Protein folding and misfolding: mechanism and principles

Probing the energy barriers and stages of membrane protein unfolding using solid-state NMR spectroscopy

Understanding how the amino acid sequence dictates protein structure and defines its stability is a fundamental problem in molecular biology. It is especially challenging for membrane proteins that reside in the complex environment of a lipid bilayer. Here, we obtain an atomic-level picture of the thermally induced unfolding of a membrane-embedded α-helical protein, human aquaporin 1, using solid-state nuclear magnetic resonance spectroscopy. Our data reveal the hierarchical two-step pathway that begins with unfolding of a structured extracellular loop and proceeds to an intermediate state with a native-like helical packing. In the second step, the transmembrane domain unravels as a single unit, resulting in a heterogeneous misfolded state with high helical content but with nonnative helical packing. Our results show the importance of loops for the kinetic stabilization of the whole membrane protein structure and support the three-stage membrane protein folding model.

Interpretation of Hydrogen/Deuterium Exchange Mass Spectrometry

This paper sheds light on the meaning of hydrogen/deuterium exchange-mass spectrometry (HDX-MS) data. HDX-MS data provide not structural information but dynamic information on an analyte protein. First, the reaction mechanism of backbone amide HDX reaction is considered and the correlation between the parameters from an X-ray crystal structure and the protection factors of HDX reactions of cytochrome c is evaluated. The presence of H-bonds in a protein structure has a strong influence on HDX rates which represent protein dynamics, while the solvent accessibility only weakly affects the HDX rates. Second, the energy diagrams of the HDX reaction at each residue in the presence and absence of perturbation are described. Whereas the free energy change upon mutation can be directly measured by the HDX rates, the free energy change upon ligand binding may be complicated due to the presence of unbound analyte protein in the protein-ligand mixture. Third, the meanings of HDX and other biophysical techniques are explained using a hypothetical protein folding well. The shape of the protein folding well describes the protein dynamics and provides Boltzmann distribution of open and closed states which yield HDX protection factors, while a protein's crystal structure represents a snapshot near the bottom of the well. All biophysical data should be consistent yet provide different information because they monitor different parts of the same protein folding well.

Mapping Protein-Protein Interactions at Birth: Single-Particle Cryo-EM Analysis of a Ribosome-Nascent Globin Complex

Interactions between ribosome-bound nascent chains (RNCs) and ribosomal components are critical to elucidate the mechanism of cotranslational protein folding. Nascent protein-ribosome contacts within the ribosomal exit tunnel were previously assessed mostly in the presence of C-terminal stalling sequences, yet little is known about contacts taking place in the absence of these strongly interacting motifs. Further, there is nearly no information about ribosomal proteins (r-proteins) interacting with nascent chains within the outer surface of the ribosome. Here, we combine chemical cross-linking, single-particle cryo-EM, and fluorescence anisotropy decays to determine the structural features of ribosome-bound apomyoglobin (apoMb). Within the ribosomal exit tunnel core, interactions are similar to those identified in previous reports. However, once the RNC enters the tunnel vestibule, it becomes more dynamic and interacts with ribosomal RNA (rRNA) and the L23 r-protein. Remarkably, on the outer surface of the ribosome, RNCs interact mainly with a highly conserved nonpolar patch of the L23 r-protein. RNCs also comprise a compact and dynamic N-terminal region lacking contact with the ribosome. In all, apoMb traverses the ribosome and interacts with it via its C-terminal region, while N-terminal residues sample conformational space and form a compact subdomain before the entire nascent protein sequence departs from the ribosome.

Protein structure generation via folding diffusion

The ability to computationally generate novel yet physically foldable protein structures could lead to new biological discoveries and new treatments targeting yet incurable diseases. Despite recent advances in protein structure prediction, directly generating diverse, novel protein structures from neural networks remains difficult. In this work, we present a diffusion-based generative model that generates protein backbone structures via a procedure inspired by the natural folding process. We describe a protein backbone structure as a sequence of angles capturing the relative orientation of the constituent backbone atoms, and generate structures by denoising from a random, unfolded state towards a stable folded structure. Not only does this mirror how proteins natively twist into energetically favorable conformations, the inherent shift and rotational invariance of this representation crucially alleviates the need for more complex equivariant networks. We train a denoising diffusion probabilistic model with a simple transformer backbone and demonstrate that our resulting model unconditionally generates highly realistic protein structures with complexity and structural patterns akin to those of naturally-occurring proteins. As a useful resource, we release an open-source codebase and trained models for protein structure diffusion.

Directed proton transfer from Fo to F1 extends the multifaceted proton functions in ATP synthase

The role of protons in ATP synthase is typically considered to be energy storage in the form of an electrochemical potential, as well as an operating element proving rotation. However, this review emphasizes that protons also act as activators of conformational changes in F1 and as direct participants in phosphorylation reaction. The protons transferred through Fo do not immediately leave to the bulk aqueous phase, but instead provide for the formation of a pH gradient between acidifying Fo and alkalizing F1. It facilitates a directed inter-subunit proton transfer to F1, where they are used in the ATP synthesis reaction. This ensures that the enzyme activity is not limited by a lack of protons in the alkaline mitochondrial matrix or chloroplast stroma. Up to one hundred protons bind to the carboxyl groups of the F1 subunit, altering the electrical interactions between the amino acids of the enzyme. This removes the inhibition of ATP synthase caused by the electrostatic attraction of charged amino acids of the stator and rotor and also makes the enzyme more prone to conformational changes. Protonation occurs during ATP synthesis initiation and during phosphorylation, while deprotonation blocks the rotation inhibiting both synthesis and hydrolysis. Thus, protons participate in the functioning of all main components of ATP synthase molecular machine making it effectively a proton-driven electric machine. The review highlights the key role of protons as a coupling factor in ATP synthase with multifaceted functions, including charge and energy transport, torque generation, facilitation of conformational changes, and participation in the ATP synthesis reaction.

