created: 22nd October 1995, last updated: 29th April 1998,© 1998 ABRF
The ABRF is a non-profit association which was formally organized in 1988 to represent and support laboratories providing highly specialized resources for analysis and/or synthesis of biomolecules to the research community. At the present time, the ABRF represents more than 260 facilities from around the world.
A major goal of the organization since its inception has been to define and improve the technical capabilities of core facilities for both the operators and users of the facility. This goal continues to be met by annual surveys and studies conducted by a diverse group of research committees. Anonymously returned results are analyzed by the committees and reported at appropriate national and international scientific meetings followed by publication in the literature or in the ABRF newsletter. Techniques oriented workshops have recently been presented at annual meetings of the American Society of Biochemists and Molecular Biologists, American Peptide Society, Protein Society, American Society for Mass Spectrometry, Methods in Protein Structure Analysis and today at the First European Symposium of the Protein Society. In March 1996, the ABRF will hold its own first independent meeting in San Francisco.
There is now an active community of over 300 scientists world-wide, the majority of whom are ABRF members, who contribute to or read an email-based bulletin board set up by the ABRF. To subscribe to the network, send a single message to abrf-request@aecom.yu.edu; to send a question or answer, use abrf@aeconyu.edu. The bests tips and techniques gleaned from the electronic mail are deposited at a gopher address (ABRF.bsjhu.edu) along with copies of newsletters, yellow pages and other items of interest.
The meeting organizers are grateful to PerSeptive Biosystems for sponsoring this abstract booklet
PerSeptive Biosystems
Biosearch Products
We also thank Perkin Elmer and Hewlett Packard for financial assistance with this meeting.
Greg Grant (ABRF Executive Committee)
Session 1 Chairpersons . . . Greg Grant, Peter Hunziker
Chromatography
20 min 12.40-1.00 Richard Simpson (Ludwig Institute,
Melbourne, Australia)
Capillary HPLC: a tool for protein structure analysis
l5 min 1.00-1.15 Christopher Southan (SmithKline
Beecham, Welwyn, UK)
Fast HPLC on cheap, disposable columns
I5 min 1.15- 1.30 Discussion
Quality control
15 min 1.30- 1.45 Len Packman (Cambridge University,
UK)
The detection and prevention of aspartimide formation in solid phase
peptide synthesis
20 min 1.45-2.05 Finn Kirpekar (Odense University,
Denmark)
MALDI mass spectrometry of nucleic acids
l0 min 2.05-2. 15 Discussion
20 min 2.15-2.35 Break and refreshments
Session 2 Chairpersons ...Richard Simpson, Len Packman
Post-translational modifications
20 min 2.35-2.55 Alastair Aitken (National Institute for
Medical Research, Mill Hill, UK)
Identification of post-translational modifications; synthesis of
modified peptides and production of specific antisera
15 min 2.55-3.10 Christopher Starr (Glyko Inc., Novata, CA) Characterization of the carbohydrate moieties attached to glycoconjugates using fluorophore-assisted-carbohydrate electrophoresis (FACE)
l0 min 3.10-3.20 Discussion
Mass spectrometry
20 min 3.20-3.40 Darryl Pappin (Imperial Cancer Research Fund, London, UK) Chemical/MALDI approaches to the analysis of peptides at the sub-picomole level
20 min 3.40-4.00 Peter Roepstorff (Odense University, Denmark) Full characterization of the covalent structure of proteins by mass spectrometry
20 min 4.00 4.20 Christoph Eckerskorn (Max-Planck Institute, Martinsreid, Germany) MALDI mass spectrometry of proteins electroblotted after polyacrylamide gel electrophoresis
l5 min 4.20-4.35 Discussion
Databases
20 min 4.35-4.55 Peter James (Federal Institute of Technology, Zurich, Switzerland) Database searches using MS and MS/MS data and their relevance to core facilities
5 min 4.55-5.00 Discussion
60 min 5.00-6.00 Open forum and refreshments
Richard J. Simpson, Gavin E. Reid and Robert L. Moritz
Joint Protein Structure Laboratory, Ludwig Institute for Cancer Research (Melbourne) and the
Walter and Eliza Hall Institute of Medical Research, Parkville, Victoria 3050 Australia (Tel: +613-347-6389; Fax: +61-3-348-1925; Email: simpson@licre.ludwig.edu.au)
Capillary or microcolumn (<0.5-mm internal diameter) HPLC is an extremely powerful microseparation technique for proteins and peptides. It has distinct advantages over wider-bore column HPLC, such as increased mass sensitivity and lower flow rates, which make it particularly attractive for coupling to hyphenated detection techniques such as mass spectrometry. Although the advantages of miniaturized columns are obvious, progress with this technology has been restricted by the availability of packed capillary columns and instrumentation designed to facilitate the operation of such columns. To this end, we have developed a procedure for slurry packing capillary columns and a continuous gradient elution method, using conventional HPLC systems, for performing capillary column liquid chromatography.
