New Kinome Database Promises Wealth of Unexpected Cancer Drug Targets

January 5, 2011

"Kinome" is the word to know this year in oncology, because it has begun to reveal molecules that some tumors are relying on to survive, which until now we had no idea were involved in cancer or which we hardly knew at all.

A word for oncologists to know this year (if they don't already) is "kinome." Analagous to "genome," it's the complete set of kinases in an organism. The human kinases, of course, are the class of enzymes whose inhibitors have proven so promising and so frustrating in recent clinical cancer trials.

Over the past year or so, study of the kinome has revealed molecules that some tumors are relying on to survive, many of which until now we had no idea were involved in cancer or which we hardly knew at all. Numerous reports have suggested novel kinases that might be useful drug targets in particular malignancies. 

In September, for instance, a team from Leiden in the Netherlands reported in Molecular Cancer that they had used kinome analysis to identify two kinases (Src and nuclear factor kappaB) involved in myxoid liposarcoma. Earlier last year, two different teams in the United States found new kinases involved in the development of ovarian cancer and in its response to cis-platin treatment.

From tidbits to a feast

Other teams have been serving up tidbits about new kinases, piecemeal. Now a team from Finland has offered drug developers a feast: A "functional taxonomy" of the kinome, revealing which of 459 kinase proteins show either significant gains or losses of activity in human tumors compared to normal tissues. The data are based on analysis of 5,681 human tissue samples.

As an appetizer, the report by Sami Kilpinen et al in PLoS One  last month provides the identities of several hitherto obscure kinases that appear to be involved in prostate and gynecological cancers as well as some lymphomas. It identifies altogether 22 kinases not previously associated with prostate cancer. Far more is certain to come, because the database is published openly on the Internet.

More than 10,000 patent applications for kinase inhibitors have been filed in the last decade in the United States, because many of the 518 human kinases and other molecules in their pathways are among the most frequently mutated in cancer. Although kinase inhibitors have revolutionized the treatment of a few cancers such as chronic myeloid leukemia (CML) and gastrointestinal stromal tumor (GIST), "clinical progress has been uneven," Zachary Knight of Rockefeller University and coauthors observed in a review about "kinomics" and cancer in Nature Reviews Cancer last February.

Kinase inhibitors have achieved significant survival gains for only a few cancers, because relatively few malignancies are "addicted" to only one kinase mutation for their survival. Many develop resistance to a particular kinase inhibitor by mutating, reverting to a different kinase signaling pathway, or making molecular adjustments downstream of the inhibitor. Rationally designed combinations of kinase inhibitors have failed in malignancy after malignancy, at least in part because of insufficient understanding of the kinome.

Expression patterns match side effects

The mass of information from Kilpinen et al also promises to offer insights into the side effects that have hindered widespread use of some kinase inhibitors, because it shows kinase expression patterns for a wide range of normal tissues. The skin rash associated with epidermal growth factor receptor (EGFR) is echoed in the discovery of elevated transcriptional activity in hair follicles in the Finnish study. ERBB2/HER-2 was shown to be active in healthy tissues including heart, lung, and colorectal tissues, corresponding to the most common side effects of anti-ERBB2 drugs: cardiovascular and respiratory problems and diarrhea.

The dataset also contributes important information by revealing co-expression of different kinases in similar tumors, offering insights for combination therapies. As Knight et al observed in their review of the field, "There are clearly clusters of kinases that tend to be inhibited by similar drugs, but ... there are also many target combinations that should be accessible but remain undiscovered."

They speculate that some pharmaceutical companies already try to optimize drug profiles against complex patterns of kinases, rather than individually, in order to maximize their specificity. The public availability of the Finnish data can only speed this process.