Search

Hsp70s are molecular chaperones involved in the folding and assembly of proteins. They recognize hydrophobic amino acid stretches in their substrate binding groove. However, a detailed understanding of substrate specificity is still missing. Here, we use the endoplasmic reticulum-resident Hsp70 BiP to identify binding sites in a natural client protein. Two sites are mutually recognized and form stable Hsp70-substrate complexes. In silico and in vitro analyses revealed an extended substrate conformation as a crucial factor for interaction and show an unexpected plasticity of the substrate binding groove. The basic binding mechanism is conserved among different Hsp70s.

The tumor suppressor protein p53 is often referred to as the guardian of the genome. In the past, controversial findings have been presented for the role of the C-terminal regulatory domain (RD) of p53 as both a negative regulator and a positive regulator of p53 activity. However, the underlying mechanism remained enigmatic. To understand the function of the RD and of a dominant phosphorylation site within the RD, we analyzed p53 variants in vivo and in vitro. Our experiments revealed, surprisingly, that the p53 RD of one subunit interacts with the DNA binding domain of an adjacent subunit in the tetramer. This leads to the formation of intersubunit contacts that stabilize the tetrameric state of p53 and enhance its transcriptional activity in a cooperative manner. These effects are further modulated by phosphorylation of a conserved serine within the RD.

Heat shock protein 90 (Hsp90) is an ATP-dependent molecular chaperone responsible for the activation, maturation, and trafficking of several hundred client proteins in the cell. It is well known that (but not understood how) residues far away from Hsp90's nucleotide binding pocket can regulate its ATPase activity, a phenomenon called allosteric regulation. Here, the computational design of allosteric mutations was combined with in vitro and in vivo experiments to unravel nucleotide-responsive hot spots in the regulation of Hsp90. With this approach, we identified both activating and inhibiting regulation points and show that changes in those amino acids affect the conformational dynamics and ATPase activity of Hsp90 in vitro. Our observations that activating mutations loosen and inhibiting mutations rigidify the protein explain for the first time how Hsp90 changes in response to allosteric mutations. Additionally, mutations of these allosteric regulation points can be controlled by the interplay with Hsp90 co-chaperones, thus providing cells with an efficient mechanism of modifying Hsp90's intrinsic properties via different layers of regulation. Altogether, our results show that a framework for transmitting conformational information exists in the Hsp90 structure.

In biological systems, proteins rarely act as isolated monomers. Association to dimers or higher oligomers is a commonly observed phenomenon. As an example, small heat shock proteins form spherical homo-oligomers of mostly 24 subunits, with the dimeric alpha-crystallin domain as the basic structural unit. The structural hierarchy of this complex is key to its function as a molecular chaperone. In this article, we analyze the folding and association of the basic building block, the alpha-crystallin domain dimer, from the hyperthermophilic archaeon Methanocaldococcus jannaschii Hsp16.5 in detail. Equilibrium denaturation experiments reveal that the alpha-crystallin domain dimer is highly stable against chemical denaturation. In these experiments, protein dissociation and unfolding appear to follow an "all-or-none" mechanism with no intermediate monomeric species populated. When the mechanical stability was determined by single-molecule force spectroscopy, we found that the alpha-crystallin domain dimer resists high forces when pulled at its termini. In contrast to bulk denaturation, stable monomeric unfolding intermediates could be directly observed in the mechanical unfolding traces after the alpha-crystallin domain dimer had been dissociated by force. Our results imply that for this hyperthermophilic member of the small heat shock protein family, assembly of the spherical 24mer starts from folded monomers, which readily associate to the dimeric structure required for assembly of the higher oligomer.

