Fragment based discovery of DsbA inhibitors

Abstract

DsbA’s are dithiol-disulfide oxidoreductases found in many Gram-negative bacteria. They catalyse the formation of disulfide bonds in many bacterial proteins. Bacteria lacking DsbA display phenotypes associated with the incorrect folding of DsbA substrates, often rendering the bacteria avirulent. DsbA is not present in eukaryotes and several crystal structures are available, thus it presents itself as an attractive antibacterial target and is the protein of interest for this project. To date there are no reported small molecule inhibitors. Previous NMR based screening conducted within the group has identified fragments that bind to Escherichia coli (Ec) DsbA, some of which gave crystal structures. This provided a platform for SBDD and FBDD of a phenylthiophene fragment hit. In the current work, design, synthesis and screening of multiple series of phenylthiophene fragment analogues was conducted. Initial optimisation of the core followed by expansion led to two additional series of compounds that also gave X-ray crystal structures. Their binding mode was found to change significantly from the parent compound highlighting the difficulty in developing EcDsbA inhibitors. Further exploration of the SAR was conducted however with increasingly poor physicochemical properties an accurate affinity measurement is required to assess if ongoing development would lead to suitable compounds. Towards this, a high concentration functional assay for inhibitor screening was developed through synthesis and optimisation of modified peptide substrates employing a fluorescence resonance energy transfer (FRET) lanthanide/coumarin readout. Investigation of the biochemistry of the assay shows weaker than expected inhibition on comparison to NMR data and is currently under further investigation. An alternate strategy for fragment development employing a reactive functional group to produce selective covalent inhibitors was investigated. Synthesis of a clickable fluoromethylketone (FMK) as a thiol selective reactive functional group allowed for the synthesis of a fragment-FMK conjugate which showed selectivity for EcDsbA as well as increased reaction rate due to the incorporation of a low affinity fragment binder. The covalent adduct formed was demonstrated to spontaneously resolve to give oxidised EcDsbA and the corresponding methyl ketone, which is the first report of reduction of a C-F bond by DsbA. Subsequently, the catalytic reduction of the well known thiol-reactive covalent modifier, maleimide was also demonstrated. Screening of additional reactive functional groups support that EcDsbA has a unique thiolate reactivity and this should be taken into account in design of future DsbA covalent inhibitors.