Search

Recent advances in targeted protein degradation have enabled chemical hijacking of the ubiquitin-proteasome system to treat disease. The catalytic rate of cereblon (CRBN)-dependent bifunctional degradation activating compounds (BiDAC), which recruit CRBN to a chosen target protein, resulting in its ubiquitination and proteasomal degradation, is an important parameter to consider during the drug discovery process. In this work, an in vitro system was developed to measure the kinetics of BRD4 bromodomain 1 (BD1) ubiquitination by fitting an essential activator kinetic model to these data. The affinities between BiDACs, BD1, and CRBN in the binary complex, ternary complex, and full ubiquitination complex were characterized. Together, this work provides a new tool for understanding and optimizing the catalytic and thermodynamic properties of BiDACs.

Protein kinases are major drug targets, but the development of highly-selective inhibitors has been challenging due to the similarity of their active sites. The observation of distinct structural states of the fully-conserved Asp-Phe-Gly (DFG) loop has put the concept of conformational selection for the DFG-state at the center of kinase drug discovery. Recently, it was shown that Gleevec selectivity for the Tyr-kinases Abl was instead rooted in conformational changes after drug binding. Here, we investigate whether protein dynamics after binding is a more general paradigm for drug selectivity by characterizing the binding of several approved drugs to the Ser/Thr-kinase Aurora A. Using a combination of biophysical techniques, we propose a universal drug-binding mechanism, that rationalizes selectivity, affinity and long on-target residence time for kinase inhibitors. These new concepts, where protein dynamics in the drug-bound state plays the crucial role, can be applied to inhibitor design of targets outside the kinome.eLife digestThe Ser/Thr kinase Aurora A is an important target for the development of new anticancer therapies. A longstanding question is how to specifically and effectively inhibit only this kinase in a background of over 550 protein kinases with very similar structures. To this end, understanding the inhibition mechanism of Aurora A by different drugs is essential. Here, we characterize the kinetic mechanism of three distinct kinase drugs, Gleevec (Imatinib), Danusertib (PHA739358) and AT9283 (Pyrazol-4-yl Urea) for Aurora A. We show that inhibitor affinities do not rely exclusively on the recognition of a specific conformation of the Asp-Phe-Gly loop of the kinase. Our quantitative kinetics data put forward an opposing mechanism in which a slow conformational change after drug binding (i.e., induced-fit step) dictates drug affinity.

Kinases perform phosphoryl-transfer reactions in milliseconds; without enzymes, these reactions would take about 8,000 years under physiological conditions. Despite extensive studies, a comprehensive understanding of kinase energy landscapes, including both chemical and conformational steps, is lacking. Here we scrutinize the microscopic steps in the catalytic cycle of adenylate kinase, through a combination of NMR measurements during catalysis, pre-steady-state kinetics, molecular-dynamics simulations and crystallography of active complexes. We find that the Mg(2+) cofactor activates two distinct molecular events: phosphoryl transfer (>10(5)-fold) and lid opening (10(3)-fold). In contrast, mutation of an essential active site arginine decelerates phosphoryl transfer 10(3)-fold without substantially affecting lid opening. Our results highlight the importance of the entire energy landscape in catalysis and suggest that adenylate kinases have evolved to activate key processes simultaneously by precise placement of a single, charged and very abundant cofactor in a preorganized active site.

For many proteins, especially for molecular motors and other enzymes, the functional mechanisms remain unsolved due to a gap between static structural data and kinetics. We have filled this gap by detecting structure and kinetics simultaneously. This structural kinetics experiment is made possible by a new technique, (TR) 2 FRET (transient time-resolved FRET), which resolves protein structural states on the submillisecond timescale during the transient phase of a biochemical reaction. (TR) 2 FRET is accomplished with a fluorescence instrument that uses a pulsed laser and direct waveform recording to acquire an accurate subnanosecond time-resolved fluorescence decay every 0.1 ms after stopped flow. To apply this method to myosin, we labeled the force-generating region site specifically with two probes, mixed rapidly with ATP to initiate the recovery stroke, and measured the interprobe distance by (TR) 2 FRET with high resolution in both space and time. We found that the relay helix bends during the recovery stroke, most of which occurs before ATP is hydrolyzed, and two structural states (relay helix straight and bent) are resolved in each nucleotide-bound biochemical state. Thus the structural transition of the force-generating region of myosin is only loosely coupled to the ATPase reaction, with conformational selection driving the motor mechanism.

