Publications

Inositol-requiring enzyme 1 alpha (IRE1α) is a transmembrane sensor protein with both kinase and ribonuclease activity, which plays a crucial role in the unfolded protein response (UPR). Protein misfolding in the endoplasmic reticulum (ER) lumen triggers dimerization and subsequent trans-autophosphorylation of IRE1α. This leads to the activation of its endoribonuclease (RNase) domain and splicing of the mRNA of the transcriptional activator XBP1, ultimately generating an active XBP1 (XBP1s) implicated in multiple myeloma survival. Previously, we have identified human IRE1α as a target for the development of kinase inhibitors that could modulate the UPR in human cells, which has particular relevance for multiple myeloma and other secretory malignancies. Here we describe the development and validation of a 384-well high-throughput screening assay using DELFIA technology that is specific for IRE1α autophosphorylation. Using this format, a focused library of 2312 potential kinase inhibitors was screened, and several novel IRE1α kinase inhibitor scaffolds were identified that could potentially be developed toward new therapies to treat multiple myeloma.

Two closely related binding modes have previously been proposed for the ATP-competitive benzimidazole class of checkpoint kinase 2 (CHK2) inhibitors; however, neither binding mode is entirely consistent with the reported SAR. Unconstrained rigid docking of benzimidazole ligands into representative CHK2 protein crystal structures reveals an alternative binding mode involving a water-mediated interaction with the hinge region; docking which incorporates protein side chain flexibility for selected residues in the ATP binding site resulted in a refinement of the water-mediated hinge binding mode that is consistent with observed SAR. The flexible docking results are in good agreement with the crystal structures of four exemplar benzimidazole ligands bound to CHK2 which unambiguously confirmed the binding mode of these inhibitors, including the water-mediated interaction with the hinge region, and which is significantly different from binding modes previously postulated in the literature.

Pyrazolopyridine inhibitors with low micromolar potency for CHK1 and good selectivity against CHK2 were previously identified by fragment-based screening. The optimization of the pyrazolopyridines to a series of potent and CHK1-selective isoquinolines demonstrates how fragment-growing and scaffold morphing strategies arising from a structure-based understanding of CHK1 inhibitor binding can be combined to successfully progress fragment-derived hit matter to compounds with activity in vivo. The challenges of improving CHK1 potency and selectivity, addressing synthetic tractability, and achieving novelty in the crowded kinase inhibitor chemical space were tackled by multiple scaffold morphing steps, which progressed through tricyclic pyrimido[2,3-b]azaindoles to N-(pyrazin-2-yl)pyrimidin-4-amines and ultimately to imidazo[4,5-c]pyridines and isoquinolines. A potent and highly selective isoquinoline CHK1 inhibitor (SAR-020106) was identified, which potentiated the efficacies of irinotecan and gemcitabine in SW620 human colon carcinoma xenografts in nude mice.

Deregulation of protein kinases is associated with numerous diseases, making them important targets for drug discovery. The majority of drugs target the catalytic site of these proteins, but due to the high level of similarity within the ATP binding sites of protein kinases, it is often difficult to achieve the required pharmacological selectivity. In this study, we describe the identification and subsequent analysis of water patterns in the ATP binding sites of 171 protein kinase structures, comprising 19 different kinases from various branches of the kinome, and demonstrate that structurally similar binding sites often have significantly different water patterns. We show that the observed variations in water patterns of different, but structurally similar kinases can be exploited in the structure-based design of potent and selective kinase inhibitors.

We describe herein the structure-activity relationship (SAR) and cocrystal structures of a series of Nek2 inhibitors derived from the published polo-like kinase 1 (Plk1) inhibitor (R)-1. Our studies reveal a nonlinear SAR for Nek2 and our cocrystal structures show that compounds in this series bind to a DFG-out conformation of Nek2 without extending into the enlarged back pocket commonly found in this conformation. These observations were further investigated, and structure-based design led to Nek2 inhibitors derived from (R)-1 with more than a hundred-fold selectivity against Plk1.

