The metastable states of proteins

Abstract

The intriguing process of protein folding comprises discrete steps that stabilize the protein molecules in different conformations. The metastable state of protein is represented by specific conformational characteristics, which place the protein in a local free energy minimum state of the energy landscape. The native‐to‐metastable structural transitions are governed by transient or long‐lived thermodynamic and kinetic fluctuations of the intrinsic interactions of the protein molecules. Depiction of the structural and functional properties of metastable proteins is not only required to understand the complexity of folding patterns but also to comprehend the mechanisms of anomalous aggregation of different proteins. In this article, we review the properties of metastable proteins in context of their stability and capability of undergoing atypical aggregation in physiological conditions.

1. INTRODUCTION

The phenomenon of protein folding is a cooperative process that involves spontaneous intramolecular interactions. The composition and sequence of amino acids of protein determine the primary quality of its structure in terms of acquiring and retaining the favorable native form. The sequence‐driven primary folding manages the conformational flexibility and dynamics of cellular proteins. This process is also assisted by the cellular machineries of folding and assembly pathways. The stages of protein folding are not limited only in the early steps of protein biogenesis, but they also mature in downstream events such as chaperone‐assisted folding and posttranslational modifications (PTMs).

In recent times, the phenomenon of protein folding is no more described as the mechanism of how an individual protein molecule folds depending on its primary, secondary, and tertiary interactions. Instead, it is proposed that the entire population of protein molecules is a conglomerate of several ensemble classes, which undergo folding in similar or different pathways. As a result of that, different molecules of a protein can exist in different folded conformers. Although the formation of differentially folded conformations relies on the free energy and entropic variations of the molecules, their lifetime in a folded intermediate form is fixed by kinetic parameters. ¹ The transition of one folded conformer to other is driven by energy jumps. The energy, usually in the form of heat, is supplied by the protein's surrounding environments.

Attainment of the most stable (global energy minimum) native structure of protein in physiological condition occurs through one or more folding pathways. Proper folding of protein is represented by a series of stochastic fluctuation events of the region‐specific or overall structure. There are principally two schools of thought that describe the folding of proteins into their native structures. The first theory embraces the concept that protein rapidly acquires a primitive native‐like conformation from its unfolded state. ² The stabilization events are gradually incorporated into the primitive native‐like structure. ² The “domain folding” concept has emerged as the second acceptable notion of protein folding. ³ It is observed that individual domains of proteins fold independently from the rest of the structure. The segmental folding precedes the formation of the final structure. In either of the situations, the hierarchy of folding process depends on various intra‐ and intermolecular contacts, dynamics of chemical groups of amino acid side chains, and the changing environmental conditions. The elemental forces of short‐ and long‐range interactions, which are hydrogen bonds, hydrophobic, polar, and van der Waal interactions, of the side chains induce bends and tilts in the structure that finally make a functional folding event successful. ⁴

2. METASTABILITY OF PROTEINS

Metastable state of protein is a kinetically trapped structure that has a local free energy minimum, and it is separated from the global free energy minimum conformation by an energy barrier (Figure 1). The transient, yet finite, lifetime of metastable state depends on the energy barrier that separates its energy from the energy minima of other conformations. However, it would be incorrect to identify any intermittent transient state of protein as metastable state. A random conformational fluctuation does not qualify any intermittent state as a true metastable state. The theoretical state of metastability is identified by obtaining a probability distribution of the free energy surface from a fully equilibrated conformational topography of protein and pointing the state that satisfies the following conditions: (a) a state that has a stability difference compared with the native state; (b) energy barrier height that separates this state from the other nearby states and the most stable state. However, a metastable state can sometimes be considered as one of the misfolded states or a partially folded state depending on whether it satisfies the abovementioned conditions. The cause of metastability can be the intrinsic features of protein such as heterogeneity of complexity regions, mutation, and folding anomaly and/or the environmental variations such as the change of pH, ion concentrations, temperature, and pressure.

FIGURE 1.

