C-terminal site-specific PEGylation of a truncated thrombomodulin mutant with retention of full bioactivity

Addition of poly(ethylene glycol) to bioactive proteins (PEGylation) improves their plasma half-life, enhances stability against proteolytic cleavage, and may also decrease protein immunogenicity. Characteristically, PEGylation usually involves a reaction to available lysine amino groups, some of which may be within or near a bioactive site. Thus, most protocols are nonspecific and result in a loss of protein activity. We report herein a strategy for site-specific PEGylation of a thrombomodulin (TM) derivative at the C terminus. A truncated TM mutant consisting of epidermal growth factor (EGF)-like domains 4-6 was expressed in Escherichia coli with a C-terminal azido-methionine. The TM mutant was site-specifically conjugated to a methyl-PEG-triarylphosphine compound via the Staudinger reaction. Enzymatic activity of the TM construct before and after PEGylation was unchanged, which confirms the utility of this site-specific PEGylation scheme.     

PEGylation of native disulfide bonds in proteins

PEGylation has turned proteins into important new biopharmaceuticals. The fundamental problems with the existing approaches to PEGylation are inefficient conjugation and the formation of heterogeneous mixtures. This is because poly(ethylene glycol) (PEG) is usually conjugated to nucleophilic amine residues. Our PEGylation protocol solves these problems by exploiting the chemical reactivity of both of the sulfur atoms in the disulfide bond of many biologically relevant proteins. An accessible disulfide bond is mildly reduced to liberate the two cysteine sulfur atoms without disturbing the protein's tertiary structure. Site-specific PEGylation is achieved with a bis-thiol alkylating PEG reagent that sequentially undergoes conjugation to form a three-carbon bridge. The two sulfur atoms are re-linked with PEG selectively conjugated to the bridge. PEGylation of a protein can be completed in 24 h and purification of the PEG-protein conjugate in another 3 h. We have successfully applied this approach to PEGylation of cytokines, enzymes, antibody fragments and peptides, without destroying their tertiary structure or abolishing their biological activity.     

Mutants of immunotoxin anti-Tac(dsFv)-PE38 with variable number of lysine residues as candidates for site-specific chemical modification. 1. Properties of mutant molecules

Chemical modification of proteins with substances such as poly(ethylene glycol) can add useful properties to proteins. Currently PEGylation is done in a random manner utilizing amino residues dispersed throughout a protein. For proteins such as immunotoxins, which have several different functional domains, random modification leads to inactivation. To determine if we could produce an immunotoxin with a diminished number of lysine residues so that chemical modification could be restricted to certain regions of the protein, we chose the recombinant immunotoxin anti-Tac(dsFv)-PE38 that has 13 lysine residues in the Fv portion and 3 in the toxin. We prepared a series of mutants with 0-12 lysines in the Fv and 0 or 3 in the toxin. Almost all of these molecules retain full biological activity. Our data indicate that replacement of lysine residues can be achieve without loss of biological potency. These molecules are a useful starting point to carry out site-specific PEGylation experiments.     

Spatially well-defined binary brushes of poly(ethylene glycol)s for micropatterning of active proteins on anti-fouling surfaces

We report a novel method for micropatterning of active proteins on anti-fouling surfaces via spatially well-defined and dense binary poly(ethylene glycol)s (PEGs) brushes with controllable protein-docking sites. Binary brushes of poly(poly(ethylene glycol) methacrylate-co-poly(ethylene glycol)methyl ether methacrylate), or P(PEGMA-co-PEGMEMA), and poly(poly(ethylene glycol)methyl ether methacrylate), or P(PEGMEMA), were prepared via consecutive surface-initiated atom transfer radical polymerizations (SI-ATRPs) from a resist-micropatterned Si(100) wafer surface. The terminal hydroxyl groups on the side chains of PEGMA units in the P(PEGMA-co-PEGMEMA) microdomains were activated directly by 1,1'-carbonyldiimidazole (CDI) for the covalent coupling of human immunoglobulin (IgG) (as a model active protein). The resulting IgG-coupled PEG microdomains interact only and specifically with target anti-IgG, while the other PEG microregions effectively prevent specific and non-specific protein fouling. When extended to other active biomolecules, microarrays for specific and non-specific analyte interactions with a high signal-to-noise ratio could be readily tailored.     

