Polyethylene glycol enhanced refolding of bovine carbonic anhydrase B. Reaction stoichiometry and refolding model.

Polyethylene glycol (PEG) inhibited aggregation during refolding of bovine carbonic anhydrase B (CAB) through the formation of a nonassociating PEG-intermediate complex. Stoichiometric concentrations of PEG were required for complete recovery of active protein during refolding at aggregating conditions. For example, a PEG (Mr = 3350) to CAB molar ratio ([PEG]/[CAB]) of 2 was sufficient to inhibit aggregation during refolding at 1.0 mg/ml (33.3 microM) protein and 0.5 M guanidine hydrochloride. In addition, the PEG concentration required for enhancement was dependent upon the molecular weight and only molecular weights between 1000 and 8000 were effective in inhibiting aggregation. In the presence of PEG, the rate of refolding was the same as that observed for refolding without the formation of associated species. Refolding in the presence of PEG resulted in the rapid formation of a PEG complex with the molten globule first intermediate, and this PEG-intermediate complex did not aggregate. The CAB refolding kinetics in the presence of PEG were determined and used to develop a model of the PEG enhanced refolding pathway. The mathematical model was validated by independent activity measurements of CAB refolding. This model predicted that PEG enhanced refolding of CAB occurred by a specific interaction of PEG with the molten globule first intermediate to form a nonassociating complex which continued to fold at the same rate as the first intermediate. The predicted pathway and binding properties of PEG indicate that PEG enhanced refolding may be analogous to chaperonin mediated protein folding.     

Protein refolding is improved by adding nonionic polyethylene glycol monooleyl ethers with various polyethylene glycol lengths

Protein refolding from bacterial inclusion bodies is a crucial step for the production of recombinant proteins, but the refolding step often results in significantly lower yields due to aggregation. To prevent aggregation, chemical additives are often used. However, the ability of additives to effectively increase refolding yields are protein dependent, and therefore, it is important to understand the manner in which the substructures of additives confer suitable properties on protein refolding. We focused attention on nonionic detergents, the polyethylene glycol monooleyl ether (PGME) series, and systematically studied the influence of two to 90 polyethylene glycol (PEG) lengths of PGMEs on the refolding of pig muscle lactate dehydrogenase (LDH), hen egg white lysozyme, and yeast α‐glucosidase. PGMEs with longer PEG lengths such as PGME20, 50, and 90 suppressed aggregation, and increased refolding yields. Notably, PGME20 increased the LDH yield to 56.7% from 2.5% without additives. According to the refolding kinetic analysis of LDH, compared with PGME50 and 90, the refolding rate constant in PGME20 solutions remained relatively high at a broad range of concentrations because of its weaker steric hindrance of intramolecular interactions involved in folding, leading to a preference for refolding over aggregation. These findings should provide basic guidelines to identify appropriate PEG‐based nonionic detergents for protein refolding.     

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

A point mutation, lysine 97 --> isoleucine (K97I), in a surface loop in the beta-sheet protein interleukin 1beta (IL-1beta), exhibits increased levels of inclusion body (IB) formation relative to the wild-type protein (WT) when expressed in Escherichia coli. Despite the common observation that less stable proteins are often found in IBs, K97I is more stable than WT. We examined the folding pathway of the mutant and wild-type proteins at pH 6.5 and 25 degrees C with manual-mixing and stopped-flow optical spectroscopy to determine whether changes in the properties of transiently populated species in vitro correlate with the observation of increased aggregation in vivo. The refolding reactions of the WT and K97I proteins are both described by three exponential processes. Two exponential processes characterize fast events (0.1-1.0 s) in folding while the third exponential process correlates with a slow (70 s) single pathway to and from the native state. The K97I replacement affects the earlier steps in the refolding pathway. Aggregation, absent in the WT refolding reaction, occurs in K97I above a critical protein concentration of 18 microM. This observation is consistent with an initial nucleation step mediating protein aggregation. Stopped-flow kinetic studies of the K97I aggregation process demonstrate that K97I aggregates most rapidly during the earliest refolding times, when unfolded protein conformers remain highly populated and the concentration of folding intermediates is low. Folding and aggregation studies together support a model in which the formation of stable folding intermediates afford protection against further K97I aggregation.     

