The workshop consisted of 4 short presentations followed by a period of open discussion between the speakers and the audience. Three of the presentations focused on capillary electrophoresis (Mike Rohde, Amgen; Gary Hathaway, U.C. Riverside; Alan Smith, Stanford), and one on capillary liquid chromatography (Ray Paxton, Immunex). A major attribute of capillary electrophoresis(CE) is the flexibility of application and a majority of the workshop was organized to present a spectrum of these applications.
Mike Rohde presented a brief overview of CE separation theory and then described the first use of CE in his laboratory which was to screen HPLC-isolated peptides from proteolytic digests of proteins. These peptides were candidates for sequence analysis, and often in limited supply. Capillary electrophoresis screening often consumes less than 1% of the sample and can provide confirmation of peak purity, or lack thereof. Since CE is an orthogonal separation method to reverse phase HPLC, it is an excellent choice in most cases, but instances were observed in which a peptide did not pass the CE detector under the screening conditions.
Also discussed was a brief review of a method in which capillary electrophoresis can estimate the pH of neutral charge (pI) for a protein. The method examines the direction and magnitude of protein migration with respect to a neutral marker. By examining several pH conditions, an extrapolation can be made to neutral charge. If the method is used in the absence of denaturants, one can obtain the true solution pl of the protein (M.F. Rohde, K.S. Stoney and J. Wiktorowicz, in "Techniques in Protein Chemistry III" (1992) pp. 121-129).
The last topic discussed was peptide mapping of complete proteolytic digests in the presence of ion pairing agents. Examples were shown which considered the effect of different ion pairing agents, the ion pairing agent concentration, and equilibration time between runs. When all these parameters were optimized and peaks were baseline resolved, reproducible peptide maps were obtained. Furthermore, it was also possible to resolve glycopeptide isoforms that eluted as a single peak under reverse phase conditions. (R.S. Rush, P.L. Derby, T.W. Strickland and M.F. Rohde. (1993) Anal. Chem. 65, 1834-1842.
Gary Hathaway presented an example of the use of CE as a preparative tool for characterizing a synthetic peptide. A 25 amino acid synthetic peptide, refractory to separation by reverse phase HPLC, was resolved into multiple components with capillary electrophoresis at pH 2.5. The two most prominent peaks (70 and 20 area per cent by U.V. absorption) were isolated using preparative CE (75 micron capillary; peptide concentration 0.3 mg/ml). Reiterating the collection process thirty times yielded 200 picomoles of the major and 80 picomoles of the minor species as estimated by amino acid analysis. MALDI time of flight mass spectrometry gave 2930 for the mass of the major peak, corresponding to that calculated for the synthetic sequence (2930.1), while the minor species was 42 mass units smaller (2887.5). Mass analysis, Edman sequencing, and composition (precolumn, PITC) were performed on 80 pmoles of the smaller peptide.
This topic was followed by an evaluation of denaturing CE in an entangled polymer matrix for estimating protein molecular weights. Experimental error was evaluated in a method for the determination of molecular weights of proteins (ProSort). The procedure uses CE in a denaturing solution (SDS) of a linear polymer of acrylamide, and standard protein markers. A plot of relative mobility versus log molecular weight yields a straight line from which the molecular weight of an unknown can then be estimated. Statistical analysis showed reproducibility was +/- 5 parts per thousand for relative mobilities (Rm) which translated to +/- 3% of the molecular weight. Electrophoresis must be carried out with standards and unknowns together, and SDS complexes must possess similar charge to mass ratios, with no gel interaction. A Ferguson plot may be used to correct for such non-ideal behavior. At 215 nm sensitivity is limited to a concentration of about 0.1 ug/ul. Automation of the semiquantitative procedure and the ability to spectrally differentiate impure preparations are major benefits of this technique.
Al Smith discussed the characterization of synthetic oligos using polyacrylamide gel filled capillaries. Examples were presented of CE separations of oligos from 10 to 80 bases which showed single base resolution. CE can accurately measure the percentage of failure sequences, the percentage of oligo modification such as biotinylation and phosphorylation, and is unaffected by the presence of mixed bases, inosine, or halogenated bases. CE appears to be equal in all respects to PAGE for the characterization of synthetic oligos. Capillary electrophoresis has some clear advantages as an analytical tool such as: high resolution, rapid analysis time, automated operation, essentially quantitative recovery and submicrogram sample requirements. The rapid turnaround time and autosampler availability make CE a viable alternative to gel electrophoresis when screening 10-30 oligos per day.
The preliminary results from a recent survey by the Nucleic Acid Committee indicated that a relatively small minority of synthetic oligos made by our membership were characterized by gel electrophoresis and less than 10% by CE. During the discussion period approximately half of the audience indicated that they possess a CE; however, fewer than half of these instruments were in use at least one day per week. These instruments may not have all been in facilities that perform DNA synthesis, but it is possible that the negative aspects of screening oligos by CE, namely high cost and relatively short capillary life (~50 runs), may influence their use in this capacity.
In the final presentation Ray Paxton provided an overview of capillary liquid chromatography as an analytical tool for proteins and peptides. He indicated that the use of capillary LC columns is an increasingly popular method for bringing greater sensitivity to protein and peptide separations. In principle, the use of a 530 um ID column can give 16- and 75-fold increases in sensitivity over the use of conventional 2.1 and 4.6 mm ID columns, respectively; even greater sensitivities can be obtained with 320, 250 and 180 um ID columns. Whereas 2 ug of protein may give a full-scale peak on a 2.1 mm column, 40 ng of protein can give the same peak height on a 320 um ID capillary column.
The increased sensitivity of capillary columns, however, places new demands on the instrumentation needed for their use and for the subsequent analysis of the purified sample. Several laboratories have adapted existing HPLC instrumentation for use with capillary columns [R.L. Moritz and R.J. Simpson (1992) J Chromatogr. 599, 119-130; M.T. Davis and T.D. Lee (1992) Protein Sci. 1, 935-944]. Companies also offer instrumentation for converting standard HPLC systems to capillary systems, and in some cases dedicated capillary LC systems are available. The above references offer a good starting point for investigating the use of capillary chromatography without having to make a large capital equipment purchase.
Capillary chromatography is an excellent complement to some of the newer methods that have been developed for isolating and sequencing proteins. For example, a number of proteins are being isolated by SDS-PAGE in the low ug to sub ug range and then being digested either in-gel or after transfer to PVDF or nitrocellulose. These procedures typically yield only a few pmol of each peptide, which may be difficult to detect and isolate by standard HPLC. The use of capillary LC offers a greater chance of detecting the peptides which can be collected in ca. 5- 10 ul of solvent or directly onto disks for sequencing, thus minimizing sample handling losses.
Return to the The ABRF Home Page