Kinetic analysis of the folding and unfolding of a mutant form of bovine pancreatic trypsin inhibitor lacking the cysteine-14 and -38 thiols

The kinetics of the disulfide-coupled unfolding-refolding transition of a mutant form of bovine pancreatic trypsin inhibitor (BPTI) lacking Cys-14 and -38 were measured and compared to previous results for the wild-type protein and other modified forms. The altered cysteines, which were changed to serine in the mutant protein, are normally paired in a disulfide in the native protein but from disulfides with Cys-5 in two-disulfide kinetic intermediates during folding. Although the mutant protein could fold efficiently, the kinetics of both folding and unfolding were altered, reflecting the roles of these cysteines in the two-disulfide intermediates with "wrong" disulfides. The intramolecular rate constant for the formation of the second disulfide of the native mutant protein was more than 10(3)-fold lower than that for the formation of a second disulfide during the refolding of the wild-type protein. The observed rate of unfolding of the mutant protein was also lower than that of the wild-type protein, demonstrating that the altered cysteines are involved in the intramolecular rearrangements that are the rate-determining step in the unfolding of the wild-type protein. These results confirm the previous conclusion [Creighton, T.E. (1977) J. Mol. Biol. 113, 275-293] that the energetically preferred pathway for folding and unfolding of BPTI includes intramolecular rearrangements of intermediates in which Cys-14 and -38 are paired in disulfides not present in the native protein. The present results are also consistent with other, less detailed, studies with similar mutants lacking Cys-14 and -38 [Marks, C.B., Naderi, H., Kosen, P.A., Kuntz, I.D., & Anderson, S. (1987) Science (Washington, D.C.) 235, 1370-1371].     

A genetic screen to identify variants of bovine pancreatic trypsin inhibitor with altered folding energetics.

A genetic screening procedure has been developed to identify mutant forms of bovine pancreatic trypsin inhibitor (BPTI) that can fold to an active conformation but are inactivated more rapidly than the wild-type protein. Small cultures of Escherichia coli containing plasmids with mutagenized BPTI genes were grown in microtiter plates, lysed, and treated with dithiothreitol (DTT). Under these conditions, unfolding and inactivation of the wild-type protein has a half-time of about 10 hours. Variants of BPTI that are inactivated within 1 hour were identified by adding trypsin and a chromogenic substrate. Approximately 11,000 mutagenized clones were screened in this way and 75 clones that produce proteins that can fold but are inactivated by DTT were isolated. The genes coding for 68 "DTT-sensitive" mutant proteins were sequenced, and 25 different single amino acid substitutions at 15 of the 58 residues of the protein were identified. Most of the altered residues are largely buried in the core of the native wild-type structure and are highly conserved among proteins homologous to BPTI. These results indicate that a large fraction of the sequence of the protein contributes to the kinetic stability of the active conformation, but it also appears that substitutions can be tolerated at most sites without completely preventing folding. Because this genetic screen is based on changes in folding energetics, further studies of the isolated mutants are expected to provide information about the roles of the altered residues in folding and unfolding.     

Different circular permutations produced different folding nuclei in proteins: a computational study

There have been many studies about the effect of circular permutation on the transition state/folding nucleus of proteins, with sometimes conflicting conclusions from different proteins and permutations. To clarify this important issue, we have studied two circular permutations of a lattice protein model with side-chains. Both permuted sequences have essentially the same native state as the original (wild-type) sequence. Circular permutant 1 cuts at the folding nucleus of the wild-type sequence. As a result, the permutant has a drastically different nucleus and folds more slowly than wild-type. In contrast, circular permutant 2 involves an incision at a site unstructured in the wild-type transition state, and the wild-type nucleus is largely retained in the permutant. In addition, permutant 2 displays both two-state and multi-state folding, with a native-like intermediate state occasionally populated. Neither the wild-type nor permutant 1 has a similar intermediate, and both fold in an apparently two-state manner. Surprisingly, permutant 2 folds at a rate identical with that of the wild-type. The intermediate in permutant 2 is stabilised by native and non-native interactions, and cannot be classified simply as on or off-pathway. So we advise caution in attributing experimental data to on or off-pathway intermediates. Finally, our work illuminates the results on alpha-spectrin SH3, chymotrypsin inhibitor 2 and beta-lactoglobulin, and supports a key assumption in the experimental efforts to locate potential nucleation sites of real proteins via circular permutations.     

