James Netterwald, PhD, MT (ASCP)
Senior Editor
Kinomics is being used for drug discovery, but has it followed other “omics” in being a bona fide high-throughput method?
Kinases are one of the most important classes of protein in cells. They play a role in cellular signaling, regulating everything from cell growth to inflammation. But, more
 click to enlarge Kinomics analysis reveals the functional activation state of proteins in complex signaling networks. (Source: Gavin Boswick, PhD) |
importantly, when their function goes awry, kinases have been implicated in certain human cancers. Because of this role in cancer, kinases have been on the priority list that drug companies draw up while looking for new drug targets for years.
Kinomics is a fairly new science. It basically relies on high-throughput screening assays to study the entire kinome (collection of all kinases in the cell) rapidly and accurately. But how this new science is being used differs from researcher to researcher. And, of course, the technologies they use also differ.
Now, without further ado, please welcome kinomics to the drug discovery stage.
Kinomic technologies
With the human kinome enumerated, drug companies are setting the stage for new kinase inhibitors. “There are 518 [human] kinases out there and you want to do high-throughput screening on them and identify compounds that can inhibit the different kinases relatively specifically or in a family fashion, depending on what your indication is,” says David Goldstein, PhD, director of chemistry for Roche Pharmaceuticals, Palo Alto, Calif.
According to Goldstein, Roche is using a high-throughput screening assay for kinase inhibitors called KinomeScan, manufactured by Ambit Biosciences in San Diego, Calif. “We can send [Ambit] a compound tomorrow and within 10 days, I get back the inhibition profile against 317 kinases on their panel. That’s 317 high-throughput screens that were run against a single compound in a very accurate and rapid manner,” says Goldstein. Other companies providing such assays include Millipore Corporation and Invitrogen Corporation. (See related article)
“The advantage of the Ambit technology is that they don’t clone, express, and purify each individual and then set up an assay to be able to screen for inhibition,” says Goldstein. Ambit does not subscribe to the traditional system, which requires the purification of a kinase to high enough levels to yield kinetically-meaningful data. On its Web site, Ambit describes how KinomeScan works. “KinomeScan is an ATP-site dependent competition binding assay in which human kinases of interest are fused to an Ambit-proprietary T7 bacteriophage and a known ligand is immobilized on solid support. The amount of kinase bound to an immobilized ligand is measured in the presence and absence of test compound.”
Roche has a few compounds currently in clinical trials that were discovered using this kinomics approach, indicating its power and usefulness in the drug discovery process. Goldstein did not disclose the names of the compounds, but did say that they were the product of the high specificity so characteristic of kinomics. But because kinomics is a fairly new science, not all kinomics-driven drug discoveries are at the clinical stage just yet. In fact, as discussed below, some are still at the basic research stage.
Kinomics to fight malaria
Some sources say that kinomics actually predated genomics because high-throughput kinase assays were available before the human genome was completed. Under this assumption, kinomics might
 |
have aided in drug design pre-genomics. One scientist who tried to target kinases for drug discovery before the genomics era is Christian Doerig, PhD, director of INSERM U609 at the Wellcome Center for Molecular Parasitology, University of Glasgow Biomedical Research Center, Glasgow, Scotland, UK. “Of course, when the genomics era came in and the malarial parasite, Plasmodium falciparum, was sequenced, this opened up new opportunities and we quickly published a study of the kinome,” says Doerig.
Indeed, the sequencing of the malarial genome gave Doerig an exhaustive list of potential drug targets. One key fact about the malarial parasite that brought Doerig’s search to the fore is that the lack of homology between malarial kinases and those of its human host. This lead to greater opportunities for chemotherapeutic intervention, says Doerig. “A good percentage of kinases we have found in the malarial parasite are called orphan kinases, which means that they don’t mesh with any clearly established kinase groups found in human cells or in yeast.”
Of course, Doerig must first select a kinase target, and this selection process is based on biology. Part of the selection process involves systematically knocking out all of the malarial kinases, 65 in total, using reverse genetics. “This
| Old School “Omics” It’s difficult to pinpoint when kinomics became a science in and of itself. The main reason for this is in the way kinomics is defined. A quick search on the Web reveals that kinomics is defined as a high-throughput method for screening a single inhibitor against a multitude of kinases. But other people have their own way of defining kinomics. “To me, kinomics is about the generation of a knowledge base that you can use to systematically explore and classify therapeutically-relevant kinase targets for drug discovery,” says David Goldstein, PhD, director of chemistry for Roche Pharmaceuticals, Palo Alto, Calif. “Essentially what we are doing is looking at each of these kinases and asking: ‘Are these actually desirable targets for drug discovery?’” Goldstein explains. “Every single one of these kinases plays a role in signal transduction, but not every one of them makes a good drug target.” He cites the example of Gleevec (Novartis), a cancer drug discovered in the pre-genomic era. He purports that kinomics was used in the discovery of Gleevec. But that would make kinomics the first omic. Is this possible? “I think so,” says Goldstein, “When you think about the discovery of Gleevec, it was before the human genome was even solved.” So, although this discovery did not involve screening for particular kinase genes, looking for polymorphisms, microarray screens, or any other typical omic-like characterics, it did involve the targeting of a specific kinase. And this was backed up by extensive research on the kinase. “People looked at that particular target because it was an oncogene. It was a kinase that was always on. And the always-on form of that kinase was correlated to a particular type of cancer. And now people taking that drug are living eight to 10 years longer. That’s exactly what kinomics is trying to do on a broader scale.” So, no matter how kinomics is defined, it appears to be doing its job, and that is, providing solid data to support drug discovery efforts. |
is target validation in which we determine which malarial kinases are essential for its asexual life cycle and pathogenesis in human red blood cells,” says Doerig. He also expresses the kinases in Escherichia coli and then tests their activity in a high-throughput kinomics assay. But genetic manipulation in the malarial parasite is not straightforward and often takes months to identify the functions of a handful of kinases. However, Doerig is moving along, characterizing about a third of the malarial kinases thus far. Each time he obtains an active kinase, he sends them to an undisclosed industry contact, who performs high-throughput screening on them to test out new inhibitors. Currently, they have malarial kinases from every major family in the pipeline.
