Structure-Based Drug Design and Synthesis of PI3Kα-Selective Inhibitor (PF-06843195).

The phosphoinositide 3-kinase (PI3K)/mammalian target of rapamycin (mTOR) signaling pathway is a frequently dysregulated pathway in human cancer, and PI3Kα is one of the most frequently mutated kinases in human cancer. A PI3Kα-selective inhibitor may provide the opportunity to spare patients the side effects associated with broader inhibition of the class I PI3K family. Here, we describe our efforts to discover a PI3Kα-selective inhibitor by applying structure-based drug design (SBDD) and computational analysis. A novel series of compounds, exemplified by 2,2-difluoroethyl (3S)-3-{[2'-amino-5-fluoro-2-(morpholin-4-yl)-4,5'-bipyrimidin-6-yl]amino}-3-(hydroxymethyl)pyrrolidine-1-carboxylate (1) (PF-06843195), with high PI3Kα potency and unique PI3K isoform and mTOR selectivity were discovered. We describe here the details of the design and synthesis program that lead to the discovery of 1.

The Axl kinase domain in complex with a macrocyclic inhibitor offers first structural insights into an active TAM receptor kinase.

The receptor tyrosine kinase family consisting of Tyro3, Axl, and Mer (TAM) is one of the most recently identified receptor tyrosine kinase families. TAM receptors are up-regulated postnatally and maintained at high levels in adults. They all play an important role in immunity, but Axl has also been implicated in cancer and therefore is a target in the discovery and development of novel therapeutics. However, of the three members of the TAM family, the Axl kinase domain is the only one that has so far eluded structure determination. To this end, using differential scanning fluorimetry and hydrogen-deuterium exchange mass spectrometry, we show here that a lower stability and greater dynamic nature of the Axl kinase domain may account for its poor crystallizability. We present the first structural characterization of the Axl kinase domain in complex with a small-molecule macrocyclic inhibitor. The Axl crystal structure revealed two distinct conformational states of the enzyme, providing a first glimpse of what an active TAM receptor kinase may look like and suggesting a potential role for the juxtamembrane region in enzyme activity. We noted that the ATP/inhibitor-binding sites of the TAM members closely resemble each other, posing a challenge for the design of a selective inhibitor. We propose that the differences in the conformational dynamics among the TAM family members could potentially be exploited to achieve inhibitor selectivity for targeted receptors.

Lessons from (S)-6-(1-(6-(1-methyl-1H-pyrazol-4-yl)-[1,2,4]triazolo[4,3-b]pyridazin-3-yl)ethyl)quinoline (PF-04254644), an inhibitor of receptor tyrosine kinase c-Met with high protein kinase selectivity but broad phosphodiesterase family inhibition leading to myocardial degeneration in rats.

The hepatocyte growth factor (HGF)/c-Met signaling axis is deregulated in many cancers and plays important roles in tumor invasive growth and metastasis. An exclusively selective c-Met inhibitor (S)-6-(1-(6-(1-methyl-1H-pyrazol-4-yl)-[1,2,4]triazolo[4,3-b]pyridazin-3-yl)ethyl)quinoline (8) was discovered from a highly selective high-throughput screening hit via structure-based drug design and medicinal chemistry lead optimization. Compound 8 had many attractive properties meriting preclinical evaluation. Broad off-target screens identified 8 as a pan-phosphodiesterase (PDE) family inhibitor, which was implicated in a sustained increase in heart rate, increased cardiac output, and decreased contractility indices, as well as myocardial degeneration in in vivo safety evaluations in rats. Compound 8 was terminated as a preclinical candidate because of a narrow therapeutic window in cardio-related safety. The learning from multiparameter lead optimization and strategies to avoid the toxicity attrition at the late stage of drug discovery are discussed.

Discovery of a novel class of exquisitely selective mesenchymal-epithelial transition factor (c-MET) protein kinase inhibitors and identification of the clinical candidate 2-(4-(1-(quinolin-6-ylmethyl)-1H-[1,2,3]triazolo[4,5-b]pyrazin-6-yl)-1H-pyrazol-1-yl)ethanol (PF-04217903) for the treatment of cancer.

