Equilibrium and kinetics of the allosteric transition of GroEL studied by solution X-ray scattering and fluorescence spectroscopy

Chaperonin-assisted protein folding: a chronologue

This chronologue seeks to document the discovery and development of an understanding of oligomeric ring protein assemblies known as chaperonins that assist protein folding in the cell. It provides detail regarding genetic, physiologic, biochemical, and biophysical studies of these ATP-utilizing machines from both in vivo and in vitro observations. The chronologue is organized into various topics of physiology and mechanism, for each of which a chronologic order is generally followed. The text is liberally illustrated to provide firsthand inspection of the key pieces of experimental data that propelled this field. Because of the length and depth of this piece, the use of the outline as a guide for selected reading is encouraged, but it should also be of help in pursuing the text in direct order.

Assessing heterogeneity in oligomeric AAA+ machines

ATPases Associated with various cellular Activities (AAA+ ATPases) are molecular motors that use the energy of ATP binding and hydrolysis to remodel their target macromolecules. The majority of these ATPases form ring-shaped hexamers in which the active sites are located at the interfaces between neighboring subunits. Structural changes initiate in an active site and propagate to distant motor parts that interface and reshape the target macromolecules, thereby performing mechanical work. During the functioning cycle, the AAA+ motor transits through multiple distinct states. Ring architecture and placement of the catalytic sites at the intersubunit interfaces allow for a unique level of coordination among subunits of the motor. This in turn results in conformational differences among subunits and overall asymmetry of the motor ring as it functions. To date, a large amount of structural information has been gathered for different AAA+ motors, but even for the most characterized of them only a few structural states are known and the full mechanistic cycle cannot be yet reconstructed. Therefore, the first part of this work will provide a broad overview of what arrangements of AAA+ subunits have been structurally observed focusing on diversity of ATPase oligomeric ensembles and heterogeneity within the ensembles. The second part of this review will concentrate on methods that assess structural and functional heterogeneity among subunits of AAA+ motors, thus bringing us closer to understanding the mechanism of these fascinating molecular motors.

Analytical methods for kinetic studies of biological interactions: A review

The rates at which biological interactions occur can provide important information concerning the mechanism and behavior of these processes in living systems. This review discusses several analytical methods that can be used to examine the kinetics of biological interactions. These techniques include common or traditional methods such as stopped-flow analysis and surface plasmon resonance spectroscopy, as well as alternative methods based on affinity chromatography and capillary electrophoresis. The general principles and theory behind these approaches are examined, and it is shown how each technique can be utilized to provide information on the kinetics of biological interactions. Examples of applications are also given for each method. In addition, a discussion is provided on the relative advantages or potential limitations of each technique regarding its use in kinetic studies.

The H/D-exchange kinetics of the Escherichia coli co-chaperonin GroES studied by 2D NMR and DMSO-quenched exchange methods

We studied hydrogen/deuterium-exchange reactions of peptide amide protons of GroES using two different techniques: (1) two-dimensional (1)H-(15)N transverse-optimized NMR spectroscopy and (2) the dimethylsulfoxide-quenched hydrogen-exchange method combined with conventional (1)H-(15)N heteronuclear single quantum coherence spectroscopy. By using these techniques together with direct heteronuclear single quantum coherence experiments, we quantitatively evaluated the exchange rates for 33 out of the 94 peptide amide protons of GroES and their protection factors, and for the remaining 61 residues, we obtained the lower limits of the exchange rates. The protection factors of the most highly protected amide protons were on the order of 10(6)-10(7), and the values were comparable in magnitude to those observed in typical small globular proteins, but the number of the highly protected amide protons with a protection factor larger than 10(6) was only 10, significantly smaller than the numbers reported for the small globular proteins, indicating that significant portions of free heptameric GroES are flexible and natively unfolded. The highly protected amino acid residues with a protection factor larger than 10(5) were mainly located in three β-strands that form the hydrophobic core of GroES, while the residues in a mobile loop (residues 17-34) were not highly protected. The protection factors of the most highly protected amide protons were orders of magnitude larger than the value expected from the equilibrium unfolding parameters previously reported, strongly suggesting that the equilibrium unfolding of GroES is more complicated than a simple two-state or three-state mechanism and may involve more than a single intermediate.