Spin-labeling and multifrequency EPR spectroscopy were used to probe the dynamic local structure of skeletal myosin in the region of force generation. Subfragment 1 (S1) of rabbit skeletal myosin was labeled with an iodoacetamide spin label at C707 (SH1). X- and W-band EPR spectra were recorded for the apo state and in the presence of ADP and nucleotide analogs. EPR spectra were analyzed in terms of spin-label rotational motion within myosin by fitting them with simulated spectra. Two models were considered: rapid-limit oscillation (spectrum-dependent on the orientational distribution only) and slow restricted motion (spectrum-dependent on the rotational correlation time and the orientational distribution). The global analysis of spectra obtained at two microwave frequencies (9.4 GHz and 94 GHz) produced clear support for the second model and enabled detailed determination of rates and amplitudes of rotational motion and resolution of multiple conformational states. The apo biochemical state is well-described by a single structural state of myosin (M) with very restricted slow motion of the spin label. The ADP-bound biochemical state of myosin also reveals a single structural state (M*, shown previously to be the same as the post-powerstroke ATP-bound state), with less restricted slow motion of the spin label. In contrast, the extra resolution available at 94 GHz reveals that the EPR spectrum of the S1.ADP.V(i)-bound biochemical state of myosin, which presumably mimics the S1.ADP.P(i) state, is resolved clearly into three spectral components (structural states). One state is indistinguishable from that of the ADP-bound state (M*) and is characterized by moderate restriction and slow motion, with a mole fraction of 16%. The remaining 84% (M**) contains two additional components and is characterized by fast rotation about the x axis of the spin label. After analyzing EPR spectra, myosin ATPase activity, and available structural information for myosin II, we conclude that post-powerstroke and pre-powerstroke structural states (M* and M**) coexist in the S1.ADP.V(i) biochemical state. We propose that the pre-powerstroke state M** is characterized by two structural states that could reflect flexibility between the converter and N-terminal domains of myosin.

Protein phosphorylation is an essential regulatory mechanism that affects all aspect of cellular life from division and growth to aging and death. Misregulation of the signaling cascades leads to severe detrimental effects, and in humans often associated with cancer and other diseases. Phosphorylation is performed by a class of protein called kinases. Activation and deactivation of kinases is normally under tight control and is regulated via different mechanisms that are incredibly complex. In this work we combine phylogenetic resurrection techniques with biophysical and chemical approaches to analyze the regulatory mechanisms of modern tyrosine oncokinases Src and Abl, their common ancestor and the common ancestors between several other families of tyrosin kinases. Our results show how the regulatory elements appeared and developed throughout the evolution enabling selective regulation of complex modern cascades.

Proteins populate a landscape of conformations that changes upon binding native ligands or drugs, altering protein function. Different conformations reveal distinct potential drug target sites, including allosteric sites distal to the active site of the protein. It has been historically difficult to identify allosteric modulators early in the screening process since the structural methods required to reveal binding sites are relatively low throughput.

We are using site-directed spin-labeling (SDSL) and dipolar electron-electron resonance (DEER) to study the structural dynamics of proteins involved in muscle function. Muscle proteins are ideal systems for DEER applications, since these proteins form large dynamic assemblies that are inaccessible to crystallography and NMR. DEER is a pulsed ELDOR technique that detects the spin-spin dipolar interaction as a decay and modulation of a spin echo signal. While the CW EPR spectrum is sensitive only to distances on the order of 1 to 2 nm, DEER is sensitive from 1.5 to 5 nm and more. DEER is sensitive to the width of the distance distribution and can resolve multiple structural populations simultaneously. However, a limitation of DEER is the use of spin probes with flexible linkers, which adds uncertainty to the locations of protein backbone atoms. A bifunctional spin label (BSL), which attaches at position (i, i+4) of a helix, was used to measure the distance between two di-Cys pairs. BSL is rigidly coupled to the protein backbone and is a breakthrough for high resolution measurement of orientation, dynamics, and distance. These distance measurements were compared to conventional labeling of cysteine residues with MSL/MSL pairs.