Structural Analysis of Oligosaccharides and Glycoconjugates Using NMR

Carbohydrate chains play critical roles in cellular recognition and subsequent signal transduction in the nervous system. Furthermore, gangliosides are targets for various amyloidogenic proteins associated with neurodegenerative disorders. To better understand the molecular mechanisms underlying these biological phenomena, atomic views are essential to delineate dynamic biomolecular interactions. Nuclear magnetic resonance (NMR) spectroscopy provides powerful tools for studying structures, dynamics, and interactions of biomolecules at the atomic level. This chapter describes the basics of solution NMR techniques and their applications to the analysis of 3D structures and interactions of glycoconjugates in the nervous system.

Protein misfolding and related human diseases: A comprehensive review of toxicity, proteins involved, and current therapeutic strategies

Most of the cell's chemical reactions and structural components are facilitated by proteins. But proteins are highly dynamic molecules, where numerous modifications or changes in the cellular environment can affect their native conformational fold leading to protein aggregation. Various stress conditions, such as oxidative stress, mutations and metal toxicity may cause protein misfolding and aggregation by shifting the conformational equilibrium towards more aggregation-prone states. Most of the protein misfolding diseases (PMDs) involve aggregation of protein. We have discussed such proteins like Aβ peptide, α-synuclein, amylin and lysozyme involved in Alzheimer's, Parkinson's, type II diabetes and non-neuropathic systemic amyloidosis respectively. Till date, all advances in PMDs therapeutics help symptomatically but do not prevent the root cause of the disease, i.e., the aggregation of protein involved in the diseases. Current efforts focused on developing therapies for PMDs have employed diverse strategies; repositioning pre-existing drugs as it saves time and money; natural compounds that are touted as potential drug candidates have an advantage of being taken in diet normally and will induce lesser side effects. This review also covers recently developed therapeutic strategies like antisense drugs and disaggregases which has yielded therapeutic agents that have transitioned from preclinical studies into human clinical trials.

Evolutionary adaptation of the protein folding pathway for secretability

Secretory preproteins of the Sec pathway are targeted post-translationally and cross cellular membranes through translocases. During cytoplasmic transit, mature domains remain non-folded for translocase recognition/translocation. After translocation and signal peptide cleavage, mature domains fold to native states in the bacterial periplasm or traffic further. We sought the structural basis for delayed mature domain folding and how signal peptides regulate it. We compared how evolution diversified a periplasmic peptidyl-prolyl isomerase PpiA mature domain from its structural cytoplasmic PpiB twin. Global and local hydrogen-deuterium exchange mass spectrometry showed that PpiA is a slower folder. We defined at near-residue resolution hierarchical folding initiated by similar foldons in the twins, at different order and rates. PpiA folding is delayed by less hydrophobic native contacts, frustrated residues and a β-turn in the earliest foldon and by signal peptide-mediated disruption of foldon hierarchy. When selected PpiA residues and/or its signal peptide were grafted onto PpiB, they converted it into a slow folder with enhanced in vivo secretion. These structural adaptations in a secretory protein facilitate trafficking.

Protein folding in vitro and in the cell: From a solitary journey to a team effort

Correct protein folding is essential for the health and function of living organisms. Yet, it is not well understood how unfolded proteins reach their native state and avoid aggregation, especially within the cellular milieu. Some proteins, especially small, single-domain and apparent two-state folders, successfully attain their native state upon dilution from denaturant. Yet, many more proteins undergo misfolding and aggregation during this process, in a concentration-dependent fashion. Once formed, native and aggregated states are often kinetically trapped relative to each other. Hence, the early stages of protein life are absolutely critical for proper kinetic channeling to the folded state and for long-term solubility and function. This review summarizes current knowledge on protein folding/aggregation mechanisms in buffered solution and within the bacterial cell, highlighting early stages. Remarkably, teamwork between nascent chain, ribosome, trigger factor and Hsp70 molecular chaperones enables all proteins to overcome aggregation propensities and reach a long-lived bioactive state.

Nucleic Acid Thermodynamics Derived from Mechanical Unzipping Experiments

Force-spectroscopy techniques have led to significant progress in studying the physicochemical properties of biomolecules that are not accessible in bulk assays. The application of piconewton forces with laser optical tweezers to single nucleic acids has permitted the characterization of molecular thermodynamics and kinetics with unprecedented accuracy. Some examples are the hybridization reaction between complementary strands in DNA and the folding of secondary, tertiary, and other heterogeneous structures, such as intermediate and misfolded states in RNA. Here we review the results obtained in our lab on deriving the nearest-neighbor free energy parameters in DNA and RNA duplexes from mechanical unzipping experiments. Remarkable nonequilibrium effects are also observed, such as the large irreversibility of RNA unzipping and the formation of non-specific secondary structures in single-stranded DNA. These features originate from forming stem-loop structures along the single strands of the nucleic acid. The recently introduced barrier energy landscape model quantifies kinetic trapping effects due to stem-loops being applicable to both RNA and DNA. The barrier energy landscape model contains the essential features to explain the many behaviors observed in heterogeneous nucleic-acid folding.

DMSO-Quenched H/D-Exchange 2D NMR Spectroscopy and Its Applications in Protein Science

Hydrogen/deuterium (H/D) exchange combined with two-dimensional (2D) NMR spectroscopy has been widely used for studying the structure, stability, and dynamics of proteins. When we apply the H/D-exchange method to investigate non-native states of proteins such as equilibrium and kinetic folding intermediates, H/D-exchange quenching techniques are indispensable, because the exchange reaction is usually too fast to follow by 2D NMR. In this article, we will describe the dimethylsulfoxide (DMSO)-quenched H/D-exchange method and its applications in protein science. In this method, the H/D-exchange buffer is replaced by an aprotic DMSO solution, which quenches the exchange reaction. We have improved the DMSO-quenched method by using spin desalting columns, which are used for medium exchange from the H/D-exchange buffer to the DMSO solution. This improvement has allowed us to monitor the H/D exchange of proteins at a high concentration of salts or denaturants. We describe methodological details of the improved DMSO-quenched method and present a case study using the improved method on the H/D-exchange behavior of unfolded human ubiquitin in 6 M guanidinium chloride.