The HSF1-mediated stress response pathway is steadily gaining momentum as a critical source of targets for cancer therapy. Key mediators of this pathway include molecular chaperones such as heat shock protein (HSP) 90. There has been considerable progress in targeting HSP90 and the preclinical efficacy and signs of early clinical activity of HSP90 inhibitors have provided proof-of-concept for targeting this group of proteins. The HSP70 family of molecular chaperones are also key mediators of the HSF-1-stress response pathway and have multiple additional roles in protein folding, trafficking and degradation, as well as regulating apoptosis. Genetic and biochemical studies have supported the discovery of HSP70 inhibitors which have the potential for use as single agents or in combination to enhance the effects of classical chemotherapeutic or molecularly targeted drugs including HSP90 inhibitors. Here we provide a perspective on the progress made so far in discovering small molecules which target the HSP70 family, in the context of the available structural data and potential issues in drugging this key chaperone.

The phosphatidylinositide 3-kinase (PI3K) pathway is very commonly activated in a wide range of human cancers and is a major driving force in oncogenesis. One of the class I lipid kinase members of the PI3K family, p110alpha, is probably the most commonly mutated kinase in the human genome. Alongside genetic, molecular biological, and biochemical studies, chemical inhibitors have been extremely helpful tools in understanding the role of PI3K enzymes in signal transduction and downstream physiological and pathological processes, and also in validating PI3Ks as therapeutic targets. Although they have been valuable in the past, the early and still frequently employed inhibitors, wortmannin and LY294002, have significant limitations as chemical tools. Here, we discuss the case history of the discovery and properties of an increasingly used chemical probe, the pan-class I PI3K and mammalian target of rapamycin (mTOR) inhibitor PI-103 (a pyridofuropyrimidine), and its very recent evolution into the thienopyrimidine drug GDC-0941, which exhibits excellent oral anticancer activity in preclinical models and is now undergoing phase I clinical trials in cancer patients. We also illustrate the impact of structural biology on the design of PI3K inhibitors and on the interpretation of their effects. The challenges and outlook for drugging the PI3 kinome are discussed in the more general context of the role of structural biology and chemical biology in innovative drug discovery.

Vps34 is the primordial member of the PI3 kinase family involved in vesicular trafficking, nutrient signaling, and autophagy. A report in Science unveils the Vps34 structure, providing new insights into the catalytic mechanism, explaining why Vsp34 is so difficult to inhibit, and facilitating design of chemical tools and potential drugs.

We report herein the first systematic exploration of inhibitors of the mitotic kinase Nek2. Starting from HTS hit aminopyrazine 2, compounds with improved activity were identified using structure-based design. Our structural biology investigations reveal two notable observations. First, 2 and related compounds bind to an unusual, inactive conformation of the kinase which to the best of our knowledge has not been reported for other types of kinase inhibitors. Second, a phenylalanine residue at the center of the ATP pocket strongly affects the ability of the inhibitor to bind to the protein. The implications of these observations are discussed, and the work described here defines key features for potent and selective Nek2 inhibition, which will aid the identification of more advanced inhibitors of Nek2.

A range of 3,6-di(hetero)arylimidazo[1,2-a]pyrazine ATP-competitive inhibitors of CHK1 were developed by scaffold hopping from a weakly active screening hit. Efficient synthetic routes for parallel synthesis were developed to prepare analogues with improved potency and ligand efficiency against CHK1. Kinase profiling showed that the imidazo[1,2-a]pyrazines could inhibit other kinases, including CHK2 and ABL, with equivalent or better potency depending on the pendant substitution. These 3,6-di(hetero)aryl imidazo[1,2-a]pyrazines appear to represent a general kinase inhibitor scaffold.

Structure-based approaches now impact across the whole continuum of drug discovery, from new target selection through the identification of hits to the optimization of lead compounds. Optimal application of structure-based design involves close integration with other discovery technologies, including fragment-based and virtual screening. Here, we illustrate the use of structural information and of structure-based drug design approaches in the discovery of small-molecule inhibitors for cancer drug targets and provide an outlook on the exploitation of structural information in future cancer drug discovery. Examples include high profile protein kinase targets and structurally related PI3 kinases, histone deacetylases, poly(ADP-ribose)polymerase and the molecular chaperone HSP90. Structure-based design approaches have also been successfully applied to the protein-protein interaction targets p53-MDM2 and the Bcl-2 family.

