Amphotericin B is one of medicine’s mightiest drugs: a potent, broad-spectrum antifungal antibiotic that kills even some of the most dangerous fungi outright. Despite more than half a century of use, pathogens have formed little resistance to it—a virtual unicorn in the antimicrobial world.
But like many powerful things, amphotericin B is also very toxic.
“When I was doing clinical rounds in residency, we called it ‘ampho-terrible,'” Martin Burke, M.D., Ph.D., a structural chemist and physician-scientist, recalled to Fierce Biotech Research in an interview. Though it has a laundry list of side effects, the kidneys bear the brunt of the blow. Even a form of the drug that’s designed to be less toxic causes acute kidney damage in up to 20% of patients who receive a standard dose—a figure that climbs to 50% at higher doses, which are often necessary to treat life-threatening infections.
“Oftentimes doctors are left with this challenge of killing the infection, but with kidney damage,” Burke said. “So the question has been, how do we make it not toxic to the kidneys but keep all the good stuff?”
Now, his lab at the University of Illinois Urbana-Champaign and collaborators at the University of Wisconsin-Madison may have found the answer. In the results of a study published Nov. 8 in Nature, they showed that their compound AM-2-19 can treat severe systemic fungal infections in mice without harming the kidneys or losing its forerunner’s superpowers. The drug has been formulated into the candidate SF001 and is currently being tested in a phase 1 clinical trial under the still-in-stealth biotech Sfunga Therapeutics, which Burke founded.
“I think this is going to be a very important product and anti-fungal,” Kieren Marr, M.D., co-founder and chief medical officer at Sfunga, told Fierce. Marr retired from an accomplished career in academic medicine to develop SF001, pulling in industry veterans from the antifungal world to join her.
“People just kind of hopped on and no one batted an eye, because this is the opportunity to work on what is, in my opinion, one of the most impactful drugs that we’ll make in our lifetimes,” Marr said.
Holes in the hole theory
Amphotericin B’s medicinal history begins in the 1950s, when the compound was isolated from Venezuelan soil by researchers with the Squibb Institute for Medical Research, now part of Bristol Myers Squibb. Until the 1980s, the drug was the only one that could be used to treat systemic fungal infections, despite its gnarly side effects. Researchers tried wrapping the compound in lipids to make it less toxic, leading to the launch of liposomal amphotericin B in the 1990s.
“Basically, there’s been no other advancements in this field since the lipid formulations of amphotericin B were introduced,” Marr said. Use of azole antifungals, such as fluconazole, became more widespread in the late 1990s and early 2000s, she added, “and we just kind of reserved these amphotericin B products to situations where the clinical scenario was pretty dire.”
Meanwhile, scientists tried to find ways to improve upon amphotericin B. For half a century, the prevalent thinking was that it killed fungi by boring holes in their cell membrane—which it also did to kidney cells, so the theory went, hence its propensity for damaging them. Efforts to ameliorate the compound’s toxicity thus focused on tuning that mechanism.
But through his lab’s research on the structure of amphotericin B, Burke realized that the old theory might be incomplete. Dissecting different aspects of the drug’s activity through what’s known as a modular synthesis approach showed that even if its ability to create holes in the cell membrane was removed, its antifungal activity remained.
“That showed us that there’s something else going on, that the mechanism was wrong,” Burke said. “Amphotericin B does form ion channels, but that’s not its primary mechanism of killing.”
Additional experiments by Burke’s lab showed that the drug’s main killing MO came from its ability to bind to a fatty substance called ergosterol, the fungi form of cholesterol, in fungal cells. His team then joined up with the lab of Chad Rienstra, Ph.D., at the University of Wisconsin-Madison, where they leveraged Rienstra’s expertise on solid-state nuclear magnetic resonance (NMR)—MRI for molecules, as Burke put it—to look more closely at what the compound was doing.
As it turns out, amphotericin B seems to take its inspiration not from a weapon’s arsenal, but from cleaning supplies. The compound self-assembles into a structure on the fungus’ surface, then extracts ergosterol from it like a sponge. “It’s this loss of ergosterol that causes the fungicidal effects,” Burke explained. Their findings were published in Nature Chemical Biology in 2014.
