A cancer drug that wears many hats
Nearly a decade ago, the FDA approved the drug lenalidomide to treat patients with deletion-5q myelodysplastic syndrome (del(5q) MDS), a cancer of the myeloid cells in the bone marrow that form several types of blood cells. In this condition, some bone marrow cells are missing a portion of chromosome 5 – hence, the “del(5q)” – on one copy of their genome (the human genome has two copies of each chromosome, one from each parent), and this deletion causes malignant cells to grow unchecked.
Lenalidomide is very effective in treating patients with del(5q) MDS. Until recently, however, scientists didn’t have a good handle on how the drug actually fights cancer. A new study, published in Nature and led by Benjamin Ebert, an institute member at the Broad Institute of MIT and Harvard, offers insight into the versatility of lenalidomide and its effects on cancer cells.
Lenalidomide is a derivative of the drug thalidomide that infamously made headlines in decades past for causing birth defects. In addition to del(5q) MDS, lenalidomide is also approved (with strong restrictions against use in pregnancy) to treat other blood disorders, including multiple myeloma. In January 2014, a team led by Ebert showed that the drug works in a novel, unpredicted way to fight that cancer. In myeloma cells, lenalidomide binds to and changes the activity of an enzyme known as the CRL4CRBN E3 ubiquitin ligase, so that it recruits two B-cell transcription factors, IKZF1 and IKZF3, and “ubiquitinates” them, marking them for degradation within the cell. Because the transcription factors are essential for the malignant cells’ survival, the drug makes cancer cells die.
MDS, however, is a disease of myeloid cells, not the lymphoid cells that form plasma cells and are altered in multiple myeloma. So Ebert and his colleagues suspected that lenalidomide plays a slightly different role in MDS, mediating the ubiquitination – and degradation – of a different protein as it halts the overgrowth of malignant myeloid cells in the bone marrow.
For a closer look at how lenalidomide works in MDS, the researchers teamed up with members of the Broad’s Proteomics Platform to track proteins in myeloid cells treated with the drug. They homed in on an enzyme, casein kinase, that is more ubiquitinated and less abundant after treatment. “This target makes sense, because casein kinase is within the region of the genome that is deleted in del(5q) MDS,” said Emma Fink, co-first author on the paper and an M.D./Ph.D. student at Harvard University and MIT, pursuing research in Ebert’s lab.
Through more experimental work, the researchers proved that casein kinase, not IKZF1 or IKZF3, is the important target of lenalidomide in del(5q) MDS. Because MDS cells have lost one copy of the casein kinase gene (through deletion of the 5q chromosome) they have half the normal levels of that enzyme. Cancer cells can become dependent upon genes that are present in only one copy, as with casein kinase in del(5q) MDS, and a team of Broad researchers led by William Hahn and Rameen Beroukhim identified this class of genes in 2012, deeming them “CYCLOPS (Copy number alterations Yielding Cancer Liabilities Owing to Partial losS) genes.”
The lower level of casein kinase in del(5q) MDS cells allows the malignant cells to take over and cause disease, but it also makes them more sensitive to degradation by lenalidomide than healthy cells. After lenalidomide diminishes a cell’s casein kinase, the tumor suppressor p53 becomes more active, halting growth of the malignant cells.
Because the casein kinase gene is within the deleted 5q region, Ebert’s lab had previously been interested in the enzyme and, fortunately, had already generated a knockout mouse model in which the cells expressed no casein kinase. Human MDS cells with the 5q deletion lack a copy of not only casein kinase, but also many other genes, so a mouse model with a single gene deletion allowed them to more precisely study the role of casein kinase.
“Surprisingly, the drug didn’t work in these mouse cells like it did in human cells,” said Fink. The team discovered that lenalidomide – like thalidomide – doesn’t act the same in mouse cells because of a slight difference in one protein. The mouse genome contains a subtly different code (a single amino acid difference) for the CRBN protein, which is the protein that lenalidomide and thalidomide bind in the ubiquitin ligase complex that marks casein kinase for degradation. By analyzing the sequence of the CRBN gene, they found that the single, altered amino acid is located near the site where lenalidomide and thalidomide bind CRBN, and it indirectly alters the binding of the drugs. By comparing the structures of the two forms of CRBN, the team’s work suggests that the sequence difference in mice introduces a bump on the surface of the CRBN protein that prevents casein kinase and other substrates from being recruited by lenalidomide.
Thalidomide was used in some countries during the 1950s and 60s for insomnia and morning sickness, but the drug’s unforeseen effects on developing fetuses sadly resulted in thousands of children born with birth defects. The findings about the mouse CRBN gene may explain how the drug was initially approved as safe after preclinical studies in mice. “There is no teratogenicity (birth defects) in mice and presumably that could be because the substrates can’t get to the ligase,” said Fink. “Today, you can’t properly study these drugs in mice, as neither IKZF1, IKZF3, nor CK1 are degraded in response to lenalidomide treatment.” A better animal model could one day serve as a more appropriate preclinical framework to test these drugs. For now, lenalidomide and related drugs are used in treating cancer, but with stringent restrictions during pregnancy to avoid the devastating outcomes seen with thalidomide.
With the new evidence, it was clear that lenalidomide is a versatile modulator of the CRL4CRBN E3 ubiquitin ligase, causing the degradation of two classes of substrates. “It’s a really cool mechanism,” said Fink, “and we wondered if the drugs could be useful for other conditions, if we could subtly alter them to target other substrates.”
To test this hypothesis, the researchers made slight modifications to thalidomide and were able to generate an altered drug that exclusively degrades IKZF1 and IKZF3, but not casein kinase. “This could represent a more effective drug for multiple myeloma, with no off-target effects due to degradation of casein kinase,” said Fink.
Building upon the earlier paper revealing lenalidomide’s role in multiple myeloma, this work suggests that it and other related drugs could be tailored to offer precise and powerful new therapeutic options. “These results give us a better understanding of how lenalidomide works in del(5q) MDS,” said Ebert, also an associate professor of medicine at Harvard Medical School, leader of the leukemia program for the Dana-Farber/Harvard Cancer Center, and co-director of the cancer program at the Harvard Stem Cell Institute. “Importantly, it offers key insights about the versatility of this class of drugs, suggesting that we can build better preclinical models and, one day, potentially develop drugs tailored to specific substrates relevant to a variety of human diseases.”
Other Broad researchers on the work include co-first author Jan Krönke, Namrata Udeshi, D.R. Mani, Tanya Svinkina, and Steven Carr.
Paper cited: Lenalidomide induces ubiquitination and degradation of CK1α in del(5q) MDS. Nature. (2015) July 1, 2015. DOI: 10.1038/nature14610.