(16k)Five speakers returned to this tenth ABRF meeting to present talks that reviewed advances since the first Research Resource Facilities Group in 1986. The first presentation gave an overview of the progress of the Association and its membership and highlighted the many accomplishments that have resulted from the efforts of ABRF members. Two speakers (Simpson, Keyt) gave updates of presentations they made at the original meeting.
Ronald L. Niece, University of Wisconsin Biotechnology Center, "Reflections on Ten Years of the ABRF".
In 1986 several scientists met to discuss issues of operating laboratories that provided service for colleagues and collaborators. The Association of Biomolecular Resource Facilities evolved from those discussions on sample handling, protocols, laboratory management, and finances. Membership in ABRF has grown continuously without ABRF mounting an organized membership drive. See Figure.
(16k)
Corporate sponsorship has increased regularly since the program began in 1989, and NSF and DOE have funded research proposals from ABRF and its Research Committees.
Education is an important goal of ABRF that has been pursued on several fronts including publications, workshops and symposia. ABRF has organized meetings jointly with many other scientific societies. By the repeated and new invitations to organize and present joint sessions, ABRF is clearly recognized for its interests in quality science, technology, and education.
The members of ABRF and its sponsors deserve credit for the successes that ABRF has. The accomplishments have resulted from the efforts of volunteer members of ABRF; hundreds of individuals have contributed their time and expertise. The results have included the many sources of information and evaluation that we all use: unknowns from the Research Committees for self-evaluation, publications for reference, ABRF News for timely information, membership directory and yellow pages for contacts, and the ABRF electronic mailing list and Gopher server for communication of problems and solutions. Annual meetings have brought members together to exchange ideas. In 1996, ABRF '96: Biomolecular Techniques will provide a broader forum allowing ABRF members from all fields to meet.
Richard Simpson, Ludwig Institute for Cancer Research, "Isolation of Low Picomole Levels of Protein and Peptide for Structural Analysis: An Overview".
If we compare current technologies with what we expected we would be using ten years ago, we would have predicted some of these but not others. We expected that today it would be possible to rely on totally automated amino acid analysis and protein sequencing, but this automation is not currently possible. However, other techniques have moved forward, including NMR studies of peptide and protein structure, X-ray crystallography and computer modeling of the structures of many types of molecules, as well as interactions between proteins and other molecules.
In the 1980's, the strategy was to start with a biological activity, isolate the protein, retrieve the gene, and make a recombinant protein. In the 1990's, the human genome project and especially PCR have influenced strategies that lead to the understanding of biological activities through mapping and nucleic acid sequencing. We now start with the gene and make a recombinant protein to understand the biological activity. In both cases, it is necessary to understand the biological activity and the factors influencing the expression and function of proteins.
Important refinements in protein sequencing since the development of the automated sequencer by Edman include miniaturization, delivery of reactants (gas phase and pulsed-liquid phase), and improved purification of reagents. The level of routine operation is currently at 5 pmol of protein or peptide sample, down 6-9 orders of magnitude. Still, sample preparation and manipulation at levels under 1 nmol become limiting.
Microtechniques currently in use today permit the purification of proteins found in low abundance, for example by polyacrylamide gel electrophoresis, affinity chromatography, and microbore HPLC. It is now a standard procedure to recover proteins from PAGE by electroelution or electroblotting. Microbore HPLC improved sensitivity by 20-fold, so that 25 ng of protein is currently sufficient for a discernible peak, and the chromatography takes only minutes. Available HPLC hardware requires little change except for use of 1 mm internal diameter columns; 10 pmol of peptide can easily be detected at a peak height of 80 mAU.
Some modern sequencers using biphasic column technology permit sequencing 60 or more residues. The sensitivity has remained about the same but repetitive yield has improved. The present PTH-amino acid columns have 2 mm internal diameters that limit sensitivity of detection to 25-500 fmol.
Removal of SDS from electroeluted samples represents a challenge that was solved by reconsidering how reversed phase chromatography can be used. By loading the sample onto the column in organic solvent, SDS and salts are not retarded by the stationary phase and pass through, but proteins are retained. A clean protein sample can be eluted by decreasing the organic solvent concentration. The purified protein can then be reduced, alkylated, and digested, and its fragments can be separated on the same column. For example, 20 ug of GnT1 in 1.5 ml of 1% Triton X100 can be effectively cleaned up and recovered using this "inverse phase" technique on a 2.1 mm internal diameter column.
Over the past ten years, two-dimensional gels have gained a prominent position in protein analysis. Immobilized pH gradients have led to better reproducibility, and roughly 2,000 spots can be seen on a single gel. Computer software has developed to allow accurate quantitation of proteins and gel-to-gel comparisons of patterns. It is estimated that the human genome has approximately 50,000-100,000 genes coding for proteins, and that any one cell type might express 5,000-10,000 different proteins. If we can detect only 2,000 of these, where are the other 3,000-8,000 proteins? They are also on the gel, but because they are expressed in lower amounts, they may not be visible by the staining techniques used. Their biological activities may still be measured. More sensitive techniques may be required to identify them.
