Proteins

The way in which a protein molecule is distributed between the two phases in a given PEG/dextran system depends on its three-dimensional structure and general surface properties and is described by its partition coefficient, Kc (the ratio of the concentration of the molecule in the top phase to that in the bottom phase). Thus, partitioning in aqueous PEG/dextran

Figure 2 Influence of the support on LLPC of polyclonal IgG: (A) LiParGel, (B) LiChrospher, (C) Superdex. Data adapted from Wingren eta/., JournalofChromatography603, 73 (1992); Journal of Chromatography 668, 65 (1994), with permission from Elsevier Science.

Figure 2 Influence of the support on LLPC of polyclonal IgG: (A) LiParGel, (B) LiChrospher, (C) Superdex. Data adapted from Wingren eta/., JournalofChromatography603, 73 (1992); Journal of Chromatography 668, 65 (1994), with permission from Elsevier Science.

two-phase systems offers a unique means of separating proteins in solution with respect to their overall exposed surface properties. In addition, the method is mild so that the conformation of a partitioned protein is not likely to be disturbed.

As in the case of DNA separation, the supports may (strongly) influence the fractionation of proteins. This is illustrated in Figure 2, where LLPC on these supports resulted in different elution profiles for polyclonal IgG, i.e. the supports influenced the partitioning of im-munoglobulins in different ways. The exact nature of the observed interactions is still not clear, but they may be determined by the conformation of the molecules together with the properties of the support with respect to pore size and polyacrylamide/dextran coating. Although ideal partitioning is desirable, the general applicability of the matrices is not invalidated as the interactions can be exploited as previously mentioned.

Separation

LLPC can be used to separate a wide variety of proteins including enzymes, hormones, immunoglobulins, serum proteins and transport proteins, some of which are shown in Table 5. LLPC has been successfully used for separations that have previously been difficult, for instance the separation of molecules with similar molecular weight and pI values such as a-1-antitrypsin and albumin, and corticosteroid-binding globulin and a sex hormone-binding globulin. Moreover, LLPC may replace the need of other conventional separation techniques, since LLPC, in contrast to most of these methods, is able to separate proteins owing to differences in several physicochemical properties. As for example, proteins in human sera, which can be separated in a multistep procedure by using a combination of several other techniques such as precipitation, ion exchange chromatography, gel filtration and electrophoresis, can be separated in a single step by LLPC. Furthermore, proteins that can be separated by differences in pI values, as for example isoforms of lactate dehydrogenase and malate dehydrogenase, can be separated by LLPC. In addition, monomeric and dimeric forms of HSA that can be separated by gel filtration and proteins that may be separated by hydrophobic chromatography can also be separated by LLPC. Thus, LLPC in aqueous polymer two-phase systems offers new possibilities to separate proteins in a single step and/or to obtain fractionation that is not readily achieved by other techniques.

Fractionation

LLPC can be used to fractionate proteins and may provide a selectivity and sensitivity unattainable by any other chromatographic method. Some

Table 5 Separation of proteins by LLPC

Separation problem

Example

LLPC results/comments

Complex system

Proteins with similar Mr and pi Proteins with different pi

Proteins with different Mr Proteins with different hydrophobicity

General separations

Proteins in human serum

Proteins in goat, sheep, dog and horse serum

Proteins in mouse ascites a-1 antitypsin and HSA

CBG and SHBG from human serum

Isoforms of LDH Isoforms of MDH

Monomers and dimers of HSA Human IgG with different relative hydrophobicity

Myoglobin, peroxidase, BSA and OVA Cyt c, HSA, p-lactoglobulin, lysozyme, MB, ovotransferrin, OVA and peroxidase ADH, GDH and ferritin ADH, FDH and LDH Transferrin, IgG and HSA

A single step separation achieved; previously separated in a multistep procedure using a combination of several other methods