Statistical mechanics of Monod-Wyman-Changeux (MWC) models

The 50th anniversary of the classic Monod-Wyman-Changeux (MWC) model provides an opportunity to survey the broader conceptual and quantitative implications of this quintessential biophysical model. With the use of statistical mechanics, the mathematical implementation of the MWC concept links problems that seem otherwise to have no ostensible biological connection including ligand-receptor binding, ligand-gated ion channels, chemotaxis, chromatin structure and gene regulation. Hence, a thorough mathematical analysis of the MWC model can illuminate the performance limits of a number of unrelated biological systems in one stroke. The goal of our review is twofold. First, we describe in detail the general physical principles that are used to derive the activity of MWC molecules as a function of their regulatory ligands. Second, we illustrate the power of ideas from information theory and dynamical systems for quantifying how well the output of MWC molecules tracks their sensory input, giving a sense of the "design" constraints faced by these receptors.

Nucleotide-induced conformational changes of tetradecameric GroEL mapped by H/D exchange monitored by FT-ICR mass spectrometry

Here we employ hydrogen/deuterium exchange mass spectrometry (HDX-MS) to access E. coli chaperonin GroEL conformation. The ~800 kDa tetradecameric GroEL plays an essential role in the proper folding of many proteins. Previous studies of the structural dynamics of GroEL upon ATP binding have been inconsistent, showing either minimal or major allosteric changes. Our results, based on the native, non-mutated, protein under physiological conditions in solution demonstrate substantial changes in conformation and/or flexibility upon ATP binding. We capture the pivotal step in its functional cycle by use of a non-hydrolyzable ATP analog, ATPγS, to mimic the ATP-bound GroEL state. Comparison of HDX-MS results for apo GroEL and GroEL-ATPγS enables the characterization of the nucleotide-regulated conformational changes throughout the entire protein with high sequence resolution. The 14-mer GroEL complex is the largest protein assembly yet accessed by HDX-MS, with sequence resolution of segments of as few as five amino acids.

Dissecting a bimolecular process of MgATP²- binding to the chaperonin GroEL

Although allosteric transitions of GroEL by MgATP(2)(-) have been widely studied, the initial bimolecular step of MgATP(2-) binding to GroEL remains unclear. Here, we studied the equilibrium and kinetics of MgATP(2)(-) binding to a variant of GroEL, in which Tyr485 was replaced by tryptophan, via isothermal titration calorimetry (ITC) and stopped-flow fluorescence spectroscopy. In the absence of K(+) at 4-5 °C, the allosteric transitions and the subsequent ATP hydrolysis by GroEL are halted, and hence, the stopped-flow fluorescence kinetics induced by rapid mixing of MgATP(2)(-) and the GroEL variant solely reflected MgATP(2)(-) binding, which was well represented by bimolecular noncooperative binding with a binding rate constant, k(on), of 9.14×10(4) M(-1) s(-1) and a dissociation rate constant, k(off), of 14.2 s(-1), yielding a binding constant, K(b) (=k(on)/k(off)), of 6.4×10(3) M(-1). We also successfully performed ITC to measure binding isotherms of MgATP(2)(-) to GroEL and obtained a K(b) of 9.5×10(3) M(-1) and a binding stoichiometric number of 6.6. K(b) was thus in good agreement with that obtained by stopped-flow fluorescence. In the presence of 10-50 mM KCl, the fluorescence kinetics consisted of three to four phases (the first fluorescence-increasing phase, followed by one or two exponential fluorescence-decreasing phases, and the final slow fluorescence-increasing phase), and comparison of the kinetics in the absence and presence of K(+) clearly demonstrated that the first fluorescence-increasing phase corresponds to bimolecular MgATP(2)(-) binding to GroEL. The temperature dependence of the kinetics indicated that MgATP(2)(-) binding to GroEL was activation-controlled with an activation enthalpy as large as 14-16 kcal mol(-1).