Gradient eluent from a microbore liquid chromatograph was split ahead of the injector so that an accurate percentage (2-3%) of the mobile phase delivered by the pump flowed through the capillary column. The outlet of the column was connected to a length of 0.075-mm I.D. fused-silica capillary tubing which, in turn, was connected to 6-mm optical path length longitudinal capillary flow cell. Fused-silica capillary columns of 0.1-0.32-mm I.D. were slurry-packed efficiently with 7-um spherical, 300 Å pore size, C8 bonded phase particles, and evaluated in terms of their ability to resolve mixtures of proteins, peptides or phenylthiohydantoin (PI H) amino acid derivatives. The gradient elution profiles agreed with those obtained using microbore (<2.1 mm I.D.) and larger bore columns. The minimum detection amounts for protein and PTH-amino acids on 0.32 mm I.D. capillary columns were 50 pg and 25 fmol, respectively. At a flow rate of 3.6 pl/min, proteins and peptides were recovered from the capillary columns in volumes of about 2-8 p1. By using a multiple-wavelength, forward optics detector, tryptophan and tyrosine-containing peptides can be identified.
With the recent improvements of proteolytic digestion procedures for acrylamide gel-resolved proteins, both in-gel and on-membrane, and the exciting development of peptide mass fingerprinting algorithms for identifying proteins based on accurate mass analysis of peptides, there has been an increased desirability for rapid, highly-sensitive peptide isolation procedures. Although fast macroporous stationary-phase packings with their minimal stagnant mobile phase mass transfer properties are now commercially available, these packings can be prohibitively expensive. To this end, we have developed a procedure utilizing conventional silica-based reversed-phase packings, applicable to both microbore and capillary columns, which decreases the separation time of a typical 60-min peptide map to ~ 12 min; little difference in peak resolution was observed with this high speed chromatographic approach when compared to conventional peptide mapping procedures.
References:
Moritz, RL & Simpson, RJ (1992) Application of capillary reversed-phase high-performance liquid chromatography to high-sensitivity protein sequence analysis, J. Chromatogr. 599,119-130.
Moritz, RL & Simpson, RJ (1992) Purification of proteins and peptides for sequence analysis using microcolumn liquid chromatography, J. Microcol. Sep. 4, 485-489.
Moritz, RL., Eddes, JS., & Simpson, RJ. (1994) High speed chromatographic separation of proteins and peptides: Application to rapid peptide mapping of in-gel digested proteins. J. Prot. Chem. 13, 486-487.
Moritz, RL., Reid, GE., Ward, LD., & Simpson, RJ. (1994) Capillary HPLC: A Method for Protein Isolation and Peptide Mapping, METHODS: A Companion to Methods in Enzymology 6, 213-226.
Christopher Southan
Department of Protein Chemistry, SmithKline Beecham Pharmaceuticals, The Frythe, Welwyn, Hertfordshire, England, AL6 9AR. Tel. 0438-782094, Fax. 2550
E-mail southan_c%frgen.dnet@sb.com
Microbore HPLC is clearly essential for sample-limited investigations but even with abundant proteins the scaling down of separations can reduce solvent consumption and analysis time. This work describes the design of disposable, high-speed, microbore columns simply constructed from 1/16" and 1/8'' HPLC fittings. With internal diameters from 0.5 to 2.5 mm they cover a 25-fold range of cross sectional area. The use of glass, TEFZEL or PEEK tubing limits operating pressures to below 2000 psi but useful separations of proteins and peptides on a wide range of chromatographic packings can be obtained by simple packing techniques. The home packed columns have proved to be effective for any of the common modes of separation, reverse-phase (RP) ion-exchange (IE) and size-exclusion (SE).
Recently, non-silica chromatographic packings with large particle and pore sizes have become available designed to operate at high horizontal velocities with low operating pressures. These are particularly well suited to microcolumn construction. For example a 0.5 mm x 50 mm PEEK column packed with 10 um Poros R can be run at 200 ul/min. This facilitates microbore-scale separations using conventional reciprocating pumps. Lowering [TFA] to 0.08-06% allows some UV detectors to zero down to 205 nm. Thus with long path-length flow cells the sensitivity can exceed those of microbore instruments. For high speed protein separations acetonitrile or salt gradients as steep as 10% per/min can be used and complete run times (i.e. with re-equilibration) of between 1 and 10 min can be achieved.