Intact antibodies and antigen binding fragments (Fab) have been previously shown to form an alternatively folded state (AFS) at low pH. This state consists primarily of secondary structure interactions, with reduced tertiary structure content. The AFS can be distinguished from the molten globule state by the formation of nonnative structure and, in particular, its high stability. In this study, the isolated domains of the MAK33 (murine monoclonal antibody of the subtype kappa/IgG1) Fab fragment were investigated under conditions that have been reported to induce the AFS. Surprising differences in the ability of individual domains to form the AFS were observed, despite the similarities in their native structures. All Fab domains were able to adopt the AFS, but only for V(H) (variable domain of the heavy chain) could a significant amount of tertiary structure be detected and different conditions were needed to induce the AFS. V(H), the least stable of the domains under physiological conditions, was the most stable in the AFS, yet all domains showed significant stability against thermal and chemical unfolding in their AFS. Formation of the AFS was found to generally proceed via the unfolded state, with similar rates for most of the domains. Taken together, our data reveal striking differences in the biophysical properties of the AFS of individual antibody domains that reflect the variation possible for domains of highly homologous native structures. Furthermore, they allow individual domain contributions to be dissected from specific oligomer effects in the AFS of the antibody Fab fragment.

The aggregation of mostly antibody light chain variable (VL) domains into amyloid fibrils in various tissues is the main cause of death in systemic amyloid light chain amyloidosis. Point mutations within the domain are important to shift the VL into the fibrillar pathway, but why and how only some site-specific mutations achieve this still remains elusive. We show here that both destabilizing and surprisingly stable mutants readily predispose an amyloid-resistant VL domain to amyloid formation. The decreased thermodynamic stability of the destabilizing mutant results in the accumulation of non-native intermediates that readily populate the amyloid state. Interestingly, the stable mutants establish site-specific non-native interactions with especially nearby serine/threonine residues that unexpectedly do not affect the folding behavior of the VL domain but rather readily induce and stabilize the fibril structure, a previously unrecognized mechanism. These findings provide a new concept for the molecular mechanism of amyloid fibril formation.

The association of light chains (LCs) and heavy chains is the basis for functional antibodies that are essential for adaptive immune responses. However, in some cases, LCs and especially fragments consisting of the LC variable (VL) domain are pathologically deposited in fatal aggregation diseases. The two domains of the LC are connected by a highly conserved linker. We show here that, unexpectedly, the linker residue Arg108 affects the conformational stability and folding of both VLκ and LC constant (CLκ) domains. Interestingly, the extension of VL by Arg108 results in its resistance to amyloid formation, which suggests that the nature of the truncation of the LC plays a crucial role in disease progression. Increased solvation due to the exposed charged C-terminal Arg108 residue explains its stabilizing effects on the VL domain. For the CL domain, the interaction of N-terminal loop residues with Arg108 is important for the integrity of the domain, as the disruption of this interaction results in fluctuation, partial opening of the protein's interior and the exposure of hydrophobic residues that destabilize the domain. This establishes new principles for antibody domain architecture and amyloidogenicity.

Hsp90 is an essential chaperone protein in the cytosol of eukaryotic cells. It cooperates with the chaperone Hsp70 in defined complexes mediated by the adaptor protein Hop (Sti1 in yeast). These Hsp70/Hsp90 chaperone complexes play a major role in the folding and maturation of key regulatory proteins in eukaryotes. Understanding how non-native client proteins are transferred from one chaperone to the other in these complexes is of central importance. Here, we analyzed the molecular mechanism of this reaction using luciferase as a substrate protein. Our experiments define a pathway for luciferase folding in the Hsp70/Hsp90 chaperone system. They demonstrate that Hsp70 is a potent capture device for unfolded protein while Hsp90 is not very efficient in this reaction. When Hsp90 is absent, in contrast to the in vivo situation, Hsp70 together with the two effector proteins Ydj1 and Sti1 exhibits chaperone activity towards luciferase. In the presence of the complete chaperone system, Hsp90 exhibits a specific positive effect only in the presence of Ydj1. If this factor is absent, the transferred luciferase is trapped on Hsp90 in an inactive conformation. Interestingly, identical results were observed for the yeast and the human chaperone systems although the regulatory function of human Hop is completely different from that of yeast Sti1.

