Mass Spectrometry and Machine Learning Reveal Determinants of Client Recognition by Antiamyloid Chaperones

A Perspective on the Prospective Use of AI in Protein Structure Prediction

AlphaFold2 (AF2) and RoseTTaFold (RF) have revolutionized structural biology, serving as highly reliable and effective methods for predicting protein structures. This article explores their impact and limitations, focusing on their integration into experimental pipelines and their application in diverse protein classes, including membrane proteins, intrinsically disordered proteins (IDPs), and oligomers. In experimental pipelines, AF2 models help X-ray crystallography in resolving the phase problem, while complementarity with mass spectrometry and NMR data enhances structure determination and protein flexibility prediction. Predicting the structure of membrane proteins remains challenging for both AF2 and RF due to difficulties in capturing conformational ensembles and interactions with the membrane. Improvements in incorporating membrane-specific features and predicting the structural effect of mutations are crucial. For intrinsically disordered proteins, AF2's confidence score (pLDDT) serves as a competitive disorder predictor, but integrative approaches including molecular dynamics (MD) simulations or hydrophobic cluster analyses are advocated for accurate dynamics representation. AF2 and RF show promising results for oligomeric models, outperforming traditional docking methods, with AlphaFold-Multimer showing improved performance. However, some caveats remain in particular for membrane proteins. Real-life examples demonstrate AF2's predictive capabilities in unknown protein structures, but models should be evaluated for their agreement with experimental data. Furthermore, AF2 models can be used complementarily with MD simulations. In this Perspective, we propose a "wish list" for improving deep-learning-based protein folding prediction models, including using experimental data as constraints and modifying models with binding partners or post-translational modifications. Additionally, a meta-tool for ranking and suggesting composite models is suggested, driving future advancements in this rapidly evolving field.

Updates on Aβ Processing by Hsp90, BRICHOS, and Newly Reported Distinctive Chaperones

Alzheimer's disease (AD) is an extremely devastating neurodegenerative disease, and there is no cure for it. AD is specified as the misfolding and aggregation of amyloid-β protein (Aβ) and abnormalities in hyperphosphorylated tau protein. Current approaches to treat Alzheimer's disease have had some success in slowing down the disease's progression. However, attempts to find a cure have been largely unsuccessful, most likely due to the complexity associated with AD pathogenesis. Hence, a shift in focus to better understand the molecular mechanism of Aβ processing and to consider alternative options such as chaperone proteins seems promising. Chaperone proteins act as molecular caretakers to facilitate cellular homeostasis under standard conditions. Chaperone proteins like heat shock proteins (Hsps) serve a pivotal role in correctly folding amyloid peptides, inhibiting mitochondrial dysfunction, and peptide aggregation. For instance, Hsp90 plays a significant role in maintaining cellular homeostasis through its protein folding mechanisms. In this review, we analyze the most recent studies from 2020 to 2023 and provide updates on Aβ regulation by Hsp90, BRICHOS domain chaperone, and distinctive newly reported chaperones.

Advances in mass spectrometry to unravel the structure and function of protein condensates

Membrane-less organelles assemble through liquid-liquid phase separation (LLPS) of partially disordered proteins into highly specialized microenvironments. Currently, it is challenging to obtain a clear understanding of the relationship between the structure and function of phase-separated protein assemblies, owing to their size, dynamics and heterogeneity. In this Perspective, we discuss recent advances in mass spectrometry (MS) that offer several promising approaches for the study of protein LLPS. We survey MS tools that have provided valuable insights into other insoluble protein systems, such as amyloids, and describe how they can also be applied to study proteins that undergo LLPS. On the basis of these recent advances, we propose to integrate MS into the experimental workflow for LLPS studies. We identify specific challenges and future opportunities for the analysis of protein condensate structure and function by MS.