Energy landscape of different folded structures of protein. The process of folding of an unfolded protein to its native structure occurs through different stages involving formation of intermediate conformations, such as metastable state, which are partially or nearly complete folded with different energy states

Since metastable entities are usually short‐lived molecules, it is rather difficult to identify them through conventional experimental techniques. In many of the studies, identification of protein's metastable state has been done by computational studies, such as replica‐exchange molecular dynamics simulation, ⁵ reactive molecular dynamics simulation, ⁶ discontinuous molecular dynamics simulation, ⁷ coarse‐grained molecular dynamics simulation, ⁸ simulated force quench dynamics, ⁹ Brownian dynamics simulation, ⁹ Monte Carlo simulation. ¹⁰ The experimental methods that are invaluable in the study of metastable proteins are circular dichroism spectroscopy, ¹¹ fluorescence spectroscopy, ¹² multivariate curve resolution alternating least square chemometric analysis, ¹³ hydrogen/deuterium exchange nuclear magnetic resonance spectroscopy, ¹⁴ oxidative labeling/mass spectrometry, ¹⁵ complementary structural mass spectrometry, ¹⁶ single molecule fluorescence resonance energy transfer spectroscopy, ¹⁷ and single molecule force spectroscopy. ¹⁸

3. STRUCTURAL PROPERTIES OF METASTABLE PROTEINS

The principle of formation of metastability in proteins is embedded in their nonnative structures. Although some metastable states are structurally close to the native states of proteins, most of the identified metastable structures are different from the native state structures (Figure 2). There are two main mechanisms of native‐to‐metastable structural transitions of proteins. Firstly, induction of metastable states in many of the proteins is linked to the conversion of secondary structures to disordered regions or vice versa. The number of such transiently formed or lost disordered regions and their positioning in the proteins determine the degree of metastability. For example, the intrinsically unstructured kinase inhibitor domain of Sic1 protein forms a collapsed helical structure that converts the protein into a metastable form. ¹⁹ In the other mechanism, metastable states are prompted in proteins by the conversion of one form of secondary structure to another form. Transformation of alpha helices to beta sheets is observed in the native‐to‐metastable transitions of several proteins. It is understood that the unstructured region is formed as an intermediate during the switching of alpha‐helical patches to beta sheets. The time‐dependent repositioning of specific segments can also induce the characteristics of metastability in proteins. The temporal fluctuation of the T54R mutation‐harboring disordered loop of superoxide dismutase 1 (SOD1^T54R) invokes metastability in the protein, leading to destabilization of its dimers. ²⁰ Despite many conformational alterations, metastable proteins show compact structures. Even though many of the intramolecular interactions are lost during the transition of native‐to‐metastable state, several nonnative ionic and hydrogen bonds, hydrophobic, and van der Waal interactions stabilize the metastable proteins. The newly formed interactions are either intramolecular or they bridge the side chains of protein and its environment. These interactions are metastable contacts. Although the short‐range metastable connections exist within the secondary structured regions, the long‐range metastable tethering involves the disordered regions. As a proof of this notion, a study shows that the isoleucine‐33 of N‐terminal disordered transactivation domain of estrogen receptor shows metastable interactions with the serine‐118 residues. ²¹

FIGURE 2.

Different stages of protein folding pathway. Differential folding events of protein lead to native or aberrant conformations that take different fate to form various final structures

Globular metastable proteins can gain specific structural forms, such as molten globule (MG), which are compact and thermodynamically stable. Metastable molten globules have near‐native secondary structural characteristics, but they lose the tertiary contacts. ²² Studies report that the huntingtin interacting protein K (HYPK) can transiently achieve an MG state, which is indistinguishable from its metastable structure11, 23, 24 (Figure 3). Loss of the intramolecular interactions between the N‐terminal disordered nanostructure and the C‐terminal charge‐rich low complexity region induces MG state in the HYPK protein. ²³ In the MG state, ubiquitin‐associated (UBA) domain of HYPK is partially unfolded to resemble what appears like a metastable state of HYPK. ¹¹ pH‐dependent metastable molten globule forms of immunoglobulin G (IgG) ²⁵ and papain ²⁶ are also reported.

FIGURE 3.