PEGylated protein separation using different hydrophobic interaction supports: Conventional and monolithic supports

Protein hydrophobicity can be modified after a PEGylation process. However, hydrophobic interaction chromatography (HIC) has been used to separate PEGylation reaction products less frequently than other techniques. In this context, chromatographic monoliths represent a good alternative to continue exploring the separation of PEGylated proteins with HIC. In this work, the separation of PEGylated proteins using C4 A monolith as well as Toyopearl Butyl 650C and Butyl Sepharose was analyzed. Three proteins were used as models: RNase A, β‐lactoglobulin, and lysozyme. All proteins were PEGylated in the N‐terminal amino groups with 20 kDa methoxy poly(ethylene glycol) propionaldehyde. The concentration of ammonium sulfate (1 M) used was the same for all stationary phases. The results obtained demonstrated that the C4 A monolith could better resolve all protein PEGylation reaction mixtures, since the peaks of mono‐ and di‐PEGylated proteins can be clearly distinguished in the chromatographic profiles. On the contrary, while using Butyl Sepharose media only the PEGylation reaction mixtures of RNase A could be partially separated at 35 and 45 CVs. PEGylated proteins of β‐lactoglobulin and lysozyme could not be resolved when Toyopearl Butyl 650C and Butyl Sepharose were used. It is then clear that monoliths are an excellent choice to explore the purification process of PEGylated proteins exploiting the advantages of HIC. © 2016 American Institute of Chemical Engineers Biotechnol. Prog., 32:702–707, 2016     

Preparation and characterization of PEGyated Concanavalin A for affinity chromatography with improved stability

In order to improve its stability, immobilized Concanavalin A (Con A) on Toyopearl adsorbents was conjugated with monomethoxy poly(ethylene glycol) succinimidyl propionate (mPEG-SPA) with different molecular weight. A colorimetric method using ninhydrin is proposed to determine the degree of PEGylation; this method has proved to be easy applicable and reproducible. The PEGylation reaction was studied in detail to elucidate how parameters such as molar ratio of mPEG-SPA to Con A and molecular weight of mPEG-SPA affect the degree of PEGylation. The adsorption isotherms of glucose oxidase (GOD) onto native and PEGylated Con A adsorbents showed that the modification did not alter substantially the specificity of the carbohydrate binding ability of Con A. However, the binding capacity for GOD was slightly reduced probably due to the steric hindrance caused by mPEG chains. Adsorption kinetic studies revealed a lower adsorption rate after PEGylation which was attributed to the steric effect. The dynamic adsorption capacity for modified Con A depended very much on the degree of PEGylation and the molecular weight of mPEG derivatives. The adsorption capacity could be highly preserved for Toyopearl Con A modified by mPEG2k (90% of the original adsorption capacity) even with a degree of PEGylation up to 20% (the ratio of primary amino groups of PEGylated immobilized Con A to that of native immobilized Con A). Studies show that the binding capacity of PEGylated Con A was highly preserved under mild process conditions. PEGylated Con A also exhibited obviously higher stability against more stressful conditions such as the exposure to organic solvents and high temperatures. Conjugation of Con A with mPEG2k provided better adsorption performance thus has greater potential for application in affinity separation processes compared with mPEG5k. The fact that PEGylation stabilizes the properties of Con A may greatly expand the range of applications of unstable proteins to bioprocessing (e.g. biocatalysis and downstream separation) as well as other protein applications (e.g. medication, industrial use, etc.).     

Hydrophobic interaction determined by partition in aqueous two-phase systems. Partition of proteins in systems containing fatty-acid esters of poly(ethylene glycol).

In this report we describe a new method which is useful for measuring hydrophobic interactions between aliphatic hydrocarbon chains and proteins in aqueous environment. The method is based on partition of proteins in an aqueous two-phase system containing dextran and poly(ethylene glycol) and different fatty acid esters of poly(ethylene glycol). The partition is measured under conditions where contributions from electrostatic interactions are eliminated. The difference in partition of proteins in phase systems with and without hyrocarbon groups bound to poly(ethylene glycol), deltalog K, where K is the partition coefficient, is taken as a measure of hydrophobic interaction. Deltalog K varies with size of hydrocarbon chain and type of protein. The length of the aliphatic chain should be greater than 8 carbon atoms in order to get a measurable effect in terms of deltalog K. Bovine serum albumin, beta-lactoglobulin, hemoglobin and myoglobin have been shown to have different affinities for palmitic acid ester of poly(ethylene glycol). No hydrophobic effect could be observed for ovalbumin, cytochrome c or alpha-chymotrypsinogen A.     