Column construction is straightforward for those experienced in manipulating plastic HPLC tubing and fittings. A full range of these, including the essential 1/16" frits, (Upchurch part no. C407) are available from Upchurch Scientific (USA) and Jour Research (Sweden). Agents for these include Anachem (U.K.) and Alltech (US and Europe). Basic dry-packing into glass tubes is described in (1) and an outline of high-speed applications with PEEK tubing in (2). Although expensive, Perceptive Biosystems supply a "Poros Self-Pack" steel packer suitable for medium pressure (up to 1000 psi) slurry packing. For this is useful to use HPLC pumps that flow -ramp down to hold a constant maximum pressure setting rather than cutting out.
A variety of high-velocity packings are now available for RP and E, although it can be difficult to negotiate 10 um loose material. The list would include Poros (Perceptive Biosystems) PLRPS (Polymer Laboratories) Resorce (Pharmacia) and Hyper D (Sepracore) For size-exclusion columns G25 works well but the newer Superdex (Pharmacia) gels are more robust. Typically a top frit is unnecessary. These columns tend to pack down with use but the resulting dead-volume and any surface clogging can easily be removed by cutting back the tubing. Do not expect high plate nos. or good isocratic resolution with these simple packing methods. As fittings are expensive it is possible to disassemble and end-cap the packed tubing for storage. Feed back on the utility of these columns is welcome and queries can be posted on the ABRF User Group list.
References:
1. Southan C, "The use of glass capillary tubes as disposable microbore columns for the RP HPLC of proteins and peptides", in: Techniques in Protein Chemistry, T. Hugli, Ed, Academic Press, pp 392-398
2. Southan C, Fantom, K, & Lavery, P. "Fast, Flexible, Sensitive and Cheap: The Use of Home Made Microcolumns for the Separation of Proteins and Peptides." Journal of Protein Chemistry, 13 (5) 461-462, 1994 (MPSA Short Communications)
Leonard C. Packman, Martin Quibell+ and Tony Johnson+.
Centre for Molecular Recognition, Department of Biochemistry, University of Cambridge, Tennis Court Road, Cambridge CB2 lQW, U.K. Phone +44 1223 333639; Fax: +44 1223 333345; email lcp2@mole.bio.cam.ac.uk and +MRC Laboratory of Molecular Biology, Cambridge, UK
Aspartic acid is the source of a number of unwanted peptide transformations which may arise during both the synthesis and purification of a peptide. It is susceptible to dehydration, transpeptidation and epimerisation, and these processes can occur under both acidic and basic conditions. In Fmoc-based syntheses, tertiary butyl protection of the side chain has generally been regarded as sufficient to prevent undesired side reactions. It is now clear that certain Asp(OBut)-X sequences are particularly susceptible to aspartimide (cyclic imide) formation, the result of the aspartyl amide nitrogen attacking the [[beta]]-carboxyl under basic conditions [l,2,3]. This reaction (Fig. l) can occur in extent from <1% to a few % per piperidine deprotection cycle and it therefore becomes more significant the closer the susceptible sequence is to the C-terminus of the peptide. Detection of aspartimide by mass analysis can only be done after cleavage of the peptide and manifests itself as -18 Da from the expected mass (succinimide) or +67 Da (&-piperidide) depending, respectively, on whether the ring is intact or has been opened by reaction with piperidine. HPLC profiles alone or amino acid analysis cannot identify an aspartimide problem and are unreliable guides as to the integrity of the peptide. If the ring has been opened by hydrolysis, then there is no mass change but a mixture of a and H peptides will result, and in enantiomeric forms.
A convenient solution to this problem is to protect reversibly the aspartyl amide nitrogen. This can be done with the 2-hydroxy, 4-methoxybenzoyl (Hmb) group [4] (Fig. 2). The Hmb group can be removed when the peptide is cleaved and deprotected using TFA-scavenger mixes. In its 2O-acetyl form, it remains attached to the peptide and can act as an aid to solubility during the purification, and then removed. Data will be presented showing the utility of Hmb in preventing aspartimide formation as well as its influence on preventing chain folding and aggregation during chain assembly.
References:
1. Nicolas, E.., Pedroso, E. and Ghalt, E. (1989) Tetrahedron Lett. 30, 497-500
2. Dolling, R., Beyermann, M., Hacnel, J., Kernchen, F., Krause, E., Brudel, M and Bienert, M. (1994) J. Chem. Soc. Chem. Commun. 853-854
3. Yang, Y., Sweeney, W.V., Schneider, K., Thornqvist, S., Chait, B.T. and Tam, J.P. (1994) Tetrahedron Lett. 35, 9689-9692
4. Quibell, M., Owen, D., Packman, L.C. and Johnson, T. (1994) J. Chem. Soc. Chem. Commun. 2343-2344
F. Kirpekar, E. Nordhoff, P. Roepstorff, S. Hahner1 & F. Hillenkamp1
Department of Molecular Biology., Odense University, Campusvej 55, DK-5230 Odense M, Denmark (fax: +45 65 932781) and 1Department of Medical Physics and Biophysics, University of Münster, Robert-Koch-Str. 31, D-48149 Münster, Germany (fax: +49 251 835121)
Background:
The possibility for generating ions for mass spectrometry by electrospray ionization (ESI) and matrix assisted laser desorption/ionization (MALDI) has within the last years meant a major break-through in the analysis of macromolecules of both synthetic and biologic origin. An exception is the mass spectrometry of nucleic acids which have proven difficult to analyze, mainly due their physio-chemical properties.
