Five years ago, the cell's trash disposal system was the toast of science when the three researchers who discovered it shared the 2004 Nobel Prize in Chemistry. In 2003, the first drug targeting it was approved by the Food & Drug Administration to treat patients with multiple myeloma, a type of white-blood-cell cancer. The achievements were a boost for researchers studying the system, called the ubiquitin-proteasome pathway. But the next stage of research hasn't exactly proved to be a snap.
The first-in-class myeloma treatment, called bortezomib (Velcade), validated the ubiquitin-proteasome pathway as a target for drugs: It blocks the proteasome, the sophisticated protein machine in every cell that acts as a trash compactor at the end of that pathway. Ubiquitin, a small regulatory protein, is the proverbial kiss of death, a cellular marker that the proteasome recognizes as a destruction signal.
Bortezomib achieved sales of more than $1 billion in 2008. The drug extends sick patients' lives but has side effects, including nerve impairment, that are tough to tolerate.
"Proteasome inhibitors are blunt instruments," affecting myriad cellular pathways, whereas only one or two pathways might be defective, says biochemist Keith D. Wilkinson of Emory University.
As a result, although many groups are developing next-generation proteasome inhibitors, others are also looking upstream of the proteasome, at steps in which ubiquitin gets attached to doomed proteins, with an eye toward developing more targeted therapies. Last month, several companies and research groups discussed their efforts at the Ubiquitin Drug Discovery & Diagnostics Conference, held in Philadelphia. Although the presentations showed they're making progress, targeting earlier parts of the pathway isn't easy.
Before ubiquitin can impart its deadly "kiss," two enzymes ready it for attachment and a third fastens it to the doomed substrate. The enzymes are called E1, E2, and E3, where E stands for enzyme and the number is its order in the ubiquitin pathway. "It's the easiest nomenclature you'll ever find in biology," quips Claudio A. Joazeiro, a cell biologist at Scripps Research Institute.
First, an E1 attaches ubiquitin to its active-site cysteine residue, activating the marker protein for transfer to the cysteine on an E2 ubiquitin-conjugating enzyme. Now loaded with its ubiquitin cargo, E2 binds to an E3 ubiquitin ligase. E3 then acts as a go-between: It binds a protein substrate and helps transfer ubiquitin from E2 to a lysine on that substrate.
Adding to the complexity, after E3s put ubiquitin on a substrate, enzymes called deubiquitinases can remove it. Several of these proteins could make good drug targets, but each has its own challenges.
Ubiquitin activation requires adenosine triphosphate (ATP), which is housed in an ATP-binding site in E1. In terms of drug discovery, an ATP-binding site is a nice place to aim for because the strategy has worked for other proteins. For instance, imatinib (Gleevec), an FDA-approved treatment for chronic myelogenous leukemia, occupies the ATP-binding site in its target, a kinase enzyme.
But a drug that targets an E1 might not be very specific, Wilkinson says. A given cell contains a few E1s, a greater variety of E2s, and many E3s. Because each E1 is intertwined with multiple cellular pathways, disabling an E1 could lead to side effects.
Millennium Pharmaceuticals, the outfit that ultimately brought bortezomib to market, is the first company to test an E1-targeted drug in patients. Its drug blocks an E1 that activates NEDD8, a ubiquitin-like regulatory protein.
Millennium's drug candidate, MLN4924, resembles adenosine monophosphate, a product of the ATP-mediated activation reaction. The agent stops growth of a variety of tumors in mice, and this helped convince the company to move it into human clinical trials (Nature 2009, 458, 732). Today, MLN4924 is in Phase I trials to evaluate its safety and dosing in patients with a range of cancers. It's too soon to tell whether blocking E1s will pay off, and researchers will be looking at Millennium's results for clues.
"You could argue that E1s are too far upstream in the pathway to be good targets and that there would be many biological consequences if they were blocked," James E. Brownell, a scientific fellow at Millennium, told C&EN at the ubiquitin conference. "But maybe for oncology that's exactly where you want to be—shutting down multiple pathways," he said.
Pathway specificity, scientists believe, can be found further down the ubiquitin pathway, at the E3 stage. Each E3 is thought to be responsible for only a few substrates, and many E3s have been linked to particular diseases, such as heart attacks. By targeting E3s, a drug that alters degradation of just a couple of proteins becomes a tantalizing possibility.
Despite decades of research, "it's fair to say we still know very little about how E3s work," Joazeiro said at the meeting. They interact with E2s, but what happens during that protein embrace isn't clear. A few X-ray crystal structures of E3/E2 complexes provide hints, but they don't tell the whole story, he says. None of the structures published so far includes ubiquitin, a vital piece of the puzzle.
The fact that E3s are involved in a protein-protein interaction doesn't help matters, says Ben Nicholson, director of biology at Progenra, a small biotech company focused on the ubiquitin pathway. "Classically, it's difficult to drug protein-protein interactions with any specificity, though many, including Progenra, are trying," he says.
Drugmakers may not know much about E3s, but another part of the pathway, deubiquitinating enzymes (DUBs), might look more familiar. DUBs take off the ubiquitins that E3s stick on and may offer another way of manipulating E3-related pathways, says Wilkinson, whose group has studied DUBs for more than 20 years.
DUBs could be more manageable drug targets to boot, says Harvard Medical School's Alfred Goldberg, whose work helped lay the foundation for the development of bortezomib. "DUBs are a class of cysteine proteases, and we know a lot about how they work," he says.
"With DUBs, you're looking at blocking catalytic activity, which is easier to measure" than blocking a protein-protein interaction, adds Nicholson, whose team is working on DUB inhibitors. "We and others have developed assays to look at this activity," he says.
No DUB-targeted drug has reached clinical trials. But at the conference, a few parties, including Progenra and French biotech company Hybrigenics, discussed preclinical work on a DUB called USP7, which regulates degradation of the well-known tumor suppressor p53. Progenra hopes to nominate a clinical candidate in 2010, Nicholson says.
Right now, the pipeline based on the ubiquitin-proteasome pathway is heavy with proteasome-targeted drugs. Some firms are testing molecules that block the proteasome irreversibly, unlike bortezomib, whose proteasome-blocking activity is reversible. One example of an agent with irreversible activity is carfilzomib, a drug candidate from biopharmaceutical company Proteolix. Carfilzomib has a ketoepoxide moiety that irreversibly modifies the proteasome. That functional group is a major reason that carfilzomib is highly selective for the proteasome, says Mark K. Bennett, vice president of research at Proteolix. Nereus Pharmaceuticals' NPI-0052 also acts irreversibly at the proteasome.
Other companies, including Millennium, are focusing on making proteasome-targeted drugs that can be taken orally, unlike bortezomib, which must be given intravenously. Still more firms are conducting clinical trials of bortezomib in conjunction with other cancer therapies from small molecules to antibodies.
But bortezomib's side effects suggest that moving past the proteasome might be necessary, especially for treating diseases besides cancer. "Cancer's a condition that tolerates toxic drugs because they save lives," Wilkinson says. It's harder to justify severe side effects when treating other diseases, he adds.
The pipeline for this drug class may not get fleshed out much further without a better understanding of the ubiquitin-proteasome pathway. But researchers are optimistic that with better assays and biochemical tools, they'll get there.
That's the same situation once faced by protein phosphorylation, now a well-recognized target for drug development, says Joseph B. Bolen, chief scientific officer at Millennium. "The ubiquitin field today looks like the kinase field did back around 1983," he said at the meeting. "No matter how daunting it looks now, this field's going to make a lot of progress."