Our work aims at increasing the effectiveness of electrophoretic separations of biological macromolecules and subcellular particle-sized species by application of separation theory through development of computer programs and by instrumental and procedural optimization at a level that is practical for the vast majority of biochemical laboratories. We continued to work on improving the method of “direct electroelution” of proteins from polyacrylamide gels and that of “static noncovalent” coating of capillaries for reducing electroosmotic flow in capillary electrophoresis (CE) to a negligible value. Using rat synaptic vesicles as a model, we have been attempting to develop a novel approach to deciphering protein complexes based on their cross-linking and fractionation by capillary electrophoresis. The complexes are to be continuously blotted to membrane, identified as SNARE protein complexes, followed by a mass-spectrometric (MS) characterization of interacting proteins. We have also been attempting to improve the resolution at low gel concentration of fluorescently labeled saccharides by using acetate as the trailing constituent of a disc electrophoresis buffer system and to replace 2D-PAGE in proteomic analysis of a representative membrane protein by a three-stage procedure of SDS-PAGE, electroelution, IPG-IEF, and band transfer into MS.

Direct Electroelution of Protein Bands from SDS-PAGE
Radko, Zakharov, Chrambach; in collaboration with Chang, Bezrukov, Yergey, Vieira
We further developed direct protein elution from SDS-PAGE bands without gel sectioning to a sensitivity toward 0.1-0.2 mg of protein through use of a green laser to detect the SYPRO-red labeled protein, providing a five-fold increase in sensitivity over that previously achieved by side-ways UV illumination of the band. Electroelution time was inversely proportional to protein mobility; at 1 kV, it ranged from 1 to 12 minutes for various proteins to attain 80 percent or more electroelution from 1-mm-thick Novex gels. After three steps of ultrafiltration in water to deplete SDS, protein yield decreased to 25 to 33 percent for various proteins. To obtain a signal/noise ratio of six, a load of 2 picomoles of protein on SDS-PAGE was required. Instrument developments included replacement of the wick between electroelution tube and anolyte by direct connection of a large anolyte reservoir to the tube and a scaled-down procedure that, in lieu of the electroelution tube, employs a capillary of a custom-made CE apparatus equipped with a laser-induced fluorescence (LIF) detector. In addition to the transfer of electroeluate into MS, we demonstrated the transfer onto immobilized pH gradient/iso-electric focusing (IPG-IEF) strips.

Resolving Capacity of Capillary Electropho-resis as a Function of the Hydrophilicity of Adsorbed Static Coating
Chrambach; in collaboration with Cretich, Chiari, Stastna
To obviate electroosmotic flow, capillaries in CE are conventionally covalently coated internally with polyacrylamide. We were able to show previously that the noncovalent coating with epoxy-poly-dimethylacrylamide (EPDMA) is a labor- and time-saving alternative to that procedure. However, EPDMA coating is unstable at very low ionic strength and in the presence of SDS, suggesting a hydrophilic binding between the coating and the silanol of the inner capillary wall. Thus, we attempted to improve the coating by derivatizing EPDMA so as to make it more hydrophilic. We demonstrated that a moderate degree of such derivatization indeed gave rise to less protein adsorption onto the capillary wall (decreased peak asymmetry) and increased resolving capacity (decreased peak width). Excessive hydrophilicity, however, led to electrostatic repulsion between silanol and coating and therefore lability of the coating.

