Aggregation events occur prior to stable intermediate formation during refolding of interleukin 1beta

Factors influencing amyloid fibril formation

Protein aggregation is a complex process with several stages that lead to the formation of complex structures and shapes with a broad variability in stability and toxicity. The aggregation process is affected by various factors and environmental conditions that disrupt the protein's original state, including internal factors like mutations, expression levels, and polypeptide chain truncation, as well as external factors, such as dense molecular surroundings, post-translation modifications, and interactions with other proteins, nucleic acids, small molecules, metal ions, chaperones, and lipid membranes. During the aggregation process, the biological activity of an aggregating protein may be reduced or eliminated, whereas the resulting aggregates may have the potential to be immunogenic, or they may have other undesirable properties. Finding the cause(s) of protein aggregation and controlling it to an acceptable level is among the most crucial topics of research in academia and biopharmaceutical companies. This chapter aims to review intrinsic pathways of protein aggregation and potential extrinsic variables that influence this process.

Prediction of Protein Aggregation Propensity via Data-Driven Approaches

Protein aggregation occurs when misfolded or unfolded proteins physically bind together and can promote the development of various amyloid diseases. This study aimed to construct surrogate models for predicting protein aggregation via data-driven methods using two types of databases. First, an aggregation propensity score database was constructed by calculating the scores for protein structures in the Protein Data Bank using Aggrescan3D 2.0. Moreover, feature- and graph-based models for predicting protein aggregation have been developed by using this database. The graph-based model outperformed the feature-based model, resulting in an R2 of 0.95, although it intrinsically required protein structures. Second, for the experimental data, a feature-based model was built using the Curated Protein Aggregation Database 2.0 to predict the aggregated intensity curves. In summary, this study suggests approaches that are more effective in predicting protein aggregation, depending on the type of descriptor and the database.

Enhancing the thermostability of a mono- and diacylglycerol lipase from Malassizia globose by stabilizing a flexible loop in the catalytic pocket

A lipase from Malassizia globose, named SMG1, is highly desirable for industrial application due to its substrate specificity towards mono- and diacylglycerol. To improve its thermostability, we constructed a mutant library using an error-prone polymerase chain reaction, which was screened for both initial and residual enzymatic activity. Selected mutants were further studied using purified proteins for their kinetic thermostability at 45 ℃, T₅₀ (the temperature at which the enzyme loses half of its activity), and the optimal reaction temperature. Results showed that the majority of mutations with improved thermostability were on the protein surface. D245N and L270P showed the most significant thermostability enhancement with an approximately 3 ℃ increase in T₅₀ compared to wild-type (WT). In addition, combining these two mutations resulted in an increase of T₅₀ by 5 °C. Also, the optimal reaction temperatures of L270P and this double mutant are 10 ℃ higher than WT. The double mutant showed an approximately 100-fold increase in half-life at 45 ℃ and higher enzymatic activities at 30 ℃ and above compared to WT. High-temperature unfolding molecular dynamics simulation suggested that the double mutant stabilized a flexible loop in the catalytic pocket.

Effect of backbone circularization on colloidal stability: Compaction of unfolded structures improves aggregation resistance of granulocyte colony-stimulating factor

Aggregation of protein therapeutics can lead to immunogenicity and loss of function in vivo. Its effective prevention requires an understanding of the conformational and colloidal stability of protein and the improvement of both. Granulocyte colony-stimulating factor (G-CSF), which is one of the most widely used protein therapeutics, was previously shown to be conformationally stabilized by connecting its N- and C-termini with amide bonds (backbone circularization). In this study, we investigated whether circularization affects the colloidal stability of proteins. Colloidal stability was indirectly assessed by analyzing the aggregation behavior of G-CSF variants using analytical ultracentrifugation (AUC) and small-angle X-ray scattering (SAXS). Consequently, we found that the unfolded structure of circularized G-CSF was more compact than non-circularized G-CSF, and that backbone circularization improved its aggregation resistance against chemical denaturation by guanidine hydrochloride (GdnHCl). The improved aggregation resistance suggests that the expansion tolerance of circularized G-CSF in the unfolded state increased its colloidal stability. Thus, backbone circularization is an excellent method for enhancing the colloidal and the conformational stability of protein with minimal sequence changes. It is therefore expected to be effective in extending the storage stability of protein therapeutics, enhancing their biological stability.

Pharmaceutical modulation of the proteolytic profile of Transforming Growth Factor Beta induced protein (TGFBIp) offers a new avenue for treatment of TGFBI-corneal dystrophy

•

Corneal stromal dystrophies are a group of hereditary disorders caused by mutations in the TGFBI gene and affect the corneal stroma and epithelium.

•

The disease is characterized by the accumulation of insoluble deposits of the mutant TGFBIp leading to poor visual acuity in patients.

•

Mutations are hypothesized to disrupt the protein folding and stability, leading oligomerization of the mutant protein.

•

Current treatment relies on surgical intervention, either tissue removal or substitution, both of which are associated with disease recurrence.

•

The lead compounds reported here prevent/delay the atypical proteolysis of the mutant protein and the generation of amyloidogenic fragments.

Corneal dystrophies are a group of genetically inherited disorders with mutations in the TGFBI gene affecting the Bowman’s membrane and the corneal stroma. The mutant TGFBIp is highly aggregation-prone and is deposited in the cornea. Depending on the type of mutation the protein deposits may vary (amyloid, amorphous powdery aggregate or a mixed form of both), making the cornea opaque and thereby decreases visual acuity. The aggregation of the mutant protein is found to be specific with a unique aggregation mechanism distinct to the cornea. The proteolytic processing of the mutant protein is reported to be different compared to the WT protein. The proteolytic processing of mutant protein gives rise to highly amyloidogenic peptide fragments. The current treatment option, available for patients, is tissue replacement surgery that is associated with high recurrence rates. The clinical need for a simple treatment option for corneal dystrophy patients has become highly essential either to prevent the protein aggregation or to dissolve the preformed aggregates. Here, we report the screening of 2500 compounds from the Maybridge RO3 fragment library using weak affinity chromatography (WAC). The primary hits from WAC were validated by ¹⁵N-HSQC NMR assays and specific regions of binding were identified. The recombinant mutant proteins (4th FAS-1 domain of R555W and H572R) were subjected to limited proteolysis by trypsin together with the lead compounds identified by NMR assays. The lead compounds (MO07617, RJF00203 and, BTB05094) were effective to delay/prevent the generation of amyloidogenic peptides in the R555W mutant and compounds (RJF00203 and BTB05094) were effective to delay/prevent the generation of amyloidogenic peptides in the H572R mutant. Thus the lead compounds reported here upon further validation and/or modification might be proposed as a potential treatment option to prevent/delay aggregation by inhibiting the formation of amyloidogenic peptides in TGFBI-corneal dystrophy.