Multimodal approaches for the improvement of the cellular folding of a recombinant iron regulatory protein in E. coli

Background

During the recombinant protein expression, most heterologous proteins expressed in E. coli cell factories are generated as insoluble and inactive aggregates, which prohibit E. coli from being employed as an expression host despite its numerous advantages and ease of use. The yeast mitochondrial aconitase protein, which has a tendency to aggregate when expressed in E. coli cells in the absence of heterologous chaperones GroEL/ES was utilised as a model to investigate how the modulation of physiological stimuli in the host cell can increase protein solubility. The presence of folding modulators such as exogenous molecular chaperones or osmolytes, as well as process variables such as incubation temperature, inducer concentrations, growth media are all important for cellular folding and are investigated in this study. This study also investigated how the cell's stress response system activates and protects the proteins from aggregation.

Results

The cells exposed to osmolytes plus a pre-induction heat shock showed a substantial increase in recombinant aconitase activity when combined with modulation of process conditions. The concomitant GroEL/ES expression further assists the folding of these soluble aggregates and increases the functional protein molecules in the cytoplasm of the recombinant E. coli cells.

Conclusions

The recombinant E. coli cells enduring physiological stress provide a cytosolic environment for the enhancement in the solubility and activity of the recombinant proteins. GroEL/ES-expressing cells not only aided in the folding of recombinant proteins, but also had an effect on the physiology of the expression host. The improvement in the specific growth rate and aconitase production during chaperone GroEL/ES co-expression is attributed to the reduction in overall cellular stress caused by the expression host's aggregation-prone recombinant protein expression.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s12934-022-01749-w.

The A39G FF domain folds on a volcano-shaped free energy surface via separate pathways

Conformational dynamics play critical roles in protein folding, misfolding, function, misfunction, and aggregation. While detecting and studying the different conformational states populated by protein molecules on their free energy surfaces (FESs) remain a challenge, NMR spectroscopy has emerged as an invaluable experimental tool to explore the FES of a protein, as conformational dynamics can be probed at atomic resolution over a wide range of timescales. Here, we use chemical exchange saturation transfer (CEST) to detect "invisible" minor states on the energy landscape of the A39G mutant FF domain that exhibited "two-state" folding kinetics in traditional experiments. Although CEST has mostly been limited to studies of processes with rates between ∼5 to 300 s^-1 involving sparse states with populations as low as ∼1%, we show that the line broadening that is often associated with minor state dips in CEST profiles can be exploited to inform on additional conformers, with lifetimes an order of magnitude shorter and populations close to 10-fold smaller than what typically is characterized. Our analysis of CEST profiles that exploits the minor state linewidths of the 71-residue A39G FF domain establishes a folding mechanism that can be described in terms of a four-state exchange process between interconverting states spanning over two orders of magnitude in timescale from ∼100 to ∼15,000 μs. A similar folding scheme is established for the wild-type domain as well. The study shows that the folding of this small domain proceeds through a pair of sparse, partially structured intermediates via two discrete pathways on a volcano-shaped FES.

Mechanical couplings of protein backbone and side chains exhibit scale-free network properties and specific hotspots for function

•

Statistical learning from protein dynamics unravels rigidities in interaction network.

•

Backbone and side-chain mechanical couplings exhibit scale-free network properties.

•

Graphical depiction of network rigidities captures sequence co-evolution patterns.

•

Functional sites at secondary structure peripheries are mechanical hotspots.

•

Our rigidity scores are compelling metrics for residue biological significance.

A backbone-side-chain elastic network model (bsENM) is devised in this contribution to decipher the network of molecular interactions during protein dynamics. The chemical details in 5 μs all-atom molecular dynamics (MD) simulation are mapped onto the bsENM spring constants by self-consistent iterations. The elastic parameters obtained by this structure-mechanics statistical learning are then used to construct inter-residue rigidity graphs for the chemical components in protein amino acids. A key discovery is that the mechanical coupling strengths of both backbone and side chains exhibit heavy-tailed distributions and scale-free network properties. In both rat trypsin and PDZ3 proteins, the statistically prominent modes of rigidity graphs uncover the sequence-specific coupling patterns and mechanical hotspots. Based on the contributions to graphical modes, our residue rigidity scores in backbone and side chains are found to be very useful metrics for the biological significance. Most functional sites have high residue rigidity scores in side chains while the biologically important glycines are generally next to mechanical hotspots. Furthermore, prominent modes in the rigidity graphs involving side chains oftentimes coincide with the co-evolution patterns due to evolutionary restraints. The bsENM specifically devised to resolve the protein chemical character thus provides useful means for extracting functional information from all-atom MD.

A Review on Computational Approaches for Analyzing Hydrogen- Deuterium (H/D) Exchange of Proteins

Native state Hydrogen-Deuterium (H/D) exchange method has been used to study the structures and the unfolding pathways for quite a number of proteins. The H/D exchange method is generally monitored using nuclear magnetic resonance (NMR) spectroscopy and mass spectrometry (MS) techniques. NMR-assisted H/D exchange methods primarily monitor the residue level fluctuation of proteins, whereas MS-assisted H/D exchange methods analyze multifold ensemble conformations of proteins. In this connection, quite a large number of computational tools and algorithms have been developed for processing and analyzing huge amount of the H/D exchange data generated from these techniques. In this review, most of the freely available computational tools associated with the H/D exchange of proteins have been comprehensively reviewed and scopes to improve/ develop novel computational approaches for analyzing the H/D exchange data of proteins have also been brought into fore.

Cryptic intermediates and metastable states of proteins as predicted by OneG computational method

Understanding structural excursions of proteins under folding conditions is crucial to map energy landscapes of proteins. In the present study, OneG computational tool has been used for analyzing possible existence of cryptic intermediates and metastable states of 26 proteins for which three prerequisite inputs of the OneG such as atomic coordinates of proteins, free energy of unfolding (ΔG_U) and free energy of exchange (ΔG_HX) determined in the absence of denaturant were available during the course of the study. The veraciousness of the tool on predicting the partially folded states of the proteins has been comprehensively described using experimental data available for 15 of the 26 proteins. Meanwhile, possible existence of partially structured states in the folding pathways of 11 other proteins has also been delineated as predicted by the OneG. In addition to mapping the folding pathways of proteins, the salient merits of the tool on systematically addressing the discrepancy between the ΔG_U and the ΔG_HX of the proteins have also been dealt.Communicated by Ramaswamy H. Sarma.

