On the reproducibility of band position in electrophoretic separations

Uncertainty Estimation for Quantitative Agarose Gel Electrophoresis of Nucleic Acids

This paper considers the evaluation of uncertainty of quantitative gel electrophoresis. To date, such uncertainty estimation presented in the literature are based on the multiple measurements performed for assessing the intra- and interlaboratory reproducibility using standard samples. This paper shows how to estimate the uncertainty in cases where we cannot study scattering components of the results. The first point is dedicated to a case where we have standard samples (the direct expressions are shown). The second point considers the situation when standard samples are absent (the algorithm for estimating the lower bound for uncertainty is discussed). The role of the data processing algorithm is demonstrated.

Identification of non-idealities in gel electrophoresis

The molecular weight separation, which is the second dimension of two-dimensional gel electrophoresis, is studied quantitatively with the goal of improving positional predictability and reproducibility. Mathematical modeling of carrier electrolyte dynamics is used to track the progress of a stacking front as a function of coulombs passed. In all test cases, the front moves more slowly than predicted and shows both curvature and tilt. These systematic deviations from the model are found to be influenced by a variety of factors, including both design and operating features. These factors are largely explained, and suggestions are made for improvements.

Biochemical changes during regeneration of sunflower (Helianthus annuus L.)

The main developmental stages in Helianthus annuus organogenesis have been studied in the sunflower hybrid "Giove". Shoot regeneration was obtained with high efficiency from mature seed cotyledons. Two-dimensional electrophoresis of protein extracts as well as the isozyme patterns of acid phosphatase, alcohol dehydrogenase, esterase, gluconate-6-phosphate dehydrogenase and phosphoglucomutase were compared during growth, callusing and regeneration. Two-dimensional protein patterns were similar, although polypeptides specific for each developmental phase could be identified. Different 2,4-dichlorophenoxyacetic acid concentrations or the sampling of specific regions of the seed did not result in significant differences in protein patterns. The activity of alcohol dehydrogenase and phosphoglucomutase appeared very low. For gluconate-6-phosphate dehydrogenase no difference, related either to the genotype or to different morphological stages, could be observed; the expression of acid phosphatase varied in a nonsystematic fashion. The isozyme pattern of esterase was related to the genotype as well as to the morphogenic phase.

Interlaboratory reproducibility of yeast protein patterns analyzed by immobilized pH gradient two-dimensional gel electrophoresis

An interlaboratory comparison was conducted on the positional and quantitative reproducibility of yeast proteins resolved by two-dimensional polyacrylamide gel electrophoresis (2-D PAGE) using isoelectric focusing with immobilized pH gradient (pH 4-7) in the first dimension. The basic experimental set-up was as follows: one laboratory prepared and distributed a [35S]methionine-labeled total yeast protein extract (Göteborg, Sweden), another laboratory prepared the IPG strips to be used by all labs in this study (Munich, Germany), the third laboratory (Aarhus, Denmark) circulated the protocols and coordinated the modest attempts to unify them. Samples were run horizontally in the first dimension and vertically in the second. The gels were sent to Göteborg for processing by phosphoimager technology and computerized image analysis (PDQuest), and the 2-D PAGE resolved proteins were located and quantified automatically. A subset of 470 spots was manually matched in all gels out of an average of 1328 resolved proteins. The positional interlaboratory comparison revealed great pattern reproducibility, the correlation coefficient in no case being less than 0.9994. In absolute terms an average deviation of 2.8 mm (x-position) and 1.8 mm (y-position) were obtained for all nine gels (three gels per lab). The interlaboratory comparison of protein quantitation displayed higher variability, and the best correlation coefficient generated was 0.975. An average standard deviation of 34.5% was calculated for protein quantitation including all three labs, a value slightly higher than the intralaboratory variation (range 20-28%). Thus, despite differences in protocols, chemicals and equipment, the immobilized pH gradient technology gave extremely high positional and quantitative reproducibility. This will greatly facilitate the exchange of data and the establishment of multi-user image-based 2-D gel databases.