The Nek family of serine/threonine kinases regulates centrosome and cilia function; in addition, several of its members are potential targets for drug discovery. Nek2 is dimeric, is cell cycle regulated and functions in the separation of centrosomes at G2/M. Here, we report the crystal structures of wild-type human Nek2 kinase domain bound to ADP at 1.55-angstrom resolution and T175A mutant in apo form as well as that bound to a non-hydrolyzable ATP analog. These show that regions of the Nek2 structure around the nucleotide-binding site can adopt several different but well-defined conformations. None of the conformations was the same as that observed for the previously reported inhibitor-bound structure, and the two nucleotides stabilized two conformations. The structures Suggest mechanisms for the auto-inhibition of Nek2 that we have tested by mutagenesis. Comparison of the structures with Aurora-A and Cdk2 gives insight into the structural mechanism of Nek2 activation. The production of specific inhibitors that target individual kinases of the human genome is all urgent challenge in drug discovery, and Nek2 is especially promising as a cancer target. We not only identify potential challenges to the task of producing Nek2 inhibitors but also propose that the conformational variability provides an opportunity for the design of Nek2 selective inhibitors because one of the conformations may provide a unique target. (C) 2008 Elsevier Ltd. All rights reserved.

Checkpoint kinase 1 (CHK1) is an oncology target of significant current interest. Inhibition of CHK1 abrogates DNA damage-induced cell cycle checkpoints and sensitizes p53 deficient cancer cells to genotoxic therapies. Using template screening, a fragment-based approach to small molecule hit generation, we have identified multiple CHK1 inhibitor scaffolds suitable for further optimization. The sequential combination of in silico low molecular weight template selection, a high concentration biochemical assay and hit validation through protein-ligand X-ray crystallography provided 13 template hits from an initial in silico screening library of ca. 15000 compounds. The use of appropriate counter-screening to rule out nonspecific aggregation by test compounds was essential for optimum performance of the high concentration bioassay. One low molecular weight, weakly active purine template hit was progressed by iterative structure-based design to give submicromolar pyrazolopyridines with good ligand efficiency and appropriate CHK1-mediated cellular activity in HT29 colon cancer cells.

The screening of fragments is an alternative approach to high-throughput screening for the identification of leads for therapeutic targets. Fragment hits have been discovered using X-ray crystallographic screening of protein crystals of the serine protease enzyme thrombin. The fragment library was designed to avoid any well-precedented, strongly basic functionality. Screening hits included a novel ligand (3), which binds exclusively to the S2-S4 pocket, in addition to smaller fragments which bind to the S1 pocket. The structure of these protein-ligand complexes are presented. A chemistry strategy to link two such fragments together and to synthesize larger drug-sized compounds resulted in the efficient identification of hybrid inhibitors with nanomolar potency (e.g., 7, IC50 = 3.7 nM). These potent ligands occupy the same area of the active site as previously described peptidic inhibitors, while having very different chemical architecture.

Protein tyrosine phosphatases regulate signal transduction pathways involving tyrosine phosphorylation(1) and have been implicated in the development of cancer, diabetes, rheumatoid arthritis and hypertension(2). Increasing evidence suggests that the cellular redox state is involved in regulating tyrosine phosphatase activity through the reversible oxidization of the catalytic cysteine to sulphenic acid (Cys-SOH)(3-6). But how further oxidation to the irreversible sulphinic (Cys-SO2H) and sulphonic (Cys-SO3H) forms is prevented remains unclear. Here we report the crystal structures of the regulatory sulphenic and irreversible sulphinic and sulphonic acids of protein tyrosine phosphatase 1B (PTP1B), an important enzyme in the negative regulation of the insulin receptor(7,8) and a therapeutic target in type II diabetes and obesity(9). We also identify a sulphenyl-amide species that is formed through oxidation of its catalytic cysteine. Formation of the sulphenyl-amide causes large changes in the PTP1B active site, which are reversible by reduction with the cellular reducing agent glutathione. The sulphenyl-amide is a protective intermediate in the oxidative inhibition of PTP1B. In addition, it may facilitate reactivation of PTP1B by biological thiols and signal a unique state of the protein.