We have used transient time-resolved FRET (TR2-FRET) to detect structural changes within the relay helix in real time during the myosin recovery stroke. The most intriguing problem in myosin's function is the mechanism of signal transduction between the nucleotide binding site and the force-generating domain of myosin during the recovery stroke. Atomic structures of myosin provide clues, but structural measurements during transient kinetics are needed to reveal the details of this mechanism. Kinetics of nucleotide binding is reliably determined from the fluorescence of mant nucleotides or single Trp myosin mutants (W129, W131). Single Trp mutant W501 has provided insight into structural kinetics in the force-generating domain, but the complex fluorescence signals observed during the recovery stroke do not yield clear structural interpretation. To overcome this problem, we have performed FRET measurements that monitor well-defined structural changes in the relay helix, which is proposed to be a key structural element in myosin's force-generating domain. We engineered two double-Cys myosin mutants in the Cys-lite D. discoideum myosin construct (gift from J. Spudich), with one Cys placed at the C-terminus of the relay helix (K498C) and the other at one of two stable positions in the lower 50K domain (D515C or A639C). We labeled these mutants with a donor-acceptor FRET pair (IAEDANS-DABSYL), then used time-resolved FRET (TR-FRET, detects donor lifetime at nanosecond timescale) to detect interprobe distances in equilibrium states trapped by nucleotide analogs, reflecting pre- and post-powerstroke structures of myosin. We then performed these nanosecond time-resolved measurements repetitively every 0.1 ms, in the transient phase after ATP addition using stopped flow. These transient time-resolved FRET (TR2-FRET) signals showed directly that the relay helix changes from straight to bent conformation in the recovery stroke, from the post-powerstroke to the pre-powerstroke state.

We have engineered a mutant of Dictyostelium discoideum ( Dicty ) myosin II that contains the same fast-reacting “SH1” thiol as in muscle myosin, spin-labeled it, and performed electron paramagnetic resonance (EPR) to compare the structure of the force-generating region of the two myosins. Dicty myosin serves as a model system for muscle myosin because of greater ease of mutagenesis, expression, and crystallization. The catalytic domains of these myosins have nearly identical crystal structures in the apo state, but there are significant differences in ATPase kinetics, and there are no crystal structures of skeletal muscle myosin with bound nucleotides, so another structural technique is needed. Previous EPR studies, with a spin label attached to SH1 in muscle myosin, have resolved the key structural states of this region. Therefore, we have performed identical experiments on both myosins spin-labeled at equivalent sites. Spectra were identical for the two myosins in the apo and ADP-bound states. With bound ADP and phosphate analogs, ( i ) both proteins exhibit two resolved structural states (prepowerstroke, postpowerstroke) in a single biochemical state (defined by the bound nucleotide), and ( ii ) these structural states are essentially identical in the two myosins but ( iii ) are occupied to different extents as a function of the biochemical state. We conclude that ( i ) myosin structural and biochemical states do not have a one-to-one correspondence, and ( ii ) Dicty myosin can serve as a good analog for structural studies of muscle myosin only if differences in the coupling between biochemical and structural states are taken into account.

Nitroxide radicals are widely used as molecular probes in different fields of chemistry and biology. In this work, we describe pH-sensitive imidazoline- and imidazolidine-based nitroxides with pK values in the range 4.7-7.6 (2,2,3,4,5,5-hexamethylperhydroimidazol-1-oxyl, 4-amino-2,2,5,5-tetramethyl-2,5-dihydro-1H-imidazol-1-oxyl, 4-dimethylamino-2,2-diethyl-5,5-dimethyl-2,5-dihydro-1H-imidazol-1-oxyl, and 2,2-diethyl-5,5-dimethyl-4-pyrrolidyline-1-yl-2,5-dihydro-1H-imidazol-1-oxyl), which allow the pH-monitoring inside chloroplasts. We have demonstrated that EPR spectra of these spin-probes localized in the thylakoid lumen markedly change with the light-induced acidification of the thylakoid lumen in chloroplasts. Comparing EPR spectrum parameters of intrathylakoid spin-probes with relevant calibrating curves, we could estimate steady-state values of lumen pHin established during illumination of chloroplasts with continuous light. For isolated bean (Vicia faba) chloroplasts suspended in a medium with pHout=7.8, we found that pHin approximately 5.4-5.7 in the state of photosynthetic control, and pHin approximately 5.7-6.0 under photophosphorylation conditions. Thus, ATP synthesis occurs at a moderate acidification of the thylakoid lumen, corresponding to transthylakoid pH difference DeltapH approximately 1.8-2.1. These values of DeltapH are consistent with a point of view that under steady-state conditions the proton gradient DeltapH is the main contributor to the proton motive force driving the operation of ATP synthesis, provided that stoichiometric ratio H+/ATP is nor =4-4.7.