Effects of macromolecular crowding on refolding of recombinant human brain-type creatine kinase

In this study, we quantitatively measured the effects of the macromolecular crowding agents, polyethylene glycol 2000 (PEG 2000), dextran 70, and calf thymus DNA (CT DNA), on the refolding and aggregation of recombinant human brain-type creatine kinase (rHBCK) denatured by guanidine hydrochloride (GdnHCl). The results showed that there is more aggregation in the presence of either a single crowding agent or in a mixture of crowding agents than in the absence of crowding agents, especially in the presence of a mixture containing CT DNA and PEG 2000 (or dextran 70). In the presence of high concentrations of PEG 2000 (100 g/L), dextran 70 (100 g/L), and CT DNA (15 g/L), the refolding yield remarkably decreased from 70% to 20%, 52% and 57%, respectively. A remarkable decrease in the refolding yield and rate with mixed crowding agent containing CT DNA and PEG 2000 (or dextran 70) was also observed. In comparison to refolding in the presence of 100 g/L PEG 2000, the refolding yields and rates improved in the presence of a mixture of PEG 2000 and dextran 70. We speculate that the crowding agents can favor both correct folding and misfolding/aggregation of denatured-rHBCK. Though it is not known what combination of crowding agents most accurately reflects the physiological environment within a cell, we believe our study could contribute to the understanding of protein folding and the factors that contribute to proper conformation and function in the intracellular environment.     

Capture of monomeric refolding intermediate of human muscle creatine kinase

Human muscle creatine kinase (CK) is an enzyme that plays an important physiological role in the energy metabolism of humans. It also serves as a typical model for studying refolding of proteins. A study of the refolding and reactivation process of guanidine chloride-denatured human muscle CK is described in the present article. The results show that the refolding process can be divided into fast and slow folding phases and that an aggregation process competes with the proper refolding process at high enzyme concentration and high temperature. An intermediate in the early stage of refolding was captured by specific protein molecules: the molecular chaperonin GroEL and alpha(s)-casein. This intermediate was found to be a monomer, which resembles the "molten globule" state in the CK folding pathway. To our knowledge, this is the first monomeric intermediate captured during refolding of CK. We propose that aggregation is caused by interaction between such monomeric intermediates. Binding of GroEL with this intermediate prevents formation of aggregates by decreasing the concentration of free monomeric intermediates, whereas binding of alpha(s)-casein with this intermediate induces more aggregation.     

The effects of arginine on refolding of aggregated proteins: not facilitate refolding,but suppress aggregation

Arginine is one of the universal reagents that are effective in assisting refolding of recombinant proteins from inclusion bodies. The mechanism of the effects of arginine on refolding has remained, however, to be elucidated. Here we show that arginine does not stabilize proteins against heat treatment, as demonstrated by little change in melting temperature. It does increase reversibility of thermal melting and reduce aggregation under thermal stress. The observations suggest that arginine may not facilitate refolding, but may suppress aggregation of the proteins during refolding.     

Critical analysis of lysozyme refolding kinetics

The kinetics of lysozyme refolding and aggregation is studied using an existing competing first- and third-order reaction scheme. The existing model overestimates yield at high refolding concentrations (>1 mg/mL), thus limiting its use for reactor design at industrially relevant refolding concentrations. This study demonstrates that a pathway exists for the incorporation of refolded native protein into aggregates. Specifically, native lysozyme labeled with fluorescein isothiocyanate was added to the refolding buffer prior to dilution refolding of denatured and reduced lysozyme. Aggregates collected from these experiments showed significant fluorescence, indicating that labeled lysozyme had been incorporated into the aggregates during refolding. Although the precise pathway of incorporation has not been elucidated, it is clear from this work that the existing model for lysozyme refolding is not globally applicable. In particular, previous work has analytically demonstrated that neglect of a pathway from native to aggregate can result in the design of a grossly suboptimal reactor strategy. This study demonstrates that such a pathway can exist experimentally and emphasizes the need to critically assess refolding kinetic models before their use in reactor design equations.     

Transient association of the first intermediate during the refolding of bovine carbonic anhydrase B.