t-Butanol Enables Dual Functionality of Mannitol: A Cryoprotectant in Frozen Systems and Bulking Agent in Freeze-Dried Formulations

The effect of tertiary butyl alcohol (TBA) as a cosolvent on the phase behavior of mannitol in frozen and freeze-dried systems was characterized using differential scanning calorimetry (DSC) and X-ray diffractometry (XRD; laboratory and synchrotron sources). Solutions of mannitol (2 and 5% w/w) in TBA-water systems of different compositions (5 to 30% w/w TBA) were characterized, both during cooling and warming using DSC and XRD. At and below the TBA-water eutectic composition (22.5% w/w TBA), mannitol crystallization was completely inhibited in the frozen state, while it crystallized as anhydrous δ-mannitol in the final lyophile. The presence of mannitol did not affect the phase behavior of TBA. The ability of mannitol to serve as a cryoprotectant in frozen solutions, and as a bulking agent in final lyophile was evaluated using human serum albumin (HSA) as a model protein. When HSA in a TBA (5% w/w)-water solution containing mannitol (2% w/w) was freeze-thawed or freeze-dried, there was no evidence of HSA aggregation. Thus, when TBA was used as a cosolvent, mannitol exhibited dual functionality, serving as a cryoprotectant in frozen solutions and as a bulking agent in the final lyophile.

Extending the Detection Limit in Fragment Screening of Proteins Using Reverse Micelle Encapsulation

Detection of very weak (K_d > 10 mM) interactions of proteins with small molecules has been elusive. This is particularly important for fragment-based drug discovery, where it is suspected that the majority of potentially useful fragments will be invisible to current screening methodologies. We describe an NMR approach that permits detection of protein-fragment interactions in the very low affinity range and extends the current detection limit of ∼10 mM up to ∼200 mM and beyond. Reverse micelle encapsulation is leveraged to effectively reach very high fragment and protein concentrations, a principle that is validated by binding model fragments to E. coli dihydrofolate reductase. The method is illustrated by target-detected screening of a small polar fragment library against interleukin-1β, which lacks a known ligand-binding pocket. Evaluation of binding by titration and structural context allows for validation of observed hits using rigorous structural and statistical criteria. The 21 curated hit molecules represent a remarkable hit rate of nearly 10% of the library. Analysis shows that fragment binding involves residues comprising two-thirds of the protein's surface. Current fragment screening methods rely on detection of relatively tight binding to ligand binding pockets. The method presented here illustrates a potential to faithfully discover starting points for development of small molecules that bind to a desired region of the protein, even if the targeted region is defined by a relatively flat surface.

Retracted:Molecular characterization of excipients’ preferential interactions with therapeutic monoclonal antibodies

Retraction: Molecular characterization of excipients’ preferential interactions with therapeutic monoclonal antibodies by Jehoon Kim, Mark R. H. Krebs and Bernhardt L. Trout

The above article from the Journal of Pharmacy and Pharmacology, first published online on 4 August 2017 in Wiley Online Library (wileyonlinelibrary.com), has been retracted by agreement between the authors, the journal Editor-in-Chief, Professor David Jones, and John Wiley & Sons Ltd. The authors discovered that the analysis of simulations was faulty making the data incorrect.

Reference

Kim J et al. Molecular characterization of excipients’ preferential interactions with therapeutic monoclonal antibodies. J Pharm Pharmacol 2017. https://doi.org/10.1111/jphp.12787

Mercury(II) binds to both of chymotrypsin's histidines, causing inhibition followed by irreversible denaturation/aggregation

The toxicity of mercury is often attributed to its tight binding to cysteine thiolate anions in vital enzymes. To test our hypothesis that Hg(II) binding to histidine could be a significant factor in mercury's toxic effects, we studied the enzyme chymotrypsin, which lacks free cysteine thiols; we found that chymotrypsin is not only inhibited, but also denatured by Hg(II). We followed the aggregation of denatured enzyme by the increase in visible absorbance due to light scattering. Hg(II)-induced chymotrypsin precipitation increased dramatically above pH 6.5, and free imidazole inhibited this precipitation, implicating histidine-Hg(II) binding in the process of chymotrypsin denaturation/aggregation. Diethylpyrocarbonate (DEPC) blocked chymotrypsin's two histidines (his40 and his57 ) quickly and completely, with an IC50 of 35 ± 6 µM. DEPC at 350 µM reduced the hydrolytic activity of chymotrypsin by 90%, suggesting that low concentrations of DEPC react with his57 at the active site catalytic triad; furthermore, DEPC below 400 µM enhanced the Hg(II)-induced precipitation of chymotrypsin. We conclude that his57 reacts readily with DEPC, causing enzyme inhibition and enhancement of Hg(II)-induced aggregation. Above 500 µM, DEPC inhibited Hg(II)-induced precipitation, and [DEPC] >2.5 mM completely protected chymotrypsin against precipitation. This suggests that his40 reacts less readily with DEPC, and that chymotrypsin denaturation is caused by Hg(II) binding specifically to the his40 residue. Finally, we show that Hg(II)-histidine binding may trigger hemoglobin aggregation as well. Because of results with these two enzymes, we suggest that metal-histidine binding may be key to understanding all heavy metal-induced protein aggregation.

Investigation into alternative testing methodologies for characterization of influenza virus vaccine

The objective of this study was to explore various testing methodologies suitable for characterizing sedimented or agglomerated material. To model this, bioCSL's split influenza virus vaccine, Fluvax® was utilized. The investigation was conducted on 5 dispensed lots of commercially manufactured vaccine, formulated for the 2013 Southern Hemisphere season. Vaccine syringes were initially inspected by visual tests; the material was then aseptically pooled for characterization assessment by microscopy and several agglomeration assays. All syringes passed bioCSL's description test where any fine or large sized particles of sediment observed in the vaccine were resuspended upon shaking; inverted light microscopy verified that the sediment morphology was consistent with influenza vaccine. Electron microscopic examination of pooled vaccine material demonstrated the presence of typical influenza structures including split virus, virosomes, whole virus particles and agglomerates. An optical density turbidity assay revealed relatively high protein recoveries in the vaccine supernatant post-centrifugation treatment, thus indicative of a well-dispersed vaccine formulation. This was corroborated by particle sizing analysis using dynamic light scattering which generated reproducible volume particle size distributions of a polydisperse nature. Ultraviolet-visible absorbance profiles further confirmed the presence of some agglomerated material. Data from all methods demonstrated consistent results between all batches of vaccine. Therefore, this investigation revealed the suitability and usefulness of the various methodologies in characterizing the appearance of agglomerated vaccine material. It is suggested that such methods may be applicable and beneficial for the development of a wider spectrum of heterogeneous and agglomerated formulations to provide safe, efficacious and superior quality biopharmaceutical products.