Characterisation of a GroEL single-ring mutant that supports growth of Escherichia coli and has GroES-dependent ATPase activity

Binding and folding of substrate proteins by the molecular chaperone GroEL alternates between its two seven-membered rings in an ATP-regulated manner. The association of ATP and GroES to a polypeptide-bound ring of GroEL encapsulates the folding proteins in the central cavity of that ring (cis ring) and allows it to fold in a protected environment where the risk of aggregation is reduced. ATP hydrolysis in the cis ring changes the potentials within the system such that ATP binding to the opposite (trans) ring triggers the release of all ligands from the cis ring of GroEL through a complex network of allosteric communication between the rings. Inter-ring allosteric communication thus appears indispensable for the function of GroEL, and an engineered single-ring version (SR1) cannot substitute for GroEL in vivo. We describe here the isolation and characterisation of an active single-ring form of the GroEL protein (SR-A92T), which has an exceptionally low ATPase activity that is strongly stimulated by the addition of GroES. Dissection of the kinetic pathway of the ATP-induced structural changes in this active single ring can be explained by the fact that the mutation effectively blocks progression through the full allosteric pathway of the GroEL reaction cycle, thus trapping an early allosteric intermediate. Addition of GroES is able to overcome this block by binding this intermediate and pulling the allosteric pathway to completion via mass action, explaining how bacterial cells expressing this protein as their only chaperonin are viable.

Chaperonin-mediated protein folding: using a central cavity to kinetically assist polypeptide chain folding

The chaperonin ring assembly GroEL provides kinetic assistance to protein folding in the cell by binding non-native protein in the hydrophobic central cavity of an open ring and subsequently, upon binding ATP and the co-chaperonin GroES to the same ring, releasing polypeptide into a now hydrophilic encapsulated cavity where productive folding occurs in isolation. The fate of polypeptide during binding, encapsulation, and folding in the chamber has been the subject of recent experimental studies and is reviewed and considered here. We conclude that GroEL, in general, behaves passively with respect to its substrate proteins during these steps. While binding appears to be able to rescue non-native polypeptides from kinetic traps, such rescue is most likely exerted at the level of maximizing hydrophobic contact, effecting alteration of the topology of weakly structured states. Encapsulation does not appear to involve 'forced unfolding', and if anything, polypeptide topology is compacted during this step. Finally, chamber-mediated folding appears to resemble folding in solution, except that major kinetic complications of multimolecular association are prevented.

Kinetic analysis of conformational changes of GroEL based on the fluorescence of tyrosine 506

The conformational changes of GroEL during the ATPase cycle in the presence of GroES were studied by measuring the fluorescence intensity time course of intrinsic tyrosine Y506, which is located near the nucleotide-binding site. A GroEL solution containing GroES was mixed with an ATP solution to initiate the reaction cycle. The tyrosine fluorescence intensity relative to that without the nucleotide reached 112% within the dead time of the apparatus (>15 s(-1)) and further increased to 123% at 0.57 s(-1) followed by a decrease to 102% at 0.32 s(-1). An initial conformational change and a second intermediate state were expected to occur in ATP-bound GroEL because similar changes were observed for the ATPase-deficient D398A mutant. The conformational change to the third intermediate state corresponded to a process during or after ATP hydrolysis because D398A had no decreasing phase. The second intermediate state before ATP hydrolysis was characterized for the first time.

Asymmetry of the GroEL-GroES complex under physiological conditions as revealed by small-angle x-ray scattering

Despite the well-known functional importance of GroEL-GroES complex formation during the chaperonin cycle, the stoichiometry of the complex has not been clarified. The complex can occur either as an asymmetric 1:1 GroEL-GroES complex or as a symmetric 1:2 GroEL-GroES complex, although it remains uncertain which type is predominant under physiological conditions. To resolve this question, we studied the structure of the GroEL-GroES complex under physiological conditions by small-angle x-ray scattering, which is a powerful technique to directly observe the structure of the protein complex in solution. We evaluated molecular structural parameters, the radius of gyration and the maximum dimension of the complex, from the x-ray scattering patterns under various nucleotide conditions (3 mM ADP, 3 mM ATP gamma S, and 3 mM ATP in 10 mM MgCl(2) and 100 mM KCl) at three different temperatures (10 degrees C, 25 degrees C, and 37 degrees C). We then compared the experimentally observed scattering patterns with those calculated from the known x-ray crystallographic structures of the GroEL-GroES complex. The results clearly demonstrated that the asymmetric complex must be the major species stably present in solution under physiological conditions. On the other hand, in the presence of ATP (3 mM) and beryllium fluoride (10 mM NaF and 300 microM BeCl(2)), we observed the formation of a stable symmetric complex, suggesting the existence of a transiently formed symmetric complex during the chaperonin cycle.

