Whether it’s bacteria-killing mold in a petri dish giving rise to penicillin or a cell division-inhibiting platinum surface inspiring cisplatin, some of the most important advances in drug discovery have been made serendipitously (1,2). While these initial incidents might be considered pure luck, translating the observations to therapeutic developments requires scientists daring enough to relentlessly ask questions and seek answers. Fortune favors the bold.

In a recent study published in Nature Cancer, researchers at the National Center for Advancing Translational Sciences (NCATS) and the Massachusetts General Hospital Cancer Center stumbled upon — and boldly pursued — the discovery that a cellular enzyme can transform harmless molecules into potent cytotoxic agents (3). In doing so, they identified a compound that kills liver cancer cells upon enzymatic modification and uncovered a larger class of molecules that is activated in the same manner, providing a new strategy for enhancing the selectivity of anticancer drugs. “Where we ended up is not the answer to the question we asked, and so we followed the science in this project,” said Matthew Hall, a cancer drug discovery researcher at NCATS and coauthor of the study. 

The team was originally interested in screening molecules to identify those active against liver cancer cells carrying an oncogenic mutation in the isocitrate dehydrogenase 1 (IDH1) enzyme. They pinpointed a compound in their library called YC-1 that showed high selectivity in killing cells with the IDH1 mutation over cells with the wild type enzyme. However, when the researchers tested the compound in an expanded panel of heterogenous cell lines, they found that its activity did not correlate strongly with IDH1 mutation status. “The more data we generated, the more complicated it got, and the less we knew, which sounds like science to me,” Hall said.

To explore the biological basis for a cell’s sensitivity to YC-1, the researchers developed acquired resistance models where they cultured liver cancer cells with increasing concentrations of the compound. This exposure prompts the cell to develop changes that prevent YC-1 from killing it, allowing the researchers to infer the compound’s mechanism of action. The team analyzed the proteome in the resistant cells and found that it showed reduced levels of the cytosolic sulfotransferase enzyme SULT1A1. The researchers observed that using gene editing to knock out SULT1A1 yielded cells that were not susceptible to YC-1, while reintroducing SULT1A1 restored the cells’ responsiveness to the compound. “It was really nifty to see the absolute clear cut ability to switch on or off sensitivity based on one enzyme in a completely predictable manner,” said Nabeel Bardeesy, a pancreatic and biliary cancer researcher at Massachusetts General Hospital Cancer Center and coauthor of the study. “We didn’t expect to see such a smoking gun.”

To investigate YC-1’s SULT1A1-mediated mechanism, the researchers systematically designed and tested a series of structural analogs and observed that a pentagon-shaped furfuryl alcohol moiety is critical for the compound’s activity. They found that inhibiting SULT1A1 also blocked YC-1’s cytotoxicity, indicating that the compound serves as a substrate for the enzyme, and determined that SULT1A1 attaches a sulfate group to YC-1 at the furfuryl alcohol. “It creates a very reactive molecule when it does that. It creates a monster,” Hall said. 

It was really nifty to see the absolute clear cut ability to switch on or off sensitivity based on one enzyme in a completely predictable manner. 
- Nabeel Bardeesy, Massachusetts General Hospital Cancer Center

Based on other SULT1A1-generated molecules, the researchers suspected that the modified YC-1 compound unleashes its terror by binding to certain biomolecules in the cell. To test this hypothesis, they developed a derivative of YC-1 with an affinity tag that can be isolated from the complex cellular environment. By using this technique to analyze the biomolecular targets of YC-1, the team found that it binds to lysine residues in RNA-processing proteins, including those involved in RNA splicing, translation, and metabolism. YC-1 may interfere with these critical functions upon binding, resulting in cell death. 

The researchers then developed mouse models with liver tumors either expressing or lacking the SULT1A1 enzyme and treated them with YC-1. They observed that while the SULT1A1- negative tumors did not respond to the compound, the SULT1A1-positive tumors showed reduced growth or shrank. While SULT1A1 is expressed in the healthy liver, intestine, lung, and adrenal gland, YC-1 did not induce cell death in normal liver tissue in the mice, suggesting that rapidly dividing cancer cells may possess a unique vulnerability to YC-1’s mechanism of action.

The team plans to explore whether cancer cells show higher expression of SULT1A1 and might therefore selectively activate YC-1. Determining the SULT1A1 expression level needed to turn on YC-1 could also allow the enzyme to serve as a biomarker to predict a patient’s response to the compound.

“It's a beautiful story, both from basic science perspective, but also from a translational potential,” said Uttam Tambar, a medicinal chemist at the University of Texas Southwestern Medical Center who was not involved in the study. “It's an exciting discovery because it provides a totally different strategy for combating cancer.” Tambar and the authors agree that further research is needed to determine which of YC-1’s many binding interactions with RNA-processing proteins is responsible for triggering cell death. 

The researchers, however, weren’t satisfied stopping with YC-1. “This is a really interesting molecule. It's got really interesting activity, and it's activated in a really interesting way,” Hall said. “We started to wonder, well, maybe SULT1A1 is responsible for the activity of other molecules that have been reported.” 

To find out, the team searched a National Cancer Institute database containing data on the cytotoxicity of more than 22,000 compounds against 60 cancer cell lines for compounds with activity that correlated with either YC-1’s profile or SULT1A1 levels (4). They identified a group of 80 molecules with chemical features needed for modification by SULT1A1 and confirmed that they are only cytotoxic in SULT1A1-positive cell lines.

The researchers’ findings reveal an unexpectedly broad role for SULT1A1 — and potentially related sulfotransferase enzymes expressed in other tissue types — in the activity of previously studied anticancer compounds (5). “Sometimes when you read a paper, the penny will drop about something you couldn't fully understand,” Hall said. “I do think that there's plenty of stuff in the literature that will be reinterpreted; I think people will start looking at some of the other [sulfotransferase] enzymes and perhaps thinking about them as activating factors.”

Revisiting existing compounds or designing them from scratch with this new insight could yield an entire toolkit of selectively-activated anticancer agents. “Do[ing] this again with other very high confidence sulfotransferases that might have very restrictive expression in other difficult-to-treat cancers is something that we are looking at and quite excited about,” Bardeesy said. 

By seeing where an unexpected observation led them, the researchers provided a valuable new avenue in cancer drug development. “This study speaks to the power of unbiased approaches,” Tambar said. “As long as you're very rigorous about following through and trying to understand the mechanism of action, you can discover something really useful serendipitously.”

References

1. American Chemical Society. Discovery and development of penicillin. At < https://www.acs.org/education/whatischemistry/landmarks/flemingpenicillin.html>.
2. Monneret, C. Platinum anticancer drugs: from serendipity to rational design. Ann Pharm Fr  69, 286-295 (2011). 
3. Shi, L. et al. SULT1A1-dependent sulfonation of alkylators is a lineage-dependent vulnerability of liver cancers. Nat Cancer  4, 365-381 (2023). 
4. National Cancer Institute. NCI-60 human tumor cell lines screen. At < https://dtp.cancer.gov/discovery_development/nci-60/>. 
5. Chapman, E., Best, M.D., Hanson, S.R., & Wong, C.-H. Sulfotransferases: structure, mechanism, biological activity, inhibition, and synthetic utility. Angew Chem Int Ed Engl  43, 3526-3548 (2004).