Mutational analysis of the BPTI folding pathway: II. Effects of aromatic-->leucine substitutions on folding kinetics and thermodynamics.

The rates of the individual steps in the disulfide-coupled folding and unfolding of eight BPTI variants, each containing a single aromatic to leucine amino acid replacement, were measured. From this analysis, the contributions of the four phenylalanine and four tyrosine residues to the stabilities of the native protein and the disulfide-bonded folding intermediates were determined. While the substitutions were found to destabilize the native protein by 2 to 7 kcal/mol, they had significantly smaller effects on the intermediates that represent the earlier stages of folding, even when the site of the substitution was located within the ordered regions of the intermediates. These results suggest that stabilizing interactions contribute less to conformational stability in the context of a partially folded intermediate than in a fully folded native protein, perhaps because of decreased cooperativity among the individual interactions. The kinetic analysis also provides new information about the transition states associated with the slowest steps in folding and unfolding, supporting previous suggestions that these transition states are extensively unfolded. Although the substitutions caused large changes in the distribution of folding intermediates and in the rates of some steps in the folding pathway, the kinetically-preferred pathway for all of the variants involved intramolecular disulfide rearrangements, as observed previously for the wild-type protein. These results suggest that the predominance of the rearrangement mechanism reflects conformational constraints present relatively early in the folding pathway.     

Phage P22 tailspike protein: removal of head-binding domain unmasks effects of folding mutations on native-state thermal stability.

A shortened, recombinant protein comprising residues 109-666 of the tailspike endorhamnosidase of Salmonella phage P22 was purified from Escherichia coli and crystallized. Like the full-length tailspike, the protein lacking the amino-terminal head-binding domain is an SDS-resistant, thermostable trimer. Its fluorescence and circular dichroism spectra indicate native structure. Oligosaccharide binding and endoglycosidase activities of both proteins are identical. A number of tailspike folding mutants have been obtained previously in a genetic approach to protein folding. Two temperature-sensitive-folding (tsf) mutations and the four known global second-site suppressor (su) mutations were introduced into the shortened protein and found to reduce or increase folding yields at high temperature. The mutational effects on folding yields and subunit folding kinetics parallel those observed with the full-length protein. They mirror the in vivo phenotypes and are consistent with the substitutions altering the stability of thermolabile folding intermediates. Because full-length and shortened tailspikes aggregate upon thermal denaturation, and their denaturant-induced unfolding displays hysteresis, kinetics of thermal unfolding were measured to assess the stability of the native proteins. Unfolding of the shortened wild-type protein in the presence of 2% SDS at 71 degrees C occurs at a rate of 9.2 x 10(-4) s(-1). It reflects the second kinetic phase of unfolding of the full-length protein. All six mutations were found to affect the thermal stability of the native protein. Both tsf mutations accelerate thermal unfolding about 10-fold. Two of the su mutations retard thermal unfolding up to 5-fold, while the remaining two mutations accelerate unfolding up to 5-fold. The mutational effects can be rationalized on the background of the recently determined crystal structure of the protein.     

Folding of intracellular retinol and retinoic acid binding proteins

The folding mechanisms of cellular retinol binding protein II (CRBP II), cellular retinoic acid binding protein I (CRABP I), and cellular retinoic acid binding protein II (CRABP II) were examined. These beta-sheet proteins have very similar structures and higher sequence homologies than most proteins in this diverse family. They have similar stabilities and show completely reversible folding at equilibrium with urea as a denaturant. The unfolding kinetics of these proteins were monitored during folding and unfolding by circular dichroism (CD) and fluorescence. During unfolding, CRABP II showed no intermediates, CRABP I had an intermediate with nativelike secondary structure, and CRBP II had an intermediate that lacked secondary structure. The refolding kinetics of these proteins were more similar. Each protein showed a burst-phase change in intensity by both CD and fluorescence, followed by a single observed phase by both CD and fluorescence and one or two additional refolding phases by fluorescence. The fluorescence spectral properties of the intermediate states were similar and suggested a gradual increase in the amount of native tertiary structure present for each step in a sequential path. However, the rates of folding differed by as much as 3 orders of magnitude and were slower than those expected from the contact order and topology of these proteins. As such, proteins with the same final structure may not follow the same route to the native state.     

