March 15, 2007 (Vol. 27, No. 6)
As Scientists Grasp the Complex Biology, Success Finally Seems Possible
In the early 1980s, scientists started to make connections between kinases, general cell proliferation, and promotion of cancer growth. One of the early reviews on the subject appeared in 1982. This paper (Weinstein, I.B., et al., “Results and speculations related to recent studies on mechanisms of tumor promotion”) made the first connections between kinase activity and growth of tumor cells. Over the course of the next few years, scientific knowledge about kinases continued to evolve, prompting pharmaceutical companies to invest research dollars on kinases as potential targets for pharmacological intervention.
Two decades after the chatter about kinases began, research efforts culminated in the approval of Gleevec®. Gleevec, also known as imanitib tosylate, proved to be a spectacular success—addressing an important unmet medical need and bringing Novartis(www.novartis.com) more than $1 billion a year in revenue.
Gleevec inhibits activity of a nonreceptor tyrosine kinase, ABL, which drives proliferation of cells in chronic myelogenous leukemia. These cells contain the so-called Philadelphia chromosome (Ph) with t(9;22) translocation where a part of the abl gene is fused to a part of the bcr gene, resulting in expression of the hybrid BRC/ABL protein with constitutive kinase activity.
Treatment with Gleevec not only suppresses the kinase activity of the fusion protein, but also inhibits proliferation of the cells. In some cases, Gleevec even eliminates all cells containing the Philadelphia chromosome from patient’s blood and bone marrow, leading to apparent cures of the disease.
Based on the success of Gleevec, what are the prospects for other kinase inhibitors? From the data generated by the sequencing of the human genome we know that there are no less than 500 kinases. More then half are now cloned, expressed, and available to evaluate activity of potential inhibitors. Currently there are only seven kinase inhibitors that are approved drugs on the market.
Three of these inhibitors were designed to target BCR/ABL, two others at epidermal growth factor receptor 1 (EGFR-1), one at RAF, and one at the kinase insert domain receptor (KDR) or the vascular endothelial growth factor receptor 2 (VEGFR-2). Thus far, four kinase targets have proved to be useful for therapeutic interventions. Why so few? Are researchers doing something wrong and, if so, how can we fix it?
We have to keep in mind that the drug discovery and development process is extremely difficult due to our poor understanding of biology of the disease and biology of the host (i.e., Homo sapiens). We are making steady progress, but there is still a long way to go. The seven marketed kinase inhibitors cited in this article, in theory, target four kinases in total. In reality, these inhibitors actually target at least 17 kinases, but only the four are thought to contribute to therapeutic effect. Considering this, how targeted should a targeted therapy be? Should a compound targeting three kinases (e.g., Gleevec) be in the category of targeted therapies?
Most kinase inhibitors bind to the ATP binding sites of their targets in a competitive manner. Two decades ago chemists were extremely skeptical about the possibility of making any selective ATP-competitive kinase inhibitors. The rationale behind their skepticism was that all kinases bind the same ATP molecule, so their ATP-binding sites must be similar, if not identical, and therefore it would be impossible to find a molecule that will compete for the ATP binding of one kinase but not the other.
Well, it has since been demonstrated that it is possible to make reasonably selective ATP-competitive kinase inhibitors, but in many cases it was also difficult to “dial out” some extra activities (e.g., KDR inhibition from RAF inhibitors, and platelet derived growth factor inhibition from ABL inhibitors). Thus, at least for now, the current classes of kinase inhibitors are multitargeted.
In principle this can be an advantage, provided that the inhibition of the “extra” targets adds to the therapeutic value of the compound, rather than to the unwanted side effects. Kinase profiling, both in vitro and in vivo, is emerging as a critical step in the drug development process due to the importance of knowing the exact kinase inhibition profile for every compound entering clinical trials.
There is now a conscious effort in the drug industry to develop active site and allosteric kinase inhibitors in parallel to those competing with the ATP in order to improve selectivity and combat possible drug resistance. However, the vast legacy libraries of the ATP-competitive inhibitors will continue to tempt chemists for some time with the promise of easy success in the search for inhibitors of new kinase targets.
Another issue that chemists face is the inability of biologists to answer deceptively simple questions such as: what kinases should a chemist avoid inhibiting when optimizing leads? The scientific community’s limited knowledge of biology does not allow us an unequivocal answer to this question. Thus, chemists are left with no choice but to try to get an inhibitor that is as “clean” as possible and not to worry about potential side effects, except for those flagged by safety pharmacology assays.
This strategy helped in the development of Gleevec and a BCR/ABL inhibitor from Bristol-Myers Squibb (www.bms.com), Sprycel®. When Gleevec was initially shown to inhibit c-kit (also known as Stem Cell Factor), the concern was that this extra activity would cause serious side effects in the clinic. Fortunately, c-kit inhibition did not cause any clinical problems and, to the contrary, allowed Novartis to expand the indications of Gleevec into gastrointestinal tumors, which are primarily driven by the c-kit activity.
Similarly, c-SRC inhibition by Sprycel proved not to be a limitation for this compound and may even have some beneficial effects for patients. Similarly, Sutent®, an intended KDR inhibitor from Sugen (now Pfizer), seems to be more effective due to the fact that it hits multiple kinases. One can ask then: do we need multitargeted kinase inhibitors in order for them to be effective? Generally speaking, the answer appears to be yes.
Multicellular organisms, as well as single cells that comprise these organisms, evolved to survive hostile environments. Thus, there is significant extra capacity and redundancy built into these systems. In many cases severe compromising of the activity of single target, or even total elimination of the target, will have no effect on a cell. At the same time compromising the activity of two or more individually dispensable targets can be lethal to the cell. This concept of synthetic lethality is not entirely new, but the concept has not been easy to apply to industrial drug discovery. Nonetheless, rational combinations are on the rise, and the day when a combination of two individually inactive drugs will enter clinical trials is not far away.
Better tools to assess selectivity and cross-reactivity of kinase inhibitors, as well as rapid approaches to determining efficacy in animal models, are currently being developed. These tools include a platform to profile multiple kinase inhibitor molecules against 48 different kinases simultaneously—providing high-quality enzyme inhibition data that will allow individual therapeutic groups to profile lead compounds for selectivity and cross-reactivity quickly and easily.
Additionally, the ability to rapidly test compounds for efficacy in vivo is enabled by visible light imaging platforms, which have already played a key role in the early-stage determination of the pharmacodynamics of lead kinases inhibitors, such as Sutent and Sprycel. The future for kinase inhibitors as targeted and multitargeted therapeutics appears bright. As our understanding of the complex biology catches up to chemistry capabilities, it should be possible to continually improve these molecules and significantly impact disease.