The oncogenic potential of the viral tyrosine kinase v-Src is due to its constitutive activity. Unlike the highly homologous cellular c-Src kinase, a C-terminal deletion of the regulatory tail and numerous point mutations make the viral kinase uncontrollable. To determine the basis of these differences, we analysed the structure and stability of v-Src and c-Src in vitro. We show that the stability of v-Src against unfolding and irreversible aggregation is significantly lower than that of c-Src. Furthermore, in v-Src hydrophobic residues are more exposed already in the native state. In consequence, v-Src was found to be inactive close to physiological temperatures. We thus suggest that the ensemble of mutations that transform c-Src into the oncogenic variant cause a concomitant destabilisation of the kinase.

Small heat-shock proteins (Hsps) are ubiquitous molecular chaperones which prevent the unspecific aggregation of non-native proteins. For Hsp26, a cytosolic sHsp from of Saccharomyces cerevisiae, it has been shown that, at elevated temperatures, the 24 subunit complex dissociates into dimers. This dissociation is required for the efficient interaction with non-native proteins. Deletion analysis of the protein showed that the N-terminal half of Hsp26 (amino acid residues 1-95) is required for the assembly of the oligomer. Limited proteolysis in combination with mass spectrometry suggested that this region can be divided in two parts, an N-terminal segment including amino acid residues 1-30 and a second part ranging from residues 31-95. To analyze the structure and function of the N-terminal part of Hsp26 we created a deletion mutant lacking amino acid residues 1-30. We show that the oligomeric state and the structure, as determined by size exclusion chromatography and electron microscopy, corresponds to that of the Hsp26 wild-type protein. Furthermore, this truncated version of Hsp26 is active as a chaperone. However, in contrast to full length Hsp26, the truncated version dissociates at lower temperatures and complexes with non-native proteins are less stable than those found with wild-type Hsp26. Our results suggest that the N-terminal segment of Hsp26 is involved in both, oligomerization and chaperone function and that the second part of the N-terminal region (amino acid residues 31-95) is essential for both functions.

p53 is one of the key molecules regulating cell proliferation, apoptosis and tumor suppression by integrating a wide variety of signals. The structural basis for this function is still poorly understood. p53 appears to exercise its function as a modular protein in which different functions are associated with distinct domains. Presumably, p53 contains both folded and partially structured parts. Here, we have investigated the structure of the isolated N-terminal part of p53 (amino acid residues 1-93) using biophysical techniques. We demonstrate that this domain is devoid of tertiary structure and largely missing secondary structure elements. It exhibits a large hydrodynamic radius, typical for unfolded proteins. These findings suggest strongly that the entire N-terminal part of p53 is natively unfolded under physiological conditions. Furthermore, the binding affinity to its functional antagonist Mdm2 was investigated. A comparison of the binding of human Mdm2 to the N-terminal part of p53 and full-length p53 suggests that unfolded and folded parts of p53 function synergistically.

Antibodies undergo significant conformational changes upon acidification, leading to the formation of an alternatively folded state. Here, we analyzed the conformation of MAK 33 Fab and its light chain at acidic pH, both in the reduced and oxidized form. At acidic pH, the proteins exhibited a highly structured, but non-native conformation, corresponding to the alternatively folded state, previously described for the intact antibody. However, the requirements to form this alternative structure were different for the oxidized and reduced protein. Whereas in the oxidized form of the immunoglobulin light chain the alternatively folded state could only be detected at pH

The C H 3 domain of antibodies is characterized by two antiparallel β-sheets forming a disulfide-linked sandwich-like structure. At acidic pH values and low ionic strength, C H 3 becomes completely unfolded. The addition of salt transforms the acid-unfolded protein into an alternatively folded state exhibiting a characteristic secondary structure. The transition from native to alternatively folded C H 3 is a fast reaction. Interestingly, this reaction involves the formation of a defined oligomer consisting of 12-14 subunits. Association is completely reversible and the native dimer is quantitatively reformed at neutral pH. This alternatively folded protein is remarkably stable against thermal and chemical denaturation and the unfolding transitions are highly cooperative. With a t m of 80°C, the stability of the alternatively folded state is comparable to that of the native state of C H 3. The defined oligomeric structure of C H 3 at pH 2 seems to be a prerequisite for the cooperative unfolding transitions.