Combining High-Pressure Perturbation with NMR Spectroscopy for a Structural and Dynamical Characterization of Protein Folding Pathways

High-hydrostatic pressure is an alternative perturbation method that can be used to destabilize globular proteins. Generally perfectly reversible, pressure exerts local effects on regions or domains of a protein containing internal voids, contrary to heat or chemical denaturant that destabilize protein structures uniformly. When combined with NMR spectroscopy, high pressure (HP) allows one to monitor at a residue-level resolution the structural transitions occurring upon unfolding and to determine the kinetic properties of the process. The use of HP-NMR has long been hampered by technical difficulties. Owing to the recent development of commercially available high-pressure sample cells, HP-NMR experiments can now be routinely performed. This review summarizes recent advances of HP-NMR techniques for the characterization at a quasi-atomic resolution of the protein folding energy landscape.

Structural dynamics of pentapeptide repeat proteins

Pentapeptide repeat proteins (PRPs) represent a large superfamily with more than 38 000 sequences in nearly 3500 species, the majority belonging to cyanobacteria but represented among all branches of life. PRPs contain at least eight consecutive pentapeptide repeats with the consensus (A/C/S/V/T/L/I)(D/N/S/K/E/I/R)(L/F)(S/T/R/E/Q/K/V/D)(G/D/E/N/R/Q/K). PRPs fold into right-handed quadrilateral β helices, also known as repeat-five-residue (Rfr)-folds, with four consecutive pentapeptide repeats comprising a single coil, the ~90° change in polypeptide direction in square-shaped coils achieved by type I, II and IV β turns, and hydrogen bonds between coils establishing β ladders on each Rfr-fold face. PRPs are broadly categorized into group 1 and 2 involved in antibiotic resistance and group 3 currently having unknown functions. Motivated by their intriguing structures, we are investigating PRP biophysical characteristics, including Rfr-fold thermal stability, β turn and β ladder hydrogen bond amide exchange rates and backbone dynamics. Here, we present analysis of 20 ns molecular dynamics (MD) simulations and all atom normal mode analysis (aaNMA) calculations for four group 1 and group 2 and four group 3 PRPs whose structures have been determined by X-ray crystallography. The MD cross-correlation matrices and aaNMA indicated strong correlated motion between adjacent coils and weak coupled motion between coils separated by one or more intervening coils. Slow anticorrelated motions were detected between adjacent coils in aaNMA modes that we hypothesize are requisite to access exchange-competent states necessary to permit solvent exchange of amide hydrogens involved in β-ladder and β-turns hydrogen bonds, which can have lifetimes on the order of months.

Kinetic control of stationary flux ratios for a wide range of biochemical processes

One of the most intriguing features of biological systems is their ability to regulate the steady-state fluxes of the underlying biochemical reactions; however, the regulatory mechanisms and their physicochemical properties are not fully understood. Fundamentally, flux regulation can be explained with a chemical kinetic formalism describing the transitions between discrete states, with the reaction rates defined by an underlying free energy landscape. Which features of the energy landscape affect the flux distribution? Here we prove that the ratios of the steady-state fluxes of quasi-first-order biochemical processes are invariant to energy perturbations of the discrete states and are only affected by the energy barriers. In other words, the nonequilibrium flux distribution is under kinetic and not thermodynamic control. We illustrate the generality of this result for three biological processes. For the network describing protein folding along competing pathways, the probabilities of proceeding via these pathways are shown to be invariant to the stability of the intermediates or to the presence of additional misfolded states. For the network describing protein synthesis, the error rate and the energy expenditure per peptide bond is proven to be independent of the stability of the intermediate states. For molecular motors such as myosin-V, the ratio of forward to backward steps and the number of adenosine 5'-triphosphate (ATP) molecules hydrolyzed per step is demonstrated to be invariant to energy perturbations of the intermediate states. These findings place important constraints on the ability of mutations and drug perturbations to affect the steady-state flux distribution for a wide class of biological processes.

Life in Phases: Intra- and Inter- Molecular Phase Transitions in Protein Solutions

Proteins, these evolutionarily-edited biological polymers, are able to undergo intramolecular and intermolecular phase transitions. Spontaneous intramolecular phase transitions define the folding of globular proteins, whereas binding-induced, intra- and inter- molecular phase transitions play a crucial role in the functionality of many intrinsically-disordered proteins. On the other hand, intermolecular phase transitions are the behind-the-scenes players in a diverse set of macrosystemic phenomena taking place in protein solutions, such as new phase nucleation in bulk, on the interface, and on the impurities, protein crystallization, protein aggregation, the formation of amyloid fibrils, and intermolecular liquid-liquid or liquid-gel phase transitions associated with the biogenesis of membraneless organelles in the cells. This review is dedicated to the systematic analysis of the phase behavior of protein molecules and their ensembles, and provides a description of the major physical principles governing intramolecular and intermolecular phase transitions in protein solutions.

Effects of Pro1266Leu mutation on structure and function of glycoprotein Ib binding domain of von Willebrand factor

Glycoprotein Ibα (GpIbα) binding ability of A1 domain of von Willebrand factor (vWF) facilitates platelet adhesion that plays a crucial role in maintaining hemostasis and thrombosis at the site of vascular damage. There are both "loss as well as gain of function" mutations observed in this domain. Naturally occurring "gain of function" mutations leave self-activating impacts on the A1 domain which turns the normal binding to characteristic constitutive binding with GPIbα. These "gain of function" mutations are associated with the von Willebrand disease type 2B. In recent years, studies focused on understanding the mechanism and conformational patterns attached to these phenomena have been conducted, but the conformational pathways leading to such binding patterns are poorly understood as of now. To obtain a microscopic picture of such events for the better understanding of pathways, we used molecular dynamics (MD) simulations along with principal component analysis and normal mode analysis to study the effects of Pro1266Leu (Pro503Leu in structural context) mutation on the structure and function of A1 domain of vWF. MD simulations have provided atomic-level details of intermolecular motions as a function of time to understand the dynamic behavior of A1 domain of vWF. Comparative analysis of the trajectories obtained from MD simulations of both the wild type and Pro503Leu mutant suggesting appreciable conformational changes in the structure of mutant which might provide a basis for assuming the "gain of function" effects of these mutations on the A1 domain of vWF, resulting in the constitutive binding with GpIbα.