Isoelectric focusing as a tool for the investigation of post-translational processing and chemical modifications of proteins

It has been demonstrated that good agreement may be observed between computed and experimental isoelectric point (pI) values when proteins of known sequence are focused under denaturing conditions on immobilized pH gradient IPG slabs, at least in the pH range 4-7.5. Hence, discrepancies between expected and found in this experimental set-up may be reliably ascribed to some kind of post-transcriptional processing, or chemical modification, having taken place in the sample. This evaluation is made easier when the comparison is set between the pI of a parent molecule and that (or those) of one to several of its derivatives as resolved in a single experiment (for instance, as a spot row in two-dimensional maps); no previous knowledge is required in these cases about the amino acid composition of the primary structure. The effects on protein surface charge are discussed in this review mainly for two biologically relevant processes, glycosylation and phosphorylation. Then, the pI shifts are analysed for some protein modifications that may occur naturally but can also be artefactually elicited, such as NH2 terminus blocking, deamidation and thiol redox reactions. Finally, carboxymethylation and carbamylation are used to exemplify chemical treatments often applied in connection with electrophoretic techniques and involving charged residues. Procedures to be applied in order to verify whether a given modification has occurred, and often relying on the focusing of a treated specimen, are detailed in each section. Numerical examples on model proteins are also discussed. As an important field of application of the above concepts may be genetic engineering, an exhaustive bibliographic list dealing with pI evaluation and structural assessment on recombinant proteins is included.

Two-dimensional electrophoresis of proteins: an updated protocol and implications for a functional analysis of the genome

The two-dimensional electrophoresis (2-DE) technique developed by Klose in 1975 (Humangenetik 1975, 26, 211-234), independently of the technique developed by O'Farrell (J. Biol. Chem. 1975, 250, 4007-4021), has been revised in our laboratory and an updated protocol is presented. This protocol is the result of our experience in using this method since its introduction. Many modifications and suggestions found in the literature were also tested and then integrated into our original method if advantageous. Gel and buffer composition, size of gels, use of stacking gels or not, necessity of isoelectric focusing (IEF) gel incubation, freezing of IEF gels or immediate use, carrier ampholytes versus Immobilines, regulation of electric current, conditions for staining and drying the gels - these and other problems were the subject of our concern. Among the technical details and special equipment which constitute our 2-DE method presented here, a few features are of particular significance: (i) sample loading onto the acid side of the IEF gel with the result that both acidic and basic proteins are well resolved in the same gel; (ii) use of large (46 x 30 cm) gels to achieve high resolution, but without the need of unusually large, flat gel equipment; (iii) preparation of ready-made gel solutions which can be stored frozen, a prerequisite, among others, for high reproducibility. Using the 2-DE method described we demonstrate that protein patterns revealing more than 10 000 polypeptide spots can be obtained from mouse tissues. This is by far the highest resolution so far reported in the literature for 2-DE of complex protein mixtures. The 2-DE patterns were of high quality with regard to spot shape and background. The reproducibility of the protein patterns is demonstrated and shown to be thoroughly satisfactory. An example is given to show how effectively 2-DE of high resolution and reproducibility can be used to study the genetic variability of proteins in an interspecific mouse backcross (Mus musculus x Mus spretus) established by the European Backcross Collaborative Group for mapping the mouse genome. We outline our opinion that the structural analysis of the human genome, currently pursued most intensively on a worldwide scale, should be accompanied by a functional analysis of the genome that starts from the proteins of the organism.

Positional reproducibility of protein spots in two-dimensional polyacrylamide gel electrophoresis using immobilised pH gradient isoelectric focusing in the first dimension: an interlaboratory comparison

An intra- and interlaboratory comparison of positional reproducibility of protein spots in two-dimensional electrophoresis using immobilised pH gradients (IPG) in the first dimension (IPG-DALT) was made. Aliquots of two different samples, human cardiac and barley leaf proteins, were separated in two different laboratories (London and Munich), using 180 mm long IPG gel strips, pH 4-8, for the first dimension and homogeneous SDS-PAGE gels (12% T) for the second dimension. Subsets of 340 (cardiac) and 200 (barley) well-resolved spots distributed across the 2-D gel patterns were selected for computer analysis (PDQUEST) of positional reproducibility. The IPG-dimension was highly reproducible in each laboratory, with a mean standard deviation of about 1 mm for both types of sample. Interlaboratory comparisons revealed identical results for barley with a mean standard deviation along the x-axis of about 1 mm, whereas the cardiac matchset showed slightly more variability (mean standard deviation approximately 1.5 mm). Nevertheless, IPG-DALT provides significantly improved reproducibility of spot positions compared to conventional isoelectric focusing with synthetic carrier ampholytes.