We have developed a high-throughput time-resolved fluorescence spectrometer capable of recording a high-resolution time-domain (sub-nanosecond) fluorescence decay, with high S/N every 100 μs. Coupled with a conventional stopped-flow rapid mixer, this technology has allowed us to measure changes in time-resolved fluorescence decays occurring during the course of a millisecond-resolved biochemical transient experiment. Most instruments used in fluorescence-based kinetic studies are limited to detecting a single fluorescence intensity signal on the millisecond time scale. While this type of fluorescence intensity-based measurement is informative, it provides scant information compared to a full sub-nanosecond resolved fluorescence decay- which is exquisitely sensitive to the structure, dynamics, and interactions of molecules in the sample. Current fluorescence lifetime technology is too slow by a factor of at least 1000 to detect this decay accurately within the millisecond timescale of a typical biochemical transient. Our system, based on a direct acquisition data collection approach, records the entire fluorescence decay with S/N > 100 from a single pulse form a 10 kHz microlaser. Using this approach, we can resolve individual fluorescence lifetime components comprising as little as 10% of a complex multi-exponential fluorescence decay. When used to monitor structural transitions by time resolved FRET, we are capable of isolating individual structural states within complex structural ensembles. Using this approach we were able to simultaneously monitor the binding of myosin to different classes of binding sites on actin filaments. This type of analysis is not possible with conventional fluorescence based stopped flow measurements.

We have used pulsed EPR spectroscopy (DEER) to determine the structural effects of point mutations in the force generation region of myosin (I499A, F506A, F692A, D.discoideum sequence). As it was previously reported, these myosin mutants maintain basal ATPase activity, but completely lose their motor function. It was proposed that F506A, located in the relay loop, disrupts communication between the nucleotide binding site and the force generation region in myosin, and F692A and I499A, located on the interface with the converter domain, affect coordination of the converter domain relative to the myosin head. In this study we assayed the structure of the relay helix in these mutants and in wild-type myosin. The mutations were introduced into the A639C:K498C myosin construct, both Cys were labeled with maleimide spin probes, and interprobe distance was measured. Previous studies showed that probes at these sites detect the nucleotide-induced bending of the relay helix. Observed spin echo decays were interpreted in terms of one or two Gaussian distance distributions, corresponding to one or two myosin structural states. Nucleotide analogs were used to trap myosin in several biochemical states. In the wild type, two structural states, corresponding to pre- and post-recovery stroke, are present in a single biochemical state. The mutations affect the equilibrium between these two structural states and increase the conformational disorder, indicating that these mutants destabilize the structure of the force-generation region, particularly in the post-recovery state. These results provide a structural explanation for the functional perturbation, introduced to myosin.

We have used electron paramagnetic resonance (EPR) of a bifunctional spin label (BSL) to measure structural transitions of the catalytic domain (CD) in Dictyostelium myosin II. The use of BSL is a critical feature in this work. The bifunctional attachment of BSL eliminates most of the nanosecond motions characteristic of monofunctional labels, making it possible to measure protein structural transitions with a precision not previously achievable. Two double-Cys constructs were engineered with Cys residues at helical locations i and i+4 (494.498 and 639.643). Residues 494.498 are located on the relay helix and residues 639.643 are located on a helix adjacent to the relay helix. After BSL labeling, myosin was bound to actin in oriented muscle fibers, making it possible to measure helix orientation relative to the fiber axis. Spectra were acquired in APO and ADP states. Simulation of the 494.498.BSL spectra demonstrates that in APO and ADP states there are two highly ordered populations of the relay helix. APO and ADP spectra contain the same two spectral components, but the distribution of these two components differs between states. Similarly, spectra from the 639.643 construct reveal an ADP-induced structural transition, but the difference between APO and ADP spectra demonstrates a 3o rotation of the 639.643 helix. Our results demonstrate structural transitions of two helixes within the CD of actin-bound myosin associated with nucleotide binding. Measuring these transitions is essential to understanding the molecular mechanism of force generation. This is particularly true for the relay helix, which plays a key role in the coupling of myosin ATPase and motor function. Additionally, these results demonstrate the utility of BSL for measuring transitions in protein orientation, order, and dynamics.