Many proteins which aggregate during refolding may form transiently populated aggregated states which do not reduce the final recovery of active species. However, the transient association of a folding intermediate will result in reduced refolding rates if the dissociation process occurs slowly. Previous studies on the refolding and aggregation of bovine carbonic anhydrase B (CAB) have shown that the molten globule first intermediate on the CAB folding pathway will form dimers and trimers prior to the formation of large aggregates (Cleland, J. L.; Wang, D. I. C. Biochemistry 1990, 29, 11072-11078; Cleland, J. L.; Wang, D. I. C. In Protein Refolding; Georgiou, G., De-Bernardez-Clark, E., Eds.; ACS Symposium Series 470; American Chemical Society: Washington, DC, 1991; pp 169-179). Refolding of CAB from 5 M guanidine hydrochloride (GuHCl) was achieved at conditions ([CAB]f = 10-33 microM, [GuHCl]f = 1.0 M) which allowed complete recovery of active protein as well as the formation of a transiently populated dimer of the molten globule intermediate on the refolding pathway. A kinetic analysis of CAB refolding provided insight into the mechanism of the association phenomenon. Using the kinetic results, a model of the refolding with transient association was constructed. By adjusting a single variable, the dimer dissociation rate constant, the model prediction fit both the experimentally determined active protein and dimer concentrations. The model developed in this analysis should also be applicable to the refolding of proteins which have been observed to form aggregates during refolding. In particular, the transient association of hydrophobic folding intermediates may also occur during the refolding of other proteins.(ABSTRACT TRUNCATED AT 250 WORDS)     

Disulfide–bridging PEGylation during refolding for the more efficient production of modified proteins

Proteins that are modified by chemical conjugation require at least two separate purification processes. First the bulk protein is purified, and then after chemical conjugation, a second purification process is required to obtain the modified protein. In an effort to develop new enabling technologies to integrate bioprocessing and protein modification, we describe the use of disulfide‐bridging conjugation to conduct PEGylation during protein refolding. Preliminary experiments using a PEG‐mono‐sulfone reagent with partially unfolded leptin and unfolded RNAse T1 indicated that the cysteine thiols underwent disulfide‐bridging conjugation to give the PEGylated proteins. Interferon‐β1b (IFN‐β1b) was then expressed in E.coli as inclusion bodies and found to undergo disulfide bridging‐conjugation during refolding. The PEG‐IFN‐β1b was isolated by ion‐exchange chromatography and displayed in vitro biological activity. In the absence of the PEGylation reagent, IFN‐β1b refolding was less efficient and yielded protein aggregates. No PEGylation was observed if the cysteines on IFN‐β1b were first modified with iodoacetamide prior to refolding. Our results demonstrate that the simultaneous refolding and disulfide bridging PEGylation of proteins could be a useful strategy in the development of affordable modified protein therapeutics.     

Ion-exchange resins greatly facilitate refolding of like-charged proteins at high concentrations

Protein refolding is a crucial step for the production of therapeutic proteins expressed in bacteria as inclusion bodies. In vitro protein refolding is severely impeded by the aggregation of folding intermediates during the folding process, so inhibition of the aggregation is the most effective approach to high‐efficiency protein refolding. We have herein found that electrostatic repulsion between like‐charged protein and ion exchange gel beads can greatly suppress the aggregation of folding intermediates, leading to the significant increase of native protein recovery. This finding is extensively demonstrated with three different proteins and four kinds of ion‐exchange resins when the protein and ion‐exchange gel are either positively or negatively charged at the refolding conditions. It is remarkable that the enhancing effect is significant at very high protein concentrations, such as 4 mg/mL lysozyme (positively charged) and 2 mg/mL bovine serum albumin (negatively charged). Moreover, the folding kinetics is not compromised by the presence of the resins, so fast protein refolding is realized at high protein concentrations. It was not realistic by any other approaches. The working mechanism of the like‐charged resin is considered due to the charge repulsion that could induce oriented alignment of protein molecules near the charged surface, leading to the inhibition of protein aggregation. The molecular crowding effect induced by the charge repulsion may also contribute to accelerating protein folding. The refolding method with like‐charged ion exchangers is simple to perform, and the key material is easy to separate for recycling. Moreover, because ion exchangers can work as adsorbents of oppositely charged impurities, an operation of simultaneous protein refolding and purification is possible. All the characters are desirable for preparative refolding of therapeutic proteins expressed in bacteria as inclusion bodies. Bioeng. 2011; 108:1068–1077. © 2010 Wiley Periodicals, Inc.     

