Harvard University COVID-19 updates

Department News

PREDICTING DRUG TARGETS BY WATCHING ZEBRAFISH SLEEP

PREDICTING DRUG TARGETS BY WATCHING ZEBRAFISH SLEEP

(L to R)Alexander Schier and Jason Rihel

Many psychiatric drugs were discovered decades ago through serendipity. For example, a tuberculosis drug called iproniazid also happened to make people happier and therefore became the first marketed anti-depressant. One reason for luck’s role is that modern drug discovery methods rely on testing millions of compounds outside the body, but the complex brain cannot be modeled in a dish. Drug screeners can make educated guesses about what targets to tweak to alleviate symptoms, but often compounds fail to work as expected in the context of the intact brain. If candidate drugs could be tested on behaving animals in large numbers, this problem might be avoided. In work published in the Jan. 15th issue of Science, Jason Rihel, David Prober, Alex Schier and their collaborators discovered that zebrafish behavioral responses to drugs can be used to predict a drug’s biological target, suggesting that drug testing in zebrafish may facilitate the drug discovery process.

The research team did not start the project with drug discovery in mind. Drugs were used as a way to probe biological pathways for their role in controlling rest and wake states. But the researchers soon realized that their large dataset could be used to predict drug targets. Specifically, the group used videotracking software to observe the effects of nearly 5600 small molecules on the rest and wake of more than 60,000 zebrafish larvae and found 463 compounds that induced significantly altered behavior. They then transformed the complex, multi-dimensional behavioral data for each compound into a phenotypic fingerprint and used clustering algorithms to group the drugs based on the similarity of the induced behaviors. This organized the compounds broadly into sedating and arousing drugs, but it also grouped them into sub-categories, for example compounds that increased waking only at night, or compounds that altered only daytime behavior. Importantly, compounds that clustered together by behavioral fingerprints also often shared annotated biological targets. But not always—sometimes a poorly characterized or unexpected compound co-clustered with a set of drugs that shared a common target. The implication was that these poorly characterized drugs shared the same target as their neighbors, an insight the group used to correctly predict drug targets. The authors propose that large-scale behavioral profiling can now be used to complement traditional drug screens and characterize large classes of compounds for effectiveness, potential side effects, and combinatorial properties.

The dataset also identified several novel signaling pathways involved in the regulation of rest/wake states, the original goal of the screen. Intriguingly, the authors found that a variety of anti-inflammatory agents, including steroidal and non-steroidal anti-inflammatory drugs, increased waking activity during the day. Inflammatory signaling pathways had been thought to increase sleep amount during infection, but the zebrafish data suggests that immune signals may also regulate normal waking activity. The authors also found a wide variety of drugs that inhibited a specific class of potassium channels and selectively increased wakefulness at night. Finally, they identified calcium channel inhibitors that increased total rest with minimal effects on waking activity. Current efforts in the Schier lab aim to use techniques from genetics and neuroscience to pursue the roles of these drug-targeted pathways in zebrafish sleep/wake regulation.

The Schier group collaborated with Lee Rubin, Anthony Arvanites, and Kelvin Lam of the Harvard Stem Cell Institute; Stephen J. Haggerty of the Broad Institute of MIT and Harvard; and Randy Peterson and David Kokel of Massachusetts General Hospital. Schier lab members and co-authors Steve Zimmerman and Sumin Jang also made important contributions to the research.

Read more in Science

Read more in HarvardScience

 

 

View Alex Schier’s Faculty Profile