Advanced tools can help researchers address acquired resistance to kinase blockers for cancers and expand to other disease targets
Lori Valigra
Valigra is a freelance writer based in Cambridge, Mass.
The future looks bright for kinase inhibitors, therapeutic compounds that block kinase enzymes from catalyzing the transfer of a phosphate group from a donor such as ATP to another molecule. So far, much of the work on the 600 or so different kinases in mammals has been focused on cancers. More than 30 protein kinase inhibitors for cancer are in clinical testing or approved by the US Food and Drug Administration (FDA), including the blockbuster drugs Gleevec (imatinib mesylate), Iressa (gefitinib), and Tarceva (erlotinib). Those drugs have proven effective in blocking the action of their kinase target without causing the negative side effects of traditional chemotherapy.
Although recent reports indicate some patients have developed mutations that cause resistance to these drugs, scientists are not discouraged. Advanced tools such as high-throughput screening, single nucleotide polymorphism (SNP) arrays, exon resequencing, and structural analysis are being used to help better understand the targets, the mutations, and which patients will most likely respond to more potent, second-generation compounds. Kinase targets are expected to be broadened in the future to inflammatory, autoimmune, central nervous system, and cardiovascular diseases.
Kinases are now where G-protein-coupled receptors (GPCRs) were 20 years ago, says Andy Barker, PhD, head of chemistry and oncology at AstraZeneca Pharmaceuticals, Alderley Park, UK. “We’re at the early stage of a group of proteins that could supply some very important drugs in cancer and other diseases in the next 10 to 20 years,” says Barker. He says the majority of drugs on the market now are GPCR agonists and antagonists.
Barker and other researchers are not very surprised about the emergence of somatic mutations. “In a disease like cancer, it is a classic process where a protein changes and mutates to provide survival advantage.” George Daley, MD, PhD, associate professor of biological chemistry and molecular pharmacology at Harvard Medical School and Children’s Hospital, Boston, has developed a method to profile a drug’s activity against various mutants and the way a target might mutate to evade the drug. “I don’t believe there is a [cancer] target that is immune to mutation,” says Daley. “If we can make a drug against a target, the cancer cell will figure out a way to mutate away from the drug.”
Gleevec, a BRC-ABL inhibitor for chronic myelogenous leukemia (CML) from Novartis International AG, Basel, Switzerland, is probably the best known of the three big cancer kinase inhibitors, having been heralded by some as a “magic bullet” when it came to market. AstraZeneca makes an epidermal growth factor receptor (EGFR)-tyrosine kinase inhibitor for non-small-cell lung carcinoma (NSCLC) called Iressa, as does OSI Pharmaceuticals Inc., Melville, N.Y., with Tarceva.
The Gleevec mutations were the first to be heavily publicized, with point mutations detected in the ATP-binding domain of the ABL gene, a factor which disturbs the binding of Gleevec to its target [Z. Iqbal et al., Biol. Proced. Online, vol. 6, pp. 144-148 (2004)]. More recent papers cite mutations that cause resistance to Iressa or Tarceva [S. Kobayashi et al., N. Engl. J. Med., vol. 352, pp. 786-792 (2005); W. Pao et al., PLoS Medicine, vol. 2 (no. 3), p. e73 (2005)]. Some second-generation compounds already are in development, including a “super Gleevec” molecule called AMN107 by Novartis, which is in phase II clinical trials.
Overcoming Initial Skepticism
Initial development of kinases was fraught with naysayers, many of whom didn’t
believe it was possible to inhibit kinases, especially at the ATP binding site, says Sasha Kamb, PhD, vice president and global head of oncology at Novartis Institutes for Biomedical Research Inc., Cambridge, Mass. ATP is used broadly in cells as a source of energy and phosphate, which is used by kinases. It is present in high millimolar concentrations in cells, so any drug would have to compete with ATP. “There was much skepticism about whether that was possible,” says Kamb. “Scientists in the predecessor company of Novartis [Ciba-Geigy] took a gamble and invested in what at that time was considered to be an extremely high-risk project.”