Large peptidyl-prolyl cis/trans isomerases (PPIases) are important components of the Hsp90 chaperone complex. In mammalian cells, either Cyp40, FKBP51 or FKBP52 is incorporated into these complexes. It has been suggested that members of this protein family exhibit both prolyl isomerase and chaperone activity. Here we define the structural and functional properties of the three mammalian large PPIases. We find that in all cases two PPIase monomers bind to an Hsp90 dimer. However, the affinities of the PPIases are different with FKBP52 exhibiting the strongest interaction and Cyp40 the weakest. Furthermore, in the mammalian system, in contrast to the yeast system, the catalytic activity of prolyl isomerization corresponds well to that of the respective small PPIases. Interestingly, Cyp40 and FKBP51 are the more potent chaperones. Thus, it seems that both the affinity for Hsp90 and the differences in their chaperone properties, which may reflect their interaction with the non-native protein in the Hsp90 complex, are critical for the selective incorporation of a specific large PPIase.

Hsp90 is an abundant molecular chaperone that functions in an ATP-dependent manner in vivo. The ATP-binding site is located in the N-terminal domain of Hsp90. Here, we dissect the ATPase cycle of Hsp90 kinetically. We find that Hsp90 binds ATP with a two-step mechanism. The rate-limiting step of the ATPase cycle is the hydrolysis of ATP. Importantly, ATP becomes trapped and committed to hydrolyze during the cycle. In the isolated ATP-binding domain of Hsp90, however, the bound ATP was not committed and the turnover numbers were markedly reduced. Analysis of a series of truncation mutants of Hsp90 showed that C-terminal regions far apart in sequence from the ATP-binding domain are essential for trapping the bound ATP and for maximum hydrolysis rates. Our results suggest that ATP binding and hydrolysis drive conformational changes that involve the entire molecule and lead to repositioning of the N and C-terminal domains of Hsp90.

p23 is a co-chaperone of the heat shock protein Hsp90. p23 binds to Hsp90 in its ATP-bound state and, on its own, interacts specifically with non-native proteins. In our attempt to correlate these functions to specific regions of p23 we have identified an unstructured region in p23 that maps to the C-terminal part of the protein sequence. This unstructured region is dispensible for interaction of p23 with Hsp90, since truncated p23 can still form complexes with Hsp90. In contrast, however, truncation of the C-terminal 30 amino acid residues of p23 affects the ability of p23 to bind non-native proteins and to prevent their non-specific aggregation. The isolated C-terminal region itself is not able to act as a chaperone nor is it possible to complement truncated p23 by addition of this peptide. These results imply that the binding site for Hsp90 is contained in the folded domain of p23 and that for efficient interaction of p23 with non-native proteins both the folded domain and the C-terminal unstructured region are required.

Chaperonins are a ubiquitous class of ring-shaped oligomeric protein complexes that are of crucial importance for protein folding in vivo . Analysis of the underlying functional principles had relied mainly on model proteins the (un)folding of which is dominated by irreversible side-reactions. We used maltose-binding protein (MBP) as a substrate protein for GroEL, since the refolding of this protein is completely reversible and thus allows a detailed analysis of the molecular parameters that determine the interaction of GroEL with non-native protein. We show that MBP folding intermediates are effectively trapped by GroEL in a diffusion-controlled reaction. This complex is stabilized via unspecific hydrophobic interactions. Stabilization energies for wild-type MBP increasing linearly with ionic strength from 50 kJ/mol to 60 kJ/mol. Depending on the intrinsic folding rate and the hydrophobicity of the substrate protein, the interaction of GroEL with MBP folding intermedi- ates leads to a dramatically decreased apparent refolding rate of MBP (wild-type) or a complete suppression of folding (MBP folding mutant Y283D). On the basis of our data, a quantitative kinetic model of the GroEL-mediated folding cycle is proposed, which allows simulation of the partial reactions of the binding and release cycles under all conditions tested. In the presence of ATP and non-hydrolysable analogues, MBP is effectively released from GroEL, since the overall dissociation constant is reduced by three orders of magnitude. Interestingly, binding of nucleotide does not change the off rate by more than a factor of 3. However the on-rate is decreased by at least two orders of magnitude. Therefore, the rebinding reaction is prevented and folding occurs in solution.