Elucidation of steps in the capture of a protein substrate for efficient encapsulation by GroE

We have identified five structural rearrangements in GroEL induced by the ordered binding of ATP and GroES. The first discernable rearrangement (designated T --> R(1)) is a rapid, cooperative transition that appears not to be functionally communicated to the apical domain. In the second (R(1) --> R(2)) step, a state is formed that binds GroES weakly in a rapid, diffusion-limited process. However, a second optical signal, carried by a protein substrate bound to GroEL, responds neither to formation of the R(2) state nor to the binding of GroES. This result strongly implies that the substrate protein remains bound to the inner walls of the initially formed GroEL.GroES cavity, and is not yet displaced from its sites of interaction with GroEL. In the next rearrangement (R(2).GroES --> R(3).GroES) the strength of interaction between GroEL and GroES is greatly enhanced, and there is a large and coincident loss of fluorescence-signal intensity in the labeled protein substrate, indicating that there is either a displacement from its binding sites on GroEL or at least a significant change of environment. These results are consistent with a mechanism in which the shift in orientation of GroEL apical domains between that seen in the apo-protein and stable GroEL.GroES complexes is highly ordered, and transient conformational intermediates permit the association of GroES before the displacement of bound polypeptide. This ensures efficient encapsulation of the polypeptide within the GroEL central cavity underneath GroES.

Multiple Structural Transitions of the GroEL Subunit Are Sensitive to Intermolecular Interactions with Cochaperonin and Refolding Polypeptide

In this study we attempted to determine the specific roles of the numerous conformational changes that are observed in the bacterial chaperonin GroEL, by performing stopped-flow experiments on GroEL R231W in the presence of a refolding substrate protein. The apparent rate of one kinetic phase was decreased by approximately 25% in the presence of prebound unfolded malate dehydrogenase while another phase was suppressed completely under the same conditions, reflecting different effects of the unfolded protein on multiple structural transitions within GroEL. The addition of cochaperonin GroES counteracts the effect of the bound substrate protein in the former case, but had no effect on the latter, more extensive suppression. Using a chemically modified form of GroEL R231W which is incapable of releasing substrate proteins at low temperatures, we identified a conformational transition that is implicated in the release of substrate proteins. Parts of the actual process of substrate protein release were also observed through fluorescence resonance energy transfer experiments involving GroEL and labeled substrate protein. Analysis of the energy transfer data revealed an interesting relationship between substrate protein displacement and a specific structural transition in the GroEL apical domain.

Ionic interactions at both inter-ring contact sites of GroEL are involved in transmission of the allosteric signal: a time-resolved infrared difference study

The biological activity of the double-ring chaperonin GroEL is regulated by complex allosteric interactions, which include positive intra-ring and negative inter-ring cooperativity. To further characterize inter-ring communication, the nucleotide-induced absorbance changes in the vibrational spectrum of the chaperonin GroEL, of two single-point mutants suppressing one inter-ring ionic contact (E461K and E434K) and of a single-ring version of this protein, were investigated by time-resolved infrared difference spectroscopy. Interaction of the nucleotide with the proteins was triggered by its photochemical release from a biologically inactive caged precursor [P3-1-(2-nitro) phenylethyl nucleotide]. The results indicate that (1) ATP binding to the protein induces a conformational change that affects concomitantly both intra-ring and inter-ring communication, and (2) the experimental absorbance changes are sensitive to the double-ring structure of the protein. The characterization of the single-point, inter-ring mutants demonstrates that ionic interactions at both contact sites are involved in the transmission of the allosteric signal. However, both mutations have different effects on the inter-ring interface. While that of E461K still retains ionic contacts sensitive to ATP binding, E434K shows spectroscopic features similar to those of the single-ring version of the protein, therefore suggesting that electrostatic interactions at these contact sites contribute differently to the stability of the inter-ring interface.

Protein structural transitions and their functional role

Living cells are a collection of molecular machines which carry out many of the functions essential for the cell's existence, differentiation and reproduction. Most, though not all, of these machines are made up of proteins. Because of their complexity, an understanding of how they work requires a synergistic combination of experimental and theoretical studies. In this paper we outline our studies of two such protein machines. One is GroEL, the chaperone from Escherichia coli, which aids in protein folding; the other is F(1)-ATPase, a motor protein which synthesizes and hydrolyses ATP.

