In 2009, pharma giant GlaxoSmithKline published an article in Antiviral Research describing a promising new drug its scientists had been investigating. The drug, called GSK983, was a broad-spectrum antiviral—a drug that could fight a variety of different viruses—that seemed to be effective against HPV, mononucleosis and more. The paper described the compound’s synthesis and effects and went on to conclude that it warranted further study. But strangely, according to the study, researchers had little idea how the compound worked.
The pharma giant put a lot of resources into the drug; a corresponding article shows synthesis on the scale of kilograms, and some animal trials were conducted. Then, the company quietly ceased its experiments. GSK983 had been abandoned.
Years passed, but the drug was not forgotten. When no subsequent articles came out, a group of scientists at Stanford decided to tackle the problem themselves. “It was interesting that there was a good antiviral that industry sort of left alone, probably because they could not explain the mode of action of this drug,” says Jan Carette, who runs a virology lab at the Stanford School of Medicine. Carette collaborated with colleagues from the genetics and chemistry departments on a study, published in Nature Chemical Biology in March, which examined GSK983’s mechanism and addressed some of its problems.
Thanks to several new techniques, GSK983 may have a future after all—one that could help doctors combat emerging diseases like Zika without having to go through as much FDA red tape. But GSK983 is just one drug, applicable only to certain classes of viruses. It could be great, or it could merely be one in a line of compounds in the search for broad-spectrum antivirals—and a program of dual genetic screening pioneered in this study could be a potent tool that'll speed up the whole process.
If you have a bacterial infection, you go to the doctor, who prescribes an antibiotic. Some are more effective than others, and some are better suited to particular infections, but in general, if you throw an antibiotic at a bacterium, it’ll clear up the infection. Not so with viruses, most of which require their own targeted drugs or vaccines. The process to develop such treatments can stretch a decade or more, by which time the virus has often evolved and changed.
This is why a broad-spectrum antiviral could be so powerful. Having one drug (or a small number of drugs) that are applicable in emerging epidemics like Zika, as well as rare diseases that don’t attract enough attention to warrant specific drugs, would be hugely important to both pharmaceutical companies and public health organizations, speeding up global epidemic response and saving lives.
But typically, antiviral development a painfully slow process. Unlike bacteria, which are susceptible to general antibiotics, it’s a challenge to make compounds that will target multiple viruses because the way viruses replicate are so varied, and because they’re active within the host’s cells, explains Johan Nyets, a professor of virology at the University of Leuven, Belgium who has been advocating broad-spectrum research for decades.
The pace of drug development can be key in minimizing the scale of an outbreak. “If a new pathogen is emerging, as was the case with Zika, and you have to start developing drugs at the time that this novel pathogen emerges, you’re too late because it takes an average 8-10 years before you have a compound developed in the lab to clinical use,” says Nyets. As Congress debates how (and how much) to fund Zika research, we fall farther and farther behind.
GSK983 targets a class of viruses that hijack a host cell’s RNA and uses that replication mechanism to make more viruses. Disrupting that process (a technique known as host targeting) is one way to attack an infection, but because the enzymes the virus uses to hijack the host cell are important for the host itself, side effects often include killing or stunting the very cells we’re trying to protect.
The Stanford crew suspected that this may be what was holding GSK983 back. In the original paper, the authors mentioned that host cells would sometimes die or cease to multiply when the drug was administered. “The challenge is to separate the antiviral and growth inhibition effects,” wrote the authors. GlaxoSmithKline has confirmed that the drug never progressed to human trials due to toxicity.
“We really have no idea what GSK’s plans were for this drug, what their actual findings are, internally,” says Michael Bassik, an assistant professor whose lab ran genetic screens for the Stanford study. Bassik’s needed to discover exactly what genes the drug targeted, so that they could figure out what was killing the cells. To do this, he employed an entirely new technique—or, really, two techniques in parallel: CRISPR and RNA interference.
CRISPR is the latest gene editing technology du jour, using a protein to splice, or in this case, cut out genetic information. It’s not quite so simple as toggling a switch, but the process effectively turns off the genes one at a time, to see which change the behavior of the drug.
RNA interference, on the other hand, introduces a piece of RNA data which, when transcribed, suppresses gene action, rather than shutting it off completely. Because this modifies the genes’ function, rather than destroying them, they maintain some of their actions. Thus, the technique generates data on essential genes that, were they knocked out completely, would kill the cell.
Each technique finds a different set of genes; by cross-referencing them, the Stanford team was able to isolate probable targets—that is, the genes (and the enzymes they produce) that the drug affects.
“The point of this paper is to say, you do get, by doing these two strategies in parallel, a much more comprehensive picture of the biology of the system, and in this case, the biology of the way this particular drug works,” says Bassik.
What it showed was this: GSK983 works as an interferon—it blocks an enzyme called DHODH that is used in replication. (This was, in fact, GlaxoSmithKline’s guess as well.) Without that enzyme, neither the RNA-based virus nor the DNA-based cell can replicate. This insight gives researchers a better understanding of how to leverage the compound to fight these kinds of viruses without killing the cells they’re trying to save.
This still leaves the problem of toxicity. But by knowing what enzyme was blocked, the Stanford team was able to restore only the DNA replication by adding a compound called deoxycytidine, thus reversing toxicity but not antiviral activity. They demonstrated its efficacy with dengue, says Carette, and next steps include testing it on Zika.
This was only tested in vitro in the study, points out Bassik, and in vivo tests are in progress. It does suggest future potential for GSK983, but perhaps more importantly, it shows that the dual CRISPR/RNA screen could be useful against one of the major drug-discovery stumbling blocks. “You have a series of molecules, you don’t know what their target is,” says Bassik. “[If] we can come in with this technology and identify the actual target, it really should facilitate the development of those drugs.”
GlaxoSmithKline, for its part, is listening. “The renewed interest has motivated us to look again at how we can publish those data and make the information available to the scientific community,” says spokesperson Kathleen Cuca.