The c-MET receptor tyrosine kinase is an attractive oncology target because of its critical role in human oncogenesis and tumor progression. An oxindole hydrazide hit 6 was identified during a c-MET HTS campaign and subsequently demonstrated to have an unusual degree of selectivity against a broad array of other kinases. The cocrystal structure of the related oxindole hydrazide c-MET inhibitor 10 with a nonphosphorylated c-MET kinase domain revealed a unique binding mode associated with the exquisite selectivity profile. The chemically labile oxindole hydrazide scaffold was replaced with a chemically and metabolically stable triazolopyrazine scaffold using structure based drug design. Medicinal chemistry lead optimization produced 2-(4-(1-(quinolin-6-ylmethyl)-1H-[1,2,3]triazolo[4,5-b]pyrazin-6-yl)-1H-pyrazol-1-yl)ethanol (2, PF-04217903), an extremely potent and exquisitely selective c-MET inhibitor. 2 demonstrated effective tumor growth inhibition in c-MET dependent tumor models with good oral PK properties and an acceptable safety profile in preclinical studies. 2 progressed to clinical evaluation in a Phase I oncology setting.

Homodimerization of the PAS-B domains of hypoxia-inducible factors.

The Per-Arnt-Sim (PAS) domains of hypoxia-inducible transcription factors (HIF) mediate heterodimer formation between the HIF-α forms that are induced in the event of cellular hypoxia and the constitutive HIF-ß variants. Previous efforts toward structural characterization of the HIF-1α PAS domains were limited by protein stability. Using homology modeling based on the published crystal structure of the PAS-B domain of the homologous protein HIF-2α in complex with the partner HIF-ß (also known as ARNT), we have identified a variant of HIF-1α with improved solubility, monodispersity, and stability. Purified solutions of the PAS-B domains of HIF-1α and HIF-2α differ in their propensity for homodimer formation. In an attempt to understand the structural basis for this difference, and to document the structural changes that accompany homodimer formation, we have undertaken a comparative NMR study of the PAS-B domains of HIF-1α and HIF-2α and mutants of HIF-1α that mimic the behavior of HIF-2α. The NMR spectra of all of these domains are very similar, consistent with the similarity of their amino acid sequences. However, the greater propensity of the HIF-1α PAS-B domain to form dimers as the concentration was increased allowed us to determine the site of homodimerization and pointed toward possible sequence changes in HIF-1α that might discourage the formation of homodimers.

Structure based drug design of crizotinib (PF-02341066), a potent and selective dual inhibitor of mesenchymal-epithelial transition factor (c-MET) kinase and anaplastic lymphoma kinase (ALK).

Because of the critical roles of aberrant signaling in cancer, both c-MET and ALK receptor tyrosine kinases are attractive oncology targets for therapeutic intervention. The cocrystal structure of 3 (PHA-665752), bound to c-MET kinase domain, revealed a novel ATP site environment, which served as the target to guide parallel, multiattribute drug design. A novel 2-amino-5-aryl-3-benzyloxypyridine series was created to more effectively make the key interactions achieved with 3. In the novel series, the 2-aminopyridine core allowed a 3-benzyloxy group to reach into the same pocket as the 2,6-dichlorophenyl group of 3 via a more direct vector and thus with a better ligand efficiency (LE). Further optimization of the lead series generated the clinical candidate crizotinib (PF-02341066), which demonstrated potent in vitro and in vivo c-MET kinase and ALK inhibition, effective tumor growth inhibition, and good pharmaceutical properties.

Structure of the catalytic domain of human protein kinase C beta II complexed with a bisindolylmaleimide inhibitor.