Regardless of the ionization method, the polyanionic phosphodiester backbone of nucleic acids is a potential source of problems. Its strong tendency to form salts with e.g. alkali ions leads to a molecule ion distribution with a variation in both the number and species of counter ions. For both ESI and MALDI, the problem is overcome by variations over the theme: addition of volatile ammonium salts. The phosphodiester groups will form ion pairs with the ammonium cation and these ion pairs dissociated into the free ammonia and the free acid of the phosphodiester groups during the desorption/ionization process. For MALDI-MS, we favour the use of NH4+-loaded cation exchange beads (1) which are added to the final sample preparation.
Instability of the analyte ion is the major limitation in MALDI-MS of DNA above 100 nucleotides. The key fragmentation reaction is the loss of nucleobases from the deoxyribose phosphate backbone. This loss may be followed by cleavage of the backbone and it results in decreased detection efficiency for large species. When investigating larger species, a linear time-of-flight mass analyzer should be used because metastable fragmentation severely decreases the signal and the mass resolution when using a reflector time-of-flight analyzer (2). In contrast to DNA, RNA is much less prone to fragmentation because it harbours a 2'-OH group which stabilizes the bond between the nucleobase and the ribose moiety. We have detected RNA species up to 150 kDa. using a reflector time-of-flight analyzer (3).
Applications:
- Verification of the synthesis of oligonucleotides up to approx. 15 kDa
- Analysis of mixtures generated by exo- or endo-nuclease sequencing of nucleic acids.
- Analysis of covalently bound nucleic acid - protein complexes.
- Verification of simple enzymatic reactions on nucleic acids (phosphatases, ligase, kinases).
Method:
With the procedure given below, we are routinely analyzing oligonucleotides up to 50 nucleotides with a mass accuracy of 0.05% or better.
Ion-exchange beads (BioRad, SOW-X8, mesh size 100 - 200 um) are incubated over night with a saturated solution of ammonium-acetate and subsequently washed 5 - 10 times with, and resuspended in double-destilled water. Prepare the sample for MALDI-MS by mixing 1 ul of matrix solution (50 g/l 3-hydroxypicolinic acid in 20% acetonitrile) with approx. 10 pmoles of analyte (in 1 ul or less) and add 10 - 20 NH4+-loaded cation exchange beads. Air-dry and remove the beads under a microscope with a pipette tip. For an accurate calibration of the spectra, the addition of internal standards to the sample preparation is recommended.
References:
1. E. Nordhoff et al. (1993). Nucleic Acids Res. 21, 3347 - 3357.
2. E. Nordhoff et al. (1994). Nucleic Acids Res. 22, 2460 - 2465.
3. F. Kirpekar et al. (1994). Nucleic Acids Res. 22, 3866 - 3870.
Alastair Aitken. Peter Fletcher and Steve Howell
National Institute for Medical Research, Mill Hill, London NW7 1AA, U.K
tel (01)81 959 3666 ext 2158; fax (01)81 906 4477; a-aitkeni@uk.ac.mrc.nimr
Methods will be presented for (1) analysis of a range of post-translational modifications, including endogenous sites of phosphorylation, glycosylation, acetylation and acylation. On-line trapping applications 1 have been improved and extended - to include use of a novel SDS trapping column and direct elution from SDS PAGE into ESMS; (2) synthesis of modified peptides and production of specific antisera 2. Analysis of modified peptides, where no 32p or other radiolabel is present, may be particularly difficult. Even in the identification of phosphorylation sites where 32P-radioactivity is present there may be problems - if one relies on phosphopeptide map analysis. We have encountered two examples recently which were thought to be phosphoserine but turn out to be something quite distinct. Has anyone successfully used the commercial antibodies against phospho-serine or -threonine? In contrast, the use of anti-phosphotyrosine antisera is well established. Antisera specific for the phosphoform of a peptide are usually successful therefore good methods for synthesis of phosphopeptides are required. Best results are obtained with the addition of a cysteine to the N- or C- terminus, for coupling to carrier protein by the maleimide method.