Reference points for comparisons of two-dimensional maps of proteins from different human cell types defined in a pH scale where isoelectric points correlate with polypeptide compositions

A highly reproducible, commercial and nonlinear, wide-range immobilized pH gradient (IPG) was used to generate two-dimensional (2-D) gel maps of [35S]methionine-labeled proteins from noncultured, unfractionated normal human epidermal keratinocytes. Forty one proteins, common to most human cell types and recorded in the human keratinocyte 2-D gel protein database were identified in the 2-D gel maps and their isoelectric points (pI) were determined using narrow-range IPGs. The latter established a pH scale that allowed comparisons between 2-D gel maps generated either with other IPGs in the first dimension or with different human protein samples. Of the 41 proteins identified, a subset of 18 was defined as suitable to evaluate the correlation between calculated and experimental pI values for polypeptides with known composition. The variance calculated for the discrepancies between calculated and experimental pI values for these proteins was 0.001 pH units. Comparison of the values by the t-test for dependent samples (paired test) gave a p-level of 0.49, indicating that there is no significant difference between the calculated and experimental pI values. The precision of the calculated values depended on the buffer capacity of the proteins, and on average, it improved with increased buffer capacity. As shown here, the widely available information on protein sequences cannot, a priori, be assumed to be sufficient for calculating pI values because post-translational modifications, in particular N-terminal blockage, pose a major problem. Of the 36 proteins analyzed in this study, 18-20 were found to be N-terminally blocked and of these only 6 were indicated as such in databases. The probability of N-terminal blockage depended on the nature of the N-terminal group. Twenty six of the proteins had either M, S or A as N-terminal amino acids and of these 17-19 were blocked. Only 1 in 10 proteins containing other N-terminal groups were blocked.

A nonlinear wide-range immobilized pH gradient for two-dimensional electrophoresis and its definition in a relevant pH scale

A new nonlinear immobilized pH gradient (IPG) is proposed as the first dimension for two-dimensional electrophoresis. In comparison to conventional carrier ampholyte techniques, it offers better resolution and greater reproducibility whilst allowing application of higher protein loads. Furthermore, we have checked and supplemented existing data on pK values for the immobilized groups in the presence of 8M urea. This is necessary for pH gradients to be defined in a pH scale relevant to the focusing conditions such that spot positions can be related to amino acid compositions. The data will allow definition of pH scales for the temperature range 10-25 degrees C and for a pH range covering the major part of the nonlinear pH gradient. With the latter, focusing positions are neither influenced by urea concentration nor by the choice or the concentration of detergent or carrier ampholyte. Temperature is the only parameter affecting focusing reproducibility and here any changes in focusing positions can be related to the amino acid compositions of peptides.

Biomedical relevance of two-dimensional protein mapping

State-of-the-art and future perspectives are discussed for the application of two-dimensional protein maps to basic medical research and routine clinical chemistry problems. Despite the technical advances that allow effective processing of a large number of samples and the refinement of devices and procedures for image analysis, at present two-dimensional maps are mostly confined to research purposes, i.e. to the inventory of normal constituents of body fluids and tissues on the one hand, and to qualitative-quantitative alterations of some protein spots in a number of instances (genetic, degenerative, infectious or xenobiotic diseases) on the other. It is hoped that in some instances a single primarily affected component will be able to be identified and then specifically tested (for instance by immunological means) as a diagnostic marker, but complex pathological patterns would still require the analysis of a large number of peptides at the resolution level only afforded by two dimensions. Further simplification of the protocols, for example with ready-made gels, and data reduction systems might then allow the application of the technique to be extended to general clinical laboratories.

