The modulation of protease activity has been pursued to treat a large number of diseases including cancer, cardiovascular, autoimmune, and neurodegenerative diseases, as well as viral and parasitic infections. Notable protease inhibitors that have been developed as drugs include captopril for the treatment of high blood pressure (angiotensin-converting enzyme inhibitor, Bristol-Myers Squibb), saquinavir for the treatment of HIV/AIDS (HIV protease inhibitor, Roche), bortezomib for the treatment of bone cancer (proteosome inhibitor, Millenium), aliskiren for the treatment of hypertension (rennin inhibitor, Novartis), and sitagliptin for the treatment of diabetes (dipeptidyl peptidase-4 inhibitor, Merck). However, despite these successful demonstrations, the development of drugs targeting human and pathogenic proteases remains a significant challenge.
Typically, protease inhibitors have been developed on the basis of the preferred peptide substrate of the protease target. As exemplified by the identification of a caspase-1 substrate, a potent peptidic inhibitor is obtained in a straightforward manner. From a tetrapeptide substrate library, the preferred peptide sequence was rapidly identified. An aldehyde pharmacophore was then incorporated to turn the preferred substrate into a peptidic inhibitor. While this approach often rapidly provides potent peptidic inhibitors, nonpeptidic inhibitors are often necessary in drug development in order to achieve good oral bioavailability, relevant tissue penetration, and long circulating half-lives. And while these peptidic inhibitors can be converted to peptidomimetic inhibitors with improved pharmacokinetic and pharmacodynamic properties, in practice this has proven incredibly challenging. High throughput screening (HTS) of libraries of drug-like compounds is currently the standard drug discovery method for identifying small molecule ligands for receptors and enzymes.
For challenging protein targets such as proteases, fragment-based approaches have been developed wherein low molecular weight fragments are identified that bind to the desired target with modest affinity and are subsequently optimized to yield more potent compounds. Several fragment-based approaches have been developed, which includes functional assay, mass spectrometry, nuclear magnetic resonance (NMR), and X-ray crystallographic screening methods.
The functional assay method, in which a fragment library is assayed against the protein target and the binding affinity is measured, is high throughput, straightforward to conduct, and typically uses small amounts of protein. However, because fragments often have low binding affinities, high fragment concentrations are required in the assay, which can lead to false positives due to aggregation and non-specific protein inhibition. Additionally, this method gives no information on the site or mode of binding, thus complicating the optimization of identified fragments.
One of the most successful mass spectrometry-based methods is Tethering™, which requires a disulfide-containing fragment library to be prepared and screened with a protein that has a cysteine residue in the binding site. Fragment binding is achieved by disulfide exchange between the fragment and the binding site cysteine thiol, with detection by mass spectrometry. With NMR-based screening methods, such as SAR by NMR, fragments are screened for interaction with an 15N-labeled protein. Fragments that bind to the target are detected by monitoring changes in chemical shifts of amide moieties in the protein using NMR. X-ray crystallography screening methods, in which fragments are crystallized in the protein binding site, provide complete information about binding.