Doerig is also using human kinomics to identify human signal transduction targets where the function is hijacked by the parasite during infection. This opens up new opportunities for drug targets, including the possibility of using pre-existing cancer drugs that inhibit kinases common to both cancer and malarial pathogenesis.
“We are characterizing a number of inhibitors right now, but it is a bit premature to talk about,” says Doerig. “The next step will be to identify inhibitors that can be pushed forward into the development stage in the drug pipeline.”
Pepped up
It seems like a rational credo would be that the more one knows about the biology of a kinase, the easier it will be to design a drug against it. And here’s a case-in-point proving that this credo is indeed rational. “Designing drugs to interact with kinases might be more surgical than to design them to disrupt the global down-regulation of a gene, which can have wider off-target or side effects,” says Norbert Herzog, PhD, associate professor of pathology at the University of Texas Medical Branch, Galveston, Texas. He explains that there would be less off-target effects if the inhibitor interacts with a single, specific kinase. And having the ability to design an inhibitor to block the function of a single kinase requires a deep understanding of that kinase’s function—in other words—its biology.
On the subject of biology, Herzog points out that “the beauty of kinases is that they are conserved [among different species].” This simple fact allows researchers to test inhibitors designed against human kinases to be reliably tested on non-human animal models. Herzog uses a guinea pig model of human hemorrhagic fever virus infection. By comparing the kinomic profiles of attenuated and virulent strains of the South American Pichindé virus, Herzog and his team are attempting to understand the host response to the virus. More importantly, they have found that the response is mediated by kinases.
To do their kinomics, Herzog’s group uses The Pep Chip, manufactured by Pepscan Systems in Lelystad, Netherlands. Each chip is composed of 12,000 peptide substrates on a glass slide array. By looking for the phosphorylation pattern of substrates on this chip, Herzog can identify a variety of kinases that are differentially activated when comparing animals infected with virulent versus the attenuated viral strains.
“Using the Pep Chip opened our eyes to all of the kinase activity that we never anticipated as being part of the immune signaling to hemorrhagic fever,” says Herzog. He adds that, in the process, he has been able to find some kinase inhibitors that have been used as cancer treatment, but may also block the same kinase culprits in infectious disease. In fact, epidermal growth factor receptor kinase, platelet-derived growth factor receptor kinases, and a few other well-known cancer protein kinase biomarkers came up positive in Herzog’s Pep Chip scans.
“The problem with our kinomics scan is that there are lots of phosphorylation sites on the chip,” says Gavin Boswick, PhD, a postdoctoral fellow in Herzog’s lab. And Herzog adds that one of the challenges of doing kinomics-based drug discovery is that many of the phosphorylation sites have not been put into any kind of biological context. And for this reason, it is difficult to know what a kinase is doing to its target, he says. In other words: Is the kinase turning its target “on” or “off”?
After the kinomics scan, Boswick says that he often has to “go back into the literature and see if the roles of those phosphorylation sites have been characterized.” And this surely slows down the high-speed analysis so typical of other “omics.” But, of course, part of the lack of speed is due to more of a bioinformatic limitation than to a problem with the high-throughput nature of the kinomics screen. Of course, like any good science, every one of the positives in a kinomics scan must be validated by an independent kinase assay.
Although no new inhibitors have come out of Herzog’s kinomics scans, he is confident there will be in the future. Herzog purports that once an investigator knows a particular kinase is involved and what its substrate is, they have the beginning of a drug design to inhibit that kinase.
An emerging star?
Although it is a relatively new science, kinomics-driven drug discovery has a lot of credibility. But what of the future of kinomics-driven drug discovery? One researcher put it this way. “If you look at the situation of kinases, you’ll see that use of kinase inhibitors has had success in some cancers. But this success has come with the price of a high degree of side effects,” says Goldstein. “Kinase inhibitors also correlate with congestive heart failure. Can we predict that? Maybe in the future, we can predict that using kinomics. If we can engineer our future molecules to avoid those kinases, we can avoid side effects.” And judging by the public concerns about industry’s ability to deliver perfectly safe drugs, if kinomics can deliver on this promise, it would be a giant leap for drug discovery.
This article was published in Drug Discovery & Development magazine: Vol. 10, No. 7, July, 2007, pp. 22-25.
Filed Under: Drug Discovery
