‘Heroic’ Korean Skincare Ingredient Could Help Fight Antibiotic-Resistant Bacteria

Centella asiatica is a scrubby, unremarkable-looking herb that creeps across wet ground in Vietnam, India, and much of tropical Asia. You might never look twice at it. But if you’ve spent any time browsing Korean skincare products, you’ve almost certainly rubbed its chemical offspring into your face; madecassic acid, one of the plant’s key compounds, has become something of a hero ingredient in serums and moisturizers marketed for calming irritated skin. What nobody in the beauty aisle quite expected was that the same molecule might also help solve one of medicine’s most urgent problems.

Antibiotic resistance is, by most credible estimates, on track to kill 39 million people between 2025 and 2050. Developing new antibiotics from scratch remains painfully slow and expensive, so researchers have increasingly turned to natural compounds, the chemical arsenals that plants evolved over millions of years to fend off microbial threats of their own.

A team at the University of Kent and University College London has now shown that madecassic acid can inhibit the growth of multidrug-resistant E. coli by targeting a piece of bacterial machinery that humans simply don’t have. The target is called the cytochrome bd oxidase, a respiratory protein complex embedded in the inner membrane of many pathogenic bacteria, where it helps generate the energy cells need to survive during infection. Block it, and bacteria lose access to a critical energy pathway. Because this protein exists only in prokaryotes (it is completely absent in human and animal cells), drugs that disable it could, in principle, kill bacteria without harming the patient.

That selectivity is the sort of thing drug developers dream about. And the computational modelling suggested the fit was, perhaps, even better than expected.

Using molecular docking simulations, the researchers found that madecassic acid binds to the quinol site of E. coli’s cytochrome bd-I oxidase with a predicted affinity roughly 95 times stronger than the enzyme’s own natural substrate, ubiquinol-8. The pentacyclic structure of the molecule (five interconnected carbon rings, if you’re counting) slots neatly into the protein’s hydrophobic cleft, while its polar groups point outward toward the surrounding water. Laboratory oxygen-consumption assays on isolated bacterial membranes confirmed the prediction, yielding an IC50 of about 34 micromolar.

“Plants have been a source of natural medicines for millennia, and now contemporary research approaches can reveal the mechanisms of action,” said Mark Shepherd, a reader in microbial biochemistry at Kent who led the study. “This is an exciting time, and we hope to further our understanding of natural antimicrobials from plants, nature’s great chemical factories.”

But the story got more complicated once the team started tinkering with the molecule’s architecture. They isolated madecassic acid from Centella asiatica leaves grown in Hue Province, Vietnam, then synthesized three chemical variants, each modified at different parts of the molecular scaffold. One derivative, dubbed MA-2, had its alcohol groups acetylated, making it less polar. Another, MA-3, swapped the carboxylic acid tail for a long hydrocarbon chain tipped with a positively charged amine group, a modification that dramatically increased the compound’s lipophilicity and shifted its calculated partition coefficient by nearly four log units. A third variant, MA-4, carried the amine chain but retained the original hydroxyl groups. All three were predicted to outcompete the natural substrate for the cytochrome bd-I binding site, and all three did inhibit the enzyme in membrane assays, though the relationship between binding affinity and actual biological effect turned out to be anything but straightforward.

Here is where it gets properly interesting. MA-2, which docked most tightly to the protein, performed well in isolated membranes but was rather poor at stopping whole bacterial cells from growing. MA-3, predicted to bind the weakest of the lot, was surprisingly effective against both membranes and live bacteria. A reminder that what happens in a test tube and what happens inside an intact cell are quite different propositions.

And then there was MA-4. Last in predicted binding strength, weakest in enzymatic inhibition, and yet it was the only compound in the entire study that actually killed E. coli cells (albeit at relatively high concentrations, with an LC50 around 304 micromolar). The others merely stopped bacteria from growing without finishing them off. Nobody anticipated that result.

The researchers reckon several mechanisms are probably at work simultaneously. Madecassic acid and its relatives don’t just block cytochrome bd oxidases; previous studies have implicated membrane disruption, interference with protein synthesis, and topoisomerase inhibition. Making the molecule more hydrophobic, as with MA-3, likely drives it deeper into the lipid bilayer, which could concentrate it near membrane-bound proteins, or alternatively trap it there, sequestered away from its target. Teasing apart these overlapping effects is, to put it mildly, a challenge. But it’s also an opportunity, because the chemical structure of madecassic acid is amenable to all sorts of modifications, giving researchers a flexible scaffold they can tune toward specific bacterial targets.

The broader implications stretch in two directions at once. For drug development, cytochrome bd oxidases remain an attractive class of targets precisely because they are absent from human biology, and madecassic acid offers a tractable starting point for optimization. For the skincare world, there’s a rather different question: what is this ingredient doing to the bacterial communities living on your skin every time you apply that calming serum? That part of the story, for now, remains mostly unwritten.

DOI: 10.1039/d5md01116g