The screening for protein-ligand interactions represents an initial key step in the modern drug discovery process. Various techniques exist for this purpose, each of which with a varying level of selectivity and throughput. Nuclear magnetic resonance (NMR) already is a popular screening technique because of the structural specificity imparted by chemical shift, the primary NMR observable parameter. NMR is also unique in the ability to specifically distinguish signals from the bound form of the ligand. NMR, however, suffers from limited sensitivity. The throughput in screening applications is reduced compared to other techniques. Furthermore, stringent conditions on sample quality and concentration are imposed. For example, NMR based screening often involves the use of a large excess of ligand over target protein, in which case observation is performed on the free ligand. As a result, the most desirable ligands with low off-rates may be missed, reducing the specificity of the method. Sensitivity in a liquid state NMR experiment can however be greatly enhanced using newly emerging hyperpolarization techniques, foremost dissolution dynamic nuclear polarization (DNP). While this technique has recently been demonstrated to remove many of the above mentioned limitations, NMR instrumentation that enables utilization of its benefits for high-throughput screening does not exist. Here, we will develop a multiplexed NMR probe and dedicated spectrometer that allows for parallel screening using hyperpolarized ligands. The probe and spectrometer will be designed for ligands hyperpolarized on fluorine atoms, with the added capability of detection of proton signals. Hyperpolarized fluorine is a particularly attractie target because this nucleus allows for the acquisition of background free spectra and at the same time is subject to comparatively large changes in chemical shift and relaxation behavior when a ligand binds to a target. These properties allow for the rapid detection of binding via single scan NMR measurement. The capability for detection of proton as a secondary nucleus will allow for an on-line assessment of sample quality. Fluorine is an abundant nucleus in pharmaceuticals; approximately 20% of marketed drug molecules contain a fluorine atom. Furthermore, other drug molecules can be identified by competitive binding to a target site together with a fluorinated reporter ligand. Parallelized detection can therefore be quite generally used for simultaneous observation of binding of multiple different ligands to a target, for determination of dissociation constants by titration of ligand concentration, and for target/anti-target screens. In a final step, the NMR methods for these applications will be developed and demonstrated specifically on the new hardware, using pharmaceutically relevant targets. These developments will allow for the sensitive and rapid detection of ligand binding by NMR, improving access to the benefits of this versatile spectroscopic technique for screening applications.
The identification of small molecule ligands that bind to a protein is the first step in the process of finding new drugs. Here, we are developing a new type of instrument that will increase the speed and fidelity of this part of the drug discovery process.
