Engineering Archeal Surrogate Systems for the Development of Protein–Protein Interaction Inhibitors against Human RAD51

Abstract

Protein–protein interactions (PPIs) are increasingly important targets for drug discovery. Efficient fragment-based drug discovery approaches to tackle PPIs are often stymied by difficulties in the production of stable, unliganded target proteins. Here, we report an approach that exploits protein engineering to “humanise” thermophilic archeal surrogate proteins as targets for small-molecule inhibitor discovery and to exemplify this approach in the development of inhibitors against the PPI between the recombinase RAD51 and tumour suppressor BRCA2. As human RAD51 has proved impossible to produce in a form that is compatible with the requirements of fragment-based drug discovery, we have developed a surrogate protein system using RadA from Pyrococcus furiosus. Using a monomerised RadA as our starting point, we have adopted two parallel and mutually instructive approaches to mimic the human enzyme: firstly by mutating RadA to increase sequence identity with RAD51 in the BRC repeat binding sites, and secondly by generating a chimeric archaeal human protein. Both approaches generate proteins that interact with a fourth BRC repeat with affinity and stoichiometry comparable to human RAD51. Stepwise humanisation has also allowed us to elucidate the determinants of RAD51 binding to BRC repeats and the contributions of key interacting residues to this interaction. These surrogate proteins have enabled the development of biochemical and biophysical assays in our ongoing fragment-based small-molecule inhibitor programme and they have allowed us to determine hundreds of liganded structures in support of our structure-guided design process, demonstrating the feasibility and advantages of using archeal surrogates to overcome difficulties in handling human proteins.

Improved RAD51 binders through motif shuffling based on the modularity of BRC repeats

“…5b, g). These values fall within the range of previously measured affinities (6.2-64 nM) for this protein-protein interaction 24 . Also, BRC4 was found to be the tightest binder of all the natural repeats, in agreement with previous studies 29,30 .…”

“…The GB1 domain does not interfere with the BRC-repeat interaction: an isothermal titration calorimetry (ITC) affinity measurement was carried out to confirm that the BRC4 peptide, upon fusion to the GB1 C-terminus, maintained its ability to bind HumRadA22 (a faithful yet monomeric model of RAD51 and for simplicity we will refer to this as monomeric RAD51 24 ; Supplementary Figs. 5, 6 and accompanying Supplementary Text 3) and matched affinities previously measured for BRC4 peptide [24][25][26] .…”

“…1, 2 and Supplementary Table 1). The E. coli expression construct for monomeric RAD51 (pBAT4-HumRadA22), has been described previously 24 . Cloning of plasmids GFP and GFP-BRC8-2 for mammalian cell transfection is described in SI ( Supplementary Fig.…”

“…4). Monomeric RAD51 expression and purification for fluorescence anisotropy measurements was carried out as described previously 24 , where monomeric RAD51 was called 'HumRadA22'. Fluorescence polarization competition assay and microfluidic based measurements.…”

Small-Molecule Inhibitors Targeting DNA Repair and DNA Repair Deficiency in Research and Cancer Therapy

“…The identified small molecule fragments can potentially be grown and/or linked to yield a more potent, specific inhibitor with drug-like properties (Scott et al, 2015). These endeavors will depend on the structural information on the RAD51-BRC peptide interaction (Pellegrini et al, 2002; Subramanyam et al, 2013) and the development of an experimental system compatible with the fragment analysis (Moschetti et al, 2016). …”