LLPC is able to separate proteins due to differences in 'several' physicochemical properties and may thus replace the need of many conventional separation techniques

ADH, alcohol dehydrogenase; BSA, bovine serum albumin; CBG, corticosteroid-binding globulin; Cyt c, cytochrome c; FDH, formate dehydrogenase; GHD, glucose 6-phosphate dehydrogenase; HSA, human serum albumin; LDH, lactate dehydrogenase; LLPC, liquid-liquid partition chromatography; MB, myoglobin; MDH, malate dehydrogenase; OVA, ovalbumin; SHBG, sex hormone-binding globulin.

representative examples, including enzymes and antibodies, are given in Table 6. The usefulness of LLPC for this purpose was highlighted in recent studies where several enzymes known to exist in equilibrium between two allosteric forms were analysed. Yeast alcohol dehydrogenase, liver alcohol dehydrogenase, the heart and muscle isoforms of lactate dehydrogen-ase and two types of hexokinase (fraction I and II of Kaji), are all eluted as two components by LLPC, in spite of the fact that the enzyme preparations appear

Table 6 Fractionation of proteins by LLPC

to be homogeneous when analysed by other conventional methods such as isoelectric focusing, sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and high performance liquid chromato-graphy-size exclusion (HPLC-SE). Remarkably, an equilibrium between the two components with different surface properties for each of these enzymes can be demonstrated by LLPC. These results show that LLPC can be used to detect and even separate different con-formational isometric forms of unliganded enzymes. It

Proteins

LLPC results/comments

LADH YADH LDH-H4

ldh-m4

Hexokinase3 sMDH

mMDH j

Polyclonal IgG ] Polyclonal IgA Polyclonal IgM J

Monoclonal IgG

An equilibrium between two components with different surface properties detected (i.e. conformational isomeric forms)

Two components with different surface properties detected

Several components detected

An equilibrium between at least two components with different surface properties detected (for six of 57 antibodies; conformational isomeric forms?)

aSimilar results were obtained for two types of hexokinase, fraction I and II of Kaji. LADH, liver alcohol dehydrogenase; mMDH, mitochondrial malate dehydrogenase; sMDH, cytoplasmic malate dehydrogenase; LDH-H4, heart lactate dehydrogenase; LDH-M4, muscle lactate dehydrogenase; LLPC, liquid-liquid partition chromatography; YADH, yeast alcohol dehydrogenase.

Figure 3 LLPC of 57 monoclonal IgG antibodies containing 95% IgG and homogeneous with respect to their physicochemical properties as determined by HPLC-SE, SDS-PAGE and IEF. The LLPC chromatograms are schematically illustrated in (A), which represents 51 of the IgGs, and (B), which represents six of the IgGs.

Figure 3 LLPC of 57 monoclonal IgG antibodies containing 95% IgG and homogeneous with respect to their physicochemical properties as determined by HPLC-SE, SDS-PAGE and IEF. The LLPC chromatograms are schematically illustrated in (A), which represents 51 of the IgGs, and (B), which represents six of the IgGs.

may be of interest to note that the LLPC fractionations of the mitochondrial and cytoplasmic isoforms of malate dehydrogenase, the allosteric mechanisms of which are still unclear, also result in two components with different surface properties.

Immunoglobulins are another group of proteins that have been successfully fractionated using LLPC (Table 6 and see below). Purified polyclonal immunoglobulins of IgG, IgM or IgA class can be fractionated into several components using LLPC. In the case of IgG and IgA, it is of a special interest to note that the fractionation is not subclass-related since the elution profile of each of the subclasses is similar to that of the applied polyclonal antibody population. Surprisingly, some monoclonal antibodies have been found to be fractionated into at least two components by LLPC, in spite of the fact that they are homogeneous with respect to their immunochemical and physicochemical properties (Figure 3). This phenomenon was observed not only for IgG antibodies but also for IgA and IgM myeloma proteins. Recent LLPC experiments have demonstrated that six of 57 monoclonal IgG antibodies seem to exist in an equilibrium between at least two components with different surface properties. Thus, LLPC seems to be able to detect and separate conformational isomeric forms of unliganded antibodies. It has been suggested (using different kinetic techniques) that a tenth of all antibodies may indeed display conformational isomerism.