Fingerprinting using mass spectrometry techniques may permit research on these more limited proteins due to its higher sensitivity. Several new methods are being developed that combine in-gel or on-membrane digestions of proteins with microbore or capillary column chromatography and mass spectrometry. New software is being developed for quick identification of proteins by comparing their chromatography and mass parameters with proteins in databases. Coupling fingerprinting with other orthogonal techniques such as MS/MS will be needed for confidence in protein structure determination. Structural analysis using capillary column and electrospray mass spectrometry with only 200 fmol of protein will soon be possible. Needed improvements include methods to simplify fragment patterns.
Sample preparation remains the limiting factor in improved sensitivity. In the last ten years, HPLC technology has gone from 4.6 mm ID columns (requiring 100 ug protein) to 2.1 mm ID columns (2-10 ug) to 1 mm ID columns (0.05-0.2 ug). The technology for microcapillary columns (less than 10 ng protein), CZE, and open-tubular columns promises more sensitivity. Similarly, sequencing has progressed from the spinning cup to gas phase to pulsed-liquid phases instruments. Mass spectrometry has progressed from FAB to electrospray and MALDI-TOF and now ion-trap instrumentation. The increased sensitivity of mass spectrometry now allows one to characterize an amount of protein that can be conveniently purified. With all these improved methods, sample preparation remains a very important part of protein sequencing.
Bruce Keyt, Genentech, Inc., "Analysis of Tissue Plasminogen Activator: Natural Variants and Site-Directed Mutants".
Recombinant tumor plasminogen activator (tPA) has been characterized thoroughly by standard structural techniques including amino acid analysis and protein sequencing. Because tPA is glycosylated, it required additional characterization. The natural product has many good properties but it had several properties that were considered possibilities for improvement, including its time for clearance from the blood and its overall biological activity. Recombinant engineering was envisioned as the most direct route to discovering what alterations were necessary, which domains of the protein would have to be altered, and how the alteration affected the overall structure of the protein. Mutations were engineered in restricted regions of single domains to avoid gross alterations. Charged residues were changed to alanine in groups of 1-4 avoiding spanning cysteine residues.
Wild type tPA appeared to be a heterogeneous molecule on analysis by SDS PAGE. Following mild trypsin digestion and chromatographic separation, sequencing demonstrated multiple amino-termini. While the structure of tPA is well known, the three-dimensional structure of tPA has not been determined. For this reason, domains of other known proteins that showed similarity with domains within tPA were used to model the structure in order to design strategies for mutagenizing tPA. For example, trypsin was used as an example for designing the sites of mutation in the protease domain.
The first changes in tPA involved introducing a different glycosylation site in the first kringle domain. This alteration was made to reduce the clearance of the recombinant molecule from the blood stream. The strategy was to look at other proteins with kringle domains to see where they were naturally glycosylated. It was proposed that engineering a similar site in tPA would minimize the chance of disturbing the tertiary structure. A threonine residue was changed to an asparagine, which introduced a second glycosylation site in the protein. In addition, this changed the original glycosylation moiety from a high-mannose type to a complex carbohydrate. As predicted, the recombinant protein had a much longer clearance time in serum, but unfortunately its biological activity was much reduced. To boost the activity to wild type levels, additional engineering steps were necessary. The next procedure was to remove the original glycosylation site, again by site specific mutagenesis. This double mutant had activity that was comparable to normal levels.
The glycosylation site studies were facilitated by development of LC/MS procedures. In general, a protein is digested with a specific protease, and the peptides are separated by reversed-phase HPLC and detected by electrospray ionization mass spectrometry. Glycosylation sites are identified from contour plots. Glycopeptides are observed as a characteristic series of masses, "streaks", at different elution times. Mass shifts after glycosidase treatment contribute to the identification of structures. This technique allows for the rapid identification of glycosylation sites and yields structural information. Furthermore, heterogeneity in the carbohydrate moiety can be observed.
The final change in tPA structure was to replace a set of four basic residues, KHRR, close to the active site with four alanines. The highly positively charged site (KHRR) on the proteolysis domain suggested it was surface accessible and likely to be important for proteolysis. Changing these residues to alanine significantly increased the affinity for the inhibitor PAI-1 and decreased the affinity for fibrin except when fibrin is a part of a clot. Clinically, the tetra-alanine protein is much more potent.
This combination of analysis and protein engineering led to a novel protein with high biological activity that had better properties--including a longer clearance time--by changing the position of the glycosylation site and better specificity by replacing four charged residues.
References
Bennett et al., J. Biol. Chem. 266: 5191-5201, 1991.
Guzzetta et al., Anal. Chem. 65: 2953-2962, 1993.
Keyt et al., Proc. Natl. Acad. Sci. USA 91: 3670-3674, 1994.
Return to the The ABRF Home Page