Over the past five years, researchers have pieced together how a mutant version of an enzyme known as isocitrate dehydrogenase (IDH) churns out copy after copy of a compound that drives tumor growth in many human cancers, including more than half of all adult gliomas and around one-sixth of all cases of acute myelogenous leukemia. Because the aberrant enzyme acts only in healthy cells, researchers have hypothesized that targeting it could choke tumor cells without harming other tissue. It's the perfect recipe for a cancer drug, but until now that idea has been little more than a theory.

Mutations on the brain: Frequency of IDH1 mutations in various glial tumors. Credit: Neurosurg. Clin. N. Am. 23, 471–480 (2012)

Reporting last month, scientists from the drug company Agios Pharmaceuticals, together with partners at academic institutions, showed that a pair of experimental drugs—one that blocks mutant versions of IDH1, another that blocks IDH2—seem to curtail tumor growth in mouse models and in human cells taken from patients with glioma and leukemia (Science http://doi.org/k7w and http://doi.org/k7x, 2013). “The pleasant surprise, at least with the caveat that these are preclinical studies, is that in fact one can measure reversibility [in tumor growth],” says William Kaelin, a physician-scientist at the Dana-Farber Cancer Institute in Boston who serves on Agios's scientific advisory board but was not involved in the recent studies. “So I think I'm a lot more optimistic than I was a year ago about these drugs.”

Importantly, neither drug blocked cells carrying the 'healthy' version of the IDH enzymes, demonstrating the drugs' specificity. “This approach should have an advantage in safety and tolerability, since normal cells should not be affected by mutant selective inhibitors,” says Scott Biller, chief scientific officer at Agios, which is headquartered in Cambridge, Massachusetts. The company is now working on developing clinical-grade versions of the two compounds for use in people, although Biller declined to comment on when such trials will commence. “In the clinic, we will determine whether IDH inhibitors are suitable as monotherapy but also plan to investigate rational combinations,” he told Nature Medicine.

A mean old amino: A single substitution in the IDH1 protein triggers cancerous complications. Credit: Molecular & Cellular Proteomics 9, 1703–1715

Metabolic minefield

IDH, when mutated, is an unusual cancer-causing protein. Unlike most oncoproteins, which have usually simply acquired increased activity levels, mutated IDH adopts a wholly new function. Normally, the two IDH enzymes act in the energy-generating pathway known as the citric acid cycle, converting the molecule isocitrate to alpha-ketoglutarate. But mutant forms of IDH1 and IDH2 gain the ability to also convert alpha-ketoglutarate into a different metabolite called 2-hydroxyglutarate (2HG).

As Kaelin and his colleagues showed earlier this year, mutations in IDH1 and higher levels of 2HG both promote growth and block differentiation in blood stem cells (Science 339, 1621–1625, 2013). And as 2HG builds up, it inhibits dozens of other regulatory proteins involved in much more than just the citric acid cycle. These include proteins that control chromatin structure and DNA methylation. Hypermethylation ensues, which shuts down cell differentiation, ultimately leading to cancer (Cancer Cell 18, 553–567, 2010).

“The thinking in the field right now is that the IDH mutations and the accumulation of 2HG produce a slow, progressive buildup of epigenetic changes,” explains Ingo Mellinghoff, a neuro-oncologist at the Memorial Sloan-Kettering Cancer Center in New York who co-led the recent study involving Agios's inhibitor of mutant IDH1 in glioma. Questions remain, however, about exactly which epigenetic and cellular changes are those promoting tumorigenesis. “It's now important to narrow down which of the 70 or so enzymes affected and which of the epigenetic changes are really the most important ones,” Mellinghoff says. But, he adds, knowing that information is not needed to move forward on the existing inhibitors.

Meanwhile, academic chemists continue to pursue other compounds that, like the Agios drugs, can disrupt faulty IDH enzymes in cancer cells. For instance, Yongcheng Song and his colleagues at Baylor College of Medicine in Houston are currently using medicinal chemistry and X-ray crystallography techniques to develop potent and selective inhibitors of mutant IDH1. And Takashi Tsukamoto is leading a team at the Johns Hopkins University Brain Science Institute in Baltimore to study whether blocking the formation of alpha-ketoglutarate through a mechanism independent of the citric acid cycle could reduce 2HG production and thus limit the growth of IDH-mutation–containing cancer cells. Aside from Agios, no other commercial company has currently announced publicly the pursuit of IDH inhibitors.