Application of an interpretable classification model on Early Folding Residues during protein folding

Background

Machine learning strategies are prominent tools for data analysis. Especially in life sciences, they have become increasingly important to handle the growing datasets collected by the scientific community. Meanwhile, algorithms improve in performance, but also gain complexity, and tend to neglect interpretability and comprehensiveness of the resulting models.

Results

Generalized Matrix Learning Vector Quantization (GMLVQ) is a supervised, prototype-based machine learning method and provides comprehensive visualization capabilities not present in other classifiers which allow for a fine-grained interpretation of the data. In contrast to commonly used machine learning strategies, GMLVQ is well-suited for imbalanced classification problems which are frequent in life sciences. We present a Weka plug-in implementing GMLVQ. The feasibility of GMLVQ is demonstrated on a dataset of Early Folding Residues (EFR) that have been shown to initiate and guide the protein folding process. Using 27 features, an area under the receiver operating characteristic of 76.6% was achieved which is comparable to other state-of-the-art classifiers. The obtained model is accessible at https://biosciences.hs-mittweida.de/efpred/.

Conclusions

The application on EFR prediction demonstrates how an easy interpretation of classification models can promote the comprehension of biological mechanisms. The results shed light on the special features of EFR which were reported as most influential for the classification: EFR are embedded in ordered secondary structure elements and they participate in networks of hydrophobic residues. Visualization capabilities of GMLVQ are presented as we demonstrate how to interpret the results.

Electronic supplementary material

The online version of this article (10.1186/s13040-018-0188-2) contains supplementary material, which is available to authorized users.

Self-Assembly of Ferritin: Structure, Biological Function and Potential Applications in Nanotechnology

Protein cages are normally formed by the self-assembly of multiple protein subunits and ferritin is a typical example of a protein cage structure. Ferritin is a ubiquitous multi-subunit iron storage protein formed by 24 polypeptide chains that self-assemble into a hollow, roughly spherical protein cage. Ferritin has external and internal diameters of approximately 12 nm and 8 nm, respectively. Functionally, ferritin performs iron sequestration and is highly conserved in evolution. The interior cavity of ferritin provides a unique reaction vessel to carry out reactions separated from the exterior environment. In nature, the cavity is utilized for sequestration of iron and bio-mineralization as a mechanism to render iron inert and safe from the external environment. Material scientists have been inspired by this system and exploited a range of ferritin superfamily proteins as supramolecular templates to encapsulate different carrier molecules ranging from cancer drugs to therapeutic proteins, in addition to using ferritin proteins as well-defined building blocks for fabrication. Besides the interior cavity, the exterior surface and sub-unit interface of ferritin can be modified without affecting ferritin assembly.

Complex Folding Landscape of Apomyoglobin at Acidic pH Revealed by Ultrafast Kinetic Analysis of Core Mutants

Under mildly acidic conditions (pH 4-4.5) apomyoglobin (apoMb) adopts a partially structured equilibrium state ( M-state) that structurally resembles a kinetic intermediate encountered at a late stage of folding to the native structure at neutral pH. We have previously reported that the M-state is formed rapidly (<1 ms) via a multistate process and thus offers a unique opportunity for exploring early stages of folding by both experimental and computational techniques. In order to gain structural insight into intermediates and barriers at the residue level, we studied the folding/unfolding kinetics of 12 apoMb mutants at pH 4.2 using fluorescence-detected ultrafast mixing techniques. Global analysis of the submillisecond folding/unfolding kinetics vs urea concentration for each variant, based on a sequential four-state mechanism ( U ⇔ I ⇔ L ⇔ M), allowed us to determine elementary rate constants and their dependence on urea concentration for most transitions. Comparison of the free energy diagrams constructed from the kinetic data of the mutants with that of wild-type apoMb yielded quantitative information on the effects of mutations on the free energy (ΔΔ G) of both intermediates and the first two kinetic barriers encountered during folding. Truncation of conserved aliphatic side chains on helices A, G, and H gives rise to a stepwise increase in ΔΔ G as the protein advances from U toward M, consistent with progressive stabilization of native-like contacts within the primary core of apoMb. Helix-helix contacts in the primary core contribute little to the first folding barrier ( U ⇔ I) and thus are not required for folding initiation but are critical for the stability of the late intermediate, L, and the M-state. Alanine substitution of hydrophobic residues at more peripheral helix-helix contact sites of the native structure, which are still absent or unstable in the M-state, shows both positive (destabilizing) and negative (stabilizing) ΔΔ G, indicating that non-native contacts are formed initially and weakened or lost as a result of subsequent structural rearrangement steps.

Characterizing the relation of functional and Early Folding Residues in protein structures using the example of aminoacyl-tRNA synthetases

Proteins are chains of amino acids which adopt a three-dimensional structure and are then able to catalyze chemical reactions or propagate signals in organisms. Without external influence, many proteins fold into their native structure, and a small number of Early Folding Residues (EFR) have previously been shown to initiate the formation of secondary structure elements and guide their respective assembly. Using the two diverse superfamilies of aminoacyl-tRNA synthetases (aaRS), it is shown that the position of EFR is preserved over the course of evolution even when the corresponding sequence conservation is small. Folding initiation sites are positioned in the center of secondary structure elements, independent of aaRS class. In class I, the predicted position of EFR resembles an ancient structural packing motif present in many seemingly unrelated proteins. Furthermore, it is shown that EFR and functionally relevant residues in aaRS are almost entirely disjoint sets of residues. The Start2Fold database is used to investigate whether this separation of EFR and functional residues can be observed for other proteins. EFR are found to constitute crucial connectors of protein regions which are distant at sequence level. Especially, these residues exhibit a high number of non-covalent residue-residue contacts such as hydrogen bonds and hydrophobic interactions. This tendency also manifests as energetically stable local regions, as substantiated by a knowledge-based potential. Despite profound differences regarding how EFR and functional residues are embedded in protein structures, a strict separation of structurally and functionally relevant residues cannot be observed for a more general collection of proteins.

