Harrison, HMS professor of biological chemistry and molecular pharmacology, whose group solved the structures of domains in comparable proteins in the tick-borne encephalitis virus in 1995 and in the dengue virus three years ago (see Focus, June 6, 2003). “Viruses that mutate in this region, thus escaping from interference, have a reduced likelihood of being infective, [but] the gymnastics of entry aren’t the issue—only the conservation of the peptide.”

If siRNA acts in its usual manner, then in this case the antisense strand gloms onto the viral genome, which attracts a molecular complex outfitted with a scissors enzyme. The enzyme cuts the RNA genome, destroying it.

“The next step was to take it into an animal model to see if it worked,” Kumar said. In mice, Japanese encephalitis virus is so virulent that one particle injected into the brain leads to certain death within four or five days. Here, too, the results were dramatic. At up to 50 times the fatal dose of the virus, brains first inoculated with the lentiviruses that made the siRNA were 100 percent protected. The siRNA lost its therapeutic power at the unnatural extreme of 1,000 times the lethal dose.

Lentivirus is not approved for human therapy, so the researchers encased the synthetic siRNA within a lipid known to enter neurons. This time, Kumar tried postinfection injections of the siRNA up to 18 hours after infection by Japanese encephalitis virus. This period is the time it takes the virus to multiply within neurons to the point of bursting the cells and spreading viral progeny to other brain cells. Up to 60 percent of the treated mice survived the infection.

Double Duty
Unfortunately, the same siRNA did not work very well in cell lines or in mice challenged with West Nile virus. So Swamy went back to the database and identified a neighboring stretch of short RNA in the same envelope protein with almost fully matched sequences across the two viruses. The researchers engineered a new piece of siRNA. This sequence provided cross-protection at the same high level against both Japanese encephalitis virus and West Nile virus. The researchers believe the same siRNA may also be effective against St. Louis encephalitis, which contains the same conserved sequence.

The big catch now is finding a feasible way to deliver the siRNA across the whole brain, Swamy and Shankar said. In this study, the siRNA only worked in the localized infected cells near the injection site and could not travel through the brain as the viruses could.

They also need to find another delivery system. In the clinical trials of age-related macular degeneration, siRNA is injected into eyes, said Lieberman, who was not involved in this study, but intracerebral injection to deliver siRNAs targeting the flaviviruses is not likely to be used clinically. (The clinical trials of RSV deliver siRNA by aerosol and take advantage of the propensity of siRNA to hunker down in mucosal tissue, she said.)

Research rarely proceeds, as this project did, in the logical progression that is described in the scientific paper. Except for the brief hiccup in identifying a cross-protective siRNA, the only big hitch in the story happened when the money came in from a bioterrorism grant to expand the level-3 safety facility.

“It was right after the first experiment when all the mice survived,” Kumar said. “We were excited, but we had to shut down the research for four or five months and then start over.”

—Carol Cruzan Morton

top