Inner Workings: Zebrafish assay forges new approach to drug discovery

Researchers at a University of Washington lab in Seattle are using standard 96-well plates to harbor something other than the usual layer of cultured cells. Instead, newly hatched zebrafish, smaller than a grain of rice, swim in each well. Running from the fish’s head to tail are clumps of hair cells called neuromasts, all of which are susceptible to the same drugs that affect mammalian inner-ear hair cells. That homology could offer crucial insights. Researchers hope these cells and these tiny hatchlings hold the key to preventing an insidious and tragic form of deafness spurred by otherwise beneficial antibiotics.

Five-day-old zebrafish larva populate the wells of a 96-well plate. These tiny hatchlings could hold the key to preventing an insidious and tragic form of deafness. Image courtesy of Edwin Rubel (University of Washington, Seattle).

Since the mid-2000s, three labs at the University of Washington and Fred Hutchinson Cancer Research Center have been using zebrafish assays to identify chemicals with the potential to protect essential auditory cells from toxic side effects—the ototoxicity—of certain antibiotics and chemotherapy treatments. University of Washington neurodevelopmental geneticist David Raible, the group’s zebrafish expert, identified a dye that, when absorbed by living neuromasts, makes them light up under a fluorescent microscope; those cells killed by the antibiotics go dark. By scoring the number of bright cells in each neuromast, researchers can quickly identify biologically active compounds or phenotype-changing mutations.

Between 2005 and 2006, the researchers tested 11,000 compounds from a small molecule library and discovered a promising molecule that blocked ototoxicity (1). More recently, the group used the zebrafish assay to test more than 400 chemical modifications to the compound to optimize it to be a more effective drug (2). At the same time, the group has been using the zebrafish to reveal the genes and mechanisms by which ototoxic drugs kill neuromast cells. After some careful fishing, the researchers are starting to catch some promising drug leads.

Neuromasts, used by the zebrafish to sense changes in water currents, light up green on this 5-day-old transgenic fish. Reprinted with permission from ref. 6.

Age of Aquaria

Widespread usage of zebrafish assays in the last decade has uncovered genes, cellular mechanisms, and compounds involved in human cardiovascular, neurological, and other biological processes. “They’re a living organism, with complex organs and physiological systems that can’t be modeled well in cultured cells,” says Randall Peterson, a professor of chemical biology at the University of Utah, Salt Lake City, and one of the first people to use zebrafish to screen small molecules.

With zebrafish, he says, you can perform screens on a relatively large scale with small amounts of compound. The fish can absorb the chemicals directly from the water, and their transparency means there's no need for dissections. An equivalent study in mammals would be prohibitively expensive. In Raible’s initial screen, the zebrafish neuromasts responded to the administered drugs within an hour, allowing a single researcher to complete the screen in a couple months.

Zebrafish have their own inner-ear hair cells for hearing and use the neuromasts purely to sense fluctuations in the surrounding water. But at a cellular level, the neuromast hair cells and mammalian inner-ear hair cells function similarly, explains Raible. Both cell types have exterior appendages that respond to movements in the fluids around them, and both transmit the information via electrical signals to the brain. The neuromast cells respond directly to water currents, helping the fish navigate up and downstream, find prey, and avoid predators, whereas the inner-ear hair cells detect the subtle disturbances in liquid vibrations caused by sound waves, allowing animals to hear. Genetics screens starting in the late 1980s revealed highly conserved genes between these two cell types (3).

But there is one striking difference. “Virtually all hair cells in all vertebrate species, except mammals, regenerate quite readily,” says neuroscientist Edwin Rubel, Raible’s collaborator at the University of Washington. According to Rubel, no one knows yet why mammals lost this ability. But without this ability to regenerate, age deterioration, noise damage, and ototoxic chemicals can all result in permanent hearing loss and deafness.