Evidence for a Partially Stalled γ Rotor in F1-ATPase from Hydrogen-Deuterium Exchange Experiments and Molecular Dynamics Simulations

F₁-ATPase uses ATP hydrolysis to drive rotation of the γ subunit. The γ C-terminal helix constitutes the rotor tip that is seated in an apical bearing formed by α₃β₃. It remains uncertain to what extent the γ conformation during rotation differs from that seen in rigid crystal structures. Existing models assume that the entire γ subunit participates in every rotation. Here we interrogated E. coli F₁-ATPase by hydrogen-deuterium exchange (HDX) mass spectrometry. Rotation of γ caused greatly enhanced deuteration in the γ C-terminal helix. The HDX kinetics implied that most F₁ complexes operate with an intact rotor at any given time, but that the rotor tip is prone to occasional unfolding. A molecular dynamics (MD) strategy was developed to model the off-axis forces acting on γ. MD runs showed stalling of the rotor tip and unfolding of the γ C-terminal helix. MD-predicted H-bond opening events coincided with experimental HDX patterns. Our data suggest that in vitro operation of F₁-ATPase is associated with significant rotational resistance in the apical bearing. These conditions cause the γ C-terminal helix to get "stuck" (and unfold) sporadically while the remainder of γ continues to rotate. This scenario contrasts the traditional "greasy bearing" model that envisions smooth rotation of the γ C-terminal helix. The fragility of the apical rotor tip in F₁-ATPase is attributed to the absence of a c₁₀ ring that stabilizes the rotation axis in intact F_oF₁. Overall, the MD/HDX strategy introduced here appears well suited for interrogating the inner workings of molecular motors.

Finding the needle in the haystack: towards solving the protein-folding problem computationally

Prediction of protein tertiary structures from amino acid sequence and understanding the mechanisms of how proteins fold, collectively known as "the protein folding problem," has been a grand challenge in molecular biology for over half a century. Theories have been developed that provide us with an unprecedented understanding of protein folding mechanisms. However, computational simulation of protein folding is still difficult, and prediction of protein tertiary structure from amino acid sequence is an unsolved problem. Progress toward a satisfying solution has been slow due to challenges in sampling the vast conformational space and deriving sufficiently accurate energy functions. Nevertheless, several techniques and algorithms have been adopted to overcome these challenges, and the last two decades have seen exciting advances in enhanced sampling algorithms, computational power and tertiary structure prediction methodologies. This review aims at summarizing these computational techniques, specifically conformational sampling algorithms and energy approximations that have been frequently used to study protein-folding mechanisms or to de novo predict protein tertiary structures. We hope that this review can serve as an overview on how the protein-folding problem can be studied computationally and, in cases where experimental approaches are prohibitive, help the researcher choose the most relevant computational approach for the problem at hand. We conclude with a summary of current challenges faced and an outlook on potential future directions.

Monitoring protein folding through high pressure NMR spectroscopy

High-pressure is a well-known perturbation method used to destabilize globular proteins. It is perfectly reversible, which is essential for a proper thermodynamic characterization of a protein equilibrium. In contrast to other perturbation methods such as heat or chemical denaturant that destabilize protein structures uniformly, pressure exerts local effects on regions or domains of a protein containing internal cavities. When combined with NMR spectroscopy, hydrostatic pressure offers the possibility to monitor at a residue level the structural transitions occurring upon unfolding and to determine the kinetic properties of the process. High-pressure NMR experiments can now be routinely performed, owing to the recent development of commercially available high-pressure sample cells. This review summarizes recent advances and some future directions of high-pressure NMR techniques for the characterization at atomic resolution of the energy landscape of protein folding.

Exploring the Sequence-based Prediction of Folding Initiation Sites in Proteins

Protein folding is a complex process that can lead to disease when it fails. Especially poorly understood are the very early stages of protein folding, which are likely defined by intrinsic local interactions between amino acids close to each other in the protein sequence. We here present EFoldMine, a method that predicts, from the primary amino acid sequence of a protein, which amino acids are likely involved in early folding events. The method is based on early folding data from hydrogen deuterium exchange (HDX) data from NMR pulsed labelling experiments, and uses backbone and sidechain dynamics as well as secondary structure propensities as features. The EFoldMine predictions give insights into the folding process, as illustrated by a qualitative comparison with independent experimental observations. Furthermore, on a quantitative proteome scale, the predicted early folding residues tend to become the residues that interact the most in the folded structure, and they are often residues that display evolutionary covariation. The connection of the EFoldMine predictions with both folding pathway data and the folded protein structure suggests that the initial statistical behavior of the protein chain with respect to local structure formation has a lasting effect on its subsequent states.

Competing Pathways and Multiple Folding Nuclei in a Large Multidomain Protein, Luciferase

Proteins obtain their final functional configuration through incremental folding with many intermediate steps in the folding pathway. If known, these intermediate steps could be valuable new targets for designing therapeutics and the sequence of events could elucidate the mechanism of refolding. However, determining these intermediate steps is hardly an easy feat, and has been elusive for most proteins, especially large, multidomain proteins. Here, we effectively map part of the folding pathway for the model large multidomain protein, Luciferase, by combining single-molecule force-spectroscopy experiments and coarse-grained simulation. Single-molecule refolding experiments reveal the initial nucleation of folding while simulations corroborate these stable core structures of Luciferase, and indicate the relative propensities for each to propagate to the final folded native state. Both experimental refolding and Monte Carlo simulations of Markov state models generated from simulation reveal that Luciferase most often folds along a pathway originating from the nucleation of the N-terminal domain, and that this pathway is the least likely to form nonnative structures. We then engineer truncated variants of Luciferase whose sequences corresponded to the putative structure from simulation and we use atomic force spectroscopy to determine their unfolding and stability. These experimental results corroborate the structures predicted from the folding simulation and strongly suggest that they are intermediates along the folding pathway. Taken together, our results suggest that initial Luciferase refolding occurs along a vectorial pathway and also suggest a mechanism that chaperones may exploit to prevent misfolding.