The unfolding action of GroEL on a protein substrate

A molecular dynamics simulation of the active unfolding of denatured rhodanese by the chaperone GroEL is presented. The compact denatured protein is bound initially to the cis cavity and forms stable contacts with several of the subunits. As the cis ring apical domains of GroEL undergo the transition from the closed to the more open (ATP-bound) state, they exert a force on rhodanese that leads to the increased unfolding of certain loops. The contacts between GroEL and rhodanese are analyzed and their variation during the GroEL transition is shown. The major contacts, which give rise to the stretching force, are found to be similar to those observed in crystal structures of peptides bound to the apical domains. The results of the simulation show that multidomain interactions play an essential role, in accord with experiments. Implications of the results for mutation experiments and for the action of GroEL are discussed.

Phi value analysis of an allosteric transition of GroEL based on a single-pathway model

There are currently two contradictory models for the kinetics of the ATP-induced GroEL allosteric transition occurring around 20 microM ATP. One model, proposed by Horovitz et al. demonstrates the existence of two parallel pathways for the allosteric transition and an abrupt ATP-dependent switch from one pathway to the other. The other model, which was proposed by the present authors, shows no need to assume the parallel pathways, and a combination of the transition-state theory and the Monod-Wyman-Changeux model of allostery can explain the kinetics as well as the equilibrium of the transition. The discrepancy appears to be due to whether we regard the transition as reversible or irreversible. Thus, here we have investigated the reversibility of the allosteric transition between 0 microM and 70 microM ATP by the use of a stopped-flow double-jump technique, which has allowed us to monitor the kinetics of the reverse reaction from the relaxed state at a high ATP concentration to the tense state at a low ATP concentration. The tryptophan fluorescence of a tryptophan-inserted variant of GroEL was used to follow the kinetics. As a result, the allosteric transition was shown to be a reversible process, supporting the validity of our model. We also show that the structural environment around the ATP-binding site of GroEL in the transition state is very similar to that in the relaxed state (Phi=0.9) by using a Phi value analysis in the kinetic Monod-Wyman-Changeux model, which is analogous to the mutational Phi value analysis in protein folding.

Kinetic analysis of ATP-dependent inter-ring communication in GroEL

Different concentrations of ATP were mixed rapidly with single-ring or double-ring forms of GroEL containing the Phe44-->Trp mutation and the time-resolved changes in fluorescence emission, upon excitation at 295 nm, were followed. Two kinetic phases that were previously found for double-ring GroEL are also observed in the case of the single-ring version: (i) a fast phase with a relatively large amplitude that is associated with the T-->R allosteric transition; (ii) and a slow phase with a smaller amplitude that is associated with ATP hydrolysis. In the case of weak intra-ring positive cooperativity, the rate constant corresponding to the T-->R allosteric switch of single-ring GroEL displays a bi-sigmoidal dependence on ATP concentration that may reflect parallel pathways of the allosteric transition. The slow phase is absent when double-ring or single-ring forms of GroEL are mixed with ADP or ATP without K(+), and it has a rate constant that is independent of ATP concentration. A third fast phase that is still unassigned is observed for both single-ring and double-ring GroEL when a large amount of data is collected. Finally, a fourth phase is observed in the case of double-ring GroEL that is found to be absent in the case of single-ring GroEL. This phase is here assigned to inter-ring communication by (i) determining its dependence on ATP concentration and (ii) based on its absence from single-ring GroEL and the Arg13-->Gly, Ala126-->Val GroEL mutant, which is defective in inter-ring negative cooperativity. The value of the rate constant corresponding to this phase is found to increase with increasing intra-ring positive cooperativity, with respect to ATP. This is the first report of the rate of ATP-mediated inter-ring communication in GroEL, in the presence of ATP alone, which is crucial for the cycling of this molecular machine between different functional states.