The metastable states of the monomers and oligomeric seeds of HYPK. Loss of tertiary contacts between the N‐terminal disordered nanostructure and C‐terminal low complexity region and loss of secondary structure of the UBA domain result in the formation of metastable/molten globule monomer of HYPK. The metastable oligomeric seed of HYPK shows less compact structure than the protein's native seed structure

Several lines of evidence suggest that the metastable regions of proteins are usually not conserved. These regions contain positionally suboptimal polar or hydrophobic residues that create packaging defects under certain conditions. Metastability of mutant SOD1 proteins due to the presence of specific hydrophobic residues at the mutation sites can result in the destabilization of the protein. The presence of an alanine instead of glycine at the 93rd position of SOD1^G93A results in induction of metastability in the protein due to local unfolding of an edge strand to disordered structure. ²⁷ Many of the G‐protein‐coupled receptors (GPCRs), including the β1‐adrenergic receptor, exist in different metastable forms due to the destabilizing effects of polar amino acids. ²⁸ The alteration of side chain length and chemical properties at the 129th position of M129V mutant of human prion protein allows it to acquire a metastable state. ²⁹

It is not essential that metastable states only form during the folding of unfolded proteins. Native protein can also undergo structural changes that result in the formation of metastable structure. The reorientation of the native structure to metastable state can be dependent on various physical processes, such as cleavage of the protein, ligand binding to cavity, and binding to partner proteins. The best studied case of conformational change‐coupled formation of metastable state is observed in the serine protease inhibitor (serpin) family of proteins. Representative serpin proteins, such as α1‐antitrypsin, antithrombin‐III, antichymotrypsin, plasminogen activator inhibitor 1 (PAI‐1), and C1 inhibitor, are functionally active only in a metastable state that is manifested by exposed reactive center loop (RCL) of the proteins. ³⁰ The primary folding of serpin proteins spontaneously forms long‐lived metastable (active) form. Mechanistically, the differential rate of folding of different beta strands of serpins results in the formation of the metastable states. Although the B and C beta strands of the protein undergo rapid folding and the RCL remains incorporated in the B–C barrel, the beta strand A folds in a deferred timescale. ³¹ This event keeps the RCL in solvent‐exposed state for a longer period, thereby stabilizing the metastable form of serpins. Moreover, the F helix is also very dynamic in the metastable state. ¹⁵ Other specific events, such as early hydrophobic collapse of short beta sheet regions of α1‐antitrypsin, ³² and differential order of formation of intramolecular disulfide bonds in antithrombin‐III, ³³ also facilitate the formation of metastable states in different serpin proteins. An extended N‐terminal region shields a hydrophobic patch in the tengpin (a serpin of Thermoanaerobacter tengcongensis), resulting in stabilization of the metastable state of the protein. ³⁴ Some critical residues of serpin proteins have functional implications in steadying the metastable states. The lysine‐335 of α1‐antitrypsin and glutamic acid‐381 of antithrombin‐III have stabilizing effects on metastable structures of the corresponding proteins.35, 36 However, certain residues can also act as negative modulators of serpin metastability. The glycine‐117 of α1‐antitrypsin is such a residue. ³⁷ Mutation of this residue with phenylalanine has a cavity‐filling effect at the interface of F helix and beta sheet A, thus destabilizing the metastable structure of α1‐antitrypsin. The intramolecular disulfide bonds involving the extended N‐terminal region and serpin domain of the C1‐inhibitor boost the stabilization of the native state protein. ³⁸ Upon binding to the target proteases, the metastable serpins are converted to the inactive (latent) state by a series of cleavage and folding events. Cleavage of the scissile P1–P1′ bond and the subsequent insertion of the RCL interior of the beta sheets 3A and 5A generate the relaxed conformation of the serpin proteins. The rigidity of the beta sheets 3A, 5A, and 1C ensures that the RCL remains buried in the latent serpins. ³¹ A study reports that serpin proteins invariably form a stressed (intermediate) structure during its transition from metastable to latent form. ³⁹ This intermediate is a molten globule that is formed by the noncooperative changes in the structure of the serpins. ³⁹ Moreover, certain serpin proteins, like antichymotrypsin, are not completely unfolded by strong denaturants. ⁴⁰ The residual structures in antichymotrypsin influence the formation of metastable structure of this protein. Although serpins are glycosylated at different residues, this PTM does not affect the generation of metastability in these proteins. ⁴¹ The metastable structure of an engineered serpin protein (conserpin) has been reported to drive its oligomerization. ⁴² The oligomerization, but not the transition to latent state, is the cause of inactivation of conserpin protein.