Biochemical characterization by two-dimensional electrophoresis of lymphocyte antigens involved in cell-to-cell or cell-to-matrix adhesion

We have exemplified three cases of application of two-dimensional (2-D) electrophoresis to the characterization of lymphocyte membrane antigens. We could show that the proteins recognized by two monoclonal antibodies, LAK1 and LAK2, on the surface of large granular cells mediating natural- and lymphokine-activated killing are distinct molecules. LAK1 is expressed without any structural modification, even on the surface of endothelial cells. Another membrane antigen, recognized by the monoclonal antibody FB12, was shown to have the overall structure of the integrins of the very late activation (VLA) class, being composed of an alpha and of a beta subunit. The latter corresponded to the beta 1 type as already characterized for other VLAs, whereas the alpha chain was different from alpha 1 through alpha 6. The 2-D protocol using immobilized pH gradients for the first dimension allows reliable assessment of the identity of individual components because of the reproducibility of the absolute coordinates for spot position.

Capillary zone electrophoresis for monitoring r-DNA protein purification in multi-compartment electrolysers with immobiline membranes

Isoforms of human monoclonal antibodies against the gp-41 of AIDS virus and of human recombinant superoxide dismutase have been purified to homogeneity by isoelectric focusing (IEF) in a multi-compartment electrolyser with isoelectric, immobiline membranes. This system allows the processing of large sample volumes and gram-scale protein loads and can resolve isoforms as close as 0.001 in pI difference. The purification progress was usually monitored by analytical IEF in immobilized pH gradients (IPG). Capillary zone electrophoresis (CZE) was applied to the monitoring of the content of each chamber of the electrolyser. CZE was found to be superior in terms of speed of analysis and quantification (but only by UV reading at 200-210 nm, i.e., in the region of the peptide bond) but, notwithstanding the millions of theoretical plates reported, was no match for the resolving power of IPGs, at least for protein analysis. When compared also with chromatofocusing, the resolving power decreases in the order IPG greater than CZE much greater than chromatofocusing.

Two-dimensional maps in the most extended (pH 2.5-11) immobilized pH gradient interval

In conventional isoelectric focusing in soluble, amphoteric buffers, it has been quite difficult to produce two-dimensional (2-D) separations in pH intervals greater than pH 4-8. In general more alkaline proteins were analyzed by non-equilibrium IEF in the first dimension. Even with the advent of immobilized pH gradients (IPG), separations could be extended to pH gradients not wider than pH 3-10, due to a lack of suitable buffers. Since more acidic and more alkaline acrylamido buffers have recently been synthesized, we have been able to optimize what is believed to be the widest possible immobilized pH gradient, a pH 2.5-11 span. We report here for the first time 2-D separations of total tissue lysates in such extended pH 2.5-11 gradients. It appears that, with the IPG technique, close to 100% of all possible cell products can be displayed in a single 2-D map.

Analysis by high-resolution two-dimensional electrophoresis of differentiation-dependent alterations in cytosolic protein pattern of HL-60 leukemic cells

HL-60 leukemic cells were differentiated along the neutrophilic pathway with retinoic acid (RA) or along the monocytic pathway with 1,25-dihydroxyvitamin D3 (1,25(OH)2D3). Using a high-resolution two-dimensional electrophoresis technique and subsequent silver staining, differentiation-dependent changes in cytosolic protein pattern of HL-60 cells were analysed and were compared with the cytosolic protein pattern of human neutrophils. The amount of 64 and 50 out of a total of 632 proteins studied was increased or decreased in RA- and 1,25(OH)2D3-differentiated HL-60 cells, respectively, in comparison to undifferentiated HL-60 cells. Thirty-three of these proteins were similarly altered in RA- and 1,25(OH)2D3-differentiated HL-60 cells. Twenty-two and 25 of the proteins altered in amount in RA- or 1,25(OH)2D3-differentiated HL-60 cells versus undifferentiated HL-60 cells were similarly altered in human neutrophils in comparison to undifferentiated HL-60 cells. Seven and 10 of the proteins altered in amount in RA- or 1,25(OH)2D3-differentiated HL-60 cells had specific equivalents in neutrophil cytosol. Our results show (i) that neutrophilic and monocytic differentiation is associated with decreases and increases in amount of cytosolic proteins; (ii) that both differentiation processes share a common set of alterations; and (iii) are associated with specific alterations in protein amount.

