Affinity partitioning of proteins in aqueous two-phase systems containing polymer-bound fatty acids. I. Effect of polyethylene glycol palmitate on the partition of human serum albumin and alpha-lactalbumin

Protein partitioning in two-phase aqueous polymer systems

Theories of protein partitioning in two-phase polymer systems which account for the effects of different aspects of system composition-such as the choice of materials, protein size, polymer molecular weight, polymer concentration, salt concentration, and affinity ligands-are reviewed. Although the present models provide some information about specific aspects of partitioning, a comprehensive and fundamental theory which can be used to predict protein partitioning behavior has not yet been developed. Some recommendations for future work are given.

Aqueous two-phase systems for biomolecule separation

Over the past thirty years, aqueous polymer two-phase technology has evolved, both experimentally and theoretically, into a separation science with many useful applications in biomolecule purification and bioconversion. This paper summarizes the developments in the applications of aqueous two-phase systems to biotechnology. The main topics to be considered are the phase diagram and its characteristics, fundamentals of biomolecule partition, large-scale and multi-stage aqueous two-phase biomolecule purification, and extractive bioconversions. The first topic involves a discussion of the thermodynamics of aqueous polymer two-phase formation and how it is influenced by such factors as polymer molecular weight and concentration, temperature, and salt type and concentration. Next, the theoretical and experimental aspects of biomolecule partition in aqueous two-phase systems will be discussed in light of the factors which influence biomolecule partition: polymer concentration and molecular weight; temperature; salt type and concentration; the addition of charged, hydrophobic and affinity derivatives. Having reviewed the fundamentals of phase diagram formation and biomolecule partition, the next two topics are applications of aqueous two-phase technology. The first set of applications involve the large-scale extraction of proteins using one to three equilibrium stages and multi-stage purifications using countercurrent distribution, liquid-liquid partition chromatography and continuous countercurrent chromatography. The second application, and very promising area for future aqueous two-phase technology, is the extractive bioconversion which permits the simultaneous production and purification of a biomolecule.

Effects of specific binding reactions on the partitioning behavior of biomaterials

Affinity partitioning is a special branch of biomaterials separations using aqueous two-phase systems. It combines the capability of diverse biomolecules to partition in aqueous two-phase systems using the principle of biorecognition. As a result, the macromolecule exhibiting affinity for a certain ligand is transferred to that phase where the ligand is present. This chapter describes the present status of the theoretical background of this approach and the properties of various natural and artificial compounds which act as affinity ligands in liquid-liquid systems. The affinity partitioning of proteins (enzymes and plasma proteins), cell membranes, cells, and nucleic acids are described as typical examples. The results are discussed in terms of theoretical understanding and practical application.

Partitioning and concentrating biomaterials in aqueous phase systems

Aqueous phase separation is a general phenomenon which occurs when structurally distinct water-soluble macromolecules are dissolved, above certain concentrations, in water. The number of aqueous phases obtained depends on the number of such distinct macromolecular species used. Aqueous two-phase systems, primarily those containing poly(ethylene glycol) and dextran, have been widely used for the separation of biomaterials (macromolecules, membranes, organelles, cells) by partitioning. The polymer and salt compositions and concentrations chosen greatly affect the physical properties of the phases. These, in turn, interact with the physical properties of biomaterials included in the phases and affect their partitioning. Specific extractions of biomaterials can be effected by including affinity ligands in the systems. The phase systems can also be used to obtain information on the surface properties of materials partitioned in them; to study interactions between biomaterials; and to concentrate such materials.

Polymer-protein interactions in aqueous two phase systems: fluorescent studies of the partition behaviour of human serum albumin

This study describes the partitioning of fluorescent macromolecules in aqueous two-phase systems (ATPS) comprising phosphate salt and poly(ethylene glycol) of three different molecular masses (i.e. 1000, 1450 and 2000 Da). The impact of system assembly was studied with fluorescent macromolecules introduced in contact with either (i) first salt, then polymer or (ii) first polymer, then salt, or (iii) with both salt and polymer simultaneously. Native human serum albumin (HSA) and derivatives labelled with N-(iodoacetylaminoethyl)-5-naphthylamine-1-sulphonic acid (1,5-IAEDANS) were partitioned using selected ATPS. Partitioning behaviour was characterised by molecular rotational studies of recovered proteins based upon changes of depolarisation. Measurements were undertaken by steady-state fluorescence or time-decay fluorescence using a single-photon counting system. In addition, circular dichroism was used as a tool for the study of macromolecular secondary structure. Two discrete categories of stable molecular structure have been identified that exist irrespective of the phase environment. The findings form the basis for a discussion of polymer protein interactions and the molecular micro-environment of proteins in ATPS.