When those ototoxic chemicals are also life-saving drugs, such as the anticancer drug cisplatin or powerful broad-spectrum aminoglycoside antibiotics, doctors and their patients must balance the risk of hearing loss against the effectiveness of the drugs. “At this time, there isn’t a reliable drug on the market that protects from any form of hearing loss,” says Rubel. With no drugs to counter ototoxicity, aminoglycosides are used only as a last resort against antibiotic-resistant bacteria or for life-threating chronic lung infections that often befall children and young adults who have cystic fibrosis (4). Aminoglycosides, which disrupt bacterial ribosomal protein synthesis, target inner-ear cells by a mechanism that isn’t completely understood, and their effect is not consistent between individuals. Some people will exhibit extreme loss of hearing, Rubel notes. Others taking the same doses of antibiotic “will show no hearing loss at all,” he says.

Suspecting there was an underlying genetic component, Rubel spent several years considering how he could get at this problem. Inner-ear hair-cell death is a major component of acquired hearing loss, and the zebrafish’s neuromast sensitivity to ototoxic drugs seemed like a good system to study what might be causing those cells’ demise.

Casting a Wide Net

In 2001, Rubel approached Raible, then a new hire with zebrafish expertise, in hopes of setting up genetic screens that identified mutants with increased or decreased sensitivity to aminoglycoside antibiotics. After identifying the vital dyes that lit up living neuromast cells in zebrafish, they discovered the cells were reliably sensitive to the antibiotics’ concentrations. By varying the amount of antibiotic added to the well, they could consistently kill all, some, or only a few cells in each neuromast. This consistency suggested the zebrafish assay would be rigorous enough to serve as a platform for phenotypic drug discovery.

Rubel and Raible then teamed up with Julian Simon, a molecular pharmacologist at Fred Hutchinson Cancer Research Center. By exposing the fish to 5 to 10 small molecules at a time, Raible and Simon were quickly able to identify a molecule that countered the ototoxic effects of one drug, neomycin (1). They didn’t know anything about the biology of PROTO-1—their name for the discovered compound—except that it kept hair cells alive, both in zebrafish and in a time-intensive follow-up study in rats. The latter required a baseline hearing test, 2 weeks of exposure to the drugs, followed by two additional weeks for a full cellular response, and finally a second test to determine if there was hearing loss.

But when researchers ran PROTO-1 through a standard toxicology screen, they found an unfortunate side effect: it inhibited the human ether-a-go-go–related gene (hERG) potassium ion channel protein, which is critical to functioning cardiac muscles. Although the rats seemed to tolerate PROTO-1, says Simon, such activity would raise red flags in any drug-approval process. A further issue was PROTO-1’s short half-life and low solubility, making it unsuitable to be packaged in a pill that might sit on a shelf for months.

The researchers needed an alternative, so it was back to the zebrafish platform for testing variants of PROTO-1. Simon’s lab synthesized around 400 alterations and identified the concentration at which the new molecules were 50% effective in protecting zebrafish neuromast hair cells. The team found a much more stable compound that was nearly 100 times more potent than PROTO-1 in its inhibition of the ototoxic effect. It also exhibited less inhibition to the hERG pathway in a toxicity screen. Once again, they laboriously confirmed its efficacy in rats.

The synthesis of this new compound was published in the Journal of Medicinal Chemistry in October (2), and the FDA approved it as a new drug compound earlier this year. The drug’s target, however, remains a mystery. “Some of the best and longest-used pharmaceutical drugs were originally discovered through phenotypic screens, not through some sort of rational drug design,” notes Raible. The group is planning a Phase I clinical trial this year. Positive results could point to a new drug that protects against aminoglycoside ototoxicity, a boon especially for those who have cystic fibrosis.

“It’s very rewarding when a phenotypic screen yields a biologically active compound with a rich biology behind it,” says Simon. He’s now synthesizing versions of the protective compound that will bind irreversibly to its as-yet unknown target, eventually allowing this target molecule to be isolated and identified.

The implications could extend beyond patients with cystic fibrosis. That drug target is likely critical to normal neuromast function, as is its homolog in inner-ear hair cells. And disruption of this pathway may be part of the mechanism of age-related hearing loss, which according to the National Institutes of Health, affects nearly half of Americans aged 75 years and older (5). To prevent and maybe even reverse hearing loss, Rubel says researchers need critical pathways and genes to target. Zebrafish hatchlings, he and his collaborators contend, could point them in the right direction.