A small molecule–kinase interaction map for clinicalkinase inhibitorsMiles A Fabian1,3, William H Biggs III1,3, Daniel K Treiber1,3, Corey E Atteridge1, Mihai D Azimioara1,2,Michael G Benedetti1,3, Todd A Carter1, Pietro Ciceri1, Philip T Edeen1, Mark Floyd1, Julia M Ford1,Margaret Galvin1, Jay L Gerlach1, Robert M Grotzfeld1, Sanna Herrgard1, Darren E Insko1, Michael A Insko1,Andiliy G Lai1, Jean-Michel Le´lias1, Shamal A Mehta1, Zdravko V Milanov1, Anne Marie Velasco1,Lisa M Wodicka1, Hitesh K Patel1, Patrick P Zarrinkar1& David J Lockhart1Kinase inhibitors show great promise as a new class of therapeutics. Here we describe an efficient way to determine kinaseinhibitor specificity by measuring binding of small molecules to the ATP site of kinases. We have profiled 20 kinase inhibitors,including 16 that are approved drugs or in clinical development, against a panel of 119 protein kinases. We find thatspecificity varies widely and is not strongly correlated with chemical structure or the identity of the intended target. Many novelinteractions were identified, including tight binding of the p38 inhibitor BIRB-796 to an imatinib-resistant variant of the ABLkinase, and binding of imatinib to the SRC-family kinase LCK. We also show that mutations in the epidermal growth factorreceptor (EGFR) found in gefitinib-responsive patients do not affect the binding affinity of gefitinib or erlotinib. Our resultsrepresent a systematic small molecule-protein interaction map for clinical compounds across a large number of related proteins.Protein kinases are critical components of cellular signal transductioncascades. They are directly involved in many diseases, including cancerand inflammation, and have become one of the most important targetclasses for drug development1,2. The approval of imatinib (Gleevec) forchronic myeloid leukemia (CML), and gefitinib (Iressa) and erlotinib(Tarceva) for non-small cell lung cancer (NSCLC) has provided proof-of-principle that small molecule kinase inhibitors can be effectivedrugs. Over 30 kinase inhibitors are currently in clinical development,and many more are in preclinical studies. The vast majority of thesecompounds target the kinase ATP site, and because all of the morethan 500 protein kinases identified in the human genome have an ATPsite3, there is great potential for cross-reactivity. Compounds must betested experimentally against many kinases to assess molecular speci-ficity and to identify off-target interactions4,5. Binding specificity andaffinity are not readily predicted based on available sequence orstructural information, and conventional profiling methods based onin vitro activity are limited by the difficulty of building and runninglarge numbers of kinase activity assays.We describe an experimental approach to assessing the specificity ofkinase inhibitors that directly and quantitatively measures binding tothe ATP site. Importantly, the method does not require chemicallinking, labeling or immobilization of tested compounds (Supple-mentary Notes online). The approach circumvents many of thedifficulties of conventional enzyme activity assays and allows rapiddevelopment and efficient use of assays for large numbers of kinases.We have applied the technology to develop assays for 113 distinctprotein kinases and six of the clinically observed imatinib-resistantvariants of the ABL kinase, and have determined quantitative bindingprofiles for 20 kinase inhibitors, including staurosporine, imatinib,gefitinib and 14 compounds which are now or have been in clinicaldevelopment. We have also assessed the effect of nine recentlyidentified gefitinib-sensitizing EGFR mutations6,7on the interactionof EGFR with eight known EGFR inhibitors, including gefitinib anderlotinib. The data constitute a small molecule–kinase interaction mapand represent a systematic exploration of binding behavior of clinicalcompounds across a large protein class.RESULTSBinding assays for small molecule–kinase interactionsThe approach uses ATP site–dependent competition binding assays(Fig. 1a)8. The key assay components are human kinases expressed asfusions to T7 bacteriophage and a small set of immobilized probeligands that bind to the ATP site of one or more kinases. The small setof immobilized ligands is used to build the assays, but the ‘free’ testcompounds (e.g., the 20 molecules profiled here) are not linked,labeled or immobilized. The kinases used in the assays can be viewedas fusion proteins that are tagged to facilitate expression, purificationand detection. In this scheme the T7 phage particle is not unlike morePublished online 13 February 2005; doi:10.1038/nbt10681Ambit Biosciences, 4215 Sorrento Valley Blvd., San Diego, California 92121, USA.2Present addresses: Genomics Institute of the Novartis Research Foundation, 10675John Jay Hopkins Dr., San Diego, California 92121, USA (M.D.A.), Buck Institute, 8001 Redwood Blvd., Novato, California 94945, USA (M.G.B.), SeattleBiomedicalResearch Institute, 307 Westlake Ave. N., Ste. 500, Seattle, Washington 98109, USA (J.L.G.) and Metabasis Therapeutics, 9390 Towne Centre Dr., San Diego, California92121, USA (M.A.I.).3These authors contributed equally to this work. Correspondence should be addressed to D.J.L. ([email protected]) or P.P.Z.([email protected]).NATURE BIOTECHNOLOGY VOLUME 23 NUMBER 3 MARCH 2005 329ARTICLES© 2005 Nature Publishing Group http://www.nature.com/naturebiotechnologyconventional protein tags such as glutathione S-transferase or greenfluorescent protein, except that it renders the attached proteinamplifiable and amenable to very sensitive and versatile detection.Kinases that have been cloned into the phage vector can be producedrapidly by simply growing phage in Escherichia coli.T7phagereplication leads to lysis of the bacterial host, and lysates containingproperly folded, tagged kinases are used directly in the assay with noneed for conventional protein purification. The small number ofimmobilized small molecule ligands used to build the assays bindkinases with high affinity (Kdo 1 mM), and were amenable toattachment of biotin through a flexible chemical linker. For the assay,tagged kinases and immobilized ‘bait’ ligands were combined with the‘free’ test compound (Fig. 1a). If the ‘free’ test compound binds thekinase and directly or indirectly occludes the ATP site, fewer proteinmolecules bind the immobilized ligand on the solid support. If the‘free’ test compound does