Amyloid fibril formation by circularly permuted and C-terminally deleted mutants

The natural N- and C-termini, i.e., the given order of secondary structure segments, are critical for protein folding and stability, as shown by several studies using circularly permuted proteins, mutants that have their N- and C-termini linked and are then digested at another site to create new termini. A previous work showed that circularly permuted mutants of sperm whale myoglobin (Mb) are functional, have native-like folding and bind heme, but are less stable than the wild-type protein and aggregate. The ability of wild-type myoglobin to form amyloid fibrils has been established recently, and because circularly permuted mutations are destabilizing, we asked whether these permutations would also affect the rate of amyloid fibril formation. Our investigations revealed that, indeed, the circularly permuted mutants formed cytotoxic fibrils at a rate higher than that of the wild-type. To further investigate the role of the C-terminus in the overall stability of the protein, we investigated two C-terminally deleted mutant, Mb(1-123) and Mb(1-99), and found that Mb(1-123) formed cytotoxic fibrils at a higher rate than that of the wild-type while Mb(1-99) formed cytotoxic fibrils at a similar rate than that of the wild-type. Collectively, our findings show that the native position of both the N-and C-termini is important for the precise structural architecture of myoglobin.     

Circular permutation and deletion studies of myoglobin indicate that the correct position of its N-terminus is required for native stability and solubility but not for native-like heme binding and folding

We studied the effect of deleted and circularly permuted mutations in sperm whale myoglobin and present here results on three classes of mutants: (i) a deletion mutant, Mb(1)(-)(99), in which the C-terminal helices, G and H, were removed; (ii) two circular permutations, Mb-B_GHA, in which helix B is N-terminal and helix A is C-terminal, and Mb-C_GHAB, in which helix C is N-terminal and helices A and B are C-terminal; and (iii) a deleted circular permutation, Mb-HAB_F, in which helix H is N-terminal, helix F is C-terminal, and helix G is deleted. The conformational characteristics of the apo and holo forms of these mutants were determined at neutral pH, by spectroscopic and hydrodynamic methods. The apo form of the deleted and permuted mutants exhibited a stronger tendency to aggregate and had lower ellipticity than the wild type. The mutants retained the ability to bind heme, but only the circularly permuted holoproteins had native-like heme binding and folding. These results agree with the theory that myoglobin has a central core that is able to bind heme, but also indicate that the presence of N- and C-terminal helices is necessary for native-like heme pocket formation. Because the holopermuteins were less stable than the wild-type protein and aggregated, we propose that the native position of the N-terminus is important for the precise structural architecture of myoglobin.     

Topological determinants of protein unfolding rates

For proteins that fold by two-state kinetics, the folding and unfolding processes are believed to be closely related to their native structures. In particular, folding and unfolding rates are influenced by the native structures of proteins. Thus, we focus on finding important topological quantities from a protein structure that determine its unfolding rate. After constructing graphs from protein native structures, we investigate the relationships between unfolding rates and various topological quantities of the graphs. First, we find that the correlation between the unfolding rate and the contact order is not as prominent as in the case of the folding rate and the contact order. Next, we investigate the correlation between the unfolding rate and the clustering coefficient of the graph of a protein native structure, and observe no correlation between them. Finally, we find that a newly introduced quantity, the impact of edge removal per residue, has a good overall correlation with protein unfolding rates. The impact of edge removal is defined as the ratio of the change of the average path length to the edge removal probability. From these facts, we conclude that the protein unfolding process is closely related to the protein native structure.     

Prediction of folding mechanism for circular-permuted proteins

Recent theoretical and experimental studies have suggested that real proteins have sequences with sufficiently small energetic frustration that topological effects are central in determining the folding mechanism. A particularly interesting and challenging framework for exploring and testing the viability of these energetically unfrustrated models is the study of circular-permuted proteins. Here we present the results of the application of a topology-based model to the study of circular permuted SH3 and CI2, in comparison with the available experimental results. The folding mechanism of the permuted proteins emerging from our simulations is in very good agreement with the experimental observations. The differences between the folding mechanisms of the permuted and wild-type proteins seem then to be strongly related to the change in the native state topology.     