Deciphering Protein Complexes: Search for Rat Synaptic Vesicle Proteins Participating in Membrane Fusion
Radko; in collaboration with Blank, Zimmerberg
Over the last years, accumulated observations have indicated that Ca-dependent membrane fusion cannot be accounted for by the action of SNARE proteins alone; instead, fusion appears to depend on hitherto other unknown fusigenic proteins. We initiated a search for those proteins by using the small but abundantly available rat synaptic vesicle as a membrane preparation. We planned to develop a new approach to separation of large protein complexes that are obtained by treating synaptic vesicles with various cross-linking agents under conditions promoting vesicle fusion, followed by solubilization of the vesicles. The approach is based on capillary electrofocusing and zone electrophoresis of cross-linked complexes, combined with direct continuous elution and blotting of the separated peaks onto a membrane by means of electro-osmotic flow. The material collected on a blotting membrane will then be tested immunologically for the presence of a SNARE protein. After dissociation of the cross-links and/or digestion with trypsin, the immunoactive spot(s) should allow for the mass spectrometric identification of the SNARE-associated proteins. As a first step, we isolated rat synaptic vesicles by employing differential velocity centrifugation, followed by controlled pore glass (CPG) chromatography. We investigated the degree of homogeneity of the vesicles by capillary zone electrophoresis (CZE). The vesicle preparations before and after velocity sedimentation in sucrose gradients and those after CPG chromatography all exhibit distinguishable CZE peaks, suggesting distinct surface charge densities depending on particle size. We have found that the following CZE conditions reveal size differences: 0.02M polyethylene oxide (Mr 4 million) in the CZE buffer and a field strength of 30 kV. We are also searching for the in vitro conditions to promote Ca-dependent fusion (within physiological range of calcium concentrations) of the vesicles either in the homotypic manner or to target membranes such as liposomes.

Enhanced Gel Electrophoretic Resolution by Application of an Extensive Computer Output of Discontinuous Buffer Systems
Chrambach; in collaboration with Cabanes-Macheteau, Taverna, Berna, Ashburn, Wheeler
We achieved gel electrophoretic separation of small, highly charged saccharides in discontinuous buffer systems at low gel concentration (less than 5 percent T) by selecting acetate as the trailing constituent of the moving boundary “front” in lieu of the conventionally used glycinate or borate (Fig. 14).

Figure 14

Separation by PAGE of ANTS-labeled dextran homopolymers as a function of the trailing constituent of a discontinuous buffer system. The trailing buffer constituent is, left: acetate; right: glycinate. All other buffer constituents (leading constituent chloride; common constituent Tris) and the gel concentration (4.5 %T, 2.7 %CBis) are the same in both cases. The operative pH values are 8.78 and 9.4, 25°C, the ionic strength 0.0793 and 0.0153 for the acetate and glycinate systems, respectively. ANTS = 8-aminonaphthalene-1,3,6-trisulfonic acid.

That selection was based on the use of the output of a computer program of T. M. Jovin, Max Planck Institut für Biophysikalische Chemie, Göttingen, Germany, which was hitherto restricted to the NIH mainframe computer. To make this output and its benefits, as demonstrated by the improved resolution of the representative saccharide separation problem, more widely available, we made the discontinuous buffer systems across the entire pH scale, 0 and 25C, available on the Internet (http://www.ncbi.nlm.nih.gov/Class/wheeler/jovin.html).

Three-Stage Electrophoretic Procedure Replacing 2D-PAGE in the Proteomics of Membrane Proteins
Buzas, Radko, Chrambach; in collaboration with Sarkadi
Proteomics is burdened by the frequent inapplicability of conventional 2-D PAGE to membrane proteins, which constitute one half of cellular proteins. The inapplicability is attributable to the membranes’ limited solubility in detergents compatible with first-dimensional IPG-IEF. Moreover, effective proteomic analysis is burdened by the need for multiple 2D gels with first-dimensional IEF at multiple narrow pH gradients. Both problems can be solved by replacing 2-D-PAGE with a three-stage procedure (3S-PAGE) in which SDS-PAGE is carried out as a first stage, followed by direct electroelution of the component of interest and analysis of the electroeluate by simultaneous IEF on IPG-strips of the desired pH ranges. Separated protein bands are then transferred into mass spectrometry by conventional proteomic procedures. We implemented 3S-PAGE using the membrane protein MRP1. The crude extract containing the protein was solubilized in media consisting of ASB-16, 7 M urea, 2 M thiourea, or of 4 percent CHAPS, 8 M urea. However, MRP1 immunologically detected in the latter solution fails entirely to enter into IPG-IEF while MRP1 from the former solution falls into two fractions, the major one of which fails to migrate in the electrofocusing gels. Those separation problems show MRP1 to be a representative candidate for development of the 3S-PAGE procedure, which is presently under way.