Techniques in Protein Chemistry IV

Techniques in Protein Chemistry IV compiles papers presented at the Sixth Protein Society Symposium held in San Diego, California in 1992. This book discusses the mass spectrometry in protein sequence and structural investigations; site specific heterogeneity of N-linked oligosaccharides on recombinant human erythropoietin; and modification of thiophosphorylated proteins with extrinsic probes. The cysteine and tryptophan amino acid analysis of ABRF92-AAA; development of separation strategies for proteins by capillary electrophoresis; and peptide mapping of 2-D gel proteins by capillary HPLC are also elaborated. This text likewise covers the single syringe-pump solid-phase protein sequencer; hydrophobic contact density distribution functions; and application of chemical shift calculation to protein structure determination by NMR. This publication is valuable to biologists and students concerned with the developments in mass spectrometry of proteins.

• Language: English
• Publisher: Academic Press
• Release date: Jun 28, 2014
• ISBN: 9781483268224

Book preview

GELS

Mass Spectrometry in Protein Sequence and Structural Investigations

A.L. Burlingame,     Department of Pharmaceutical Chemistry, The Mass Spectrometry Facility and the Liver Center, University of California, San Francisco, California

Publisher Summary

This chapter discusses the use of mass spectrometry in protein sequence and structural investigations. The growing realization of the importance of mass spectrometry in protein science is based upon quite recent recognition of the relative ease with which classically difficult or often intractable questions in protein biology may now be readily and successfully addressed. The progression of events that underpin the present state of mass spectrometry was triggered by discoveries of revolutionary new ionization techniques that are able to handle polar, labile macromolecules mass spectrometrically and the subsequent development of new instruments utilizing these techniques. Mass spectrometry has a variety of advantages not previously experienced by the biological research community, such as the ability to produce high quality, detailed molecular information at high absolute sensitivity. Mass spectrometry can routinely visualize the entire composition of the sample, and easily handle heterogeneous substances and mixtures of substances. Its main disadvantage at present is the unusual care required to assure suitable sample preparation, particularly in connection with problems where mass spectrometry has not been previously employed.

I Overview

The growing realization of the importance of mass spectrometry in protein science is based upon quite recent recognition of the relative ease with which classically difficult or often intractable questions in protein biology may now be readily and successfully addressed. The progression of events which underpin the present state of the art was triggered by discoveries of revolutionary new ionization techniques, viz. LSIMS or FAB, electrospray and matrix-assisted laser desorption, able to handle polar, labile macromolecules mass spectrometrically and the subsequent development of new instruments utilizing these techniques. These methods are virtually ideally suited to deal effectively with the challenges of protein sequencing and structural analysis at the picomole level. Currently active export of some of this new instrumental capability to the biochemistry laboratories themselves is under way due to the burgeoning availability of more user-friendly, relatively low cost commercial instrumentation. The most recent examples are instruments designed to permit exploitation of the two newest ionization methods, namely electrospray (ES) (1) and matrix-assisted laser desorption (maLD) (2) techniques.

Over the last decade a number of mass spectrometry laboratories have been key participants in establishing the unique power and versatility of these techniques in solving previously tedious and often impossible problems. There is a growing literature of examples of questions which have been tackled successfully representing increasingly more difficult biological problems (See recent reviews 3-6). The present excitement about mass spectrometry in the protein biochemistry community is not due to any recently recognized need to have a better structural understanding of the machinery of cells, but rather to a growing awareness that the tools are now in hand with which to discover and dissect even minute structural details with high accuracy and fidelity through application of these mass spectrometric techniques. Indeed, the time is ripe to exploit these new methodologies, both to gain a deeper understanding of protein biochemistry with a reasonable investment of time and effort, and save labor costs incurred in pursuing outmoded, ineffective biochemical strategies in pursuit of similar goals. With the arsenal of new methods and instruments, and the ease with which definitive answers can now be obtained, it is almost to the point where laboratories cannot afford not to have access to mass spectrometry in order to remain competitive. This point has been well illustrated by the immediate massive investments that several of the leading biotechnology companies have made to procure the entire suite of new instruments and methods of mass spectrometry over the last several years. Already, many traditional academic protein laboratories have begun to add mass spectrometry to their capability. These trends will certainly spread rapidly as lower cost instrumentation becomes commercially available.

II Using the Tools Now Available

How do you go about solving your problem or getting into the business? Initially the easiest way is to become a user or collaborator with an established protein mass spectrometry group. The NIH/NSF biotechnology resources (7) will provide assistance in both how to go about solving your problem and the access to necessary instrumentation and interpretative expertise without the need of making a major investment yourself. Depending on your needs, anything from obtaining a few spectra on your synthetic peptides to taking a mini-sabbatical to learn the techniques and solve your problem firsthand can be arranged.