The ESMS of products obtained on-site are shown in fig. 1. In contrast, a commercial "effort" produced no detectable levels of another phosphopeptide, which again emphasizes that routine analysis by MS is essential. Phosphopeptides run well on ESMS and MALDITOF in positive ion mode. Particular problems may be associated with ESMS of phosphopeptides, where high levels of Na+ and K+ adducts are regularly seen on species at charge states above that predicted by the theoretical number of basic groups (see fig. 1). Other methods of phosphorylation site analysis include: (a) conversion of phospho-Ser to S-ethylcysteine, which may be followed by ESMS as well as sequencing3 (b) production of phosphate-specific fragment ions by CID during negative ion LC-ESMS4.
Synthesis of phosphopeptides using Fmoc methodology
(a) The unprotected residue can be modified by a global phosphorylation after synthesis. It should be emphasized that the residues to be phosphorylated must be incorporated without side chain protection. Global phosphorylation works well if the N-terminus is protected with Boc (using di-tert butyl pyrocarbonate) or (permanently) by acetylation. We use dibenzyl-N,N-diisopropylphosphoramidite (10 equiv., Calbiochem/Novabiochem) under dry nitrogen. After removal of excess reagent, add t-butyl peroxide (20 equiv.) in DMF for 30 min.
(b) The Barany method uses the Fmoc-phosphotyrosine derivative (Novabiochem). This is expensive, and a re-couple to get >99% coupling may be required although the final yield is excellent. Dephosphorylation can occur on long peptides. Due to base lability peptides with phospho-serine/threonine can't be synthesized by this method.
(c) A third method, which we have not used involves dimethyl-protected phosphoTyr and deprotection with TMS-EKr or TFMSA.
References:
1. Aitken, A. Patel, Y., Martin, H., Jones, D., Robinson, K., Madrazo, J. and Howell, S. (1994) Electrospray mass spectrometric analysis with on-line trapping of posttranslationally modified mammalian and avian brain 14-3-3 isoforms. J. Prot. Chem. 13, 463465.
2. Martin, H., Patel, Y., Jones, D., Howell, S., Robinson, K. and Aitken, A. (1993) Antibodies against the major isoforms of 143-3 protein. An antibody specific for the N-acetylated amino terminus of a protein. FEBS Lett. 331, 296-303.
3. Aitken, A., Howell, S., Jones, D., Madrazo, J. and Patel, Y. (1995) 143-3 [[alpha]] and d are the phosphorylated forms of Raf-activating 143-3 [[beta]] and [[varsigma]]. In vivo stoichiometric phosphorylation in brain at a Ser-Pro-Glu-Lys motif. J. Biol. Chem. 270, in press.
4. Huddleston, MJ., Annan, RS. Bean, MF. and Carr, SA. (1994) Selective detection of Thr-, Ser-, and Tyr- phosphopeptides in complex digests by electrospray LC-MS In, Techniques in Protein Chemistry V.
J. van Oostrum1, E.A. Kragten1, A.A. Bergwerff1, K.O. Börnsen2, D.R. Müller3 and W.J. Richter3
1Core Drug Discovery Technology, 2Corporate Analytical Research, 3Central Research Services, Ciba, Basel, Switzerland
Electrospray LC/MS was used to investigate the carbohydrate chains on recombinant chimeric human/murine monoclonal immunoglobulin G1 antibodies. Recombinant IgG antibodies are generally produced in a batch process and expressed in cells having a glycosylation machinery different from that of human cell types. In order to identify any aberrant, possibly immunogenic carbohydrate structures as well as to confirm structural consistency between batches, glycosylation pattern analysis of recombinant IgG is necessary. Tryptic digests of three (reduced and alkylated) IgG1 antibodies were analyzed by reverse phase HPLC coupled to electrospray MS (API-III triple quadrupole mass spectrometer). Intermittent collisional excitation in the sample orifice region was used to generate fragment ions only during the first part of the scan cycle, i.e. m/z 100 to 450, whereas in the residual mass window, m/z 450 to 2400, conventional scanning without excitation was used. The elution of a glycopeptide was selectively indicated by the simultaneous appearance of sugar-specific fragment ions at m/z 204 and 366 (GlcNAc and LacNAc C1-carbenium ions, respectively) in the pertinent reconstructed mass chromatograms. For each of the three IgG1 hydrolysates, two glycopeptide-containing fractions were thus located and found to correspond to overlapping g and 13 amino acid (aa) sequences comprising the same glycosylation site. Each contained fucosylated di-antennary N-Glycans comprising zero, one or two terminal galactose residues in varying ratios. Exclusively on the 13-aa peptide of one of the IgG1 preparations, two minor congeners of the mono- and digalactosylated main structures were found to be monosialylated with N-glycolyineuraminic acid. The smallest structure found exclusively on the respective 9-aa peptides as a minor component, was a fucosylated Mglywan pentasaccharide core carrying only one of the customary two GlcNAc residues. The observed differences in the glycosylation patterns of the 9- and 13-aa glycopeptides for the individual IgG1's suggest an influence of the carbohydrate structures on the proteolytic ability of trypsin. An independent confirmation of the derived neutral oligosaccharide structures and proportions in the glycopeptides was obtained by high-pH anion chromatography and MALD-MS of the enzymatically released N-glycans.
