We have developed the Allosteric Library based on published allosteric enzyme inhibitors and modulators compounds. The Allosteric Library contains more than 1400 compounds. It has been built around more than 40 published allosteric enzyme inhibitors and modulators and we have built small focused libraries around these structures including several novel patentable structures.
Recently, X-ray diffraction experiments with co-crystallized inhibitors like imatinib (inhibitor of Bcr-Abl) and BIRB-796 (inhibitor of p38 MAP kinase) proved allosteric-type binding mode or DFG-out binding for these drugs. Several published kinase inhibitors with allosteric binding mode are now known: imatinib for Bcr-Abl, BIRB-796 for p38, BMS-345541 for IKK, lapatinib for EGFR/Her2, nilotinib for imatinib-resistant Bcr-Abl and sorafenib for Raf. Allosteric inhibitors which block the enzyme in inactive conformation or inhibit kinase activation via protein-protein interaction inhibition have also been included in the library. For example, allosteric inhibitors which bind to MEK1 and prevent its activation or allosteric inhibtors which bind to Akt kinase and prevent it from being activated by PDK1.
DFG-out binders have an elevated residence time on the target kinase, and are therefore ‘better’ drugs, since they can cope with physiochemical deficiencies. In addition to the DFG-out binders, there are other allosteric inhibitors known, such as other Erk1/2 inhibitors, other Bcr-Abl inhibitors (like GNF-1), some Akt inhibitors, PLK1 inhibitors, JAK2 inhibitors, JNK inhibitors, some allosteric multiple kinase inhibitors, Vichem’s PDGFR inhibitory HDL series and some tyrphostins developed earlier for Sugen. In total, the Allosteric Library has been built around more than 60 published reference compounds, and the library is continuously increasing and improving.
Applying Vichem’s special ‘theoretical interactive surface’ pharmacophore modeling on pre-filtered ‘non-ATP-like’ kinase inhibitors, we were able to identify the main structural features needed for allosteric binding. Appropriate molecular shape, size and electronic properties in suitable positions can afford selective and high-affinity binding at allosteric binding sites of kinases. We have developed a series of allosteric inhibitor analogs around such proven allosteric-type core structures, which are available for testing. These structures have been tested and have been found to be active on several kinases in low micromolar or nanomolar concentrations.
Type II inhibitors occupy the nucleotide pocket, extend into an adjacent ‘allosteric pocket’ and induce DFG-out configuration. Type III inhibitors exclusively bind to this ‘allosteric pocket’ and block kinase activity without displacing ATP. Type IV inhibitors bind to sites several Angstroms away from the nucleotide binding pocket. Kinase activation inhibitors bind to specific motifs on the kinase and inhibit activation or protein-protein interactions.
For the determination of binding type we use Reporter Displacement Binding Assay (with Proteros) and environmental sensitive fluorophore assay for type II and III inhibitors.
Developing Novel Allosteric Inhibitors
In order to develop novel non-ATP binding site- or allosteric binding site-directed inhibitors, 3D structure-based drug design provides a key strategy to generate promising lead compounds. Protein kinase drug discovery will exploit a variety of high-powered in silico tools to visualize, analyze and probe the molecular properties and interactions of novel small-molecule inhibitors with possible allosteric binding sites. Integration of in silico tools (e.g., molecular modeling and virtual screening) with structural biology (e.g., X-ray and/or NMR proven structures of a lead compound complexed with its therapeutic target), chemical synthesis, biological screening (e.g., biochemical kinase assays with kinetic analysis, in vitro and in vivo assays), cheminformatics, and bioinformatics provides the background for novel non-ATP binding site-directed or allosteric kinase inhibitor drug discovery. The use of the Allosteric Library for hit finding on various kinase targets could be a very important (first) step for the required structural biological studies.