Thermal stability of proteins in the presence of poly(ethylene glycols)

Thermal unfolding of ribonuclease, lysozyme, chymotrypsinogen, and beta-lactoglobulin was studied in the absence or presence of poly(ethylene glycols). The unfolding curves were fitted to a two-state model by a nonlinear least-squares program to obtain values of delta H, delta S, and the melting temperature Tm. A decrease in thermal transition temperature was observed in the presence of poly(ethylene glycol) for all of the protein systems studied. The magnitude of such a decrease depends on the particular protein and the molecular size of poly(ethylene glycol) employed. A linear relation can be established between the magnitude of the decrease in transition temperature and the average hydrophobicity of these proteins; namely, the largest observable decrease is associated with the protein of the highest hydrophobicity. Further analysis of the thermal unfolding data reveals that poly(ethylene glycols) significantly effect the relation between delta H degrees of unfolding and temperature for all the proteins studied. For beta-lactoglobulin, a plot of delta H versus Tm indicates a change in slope from a negative to a positive value, thus implying a change in delta Cp in thermal unfolding caused by the presence of poly(ethylene glycols). Results from solvent-protein interaction studies indicate that at high temperature poly(ethylene glycol) 1000 preferentially interacts with the denatured state of protein but is excluded from the native state at low temperature. These observations are consistent with the fact that poly(ethylene glycols) are hydrophobic in nature and will interact favorably with the hydrophobic side chains exposed upon unfolding; thus, it leads to a lowering of thermal transition temperature.     

The conformation of the poly(ethylene glycol) chain in mono-PEGylated lysozyme and mono-PEGylated human growth hormone

Covalent conjugation of poly(ethylene glycol) or "PEGylation" has proven an effective strategy to improve pharmaceutical protein efficacy by hindering recognition by proteases, inhibitors, and antibodies and by retarding renal clearance. Because it determines the strength and range of intermolecular steric forces and the hydrodynamic properties of the conjugates, the configuration of protein-conjugated PEG chains is the key factor determining how PEGylation alters protein in vivo circulation time. Mono-PEGylated proteins are typically described as having a protective PEG shroud wrapped around the protein, but recent dynamic light scattering studies suggested that conjugates adopt a dumbbell configuration, with a relatively unperturbed PEG random coil adjacent to the globular protein. We used small-angle neutron scattering (SANS) to distinguish between the dumbbell model and the shroud model for chicken-egg lysozyme and human growth hormone covalently conjugated to a single 20 kDa PEG chain. The SANS contrast variation technique was used to isolate the PEG portion of the conjugate. Scattering intensity profiles were well described by the dumbbell model and inconsistent with the shroud model.     

Mechanism of poly(ethylene glycol) interaction with proteins

Poly(ethylene glycol) (PEG) is one of the most useful protein salting-out agents. In this study, it has been shown that the salting-out effectiveness of PEG can be explained by the large unfavorable free energy of its interaction with proteins. Preferential interaction measurements of beta-lactoglobulin with poly(ethylene glycols) with molecular weights between 200 and 1000 showed preferential hydration of the protein for those with Mr greater than or equal to 400, the degree of hydration increasing with the increase in poly(ethylene glycol) molecular weight. The preferential interaction parameter had a strong cosolvent concentration dependence, with poly(ethylene glycol) 1000 having the sharpest decrease with an increase in concentration. The preferential hydration extrapolated to zero cosolvent concentration increased almost linearly with increasing size of the additive, suggesting steric exclusion as the major factor responsible for the preferential hydration. The poly(ethylene glycol) concentration dependence of the preferential interactions could be explained in terms of the nonideality of poly(ethylene glycol) solutions. All the poly(ethylene glycols) studied, when used at levels of 10-30%, decreased the thermal stability of beta-lactoglobulin, suggesting that caution must be exercised in the use of this additive at extreme conditions such as high temperature.     