Darryl J. C. Pappin
Protein Sequencing Laboratory, Imperial Cancer Research Fund, 44 Lincoln's Inn Fields, London WC2A 3PX, UK Phone: 144 171 269 3071 Fax: +44 171 269 3093
email: pappin@europa.lif.icnet.uk
Large-format 2D gel electrophoresis systems have been developed that are capable of resolving several thousand cellular proteins in a matter of days [1,2]. For a number of years, a combination of Edman microsequence analysis and identification of proteins by staining with specific antibodies has been used systematically to categorize proteins and establish cellular databases [3,4]. There are, however, significant problems associated with these approaches. Most resolved proteins are only present in the low- to upper-femtomole range, significantly below the level at which automated sequencers can reliably operate. The relatively slow speed of the Edman process (one or two samples per machine per day) also means that the sheer number of proteins is too great to permit large-scale characterization within any useful period of time. The use of monoclonal antibodies, whilst both rapid and extremely sensitive, requires the ready availability of a large pool of antibody probes.
New methods have recently been developed, using a combination of protease digestion, matrix-assisted laser-desorption ionization (MALDI) mass spectrometry and screening of peptide-mass databases, that offer significant increases in the speed at which proteins can be identified, with sensitivities extending into the low-femtomole range. Search discrimination can be dramatically improved by including partial linear sequence or peptide compositional information derived from mass measurement following specific chemical modification (e.g. esterification, deuteration, alkylation). Novel reagents and micro-scale techniques have been developed that allow these modifications to be performed (in series, if required) on sub-picomole amounts of peptides and proteins to provide such compositional data. Complete or partial linear sequence data seems increasingly likely to be provided by tandem MS techniques [5] or post- source decay (PSD) analysis [6] rather than by automated Edman degradation, and we report on the development of derivatisation reagents that may be useful for the simplification of collision spectra for peptide sequencing by low-energy tandem MS/MS or PSD methods.
The use of compositional information or short stretches of linear sequence can increase peptide-mass search discrimination by several orders of magnitude, and usually eliminates any requirement to perform parallel digests with different enzymes. Incorporation of such additional information more than compensates for the lower precision of mass determination when using the simplest time-of-flight mass spectrometers.
References:
1. O'Farrell, P. (1975) J. Biol. Chem. 250, 4007-4021.
2. Patton, W.F., Pluskal, M.G., Skea, W.M., Buecker, J.L., Lopez, M.F., Zimmelmann, R., Belanger, L.M and Hatch P.D (1990) Biotechniques 8, 518-527.
3. Celis, J.E., Gesser, B., Rasmussen, H.H., Madsen, P., Leffers, H., Dejgaard, K, Honore, B., Olsen, E., Ratz, G., Lauridsen, J.B., Basse, B., Mouritzen, S., Hellerup, M., Andersen, A., Walbum, E., Celis, A., Bauw, G., Puype, M., Van Damme, J. and Vandekerckhove, J. (1990) Electrophoresis 11, 989-1071.
4. Garrels, J. and Franza, B. (1989) J. Biol. Chem. 264, 5283-5298.
5. Hunt, D.F., Henderson, RA., Shabanowitz, J., Sakaguchi, K., Michel, H., Sevilir, N., Cox, A.L., Appella, E. and Engelhard, V.H. (1992) Science 255, 1261-1263.
6. Kaufmann, R., Spengler, B. and Lutzenkirchen, F. (1993) Rapid Commun. Mass Spectrom. 902-910.
Peter Roepstorff
Department of Molecular Biology, Odense University, DK-5230 Odense M, Denmark
Phone: +45 66158600 ext. 2404, Fax: +4565932781, E-mail: roe@pr-group.ou.dk
The need for characterization of the complete covalent structure of proteins have dramatically increased since the introduction of DNA recombinant technology. A majority of the primary structure information of proteins is presently being derived from cDNA or genome sequences. Unfortunately, these do not yield any information about processing of the peptide chain, disulfide bridging and post translational modification. Traditional protein chemical methods are at the best laborious, time and sample demanding but more often insufficient for characterizing these phenomena The task is huge and needs new and improved technology in terms of speed, sensitivity and specificity. New hitherto unknown proteins predicted based on the genome sequences are identified and need characterization of the covalent structure of the functional protein. Recombinant proteins are produced and these as well as their natural counterparts need full characterization of all post translational events. Engineered proteins must be checked. Industrially produced recombinant proteins for therapeutic and industrial use must be checked for gene stability and batch to batch variations. Fortunately, the recent development of electrospray ionization (ESI)1 and matrix assisted laser desorption/ionization (MALDI)2 mass spectrometry (MS) has provided the needed technology.