Another classic example of metastability is observed in the α‐lytic protease (α‐LP) of Lysobacter enzymogenes bacterium. The metastable state‐coupled autoactivation of α‐LP is accomplished by the cleavage of its N‐terminal 32 amino acid pro‐region. ⁴³ The cleaved α‐LP develops a molten globule‐like intermediate form, which is also recognized as the catalytically active metastable state of the protein. The pro‐region makes high‐affinity interactions with the native protease, thereby accelerating the rigid and more stable folding of the protein. ⁴⁴ The high glycine content of the protein is also believed to contribute in providing stability to the full‐length protein.

An active metastable state is found in the pyruvate dehydrogenase kinase isoform‐4 (PDK4) protein. ⁴⁵ Instead of acquiring a closed structure like the other isoforms of PDK, the PDK4 dimer remains in an open conformation that has a wider ligand binding cleft and less activity of the core region. Thus, the metastable PDK4 has spatiotemporally reduced the binding ability to ADP. In the metastable state, the C‐terminal region of each monomer of PDK4 slips into and interacts with the N‐terminal region of the other monomer.

The dynamic nature of alpha helices and beta sheets in the metastable state of αxβ2 integrin is also reported. ⁴⁶ The crystal structure shows unwinding of the alpha‐helical regions and movement of beta sheets that allow the acquisition of a specific ligand‐binding compatible metastable state of the complex.

Binding of small molecules can modulate the formation of metastable proteins. Nitric oxide (NO) and carbon monoxide (CO) binding to myoglobin results in a very short‐lived metastable myoglobin–NO or myoglobin–CO complexes.47, 48 These metastable complexes are defined by a structure in which NO or CO is very close (3 Å) to the iron (Fe) atom of heme group. Noncovalent binding of zinc ion (Zn⁺²) to amyloid beta peptide (Aβ) potentiates the formation of “quasi‐spherical” metastable complex of Zn⁺²‐Aβ40. ⁴⁹ In contrast, binding of ligand can reduce the metastability of the proteins. For example, nucleotide binding can reduce the quality of protein metastability. The empty pocket of the α‐subunit of GPCR achieves a metastable intermediate state due to the influence of associated receptor protein. ⁵⁰ The metastable state of the α‐subunit determines the gating of guanosine triphosphate (GTP) binding to the GPCR. GTP binding to the pocket leads to a drastic decrease of the metastability and increment in the activity of the pocket region.

Thermal stress can stimulate the induction of metastability in various cellular proteins, including the toxic aggregation‐prone proteins. The metastable species that are generated during the folding transition of prion protein PrpC to PrPSC are heat inducible. ⁵¹ Heat‐induced metastable forms of nonprion proteins can divert to an off‐pathway to become a prion‐like protein. For example, an increase in temperature causes native Las17‐binding protein 2 (Lsb2, an actin‐associated protein) to become metastable, which acquires prion‐like properties. ⁵² Heat‐induced metastable oligomers of β‐lactoglobulin are also observed. ⁵³ Like the heat, pH can also drive the formation of metastability in certain proteins. A specific helical region of the membrane‐associated lipoprotein surfactant protein C (SP‐C) is transformed to beta sheet in its metastable form. ⁵⁴ The formation of metastable SP‐C is (higher) pH dependent, and it is mostly accelerated in its nonacylated (nonpalmitoylated) form.

4. FUNCTIONS OF METASTABLE PROTEINS IN CELLULAR PROCESSES

The formation of metastable proteins in terms of their functional perspectives is a much debated topic. Although the function of metastable serpins is widely established, the functions of other metastable proteins are yet to be identified. A hypothesis argues that all conformations of proteins represent to one or the other kind of metastable form of the respective proteins. ⁵⁵ In that definition, the functionally active native protein is also regarded as a long‐lived metastable form. However, the conventional explanation of metastable condition classifies a state of the protein that has nonnative structural and free energy parameters. Due to the experimental limitations of monitoring such metastable states of proteins under in vivo conditions, the functional implication of intricate metastable proteins has remained vastly elusive. In spite of this, some studies have shown that the native‐like metastable states are required for proper functioning of many cellular proteins in signaling, catalysis, binding, and scaffolding.