The glycophosphatidylinositol anchor oppositely affects unfolding and refolding of alkaline phosphatase

Regarding the world wide success of artificial chaperone-assisted protein refolding technique and based on its well worked-out mechanism, it is anticipated that the lipid moieties of glycosylphosphatidylinositol (GPI) group, which is present in some membrane proteins, might interfere with the capturing step of the technique. To find an answer, we evaluated the chemical denaturation and also the refolding behavior of insoluble and soluble alkaline phosohatase (ALP), with or without GPI group, respectively. The results indicated that the presence of GPI in the enzyme increased the stability of the protein against chemical denaturation while it decreased its refolding yield by the artificial chaperone refolding technique. The lower refolding yield, compared to soluble ALP (sALP), might be due to a less efficient stripping step caused by new interactions imparted to the refolding elements of the system especially those among the hydrophobic tails of GPI and the capturing agent of the technique. These new interactions will interrupt the kinetics of detergent stripping from the captured molecules by the stripping agent (i.e., cyclodextrins). This situation will lead to higher intermolecular hydrophobic interactions among the refolding protein intermediates leading to their higher misfolding and aggregation.     

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.     

Protein refolding assisted by self-assembled nanogels as novel artificial molecular chaperone

Molecular chaperone-like activity for protein refolding was investigated using nanogels of self-assembly of cholesterol-bearing pullulan. Nanogels effectively prevented protein aggregation (i.e. carbonic anhydrase and citrate synthase) during protein refolding from GdmCl denaturation. Enzyme activity recovered in high yields upon dissociation of the gel structure in which the proteins were trapped, by the addition of cyclodextrins. The nanogels assisted protein refolding in a manner similar to the mechanism of molecular chaperones, namely by catching and releasing proteins. The nanogels acted as a host for the trapping of refolded intermediate proteins. Cyclodextrin is an effector molecule that controls the binding ability of these host nanogels to proteins. The present nanogel system was also effective at the renaturation of inclusion body of a recombinant protein of the serine protease family.     

Unassisted refolding of urea-denatured arginine kinase from shrimp Feneropenaeus chinensis: evidence for two equilibrium intermediates in the refolding pathway

The refolding process and the equilibrium intermediates of urea-denatured arginine kinase (AK) were investigated by 1-anilino-8-naphthalenesulfonate (ANS) intrinsic fluorescence, far-UV circular dichroism (CD), size-exclusion chromatography (SEC), and enzymatic activity. In dilute denaturant, two equilibrium refolding intermediates (I and N') were discovered, and a refolding scheme of urea-denatured AK was proposed. During the refolding of urea-denatured AK, the fluorescence intensity increased remarkably, accompanied by a significant blue shift of the emission maximum and a pronounced increase in molar ellipticity of CD at 222 nm. The first folding intermediate (I) was inactive in urea solution ranging between 2.4 and 3.0 M. The second (N') existed between a 0.4- and 0.8-M urea solution, with slightly increased activity. Neither the blue shift emission maximum nor the molar ellipticity of CD at 222 nm showed significant changes in these two regions. The two intermediates were characterized by monitoring the ANS binding ability in various residual urea solutions, and two peaks of the emission intensity were observed in urea solutions of 0.6 and 2.8 M, respectively. The SEC results indicated that a distribution coefficient (K(D)) platform existed in urea solutions ranging between 2.4 and 3.0 M urea, suggesting that there was a similarly apparent protein profile and size in the urea solution region. The refolding kinetics showed that the urea-denatured AK was in two-phase refolding. Proline isomerization occurred in the unfolding process of AK, which blocked the slow phase of refolding. These results suggested that the refolding process of urea-denatured AK contained at the least two equilibrium refolding intermediates.     