Regardless of how you choose to proceed, it is of prime importance to discuss how mass spectrometry may eventually be used in solving your problem with an experienced protein mass spectrometrist prior to starting isolation of your protein. This dialogue can save both parties time and considerable mutual frustration by determining at the outset what the best stage in the problem would be to involve the use of the different capabilities of mass spectrometry (including LC/MS) and particularly to avoid use of common biochemical experimental materials which could preclude eventual successful mass spectrometric analysis. These include non-volatile buffers, detergents which are mixtures of polymers (e.g. Triton X-100, NP-40, etc.), plasticized labware, ion exchange columns which bleed polymers, and impure solvents, to mention a few. Volatile buffers such as ammonium acetate or bicarbonate and N-ethylmorpholine acetate, octyl beta glucoside and SDS can be successfully removed (8). You will recall that matrix-assisted laser desorption ionization is said to be more tolerant of the presence of common biochemical cocktails (2) used to preserve biological activity, including SDS. While that is true to some extent, one of maLD’s real practical advantages will probably be felt in the discovery of how to experimentally handle new difficult problems, such as heterogeneous membrane-bound glycoproteins (9), hydrophobic transmembrane components (10), etc. Bear in mind, for sequence and structural analysis of components in your chemical or enzymatic digest, it is likely that it will still be necessary to make use of other ionization techniques, vibrational excitation and ion optical systems possessing different capabilities which will not, unfortunately, tolerate these ingredients in the sample (11).

III The Tools

A Ionization: Determination of Molecular Weight, Purity and Heterogeneity

Three ionization methods are in common use; namely (1) sputtering of surface active samples from a viscous liquid surface by irradiation with kiloelectron volt kinetic energy cesium ion (liquid secondary ion mass spectrometry, LSIMS) or xenon neutral atom(fast atom bombardment, FAB) beam; (2) nebulization of an acidified solution of a sample into an atmospheric pressure desolvation bath gas (electrospray); and (3) irradiation of a solid solution of sample with a laser frequency tuned to the chromophore of the solvent matrix.

These ionization/ejection techniques can be thought of as sampling of acid-base type ionization equilibria existing in solution, either by protonation of a basic site in a given molecule (M) under acidic conditions to form MH+ or conversely, deprotonation of an acidic site to form (M-H)−. These processes are sometimes referred to as soft ionization due to the fact that the ejection (desorption and desolvation) of these naked MH+ or (M-H)− species into the gas phase occurs while maintaining their ambient vibrational energy content, thus creating a preponderance of low internal energy intact molecular species. This is an ideal situation for very sensitive detection of molecules and their mixtures, but obviously not for investigation of sequence or structural questions as discussed below. It should be mentioned in passing that the accuracy with which mass can be measured using any ionization method depends on the quality and performance as well as the proper calibration of the ion optical system used (4). It is crucial to be able to routinely measure the nominal mass of peptides up to 3000-4000 Da to ± 0.2 Da on the ¹²C mass scale in order to be sure of a proper composition assignment (for example the mass difference of asn vs. asp or a C-terminal carboxyl vs. C-terminal amide is only 1.0 Da).

B Collision-Induced Dissociation of MH+ or (M-H)- (CID)

Sequence and structural information can be obtained readily by interpretation of the mass spectra obtained by using tandem mass spectrometers. Tandem mass spectrometers function both to separate (in MS-1) a selected component from impurities or mixtures as well as to vibrationally activate that selected component to undergo gas phase unimolecular dissociation reactions then recorded by MS-2 as its collision-induced dissociation (CID) mass spectrum. There are two ion activation energies currently in common use, namely high and low, which are carried out on four sector and triple quadrupole instruments, respectively (12). As might be expected intuitively, the high energy mode deposits more vibrational energy in the MH+ ion selected and hence in addition to peptide backbone cleavages (also observed under low energy conditions), both amino acid side chain cleavages and immonium ions are formed as well. This class of higher energy fragmentation processes is of considerable importance in permitting the distinction between and assignment of the isobaric pair leu/ile (13), the location (14–16) and structure (15, 16) of covalent modifications and induction of saccharide ring cleavages which permit branching assignments in C-terminal glycan anchors (17) to give a few examples.

C LC Coupled ES-MS

The combination of reverse phase liquid chromatographic systems with electrospray mass spectrometers has become a routine operation in many laboratories. It is the first LC/MS combination which is technically virtually ideally suited for the molecular weight mapping of protein and glycoprotein digests. It has replaced flow FAB due to superior performance in the size range accessible, and the sensitivity and faithfulness of detection of most components ranging from the very hydrophilic tri-mer sizes (but retained on the RP-HPLC column) to all but the most hydrophobic in a single gradient (18, 19). For example, detection of larger peptides from incomplete digestion and large glycopeptides may also be readily achieved. From selected reaction monitoring of monosaccharide ions formed under collisional activation conditions (20, 21), the retention times of both small and large glycopeptides may be detected.

IV Examples Using the Tools

A Molecular Weight Determination, Purity

At this writing electrospray ionization employing either quadrupole or sector mass analyzers offers the highest accuracy of mass measurement at the highest mass resolutions (18, 19, 22). In addition, sensitivities are adequate for quadrupoles [low picomole quantities (18, 19, 23)] and are significantly better (low femtomoles) using sector instruments with multichannel array detection systems (24).

