Skeletal editing can transform molecules’ biological activity

Researchers in Korea have used an oxygen-to-nitrogen atom swap reaction to dramatically alter the biological activity of natural products (J. Am. Chem. Soc. 2026, DOI: 10.1021/jacs.6c14447). The work highlights how single-atom changes can have big effects on molecules’ activity and the potential of skeletal-editing methods to make such changes.

“Our question was, by changing that single atom in a very tiny portion of the compound, can we kind of change the fate of [its] biological function?” says project leader Yoonsu Park of the Korea Advanced Institute of Science and Technology.

Atom-swapped analogs of drugs and natural products aren’t new, but chemists usually must backtrack many steps in a synthesis to make them. Skeletal-editing methods aim to give chemists tools to strategically alter molecules without such backtracking.

The work is an extension of chemistry that Park’s lab developed in 2024. The researchers updated the protocol to use ammonium chloride as a nitrogen source instead of organic amines or ammonium carbamate. The insoluble NH₄Cl provides a slow release of ammonia so that it doesn’t interfere with the photochemical oxidation step, thus making the reaction compatible with a broader range of molecules.

Switching the ammonia source also enabled the researchers to insert isotopically enriched nitrogen atoms into molecules to observe how they behave in biological systems. The team created a library of ¹⁵N-labeled molecules, including analogs of several commercially available natural products that contain a furan group. They then partnered with Ho Jeong Kwon’s lab at Yonsei University to explore how the edited molecules interact with proteins.

Using a thermal binding assay, the researchers found that ¹⁵N-LIMO, the pyrrole-bearing version of the citrus alkaloid limonin, engages with a very different set of proteins than its parent compound. Limonin prefers binding to kinases, while ¹⁵N-LIMO goes for metabolic and mitochondrial proteins such as nicotinamide nucleotide transhydrogenase (NNT).

“Just by changing a minimal portion of the whole molecule, we could change the binding affinity,” Park says.

Further cell studies showed that both limonin and ¹⁵N-LIMO have antifibrotic activity, via completely different mechanisms. Limonin inhibits a signaling pathway, while ¹⁵N-LIMO acts on mitochondrial redox systems to reduce reactive oxygen species. ¹⁵N-LIMO is also less cytotoxic than limonin.

“This is a very nice application of skeletal editing in a complex setting,” says Mingji Dai, an organic chemist at Emory University who was not involved in the work. Skeletal editing is getting a lot of attention in the synthesis community, but “we really need to see some major applications,” Dai says, and this paper represents a concrete step toward those applications. That the reaction works so well on molecules with such a high density of sensitive functional groups is “remarkable,” he adds.

Much more work would need to be done to translate the findings to drug discovery, Park says, but the study serves as proof of concept for how skeletal editing can reprogram molecules’ activity. It also illustrates why synthetic chemists should collaborate more with chemical biologists. Park and his team plan to take their natural product alterations further by using multiple editing methods in sequence, he says. “The value here is that we have an easy way to diversify preexisting libraries.”

Indrajeet Sharma of the University of Oklahoma says the thing that impressed him most about the work is that it gives researchers an inexpensive way to make isotopically labeled pyrroles. Precision edits to fine-tune a molecule’s activity are still a ways off, but “at least we can make a change,” Sharma says. “That is a big advancement . . . and if those molecules are labeled, then it’s icing on the cake.”