Two-dimensional maps in very acidic immobilized pH gradients

Up to the present time it has been impossible to perform two-dimensional (2-D) separations in very acidic immobilized pH gradients (IPG), due to the lack of suitable buffering acrylamido derivatives to be incorporated into the polyacrylamide matrix. The advent of the pK 3.1 buffer (2-acrylamido glycolic acid; Righetti et al., J. Biochem. Biophys. Methods 16, 1988, 185-192) allowed the formulation of such acidic gradients. We report here separations in IPG pH 2.8-5.0 intervals of polypeptide chains from total lysates of rat intestinal and liver cells and 30S and 50S ribosomal proteins from Halobacterium marismortui. Conditions are given for highly reproducible first and second dimension gels and for a proper silver staining of 2-D maps with practically no background deposition.

Isoelectric focusing of alpha-1 acid glycoprotein (orosomucoid) in immobilized pH-gradients with 8M urea: detection of its desialylated variants using an alkaline phosphatase-linked secondary antibody system

A method allowing a clear separation of the different variants of desialylated alpha 1-acid glycoprotein (orosomucoid) has been developed using isoelectric focusing in immobilized pH gradients, supplemented with 8 M urea and 2% v/v 2-mercaptoethanol. Immunoblotting with two antibody-steps afforded high sensitivity and permitted the detection of about 700 pg of alpha 1-acid glycoprotein in a 20 microL plasma sample diluted 1:28 672. A one year old bloodstrain, kept at room temperature, could easily be phenotyped.

Orosomucoid (alpha-1 acid glycoprotein) phenotyping by use of immobilized pH gradients with 8 M urea and immunoblotting. A new variant encountered in a population study

Orosomucoid (ORM) phenotyping has been performed on 329 unrelated Swiss subjects, using immobilized pH gradients with 8 M urea and 2% v/v 2-mercaptoethanol followed by immunoblotting. After desialylation the band patterns of ORM confirmed that the polymorphism of the structural locus ORM1 is controlled by three codominant autosomal alleles (ORM1*F1, ORM1*S and ORM1*F2). One rare and one new allele were detected. The rare variant, tentatively assigned to the second structural locus ORM2, is observed in a cathodal position and named ORM2 B1. The new variant, tentatively assigned to the first structural locus ORM1, is observed in a region located between ORM1 S and ORM1 F2, and named ORM1 F3. Moreover, the pI values of the ORM variants have been measured accurately with Immobiline Dry Plates (LKB): they were found to be within the pH range 4.93-5.14.

Casting immobilized pH gradients into cylindrical polyacrylamide gels

A novel method is described for casting immobilized pH gradients in polyacrylamide gel rods of small diameter (2 mm), based on the principle of rotational centrifugation. The tubes are filled vertically with equal volumes of dense and light solution (250 microliter each) titrated to the extremes of the desired pH gradient, and then tilted at 2.5 degrees to the level. After 5 min at rest, to allow for sliding of the two menisci to equilibrium position, the glass tubes are rotated for 3 min at 180 rpm, followed by an additional 3 min at 180 rpm by reversing the sense of rotation. A homogeneous linear gradient is thus produced. The rotating platform is then raised to 90 degrees and the gels allowed to polymerize under standard conditions. Formation of linear and reproducible pH gradients is ensured by using stabilizing density gradients of low viscosity (0-5% glycerol, having a maximal ratio viscosity/density of 1.1).

A motor-driven syringe-type gradient maker for forming immobilized pH gradient gels

A motor driven gradient maker based on the commercial model (Jule Inc., Trumbull, CT) was designed for immobilized pH gradient gels to provide small volumes, rapid stirring and delivery, strict volume and temperature control and air exclusion. The device was constructed and by a convenient procedure yields highly reproducible gradients either in solution or on polyacrylamide gels.

Properties of thin-rod immobilized pH gradients

Immobilized pH gradient gel rods, 1.5 mm in diameter, were cast with a manifold connected to high-precision burettes. The reproducibility of gel length was ca. 1.7 mm. The average standard deviation sigma x for spot position was 2 mm after one-dimensional and 5.8 mm after two-dimensional runs. In order to bring to completion the elution of the salt fronts into the electrode compartments, carrier ampholytes had to be included in the gel formulation at concentrations of at least 0.5-1%, depending on the pH range. The presence of carrier ampholytes, however, was troublesome in two respects: the gel tended to shrink and the cathodic bands drifted with time. Ionic components in the sample were tolerated up to the following concentrations: NaCl 8 mumoles, sodium dodecyl sulfate 10 micrograms per tube. In presence of non-ionic detergents, the gels moved as a whole towards the cathode.