Crystallins are long-lived proteins packed inside eye lens fiber cells that are essential in maintaining the transparency and refractive power of the eye lens. Members of the two-domain betagamma-crystallin family assemble into an array of oligomer sizes, forming intricate higher-order networks in the lens cell. Here we describe the 1.4 Angstrom resolution crystal structure of a truncated version of human betaB1 that resembles an in vivo age-related truncation. The structure shows that unlike its close homolog, betaB2-crystallin, the homodimer is not domain swapped, but its domains are paired intramolecularly, as in more distantly related monomeric gamma-crystallins. However, the four-domain dimer resembles one half of the crystallographic bovine betaB2 tetramer and is similar to the engineered circular permuted rat betaB2. The crystal structure shows that the truncated betaB1 dimer is extremely well suited to form higher-order lattice interactions using its hydrophobic surface patches, linker regions, and sequence extensions.

Eta-crystallin is a retinal dehydrogenase that has acquired a role as a structural protein in the eye lens of elephant shrews, members of an ancient order of mammals. While it retains some activity toward retinal, which is oxidized to retinoic acid, the protein has acquired a number of specific sequence changes that have presumably been selected to enhance the lens role. The crystal structure of eta-crystallin, in common with class 1 and 2 ALDHs, is a dimer of dimers. It has a better-defined NAD binding site than those of related mammalian ALDH1 enzymes with the cofactor bound in the "hydride transfer" position in all four monomers with small differences about the dimer dyads. Although the active site is well conserved, the substrate-binding site is larger in eta-crystallin, and there are some mutations to the substrate access tunnel that might affect binding or release of substrate and product. It is possible that eta-crystallin has lost flexibility to improve its role in the lens. Enhanced binding of cofactor could enable it to act as a UV/blue light filter in the lens, improving visual acuity. The structure not only gives a view of a "natural mutant" of ALDH1 illustrating the adaptive conflict that can arise in multifunctional proteins, but also provides a well-ordered NAD binding site structure for this class of enzymes with important roles in development and health.

Background: The bacterial phosphoenolpyruvate-dependent phosphotransferase system (PTS) catalyses the cellular uptake and subsequent phosphorylation of carbohydrates. Moreover, the PTS plays a crucial role in the global regulation of various metabolic pathways, The PTS consists of two general proteins, enzyme I and the histidine-containing protein (HPr), and the carbohydrate-specific enzyme II [E-II]. E(II)s are usually composed of two cytoplasmic domains, IIA and IIB, and a transmembrane domain, IIC. The IIA domains catalyse the transfer of a phosphoryl group from HPr to IIB, which phosphorylates the transported carbohydrate. Knowledge of the structures of the IIA proteins may provide insight into the mechanisms by which the PTS couples phosphorylation reactions with carbohydrate specificity.Results: We have determined the crystal structure of the Escherichia coli mannitol-specific IIA domain, IIA(mtl) (M-r 16.3 kDa), by multiple anomalous dispersion analysis of a selenomethionine variant of IIA(mtl). The structure was refined at 1.8 Angstrom resolution to an R factor of 19.0% (R-free 24.2%). The enzyme consists of a single five-stranded mixed beta sheet, flanked by helices on both sides. The phosphorylation site (His65) is located at the end of the third beta strand, in a shallow crevice lined with hydrophobic residues. The sidechains of two conserved active-site residues, Arg49 and His111, adopt two different conformations in the four independent IIA(mtl) molecules. Using a solution structure of phosphorylated HPr, and a combination of molecular modelling and NMR binding experiments, structural models of the HPr-IIA(mtl) complex were generated.Conclusions: The fold of IIA(mtl) is completely different from the structures of other IIA proteins determined so far, The two conformations of Arg49 and His111 might represent different states of the active site, required for the different phosphoryl transfer reactions in which IIA(mtl) is involved. A comparison of the HPr-IIA(mtl) model with models of HPr in complex with other IIA enzymes shows that the overall interaction mode between the two proteins is similar. Differences in the stabilisation of the invariant residue Arg17 of HPr by the different IIA proteins might be part of a subtle mechanism to control the hierarchy of carbohydrate utilisation by the bacterium.