Two populations of the estrogen receptor separated and characterized using aqueous two-phase partitioning

Two populations of the rat uterine estrogen receptor (ER) were separated and characterized using aqueous two-phase partitioning. Countercurrent distribution of rat uterine cytosolic ER allowed rapid and efficient separation of two populations, one population partitioned preferentially into the top phase (T, K(obs) = 3-6) and the other into the bottom phase (B, K(obs) = 0.01-0.03). The majority of unoccupied cytosolic ER is in the T population. Upon estrogen binding and/or heating to 30 degrees C in vitro the T population is converted to the B population. The transition from T to B does not exclusively involve loss of heat shock protein 90 and does not alter the ligand binding ability of the steroid binding domain. Using the human ER steroid binding domain overproduced in Escherichia coli and the steroid binding domain generated by partial trypsinization of the rat uterine ER, we demonstrate that the characteristic distinguishing T and B populations is not localized to this domain alone but may be associated with the amino terminal half of the ER (the A/B and DNA binding domains). The T to B transition of the ER also occurs in human MCF-7 breast cancer cells upon treatment with estrogen at 37 degrees C.

Protein partitioning in weakly charged polymer-surfactant aqueous two-phase systems

The study includes partitioning of proteins in aqueous two-phase systems consisting of the polymer dextran and the non-ionic surfactant C12E5 (pentaethylene glycol mono-n-dodecyl ether). In this system a micelle-enriched phase is in equilibrium with a polymer-enriched phase. Charges can be introduced into the micelles by the addition of charged surfactants. The charge of the mixed micelles is easily varied in sign and magnitude independently of pH, by the addition of different amounts of negatively charged surfactant, sodium dodecyl sulphate (SDS), or positively charged surfactant dodecyl trimethyl ammonium chloride (DoTAC). A series of water-soluble model proteins (BSA, beta-lactoglobulin, myoglobin, cytochrome c and lysozyme), with different net charges at pH 7.1, have been partitioned in non-charged systems and in systems with charged mixed micelles or charged polymer (dextran sulphate). It is shown that partition coefficients for charged proteins in dextran-C12E5 systems can be strongly affected by addition of charged surfactants (SDS, DoTAC) or polymer (dextran sulphate) and that the effects are directly correlated to protein net charge.

Effects of ionic strength and pH on the binding of medium-chain fatty acids to human serum albumin

Binding equilibria for the interactions of the medium-chain fatty acid anions, laurate and myristate, with defatted human serum albumin have been investigated under varying environmental conditions such as ionic strength and pH. Since these ligands bind strongly to albumin (Kass approximately 10(7) M-1), conventional equilibrium dialysis is not a feasible method for these investigations. Consequently, we employed a dialysis method, allowing determination of very low concentrations of unbound ligand by measuring the rate of exchange of labelled ligand across a dialysis membrane under conditions of chemical equilibrium. Over a range of ionic strength, 8-68 mM, the binding of the first few molecules of laurate to albumin was weakened with increasing ionic strength, whereas the binding of subsequent molecules seemed to proceed independently of ionic strength. The binding of myristate by albumin, however, appeared to be independent of ionic strength in the observed range of concentrations. The influence of pH in the range 5.1-9.0 on the binding of the two fatty acid anions by albumin was more complicated. The first molecule of laurate appeared to bind with a slightly weaker affinity to albumin at low pH, compared to pH 7 and high pH, while the trends for the following molecules varied. The binding of myristate (irrespective of concentration) seemed to strengthen monotonously with pH, but this conclusion depends critically on the interpretation of the kinetic behaviour of the myristate anion. We have previously shown [Pedersen, A. O., Honoré, B. & Brodersen, R. (1990) Eur. J. Biochem. 190, 497-502] that the strength of binding of the first few molecules of the two fatty acid anions to albumin decreases with increasing temperature, whereas binding of subsequent molecules seems to proceed independently of temperature. We explain these findings as follows. The binding of the first few (3 or 4) molecules of the C12 laurate anion is clearly driven by formation of ionic bonds between the fatty acid anion and positively charged groups, such as lysine residues, in the albumin molecule, whereas the binding of subsequent molecules of laurate seems to depend more on hydrophobic interactions. In the case of the C14 myristate anion, the binding of the first few (only 1 or 2) molecules may depend on ionic forces, but binding of the following molecules of myristate seems to depend on hydrophobic interactions only.(ABSTRACT TRUNCATED AT 250 WORDS)

Interactions in affinity partition studied using fluorescence spectroscopy

Fluorescence titration has been used to determine the binding constant and number of binding sites for the textile triazine dye Procion Yellow HE-3G to lactate dehydrogenase from rabbit muscle (E.C. 1.1.1.27). Triazine dye was either free in solution or attached to one of the polymer carriers, polyethylene glycol or dextran. Titrations were performed in solutions of buffer, dextran, and polyethylene glycol. Aqueous two-phase systems composed of polyethylene glycol and dextran were prepared and the binding constant and number of binding sites for ligand polyethylene glycol-Procion Yellow to lactate dehydrogenase were determined in both upper and lower phases of these systems. Affinity partition of lactate dehydrogenase in a PEG-dextran system was also performed using PEG-Procion Yellow as ligand, and partition coefficients of lactate dehydrogenase showed good agreement with theoretical partition coefficients calculated from the binding constant and number of binding sites obtained from fluorescence titration. The advantage of using fluorescence titration to determine affinity of a polymer ligand for a protein is that measurement of binding strength can be made in the actual environment encountered by protein-ligand complex during the purification process.

