Establish an automated flow injection ESI-MS method for the screening of fragment based libraries: Application to Hsp90.

ESI-MS is a well established technique for the study of biopolymers (nucleic acids, proteins) and their non covalent adducts, due to its capacity to detect ligand-target complexes in the gas phase and allows inference of ligand-target binding in solution. In this article we used this approach to investigate the interaction of ligands to the Heat Shock Protein 90 (Hsp90). This enzyme is a molecular chaperone involved in the folding and maturation of several proteins which has been subjected in the last years to intensive drug discovery efforts due to its key role in cancer. In particular, reference compounds, with a broad range of dissociation constants from 40pM to 100µM, were tested to assess the reliability of ESI-MS for the study of protein-ligand complexes. A good agreement was found between the values measured with a fluorescence polarization displacement assay and those determined by mass spectrometry. After this validation step we describe the setup of a medium throughput screening method, based on ESI-MS, suitable to explore interactions of therapeutic relevance biopolymers with chemical libraries. Our approach is based on an automated flow injection ESI-MS method (AFI-MS) and has been applied to screen the Nerviano Medical Sciences proprietary fragment library of about 2000 fragments against Hsp90. In order to discard false positive hits and to discriminate those of them interacting with the N-terminal ATP binding site, competition experiments were performed using a reference inhibitor. Gratifyingly, this group of hits matches with the ligands previously identified by NMR FAXS techniques and confirmed by X-ray co-crystallization experiments. These results support the use of AFI-MS for the screening of medium size libraries, including libraries of small molecules with low affinity typically used in fragment based drug discovery. AFI-MS is a valid alternative to other techniques with the additional opportunities to identify compounds interacting with unpredicted or allosteric sites, without the need of any binding probes.

Fragment-based hit discovery and structure-based optimization of aminotriazoloquinazolines as novel Hsp90 inhibitors.

In the last decade the heat shock protein 90 (Hsp90) has emerged as a major therapeutic target and many efforts have been dedicated to the discovery of Hsp90 inhibitors as new potent anticancer agents. Here we report the identification of a novel class of Hsp90 inhibitors by means of a biophysical FAXS-NMR based screening of a library of fragments. The use of X-ray structure information combined with modeling studies enabled the fragment evolution of the initial triazoloquinazoline hit to a class of compounds with nanomolar potency and drug-like properties suited for further lead optimization.

Discovery of NMS-E973 as novel, selective and potent inhibitor of heat shock protein 90 (Hsp90).

Novel small molecule inhibitors of heat shock protein 90 (Hsp90) were discovered with the help of a fragment based drug discovery approach (FBDD) and subsequent optimization with a combination of structure guided design, parallel synthesis and application of medicinal chemistry principles. These efforts led to the identification of compound 18 (NMS-E973), which displayed significant efficacy in a human ovarian A2780 xenograft tumor model, with a mechanism of action confirmed in vivo by typical modulation of known Hsp90 client proteins, and with a favorable pharmacokinetic and safety profile.

Polyfluorinated amino acids for sensitive 19F NMR-based screening and kinetic measurements.

Two novel series of polyfluorinated amino acids (PFAs) were designed and synthesized according to a very short and scalable synthetic sequence. The advantages and limitations of these moieties for screening purposes are presented and discussed. The potential applications of these PFAs were tested with their incorporation into small arginine-containing peptides that represent suitable substrates for the enzyme trypsin. The enzymatic reactions were monitored by 19F NMR spectroscopy, using the 3-FABS (three fluorine atoms for biochemical screening) technique. The high sensitivity achieved with these PFAs permits a reduction in substrate concentration required for 3-FABS. This is relevant in the utilization of 3-FABS in fragment-based screening for identification of small scaffolds that bind weakly to the receptor of interest. The large dispersion of 19F isotropic chemical shifts allows the simultaneous measurement of the efficiency of the different substrates, thus identifying the best substrate for screening purposes. Furthermore, the knowledge of KM and Kcat for the different substrates allows the identification of the structural motifs responsible for the binding affinity to the receptor and those affecting the chemical steps in enzymatic catalysis. This enables the construction of suitable pharmacophores that can be used for designing nonpeptidic inhibitors with high affinity for the enzyme or molecules that mimic the transition state. The novel PFAs can also find useful application in the FAXS (fluorine chemical shift anisotropy and exchange for screening) experiment, a 19F-based competition binding assay for the detection of molecules that inhibit the interaction between two proteins.

NMR-based quality control approach for the identification of false positives and false negatives in high throughput screening.

The quality of the data generated in a high throughput screening (HTS) run is fundamental for selecting bona fide inhibitors and for ensuring the capture of the full richness of inhibitors present in a chemical library. For this purpose a quality control filter based on three one dimensional (1D) proton NMR experiments is proposed. The approach called SPAM (Solubility, Purity and Aggregation of the Molecule) Filter requires the acquisition of a 1D reference spectrum, a WaterLOGSY spectrum and/or a selective longitudinal relaxation filter spectrum for the identified hits dissolved in aqueous solution and in the presence of a water soluble reference molecule. This palette of experiments permits the rapid characterization of the identity, purity, solubility and aggregation state of the active compound. This knowledge is crucial for deriving accurate IC(50) and K(1) values of the inhibitors, for identifying false negatives and for detecting promiscuous inhibitors. Only compounds that pass through the SPAM Filter can be considered as starting points for medicinal chemistry efforts directed toward lead optimization. Examples of this approach in the identification of false positives in a screening run against the enzyme thymidine phosphorylase (TP) and the rescue of a false negative in a screening run against the Ser/Thr kinase AKT1 are presented.