Specificity

LLPC can be used to examine differences in surface properties of proteins (mainly immunoglobulins) related to their biological specificities; some illustrative examples are given in Table 7.

In early LLPC experiments, polyclonal rabbit IgG antibodies with different specificities were found to exhibit different partition properties (expressed as a partition coefficient, Kc). Since no IgG subclasses have been reported for rabbit IgG, i.e. their Fc parts are likely to be the same, these results indicated that LLC had detected differences in surface properties between IgGs located on structures in their Fab parts. This exciting finding was followed up by more extensive studies in which large sets of well-characterized monoclonal IgG antibodies were analysed by LLPC. These studies clearly showed that there was no correlation between Kc and a single immunochemical and physicochemical property such as light-chain and

Table 7 Examination of surface properties of proteins related to biological specificities using LLPC

Proteins

Example

LLPC results pAbs with different specificities mAbs with different affinities for the same Ag mAbs with identical affinity for one anti-IgG, anti-HSA and anti-transferrin

Protein Ag 5 mAbs against cytomegalovirus

5 mAbs against albumin Hapten Ag 3 mAbs against DNP

2 mAbs against T3

3 mAbs against T4

Hapten 2 mAbs against T3

2 mAbs against T4 Epitope 2 mAbs against IgGFc

3 chimaeric mAbs against NIP

Different K values

Different K values Different K values Different K values Different K values Different K values

Identical K values Identical Kc values Identical K values Identical K values pAbs, polyclonal antibodies; mAbs, monoclonal antibodies; Ag, antigen; HSA, human serum albumin; DNP, 2,4-dinitrophenol; T3, triiodothyronine; T4, thyroxine; NIP, 5-iodo-4-hydroxy-3-nitrophenacetyl; K, the ratio of the concentration of the molecule in the top phase to that in the bottom phase; LLPC, liquid-liquid partition chromatography.

• IgG 1

x IgG 2

.A IgG 3

□ IgG 4

a^^ A

Subclass Kc, Fc

IgG 1

1.22 ± 0.01

1.23 ± 0.01

1.21 ± 0.02

IgG 2

1.22 + 0.01

1.22 ± 0.01

1.21 ± 0.02

IgG 4

1.22 ± 0.01

1.22 ± 0.01

1.21 ± 0.02

Figure 4 Comparison of the partition properties of (A) intact IgG with those of the corresponding Fab fragments and (B) Fc fragments from one IgG subclass with those of the other subclasses. The partition properties are expressed as a partition coefficient, Kc. The correlation coefficient was (A) 0.80 (P < 0.05). (B) The 95% confidence limits of the K values are given.

heavy-chain isotypes, charge, size and shape. However, a significant linear correlation between the partition properties of intact IgG1, 2, 3 and 4 and their corresponding Fab fragments was detected (Figure 4). In contrast to the Fabs, Fc fragments from IgG1, 2 and 4 displayed almost identical surface properties (Figure 4). Taken together, the results showed that the differences detected by LLPC between intact IgGs were indeed located on their Fabs.