We have used transient time-resolved FRET (TR2FRET) to monitor the conformation of the relay helix in a myosin II functional mutant during the recovery stroke in real time. Myosin was perturbed with the F506A mutation (Dictyostelium discoideum sequence), located within the relay loop in the force-generating region. F506 is a highly conserved residue in myosin II and is a hypertrophic cardiomyopathy mutation site. Previous studies [Tsiavaliaris, EMBO Rep, 2002, 3(11), 1099] showed a significant effect of the F506A mutation on myosin function. Actin affinity in the presence of ATP was increased, and the mutant did not move actin filaments in in vitro motility assays. A small decrease in intrinsic fluorescence was observed upon addition of excess ATP, but ATP binding and hydrolysis were not affected by the mutation. It was proposed that the F506A disrupts the communication between the active site and the lever arm. We engineered a double-Cys myosin mutant (A639C:K498C) in the Cys-less background with the F506A functional mutation, and labeled the mutant with optical probes. We used TR-FRET to determine the interprobe distance, and TR2FRET measurements after rapid mixing with ATP revealed changes in the relay helix conformation during the recovery stroke in real time. The mutation induced significant disorder of the relay helix in the force-generating region, but myosin still produces a recovery stroke, changing the relay helix conformation from straight to bent. We conclude that (a) the relay helix is disordered in myosin functional mutant F506A, which demonstrates the importance of the relay loop - relay helix interaction in the relay helix stabilization, and (b) the relay helix is the major structural element in the force-generating region of myosin, responsible for communication from the active site to the converter domain and the lever arm.

We have used dipolar electron-electron resonance (DEER) and time-resolved fluorescence resonance energy transfer (TR-FRET) to investigate the role of the myosin relay helix in coupling between the active site and the force-generating region of myosin II. Two double-Cys Dictyostelium myosin constructs have been engineered, and the structure of the relay helix was monitored by measuring interprobe distances in MSL/MSL or IAEDANS/Dabcyl-labeled myosin. Experiments were performed on WT myosin and on F506A, a functional mutant that has close to normal enzymatic activity but completely lacks motor functions (e.g. unable to move actin filaments or support cell development). We found that the WT myosin relay helix adopts two distinct states (straight and bent), with the bent conformation populated when ATP and ADP.Pi analogs are bound at the active site. In contrast, binding of nucleotide analogs had very little effect on relay helix conformation in the F506A mutant. In addition, the width of the distance distribution was significantly larger in the F506A compared to WT myosin, indicating loss of structural organization. Our results demonstrate that the relay helix plays a key role in coupling of myosin ATPase and motor activities. Loss of functionality observed in F506A myosin can be explained by the disruption of the relay helix-relay loop interactions that normally stabilize well-defined conformations of the myosin force-generating region allowing it to switch between distinct structural states.

We have used three complementary spectroscopic techniques, pulsed EPR (DEER), time-resolved fluorescence resonance energy transfer (TR-FRET) and transient time-resolved FRET [(TR)2FRET], to elucidate the mechanisms of energy transduction in myosin - an important motility protein involved in a variety of processes including muscle contraction. We engineered two double-Cys myosin mutants that were labeled with optical probes or spin labels, and determined structural changes in a single structural element of the myosin motor domain, the relay helix, as affected by nucleotide binding and hydrolysis. In all cases, time-domain detection permitted the reliable determination of the interprobe distance distribution, quantitating both order and disorder, and resolving coexisting structural states. While DEER offered superior distance resolution, (TR)2FRET permitted detection of transient structural changes after rapid mixing (stopped-flow) with ATP. All methods demonstrated two structural states of myosin during the recovery stroke, corresponding to straight and bent conformations of the relay helix, in parallel with straight and bent conformations of the entire myosin head. A narrower interprobe distance distribution in the post-recovery state shows ordering of the relay helix structure during the recovery stroke. These experiments reveal changes in both structure and dynamics of the relay helix and identify it as a key player in the interdomain coupling mechanism. This methodology is applicable to any enzyme in which double Cys mutants can be engineered in a key region of energy transduction, permitting resolution of structural states and dynamics in real time during the transient phase of a biochemical reaction.