Lysine-Based Site-Directed Mutagenesis Increased Rigid β-Sheet Structure and Thermostability of Mesophilic 1,3-1,4-β-Glucanase

1,3-1,4-β-Glucanase is widely applied in the food industry, while its low thermostability often reduces its performance. In a previous study, chemical modification of surface lysine residues was proved to increase the thermostability of β-glucanase. To improve the thermostability, the mesophilic β-glucanase from Bacillus terquilensis was rationally engineered through site-directed mutagenesis of the 12 lysines into serines. The results showed that the K20S, K117S, and K165S mutants could both enhance the specific activities and thermostability of β-glucanase. The triple mutant (K20S/K117S/K165S) could increase the optimal temperature and T50 value by 15 and 14 °C, respectively. Five percent more structured residues were observed in the mutant, which formed new β-sheet structures in the concave side. Molecular dynamics simulation analysis showed that the flexibility in the mutation regions was decreased, which resulted in the overall rigidity of the β-glucanase. Therefore, the lysine-based site-directed mutagenesis is a simple and effective method for improving the thermostability of β-glucanase.

Altered backbone and side-chain interactions result in route heterogeneity during the folding of interleukin-1β (IL-1β)

Deletion of the β-bulge trigger-loop results in both a switch in the preferred folding route, from the functional loop packing folding route to barrel closure, as well as conversion of the agonist activity of IL-1β into antagonist activity. Conversely, circular permutations of IL-1β conserve the functional folding route as well as the agonist activity. These two extremes in the folding-functional interplay beg the question of whether mutations in IL-1β would result in changes in the populations of heterogeneous folding routes and the signaling activity. A series of topologically equivalent water-mediated β-strand bridging interactions within the pseudosymmetric β-trefoil fold of IL-1β highlight the backbone water interactions that stabilize the secondary and tertiary structure of the protein. Additionally, conserved aromatic residues lining the central cavity appear to be essential for both stability and folding. Here, we probe these protein backbone-water molecule and side chain-side chain interactions and the role they play in the folding mechanism of this geometrically stressed molecule. We used folding simulations with structure-based models, as well as a series of folding kinetic experiments to examine the effects of the F42W core mutation on the folding landscape of IL-1β. This mutation alters water-mediated backbone interactions essential for maintaining the trefoil fold. Our results clearly indicate that this perturbation in the primary structure alters a structural water interaction and consequently modulates the population of folding routes accessed during folding and signaling activity.

Competing aggregation pathways for monoclonal antibodies

Aggregation is mediated by local unfolding to allow aggregation "hot spot(s)" to become solvent exposed and available to associate with a hot spot on another partially unfolded protein. Historically, the unfolding of either the crystallizable fragment (Fc) or the antigen binding fragment (Fab) regions of a given monoclonal antibody (MAb) has been implicated in aggregation, with differing results across different proteins. The present work focuses on separately quantifying the aggregation kinetics of isolated Fc, isolated Fab, and intact MAb as a function of pH under accelerated (high temperature) conditions. The results show that both Fab and Fc are aggregation prone and compete within the same MAb.

Stability engineering of the human antibody repertoire

Human monoclonal antibodies often display limited thermodynamic and colloidal stabilities. This behavior hinders their production, and places limitations on the development of novel formulation conditions and therapeutic applications. Antibodies are highly diverse molecules, with much of the sequence variation observed within variable domain families and, in particular, their complementarity determining regions. This has complicated the development of comprehensive strategies for the stability engineering of the human antibody repertoire. Here we provide an overview of the field, and discuss recent advances in the development of robust and aggregation resistant antibody therapeutics.

Investigating monoclonal antibody aggregation using a combination of H/DX-MS and other biophysical measurements

To determine how structural changes in antibodies are connected with aggregation, the structural areas of an antibody prone to and/or impacted by aggregation must be identified. In this work, the higher-order structure and biophysical properties of two different monoclonal antibody (mAb) monomers were compared with their simplest aggregated form, that is, dimers that naturally occurred during normal production and storage conditions. A combination of hydrogen/deuterium exchange mass spectrometry and other biophysical measurements was used to make the comparison. The results show that the dimerization process for one of the mAb monomers (mAb1) displayed no differences in its deuterium uptake between monomer and dimer forms. However, the other mAb monomer (mAb2) showed subtle changes in hydrogen/deuterium exchange as compared with its dimer form. In this case, differences observed were located in specific functional regions of the CH 2 domain and the hinge region between CH 1 and CH 2 domains. The importance and the implications of these changes on the antibody structure and mechanism of aggregation are discussed.

Early folding events protect aggregation-prone regions of a β-rich protein

Protein folding and aggregation inevitably compete with one another. This competition is even keener for proteins with frustrated landscapes, such as those rich in β structure. It is interesting that, despite their rugged energy landscapes and high β sheet content, intracellular lipid-binding proteins (iLBPs) appear to successfully avoid aggregation, as they are not implicated in aggregation diseases. In this study, we used a canonical iLBP, cellular retinoic acid-binding protein 1 (CRABP1), to understand better how folding is favored over aggregation. Analysis of folding kinetics of point mutants reveals that the folding pathway of CRABP1 involves early barrel closure. This folding mechanism protects sequences in CRABP1 that comprise cores of aggregates as identified by nuclear magnetic resonance. The amino acid conservation pattern in other iLBPs suggests that early barrel closure may be a general strategy for successful folding and minimization of aggregation. We suggest that folding mechanisms in general may incorporate steps that disfavor aggregation.