Different proteins of cell signaling pathways assume different conformations depending on their PTMs and binding to other proteins. Conversion of these signal transducers from the inactive to active state does not follow a simple two‐stage process but requires the formation of gradual intermediate or metastable states. In the study of a bacterial two‐component signal‐relaying protein, response regulator 468 (RR468), it was observed that the protein could remain in various metastable forms that show alteration in conformation of specific loops and a helix. ⁵⁶ The upstream activators of signal transduction pathways, such as the cell surface receptor proteins, can also be metastable. The dimer of CD2 protein has a distinct metastable folded state, which is marked by the reciprocal intercalation of the IgSF domain of each monomer to the other one. ⁵⁷ The N‐terminal region of the CD2 protein can exist in an ensemble class of metastable states, ⁵⁸ and the differential fold transitions of the metastable N‐terminal region are independent of folding of the protein's other domains. The receptor of relaxin‐3 peptide, RXFP3, is also able to remain in metastable conformations. ⁵⁹ A single aspartate residue (aspartate‐128) in the RXFP3 determines the metastability of the protein, which, in turn, regulates the conformational fluctuation of the protein during its ligand or agonist binding. Another example of metastability is found in the transmembrane (TM) helix of the cell surface receptor ephrin type‐A receptor 1 (EPHA1). Although the stable dimerization of the TM region of EPHA1 is dependent on the glycine zipper motif of the protein, the metastable state‐driven strained dimerization is independent of the glycine zipper. ⁶⁰

The coiled‐coil domain (CCD) of beclin‐1/2 protein preferentially exists in a native metastable state, ⁶¹ which helps in the formation of an autophagy modulatory scaffolding complex. The metastable CCD destabilizes the beclin‐1/2 homodimer and facilitates the heterodimerization of beclin‐1/2 with ATG14 ⁶² or UVRAG. ⁶¹ The CCD metastability causes unfavorable packing of the two monomers of beclin‐1/2. However, the same metastable CCD of beclin‐1/2 is tightly packaged with the CCD of ATG14 in a parallel orientation, which is held together by multiple polar interactions. The metastable state of topoisomerase‐I is also linked to its scaffolding function during the cleavage and re‐ligation of superhelical dsDNA molecules. ⁶³ Thus, the metastable topoisomerase‐I relieves the torsional stress of complex DNA structures.

Many of the viral envelop proteins tend to remain in metastable state during the endosomal prefusion stage. After binding to the cell surface receptors, the viral entry proteins undergo robust structural transition from stable to metastable state in the low pH of endosomal vesicles. In this context, the best studied example is the metastability of hemagglutinin (HA) protein of influenza virus. HA forms homotrimeric complex in which each monomer is cleaved at a specific site but held together by disulfide linkage. In the acidic pH of endosome, native HA undergoes a transition to a metastable structure that results in displacement of the head structures of each monomer distant from its initial position (resembling an open conformation) (Figure 4). In this conformation, the previously buried fusion peptide of each monomer pops out and gets inserted into the endosomal membrane, thereby mediating the collapse and hemifusion of the viral membrane to endosomal membrane, and subsequent release of viral contents into the host cell cytoplasm. Thus, metastable forms of the viral proteins, like HA of influenza virus, ⁶⁴ envelope glycoprotein of avian leukosis virus A, ⁶⁵ are indispensable for the delivery of viral genome to the receptor cells.

FIGURE 4.

Native and metastable states of trimeric hemagglutinin (HA) complex. Acidic environment of endosome causes transition of native trimeric HA complex to a metastable complex in which the monomers show drastic changes of structural conformational of the three regions of the protein, namely HA1 segment, HA2 segment, and fusion peptide