Chaperone-assisted refolding of Escherichia coli maltodextrin glucosidase

In vitro refolding of maltodextrin glucosidase, a 69 kDa monomeric Escherichia coli protein, was studied in the presence of glycerol, dimethylsulfoxide, trimethylamine-N-oxide, ethylene glycol, trehalose, proline and chaperonins GroEL and GroES. Different osmolytes, namely proline, glycerol, trimethylamine-N-oxide and dimethylsulfoxide, also known as chemical chaperones, assist in protein folding through effective inhibition of the aggregation process. In the present study, it was observed that a few chemical chaperones effectively reduced the aggregation process of maltodextrin glucosidase and hence the in vitro refolding was substantially enhanced, with ethylene glycol being the exception. Although, the highest recovery of active maltodextrin glucosidase was achieved through the ATP-mediated GroEL/GroES-assisted refolding of denatured protein, the yield of correctly folded protein from glycerol- or proline-assisted spontaneous refolding process was closer to the chaperonin-assisted refolding. It was also observed that the combined application of chemical chaperones and molecular chaperone was more productive than their individual contribution towards the in vitro refolding of maltodextrin glucosidase. The chemical chaperones, except ethylene glycol, were found to provide different degrees of protection to maltodextrin glucosidase from thermal denaturation, whereas proline caused the highest protection. The observations from the present studies conclusively demonstrate that chemical or molecular chaperones, or the combination of both chaperones, could be used in the efficient refolding of recombinant E. coli maltodextrin glucosidase, which enhances the possibility of identifying or designing suitable small molecules that can act as chemical chaperones in the efficient refolding of various aggregate-prone proteins of commercial and medical importance.     

Protein refolding using chemical refolding additives

In laboratories and manufacturing settings, a rapid and inexpensive method for the preparation of a target protein is crucial for promoting resesrach in protein science and engineering. Inclusion-body-based protein production is a promising method because high yields are achieved in the upstream process, although the refolding of solubilized, unfolded proteins in downstream processes often leads to significantly lower yields. The most challenging problem is that the effective condition for refolding is protein dependent and is therefore difficult to select in a rational manner. Accordingly, considerable time and expense using trial-and-error approaches are often needed to increase the final protein yield. Furthermore, for certain target proteins, finding suitable conditions to achieve an adequate yield cannot be obtained by existing methods. Therefore, to convert such a troublesome refolding process into a routine one, a wide array of methods based on novel technologies and materials have been developed. These methods select refolding conditions where productive refolding dominates over unproductive aggregation in competitive refolding reactions. This review focuses on synthetic refolding additives and describes the concepts underlying the development of reported chemical additives or chemical-additive-b     

Non-chromatographic strategies for protein refolding

Overexpression of recombinant proteins in bacterial systems (such as E. coli) often leads to formation of inactive and insoluble ' inclusion bodies' . Protein refolding refers to folding back the proteins after solubilizing/unfolding the misfolded proteins of the inclusion bodies. Protein aggregation, a concentration dependent phenomenon, competes with refolding pathway. The refolding strategies largely aim at reducing aggregation and/or promoting correct folding. This review focuses on non-chromatographic strategies for refolding like dilution, precipitation, three phase partitioning and macro-(affinity ligand) facilitated three phase partitioning. The nanomaterials which disperse well in aqueous buffers are also discussed in the context of facilitating protein refolding. Apart from general results with these methods, the review also covers the use of non-chromatographic methods in protein refolding in the patented literature beyond 2000. The patented literature generally describes use of cocktail of additives which results in increase in refolding yield. Such additives include low concentration of chaotropic agents, redox systems, ions like SO4(2-) and Cl-, amines, carboxylic acids and surfactants. Some novel approaches like use of a "pressure window" or ionic liquids for refolding and immobilized diselenide compounds for ensuring correct -S-S- bonds pairing have also been discussed in various patents. In most of the patented literature, focus naturally has been on refolding in case of pharmaceutical proteins.     