Additives for immobilized pH gradient two-dimensional separation of particulate material: comparison between commercial and new synthetic detergents

We describe the synthesis of two detergents, L and A15, whose performances as solubilizing agents and as additives in the first-dimension step of a two-dimensional separation are compared with those of some commercial compounds, i.e., Nonidet P-40, 3-[(3-cholamidopropyl)dimethylammonio]propanesulfonate(Chaps), and sulfobetaine, on three membrane protein preparations: rat RBC ghosts, beef kidney microvilli, and spinach thylakoids. L is 3-]3-dodecylamidoprophylcbdimethylammonio propane-1-sulfonate; owing to the substitution of a dodecylamido for the dodecyl residue of SB 3-12, the concentration of urea compatible with 2% detergent increases from 4.5 M for the parent molecule up to 7 M. With all three biological samples on which the panel of different detergents has been tested in parallel, L + urea scores as the most effective solubilization medium. On red blood cells a notable qualitative difference is observed with the selective extraction by L as well as by N-dodecyl-N,N-dimethylammonio-3-propanesulfonate of a major protein (pI = ca. 5.5, Mr = ca. 100,000). A15 is derived from a tertiary amine, with one alkylic substituent (either C11 or C13) and two poly(ethylene oxide) tails (totaling 15 ethoxy residues), which is reacted with propane sultone. Approximately 30% of the product corresponds to the N-adduct and is a truly zwitterionic detergent, while 60% is an O-derivative and still contains a titratable amino group (with a pK of 7.2). A15 can thus be used for isoelectric focusing on immobilized pH gradients, as in this work, but would not be compatible with carrier ampholyte isoelectric focusing.(ABSTRACT TRUNCATED AT 250 WORDS)

Two-dimensional gel electrophoresis of proteins

The high-resolution capacity of two-dimensional polyacrylamide gel electrophoresis (2-D PAGE) makes it an excellent tool for the analysis and characterisation of complex protein mixtures. The evolution of two-dimensional electrophoresis is briefly described. The various steps involved in 2-D PAGE, the identification and characterization of proteins separated by 2-D PAGE and the quantitative and qualitative analysis of 2-D patterns are discussed in detail and some new approaches are described. In the final section a brief outline of some of the biomedical applications of 2-D PAGE to screening of body fluids, genetic diseases, inborn errors of metabolism, cancer and neoplastic transformation are discussed.

Hydrophobic interaction between alkaline immobilines and ferritin during isoelectric focusing in immobilized pH gradients

It has been found that three alkaline Immobilines (out of seven weak acids and bases used to generate immobilized pH gradients), having pK values of 6.2, 7.0 and 9.3, act as cross-linking agents, aggregating and precipitating out of solution ferritin and other large macromolecules (e.g., from serum and tissue extracts) present in body fluids and human biopsies. All the acidic Immobilines (pK 3.6, 4.4 and 4.6) and the basic species of pK 8.5 appear to be unreactive. The three precipitin Immobilines mimic cationic detergents, acting on the basis of two different principles at the opposite extremes, by ionic interaction at one end and by hydrophobic bonding at the other end of the molecule. The ionic type of interaction was clearly demonstrated, owing to its sensitivity to pH extremes and to progressively increasing ionic strength. The hydrophobic interaction in the region of the double bond (Immobilines are N-substituted acrylamido acids and bases) was deduced on the basis of the following observations: (a) oxidation of the double bond with introduction of a vicinal diol totally inhibited ferritin aggregation; (b) addition of SH groups to the double bond increased protein precipitation and (c) the protein-Immobiline aggregates were found to be sensitive to alkyl-substituted ureas (especially ethyl- and propylurea), which are known to bind to hydrophobic regions of proteins, and insensitive to urea, which is known to split only hydrogen bonds. Interestingly, neutral and zwitterionic detergents were unable to split the Immobiline-ferritin complexes, suggesting that their large micelles could not have access to the tightly packed Immobiline cross-linking region.