5. QUALITY CONTROL OF METASTABLE PROTEIN

In the structural landscape, metastable entities are generally put in the category of partially unfolded or misfolded proteins. ⁶⁶ Although less frequent, yet some metastable proteins also attain near‐native structures. ⁶⁷ Due to the folding complexity, the metastable proteins are subjected to various cellular quality control mechanisms and machineries that include but not limited to chaperones, disaggregases, and degradation systems. For example, the misfolded metastable monomer of transthyretin (TTR) protein, which is formed by the proteolytic cleavage of its C‐terminal beta strand and the reorientation of its AB loop, is selectively captured by Hsp90. ⁶⁸ Binding of Hsp90 to metastable TTR prevents the amyloidogenic aggregation of TTR. However, the native form of TTR is not recognized by Hsp90. In yeast, the Hsp70 (Ssa) chaperone and its cochaperones (Sis1 and Ydj1) regulate the aggregation properties of the prion protein Swi1 by modulating the stability of metastable Swi1. ⁶⁹ Association of an Hsp110 family protein, Sse1/2, to Swi1 is also reported to regulate the metastability of Swi1. ⁶⁹ It is interesting to note that some chaperone–substrate interactions are also guided by the induction of metastability in the chaperone proteins. Conditionally disordered chaperones maintain metastable states that are derived through unfolding of the structured regions. The newly formed disordered regions either act as linker regions or bind to substrate proteins to pursue chaperone activity. ⁷⁰ A good example of metastability is found in Hsp33 chaperone. The formation of metastable linker region of Hsp33 is dependent on redox stress. A zinc coordination center of the protein responds to oxidative stress, resulting in the activation of metastable properties in the neighboring linker region. ⁷¹ The active conformation of the metastable Hsp33 presents an amphipathic helix, which is used for high‐affinity binding to the client proteins. ⁷²

Sequestrases and disaggregases control the qualitative and quantitative accumulation of aggregation‐prone exon1 of huntingtin protein. The alpha‐helical polyglutamine region of huntingtin exon1 undergoes a structural transition to beta sheet through formation of an unstructured intermediate. ²³ The helix‐to‐disorder transition accounts for the putative metastable state of huntingtin exon1. Sequestrase, such as HYPK, binds to the N‐terminal 17 amino acid region of huntingtin exon1, ⁷³ thereby preventing the formation of metastable‐like structures of huntingtin‐exon1. ²⁴ VCP/p97 functions as an ATP‐dependent disaggregate to untangle the aggregates of huntingtin exon1, ⁷⁴ and it possibly modulate the stability of monomeric huntingtin exon1.

6. AGGREGATION OF METASTABLE PROTEINS

Even though cells employ robust quality control mechanisms to restrict the aberrant interaction of metastable proteins, it is found that some metastable proteins dodge the quality control processes to undergo uncontrolled aggregation. For instance, the α‐synuclein monomer secures a metastable state from its disordered structure, and the metastable form is typically responsible for its fibrillar aggregation. ⁷⁵ α‐Synuclein can also exist in metastable tetramer and higher oligomers, in which the monomers acquire alpha‐helical structures. ⁷⁶ However, the metastable oligomers of α‐synuclein are resistant to aggregation. Another intrinsically unstructured peptide, hIAPP, generates short beta hairpins in its structure, and such transiently formed structures are recognized as the metastable forms of hIAPP. ⁷⁷ Metastable hIAPP peptides nucleate to form amyloid fibrils. The metastable monomers of IgG form aggregates, ²⁵ which can also revert to its native structure. The metastable states are inducible in the monomers and oligomeric seeds of the UBA domain of HYPK protein. ¹¹ The metastable state of HYPK‐UBA monomer is characterized by a helical‐to‐disorder transition. On the other hand, the metastable oligomeric seeds of HYPK‐UBA do not show loss of secondary structures when compared to its native state. The metastable trimeric seed of HYPK‐UBA shows loss of compactness in its structure, leading to the generation of an open conformation from the closed structure (Figure 3). Although the metastable monomers of HYPK are responsible for initiating the formation of smaller oligomeric seeds, the metastable seeds of HYPK propagate to ordered annular aggregates of the protein. Metastability, when introduced in the aggregate‐residing seeds of HYPK by application of laser light, can also dissolute the aggregates of HYPK in human neuroblastoma cells. Two metastable forms of rhodanese enzyme show similar properties like HYPK in terms of taking different fates. ⁷⁸ One metastable state of rhodanese reverts to the native state, and the other one slowly aggregates. A similar observation is also noted in sickle cell hemoglobin, which shows “metastable mesoscopic cluster” in its aggregates. ⁷⁹