Stopped-flow Fluorescence Analysis of the Conformational Changes in the GroEL Apical Domain

GroEL undergoes numerous conformational alterations in the course of facilitating the folding of various proteins, and the specific movements of the GroEL apical domain are of particular importance in the molecular mechanism. In order to monitor in detail the numerous movements of the GroEL apical domain, we have constructed a mutant chaperonin (GroEL R231W) with wild type-like function and a fluorescent probe introduced into the apical domain. By monitoring the tryptophan fluorescence changes of GroEL R231W upon ATP addition in the presence and absence of the co-chaperonin GroES, we detected a total of four distinct kinetic phases that corresponded to conformational changes of the apical domain and GroES binding. By introducing this mutation into a single ring variant of GroEL (GroEL SR-1), we determined the extent of inter-ring cooperation that was involved in apical domain movements. Surprisingly, we found that the apical domain movements of GroEL were affected only slightly by the change in quaternary structure. Our experiments provide a number of novel insights regarding the dynamic movements of this protein.

The allosteric transition of GroEL induced by metal fluoride-ADP complexes

To understand the mechanism of a functionally important ATP-induced allosteric transition of GroEL, we have studied the effect of a series of metal fluoride-ADP complexes and vanadate-ADP on GroEL by kinetic fluorescence measurement of pyrene-labeled GroEL and by small-angle X-ray scattering measurement of wild-type GroEL. The metal fluorides and vanadate, complexed with ADP, are known to mimic the gamma-phosphate group of ATP, but they differ in geometry and size; it is expected that these compounds will be useful for investigating the strikingly high specificity of GroEL for ATP that enables the induction of the allosteric transition. The kinetic fluorescence measurement revealed that aluminium, beryllium, and gallium ions, when complexed with the fluoride ion and ADP, induced a biphasic fluorescence change of pyrenyl GroEL, while scandium and vanadate ions did not induce any kinetically observed change in fluorescence. The burst phase and the first phase of the fluorescence kinetics were reversible, while the second phase and subsequent changes were irreversible. The dependence of the burst-phase and the first-phase fluorescence changes on the ADP concentration indicated that the burst phase represents non-cooperative nucleotide binding to GroEL, and that the first phase represents the allosteric transition of GroEL. Both the amplitude and the rate constant of the first phase of the fluorescence kinetics were well understood in terms of a kinetic allosteric model, which is a combination of transition state theory and the Monod-Wyman-Changeux allosteric model. From the kinetic allosteric model analysis, the relative free energy of the transition state in the metal fluoride-ADP-induced allosteric transition of GroEL was found to be larger than the corresponding free energy of the ATP-induced allosteric transition by more than 5.5kcal/mol. However, the X-ray scattering measurements indicated that the allosteric state induced by these metal fluoride-ADP complexes is structurally equivalent to the allosteric state induced by ATP. These results suggested that both the size and coordination geometry of gamma-phosphate (and its analogs) are related to the allosteric transition of GroEL. It was therefore concluded that the tetrahedral geometry of gamma-phosphate (or its analogs) and the inter-atomic distance ( approximately 1.6A) between phosphorus (vanadium, or metal atom) and oxygen (or fluorine) are both important for inducing the allosteric transition of GroEL, leading to the high selectivity of GroEL for ATP about ligand adenine nucleotides, which function as the preferred allosteric ligand.

Denaturation and reassembly of chaperonin GroEL studied by solution X-ray scattering

We measured the denaturation and reassembly of Escherichia coli chaperonin GroEL using small-angle solution X-ray scattering, which is a powerful technique for studying the overall structure and assembly of a protein in solution. The results of the urea-induced unfolding transition show that GroEL partially dissociates in the presence of more than 2 M urea, cooperatively unfolds at around 3 M urea, and is in a monomeric random coil-like unfolded structure at more than 3.2 M urea. Attempted refolding of the unfolded GroEL monomer by a simple dilution procedure is not successful, leading to formation of aggregates. However, the presence of ammonium sulfate and MgADP allows the fully unfolded GroEL to refold into a structure with the same hydrodynamic dimension, within experimental error, as that of the native GroEL. Moreover, the X-ray scattering profiles of the GroEL thus refolded and the native GroEL are coincident with each other, showing that the refolded GroEL has the same structure and the molecular mass as the native GroEL. These results demonstrate that the fully unfolded GroEL monomer can refold and reassemble into the native tetradecameric structure in the presence of ammonium sulfate and MgADP without ATP hydrolysis and preexisting chaperones. Therefore, GroEL can, in principle, fold and assemble into the native structure according to the intrinsic characteristic of its polypeptide chain, although preexisting GroEL would be important when the GroEL folding takes place under in vivo conditions, in order to avoid misfolding and aggregation.