Current views of multidrug (MD) recognition focus on large drug-binding cavities with flexible elements. However, MD recognition in BmrR is supported by a small, rigid drug-binding pocket. Here, a detailed description of MD binding by the noncanonical BmrR protein is offered through the combined use of X-ray and solution studies. Low shape complementarity, suboptimal packing, and efficient burial of a diverse set of ligands is facilitated by an aromatic docking platform formed by a set of conformationally fixed aromatic residues, hydrophobic pincer pair that locks the different drug structures on the adaptable platform surface, and a trio of acidic residues that enables cation selectivity without much regard to ligand structure. Within the binding pocket is a set of BmrR-derived H-bonding donor and acceptors that solvate a wide range of ligand polar substituent arrangements in a manner analogous to aqueous solvent. Energetic analyses of MD binding by BmrR are consistent with structural data. A common binding orientation for the different BmrR ligands is in line with promiscuous allosteric regulation.

The large yybP-ykoY family of bacterial riboswitches is broadly distributed phylogenetically. Previously, these gene-regulatory RNAs were proposed to respond to Mn. X-ray crystallography revealed a binuclear cation-binding pocket. This comprises one hexacoordinate site, with six oxygen ligands, which preorganizes the second, with five oxygen and one nitrogen ligands. The relatively soft nitrogen ligand was proposed to confer affinity for Mn, but how this excludes other soft cations remained enigmatic. By subjecting representative yybP-ykoY riboswitches to diverse cations in vitro, we now find that these RNAs exhibit limited transition metal ion selectivity. Among the cations tested, Cd and Mn bind most tightly, and comparison of three new Cd-bound crystal structures suggests that these riboswitches achieve selectivity by enforcing heptacoordination (favored by high-spin Cd and Mn, but otherwise uncommon) in the softer site. Remarkably, the Cd- and Mn-selective bacterial transcription factor MntR also uses heptacoordination within a binuclear site to achieve selectivity.

No abstract

Generative artificial intelligence (AI) has the potential to greatly increase the speed, quality and controllability of antibody design. Traditional de novo antibody discovery requires time and resource intensive screening of large immune or synthetic libraries. These methods also offer little control over the output sequences, which can result in lead candidates with sub-optimal binding and poor developability attributes. Several groups have introduced models for generative antibody design with promising in silico evidence, however, no such method has demonstrated de novo antibody design with experimental validation. Here we use generative deep learning models to de novo design antibodies against three distinct targets, in a zero-shot fashion, where all designs are the result of a single round of model generations with no follow-up optimization. In particular, we screen over 400,000 antibody variants designed for binding to human epidermal growth factor receptor 2 (HER2) using our high-throughput wet lab capabilities. From these screens, we further characterize 421 binders using surface plasmon resonance (SPR), finding three that bind tighter than the therapeutic antibody trastuzumab. The binders are highly diverse, have low sequence identity to known antibodies, and adopt variable structural conformations. Additionally, these binders score highly on our previously introduced Naturalness metric, indicating they are likely to possess desirable developability profiles and low immunogenicity. We open source the HER2 binders and report the measured binding affinities. These results unlock a path to accelerated drug creation for novel therapeutic targets using generative AI combined with high-throughput experimentation.

Multidrug resistance (MDR) refers to the acquired ability of cells to tolerate a broad range of toxic compounds. One mechanism cells employ is to increase the level of expression of efflux pumps for the expulsion of xenobiotics. A key feature uniting efflux-related mechanisms is multidrug (MD) recognition, either by efflux pumps themselves or by their transcriptional regulators. However, models describing MD binding by MDR effectors are incomplete, underscoring the importance of studies focused on the recognition elements and key motifs that dictate polyspecific binding. One such motif is the GyrI-like domain, which is found in several MDR proteins and is postulated to have been adapted for small-molecule binding and signaling. Here we report the solution binding properties and crystal structures of two proteins containing GyrI-like domains, SAV2435 and CTR107, bound to various ligands. Furthermore, we provide a comparison with deposited crystal structures of GyrI-like proteins, revealing key features of GyrI-like domains that not only support polyspecific binding but also are conserved among GyrI-like domains. Together, our studies suggest that GyrI-like domains perform evolutionarily conserved functions connected to multidrug binding and highlight the utility of these types of studies for elucidating mechanisms of MDR.