Protein folding by NMR

Protein folding is a highly complex process proceeding through a number of disordered and partially folded nonnative states with various degrees of structural organization. These transiently and sparsely populated species on the protein folding energy landscape play crucial roles in driving folding toward the native conformation, yet some of these nonnative states may also serve as precursors for protein misfolding and aggregation associated with a range of devastating diseases, including neuro-degeneration, diabetes and cancer. Therefore, in vivo protein folding is often reshaped co- and post-translationally through interactions with the ribosome, molecular chaperones and/or other cellular components. Owing to developments in instrumentation and methodology, solution NMR spectroscopy has emerged as the central experimental approach for the detailed characterization of the complex protein folding processes in vitro and in vivo. NMR relaxation dispersion and saturation transfer methods provide the means for a detailed characterization of protein folding kinetics and thermodynamics under native-like conditions, as well as modeling high-resolution structures of weakly populated short-lived conformational states on the protein folding energy landscape. Continuing development of isotope labeling strategies and NMR methods to probe high molecular weight protein assemblies, along with advances of in-cell NMR, have recently allowed protein folding to be studied in the context of ribosome-nascent chain complexes and molecular chaperones, and even inside living cells. Here we review solution NMR approaches to investigate the protein folding energy landscape, and discuss selected applications of NMR methodology to studying protein folding in vitro and in vivo. Together, these examples highlight a vast potential of solution NMR in providing atomistic insights into molecular mechanisms of protein folding and homeostasis in health and disease.

Automation of peak-tracking analysis of stepwise perturbed NMR spectra

We describe a new algorithmic approach able to automatically pick and track the NMR resonances of a large number of 2D NMR spectra acquired during a stepwise variation of a physical parameter. The method has been named Trace in Track (TINT), referring to the idea that a gaussian decomposition traces peaks within the tracks recognised through 3D mathematical morphology. It is capable of determining the evolution of the chemical shifts, intensity and linewidths of each tracked peak.The performances obtained in term of track reconstruction and correct assignment on realistic synthetic spectra were high above 90% when a noise level similar to that of experimental data were considered. TINT was applied successfully to several protein systems during a temperature ramp in isotope exchange experiments. A comparison with a state-of-the-art algorithm showed promising results for great numbers of spectra and low signal to noise ratios, when the graduality of the perturbation is appropriate. TINT can be applied to different kinds of high throughput chemical shift mapping experiments, with quasi-continuous variations, in which a quantitative automated recognition is crucial.

Evidence of Allosteric Enzyme Regulation via Changes in Conformational Dynamics: A Hydrogen/Deuterium Exchange Investigation of Dihydrodipicolinate Synthase

Dihydrodipicolinate synthase is a tetrameric enzyme of the diaminopimelate pathway in bacteria and plants. The protein catalyzes the condensation of pyruvate (Pyr) and aspartate semialdehyde en route to the end product lysine (Lys). Dihydrodipicolinate synthase from Campylobacter jejuni (CjDHDPS) is allosterically inhibited by Lys. CjDHDPS is a promising antibiotic target, as highlighted by the recent development of a potent bis-lysine (bisLys) inhibitor. The mechanism whereby Lys and bisLys allosterically inhibit CjDHDPS remains poorly understood. In contrast to the case for other allosteric enzymes, crystallographically detectable conformational changes in CjDHDPS upon inhibitor binding are very minor. Also, it is difficult to envision how Pyr can access the active site; the available X-ray data seemingly imply that each turnover step requires diffusion-based mass transfer through a narrow access channel. This study employs hydrogen/deuterium exchange mass spectrometry for probing the structure and dynamics of CjDHDPS in a native solution environment. The deuteration kinetics reveal that the most dynamic protein regions are in the direct vicinity of the substrate access channel. This finding is consistent with the view that transient opening/closing fluctuations facilitate access of the substrate to the active site. Under saturating conditions, both Lys and bisLys cause dramatically reduced dynamics in the inhibitor binding region. In addition, rigidification extends to regions close to the substrate access channel. This finding strongly suggests that allosteric inhibitors interfere with conformational fluctuations that are required for CjDHDPS substrate turnover. In particular, our data imply that Lys and bisLys suppress opening/closing events of the access channel, thereby impeding diffusion of the substrate into the active site. Overall, this work illustrates why allosteric control does not have to be associated with crystallographically detectable large-scale transitions. Our experiments provide evidence that in CjDHDPS allostery is mediated by changes in the extent of thermally activated conformational fluctuations.

Fold and flexibility: what can proteins' mechanical properties tell us about their folding nucleus?

The determination of a protein's folding nucleus, i.e. a set of native contacts playing an important role during its folding process, remains an elusive yet essential problem in biochemistry. In this work, we investigate the mechanical properties of 70 protein structures belonging to 14 protein families presenting various folds using coarse-grain Brownian dynamics simulations. The resulting rigidity profiles combined with multiple sequence alignments show that a limited set of rigid residues, which we call the consensus nucleus, occupy conserved positions along the protein sequence. These residues' side chains form a tight interaction network within the protein's core, thus making our consensus nuclei potential folding nuclei. A review of experimental and theoretical literature shows that most (above 80%) of these residues were indeed identified as folding nucleus member in earlier studies.