The folding pathways of multi-domain proteins are still poorly understood due to the complexity of the reaction involving domain folding, association and, in many cases, prolyl cis/trans isomerization. Here, we have established a kinetic model for the folding of the Fab fragment of the antibody MAK 33 with intact disulfide bonds. Folding of the hetero-dimeric protein from the completely denatured, oxidized state comprises the pairwise association of the two domains of each chain with those of the partner protein. Both the reactivation of the Fab fragment in which the two constituent polypeptide chains were covalently linked via a cystine bond (Fab) and that of a mutant lacking this covalent linkage (Fab/-cys) were monitored by ELISA. Folding of the Fab fragment is a slow process, which can be described by a single exponential term. The kinetic phase reflects a folding step after the association of the two chains. The same reaction was detected in the folding of Fab/-cys but an additional rate-limiting step is involved that is due to a unimolecular step in the folding of the isolated light chain. This implies that, during Fab reactivation, Fd associates with the light chain at the stage of an earlier folding intermediate, thus eliminating the additional slow folding step of the light chain observed with Fab/-cys. Both in Fab and Fab/-cys renaturation, the folding reaction after association is determined by prolyl isomerization. Therefore, at least four different association-competent folding intermediates have to be postulated according to the folding stage of light chain and the configuration of at least one prolyl-peptide bond. Using the different substrate specificities of cyclophilin and FK506 binding protein, we have obtained evidence that Pro159 within the Fd fragment may be responsible for the observed slow folding phase after association, although three other proline residues adopt a cis configuration in the native protein. Furthermore, the data suggest that in the case of the Fab fragment, association is a prerequisite for cis/trans isomerization of prolyl peptide bonds, implying that the quaternary but not the tertiary structure determines the cis-configuration of the prolyl residue in Fd involved in the rate-limiting folding reaction.

Antibodies are modular proteins consisting of domains that exhibit a beta-sandwich structure, the so-called immunoglobulin fold. Despite structural similarity, differences in folding and stability exist between different domains. In particular, the variable domain of the light chain V(L) is unusual as it is associated with misfolding diseases, including the pathologic assembly of the protein into fibrillar structures. Here, we have analysed the folding pathway of a V(L) domain with a view to determine features that may influence the relationship between productive folding and fibril formation. The V(L) domain from MAK33 (murine monoclonal antibody of the subtype kappa/IgG1) has not previously been associated with fibrillisation but is shown here to be capable of forming fibrils. The folding pathway of this V(L) domain is complex, involving two intermediates in different pathways. An obligatory early molten globule-like intermediate with secondary structure but only loose tertiary interactions is inferred. The native state can then be formed directly from this intermediate in a phase that can be accelerated by the addition of prolyl isomerases. However, an alternative pathway involving a second, more native-like intermediate is also significantly populated. Thus, the protein can reach the native state via two distinct folding pathways. Comparisons to the folding pathways of other antibody domains reveal similarities in the folding pathways; however, in detail, the folding of the V(L) domain is striking, with two intermediates populated on different branches of the folding pathway, one of which could provide an entry point for molecules diverted into the amyloid pathway.