We have in our laboratory attempted to develop MS-based strategies which allows us to identify proteins based on data base sequence information irrespective of its source3. Once the protein is identified deviations from the expected structure can be located taking advantage of specific enzymes and the complementarity of the MS-techniques using the strategy shown below4.
A prerequisite for the use of the strategy is the high sensitivity provided by MALDI5, ability to analyze complex mixtures provided by both MALDI and LCMS using ESI, and high mass accuracy provided by ESI MS for proteins and for peptides by both techniques. Examples of the full characterization of the covalent structure of natural and recombinant proteins will be given.
References:
1. Karas, M., and Hillenkamp, F., (1988) Anal. Chem. 60, 2299.
2. Fenn, J., Mann, M., Meng, C.K., Wong, S.F., and Whitehouse, C.M., (1989) Science 246, 64.
3. Mann. M., Højrup, P., and Roepstorff, P. (1993) Biol. Mass Spectrom. 22, 338
4. Andersen, J., Søgaard, M., Svensson, B. and Roepstorff, P. (1994) Biol. Mass Spectrom. 23, 547
5. Vorm, O., Roepstorff, P., and Mann, M. (1994) Anal. Chem. 66, 3281.
Christoph Eckerskorn*, Kerstin Strupat#, Franz Hillenkamp#, Friedrich Lottspeich*
*: Max-Planck Institute for Biochemistry, 82152 Martinsried, Germany Phone: ++ 89 8578 2477, Fax: ++ 89 8578 2802
#: Institute for Medical Physics and Biophysics, University of Münster, Robert-Koch-Str.31 48149 Münster, Germany
Matrix-assisted laser desorption ionization mass spectrometry (MALDI-MS) of proteins electroblotted onto polymer membranes after polyacrylamide gel electrophoresis (PAGE) separation was reported by us recently (1,2). The proteins are desorbed directly from the blot membranes after matrix incubation. The matrix preparation immediately after the electrotransfer is of prime importance for a successful MALM analysis of blotted proteins due to strong interactions of the proteins with the surface of the blotting membrane during the electrotransfer. Desorption with infrared radiation was found to be superior to W-desorption. Several commercially available membranes tested resulted in protein signals after matrix incubation of the membranes. Systematic investigations on different PVDF based membranes showed improved results for membranes exhibiting high specific surface areas and dense pores. Staining with organic dyes results in broad protein signals shifted in mass due to addition of several dye molecules; staining with colloidals such as gold or India ink did not shift the protein mass and no significant increase of peak widths compared to signals taken from nonstained protein bands was observed. The average number of shots onto one spot of a protein band on a PVDF membrane yielding between 5 spectra at the edges and more than 50 spectra at the most intense center of the band. Most of the shots onto a given spot in a protein band gave spectra with comparable intensities and qualities with an excellent signal-to-noise ratio due to an uniform protein distribution on the blotting membrane. Assuming transfer rates during electroblotting higher than 90%, the protein consumption per laser shot (diameter 100 um) was found to be smaller than 1 femtomole. Scans across neighboring protein bands showed that the lateral resolution of the electrophoretically separated proteins is preserved after the matrix preparation procedure. Analyzing all the proteins of the paramyxovirus (Sendai D52) gave spectra with comparable intensities and qualities for the well-soluble proteins as well as for the membrane proteins. Scanning across the highly glycosylated virus membrane protein HN revealed a continuous decrease of the molecular mass for adjacent sum spectra (distance 1 mm) due to a heterogeneous glycosylation.
High-resolution two-dimensional PAGE (2D-PAGE) is commonly used as an analytical approach to resolve and detect most of the numerous protein species of an organism. The technology of 2D-PAGE has become a worthwhile tool for the analysis of gene expression effects caused by cancer, cell signals, pharmaceuticals etc. on the protein level. The isolation of microgram and submicrogram amounts of protein in a 2D-spot in a form suitable for microsequencing and amino acid composition analysis was a key step in the chemical characterization of these proteins. This was possible by development and introduction of suitable membranes which retain proteins present in low quantities from the polyacrylamide matrix with high yields during the blotting process. These immobilized proteins are suitable for direct sequence analysis and amino-acid composition analysis and now with MALDI directly to mass spectrometry. "Laser scanning" of blots would link the high separation capability of 2D-PAGE with the high mass accuracy of mass spectrometric analysis. First "Laser scanning" experiments of 2D-patterns from lymphocytes and plasma proteins will be presented.