Kinetics of peptide folding in lipid membranes

Despite our extensive understanding of water-soluble protein folding kinetics, much less is known about the folding dynamics and mechanisms of membrane proteins. However, recent studies have shown that for relatively simple systems, such as peptides that form a transmembrane α-helix, helical dimer, or helix-turn-helix, it is possible to assess the kinetics of several important steps, including peptide binding to the membrane from aqueous solution, peptide folding on the membrane surface, helix insertion into the membrane, and helix-helix association inside the membrane. Herein, we provide a brief review of these studies and also suggest new initiation and probing methods that could lead to improved temporal and structural resolution in future experiments.

Intrinsic local destabilization of the C-terminus predisposes integrin α1 I domain to a conformational switch induced by collagen binding

Integrin-collagen interactions play a critical role in a myriad of cellular functions that include immune response, and cell development and differentiation, yet their mechanism of binding is poorly understood. There is increasing evidence that conformational flexibility assumes a central role in the molecular mechanisms of protein-protein interactions and here we employ NMR hydrogen-deuterium exchange (HDX) experiments to explore the impact of slower timescale dynamic events. To gain insight into the mechanisms underlying collagen-induced conformational switches, we have undertaken a comparative study between the wild type integrin α1 I and a gain-of-function E317A mutant. NMR HDX results suggest a relationship between regions exhibiting a reduced local stability in the unbound I domain and those that undergo significant conformational changes upon binding. Specifically, the αC and α7 helices within the C-terminus are at the center of such major perturbations and present reduced local stabilities in the unbound state relative to other structural elements. Complementary isothermal titration calorimetry experiments have been performed to derive complete thermodynamic binding profiles for association of the collagen-like triple-helical peptide with wild type α1 I and E317A mutant. The differential energetics observed for E317A are consistent with the HDX experiments and support a model in which intrinsically destabilized regions predispose conformational rearrangement in the integrin I domain. This study highlights the importance of exploring different timescales to delineate allosteric and binding events.

Residual structures in the unfolded state of starch-binding domain of glucoamylase revealed by near-UV circular dichroism and protein engineering techniques

Protein folding is a thermodynamic process driven by energy gaps between the native and unfolded states. Although a wealth of information is available on the structure of folded species, there is a paucity of data on unfolded species. Here, we analyzed the structural properties of the unfolded state of the starch-binding domain of glucoamylase from Aspergillus niger (SBD) formed in the presence of guanidinium hydrochloride (GuHCl). Although far-UV CD and intrinsic tryptophan fluorescence spectra as well as small angle X-ray scattering suggested that SBD assumes highly unfolded structures in the presence of GuHCl, near-UV circular dichroism of wild-type SBD suggested the presence of residual structures in the unfolded state. Analyses of the unfolded states of tryptophan mutants (W543L, W563A, W590A and W615L) using Similarity Parameter, a modified version of root mean square deviation as a measure of similarity between two spectra, suggested that W543 and W563 have preferences to form native-like residual structures in the GuHCl-unfolded state. In contrast, W615 was entirely unstructured, while W590 tended to form non-native ordered structures in the unfolded state. These data and the amino acid sequence of SBD suggest that local structural propensities in the unfolded state can be determined by the probability of the presence of hydrophobic or charged residues nearby tryptophan residues.

Network representation of conformational transitions between hidden intermediates of Rd-apocytochrome b562

The folding kinetics of Rd-apocytochrome b562 is two-state, but native-state hydrogen exchange experiments show that there are discrete partially unfolded (PUF) structures in equilibrium with the native state. These PUF structures are called hidden intermediates because they are not detected in kinetic experiments and they exist after the rate-limiting step. Structures of the mimics of hidden intermediates of Rd-apocytochrome b562 are resolved by NMR. Based upon their relative stability and structural features, the folding mechanism was proposed to follow a specific pathway (unfolded → rate-limiting transition state → PUF1 → PUF2 → native). Investigating the roles of equilibrium PUF structures in folding kinetics and their interrelationship not only deepens our understanding of the details of folding mechanism but also provides guides in protein design and prevention of misfolding. We performed molecular dynamics simulations starting from a hidden intermediate and the native state of Rd-apocytochrome b562 in explicit solvent, for a total of 37.18 μs mainly with Anton. We validated our simulations by detailed comparison with experimental data and other computations. We have verified that we sampled the post rate-limiting transition state region only. Markov state model was used to analyze the simulation results. We replace the specific pathway model with a network model. Transition-path theory was employed to calculate the net effective flux from the most unfolded state towards the most folded state in the network. The proposed sequential folding pathway via PUF1 then more stable, more native-like PUF2 is one of the routes in our network, but it is not dominant. The dominant path visits PUF2 without going through PUF1. There is also a route from PUF1 directly to the most folded state in the network without visiting PUF2. Our results indicate that the PUF states are not necessarily sequential in the folding. The major routes predicted in our network are testable by future experiments such as single molecule experiment.

Under-folded proteins: Conformational ensembles and their roles in protein folding, function, and pathogenesis

For decades, protein function was intimately linked to the presence of a unique, aperiodic crystal‐like structure in a functional protein. The two only places for conformational ensembles of under‐folded (or partially folded) protein forms in this picture were either the end points of the protein denaturation processes or transiently populated folding intermediates. Recent years witnessed dramatic change in this perception and conformational ensembles, which the under‐folded proteins are, have moved from the shadow. Accumulated to date data suggest that a protein can exist in at least three global forms–functional and folded, functional and intrinsically disordered (nonfolded), and nonfunctional and misfolded/aggregated. Under‐folded protein states are crucial for each of these forms, serving as important folding intermediates of ordered proteins, or as functional states of intrinsically disordered proteins (IDPs) and IDP regions (IDPRs), or as pathology triggers of misfolded proteins. Based on these observations, conformational ensembles of under‐folded proteins can be classified as transient (folding and misfolding intermediates) and permanent (IDPs and stable misfolded proteins). Permanently under‐folded proteins can further be split into intentionally designed (IDPs and IDPRs) and unintentionally designed (misfolded proteins). Although intrinsic flexibility, dynamics, and pliability are crucial for all under‐folded proteins, the different categories of under‐foldedness are differently encoded in protein amino acid sequences. © 2013 Wiley Periodicals, Inc. Biopolymers 99: 870–887, 2013.