Folding mechanism of three structurally similar β-sheet proteins

The folding mechanism of cellular retinoic acid binding protein I (CRABP I), cellular retinol binding protein II (CRBP II), and intestinal fatty acid binding protein (IFABP) were investigated to determine if proteins with similar native structures have similar folding mechanisms. These mostly β-sheet proteins have very similar structures, despite having as little as 33% sequence similarity. The reversible urea denaturation of these proteins was characterized at equilibrium by circular dichroism and fluorescence. The data were best fit by a two-state model for each of these proteins, suggesting that no significant population of folding intermediates were present at equilibrium. The native states were of similar stability with free energies (linearly extrapolated to 0 M urea, ΔG) of 6.5, 8.3, and 5.5 kcal/mole for CRABP I, CRBP II, and IFABP, respectively. The kinetics of the folding and unfolding processes for these proteins was monitored by stopped-flow CD and fluorescence. Intermediates were observed during both the folding and unfolding of all of these proteins. However, the overall rates of folding and unfolding differed by nearly three orders of magnitude. Further, the spectroscopic properties of the intermediate states were different for each protein, suggesting that different amounts of secondary and/or tertiary structure were associated with each intermediate state for each protein. These data show that the folding path for proteins in the same structural family can be quite different, and provide evidence for different folding landscapes for these sequences. Proteins 33:107–118, 1998. © 1998 Wiley-Liss, Inc.     

Unfolding and folding kinetics of amyotrophic lateral sclerosis-associated mutant Cu,Zn superoxide dismutases

More than 110 mutations in dimeric, Cu,Zn superoxide dismutase (SOD) have been linked to the fatal neurodegenerative disease, amyotrophic lateral sclerosis (ALS). In both human patients and mouse model studies, protein misfolding has been implicated in disease pathogenesis. A central step in understanding the misfolding/aggregation mechanism of this protein is the elucidation of the folding pathway of SOD. Here we report a systematic analyses of unfolding and folding kinetics using single- and double-jump experiments as well as measurements as a function of guanidium chloride, protein, and metal concentration for fully metallated (holo) pseudo wild-type and ALS-associated mutant (E100G, G93R, G93A, and metal binding mutants G85R and H46R) SODs. The kinetic mechanism for holo SODs involves native dimer, monomer intermediate, and unfolded monomer, with variable metal dissociation from the monomeric states depending on solution conditions. The effects of the ALS mutations on the kinetics of the holoproteins in guanidium chloride are markedly different from those observed previously for acid-induced unfolding and for the unmetallated (apo) forms of the proteins. The mutations decrease the stability of holo SOD mainly by increasing unfolding rates, which is particularly pronounced for the metal-binding mutants, and have relatively smaller effects on the observed folding kinetics. Mutations also seem to favour increased formation of a Zn-free monomer intermediate, which has been implicated in the formation of toxic aggregates. The results reveal the kinetic basis for the extremely high stability of wild-type holo SOD and the possible consequences of kinetic changes for disease.     

On the role of some conserved and nonconserved amino acid residues in the transitional state and intermediate of apomyoglobin folding

The contributions of some amino acid residues in the A, B, G, and H helices to the formation of the folding nucleus and folding intermediate of apomyoglobin were estimated. The effects of point substitutions of Ala for hydrophobic amino acid residues on the structural stability of the native (N) protein and its folding intermediate (I), as well as on the folding/unfolding rates for four mutant apomyoglobin forms, were studied. The equilibrium and kinetic studies of the folding/unfolding rates of these mutant proteins in a wide range of urea concentrations demonstrated that their native state was considerably destabilized as compared with the wild-type protein, whereas the stability of the intermediate state changed moderately. It was shown that the amino acid residues in the A, G, and H helices contributed insignificantly to the stabilization of the apomyoglobin folding nucleus in the rate-limiting I ? N transition, taking place after the formation of the intermediate, whereas the residue of the B helix was of great importance in the formation of the folding nucleus in this transition.     

Secretion incompetence of bovine pancreatic trypsin inhibitor expressed in Escherichia coli.