Small heat shock proteins (sHsps) are a widespread and diverse class of molecular chaperones. In vivo, sHsps contribute to thermotolerance. Recent evidence suggests that their function in the cellular chaperone network is to maintain protein homeostasis by complexing a variety of non-native proteins. One of the most characteristic features of sHsps is their organization into large, sphere-like structures commonly consisting of 12 or 24 subunits. Here, we investigated the functional and structural properties of Hsp20.2, an sHsp from Archaeoglobus fulgidus, in comparison to its relative, Hsp16.5 from Methanocaldococcus jannaschii. Hsp20.2 is active in suppressing the aggregation of different model substrates at physiological and heat-stress temperatures. Electron microscopy showed that Hsp20.2 forms two distinct types of octahedral oligomers of slightly different sizes, indicating certain structural flexibility of the oligomeric assembly. By three-dimensional analysis of electron microscopic images of negatively stained specimens, we were able to reconstitute 3D models of the assemblies at a resolution of 19 A. Under conditions of heat stress, the distribution of the structurally different Hsp20.2 assemblies changed, and this change was correlated with an increased chaperone activity. In analogy to Hsp20.2, Hsp16.5 oligomers displayed structural dynamics and exhibited increased chaperone activity under conditions of heat stress. Thus, temperature-induced conformational regulation of the activity of sHsps may be a general phenomenon in thermophilic archaea.

Disulfide bridges are one of the most important factors stabilizing the native structure of a protein. Whereas the basis for their stabilizing effect is well understood, their role in a protein folding reaction still seems to require further attention. We used the constant domain of the antibody light chain (CL), a representative of the ubiquitous immunoglobulin (Ig)-superfamily, to delineate the kinetic role of its single buried disulfide bridge. Independent of its redox state, the monomeric CL domain adopts a typical Ig-fold under native conditions and does not retain significant structural elements when unfolded. Interestingly, its folding pathway is strongly influenced by the disulfide bridge. The more stable oxidized protein folds via a highly structured on-pathway intermediate, whereas the destabilized reduced protein populates a misfolded off-pathway species on its way to the native state. In both cases, the formation of the intermediate species is shown to be independent of the isomerization state of the Tyr141-Pro142 bond. Our results demonstrate that the internal disulfide bridge in an antibody domain restricts the folding pathway by bringing residues of the folding nucleus into proximity thus facilitating the way to the native state.

Small heat shock proteins (sHsps) are molecular chaperones that specifically bind non-native proteins and prevent them from irreversible aggregation. A key trait of sHsps is their existence as dynamic oligomers. Hsp26 from Saccharomyces cerevisiae assembles into a 24mer, which becomes activated under heat shock conditions and forms large, stable substrate complexes. This activation coincides with the destabilization of the oligomer and the appearance of dimers. This and results from other groups led to the generally accepted notion that dissociation might be a requirement for the chaperone mechanism of sHsps. To understand the chaperone mechanism of sHsps it is crucial to analyze the relationship between chaperone activity and stability of the oligomer. We generated an Hsp26 variant, in which a serine residue of the N-terminal domain was replaced by cysteine. This allowed us to covalently crosslink neighboring subunits by disulfide bonds. We show that under reducing conditions the structure and function of this variant are indistinguishable from that of the wild-type protein. However, when the cysteine residues are oxidized, the dissociation into dimers at higher temperatures is no longer observed, yet the chaperone activity remains unaffected. Furthermore, we show that the exchange of subunits between Hsp26 oligomers is significantly slower than substrate aggregation and even inhibited in the presence of disulfide bonds. This demonstrates that the rearrangements necessary for shifting Hsp26 from a low to a high affinity state for binding non-native proteins occur without dissolving the oligomer.

Oxidation and folding of secretory proteins in the endoplasmic reticulum (ER) depends on the presence of chaperones and oxidoreductases. Two of the oxidoreductases present in the ER of mammalian cells are protein disulfide isomerase (PDI) and ERp57. In this study, we investigated the influence of ERp57 on the in vitro reoxidation and refolding of an antibody Fab fragment. Our results show that ERp57 shares functional properties with PDI and that both are clearly different from other oxidoreductases. The reactivation of the denatured and reduced Fab fragment was enhanced significantly in the presence of ERp57 with kinetics and redox dependence of the reactivation reaction comparable to those obtained for PDI. These properties were not influenced by the presence of calnexin. Furthermore, whereas PDI cooperates with the immunoglobulin heavy chain binding protein (BiP), no synergistic effect could be observed for BiP and ERp57. These results indicate that the cooperation of the two oxidoreductases with different partner proteins may explain their different roles in the folding of proteins in the ER.

