Artemisinin is best known as a potent antimalarial drug isolated from the sweet wormwood plant, a 1,600-year-old Chinese herbal remedy. Beginning in the 1990s, researchers reported that artemisinin also showed anticancer activity. But precisely how it acted against cancer cells was a mystery. Researchers now report that artemisinin interacts with a cell-death-promoting protein present in cultured cancer cells. This is the first evidence that the molecule directly targets a tumor-related protein and suggests a route to designing novel, artemisinin-based drugs to target cancer cells (ACS Chem. Biol. 2019, DOI: 10.1021/acschembio.8b01004).
Previous studies in cancer cell lines and rodents have identified a range of possible ways that artemisinin acts on cancer cells, but none had homed in on a specific protein target for the drug. Typically, researchers identify such targets using chromatography, affinity capture mass spectrometry, or other analytical methods. But these techniques work best when cells have high concentrations of the protein target or, if a protein is less abundant, when it binds strongly to a drug, says Peter Karuso of Macquarie University, who led the study. But proteins unique to cancer cells tend to occur at low concentrations, and if such proteins don’t bind strongly to artemisinin, they might have been overlooked in earlier experiments.
So Karuso, Ho Jeong Kwon of Yonsei University, and their colleagues turned to phage display, a technique that can amplify rare proteins by expressing them in bacteria-infecting viruses known as phages. Phages can be easily multiplied by growing them on bacteria. Even if an artemisinin-binding protein is only present in minute amounts, “if its interaction with the drug is weak but specific, you’ll eventually see it,” Karuso says.
The team used four commercially available phage libraries; each is a collection of phages that express bits of DNA from a human cell line. Collectively, the library spans all the peptides encoded by that cell line’s genome. Here, the team used phage libraries of breast, colon, liver, and lung tumors, as well as a noncancerous control.
Then, they performed a series of “bio-panning” experiments to identify the phages carrying protein fragments that bound to artesunate, a semisynthetic derivative of artemisinin that is the drug commonly used to treat malaria. Each experiment began with pouring the phage library over microwell plates coated with artesunate, seeing which phage particles bound, and then amplifying that subset of viruses. The researchers then repeated the experiment with that subset.
After nine such cycles, they found that artesunate bound to a peptide found in all four cancer libraries—a protein named BAD, which, in its active form, binds to a cell-death-promoting protein. When BAD is inactive and thus not bound to this protein, it can promote the uncontrolled growth of cancerous cells.
In subsequent cell-culture experiments, the team treated a cancer cell line with gene-silencing RNA sequences that knock down BAD expression. They then treated the cells with artesunate and found that 71% of the cells lacking BAD survived, indicating that when BAD was missing, artesunate could not switch on the cell-death pathway. But in control cells that expressed BAD, artesunate increased amounts of the active form of BAD and so promoted cell death. As a result, only 20% of these cells survived.
This finding suggests that cells expressing BAD were more sensitive to the drug. The researchers saw no effect when they used a random mix of other gene-silencing RNAs or another cell-death-inducing drug that targets a different cellular protein, suggesting that artesunate indeed works to kill cancer cells by activating BAD.
Artemisinin’s target may have proved elusive in the past because BAD is an intrinsically disordered protein, meaning it changes its shape depending on environment and binding partners, Karuso explains, adding that the phage display techniques used here could extend to “other drugs where the target is not known or is suspect.”
The results are also a route to designing anticancer molecules based on artemisinin—an effort that’s been slow in the past, says pharmacology researcher Jun O. Liu of Johns Hopkins University. Now, knowing its potential protein target could help identify more potent members of the same class of compounds as potential anticancer agents. Liu adds that “this may not be the drug’s last or only target, but it certainly opens up new directions for people to study.”