Courtesy of Kinexus
Kinases, the enzymes that catalyze phosphorylation events, have been implicated in hundreds of different diseases, including cancer, inflammatory diseases, and neurological disorders. Compounds that control their activity, therefore, hold rich promise for drug development.
Consider Gleevec, Novartis Pharmaceutical's antileukemia drug. Gleevec targets the Bcr-Abl tyrosine kinase in patients with chronic myelogenous leukemia. In binding that enzyme, the drug inhibits proliferation and induces apoptosis, thereby slowing cancer progression. Though Gleevec targets a special kind of kinase (one formed specifically by a chromosomal translocation and therefore found only in patients with a specific cancer), the general strategy could be applicable to other diseases.
"I believe that every person probably has a unique form of cancer at the molecular level," says Steven Pelech, professor of neurology at the University of British Columbia, and founder and CEO of Vancouver-based Kinexus. "The pharmaceutical industry is developing a very specific arsenal of kinase...
A PROBLEM OF SPECIFICITY
With the human kinase complement, or known, it should be possible to develop assay panels against the whole collection. Yet none of the companies currently offering screening services covers even half of the 518 possible enzymes.
One reason for that is the shortage of highly effective, commercially available antibodies for use in screens that detect phosphorylated proteins. "We know from our own work how difficult it is to make antibodies," says Pelech. "Only about one out of three that you make are worth selling. Unfortunately, a lot of manufacturers sell antibodies that don't work."
The Upstate company, one of the leaders in the development of kinase profiling, offers two types of profiling services from its campus in Dundee, Scotland: a radiometric assay that uses a weakly radioactive form of ATP to measure rates of phosphorylation directly and which does not rely on antibodies; and a fluorescence-polarization assay, which does rely on antibodies.
"Many people do view the radioactive assay as the gold standard," says Andrew Paterson, general manager at the Dundee facility. But it is also expensive, he says, and difficult for labs to duplicate in-house because of the safety and environmental challenges posed by the radioactive materials. Upstate's radiometric assay service screens a panel of 208 kinases, while the fluorometric method covers 80.
Much of the technology and many of the reagents Upstate uses for its kinase-profiling service were developed in the lab of Philip Cohen, director of the protein phosphorylation unit and Royal Society Research Professor at the University of Dundee. In 2000 Cohen's lab demonstrated that many commercially available kinase inhibitors were not nearly as specific as they were previously thought to be.2 Cited more than 1,000 times to date, the study underscores the difficulty of drawing conclusions based on inhibitor studies: These reagents can be powerful probes of signal transduction pathways, yet unforeseen effects on nontarget kinases can yield misleading results.
"If you haven't done kinase profiling, you may have a compound that is claimed to be a specific kinase inhibitor, but in fact is nothing of the kind. You need to know the limitations of each inhibitor," Cohen says.
That's because such limitations have practical impact, both in the lab and in the clinic. The wider the range of pathways a kinase inhibitor disrupts, the more likely the compound is to affect normal cell processes adversely. Comprehensive data on what pathways a potential drug affects will help not only to screen out potential candidates for drug development, but also to yield better understanding of a compound's potential uses, side effects, and weaknesses.
HIGH-THROUGHPUT IN VITRO SCREENING
Courtesy of Kinexus
Cohen estimates the world market for kinase profiling at around $200 million and climbing. "There's obviously a need for this in the academic community as well as in the biotech world," he says. "It's very much a growth area."
Several companies have joined Upstate in developing that market, including Invitrogen in Carlsbad, Calif. The company's SelectScreen service uses a discriminating protease and fluorescence resonance energy transfer (FRET) to differentiate between phosphorylated and unphosphorylated kinase substrates. The list price is about $45 per data point at low volumes, though Invitrogen's prices, like those of most companies offering these services, depend greatly on the size and time frame of the project. "It's a highly robust method for calculating percent phosphorylation," says Richard Somberg, Invitrogen's research area manager for molecular tools and targets.
Currently, Invitrogen has a panel of 100 target kinases, which it plans to double by the end of the year. Invitrogen recently announced a major agreement with Plexxikon of Berkeley, Calif. The pharmaceutical company will perform all the company's kinase screening using SelectScreen.
Invitrogen also offers a CellSensor service, which tracks specific signal-transduction pathways within cells engineered using the company's GeneBLAzer reporter technology. Pharma customers frequently use the CellSensor lines to investigate promising compounds that are further along in the drug-discovery pipeline, says Chris Armstrong, Invitrogen's business area manager for molecular tools and targets.
"One of the key differentiators for Invitrogen compared with some of the other providers is that we can take projects and clients from very early understanding of kinase biology right through understanding of mechanism and action through both molecular and cellular processes," says Armstrong.
Other companies that offer selectivity screening services include ProQinase, in Freiburg, Germany, (94 kinases, from about $81 to $3.60 per data point); NovaScreen, in Hanover, Md., (either 36 or 60 kinases, $4,320 or $7,200 per compound respectively); and Kobe, Japan-based Carna Biosciences (101 kinases,). Many of these companies manufacture their own recombinant enzymes, and also offer the same kinases used in their profiling service to researchers in smaller kits or as reagents.
In February, San Diego-based Ambit Biosciences published details on a novel method for kinase selectivity screening in
PEERING INTO THE CELL
No matter how good the assay, in vitro screening will never entirely capture the complexity of how kinases work in their native habitat, the cell. "We've made a huma protein outside of a human, and we put it in a test tube, and you have a lovely view of how that inhibitor might work against an enzyme in isolation. Our service is very straightforward, very easy to interpret. But in a cell is the best approximation of what's going to happen in the human body," says Paterson.
Taking a broad view of how kinases behave in the context of a cell is Kinexus' specialty. With its Kinetworks service, the company offers one-shot immunoblot screening for 25 to 78 kinases or uniquely phosphorylated proteins at a time, thereby tracking the presence or activity of kinases in cells either experimentally treated with inhibitory compounds or left as controls.
Currently, Kinexus has the capability to screen about 180 kinases and about 250 phosphorylation sites. The Kinetworks platform can also track phosphatases, as well as proteins involved in heat shock, apoptosis, and cell-cycle progression. The resulting immunoblots from both the control and the experimental samples are combined in a single image, revealing the screened sequences that have increased in quantity, those that have decreased, and those remained unchanged.
At Louisiana State University's Health Sciences Center in New Orleans, physiology and neuroscience professor Michele Meneray uses the Kinetworks screen to examine signal transduction pathways involved in a mouse model of Sjogren syndrome. "I could spend the rest of my life looking at each one of those individual protein kinases. It takes a long time to work out a blot for one antibody," she says. "I'm very pleased with it. We've gotten an enormous amount of information out of a relatively small amount of samples, and very good data analysis."
Kinexus charges its customers two different fees, depending on whether they want their data to remain confidential. Those who don't mind sharing pay about $750 per sample; customers who opt for confidentiality pay roughly double that amount.
With the shared data, Kinexus is compiling an online database called KiNET, which so far has roughly one million data points. When KiNET is unveiled, which Pelech estimates will be late this spring, subscribers will pay about $1,000 per year, and will be able to compare proteins involved in kinase-mediated signal transduction across hundreds of different model systems in an ever-expanding compendium of data.
"Kinexus plans to apply various bioinformatics analyses to KiNET to map the composition and architecture of cell-signaling systems in diverse tissues and cell lines, in order to create an extensive atlas of these regulatory pathways," says Pelech. "We believe that this will help define useful disease markers and suitable therapeutic drug targets."