Engineering of Aerococcus viridans L-lactate oxidase for site-specific PEGylation: characterization and selective bioorthogonal modification of a S218C mutant

A defined bioconjugate of Aerococcus viridans L-lactate oxidase and poly(ethylene glycol) 5000 was prepared and characterized in its structural and functional properties in comparison to the unmodified enzyme. Because the L-lactate oxidase in the native form does not contain cysteines, we introduced a new site for chemical modification via thiol chemistry by substituting the presumably surface-exposed serine-218, a nonconserved residue in the amino acid sequence, with cysteine. The resulting S218C mutant was isolated from Escherichia coli and shown in kinetic assays to be similarly (i.e., about half as) active as the native enzyme, thus validating the structure-guided design of the mutation. Using maleimide-activated methoxypoly(ethylene glycol) 5000 in about 10-fold molar excess over protein, the S218C mutant was converted in high yield (94%) into PEGylated derivative, while the native enzyme was totally unreactive under equivalent conditions. PEGylation caused only a relatively small decrease (30%) in the specific activity of the S218C mutant, and it did not change the protein stability. PEGylation went along with enhancement of the apparent size of the homotetrameric L-lactate oxidase in gel permeation chromatography, from 170 kDa to 250 kDa. The protein hydrodynamic diameter determined by dynamic light scattering increased from 11.9 nm in unmodified S218C mutant to 16.4 nm in the PEGylated form. Site-selective PEGylation of the mutated L-lactate oxidase, using orthogonal maleimide-thiol coupling, could therefore facilitate incorporation of the enzyme into biosensors currently employed for determination of blood L-lactate levels, and it could also support different applications of the enzyme in applied biocatalysis.     

Site-specific PEGylation of bone morphogenetic protein-2 cysteine analogues

Three cysteine analogues of bone morphogenetic protein (BMP)-2, BMP2A2C, BMP2N56C, and BMP2E96C, were generated in order to enable the attachment of SH-reactive poly(ethylene glycol) (PEG) at specific sites. Three different approaches (Ap) were used for SH-specific PEGylation: (Ap1) reaction of glutathione activated proteins with thiol PEG; (Ap2) reaction of DTT reduced proteins with orthopyridyl disulfide PEG; (Ap3) reaction of DTT reduced proteins with maleimide PEG. Non-, mono-, and di-PEGylated BMP-2 analogues could be separated by RP-HPLC. Trypsin digestion of PEGylated proteins and Trypsin and GluC double-digestion of N-ethylmaleimide-labeled proteins confirmed that the modifications were site-specific. Surface plasmon resonance analysis of type I and type II receptor binding of the PEGylated BMP-2 analogues revealed that all three PEGylation approaches were equivalent. PEGylation at positions 2 and 96 caused a similar decrease in receptor affinity. PEGylation at position 56 resulted in a larger decrease in affinity for both types of receptors. Mono-PEGylated BMP-2 analogues exhibited intermediate affinities in comparison with unmodified and di-PEGylated proteins. However, the biological activity of the PEGylated BMP-2 analogues as measured in alkaline phosphatase assay was higher than BMP-2 wild-type for the PEGylated BMP2A2C, slightly reduced for the BMP2N56C, and strongly reduced for the BMP2E96C. These results taken together indicate that specific attachment of PEG at engineered sites of BMP-2 is possible and that the attachment site is critical for biological activity. Furthermore, the biological activity of PEGylated BMP-2 analogues in cell culture seems to be determined not only by receptor affinity, but also by other factors such as protein solubility and stability. It is also discussed that the attached PEG interferes with the binding of BMP-2 to modulator proteins, co-receptors, or heparinic sites of proteoglycans in the extracellular matrix.     

Toward improving selectivity in affinity chromatography with PEGylated affinity ligands: The performance of PEGylated protein A

Chemical modification of macromolecular affinity chromatography ligands with polyethylene glycol chains or “PEGylation” can potentially improve selectivity by sterically suppressing non‐specific binding interactions without sacrificing binding capacity. For a commercial protein A affinity media and with yeast extract (YE) and fetal bovine serum (FBS) serving as mock contaminants, we found that the ligand accounted for more than 90% of the media‐associated non‐specific binding, demonstrating an opportunity for improvement. The IgG static binding affinity of protein A mono‐PEGylated with 5.0 and 20.7 kDa poly(ethylene glycol) chains was found to be preserved using a biomolecular interaction screening platform. Similar in situ PEGylations of the commercial protein A media were conducted and the modified media was functionally characterized with IgG solutions spiked with YE and FBS. Ligand PEGylation reduced the mass of media‐associated contaminants by a factor of two to three or more. Curiously, we also found an increase of up to 15% in the average recovery of IgG on elution after PEGylation. Combined, these effects produced an order of magnitude increase in the IgG selectivity on average when spiked with YE and a two‐ to three‐fold increase when spiked with FBS relative to the commercial media. Dynamic binding capacity and mass‐transfer resistance measurements revealed a reduction in dynamic capacity attributed to a decrease in IgG effective pore diffusivity and possibly slower IgG association kinetics for the PEGylated protein A ligands. Ligand PEGylation is a viable approach to improving selectivity in affinity chromatography with macromolecular ligands. © 2014 American Institute of Chemical Engineers Biotechnol. Prog., 30:1364–1379, 2014     