A ligand-induced conformational change in the estrogen receptor is localized in the steroid binding domain

Upon binding estrogen, the estrogen receptor (ER) is proposed to undergo some form of conformational transition leading to increased transcription from estrogen-responsive genes. In vitro methods used to study the transition often do not separate heat-induced effects on the ER from estrogen-induced effects. The technique of affinity partitioning with PEG-palmitate was used to study the change in the hydrophobic surface properties of the ER upon binding ligand with and without in vitro heating. Upon binding estradiol (E2), the full-length rat uterine cytosolic ER undergoes a dramatic decrease in surface hydrophobicity. The binding of the anti-estrogen 4-hydroxytamoxifen (4-OHT) results in a similar decrease in surface hydrophobicity. These effects are independent of any conformational changes induced by heating the ER to 30 degrees C for 45 min. The use of the human ER steroid binding domain overproduced in Escherichia coli (ER-C) and the trypsin-generated steroid binding domain from rat uterine cytosolic ER demonstrates that the decrease in surface hydrophobicity upon binding E2 or 4-OHT is localized to the steroid binding domain. Gel filtration analysis indicates that the change in surface hydrophobicity upon binding ligand is an inherent property of the steroid binding domain and not due to a ligand-induced change in the oligomeric state of the receptor. The decrease in surface hydrophobicity of the steroid binding domain of the ER upon binding E2 or 4-OHT represents an early and possibly a necessary event in estrogen action and may be important for "tight" binding of the ER in the nucleus.

Affinity partitioning of enzymes using dextran-bound procion yellow HE-3G. Influence of dye-ligand density

The dye Procion Yellow HE-3G was bound to dextran of molecular weight 70,000 and the partitioning of this dye-polymer within an aqueous two-phase system containing dextran and poly(ethylene glycol) was studied as function of ligand density, polymer concentration, type of salt, concentration of salt and concentration of dye-dextran. Even moderate dye:dextran molar ratios (5-8) make the partitioning strongly salt-dependent. The dye-dextran can be directed to either the upper or the lower phase with partition coefficients from 0.02 to 28 by using salts. The dye-dextran in the two-phase system affects the partitioning of dye-binding enzymes (lactate dehydrogenase, glucose-6-phosphate dehydrogenase, 3-phosphoglycerate kinase) towards the dye-containing phase. Measurements of competition with nucleotide binding show an increased affinity of the dye for the enzyme with increasing ligand:dextran ratio. Theoretical considerations indicate that 1-2 dextran molecules are attached per enzyme molecule when affinity partitioning is fully developed.

Laurate binding to human serum albumin. Multiple binding equilibria investigated by a dialysis exchange method

Multiple binding of laurate (n-dodecanoate) to human serum albumin was studied by a kinetic dialysis method. With this method the concentration of unbound ligand can be determined by measuring the rate of exchange of radioactive label across a dialysis membrane under conditions of equilibrium. Determination of the rate constant of exchange, in the absence of albumin, revealed that laurate does not change its state of aggregation over a concentration range from 0.1 microM to 500 microM, indicating the prevalence of a monomer. Binding of laurate to albumin, at pH 7.5, 37 degrees C, was investigated through 220 determinations of binding equilibria in the range of 0-10 mol laurate/mol albumin, corresponding to a concentration range of unbound laurate from 1 nM to 0.1 mM. Binding data were analyzed in terms of stepwise binding. Considering the stochastic errors of the experimental data, we generated a variety of possible (equal goodness of fit) solutions to the stoichiometric binding equation. This analysis resulted in reasonably well-defined values for the first two step constants, with an indication of negative interaction. The variation of the higher step constants excluded any mechanistic conclusions, although the binding isotherms, defined by these varying sets of binding constants, fitted the experimental data well within a chosen probability limit.

Parameters determining affinity partitioning of yeast enzymes using polymer-bound triazine dye ligands

Triazine dyes, bound to polyethylene glycol, have been used to influence the partition of some enzymes within a dextran-polyethylene glycol-water two-phase system. The enzymes, present in a protein extract from baker's yeast, included glucose 6-phosphate dehydrogenase, glyceraldehyde phosphate dehydrogenase, 3-phospho-glycerate kinase and alcohol dehydrogenase. The partition coefficients of the enzymes could be changed by a factor of 10-500 in favour of the polyethylene glycol-rich phase, while the partition of bulk protein was much less affected. The influence of the concentration of polymer-bound dye and phase-forming polymers, temperature, pH, kind and concentration of salt and the presence of nucleotides on this affinity partitioning effect was studied. The extraction was effective even at high concentrations of dye and protein (40 g/l). A partial purification (32-fold) of glucose 6-phosphate dehydrogenase was carried out by an extraction in five steps.