There is increasing evidence that protein folding and protein export are competing processes in prokaryotic cells. Virtually all secretion studies reported to date, however, have employed proteins that are relatively uncharacterized in terms of their folding behavior and three-dimensional structure. In contrast, the structural and biochemical parameters governing the folding of bovine pancreatic trypsin inhibitor (BPTI) and several of its mutants have been studied intensively. We therefore undertook a study of the secretion behavior in Escherichia coli of recombinant BPTI and its mutants. Wild-type BPTI and two well-characterized folding mutants (C14A, C38A)BPTI and (C30A, C51A)BPTI (missing the 14-38 and 30-51 disulfide bonds, respectively), were investigated by analyzing their expression fused to an E. coli signal sequence or to two synthetic IgG-binding domains of staphylococcal protein A. Both disulfide mutants are destabilized relative to wild-type BPTI and exhibit markedly altered folding kinetics: one (C14A, C38A) folds more slowly than wild-type BPTI and the other (C30A, C51A) unfolds more rapidly. Both mutants were observed to be exported 3-10 times more efficiently than the wild-type molecule. Moreover, the levels of unprocessed preprotein in the cytoplasm were severalfold higher for the wild-type fusion than for the fusion to the two folding mutants. Intracellular degradation of the BPTI moiety was also observed. These results are consistent with traffic of intracellular BPTI preproteins on at least three routes along the secretory pathway: (a) facile secretion of unfolded material, (b) intracellular folding leading to secretion blockage, and (c) degradation followed by export of truncated molecules. A novel feature of these findings is the implication that disulfide bonds can form in the bacterial cytoplasm and lead to secretion incompetence.     

Equilibrium and kinetics of the folding and unfolding of canine milk lysozyme

The equilibrium and kinetics of canine milk lysozyme folding/unfolding were studied by peptide and aromatic circular dichroism and tryptophan fluorescence spectroscopy. The Ca2+-free apo form of the protein exhibited a three-state equilibrium unfolding, in which the molten globule state is well populated as an unfolding intermediate. A rigorous analysis of holo protein unfolding, including the data from the kinetic refolding experiments, revealed that the holo protein also underwent three-state unfolding with the same molten globule intermediate. Although the observed kinetic refolding curves of both forms were single-exponential, a burst-phase change in the peptide ellipticity was observed in both forms, and the burst-phase intermediates of both forms were identical to each other with respect to their stability, indicating that the intermediate does not bind Ca2+. This intermediate was also shown to be identical to the molten globule state observed at equilibrium. The phi-value analysis, based on the effect of Ca2+ on the folding and unfolding rate constants, showed that the Ca2+-binding site was not yet organized in the transition state of folding. A comparison of the result with that previously reported for alpha-lactalbumin indicated that the folding initiation site is different between canine milk lysozyme and alpha-lactalbumin, and hence, the folding pathways must be different between the two proteins. These results thus provide an example of the phenomenon wherein proteins that are very homologous to each other take different folding pathways. It is also shown that the native state of the apo form is composed of at least two species that interconvert.     

The underlying mechanism for the diversity of disulfide folding pathways

The disulfide folding pathway of bovine pancreatic trypsin inhibitor (BPTI) is characterized by the predominance of folding intermediates with native-like structures. Our laboratory has recently analyzed the folding pathway(s) of four 3-disulfide-containing proteins, including hirudin, potato carboxypeptidase inhibitor, epidermal growth factor, and tick anticoagulant peptide. Their folding mechanism(s) differ from that of BPTI by 1) a higher degree of heterogeneity of 1- and 2-disulfide intermediates and 2) the presence of 3-disulfide scrambled isomers as folding intermediates. To search for the underlying causes of these diversities, we conducted kinetic analyses of the reductive unfolding of these five proteins. The experiment of reductive unfolding was designed to evaluate the relative stability and interdependence of disulfide bonds in the native protein. It is demonstrated here that among these five proteins, there exists a striking correlation between the mechanism(s) of reductive unfolding and that of oxidative folding. Those proteins with their native disulfide bonds reduced in a collective and simultaneous manner exhibit both a high degree of heterogeneity of folding intermediates and the accumulation of scrambled isomers along the folding pathway. A sequential reduction of the native disulfide bonds is associated with the presence of predominant intermediates with native- like structures.     

