Research published this week by a team including scientists from proteomics firm Cellzome and GlaxoSmithKline suggests that the small molecule inhibitor I-BET151 could be useful as a treatment for MLL-fusion leukemia.
In a study published in the current edition of Nature, the researchers used Cellzome's Episphere screening platform to demonstrate that the bromodomain and extra terminal, or BET, family of chromatin adaptor proteins are associated with leukemia-initiating mixed lineage leukemia fusions, indicating that they may be promising therapeutic targets for the disease.
They then tested the efficacy of GSK's BET inhibitor I-BET151 in mice and human leukemia cells, finding that it provided survival benefit in two distinct mouse models and accelerated apoptosis and abrogated clonogenic efficiency in the human cells.
The work was part of an agreement that GSK and Cellzome signed in 2010 that could be worth up to $645 million to the proteomics firm (PM 3/10/2010), and provided data with which GSK intends to seek approval to begin clinical trials for the inhibitor, said Tony Kouzarides, one of the study's leaders, a professor at the University of Cambridge, and a member of both GSK's and Cellzome's scientific advisory boards.
MLL-fusion leukemia is responsible for around 80 percent of infant leukemia and up to 10 percent of adult leukemia. According to Kouzarides, the notion that I-BET151 might be useful as an drug for the disease emerged from a series of proteomic screens demonstrating that the BET family proteins BRD3 and BRD4 were associating with protein complexes known to bind MLL-fusion proteins. In particular, the BET proteins were found to interact with members of the super elongation complex — SEC — and the polymerase-associated factor complex — PAFc — which are key to malignant transformation by MLL fusions.
"It all came out of the proteomic analysis, which indicated that the BRD proteins were associating with complexes that normally bind MLL," he told ProteoMonitor. "Therefore, the hypothesis was put forward that these BRD3/4 proteins would allow targeting of these MLL fusions in leukemia. So it was like we had found the targeting molecule of these MLL fusions."
To identify the nuclear complexes associated with BRD3 and BRD4, the researchers performed three proteomic surveys. In the first, bead-immobilized analogues of the BET inhibitor I-BET762 were incubated with HL60 nuclear extracts and the pulled-down proteins were analyzed via mass spec. In the second, the scientists performed immunoprecipitation with antibodies against BRD2/3/4 as well as against proteins that had co-purified with BRD3/4 in the first step. The third screen used bead-immobilized histone acetylated peptides — which BET proteins recognize — to purify protein complexes.
The researchers used this data to generate a list of complexes identified by all three methods, the idea being, Kouzarides said, that "using three different approaches to pull down the same set of proteins would increase the chances of whatever you pulled down being [a] real [interaction]."
The emergence of SEC and PAFc proteins as BET binders suggested that displacement of the BET proteins from chromatin via an inhibitor like I-BET151 might prove effective in treating MLL leukemia.
Other potential leads emerged from the analysis as well, Kouzarides said, noting that a protein involved in multiple myeloma was also pulled out with BRD3/4. His team is now following up to investigate BET proteins' roles in a variety of other cancers as well as in inflammation, he said.
While kinase inhibitors have been a focus of cancer research for some time now, inhibitors of epigenetic targets like the BET proteins are of growing interest to pharma firms, Gerard Drewes, Cellzome's vice president of discovery research, told ProteoMonitor.
"I think you could say that epigenetic targets right now are where kinases were 15 years ago," he said. "So I think we're at the beginning of the way [with epigenetic targets]. And there are a large number of epigenetic targets — methyltransferases, de-methylases, bromodomain proteins — so it's quite a large collection of potential drug targets."
The company's Episphere platform is essentially a variety of its Kinobeads kinase profiling platform that uses probes on immobilized beads to pull out kinases in cell and tissue samples for analysis by mass spec, Drewes said. Instead of probes for kinases, Episphere uses probes for classes of epigenetic proteins, like the BET family.
"It's just adopting this principle to target [protein] classes in epigenetics that, [like] kinases, also have very structurally conserved molecular binding sites," he said.
Cellzome's work in the Nature study was done on a Thermo Scientific Orbitrap Velos instrument, but, Drewes said, the company is in the process of upgrading that machine to the new Orbitrap Elite, noting that "you need state-of-the-art equipment if you want to do accurate identification and quantification in complex protein mixtures."
The company also recently acquired a Thermo Scientific Q Exactive instrument, which Drewes said he found particularly attractive because of its combination of qualitative and quantitative capabilities, including its potential for SRM-type analyses (PM 9/16/2011).
"I think the big potential of the Exactive is that it covers applications that you would traditionally run maybe on a triple quad, like multiple-reaction monitoring-type experiments. You can cover that partially on the Exactive," he said. "It's almost like combining both worlds — both your typical LC-MS/MS and your MRM-type approaches."
Such a targeted mass spec capability is increasingly important given that Cellzome is increasingly using targeted mass spec for its assays, Drewes said.
"While obviously for many experiments we still run our [mass spec] setup in shotgun mode, we are increasingly using targeted mass spec," he said. "For instance, if we use these [BET protein-targeting] bromodomain beads, then we know the bromodomain proteins that bind to a particular set of beads, and we can use targeted mass spectrometry to get more robust identification."
Have topics you'd like to see covered in ProteoMonitor? Contact the editor at abonislawski [at] genomeweb [.] com.