Preparation and characterization of active site protected poly(ethylene glycol)–avidin bioconjugates

Avidin was modified with poly(ethylene glycol) in the presence of a biotin binding site protective agent synthesised by imminobiotin conjugation to branched 20 kDa PEG. Avidin was incubated with imminobiotin–PEG and reacted with high amounts of 5, 10 or 20 kDa PEG to modify the protein amino groups. Circular dichroism demonstrated that the extensive PEGylation does not alter the protein conformational structure. The affinity of avidin–PEG conjugates for biotin and biotinylated antibodies depended on the PEG size or the use of a protective agent. Avidin–PEG 10 and 20 kDa prepared in the presence of imminobiotin–PEG maintained 100% of the native affinity for biotin. The 5 kDa PEG derivative and the ones obtained without biotin site protection maintained 79–85% of the native affinity. The affinity for biotinylated antibodies decreased to 35% when the conjugation was performed without imminobiotin–PEG, while the conjugates obtained with high-molecular-weight PEGs in the presence of protective agent displayed high residual affinity. All conjugates possessed negligible antigenicity and immunogenicity. PEGylation greatly prolonged the avidin permanence in the circulation, reduced its disposition in the liver and kidneys and promoted accumulation into solid tumors. PEGylation was found to prevent the protein cell uptake, either by phagocytosis or pinocytosis.     

Structure-property relationships in poly(ethylene glycol)-protein hydrogel systems made from various proteins

A series of poly(ethylene glycol)-protein hydrogels were synthesized with different proteins, and the resultant structures were characterized in terms of swelling behavior and mechanical, optical, and drug release properties. Irrespectively of the protein involved in polymerization with poly(ethylene glycol), all studied systems were found to be loosely cross-linked networks, where both polymer and protein are completely solvated, enabling as high as 96% water content. Changes in the apparent transparency of the hydrogels synthesized with different proteins were attributed to the ability of the protein component to self-associate via hydrophobic interactions. The polyelectrolyte nature of the protein component governs the pH responsiveness of the network, which manifested itself in a pH-dependent mechanism of swelling and drug release. It was demonstrated that there is great opportunity to modulate the final characteristics of the hydrogel system to fit the need of specific biomedical application.     

Adsorption of lipoproteins with the aid of carboxymethylchitosan mircrospheres crosslinked with poly(ethylene glycol) bisglycidyl ether

Carboxymethylchitosan microspheres crosslinked with poly(ethylene glycol) bisglycidyl ether were prepared and then tested as an adsorbent for selective removal of low-density lipoprotein (LDL) in human plasma. The microspheres were formed by a method of electrostatic instillation and crosslinked with poly(ethylene glycol) bisglycidyl ether. FTIR spectral analyses and X-ray photoelectron spectroscopy revealed that carboxymethylchitosan was crosslinked through amino groups to poly(ethylene glycol) bisglycidyl ether. The plasma lipoprotein sorption tests showed that the adsorption properties of the crosslinked microspheres for LDL were dependent on the concentrations of carboxymethylchitosan and poly(ethylene glycol) bisglycidyl ether. When the concentrations of carboxymethylchitosan and poly(ethylene glycol) bisglycidyl ether were 3.5% and 6%, respectively, 40% LDL and lower than 10% high density lipoprotein in plasma could be removed and the adsorption could be reach an equilibrium in 30 min.     

Surface PEGylation via native chemical ligation

Native chemical ligation (NCL) is an emerging chemoselective chemistry that forms an amide bond by trans-thioesterification followed by intramolecular nucleophilic rearrangement between thioester and cysteine. The reaction is simple, occurs in a mild aqueous solution, and gives near-quantitative yields of a desired product. Since the first report in 1994, most studies involving the use of NCL have focused on the total synthesis of proteins to address fundamental questions pertaining to many aspects of protein science, such as folding, mirror images, and site-specific labeling of proteins, but applications of the NCL reaction for other areas remain largely unexplored. Herein, we present a facile strategy for surface immobilization of poly(ethylene glycol) (PEG) utilizing the NCL reaction. Surface immobilization of PEG (i.e., PEGylation) plays a key role in preventing nonspecific protein adsorption on surfaces, which is crucial in a wide variety of medical devices. Using cysteine-PEG and thioester-containing phosphonic acid conjugates, we achieved efficient surface PEGylation on titanium surfaces. Ellipsometry, goniometry, and X-ray photoelectron spectroscopy (XPS) unambiguously confirmed the presence of PEGs, which provided nonfouling effects of surfaces. This study indicates that the NCL reaction will be a useful toolkit for surface bioconjugation and functionalization.     

