New antifungal kills without toxic side effects

If approved, synthetic compound could bolster shaky arsenal of antifungal drugs

The antifungal Amphotericin B (AmB) is an old and effective drug—it saved many COVID-19 patients whose compromised immune systems failed to stop secondary fungal infections. But it sometimes causes life-threatening kidney damage. Now, after more than a decade of sleuthing into this toxicity, researchers have not only found an explanation, but used it and a robotic “chemist” to devise a powerful antifungal alternative without any obvious side effects in mice and human cells. And the strategy that led to the discovery of the compound, described today in Nature, may offer a route for detoxifying other antimicrobial drugs.

“This is really inspiring work,” says Leah Cowen, a mycologist at the University of Toronto. “They leveraged molecular insights into how the drug works to dial up the properties they wanted and dial down properties they didn’t want.”

Worldwide, fungal diseases kill some 1.5 million people annually, about the same as tuberculosis or malaria. But in contrast with antibiotics, where dozens of classes of effective drugs are available, there are only three classes of antifungals, and each faces problems of toxicity, growing resistance, or limited effectiveness. “We are very much in need of new antifungals that are safe and effective,” Cowen says.

AmB, produced by a Streptomyces bacterium, was first isolated in 1955 from soils near the Orinoco River in Venezuela. But it wasn’t until 2012 that researchers led by Martin Burke, a chemist at the University of Illinois Urbana-Champaign (UIUC), figured out that it kills fungi by stripping them of ergosterol, a key structural support in their cell membranes. Human cells don’t use ergosterol. But cholesterol, a closely related sterol, performs much the same function in human cells. And Burke and his colleagues found that AmB likely causes renal damage by stripping cholesterol out of the membranes of kidney cells and weakening them.

AmB is a large, complex compound that’s painstaking to make from scratch. But in 2015, Burke and his colleagues reported in Science that they had developed a robotic machine to synthesize novel drug compounds from hundreds of premade building blocks. The machine allowed Burke’s team to quickly build and test AmB analogs, each with a slight tweak to its chemical structure, to see whether any had reduced toxicity. One initially promising compound, abbreviated AmBMU, was safer but ultimately proved less effective in animal studies. “We pushed the pendulum too far,” Burke says. “We got rid of the toxicity but lost potency.”

Burke and his colleagues went back to the drawing board. They found that AmBMU’s lower potency wasn’t so much because of a looser binding between it and ergosterol, but the slower pace at which the molecule stripped out the membrane component. Fungi like yeast “could make new ergosterol faster than we could remove it,” Burke says.

They gained ideas on where tweaks to the molecule might speed up its action by studying high-resolution images of AmB binding to both ergosterol and cholesterol recently provided by Chad Rienstra’s group at the University of Wisconsin-Madison. “The atomic resolution models were really the key to zoom in and identify these very subtle differences,” says Corinne Soutar, a UIUC graduate student and the paper’s co-first author.

Burke’s team turned back to its synthesis machine, generating more than 200 additional analogs. In the end, swapping the position of a hydrogen atom and a hydroxyl group on the backbone of the molecule blocked its ability to bind to cholesterol. They also tweaked a carboxylic acid group at the bottom of the molecule, speeding up how quickly the antifungal removed ergosterols. Today in Nature, Burke and his colleagues report that in cell culture the compound, dubbed Am-2-19, is at least as effective as AmB, if not more so, in killing more than 500 different fungal species. Studies on mice showed that Am-2-19 thwarted three common, hard-to-treat fungal infections with no signs of toxicity, even at high doses. And tests on human blood and kidney cells produced no warning signals. Am-2-19 has been licensed to Sfunga Therapeutics, which has launched a phase 1 human safety trial in New Zealand.

Beyond the hope that Am-2-19 will make a safer form of AmB, Burke says the same strategy of teasing out the key molecular interactions could help detoxify other antifungals that are in the same class as AmB and are already on the market. Several such compounds work through the same ergosterol-stripping mechanism and are similarly toxic. And the group’s synthesis machine could help discover them quickly. Burke says: “The lights have been turned on.”

Source