Lysozyme refolding with cyclodextrins: structure-activity relationship

Cyclodextrins (CDs), in the presence or absence of detergents, have been reported to suppress aggregate formation during the refolding of a number of proteins. A structure-activity relationship study between CD chemistry and refolding of lysozyme was performed and compared to carbonic anhydrase, in order to better understand the mechanism of CD-assisted protein refolding and to identify CDs that could function as good protein folding agents. Among the natural CDs, which have only hydroxyl groups, alpha-CD, with a smaller cavity size was more effective than the oligosaccharide with a larger cavity, gamma-CD. Replacement of the hydroxyls with other functional groups did not improve, but could seriously interfere, with the lysozyme refolding ability of alpha-CD. In case of gamma-CD, substitution of its hydroxyls with other groups either enhanced or diminished its refolding capability towards lysozyme. In general, neutral CDs were better refolding agents than the charged sugars. The presence of anionic substituents like carboxyl and phosphate groups actually promoted aggregate formation and completely abolished the sugar's refolding ability. This effect was more pronounced with lysozyme than with carbonic anhydrase. CDs with cationic functional groups did not show any significant effects on lysozyme refolding. The presence of both anionic and cationic substituents on the same CD molecule was found to partially restore its renaturation ability. Electrophoresis data indicate that CDs, which promoted lysozyme refolding, arrested aggregation at the stage of smaller soluble aggregates. Interestingly, the structure-activity relationship observed with lysozyme was quite similar to that reported for a non-disulfide protein, carbonic anhydrase. These results suggest that the effects of CDs on protein refolding are attributed to their ability to suppress aggregation of proteins. CDs may show properties similar to chaotropic agents, which may help explain their anti-aggregation and protein refolding ability. Besides alpha-CD, a number of other neutral CDs were found to be effective protein folding aids.     

Improved protein refolding using hollow-fibre membrane dialysis

We have used a cellulose acetate, hollow-fibre (HF) ultrafiltration membrane to refold bovine carbonic anhydrase, loaded into the lumen space, by removing the denaturant through controlled dialysis via the shell side space. When challenged with GdnHCl-denatured carbonic anhydrase, 70% of the loaded protein reptated through the membrane into the circulating dialysis buffer. Reptation occurred because the protein, in its fully unfolded configuration, was able to pass through the pores. The loss of carbonic anhydrase through the membrane was controlled by the dialysis conditions. Dialysis against 0.05 M Tris-HCl for 30 min reduced the denaturant around the protein to a concentration that allowed the return of secondary structure, increasing the hydrodynamic radius, thus preventing protein transmission. Under these conditions a maximum of 42% of carbonic anhydrase was recovered (from a starting concentration of 5 mg/mL) with 94% activity. This is an improvement over refolding carbonic anhydrase by simple batch dilution, which gave a maximum reactivation of 85% with 35% soluble protein yield. The batch refolding of carbonic anhydrase is very sensitive to temperature; however, during HF refolding between 0 and 25 degrees C the temperature sensitivity was considerably reduced. In order to reduce the convection forces that give rise to aggregation and promote refolding the dialyzate was slowly heated from 4 to 25 degrees C. This slow, temperature-controlled refolding gave an improved soluble protein recovery of 55% with a reactivation yield of 90%. The effect of a number of additives on the refolding system performance were tested: the presence of PEG improved both the protein recovery and the recovered activity from the membrane, while the detergents Tween 20 and IGEPAL CA-630 increased only the refolding yield.     

Protein refolding in reversed micelles: Interactions of the protein with micelle components

A novel process has been developed to improve the refolding yield of denatured proteins. It uses reversed micelles to isolate denatured protein molecules from each other and thus, upon refolding, reduces the intermolecular interactions which lead to aggregation. The feasibility of this process was first demonstrated with Ribonuclease A as a model protein. In the present work, we expanded the scope of this study to better understand both the general mechanisms of protein refolding in reversed micelles and the biotechnological applicability of the process. First, we investigated the interactions between the individual components of the reversed micellar system (the protein molecule, the denaturant guanidine hydrochloride (GuHCl), and the surfactant (AOT)) during the refolding process. We then extended our studies to a more hydrophobic protein, gamma-interferon, which aggregates upon refolding in aqueous solution. However, it was also found to aggregate in our reversed micelle process during the extraction step. Since gamma-interferon is a much more hydrophobic protein than RNase, we hypothesize that interactions between hydrophobic amino acids and the surfactant layer may interfere with refolding. This hypothesis was tested by studying the refolding of chemically modified RNase. The substitution of 55% of the surface lysine residues with hydrophobic caproyl groups caused a significant decrease in the refolding yield of RNase in the reversed micellar system without affecting aqueous solution renaturation. In addition, the extraction efficiency of the enzyme from reversed micelles back into aqueous solution was severely reduced and resulted in aggregation. These experiments indicate that unfolded hydrophobic Proteinsinteract with the Surfactant molecules, which limits their ability to refold in reversed micelles.