Single amino acid substitutions can profoundly affect protein folding, dynamics, and function. The ability to discern between benign and pathogenic substitutions is pivotal for therapeutic interventions and research directions. Given the limitations in experimental examination of these variants, AlphaMissense has emerged as a promising predictor of the pathogenicity of missense variants. Since heterogenous performance on different types of proteins can be expected, we assessed the efficacy of AlphaMissense across several protein groups (e.g. soluble, transmembrane, and mitochondrial proteins) and regions (e.g. intramembrane, membrane interacting, and high confidence AlphaFold segments) using ClinVar data for validation. Our comprehensive evaluation showed that AlphaMissense delivers outstanding performance, with MCC scores predominantly between 0.6 and 0.74. We observed low performance on disordered datasets and ClinVar data related to the CFTR ABC protein. However, a superior performance was shown when benchmarked against the high quality CFTR2 database. Our results with CFTR emphasizes AlphaMissense’s potential in pinpointing functional hot spots, with its performance likely surpassing benchmarks calculated from ClinVar and ProteinGym datasets.

Transmembrane (TM) proteins are major drug targets, but their structure determination, a prerequisite for rational drug design, remains challenging. Recently, the DeepMind’s AlphaFold2 machine learning method greatly expanded the structural coverage of sequences with high accuracy. Since the employed algorithm did not take specific properties of TM proteins into account, the reliability of the generated TM structures should be assessed. Therefore, we quantitatively investigated the quality of structures at genome scales, at the level of ABC protein superfamily folds and for specific membrane proteins (e.g. dimer modeling and stability in molecular dynamics simulations). We tested template-free structure prediction with a challenging TM CASP14 target and several TM protein structures published after AlphaFold2 training. Our results suggest that AlphaFold2 performs well in the case of TM proteins and its neural network is not overfitted. We conclude that cautious applications of AlphaFold2 structural models will advance TM protein-associated studies at an unexpected level.

The ABCG2 protein has a key role in the transport of a wide range of structurally dissimilar endo‐ and xenobiotics in the human body, especially in the tissue barriers and the metabolizing or secreting organs. The human ABCG2 gene harbors a high number of polymorphisms and mutations, which may significantly modulate its expression and function. Recent high‐resolution structural data, complemented with molecular dynamic simulations, may significantly help to understand intramolecular movements and substrate handling, as well as the effects of mutations on the membrane transporter function of ABCG2. As reviewed here, structural alterations may result not only in direct alterations in drug binding and transporter activity, but also in improper folding or problems in the carefully regulated process of trafficking, including vesicular transport, endocytosis, recycling, and degradation. Here, we also review the clinical importance of altered ABCG2 expression and function in general drug metabolism, cancer multidrug resistance, and impaired uric acid excretion, leading to gout.

ABCG2/BCRP is a membrane protein, involved in xenobiotic and endobiotic transport in key pharmacological barriers and drug metabolizing organs, in the protection of stem cells, and in multidrug resistance of cancer. Pharmacogenetic studies implicated the role of ABCG2 in response to widely used medicines and anticancer agents, as well as in gout. Its Q141K variant exhibits decreased functional expression thus increased drug accumulation and decreased urate secretion. Still, there has been no reliable molecular model available for this protein, as the published structures of other ABC transporters could not be properly fitted to the ABCG2 topology and experimental data. The recently published high resolution structure of a close homologue, the ABCG5-ABCG8 heterodimer, revealed a new ABC transporter fold, unique for ABCG proteins. Here we present a structural model of the ABCG2 homodimer based on this fold and detail the experimental results supporting this model. In order to describe the effect of mutations on structure and dynamics, and characterize substrate recognition and cholesterol regulation we performed molecular dynamics simulations using full length ABCG2 protein embedded in a membrane bilayer and in silico docking simulations. Our results show that in the Q141K variant the introduced positive charge diminishes the interaction between the nucleotide binding and transmembrane domains and the R482G variation alters the orientation of transmembrane helices. Moreover, the R482 position, which plays a role the substrate specificity of the transporter, is located in one of the substrate binding pockets identified by the in silico docking calculations. In summary, the ABCG2 model and in silico simulations presented here may have significant impact on understanding drug distribution and toxicity, as well as drug development against cancer chemotherapy resistance or gout.

While scientists can often infer the biological function of proteins from their 3-dimensional quaternary structures, the gap between the number of known protein sequences and their experimentally determined structures keeps increasing. A potential solution to this problem is presented by ever more sophisticated computational protein modeling approaches. While often powerful on their own, most methods have strengths and weaknesses. Therefore, it benefits researchers to examine models from various model providers and perform comparative analysis to identify what models can best address their specific use cases. To make data from a large array of model providers more easily accessible to the broader scientific community, we established 3D-Beacons, a collaborative initiative to create a federated network with unified data access mechanisms. The 3D-Beacons Network allows researchers to collate coordinate files and metadata for experimentally determined and theoretical protein models from state-of-the-art and specialist model providers and also from the Protein Data Bank.

Intrinsically disordered proteins (IDPs) lack a stable tertiary structure, but their short binding regions termed Pre-Structured Motifs (PreSMo) can form transient secondary structure elements in solution. Although disordered proteins are crucial in many biological processes and designing strategies to modulate their function is highly important, both experimental and computational tools to describe their conformational ensembles and the initial steps of folding are sparse. Here we report that discrete molecular dynamics (DMD) simulations combined with replica exchange (RX) method efficiently samples the conformational space and detects regions populating α-helical conformational states in disordered protein regions. While the available computational methods predict secondary structural propensities in IDPs based on the observation of protein-protein interactions, our ab initio method rests on physical principles of protein folding and dynamics. We show that RX-DMD predicts α-PreSMos with high confidence confirmed by comparison to experimental NMR data. Moreover, the method also can dissect α-PreSMos in close vicinity to each other and indicate helix stability. Importantly, simulations with disordered regions forming helices in X-ray structures of complexes indicate that a preformed helix is frequently the binding element itself, while in other cases it may have a role in initiating the binding process. Our results indicate that RX-DMD provides a breakthrough in the structural and dynamical characterization of disordered proteins by generating the structural ensembles of IDPs even when experimental data are not available.