This led the researchers to wonder whether amphotericin B had an affinity for other sterols, too—specifically the ones found in humans. Experiments included in their new Nature paper showed that to be the case: The compound binds to and sucks cholesterol from kidney cells just as it does ergosterol from fungi.
“It was exciting to us to see that the same mechanism is what drives toxicity in human kidney cells,” Burke said.
Speed over strength
Now that the researchers knew how amphotericin B was really killing cells, it was time to try building a better version of the drug. They started by using NMR to look carefully at the way it binds to ergosterol and cholesterol, then tested to see whether the strength of those bonds was greater for one over the other. Seeing that it bound more tightly to ergosterol than cholesterol, the scientists performed what they call “controlled destabilization.”
“We intentionally messed it up, but just a little bit, so you still keep the ergosterol binding but lose the cholesterol binding,” Burke recalled. Yet while this took away the compound’s toxic traits, it also reduced its potency, tests on fungi showed.
“That’s a conundrum: If you make the binding stronger, you risk binding to cholesterol again,” he explained. Rather than focusing on strength, they decided to focus on the rate of extraction instead—in other words, a faster sponge. Tuning this dial solved the problem.
“We had faster ergosterol extraction and still didn’t detect any cholesterol binding,” Burke said. The final compound was dubbed AM-2-19 (which stands for “Arun Maji, lab notebook 2, page 19,” a moniker plucked from the notes of first author Arun Maji, Ph.D. In press materials from the University of Illinois, Maji said he would have “called it something else” had he known).
Next, the researchers tested AM-2-19 on kidney cells from mice and humans, including some that were lab-grown and others derived from patients. Seeing that the compound had no effect on any of them, they moved to evaluating its effects on fungi. They tested head-to-head against amphotericin B on more than 500 different strains, working across four different research labs to do so. The drug killed all of them, meeting and even surpassing the scope of amphotericin B.
Finally, the team moved to live animals. After successful in vivo toxicology studies on healthy mice, they tested AM-2-19 on four different models of mice with systemic infections of candidiasis, aspergillosis or mucormycosis. It again worked in all of them, including against a form of aspergillus that is resistant to amphotericin B.
“It was at least as potent if not better,” Burke said. “We saw no toxicity.”
More funds for Sfunga, and beyond
The animal studies were part of an investigational new drug application with the FDA, which in June 2023 gave Sfunga the green light to enter the drug into phase 1 clinical trials. Enrollment on the single ascending dose study was recently completed, and Sfunga plans to begin multiple ascending dose studies at the start of 2024. The company has dosed six cohorts of volunteers so far, and feels “very, very confident” about the results, Marr said.
Meanwhile, more fundraising is on the horizon. Deerfield Management fully funded the biotech's series A; as it moves into phase 2 next year, the company will be looking for more investment and potential partnerships, Marr added.
“[After phase 1] we’ll be at a point where we’re seeking more investment to finish our clinical developments, as well as potentially strategic partners for alignments either in late development or commercialization,” she said.
Back in the lab, Burke and his team are already looking for ways to apply their findings more broadly, especially to treating pathogens that have developed antifungal resistance. Amphotericin B is part of a large family of natural products called glycosylated polyene macrolides that likely all work the same way, as his team has shown, presenting an opportunity to optimize them.
“The fundamental science that’s been illuminated in this paper could provide a general roadmap for more broadly improving the pharmacological properties of this entire family of natural products,” Burke said. What’s more, the discovery of their sponge-like mechanism paired with the modular chemistry that ultimately made AM-2-19 possible is indicative of a new stage of drug design, he added, one that leverages compounds that work by self-assembling into multimolecular aggregates to, in his words, “become a whole that’s greater than the sum of its parts.”
“We’re able to kind of rationally design and engineer changes to molecules to make them behave the way that we want,” Burke said. “Because these types of compounds evade resistance, it’s an exciting playbook to make resistance evasive antimicrobials, which we badly need in many areas.”
Editor's note: An earlier version of this article incorrectly said that Deerfield Investments funded Sfunga's Series A. The firm was Deerfield Management.