Metastability of the Aβ exacerbates its aggregation in different ways. The shapes of metastable conformers of Aβ dictate its aggregation. Metastable monomers and oligomers of Aβ17‐41 peptide are reported to form “U‐shaped” and “S‐shaped” structures, respectively. ⁸⁰ These structures originate from the disordered monomers of Aβ17‐41 peptide in two different pathways, and they are responsible for fibrillar aggregation of the peptide. In the immature fibrillar aggregates, the metastable full‐length Aβ42 peptides show antiparallel beta sheets and beta hairpin structures. ⁸¹ Hydrophobic interactions of the isoleucine‐41 and alanine‐42 of Aβ42 stabilize the beta hairpins. ⁸¹ Exclusion of these two residues (as in Aβ40) results in change of the beta hairpin to turn‐coil structure, which is not compatible to generate the metastability in Aβ40. The 21–30 amino acid region of Aβ42 peptide is also vulnerable with respect to the formation of the metastable beta‐hairpin structures. ⁸² The 16–22 amino acid fragment of Aβ42 (Aβ16‐22) remains in a metastable particle phase in the organic solvent/water mixture system. ⁸³ The N‐terminal pyroglutamylated form of the truncated amyloid beta peptide (Aβ3‐42) associates with the naturally existing Aβ42 peptides to generate lower molecule number (n < 11) oligomers. ⁸⁴ These oligomers are metastable, and they form heterogeneous aggregates with Tau.

Mutation‐linked metastability can turn the mutant proteins into aggregation‐prone structures. Most often, such aggregation is a result of anomalous folding of the metastable protein. The best studied example of metastable protein aggregation is reported for Z mutant (E342Q mutation) of α1‐antitrypsin. The mutation renders α1‐antitrypsin in a nonnative (resembling metastable) structure by destroying a critical salt bridge and resulting in opening of the A sheet. The final consequence of Z mutation is the loop–sheet polymerization. ⁸⁵ In this type of chain oligomerization, the RCL of one α1‐antitrypsin molecule gets inserted into the A beta sheet of another molecule. A cysteine‐to‐tyrosine (C7Y) mutation drives the misfolding‐led aggregation of insulin protein. The misfolding arises due to the formation of metastable structures in the N‐terminal region of mutant insulin. ⁸⁶ Partially unfolded metastable monomers can be aggregation‐prone structures. Urea‐induced partially denaturated cytochrome P450‐2B1 forms metastable monomers that are highly aggregation prone. ⁸⁷ However, semi‐unfolded metastable monomers of other proteins, such as the UDP‐N‐acetylglucosamine enolpyruvyl transferase of Acinetobacter baumannii, do not undergo aggregation. ⁸⁸ Moreover, not all the proteins that show monomer–oligomer equilibrium in physiological condition, simultaneously, possess the metastable properties in its monomer and oligomers. For example, only the dimer, but not the monomer, of cyanovirin‐N adopts a metastable state. ⁸⁹

Contrary to the unfolding associated metastability, the disorder‐to‐structured metastable transition can also be important for biomolecular recognition. A loop‐to‐beta sheet metastable transition of malectin allows the protein to selectively bind the diglucosylated N‐glycan. ⁹⁰

Formation of metastable oligomers can also be triggered in extracellular proteins. The oligomers of amelogenin are metastable, and they can self‐associate into extracellular matrix‐like structure. ⁹¹ The N‐terminal region of amelogenin is reported to be responsible for its metastable structure and aggregation.

7. CONCLUDING REMARKS

Evolution of several structures in the folding pathways imparts variability in the three‐dimensional structures of the proteins. The metastable states ensue as the kinetically trapped species in the folding pathways. The increasing reports of metastable state in different proteins are showing that this state is not merely an inert molecule, and its formation may be purposed for definite functions. Such functions of metastable proteins are essentially connected to their appropriate structural conformations. The relevance of metastable states in proteostasis has also become prominent due to the occurrence of aggregation of metastable proteins in different diseases. Hence, computational, structural, biophysical, and clinical studies of metastable proteins are immensely important to increase the understanding of the sequential stages of protein folding in complex environments.

CONFLICT OF INTEREST

The authors declare no potential conflict of interest.

AUTHOR CONTRIBUTIONS

Debasish Kumar Ghosh: Conceptualization; data curation; writing‐original draft; writing‐review and editing. Akash Ranjan: Conceptualization; funding acquisition; supervision; writing‐original draft; writing‐review and editing.

Ghosh DK, Ranjan A. The metastable states of proteins. Protein Science. 2020;29:1559–1568. 10.1002/pro.3859