The conventional protein kinase C isoform, PKCII, is a signaling kinase activated during the hyperglycemic state and has been associated with the development of microvascular abnormalities associated with diabetes. PKCII, therefore, has been identified as a therapeutic target where inhibitors of its kinase activity are being pursued for treatment of microvascular-related diabetic complications. In this report, we describe the crystal structure of the catalytic domain of PKCbetaII complexed with an inhibitor at 2.6 A resolution. The kinase domain of PKCbetaII was cleaved and purified from full-length PKCbetaII expressed in baculovirus-infected insect cells. The overall kinase domain structure follows the classical bilobal fold and is in its fully activated conformation with three well-defined phosphorylated residues: Thr-500, Thr-641, and Ser-660. Different from the crystal structures of nonconventional PKC isoforms, the C-terminus of the PKCbetaII catalytic domain is almost fully ordered and features a novel alpha helix in the turn motif. An ATP-competitive inhibitor, 2-methyl-1H-indol-3-yl-BIM-1, was crystallized with the PKCbetaII catalytic domain as a dimer of two enzyme-inhibitor complexes. The bound inhibitor adopts a nonplanar conformation in the ATP-binding site, with the kinase domain taking on an intermediate, open conformation. This PKCbetaII-inhibitor complex represents the first structural description of any conventional PKC kinase domain. Given the pathogenic role of PKCbetaII in the development of diabetic complications, this structure can serve as a template for the rational design of inhibitors as potential therapeutic agents.

Ligands of the Mn2+ bound to porcine mitochondrial NADP-dependent isocitrate dehydrogenase, as assessed by mutagenesis.

Pig heart mitochondrial NADP-dependent isocitrate dehydrogenase requires a divalent metal ion for catalysis, and metal-isocitrate is its preferred substrate. On the basis of the crystal structure of the enzyme-Mn(2+)-isocitrate complex, Asp(252), Asp(275), and Asp(279) were selected as targets for site-directed mutagenesis to evaluate the roles of these residues as ligands of the metal ion. The circular dichroism spectra of the purified mutant enzymes are similar to that of wild-type enzyme indicating there are no appreciable conformational changes. The K(m) values for isocitrate and for Mn(2+) are increased in the asparagine and histidine mutants at positions 252 and 275; while for cysteine mutants at the same positions, the K(m)'s are not changed appreciably. Mutants at position 279 exhibit only a small change in K(m) for isocitrate. These results indicate that Asp(252) and Asp(275) are ligands of enzyme-bound Mn(2+)and influence the binding of Mn(2+)-isocitrate. Cysteine is an acceptable substitute for aspartate as a ligand of Mn(2+). The pK(aes)'s of D252C and D275C enzymes are similar to that of the wild-type enzyme (about 5.2), while the pK(aes) of D279C is a little lower (about 4.7). These findings suggest that the V(max)'s of the D252C, D275C, and D279C enzymes depend on the ionizable form of the same group as in wild-type enzyme and neither Asp(252), Asp(275), nor Asp(279) acts as the general base in the enzymatic reaction. For wild-type enzyme, the pK(aes) varies with the metal ion used with Mg(2+) > Cd(2+) > Mn(2+) > Co(2+), similar to the order of the pK's for these four metal-bound waters. We therefore attribute the pH dependence of V(max) to the deprotonation of the metal-coordinated hydroxyl group of isocitrate bound to isocitrate dehydrogenase.

Crystal structure of porcine mitochondrial NADP+-dependent isocitrate dehydrogenase complexed with Mn2+ and isocitrate. Insights into the enzyme mechanism.

The crystal structure of porcine heart mitochondrial NADP+-dependent isocitrate dehydrogenase (IDH) complexed with Mn2+ and isocitrate was solved to a resolution of 1.85 A. The enzyme was expressed in Escherichia coli, purified as a fusion protein with maltose binding protein, and cleaved with thrombin to yield homogeneous enzyme. The structure was determined by multiwavelength anomalous diffraction phasing using selenium substitution in the form of selenomethionine as the anomalous scatterer. The porcine NADP+-IDH enzyme is structurally compared with the previously solved structures of IDH from E. coli and Bacillus subtilis that share 16 and 17% identity, respectively, with the mammalian enzyme. The porcine enzyme has a protein fold similar to the bacterial IDH structures with each monomer folding into two domains. However, considerable differences exist between the bacterial and mammalian forms of IDH in regions connecting core secondary structure. Based on the alignment of sequence and structure among the porcine, E. coli, and B. subtilis IDH, a putative phosphorylation site has been identified for the mammalian enzyme. The active site, including the bound Mn2+-isocitrate complex, is highly ordered and, therefore, mechanistically informative. The consensus IDH mechanism predicts that the Mn2+-bound hydroxyl of isocitrate is deprotonated prior to its NADP+-dependent oxidation. The present crystal structure has an active site water that is well positioned to accept the proton and ultimately transfer the proton to solvent through an additional bound water.