The three-dimensional structures of the IIA domain of the mannitol-specific phosphoenol-pyruvate dependent phosphotransferase system (PTS) of Escherichia coli and the regulatory IIA(ntr) enzyme have been compared. The enzymes are 20% identical in sequence and contain the sequence motif specific for IIA proteins belonging to the mannitol-fructose family of the PTS. The fold of the enzymes is nearly identical, which confirms their evolution from a common ancestor. Moreover, the phosphorylation site of IIA(mtl) (His65) and one of the observed conformations of the active site Arg49 are extremely similar to their equivalents (His73 and Arg57) in IIA(ntr). In contrast, His120, the equivalent of the second active site His111 of IIA(mtl), is not located in the active site of IIA(ntr) but points into the solvent on the other side of the molecule. The different position of His120 makes it unlikely that this residue assists in phosphoryl transfer in the regulatory IIA(ntr)s. As His120 is conserved in all IIA(ntr) enzymes, it could have an essential role in the recognition of the target protein of IIA(ntr).

Background: The bacterial phosphoenolpyruvate-dependent phosphotransferase system (PTS) mediates the energy-driven uptake of carbohydrates and their concomitant phosphorylation, In addition, the PTS is intimately involved in the regulation of a variety of metabolic and transcriptional processes in the bacterium. The multiprotein PTS consists of a membrane channel and at least four cytoplasmic proteins or protein domains that sequentially transfer a phosphoryl group from phosphoenolpyruvate to the transported carbohydrate, Determination of the three-dimensional structure of the IIB enzymes within the multiprotein complex would provide insights into the mechanisms by which they promote efficient transport by the membrane channel IIC protein and phosphorylate the transported carbohydrate on the inside of the cell,Results: The crystal structure of the IIB enzyme specific for cellobiose, IIBcellobiose (molecular weight 11.4 kDa), has been determined to a resolution of 1.8 Angstrom and refined to an R factor of 18.7% (R(free) of 24.1%). The enzyme consists of a single four-stranded parallel beta sheet flanked by helices on both sides. The phosphorylation site (Cys10) is located at the C-terminal end of the first beta strand, No positively charged residues, which could assist in phosphoryl-transfer, can be found in or near the active site. The fold of IIBcellobiose is remarkably similar to that of the mammalian low molecular weight protein tyrosine phosphatases.Conclusions: A comparison between IIBcellobiose and the structurally similar low molecular weight protein tyrosine phosphatases provides insight into the mechanism of the phosphoryltransfer reactions in which IIBcellobiose is involved, The differences in tertiary structure and active-site composition between IIBcellobiose and the glucose-specific IIBglucose give a structural explanation why the carbohydrate-specific components of different families cannot complement each other.

The cyclodextrin glycosyltransferase (CGTase, EC 2.4.1.19) gene from Bacillus circulans strain 251 was cloned and sequenced. It was found to code for a mature protein of 686 amino acid residues, showing 75% identity to the CGTase from B. circulans strain 8. The X-ray structure of the CGTase was elucidated in a maltodextrin-dependent crystal form and refined against X-ray diffraction data to 2.0 A resolution. The structure of the enzyme is nearly identical to the CGTase from B. circulans strain 8. Three maltose binding sites are observed at the protein surface, two in domain E and one in domain C. The maltose-dependence of CGTase crystallization can be ascribed to the proximity of two of the maltose binding sites to intermolecular crystal contacts. The maltose molecules bound in the E domain interact with several residues implicated in a raw starch binding motif conserved among a diverse group of starch converting enzymes.