Early aggregated States in the folding of interleukin-1β

Kinetic data measured from folding of the protein interleukin-1β fits best to three exponential phases when studied with tryptophan fluorescence but only two exponential phases when measured using other methods. The technique of ANS fluorescence was used to determine whether the additional phase observed in tryptophan fluorescence was also detected with ANS dye binding. Unlike trytophan fluorescence, the ANS fluorescence was highly dependent on the concentration of protein present during the folding experiment. Experimental controls provide evidence that ANS binds to protein aggregates, present at higher concentrations and absent at lower concentrations. Protein concentration-dependent folding studies demonstrate that, at lower interleukin-1β concentrations, tryptophan fluorescence kinetics can be fit adequately with a two exponential fit. This study indicates that (1) measured interleukin-1β folding kinetics fit to a 2 phase model and (2) at higher protein concentrations, transient association of IL-1β may result in a kinetic fit of 3 phases.

The unique cysteine knot regulates the pleotropic hormone leptin

Leptin plays a key role in regulating energy intake/expenditure, metabolism and hypertension. It folds into a four-helix bundle that binds to the extracellular receptor to initiate signaling. Our work on leptin revealed a hidden complexity in the formation of a previously un-described, cysteine-knotted topology in leptin. We hypothesized that this unique topology could offer new mechanisms in regulating the protein activity. A combination of in silico simulation and in vitro experiments was used to probe the role of the knotted topology introduced by the disulphide-bridge on leptin folding and function. Our results surprisingly show that the free energy landscape is conserved between knotted and unknotted protein, however the additional complexity added by the knot formation is structurally important. Native state analyses led to the discovery that the disulphide-bond plays an important role in receptor binding and thus mediate biological activity by local motions on distal receptor-binding sites, far removed from the disulphide-bridge. Thus, the disulphide-bridge appears to function as a point of tension that allows dissipation of stress at a distance in leptin.

Reconciling theories of chaperonin accelerated folding with experimental evidence

For the last 20 years, a large volume of experimental and theoretical work has been undertaken to understand how chaperones like GroEL can assist protein folding in the cell. The most accepted explanation appears to be the simplest: GroEL, like most other chaperones, helps proteins fold by preventing aggregation. However, evidence suggests that, under some conditions, GroEL can play a more active role by accelerating protein folding. A large number of models have been proposed to explain how this could occur. Focused experiments have been designed and carried out using different protein substrates with conclusions that support many different mechanisms. In the current article, we attempt to see the forest through the trees. We review all suggested mechanisms for chaperonin-mediated folding and weigh the plausibility of each in light of what we now know about the most stringent, essential, GroEL-dependent protein substrates.

Non-native protein aggregation kinetics

Experimental kinetics of non-native protein aggregation are of practical importance in that they help dictate viable processing, formulation, and storage conditions for biotechnology products, and appear to play a role in determining the onset of a number of diseases. Fundamentally, aggregation kinetics provide insights into the identity of key intermediates in the process, and quantitative tests of available models of aggregation. Although aggregation kinetics often display seemingly disparate behaviors across different proteins and sample conditions, this review illustrates how many of these can be understood within a general framework that treats aggregation as a multi-stage process, and how most available kinetic models of aggregation can be grouped hierarchically in terms of which stage(s) they include. This provides an aid for workers seeking a mechanistic interpretation of in vitro aggregation kinetics, for discriminating among competing models, and in designing experiments to assess in vitro protein stability. Limitations and the utility of purely kinetic approaches to studying aggregation, clarifications of common misperceptions regarding experimental aggregation kinetics, and some outstanding challenges in the field are briefly discussed.

Solute effects on the irreversible aggregation of serum albumin

Thermal stress on bovine serum albumin (BSA) promotes protein aggregation through the formation of intermolecular beta-sheets. We have used light scattering and chromatography to study effects of (<1 M) Na(2)SO(4), NaSCN, sucrose, sorbitol and urea on the rate of the thermal aggregation. Both salts were strong inhibitors of BSA aggregation and they reduced both the size and number (concentration) of aggregate particles compared to non-ionic solutes (or pure buffer). Hence, the salts appear to suppress both nucleation- and growth rate. The non-electrolyte additives reduced the initial aggregation rate (compared to pure buffer), but did not significantly limit the extent of aggregation in samples quenched after 27 min. heat exposure (40-50% aggregation in all samples). The non-electrolytes did, however, modify the aggregation process as they consistently brought about smaller but more concentrated aggregates than pure buffer. The results are discussed along the lines of linkage- and transition state theories. In this framework, the rate of the aggregation process is governed by the equilibrium between a thermally denatured state (D) and the transition state D( not equal). Thus, the effect of a solute relies on its preferential interactions with respectively D and D( not equal). The current results do not show any correlation between the solutes' preferential interactions with native BSA and their effect on the rate of aggregation. This suggests that non-specific, "Hofmeister-type" interactions, which scale with the solvent accessible surface area, are of minor importance. Rather, salt induced suppression of aggregation is suggested to depend on the modulation of specific electrostatic forces in the D( not equal) state.

A Lumry-Eyring nucleated polymerization model of protein aggregation kinetics: 1. Aggregation with pre-equilibrated unfolding

A mathematical model is presented of the kinetics of non-native protein aggregation that combines Lumry-Eyring and nucleated polymerization (LENP) descriptions. The LENP model is solved for cases in which aggregation rates are slow compared to folding-unfolding equilibration and is shown to be a generalization of a number of previously proposed nucleation-and-growth models for non-native and native protein aggregation. The model solutions exhibit a number of qualitative kinetic regimes. Each regime has a characteristic set of experimental signatures that are related to the relative rates of growth and nucleation as well as to the threshold size at which aggregates condense to form higher-order structures or other phases. Approximate model solutions provide practical rate equations that can be regressed against typical experimental kinetic data to obtain mechanistic parameters characterizing the aggregation pathway. In all kinetic regimes, it is found that observed rate coefficients (kobs) or half-lives (t50) obtained from extent-of-reaction measurements are convolutions of more than one stage in the pathway unless purely seeded growth occurs. Despite this convolution, the combination of apparent reaction order (time domain) and the scaling of kobs or t50 with initial protein concentration provides a means to determine a value for the dominant nucleus size in each case. Additional information, such as equilibrium unfolding thermodynamics and the limiting aggregate size distribution, are required to further deconvolute kobs into intrinsic contributions from nucleation, growth, and conformational changes. The model and analysis are expected to be generally applicable to a wide range of proteins and polypeptides that form non-native aggregates.