Pathway modulation, circular permutation and rapid RNA folding under kinetic control

The thermodynamics and folding kinetics of a circularly permuted construct of the ribozyme from Bacillus subtilis RNase P are analyzed and compared with the folding properties of the wild-type ribozyme using optical spectroscopy and catalytic activity. The folding of the wild-type ribozyme is slow due to the rearrangement of kinetically trapped species containing misfolded structures. To test whether any misfolded structure arises from interactions between the two independently folding domains of the RNase P RNA, a circular permuted form was created where one of the two phosphodiester bonds connecting these domains is broken. This construct folds approximately 15-fold faster (t1/2 approximately nine seconds) than the wild-type ribozyme at 37 degreesC. While the complete folding of both domains is kinetically indistinguishable in the wild-type ribozyme, one domain folds much faster than the other domain in the circularly permuted construct. Hence, the major kinetic trap in the folding of the wild-type RNase P RNA involves interdomain interactions. This kinetic trap is avoidable at 37 degreesC in the circularly permuted RNA. However, at temperatures below 30 degreesC or when refolding begins from an equilibrium intermediate stabilized by submillimolar concentrations of Mg2+, a subpopulation containing an interdomain misfold still forms. These results indicate that the folding pathway of this large RNA is highly malleable and can be under kinetic control.     

Rapid unfolding of a domain populates an aggregation-prone intermediate that can be recognized by GroEL

Some amino acid substitutions in phage P22 coat protein cause a temperature-sensitive folding (tsf) phenotype. In vivo, these tsf amino acid substitutions cause coat protein to aggregate and form intracellular inclusion bodies when folded at high temperatures, but at low temperatures the proteins fold properly. Here the effects of tsf amino acid substitutions on folding and unfolding kinetics and the stability of coat protein in vitro have been investigated to determine how the substitutions change the ability of coat protein to fold properly. The equilibrium unfolding transitions of the tsf variants were best fit to a three-state model, N if I if U, where all species concerned were monomeric, a result confirmed by velocity sedimentation analytical ultracentrifugation. The primary effect of the tsf amino acid substitutions on the equilibrium unfolding pathway was to decrease the stability (DeltaG) and the solvent accessibility (m-value) of the N if I transition. The kinetics of folding and unfolding of the tsf coat proteins were investigated using tryptophan fluorescence and circular dichroism (CD) at 222 nm. The tsf amino acid substitutions increased the rate of unfolding by 8-14-fold, with little effect on the rate of folding, when monitored by tryptophan fluorescence. In contrast, when folding or unfolding reactions were monitored by CD, the reactions were too fast to be observed. The tsf coat proteins are natural substrates for the molecular chaperones, GroEL/S. When native tsf coat protein monomers were incubated with GroEL, they bound efficiently, indicating that a folding intermediate was significantly populated even without denaturant. Thus, the tsf coat proteins aggregate in vivo because of an increased propensity to populate this unfolding intermediate.     

pH-jump-induced folding and unfolding studies of barstar: evidence for multiple folding and unfolding pathways.

Equilibrium and kinetic characterization of the high pH-induced unfolding transition of the small protein barstar have been carried out in the pH range 7-12. A mutant form of barstar, containing a single tryptophan, Trp 53, completely buried in the core of the native protein, has been used. It is shown that the protein undergoes reversible unfolding above pH 10. The pH 12 form (the D form) appears to be as unfolded as the form unfolded by 6 M guanidine hydrochloride (GdnHCl) at pH 7 (the U form): both forms have similar fluorescence and far-UV circular dichroism (CD) signals and have similar sizes, as determined by dynamic light scattering and size-exclusion chromatography. No residual structure is detected in the D form: addition of GdnHCl does not alter its fluorescence and far-UV CD properties. The fluorescence signal of Trp 53 has been used to monitor folding and unfolding kinetics. The kinetics of folding of the D form in the pH range 7-11 are complex and are described by four exponential processes, as are the kinetics of unfolding of the native state (N state) in the pH range 10.5-12. Each kinetic phase of folding decreases in rate with increase in pH from 7 to 10.85, and each kinetic phase of unfolding decreases in rate with decrease in pH from 12 to 10.85. At pH 10.85, the folding and unfolding rates for any particular kinetic phase are identical and minimal. The two slowest phases of folding and unfolding have identical kinetics whether measured by Trp 53 fluorescence or by mean residue ellipticity at 222 nm. Direct determination of the increase in the N state with time of folding at pH 7 and of the D form with time of unfolding at pH 12, by means of double-jump assays, show that between 85 and 95% of protein molecules fold or unfold via fast pathways between the two forms. The remaining 5-15% of protein molecules appear to fold or unfold via slower pathways, on which at least two intermediates accumulate. The mechanism of folding from the high pH-denatured D form is remarkably similar to the mechanism of folding from the urea or GdnHCl-denatured U form.