The original work on Gleevec (also known as Glivec in Europe) started in the early 1960s, when it was discovered that certain leukemias such as chronic myelogenous leukemia (CML) had a telltale chromosome abnormality, the so-called Philadelphia chromosome. With the advent of cloning technology, it was possible to identify the molecule, which had a chromosomal reciprocal translocation that exchanged two arms of two human chromosomes that fused the Abelson kinase with the BCR gene. This became known as the BCR-ABL protein. Kamb says the reasonable assumption was that because BCR-ABL was present in the leukemias, it was somehow essential to the tumor growth of those leukemic cells, and it made sense to try to inhibit it. That assumption wasn’t proven correct until Gleevec was developed.
The researchers were looking for a drug that looked like ATP. At the time, there was no high-throughput screening available to help cull through potential compounds, so they used what Kamb says were low-throughput screens and assays. “A number of molecules looked at were basically eyeballed, and they looked a bit like ATP. That’s how Gleevec was first discovered,” he says. Some modifications were then made based on screening and medicinal chemistry decorating of the original molecule in different ways to see which modifications would provide higher potency. It was then screened for safety using standard techniques. “The fortunate and key thing is that ABL is a normal kinase that’s present in the body. Very few, if any, cells depend on it for survival, so it turns out to be a pretty well-tolerated cancer drug,” Kamb says.
Work on kinases can be a challenge, because they make up a large gene family. Because kinases bind to ATP, which doesn’t evolve very quickly, chemists are under pressure to come up with selective inhibitors. “If you inhibit all the kinases, obviously you’ve got a very toxic compound,” Kamb says. “That’s one of the challenges in this area, to produce compounds that have reasonable selectivity for a specific target or targets.” But that’s where all the tools of the drug discovery profession come into play, he says. It is extremely helpful to have crystals, or high-resolution structures of the targets, to help guide compound design to get affinity and selectivity. It also helps to have many of the kinases in a screening format so they can be screened against the target and counterscreened against other kinases that are not desired. Kamb says that having cellular assays as well as biochemical assays is very important when studying kinases. “It’s often seen that inhibitors that work nicely in a biochemical assay actually don’t work well in a cellular assay, and sometimes vice versa. So there’s something about the cellular context that changes the chemical or biochemical properties of the protein of target. It isn’t clear why.”
| Finding New Mutations Matthew Meyerson, PhD, assistant professor of pathology at the Dana-Farber Cancer Institute and Harvard Medical School, Boston, and his colleagues are trying to discover new mutations, primarily somatic ones, in protein kinases. The researchers are systematically using single-nucleotide polymorphism (SNP) arrays to find copy number alterations in the genome and high-throughput exon resequencing to find coactivating point mutations. The group is taking cancer samples and sequencing all of the protein tyrosine kinase genes. They perform genotyping to determine if mutations exist, then repeat the sequencing in larger samples to validate the results. Meyerson says a major discovery of his group is EGFR mutations. EGFR inhibitors such as Iressa are good candidates for treating lung-cancer patients, and a significant number of patients showed responses, but it wasn’t clear who those patients were or why they responded, he says. “Using the systematic exon resequencing approach, we discovered mutations in lung cancer and found that those mutations correlated with response to Iressa or Tarceva. It was a surprise.” |
Gleevec initially was targeted at CML patients, but Novartis researchers determined it also hits other kinases, including KIT, with reasonably high affinity. KIT is activated in gastrointestinal stromal tumors (GIST), and Gleevec is approved for use in GIST as well. Phase II trials of a second-generation Gleevec-like compound called AMN107, which hits the same target but has a different chemical structure, will begin soon. AMN107 has higher potency and a sufficiently different binding mechanism, so it can actually hit many of the resistant mutants. In retrospect, Kamb says, Novartis was lucky with Gleevec because it was a first molecule that had very few problems. Novartis used the toolkit initially developed for Gleevec and some newer technologies such as high-throughput screens to develop the follow-on compound. They also used more biochemical and cellular assays, larger chemical libraries, and more crystal structures.