Polyvinylpyrrolidone 40 Assists the Refolding of Bovine Carbonic Anhydrase B by Accelerating the Refolding of the First Molten Globule Intermediate

Protecting proteins from aggregation is one of the most important issues in both protein science and protein engineering. In this research, the mechanism of enhancing the refolding of guanidine hydrochloride-denatured carbonic anhydrase B by polyvinylpyrrolidone 40 (PVP40) was studied by both kinetic and equilibrium refolding experiments. The reactivation and refolding kinetics indicated that the rate constant of refolding the first refolding intermediate (I(1)) to the second one (I(2)) is promoted by the addition of PVP. Fluorescence quenching studies further indicated that PVP could bind to the aggregation-prone species I(1), resulting in the protection of the exposed hydrophobic surface, a minimization of the protein surface, and more importantly, an increase of the refolding rate of I(1). These properties were quite different from those of poly(ethylene glycol) (PEG), which has been shown to have a strong and stoichiometric binding to I(1) and does not interfere with the refolding pathway. Unlike PEG, the binding of PVP to I(1) does not block the aggregation pathway directly but decreases the energy barrier for I(1) to refold to I(2) and thus reduces the accumulation of I(1). These results suggested that PVP works by a quite different mechanism from those well established ones in chaperones and chemical promoters. PVP is more like a folding catalyst rather than a chemical chaperone. The distinct mechanism of enhancing protein aggregation by PVP is expected to facilitate the attempt to develop new chemical compounds as well as new strategies to protect proteins from aggregation.

A support vector machine-based method for predicting the propensity of a protein to be soluble or to form inclusion body on overexpression in Escherichia coli

MOTIVATION

Inclusion body formation has been a major deterrent for overexpression studies since a large number of proteins form insoluble inclusion bodies when overexpressed in Escherichia coli. The formation of inclusion bodies is known to be an outcome of improper protein folding; thus the composition and arrangement of amino acids in the proteins would be a major influencing factor in deciding its aggregation propensity. There is a significant need for a prediction algorithm that would enable the rational identification of both mutants and also the ideal protein candidates for mutations that would confer higher solubility-on-overexpression instead of the presently used trial-and-error procedures.

RESULTS

Six physicochemical properties together with residue and dipeptide-compositions have been used to develop a support vector machine-based classifier to predict the overexpression status in E.coli. The prediction accuracy is approximately 72% suggesting that it performs reasonably well in predicting the propensity of a protein to be soluble or to form inclusion bodies. The algorithm could also correctly predict the change in solubility for most of the point mutations reported in literature. This algorithm can be a useful tool in screening protein libraries to identify soluble variants of proteins.

Aggregation of granulocyte-colony stimulating factor in vitro involves a conformationally altered monomeric state

Aggregation of partially folded intermediates populated during protein folding processes has been described for many proteins. Likewise, partially unfolded chains, generated by perturbation of numerous proteins by heat or chemical denaturants, have also been shown to aggregate readily. However, the process of protein aggregation from native-state conditions is less well understood. Granulocyte-colony stimulating factor (G-CSF), a member of the four-helix bundle class of cytokines, is a therapeutically relevant protein involved in stimulating the growth and maturation of phagocytotic white blood cells. Under native-like conditions (37 degrees C [pH 7.0]), G-CSF shows a significant propensity to aggregate. Our data suggest that under these conditions, native G-CSF exists in equilibrium with an altered conformation, which is highly aggregation prone. This species is enriched in 1-2 M GdmCl, as determined by tryptophan fluorescence and increased aggregation kinetics. In particular, specific changes in Trp58 fluorescence report a local rearrangement in the large loop region between helices A and B. However, circular dichroism, reactivity toward cyanylation, and ANS binding demonstrate that this conformational change is subtle, having no substantial disruption of secondary and tertiary structure, reactivity of the free sulfhydryl at Cys17 or exposure of buried hydrophobic regions. There is no indication that this altered conformation is important to biological activity, making it an attractive target for rational protein stabilization.

The native energy landscape for interleukin-1beta. Modulation of the population ensemble through native-state topology

A minimalist Go-model, with no energetic frustration in the native conformation, has been shown to describe accurately the folding pathway of the beta-trefoil protein, interleukin-1beta (IL-1beta). While it appears that these models successfully model transition states and intermediates between the unfolded and native ensembles, it is unclear how accurately they capture smaller, yet biologically relevant, structural changes within the native ensemble after energetic perturbation. Here, we address the following questions. Can a simple Go-model of interleukin-1beta, based on native topology, describe changes in structural properties of the native ensemble as the protein stability is changed? Or is it necessary to include a more explicit representation of atoms, electrostatic, hydrogen bonding, and van der Waals forces to describe these changes? The native ensemble of IL-1beta was characterized using a variety of experimental probes under native (0 M NaCl, guanidine hydrochloride (Gdn-HCl)), moderately destabilized (0 M NaCl, 0.8 M Gdn-HCl), and in moderate salt concentration (0.8 M NaCl, 0 M Gdn-HCl). Heteronuclear (1)H-(15)N nuclear Overhauser effect spectroscopy (NOESY) and heteronuclear single quantum correlation (HSQC) NMR spectra confirmed that the beta-trefoil global fold was largely intact under these three conditions. However, 25 of the 153 residues throughout the chain did demonstrate (13)C and (1)H-(15)N chemical shifts when perturbed with 0.8 M NaCl or Gdn-HCl. Despite large differences in protection factors from solvent hydrogen-deuterium exchange for all residues between stable (0 M Gdn-HCl) and destabilized (0.8 M Gdn-HCl) IL-1beta, no difference in steady-state (15)N-(1)H NOE enhancements were measured. Thus, the chemical shifts correlate with a global but limited increase in residue flexibility in the presence of Gdn-HCl. Minimalist simulations highlight the regions of greatest position shift between native and 0.8 M Gdn-HCl, which were determined experimentally. This correlation demonstrates that structural changes within the native ensemble of IL-1beta are, at least partially, governed by the principle of minimal energetic frustration.