The molecular chaperone Hsp90 mediates the ATP-dependent activation of a large number of proteins involved in signal transduction. During this process, Hsp90 was found to associate transiently with several accessory factors, such as p23/Sba1, Hop/Sti1, and prolyl isomerases. It has been shown that ATP hydrolysis triggers conformational changes within Hsp90, which in turn are thought to mediate conformational changes in the substrate proteins, thereby causing their activation. The specific role of the partner proteins in this process is unknown. Using proteins from Saccharomayces cerevisiae , we characterized the interaction of Hsp90 with its partner protein p23/Sba1. Our results show that the nucleotide-dependent N-terminal dimerization of Hsp90 is necessary for the binding of Sba1 to Hsp90 with an affinity in the nanomolar range. Two Sba1 molecules were found to bind per Hsp90 dimer. Sba1 binding to Hsp90 resulted in a decreased ATPase activity, presumably by trapping the hydrolysis state of Hsp90 ATP. Ternary complexes of Hsp90 Sba1 could be formed with the prolyl isomerase Cpr6, but not with Sti1. Based on these findings, we propose a model that correlates the ordered assembly of the Hsp90 co-chaperones with distinct steps of the ATP hydrolysis reaction during the chaperone cycle.

The immunoglobulin C(H)2 domain is a simple model system suitable for the study of the folding of all-beta-proteins. Its structure consists of two beta-sheets forming a greek-key beta-barrel, which is stabilized by an internal disulfide bridge located in the hydrophobic core. Crystal structures of various antibodies suggest that the C(H)2 domains of the two heavy chains interact with their sugar moieties and form a homodimer. Here, we show that the isolated, unglycosylated C(H)2 domain is a monomeric protein. Equilibrium unfolding was a two-state process, and the conformational stability is remarkably low compared to other antibody domains. Folding kinetics of C(H)2 were found to consist of several phases. The reactions could be mapped to three parallel pathways, two of which are generated by prolyl isomerizations in the unfolded state. The slowest folding reaction, which was observed only after long-term denaturation, could be catalyzed by a prolyl isomerase. The majority of the unfolded molecules, however, folded more rapidly, on a time-scale of minutes. Presumably, these molecules also have to undergo prolyl isomerization before reaching the native state. In addition, we detected a small number of fast-folding molecules in which all proline residues appear to be in the correct conformation. On both prolyl isomerization limited pathways, the formation of partly structured intermediates could be observed.

The human tumor suppressor protein p53 is understood only to some extent on a structural level. We performed a comprehensive biochemical and biophysical structure–function analysis of p53 full-length protein and p53 fragments. The analysis showed that p53 and the fragments investigated form stable functional units. Full-length p53 and the tetrameric fragment N93p53 (residues 93–393) are, however, destabilized significantly compared to the monomeric core domain (residues 94–312) and the monomeric fragment p53C312 (residues 1–312). At the physiological temperature of 37 °C and in the absence of modifications or stabilizing partners, wild-type p53 is more than 50% unfolded correlating with a 75% loss in DNA-binding activity. Furthermore the analysis of CD spectra revealed that full-length p53 contains large unstructured regions in its N and C-terminal parts. Our results indicate that full-length p53 is a modular protein consisting of defined structured and unstructured regions. We propose that p53 belongs to the growing family of loosely folded or partially unstructured native proteins. The lack of a rigid structure combined with the low overall stability may allow the physiological interaction of p53 with a multitude of partner proteins and the regulation of its turnover.