The C-terminal domain of the antiamyloid chaperone DNAJB6 binds to amyloid-β peptide fibrils and inhibits secondary nucleation

The DNAJB6 chaperone inhibits fibril formation of aggregation-prone client peptides through interaction with aggregated and oligomeric forms of the amyloid peptides. Here, we studied the role of its C-terminal domain (CTD) using constructs comprising either the entire CTD or the first two or all four of the CTD β-strands grafted onto a scaffold protein. Each construct was expressed as WT and as a variant with alanines replacing five highly conserved and functionally important serine and threonine residues in the first β-strand. We investigated the stability, oligomerization, antiamyloid activity, and affinity for amyloid-β (Aβ42) species using optical spectroscopy, native mass spectrometry, chemical crosslinking, and surface plasmon resonance technology. While DNAJB6 forms large and polydisperse oligomers, CTD was found to form only monomers, dimers, and tetramers of low affinity. Kinetic analyses showed a shift in inhibition mechanism. Whereas full-length DNAJB6 activity is dependent on the serine and threonine residues and efficiently inhibits primary and secondary nucleation, all CTD constructs inhibit secondary nucleation only, independently of the serine and threonine residues, although their dimerization and thermal stabilities are reduced by alanine substitution. While the full-length DNAJB6 inhibition of primary nucleation is related to its propensity to form coaggregates with Aβ, the CTD constructs instead bind to Aβ42 fibrils, which affects the nucleation events at the fibril surface. The retardation of secondary nucleation by DNAJB6 can thus be ascribed to the first two β-strands of its CTD, whereas the inhibition of primary nucleation is dependent on the entire protein or regions outside the CTD.

A Hairpin Motif in the Amyloid-β Peptide Is Important for Formation of Disease-Related Oligomers

The amyloid-β (Aβ) peptide is associated with the development of Alzheimer's disease and is known to form highly neurotoxic prefibrillar oligomeric aggregates, which are difficult to study due to their transient, low-abundance, and heterogeneous nature. To obtain high-resolution information about oligomer structure and dynamics as well as relative populations of assembly states, we here employ a combination of native ion mobility mass spectrometry and molecular dynamics simulations. We find that the formation of Aβ oligomers is dependent on the presence of a specific β-hairpin motif in the peptide sequence. Oligomers initially grow spherically but start to form extended linear aggregates at oligomeric states larger than those of the tetramer. The population of the extended oligomers could be notably increased by introducing an intramolecular disulfide bond, which prearranges the peptide in the hairpin conformation, thereby promoting oligomeric structures but preventing conversion into mature fibrils. Conversely, truncating one of the β-strand-forming segments of Aβ decreased the hairpin propensity of the peptide and thus decreased the oligomer population, removed the formation of extended oligomers entirely, and decreased the aggregation propensity of the peptide. We thus propose that the observed extended oligomer state is related to the formation of an antiparallel sheet state, which then nucleates into the amyloid state. These studies provide increased mechanistic understanding of the earliest steps in Aβ aggregation and suggest that inhibition of Aβ folding into the hairpin conformation could be a viable strategy for reducing the amount of toxic oligomers.

AlphaFold, allosteric, and orthosteric drug discovery: Ways forward

Drug discovery is arguably a highly challenging and significant interdisciplinary aim. The stunning success of the artificial intelligence-powered AlphaFold, whose latest version is buttressed by an innovative machine-learning approach that integrates physical and biological knowledge about protein structures, raised drug discovery hopes that unsurprisingly, have not come to bear. Even though accurate, the models are rigid, including the drug pockets. AlphaFold's mixed performance poses the question of how its power can be harnessed in drug discovery. Here we discuss possible ways of going forward wielding its strengths, while bearing in mind what AlphaFold can and cannot do. For kinases and receptors, an input enriched in active (ON) state models can better AlphaFold's chance of rational drug design success.

AlphaFold2 and its applications in the fields of biology and medicine

AlphaFold2 (AF2) is an artificial intelligence (AI) system developed by DeepMind that can predict three-dimensional (3D) structures of proteins from amino acid sequences with atomic-level accuracy. Protein structure prediction is one of the most challenging problems in computational biology and chemistry, and has puzzled scientists for 50 years. The advent of AF2 presents an unprecedented progress in protein structure prediction and has attracted much attention. Subsequent release of structures of more than 200 million proteins predicted by AF2 further aroused great enthusiasm in the science community, especially in the fields of biology and medicine. AF2 is thought to have a significant impact on structural biology and research areas that need protein structure information, such as drug discovery, protein design, prediction of protein function, et al. Though the time is not long since AF2 was developed, there are already quite a few application studies of AF2 in the fields of biology and medicine, with many of them having preliminarily proved the potential of AF2. To better understand AF2 and promote its applications, we will in this article summarize the principle and system architecture of AF2 as well as the recipe of its success, and particularly focus on reviewing its applications in the fields of biology and medicine. Limitations of current AF2 prediction will also be discussed.