The role of detergent in refolding of GdnHCl-denatured arginine kinase from shrimpFenneropenaeus Chinensis: the solubilization of aggregate and refolding in detergent solutions

Strong aggregation occurred in the refolding route of arginine kinase (AK) denatured with 3 mol GdnHCl/L (GdnHCl, guanidine hydrochloride). The activity recovery of GdnHCl-denatured AK was very low and dependent on the protein concentration in the process of refolding. For denatured AK at 1.2 µmol/L concentration, the recovered activity yield was about 45.2% of the native enzyme, whereas at 5.2 µmol/L the activity recovery yield was only 20% of native activity. The nonionic detergent Triton X-100 and Tween 20 (≤100 mmol/L concentration) not only effectively blocked the aggregation but also enabled the denatured AK to recover most of its native activity. The kinetics of aggregate solubilization showed that there was an induction phase dependent on the detergent, but there was no dependency when detergent was absent. The apparent activity recovery had a cooperative relation with detergents in the process of refolding, which suggested the existence of some interaction between the detergent and the refolding intermediate. On the basis of the study results, a scheme of refolding was proposed.Key words: arginine kinase, guanidine-denatured, refolding, detergent, aggregation.

Non-linear effects of temperature and urea on the thermodynamics and kinetics of folding and unfolding of hisactophilin

Extensive measurements and analysis of thermodynamic stability and kinetics of urea-induced unfolding and folding of hisactophilin are reported for 5-50 degrees C, at pH 6.7. Under these conditions hisactophilin has moderate thermodynamic stability, and equilibrium and kinetic data are well fit by a two-state transition between the native and the denatured states. Equilibrium and kinetic m values decrease with increasing temperature, and decrease with increasing denaturant concentration. The betaF values at different temperatures and urea concentrations are quite constant, however, at about 0.7. This suggests that the transition state for hisactophilin unfolding is native-like and changes little with changing solution conditions, consistent with a narrow free energy profile for the transition state. The activation enthalpy and entropy of unfolding are unusually low for hisactophilin, as is also the case for the corresponding equilibrium parameters. Conventional Arrhenius and Eyring plots for both folding and unfolding are markedly non-linear, but these plots become linear for constant DeltaG/T contours. The Gibbs free energy changes for structural changes in hisactophilin have a non-linear denaturant dependence that is comparable to non-linearities observed for many other proteins. These non-linearities can be fit for many proteins using a variation of the Tanford model, incorporating empirical quadratic denaturant dependencies for Gibbs free energies of transfer of amino acid constituents from water to urea, and changes in fractional solvent accessible surface area of protein constituents based on the known protein structures. Noteworthy exceptions that are not well fit include amyloidogenic proteins and large proteins, which may form intermediates. The model is easily implemented and should be widely applicable to analysis of urea-induced structural transitions in proteins.

Protein aggregation and its inhibition in biopharmaceutics

Protein aggregation is arguably the most common and troubling manifestation of protein instability, encountered in almost all stages of protein drug development. Protein aggregation, along with other physical and/or chemical instabilities of proteins, remains to be one of the major road barriers hindering rapid commercialization of potential protein drug candidates. Although a variety of methods have been used/designed to prevent/inhibit protein aggregation, the end results are often unsatisfactory for many proteins. The limited success is partly due to our lack of a clear understanding of the protein aggregation process. This article intends to discuss protein aggregation and its related mechanisms, methods characterizing protein aggregation, factors affecting protein aggregation, and possible venues in aggregation prevention/inhibition in various stages of protein drug development.

Protein misfolding: optional barriers, misfolded intermediates, and pathway heterogeneity

To investigate the character and role of misfolded intermediates in protein folding, a recombinant cytochrome c without the normally blocking histidine to heme misligation was studied. Folding remains heterogeneous as in the wild-type protein. Half of the population folds relatively rapidly to the native state in a two-state manner. The other half collapses (fluorescence quenching) and forms a full complement of helix (CD) with the same rate and denaturant dependence as the fast folding fraction but then is blocked and reaches the native structure (695nm absorbance) much more slowly. The factors that transiently block folding are not intrinsic to the folding process but depend on ambient conditions, including protein aggregation (f(concentration)), N terminus to heme misligation (f(pH)), and proline mis-isomerization (f(U state equilibration time)). The misfolded intermediate populated by the slowly folding fraction was characterized by hydrogen exchange pulse labeling. It is very advanced with all of the native-like elements fairly stably formed but not the final Met80-S to heme iron ligation, similar to a previously studied molten globule form induced by low pH. To complete final native state acquisition, some small back unfolding is required (error repair) but the misfolded intermediate does not revisit the U state before proceeding to N. These properties show that the intermediate is a normal on-pathway form that contains, in addition, adventitious misfolding errors that transiently block its forward progress. Related observations for other proteins (partially misfolded intermediates, pathway heterogeneity) might be similarly explained in terms of the optional insertion of error-dependent barriers into a classical folding pathway.

Rational engineering of enzyme stability

During the past 15 years there has been a continuous flow of reports describing proteins stabilized by the introduction of mutations. These reports span a period from pioneering rational design work on small enzymes such as T4 lysozyme and barnase to protein design, and directed evolution. Concomitantly, the purification and characterization of naturally occurring hyperstable proteins has added to our understanding of protein stability. Along the way, many strategies for rational protein stabilization have been proposed, some of which (e.g. entropic stabilization by introduction of prolines or disulfide bridges) have reasonable success rates. On the other hand, comparative studies and efforts in directed evolution have revealed that there are many mutational strategies that lead to high stability, some of which are not easy to define and rationalize. Recent developments in the field include increasing awareness of the importance of the protein surface for stability, as well as the notion that normally a very limited number of mutations can yield a large increase in stability. Another development concerns the notion that there is a fundamental difference between the "laboratory stability" of small pure proteins that unfold reversibly and completely at high temperatures and "industrial stability", which is usually governed by partial unfolding processes followed by some kind of irreversible inactivation process (e.g. aggregation). Provided that one has sufficient knowledge of the mechanism of thermal inactivation, successful and efficient rational stabilization of enzymes can be achieved.

Kinetics of heat- and acidification-induced aggregation of firefly luciferase

The general approach to analysis of the kinetics of protein aggregation registered by the turbidimetric method has been elaborated. The terminal part of the kinetic curves is analyzed using a theoretical equation connecting the derivative of the apparent absorbance (A) with respect to time (dA/dt) and A (t is time). This analysis allows the limiting value of A at t--> infinity (A(lim)) and the order of aggregation with respect to protein (n) to be calculated. Approach proposed was applied to analysis of thermal and acidification-induced aggregation of firefly luciferase. In both cases the A(lim) value is a linear function of the protein concentration. The terminal part of the kinetic curves of thermal aggregation follows the first-order kinetics (n=1), whereas the kinetics of acidification-induced aggregation are characterized by the value of n higher than unity (n=1.29). The mechanism of nucleation-dependent aggregation has been discussed.