Honing Activity, Selectivity
Although some people think selectivity can be a problem with kinases, AstraZeneca’s Barker doesn’t agree. “I don’t think it’s any more of a problem than with any other large class of biochemical targets.” A typical drug program aimed at a GPCR would look similar in its steps to one aimed at a kinase, he says, but the levels of knowledge are different.
When Iressa was developed about 10 years ago, the company used standard biochemical assays. AstraZeneca incorporated radiolabels into its substrate or used colorimetric detection to look for changes in a protein or substrate. But some of the newer tools AstraZeneca finds helpful include the use of more structural information on kinases to design directed libraries, refine structural activity relationships, and develop selectivity between different kinases. Barker says that recent steps forward include a better understanding of some of the signal transduction pathways that kinases are involved in and their relative importance in various disease states. “A lot of that has been teased apart by new specific antibodies that recognize activated and nonactivated kinases. Those sorts of tools have proven very useful not only in understanding the fundamental bioscience, but in understanding how that bioscience relates to overall effects in a disease downstream.”
An advantage with kinases is that it is possible to model many of the mutations in structural terms and generate proteins to test against. “The bigger challenge is to know which patients have which mutation and when to treat with a drug,” Barker says. “With some cancers, there are very specific mutations that seem to occur regularly and in others there are a variety of mutations in different parts of the protein, none of which is dominant, and that makes targeting specific mutations quite difficult.” Like other researchers, he thinks cancer treatment may evolve to be like HIV treatment, where a cocktail of two or three inhibitors can cover 99% of the kinases and their mutations and stabilize the disease for a period.
“I suspect it’s possible that one molecule may not serve all needs in this field [kinases],” says Tomi Sawyer, PhD, senior vice president of drug discovery, Ariad Pharmaceuticals Inc., Cambridge, Mass.
mTORs Pose Other Challenges
To Ariad’s Sawyer, there are two different extremes of developing protein kinase inhibitors. One is the story of Iressa and Gleevec, and having to figure out a way to override the mutations. The other is Ariad’s experience with mammalian target of rapamycin (mTOR), which did not require starting development work with a dozen or more different templates to optimize into potent inhibitors. The mTOR inhibitors appear to not have the problem of kinase mutations. “We started with nature’s optimized inhibitor, rapamycin,” he says. The challenge to Ariad, and other companies, is to develop their own creative ways to advance a proprietary rapamycin analog. Ariad developed AP23464, a proof-of-concept molecule that is a potent inhibitor of SRC, ABL, and KIT, which it is now optimizing in analogs. The company is in phase II tests with AP23573, an mTOR inhibitor of hematologic malignancies and various solid tumors.
Sawyer says there is very little wiggle room from a chemistry and drug-design standpoint to provide an active rapamycin analog. In the case of rapamycin optimization to make proprietary second-generation analogs, chemists have focused on one site in the rapamycin molecule, a hydroxyl group on carbon 43 that is unstable because it is a site of metabolism. Ariad used its proprietary phosphorus chemistry to develop a potent and effective in vitro and in vivo mTOR inhibitor. “We knew exactly where to modify the molecule. We had proprietary chemistry that only had to be exploited within the framework of the rapamycin molecule, and it was a straightforward process to evaluate the initial process in cells and then go right to in vivo [testing],” Sawyer says of AP23573.
Even though Ariad hasn’t had to deal with the issue of mutations, Sawyer thinks new tools like the mutational cell analysis approach being used at Children’s Hospital in Boston will be important in future kinase work. “This is an important emerging tool to deal with these mutant kinases,” says Sawyer. “It will become more and more important for people to determine whether or not they have a novel and potentially very effective new molecule that can escape some of these key mutants.”
Filed Under: Drug Discovery