Immunological properties of uricase conjugated to neutral soluble polymers.

For a comparative study of immunological properties of protein-polymer conjugates, uricase was modified with (a) poly(N-vinylpyrrolidone) 6000 Da, (b) poly(N-acriloylmorpholine) 6000 Da, (c) branched monomethoxypoly(ethylene glycol) 10000 Da, and (d) linear monomethoxypoly(ethylene glycol) 5000 Da. Spectroscopic studies performed by UV, fluorescence, and circular dichroism did not show any relevant difference in protein conformation among the native and the conjugates. Immunological studies showed that both uricase antigenicity and immunogenicity were altered by polymer conjugation to an extent that depended upon the polymer composition; in particular, monomethoxypoly(ethylene glycol) 10000 Da remarkably reduced the protein antigenicity, while unexpectedly, the poly(N-vinylpyrrolidone) derivative presented higher antigenicity than the native protein. In Balb/c mice, the native protein elicited a rapid and intense immunoresponse whereas all the conjugates induced a lower production of anti-native uricase antibodies. The rank order of immunogenicity was native uricase > uricase-poly(N-vinylpyrrolidone) > or = uricase-poly(N-acriloylmorpholine) > uricase-monomethoxypoly(ethylene glycol) 5000 Da > uricase-monomethoxypoly(ethylene glycol) 10000 Da. The four conjugates also induced anti polymer immunoresponse. Anti poly(N-vinylpyrrolidone) and anti poly(N-acriloylmorpholine) antibodies were generated from the first immunization while low levels of anti polymer antibodies were found with both poly(ethylene glycol) conjugates only after the second immunization.     

Separation of PEGylated alpha-lactalbumin from unreacted precursors and byproducts using ultrafiltration

There is considerable clinical interest in the use of "second-generation" therapeutic proteins produced by conjugation of the native protein with various polymers including poly(ethylene glycol) (PEG). One of the challenges in the production of polymer-protein conjugates is the need to remove residual polymer, native (unreacted) protein, and any reaction byproducts from the final therapeutic formulation. The overall objective of this study was to evaluate the possibility of using ultrafiltration for the purification of a model PEGylated protein. Sieving data were obtained using PEGylated alpha-lactalbumin, the native protein, and the poly(ethylene glycol) over a range of pH, ionic strength, and filtrate flux using both neutral and charge-modified composite regenerated cellulose membranes. Purification of the PEGylated protein was achieved using a two-stage diafiltration process. The first stage used a neutral membrane to remove the unreacted protein and any small reaction byproducts while retaining the large PEGylated product. The second stage used a negatively charged membrane to remove the neutral poly(ethylene glycol) while retaining the PEGylated alpha-lactalbumin as a result of strong electrostatic interactions. These results clearly demonstrate the potential of using membrane-based separations for the purification of second-generation therapeutic proteins.     

Segregation of modified bacteriorhodopsin aggregations in reconstituted vesicle membrane induced by the change of thermodynamical parameters

It was clearly shown that the change in thermodynamical parameters could cause the segregation of membrane protein aggregations in the phospholipid membrane. At first, reconstituted vesicles were prepared with a membrane protein, bacteriorhodopsin and a constituent phospholipid of biomembranes, L-alpha-dimyristoyl phosphatidylcholine. When the temperature of the suspension was decreased or the osmotic pressure was increased by adding poly(ethylene glycol) to this vesicle suspension at 23 degrees, the circular dichroism spectra showed a typical band indicating bacteriorhodopsin trimer formation implying their aggregation. This suggests that the aggregation of trimers proceeded by adding poly(ethylene glycol) into vesicle suspension, just as it proceeded by decreasing the temperature. Next, vesicles were prepared with fluorescein isothiocyanate-labeled bacteriorhodopsin, photoemissive bacteriorhodopsin and L-alpha-dimyristoyl phosphatidylcholine. The excitation energy transfer between the two modified proteins was measured by fluorescence spectroscopy. In this case, however, when poly(ethylene glycol) was added into the suspension, the yield of the excitation energy transfer decreased. This result indicates that modified proteins aggregate separately in a segregated form in the vesicle membrane.