A new kinetic scheme for lysozyme refolding and aggregation

The competing first- and third-order reaction scheme for lysozyme is shown to not predict fed-batch lysozyme refolding when the model is parameterized using independent batch experiments, even when variations in chemical composition during the fed-batch experiment are accounted for. A new kinetic scheme is proposed that involves rapid partitioning between the alternative fates of refolding and aggregation, and which allows for aggregation via a sequential mechanism. The model assumes that monomeric lysozyme in different states, including native, is able to aggregate with intermediates, accounting for recent experimental evidence that native protein can be incorporated into aggregates and explaining why native protein in the refolding buffer reduces yield. Stopped-flow light-scattering measurements were used to measure the association rate for the sequential aggregation mechanism, and refolding rate constants were determined in a series of batch experiments designed to be "snapshots" of the composition during a fed-batch experiment. The new kinetic scheme gave a good a priori prediction of fed-batch refolding performance.

Real-time NMR kinetic studies provide global and residue-specific information on the non-cooperative unfolding of the beta-trefoil protein, interleukin-1beta

The interleukin-1beta (IL-1beta) structural motif is a beta-trefoil super fold created by six two-stranded beta-hairpins. Turns are thus particularly important in creating the topology and the arrangement of beta-strands in this structural motif. In contrast to the signals observed in optical studies, real-time NMR kinetic investigations of the denaturant-induced unfolding of interleukin-1beta provide direct, global, and residue-specific information on the structural nature of the unfolding reaction. Heterogeneity in the individual amino acid residue kinetics reveals a rugged unfolding landscape. The relative kinetic stability of native-like turns supports low temperature molecular dynamics predictions of turn-controlled unfolding.

Selection of mutations for increased protein stability

There are many ways to select mutations that increase the stability of proteins, including rational design, functional screening of randomly generated mutant libraries, and comparison of naturally occurring homologous proteins. The protein engineer's toolbox is expanding and the number of successful examples of engineered protein stability is increasing. Still, the selection of thermostable mutations is not a standard process. Selection is complicated by lack of knowledge of the process that leads to thermal inactivation and by the fact that proteins employ a large variety of structural tricks to achieve stability.

Biology of the prion gene complex

The prion protein gene Prnp encodes PrPSc, the major structural component of prions, infectious pathogens causing a number of disorders including scrapie and bovine spongiform encephalopathy (BSE). Missense mutations in the human Prnp gene, PRNP, cause inherited prion diseases such as familial Creutzfeldt-Jakob Disease. In uninfected animals, Prnp encodes a GPI-anchored protein denoted PrPC, and in prion infections, PrPC is converted to PrPSc by templated refolding. Although Prnp is conserved in mammalian species, attempts to verify interactions of putative PrP-binding proteins by genetic means have proven frustrating in that two independent lines of Prnp gene ablated mice (Prnp0/0 mice: ZrchI and Npu) lacking PrPC remain healthy throughout development. This indicates that PrPC serves a function that is not apparent in a laboratory setting or that other molecules have overlapping functions. Shuttling or sequestration of synaptic Cu(II) via binding to N-terminal octapeptide residues and (or) signal transduction involving the fyn kinase are possibilities currently under consideration. A new point of entry into the issue of prion protein function has emerged from identification of a paralog, Prnd, with 25% coding sequence identity to Prnp. Prnd lies downstream of Prnp and encodes the Dpl protein. Like PrPC, Dpl is presented on the cell surface via a GPI anchor and has three alpha-helices: however, it lacks the conformationally plastic and octapeptide repeat domains present in its well-known relative. Interestingly, Dpl is overexpressed in two other lines of Prnp0/0 mice (Ngsk and Rcm0) via intergenic splicing events. These lines of Prnp0/0 mice exhibit ataxia and apoptosis of cerebellar cells, indicating that ectopic synthesis of Dpl protein is toxic to CNS neurons: this inference has now been confirmed by the construction of transgenic mice expressing Dpl under the direct control of the PrP promoter. Remarkably, Dpl-programmed ataxia is rescued by wt Prnp transgenes. The interaction between the Prnp and Prnd genes in mouse cerebellar neurons may have a physical correlate in competition between Dpl and PrPC within a common biochemical pathway that, when misregulated, leads to apoptosis.

Conserved and nonconserved features of the folding pathway of hisactophilin, a beta-trefoil protein

Based on previous studies of interleukin-1beta (IL-1beta) and both acidic and basic fibroblast growth factors (FGFs), it has been suggested that the folding of beta-trefoil proteins is intrinsically slow and may occur via the formation of essential intermediates. Using optical and NMR-detected quenched-flow hydrogen/deuterium exchange methods, we have measured the folding kinetics of hisactophilin, another beta-trefoil protein that has < 10% sequence identity and unrelated function to IL-1beta and FGFs. We find that hisactophilin can fold rapidly and with apparently two-state kinetics, except under the most stabilizing conditions investigated where there is evidence for formation of a folding intermediate. The hisactophilin intermediate has significant structural similarities to the IL-1beta intermediate that has been observed experimentally and predicted theoretically using a simple, topology-based folding model; however, it appears to be different from the folding intermediate observed experimentally for acidic FGF. For hisactophilin and acidic FGF, intermediates are much less prominent during folding than for IL-1beta. Considering the structures of the different beta-trefoil proteins, it appears that differences in nonconserved loops and hydrophobic interactions may play an important role in differential stabilization of the intermediates for these proteins.

Three topologically equivalent core residues affect the transition state ensemble in a protein folding reaction

The single domain protein, interleukin-1beta, is representative of a distinct class of proteins characterized by their beta-trefoil topology. Each subdomain of this structural class is composed of a beta beta beta loop beta (betabetabetaLbeta) motif comprised of approximately 50 residues and gives the protein a pseudo- 3-fold axis of symmetry. A common feature of proteins in this topological family appears to be that they are slow folders, which reach the native state on the order of tens to 100s of seconds. Sequence analysis of interleukin-1beta indicates that three phenylalanine residues located at positions 42, 101, and 146 are well conserved, separated by approximately 50 residues in the primary sequence, located in similar positions in the pseudo-symmetric units of the trefoil, and are juxtaposed to one another in conformational space. These residues surround the hydrophobic cavity and "pin" the hairpin triplet cap to the core beta-barrel. To determine if cap-barrel interactions are involved in maintaining the structural stability and cooperativity or in controlling the slow formation of the native state, we performed a series of mutational studies. The results indicate that interleukin-1beta tolerates large increases in side-chain volume at these three topologically conserved sites with little effect on stability, while the kinetics show significant differences in both the unfolding and refolding rates. Taken together, our results indicate that these conserved core residues are essential contacts in the transition-state ensemble for folding.