References:
1. Eckerskorn, C., Strupat, K., Karas M., Hillenkamp, F., Lottspeich, F., Electrophoresis, 1992, 13, 664-665. Mass spectrometric analysis of blotted proteins after gel elotrophoretic separation by matrix-assisted laser desorption ionization
2. Strupat, K., Karas, M., Hillenkamp, F., Eckerskorn, C., Lottspeich, F., Anal. Chem., 1994, 66, 464470. Matrix-assisted laser desorption ionization mass spectrometry of proteins electroblotted after polyacrylamide gel electrophoresis
Gaston Gonnet*, Manfredo Ouadroni Paola Dainese, Werner Staudenmann and Peter James
Protein Chemistry Laboratory and *Institute for Scientific Computation, Federal Institute of Technology (ETH), 8092 Zurich, Switzerland
Phone 01 632 2919; Fax 01 632 1213; Email bcmass@ezrzl.vmsmail.ethz.ch
Within the next few years the sequences of entire genomes of organisms will become available; the complete sequences of several yeast chromosomes are already available. The complete sequence of the E. coli genome will probably be completed by 1997 and that of yeast by 2000. Larger projects such as the Drosophila and human genomes are targeted for completion within the next twenty years. The databases are starting to double in size each year at the end of 1989 there were 14,000 protein sequences available, five years later there are around 60,000. The Expressed Sequence Tag (EST) database will soon bring about an exponential growth rate. EST database construction involves the creation of a normalized cDNA library (one in which all mRNA's coding for the proteins being expressed in a particular cell type are present in equal amounts). The cDNA is then broken down to random fragments around 300-800 bases long and cloned into a vector. Individual clones are randomly selected and sequenced; this has be automated and centres have been established which can produce over 5,000 sequences a year. Databases are being built for human brain, Caenorhabditis elegans and Zebra fish. Within a few years protein identification will reduce to matching a protein to an entry in a database: the challenge will then be which type of search to use and decreasing the amount of material required to obtain data for a search.
Currently the most sensitive method of protein identification by database searching is sequence searches using N-terminal Edman sequence data. This approach is limited to proteins with free amino termini which are available in amounts of 2-5 picomoles (i.e. an initial yield of 500 femtomoles for the most sensitive methods). Moreover the EST databases are not accessible since the sequences present are randomly distributed over the whole length of the protein and N-terminal data is proportionally low. As the databases grow, so does the length of the N-terminal sequence required to uniquely define a protein above the background of random sequence noise. Only six residues were necessary ten years ago, now at least 10 are needed for a certain match. One solution to this problem was published independently by five groups in 1993 (Bill Henzel Matthias Mann, Darryl Pappin, Peter James, and John Yates) termed Peptide Mass Fingerprinting. The basic premise is that the set of masses of peptides produced by a specific chemical or enzymatic digestion of a protein is unique (an MS fingerprint) and can be used to search a database in which the protein sequences have been replaced by the predicted mass fingerprints. Data collection by matrix assisted matrix assisted laser desorption / ionization time of flight (MALDI-TOEi) mass spectrometry is very simple and sensitive and instrument prices are now less than $150,000.
In order to search DNA sequence databases, especially ESTs, two orthogonal fingerprints must be used to ensure the certainty of a match (1). This can be achieved using two different digestions, or by the chemical modification of a single digestion (acetylation, methylation, or deuterium exchange). A more rigorous method is to use raw MS/MS data obtained from on line HPLC MS on a triple quadrupole, or by Post Source Decay (PSD) on a MALDI-TOF to search the databases (2). Alternatively partial sequence data from PSD-MALDI-TOF (3) or ladder sequencing (4) can be combined with peptide mass data for searching The MS methods can detect peptides down into the low femtomole range and the time required for an analysis and a search is less than 20 minutes. Several instruments offer multiple sample loading platforms with reliable `autopilot' operation for the automated analysis of up to 100 samples which can easily be digested in parallel.
Peptide mass fingerprinting is a viable method for service labs both in regard to instrument price, ease of handling and throughput (the database search programs can be obtained free of charge). The main problem still remains that of sample handling at low levels.
References:
1. James, P., Quadroni, M, Carafoli, E., and Gonnet, G. (1994) Protein Science, 3, 1347-50. Protein identification in DNA databases by peptide mass fingerprinting.
2. Eng, J.K, McConnack, A.L., and Yates, J.R. (1994) J. Am. Soc. Mass. Spec. 5, 976-989. An approach to correlate tandem mass spectral data of peptides with amino acid sequences in a protein database.
3. Mann M., and Wilm, M. (1994) Anal. Chem. 66, 43904399. Error tolerant identification of peptides in sequence databases by peptide sequence tags.
4. Bartlet-Jones, M., Jeffrey, W.A., Hansen, H.F., and Pappin, D.J.C. (1994) Rapid Comm. Mass Spec. 8, 737-742. Peptide ladder sequencing by mass spectrometry using a novel, volatile degradation reagent.
Return to the The ABRF Home Page