Further studies have revealed a remarkable relationship between the partition properties of an IgG molecule and the structure of its combining site (specificity; see Table 7). In particular, monoclonal IgG antibodies with different affinities for the same antigen or directed against different antigens are eluted with different Kc values, while monoclonal IgGs with identical affinity constants for the same hapten or the same epitope on a protein are found to have identical Kc values. Moreover, chimaeric anti-NIP antibodies with identical variable regions, corresponding to the human IgG1, 2 and 4 subclasses, display identical surface properties, i.e. differences in the constant part of the heavy chains did not affect the partitioning. Thus, the surface properties of IgGs are, as detected by LLPC, dominated by those of their antigen-binding sites. Recent LLPC experiments have indicated that the antigen-binding site dominance observed for IgG is valid for all the other Ig classes and subclasses provided that antibodies only within a given class or subclass are compared with respect to exposed surfaces. Indications of a dominance of ligand-binding sites for other specific proteins, such as enzymes, have been reported. Taken together, LLPC may offer a unique possibility to screen the antigen-binding sites for differences/similarities in exposed surfaces, even in those cases when we do not know the specificity of the antibodies.

This unique property of LLPC was made use of in a recent study that showed that the LLPC profile for polyclonal IgG antibodies isolated from the sera of patients with an autoimmune disease (primary Sjogren's syndrome) differed significantly from those of polyclonal IgG isolated from the sera of healthy individuals. These antibodies with 'unique and deviating' specificities could also be isolated by LLPC. Thus, LLPC may provide us with the means to fractionate and isolate specific antibodies 'of unknown specificities'.

Ligand Binding

Finally, LLPC can also be used to study the events upon binding of a ligand by specific proteins in solution. Some typical example, including transport proteins, enzymes and polyclonal as well as monoclonal antibodies, are given in Table 8. In many cases, LLPC can simply be used to separate the ligand-protein complex from either of the free components. This is, for example, the case for most of the enzyme-ligand and antibody-ligand complexes analysed so far.

In recent studies, the capacity of LLPC for analysis of protein-ligand complexes was demonstrated using well-characterized ligand-protein complexes where ligand-induced conformational changes are known to occur. In addition, the ligands in these model complexes are small and almost completely buried upon binding, i.e. the bound ligands themselves would not be responsible for any changes in surface properties observed upon complexation. Small changes in surface properties were detected when rabbit transferrin bound ferric ions, while no changes were observed when human transferrin (which is known to be less flexible) bound the ions. The ligand-induced transition between two conformational isomeric forms could be detected by LLPC for a large number of enzymes, including alcohol dehydrogenase, citrate synthase, glutamate-oxaloacetate transaminase, hexokinase, lactate dehydrogenase, malate dehydro-

Table 8 Examination of the events upon binding of ligand by specific proteins in solution using LLPC

Protein

Ligand

LLPC results/comments

Rabbit transferrin

Fe3#

Small changes in Kc detected

Human transferrin

Fe3#

No changes in Kc detected

CS

Oxaloacetate

Hexokinase

Glucose

GAPDH

NADH or NAD #

GOT

a-Methyl-aspartate

NADH or NADH#oxamate

NAD #

Ligand-induced transition between two conformational isomeric forms detected

mMDH

Citrate # NAD #

sMDH

Citrate # NAD #

PGK

GDP # 3-PGA

YADH

NADH or NAD # J

pAbs

IgG, HSA or transferrin

mAb

T3

Differences in Kc detected, may be interpreted in terms of conformational characteristics

9 chimaeric mAbs

NIP

5 mAbs

HSA

Differences in Kc detected; different mAbs formed AgAb with identical Kc

18 mAbs

5 haptens

Linear relationship between Kc of unliganded mAbs and their corresponding hapten-mAb complex

CS, citrate synthase; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; GOT, glutamate-oxaloacetate transaminase; HSA, human serum albumin; LADH, liver alcohol dehydrogenase; LDH-H4, heart lactate dehydrogenase; LDH-M4, muscle lactate dehydrogenase; mAbs, monoclonal antibodies; LLPC, liquid-liquid partition chromatography; mMDH, mitochondrial malate dehydrogenase; sMDH, cytoplasmic malate dehydrogenase; NIP, 5-iodo-4-hydroxy-3-nitrophenacetyl; pAbs, polyclonal antibodies; PGA, 3-phospho-glycerate; PGK, 3-phosphoglycerate kinase; T3, triiodothyronine; YADH, yeast alcohol dehydrogenase.

