Genome mining meets microcrystal electron diffraction to speed up drug discovery

More than 60% of drugs are derived from natural products—chemicals made by living organisms like bacteria, fungi, and plants. But identifying and characterizing biologically active molecules in the complex chemical brew these organisms make can be a painstaking process that takes years. By combining genome mining with microcrystal electron diffraction (microED), scientists may have found what may be a faster way of identifying drugs from natural products.

Researchers led by University of California, Los Angeles chemical biologist Yi Tang and California Institute of Technology organic chemist Hosea M. Nelson developed the prospecting process (Nat. Chem. Biol. 2021, DOI: 10.1038/s41589-021-00834-2). The researchers comb through the DNA sequence of an interesting organism, looking for genes that are known to make enzymes responsible for constructing biomolecules. They then insert those genes into a host that they can grow in the lab. This host churns out those enzymes and the natural product of interest.

“It’s a very powerful way to do it because sequencing is becoming really easy and cheap and fast,” Tang says. Also, the organism that makes the interesting molecule in nature might not grow in the lab, or when it’s growing in the lab, it might not make the molecule the scientists want. “It’s estimated that over 90% of all the biosynthetic pathways are silent under laboratory conditions,” Tang says. Genome mining can help scientists tap into those unexplored regions of chemical space.

The major bottleneck in making drugs from natural products is identifying the structure of interesting molecules. This usually requires researchers to make many liters of culture from which they hope to get a few milligrams of the molecule they’re interested in. But it only takes nanograms of a molecule to determine its structure with microED, a cryo-electron microscopy technique that reveals molecular structures. “That’s just unattainable by any other method,” Nelson says. Tang says that now his group has to prepare only 5–10 mL of culture to get enough material for microED analysis.

Pairing the two methods streamlines the identification and characterization process. As a proof-of-principle, Tang and Nelson’s team identified and solved the structure of a new natural product, which they call Py-469. They also figured out the structure of fischerin, a cytotoxic natural product that’s been known for more than 25 years but had not been unambiguously assigned a chemical structure.

“As one who is always on the lookout for novel natural products to synthesize in the laboratory, it is reassuring, especially when those unique targets have been encountered on small scale, that their structures have been accurately determined,” says University of Chicago chemist Scott A. Snyder in an email. Time will tell if this method will accelerate drug discovery, he says, but “given the significant role that natural products play as medicinal agents and pharmaceutical leads, I am excited by the prospect.”