BiP is an Hsp70 homologue found in the endoplasmic reticulum of eukaryotic cells. Like other Hsp70 chaperones, BiP interacts with its substrate proteins in an ATP-dependent manner. The functional analysis has so far been performed mainly with short, synthetic peptides. Here, we present an experimental system that allows to study the partial reactions of the BiP chaperone cycle for a natural substrate protein domain in its soluble, stably unfolded conformation. This unfolded antibody domain forms a binary complex with BiP in the absence of ATP. The dissociation of the BiP dimer seems to be the rate-limiting step in this reaction. The BiP-C(H)3 complexes dissociate rapidly in the presence of ATP. The affinity for BiP-binding peptides and the non-native antibody domain was determined to be similar, suggesting that only the peptide binding site is involved in these interactions. Furthermore, these results imply that, also in the context of the antibody domain, an extended peptide sequence is recognized. However, the accessibility of the BiP-binding site in the non-native protein seems to influence the kinetics of complex formation.

The simplest naturally occurring model system for studying immunoglobulin folding and assembly is the non-covalent homodimer formed by the C-terminal domains (CH3) of the heavy chains of IgG. Here, we describe the structure of recombinant CH3 dimer as determined by X-ray crystallography and an analysis of the folding pathway of this protein. Under conditions where prolyl isomerization does not contribute to the folding kinetics, formation of the beta-sandwich structure is the rate-limiting step. beta-Sheet formation of CH3 is a slow process, even compared to other antibody domains, while the subsequent association of the folded monomers is fast. After long-time denaturation, the majority of the unfolded CH3 molecules reaches the native state in two serial reactions, involving the re-isomerization of the Pro35-peptide bond to the cis configuration. The species with the wrong isomer accumulate as a monomeric intermediate. Importantly, the isomerization to the correct cis configuration is the prerequisite for dimerization of the CH3 domain. In contrast, in the Fab fragment of the same antibody, prolyl isomerization occurs after dimerization demonstrating that within one protein, comprised of highly homologous domains, both the kinetics of beta-sandwich formation and the stage at which prolyl isomerization occurs during the folding process can be completely different.

The Escherichia coli GroE chaperones assist protein folding under conditions where no spontaneous folding occurs. To achieve this, the cooperation of GroEL and GroES, the two protein components of the chaperone system, is an essential requirement. While in many cases GroE simply suppresses unspecific aggregation of non-native proteins by encapsulation, there are examples where folding is accelerated by GroE. Using maltose-binding protein (MBP) as a substrate for GroE, it had been possible to define basic requirements for catalysis of folding. Here, we have analyzed key steps in the interaction of GroE and the MBP mutant Y283D during catalyzed folding. In addition to high temperature, high ionic strength was shown to be a restrictive condition for MBP Y283D folding. In both cases, the complete GroE system (GroEL, GroES and ATP) compensates the deceleration of MBP Y283D folding. Combining kinetic folding experiments and electron microscopy of GroE particles, we demonstrate that at elevated temperatures, symmetrical GroE particles with GroES bound to both ends of the GroEL cylinder play an important role in the efficient catalysis of MBP Y283D refolding. In principle, MBP Y283D folding can be catalyzed during one encapsulation cycle. However, because the commitment to reach the native state is low after only one cycle of ATP hydrolysis, several interaction cycles are required for catalyzed folding.

beta-D-Galactosidase from Escherichia coli is one of the largest tetrameric enzymes known at present. Although its physiological importance, the regulation of its synthesis, its enzymatic properties and its structure are well established, little is known about the stability and the folding pathway of this enzyme. Here we show that the overall folding mechanism of chemically denatured beta-galactosidase consists of three stages: (i) formation of elements of secondary structure; (ii) collapse to subdomains and structured monomers; (iii) association to the native quaternary structure via dimeric intermediates. The first rate-limiting step is the association of structured monomers to form dimers in a bi-molecular reaction, with a rate constant of 4.3x10(3) M-1 s-1 at 20 degreesC. The second rate-limiting uni-molecular folding step leads to dimers which are competent for further association, with a rate constant of 0.5x10(-3) s-1 at 20 degreesC. Tetramers form from these dimers in a fast reaction. By determining a similar mechanism for alpha-complementation of beta-galactosidase fragments it could be confirmed that beta-galactosidase follows a consecutive bi-uni-molecular mechanism of folding and association.