CS, citrate synthase; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; GOT, glutamate-oxaloacetate transaminase; HSA, human serum albumin; LADH, liver alcohol dehydrogenase; LDH-H4, heart lactate dehydrogenase; LDH-M4, muscle lactate dehydrogenase; mAbs, monoclonal antibodies; LLPC, liquid-liquid partition chromatography; mMDH, mitochondrial malate dehydrogenase; sMDH, cytoplasmic malate dehydrogenase; NIP, 5-iodo-4-hydroxy-3-nitrophenacetyl; pAbs, polyclonal antibodies; PGA, 3-phospho-glycerate; PGK, 3-phosphoglycerate kinase; T3, triiodothyronine; YADH, yeast alcohol dehydrogenase.

genase and 3-phosphoglycerate kinase. In the case of antibodies, conformational changes may occur upon binding of either hapten (T3 and NIP) or protein antigens (HSA, IgG and transferrin) that are detectable by LLPC. The surface properties of chimaeric anti-NIP antibodies with identical variable regions, corresponding to the human IgA1, IgA2, IgE, IgG1, IgG2, IgG3, IgG4 and IgM isotypes, have been compared before and after binding of hapten. The hapten-antibody complexes were found to be eluted in a considerably narrower range of Kc values than were the free antibodies. It was concluded that con-formational changes, detectable by LLPC, occurred in either IgA1, IgA2, IgE and/or IgM, but not in the IgGs, making the surfaces of the constant regions of the heavy chains of the Ig classes and subclasses more similar. Hence, LLPC may provide us with the means to examine whether ligand binding induces confor-mational changes in the antibody when in solution.

Almost all the antigen-antibody pair analysed by LLPC so far have formed complexes that are eluted as single homogeneous peaks by LLPC (Figure 5). Moreover, the complexes are eluted with the same value of Kc irrespective of the molar ratio of antigen to antibody at which they are formed (ranging from antigen to antibody excess). Thus, the results imply that each antigen-antibody pair forms one type of complex with respect to exposed dominant surfaces. This is thought to be the first study reporting such a feature for antigen-antibody complexes.

Remarkably, a linear relationship between the surface properties of unliganded IgGs and their corresponding hapten-IgG complexes has been reported, and it was concluded that the surface properties of

Ab AgAb

Figure 5 LLPC of antigen-antibody complexes (AgAb) formed at different molar ratios of antigen (Ag) to antibody (Ab) (Ag : Ab = 2 : 1-0.5 : 1). Representative results obtained for one monoclonal IgG antialbumin antibody are shown.

IgG are dominated by those of its antigen-binding sites even after the specific binding of hapten or hap-ten-carrier at the combining sites. By contrast, the surface properties of protein-antibody complexes are not related to those of the unliganded antibodies. Instead, the surface properties of protein-IgG complexes are related mainly to those of the antigens. Depending on the type of antigen (hapten or protein), LLPC may thus be used to separate antigen-IgG complexes through differences in exposed surfaces of either the antibody combining sites or the antigen. LLPC may thus provide us with new, interesting information concerning the events upon complex formation in solution and possible ligand-induced conformational changes.

See also: III/Nucleic Acids: Liquid Chromatography. Proteins: Ion Exchange. Appendix 1/Essential Guides for Isolation/Purification of Enzymes and Proteins. Essential Guides for Isolation/Purification of Im-munoglobulins. Essential Guides for Isolation/Purification of Nucleic Acids.

Solar Panel Basics

Solar Panel Basics

Global warming is a huge problem which will significantly affect every country in the world. Many people all over the world are trying to do whatever they can to help combat the effects of global warming. One of the ways that people can fight global warming is to reduce their dependence on non-renewable energy sources like oil and petroleum based products.

Get My Free Ebook


Post a comment