Kevan M. Shokat

Board Member: 2011- 2015

Current Institution
University of California, San Francisco
Department of Cellular and Molecular Pharmacology

Scholar: 1997

Awarded Institution
Princeton University


Research Interests

Research in the lab is highly interdisciplinary utilizing organic chemistry as well as molecular biology to address contemporary problems in chemistry, biology and medicine. Our major goal is to develop novel chemical tools which are complementary to existing genetic and biochemical tools to study cellular signal transduction. We have focused our efforts on the most ubiquitous form of information transfer inside cells which is protein phosphorylation and dephosphorylation. The phosphorylation of amino acid side chains causes proteins to be turned on or off in response to external stimuli. The kinases and phosphatases are the two classes of enzymes which mediate this on/off switch by phosphorylation and dephosphorylation, respectively, of hydroxyl groups in proteins. Tyrosine kinases transfer the g-phosphate from ATP to the phenolic oxygen of tyrosine residues. The tyrosine kinases are particularly important because of their involvement in growth factor response (cancer), inflammation, allergy, autoimmunity, learning, memory, viral infection, and many other normal and abnormal cellular processes.

Hundreds of tyrosine kinases have been identified using powerful genetic and biochemical approaches, yet these tools have failed to identify the most fundamental question about each known tyrosine kinase: what are its direct protein substrates. The two main reasons is difficult to identify the targets of protein kinases are: (1) the number of kinases inside each cell is enormous (it is estimated that 2% of all proteins are kinases) & (2) there is a great degree of overlap in terms of protein target specificity (i.e. nature has built in redundancies in the system of kinases and substrates). These two factors have made the identification of the important downstream targets of protein kinases virtually impossible since the first kinase was discovered over 30 years ago.

The current challenge is to use chemistry to break down the overwhelming redundancy in cellular signal transduction and to develop methods to identify new players in signal transduction cascades which remain obscure. We have developed a molecular tagging system for individual proteins in signal transduction pathways. All protein kinases use ATP as the phosphodonating cofactor in the phosphorylation reaction. Our method involves simultaneous protein engineering and chemical synthesis of unnatural ATP’s (A*TPs) to identify an orthogonal non-natural substrate of an engineered kinase. Such an engineered kinase would be unique in the cell in that it would be the only protein able to use the unnatural A*TP as a phosphodonor. The use of the engineered kinase in conjunction with radiolabeled [gamma-32P]-A*TP would lead to specific radiolabeling (tagging) of the direct substrates of the engineered kinase.

We have demonstrated the feasibility of our approach using the prototypical tyrosine kinase, v-Src, the transforming gene product of the Rous Sarcoma Virus. We have specifically reengineered this protein kinase to accept the unnatural ATP analog, N6-cyclopentyl-ATP, which is not accepted by any other wild-type protein kinase. Our rational engineering of unnatural substrate specificity was accomplished through an iterative process of site-directed mutagenesis of v-Src and synthesis of a panel of 20 ATP analogs in order to identify a suitable engineered protein and unnatural substrate.

Recently we have begun to extend our protein engineering of v-Src to other proteins in this kinase family and to more distantly related protein kinases which are associated with disease (oncogenes, viral transformation, etc.). The goal is to develop protein engineering approaches for the entire protein superfamily such that all the estimated 2000 protein kinases can be engineered to accept an unnatural ATP substrate which would in turn be used to identify the direct substrates of any kinase of interest. This type of protein engineering is extremely challenging because it must be accomplished in cases where only 30% sequence homology exists between two members of a protein family.

Our long term goal is to develop chemical tools which can be readily used to understand the role of unknown genes identified in the human genome sequencing project. This is a huge challenge to chemistry because the number of genes and thereby proteins to study will is expanding enormously as a result of the rapid sequencing of the human genome. Our approach of domain specific protein engineering which is generalizable across entire gene families will greatly accelerate the rate at which the function of individual members of proteins in families can be studied using powerful chemical tools.

A sample of other projects in the lab includes:

Designing approaches to take advantage of pre-programmed cellular regulation we hope to create novel highly selective compounds which modulate the immune system.

Designing intracellular fluorescent sensors of protein kinase activity.