Cofactor and tryptophan accessibility and unfolding of brain glutamate decarboxylase

Cofactor and tryptophan accessibility of the 65-kDa form of rat brain glutamate decarboxylase (GAD) was investigated by fluorescence quenching measurements using acrylamide, I-, and Cs+ as the quenchers. Trp residues were partially exposed to solvent. I- was less able and Cs+ was more able to quench the fluorescence of Trp residues in the holoenzyme of GAD (holoGAD) than the apoenzyme (apoGAD). The fraction of exposed Trp residues were in the range of 30-49%. In contrast, pyridoxal-P bound to the active site of GAD was exposed to solvent. I- was more able and Cs+ was less able to quench the fluorescence of pyridoxal-P in holoGAD. The cofactor was present in a positively charged microenvironment, making it accessible for interactions with anions. A difference in the exposure of Trp residues and pyridoxal-P to these charged quenchers suggested that the exposed Trp residues were essentially located outside of the active site. Changes in the accessibility of Trp residues upon pyridoxal-P binding strongly supported a significant conformational change in GAD. Fluorescence intensity measurements were also carried out to investigate the unfolding of GAD using guanidine hydrochloride (GdnHCl) as the denaturant. At 0.8-1.5 M GdnHCl, an intermediate step was observed during the unfolding of GAD from the native to the denatured state, and was not found during the refolding of GAD from the denatured to native state, indicating that this intermediate step was not a reversible process. However, at >1.5 M GdnHCl for holoGAD and >2.0 M GdnHCl for apoGAD, the transition leading to the denatured state was reversible. It was suggested that the intermediate step involved the dissociation of native dimer of GAD into monomers and the change in the secondary structure of the protein. Circular dichroism revealed a decrease in the alpha-helix content of GAD from 36 to 28%. The unfolding pattern suggested that GAD may consist of at least two unfolding domains. Unfolding of the lower GdnHCl-resisting domain occurred at a similar concentration of denaturant for apoGAD and holoGAD, while unfolding of the higher GdnHCl-resisting domain occurred at a higher concentration of GdnHCl for apoGAD than holoGAD.

Reversible formation of on-pathway macroscopic aggregates during the folding of maltose binding protein

Maltose binding protein (MBP) is widely used as a model for protein folding and export studies. We show here that macroscopic aggregates form transiently during the refolding of MBP at micromolar protein concentrations. Disaggregation occurs spontaneously without any aid, and the refolded material has structure and activity identical to those of the native, nondenatured protein. A considerable fraction of protein undergoing folding partitions into the aggregate phase and can be manually separated from the soluble phase by centrifugation. The separated MBP precipitate can be resolubilized and yields active, refolded protein. This demonstrates that both the soluble and aggregate phases contribute to the final yield of refolded protein. SecB, the cognate Escherichia coli cytosolic chaperone in vivo for MBP, reduces but does not entirely prevent aggregation, whereas GroEL and a variety of other control proteins have no effect. Kinetic studies using a variety of spectroscopic probes show that aggregation occurs through a collapsed intermediate with some secondary structure. The aggregate formed during refolding can convert directly to a near native state without going through the unfolded state. Further, optical and electron microscopic studies indicate that the MBP precipitate is not an amyloid.

Core and surface mutations affect folding kinetics, stability and cooperativity in IL-1 beta: does alteration in buried water play a role?

Interleukin-1 beta (IL-1 beta) is a cytokine and a member of the beta-trefoil superfamily of protein structures. An interesting feature in the folding of IL-1 beta, shared with some other members of the same topological family, is the existence of a slow step in folding to the native conformation from a discrete intermediate. Wanting to probe the nature of this slow step in the folding of WT IL-1 beta (tau(1)=45 seconds), we made ten sequence variants of IL-1 beta (L10A, T9Q, T9G, C8S, C8A, N7G, N7D, L6A, R4P, and R4Q), where all mutations are located along strand 1. This strand is not protected from hydrogen exchange until late in folding. Most of the mutations showed little effect on the kinetics of folding for IL-1 beta. However, C8 is clearly involved in both the late and the early steps in folding, while sequence variants at L10 and L6 affect only late events in folding. The value of the slowest relaxation time, tau(1), which is associated with the rate of native protein formation, increased for the refolding of C8S, while C8A, L6A, and L10A showed smaller but systematic increases in the value of tau(1.)For both C8S and C8A, the value of the step associated with formation of the intermediate, tau(2), was independent of denaturant concentration. In addition, mutations in the hydrophobic core (L10A, C8A, C8S, and L6A) and, surprisingly, along the surface (T9G, T9Q, and N7G) alter the stability. The most destabilizing mutations show changes in equilibrium unfolding cooperativity, which is atypical for destabilizing mutations in IL-1 beta. Crystallographic studies indicate that mutations along strand 1 may alter the number of ordered water molecules within the core. Thus, side-chain replacement in this region can disrupt essential main-chain interactions mediated by ordered water contacts in a highly cooperative network of hydrogen bonding.

Structural events during the refolding of an all beta-sheet protein

The refolding kinetics of the 140-residue, all beta-sheet, human fibroblast growth factor (hFGF-1) is studied using a variety of biophysical techniques such as stopped-flow fluorescence, stopped-flow circular dichroism, and quenched-flow hydrogen exchange in conjunction with multidimensional NMR spectroscopy. Urea-induced unfolding of hFGF-1 under equilibrium conditions reveals that the protein folds via a two-state (native <--> unfolded) mechanism without the accumulation of stable intermediates. However, measurement of the unfolding and refolding rates in various concentrations of urea shows that the refolding of hFGF-1 proceeds through accumulation of kinetic intermediates. Results of the quenched-flow hydrogen exchange experiments reveal that the hydrogen bonds linking the N- and C-terminal ends are the first to form during the refolding of hFGF-1. The basic beta-trefoil framework is provided by the simultaneous formation of beta-strands I, IV, IX, and X. The other beta-strands comprising the beta-barrel structure of hFGF-1 are formed relatively slowly with time constants ranging from 4 to 13 s.