We have used DEER (double electron-electron resonance) and TR-FRET (time-resolved fluorescence resonance energy transfer) to study conformational changes within the myosin II relay helix. Major structural changes during the myosin II ATPase cycle take place in the force-generating domain. Crystal structures show that the converter domain, the relay helix, and SH1 helix have different conformations in the proposed pre- and post- powerstroke structural states of myosin. This dramatic structural change is detected in solution studies by intrinsic fluorescence [1, 2] and EPR of a spin label attached to SH1 [3]. In the present study, we focus on the relay helix as a crucial structural element involved in coupling between the force-generating domain and the nucleotide-binding pocket. Cysteine mutations were introduced into a Cys-lite construct of Dictyostelium discoideum (Dicty) myosin in the lower 50k domain (either D515C or A639C) and the C-terminal end of the relay helix (K498C). These constructs were selectively modified with either MSL/MSL or IAEDANS/DABCYL pairs, and the distance between probes was measured in different myosin conformations trapped with nucleotides or nucleotide analogs Two conformations of the relay helix (with distinct probe-to-probe distances, presumably corresponding to the “straight” and “bent“ states of the relay helix) were resolved. Observed distances were in good agreement with existing crystal structures, but at least two distinct structural states were present in certain biochemical states (e.g, with bound ADP.BeF3, ADP.Vi, ADP.AlF4). The mole fraction of the “bent” conformation was higher with post-hydrolysis analogs (ADP.Vi, ADP.AlF4) bound at the active site. Our results reveal structural rearrangements within a single subdomain of myosin and provide insights into the coupling between ATP binding and changes in the force-generating region.

SummaryMolecular recognition plays a central role in biology, and protein dynamics has been acknowledged to be important in this process. However, it is highly debated whether conformational changes happen before ligand binding to produce a binding-competent state (conformational selection) or are caused in response to ligand binding (induced fit). Proposals for both mechanisms in protein/protein recognition have been primarily based on structural arguments. However, the distinction between them is a question of the probabilities of going via these two opposing pathways. Here, we present a direct demonstration of exclusive conformational selection in protein/protein recognition by measuring the flux for rhodopsin kinase binding to its regulator recoverin, an important molecular recognition in the vision system. Using nuclear magnetic resonance (NMR) spectroscopy, stopped-flow kinetics, and isothermal titration calorimetry, we show that recoverin populates a minor conformation in solution that exposes a hydrophobic binding pocket responsible for binding rhodopsin kinase. Protein dynamics in free recoverin limits the overall rate of binding.

Catalytic function of many enzymes is comprised of a number of microscopic steps including enzyme structural rearrangements and the chemical steps, in which bonds of the substrate(s) are broken/made to synthesize the product(s). Separation of these processes is a challenging task that hinders our understanding of enzyme catalysis. We have used nuclear magnetic resonance (NMR) and rapid-quenching techniques to independently study these processes in adenylate kinase (ADK) and to characterize their energetic contribution at atomic resolution. Adenylate kinase is an important enzyme that catalyzes a reversible reaction: 2ADP ↔ ATP +AMP, and is involved in maintaining energy homeostasis in a cell. Two binding sites of ADK and three different ligands result in a variety of biochemical states (determined by the bound ligands). Each of these states has different dynamic properties making the overall structural dynamics of ADK complex. We have tackled this problem by engineering a loss-of-function ADK mutant (with greatly reduced rate of turnover), which allowed us to “trap” the enzyme in well-defined biochemical states. Our results lead to a detailed energy profile of ADK, providing insights into the molecular mechanism of its functioning. This work reveals fundamental principles of enzyme catalysis and highlights the role of protein's intrinsic dynamics.

Cancer is a serious illness and the second leading cause of death in the US. It is caused by misregulation of the signaling cascades and uncontrolled cell proliferation. Protein kinases are the key players in the regulatory machinery, and in the last 20 years they have become an increasingly important drug targets. The main requirement for such drugs is selectivity: ideally, only one kinase should be affected by an individual drug. The prominent success story is an FDA-approved chronic myeloid leukemia therapeutic Gleevec. Gleevec is a highly specific inhibitor of the Abl kinase and has very little potency towards a closely related Src subfamily of kinases. This specificity is particularly intriguing considering that the drug-binding pockets look nearly identical in these kinases. We combined NMR and pre-steady-state kinetic experiments to monitor Gleevec binding in real time and with residue-specific precision. Our data establish a novel model that fully and quantitatively explains the observed 3000-fold difference in Gleevec's affinities to Abl and Src, solving a long-standing paradox in the field. These results reveal the general principles of kinase-drug interactions and highlight the molecular mechanisms behind nanomolar affinities found in clinically relevant therapeutics.