Data mining for a hydrogen sulfide antidote

Researchers have found a surprising use for a database of fluorescent probes: searching for antidotes to hydrogen sulfide (H₂S) poisoning. Washington State University chemist Ming Xian and a team of undergraduate helpers initially built the searchable database to help biologists sort through the multitude of available H₂S fluorophores. When he found out that H₂S is a potential chemical weapon, he found a new use for the data: searching for compounds that might be used to treat H₂S poisoning by scavening it from the body (Angew. Chem., Int. Ed. 2019, DOI: 10.1002/anie.201905580).

“I didn’t plan to do this,” Xian admits. “If we didn’t have this database, probably I would never have thought about designing scavengers.”

H₂S is a simple molecule with a surprising variety of roles. In the body it relaxes smooth muscle and causes blood vessels to dilate. It also plays a role in diseases like cancer and diabetes.

But the gas was also used as a chemical warfare agent in World War I, and the US Department of Homeland Security lists H₂S as a chemical of concern that might be used by terrorists.

At low levels, H₂S smells of rotten eggs. At higher concentrations, it can cause serious health problems such as nausea and throat irritation, or even death. And, says Xian, there isn’t an FDA approved antidote for H₂S poisoning.

Xian wondered if his database might have a solution. He considered that some of the properties of a good fluorescent reporter molecule—biostable, specific for H₂S—are also what you need for an antidote to H₂S poisoning. “We don’t need the fluorescent dye component,” he explains, “we just need the reactivity side.”

Xian searched the database for probes that reacted the fastest with H₂S, and identified 30 candidate molecules. Then, he and his team performed a variety of tests to narrow down the list. The most promising were sulfonyl azides. They then tested one of these, methanesulfonyl azide, in mice poisoned with H₂S, injecting some of them with potential antidote. All the mice treated with the compound survived, unlike in the control group.

“This paper is a nice example of using a large data set of parameters from the literature and using it to guide synthesis and screening efforts,” says Michael Pluth, who studies H₂S signaling at the University of Oregon. It can be hard to compare the reactivity of probes because most experiments are run slightly different conditions. Taking a broader approach to mine and group these data, as Xian’s group did, is a clever way to pull out trends, he adds.

A lot more tests will need to be completed before these compounds become therapeutics and Xian hopes to further optimize them for use as antidotes. He is also keen to find other uses for H₂S scavengers, such as seeing if the compounds can work with existing cancer drugs to improve their efficacy.