It is straightforward to obtain an accurate measurement of the molecular weight together with reasonably faithful detection, molecular weight measurement and semi-quantitation of other similar components present in a peptide or protein sample. For example, consider the ES-MS of recombinant IGF shown in Figure 1. As can be seen from the signals accompanying the multiply (MH7⁷+ chosen for example) protonated molecular ion region occurring around m/z 1090, six additional mass resolved IGF-related polypeptides are present at the 1-3% level. The masses observed for these correspond to components having N- and C-terminal truncations, methionine oxidation and mono- and di-mannosylation (25).

Figure 1 The ES spectrum of recombinant IGF illustrating the detection and identification of minor components at the 1-3% level. (Reproduced with permission from ref. 3.)

A further example concerns analysis of a sample of commercial ribonuclease B. The ES-MS is shown in Figure 2. Signals representing six significant molecular species are observed between 13,500 and 16,000 Daltons. These signals correspond to the presence of both ribonuclease A and several glycoforms of ribonuclease B (26). The molecular weights of such samples can be measured by maLD-TOF and maLD-sector instruments as well. Spectra of ribonuclease B obtained on both TOF and double focusing instrumentation are shown in Figures 3 and 4, respectively. Similar molecular species are observed in the maLD-TOF spectrum (Figure 3) of ribonuclease B with consumption of less sample, but recorded with significantly lower mass resolution. It is illustrative to note that this particular preparation of ribonuclease B contains additional components such as peak B, corresponding to ribonuclease A with one GlcNAc attached (27). Thus, it is readily apparent from two 5 minute experiments that these two preparations are not identical, and one can clearly discern their significant differences in molecular composition. Of course, these molecular weight measurements suggest, but certainly do not prove the precise structural nature of each component. To obtain that information would require further work, probably involving CID spectral analyses of enzymic digest components (see discussion in refs 3–6). Turning to yet another maLD spectrum (Figure 4) of a third preparation of ribonuclease B, both the TOF and sector spectra illustrate the enhanced mass resolution which may be obtained from use of the double focusing instrument (28). Note that these latter spectra indicate the presence of only the high mannose glycoforms of ribonuclease B.

Figure 2 Electrospray mass spectrum of ribonuclease B (bovine pancreatic, Sigma R 5500) purified by reverse phase HPLC. Solvent: H2O:CH3OH 1:1, 2% acetic acid. (Reproduced with permission from author.)

Figure 3 The protonated molecular ion region of the mass spectrum of a sample of bovine pancreatic ribonuclease B, using sinapinic acid as the matrix and λ = 355 nm. The peaks are: (A) ribonuclease A; (B) ribonuclease A + 1 N-acetylglucosamine; (C) - (F) different glycosylation states of ribonuclease B, i.e. ribonuclease A + di-N-acetylchitobiose + 5-9 mannose residues. (Reproduced with permission.)

Figure 4 Matrix-assisted laser-desorption spectrum of the glycoprotein ribonuclease B. [M+H]+ 14 900.4 (average). Left: TOF spectrum [obtained with a modified VESTEC 2000 instrument (VESTEC Corp., Houston, TX. USA)] showing only the molecular-ion region, acquired with nitrogen laser (337 nm) using sinapinic acid as a matrix. Right: magnetic sector spectrum acquired as a single segment using an excimer laser (353 nm) and DHB as a matrix. (Reproduced with permission from ref. 28.)

Finally, using ES with double focusing instrumentation higher mass resolution can be readily obtained. For example, in studying genetic variants of human hemoglobin by measurement of the molecular weights of both the intact a- and β-chains, it is easy to detect amino acid substitutions when the mass difference is large enough to be resolved by the mass analyzer. This is illustrated in Figure 5 for the hemoglobin of a patient with Hemoglobin Yamagata. The asn¹³²→lys in the β-chain differs by only 14 Da. at Mr 15,867.2 (normal β-chain) and is clearly mass resolved with the instrument resolution of 2,000 used in making this measurement (22). A mass difference this small at 16,000 Da. would have been difficult to fully resolve using a normal quadrupole mass analyzer.

Figure 5 Partial ESI mass spectrum of hemoglobin chains from a patient with Hemoglobin Yamagata at a resolution of 2000. The shape of the doublet of 17-charged ions was compared with the peak shape for the theoretical isotopic distribution at a resolution of 15,000. (Reproduced with permission from ref. 22.)

It should be noted that for ES-MS the sample must be soluble in the solvent (e.g., water, acetonitrile, 0.1% TFA) being employed for introduction in to the ion source. Hydrophobic and integral membrane proteins may be run in appropriate solvents (29). At this point in time, ES-MS works well up to around 100,000 Da. (1), while maLD would be the method of choice up to 250,000 Da. (2), and also for measurement of mixtures of smaller proteins (30).

B Peptide Mixture Analysis

All of these ionization methods can be used to investigate mixtures of peptides, including those obtained from protein or glycoprotein digests. Exactly how to proceed depends upon the information sought. All of these methods work fairly well on simple mixtures, especially with components of similar hydrophilicity/hydrophobicity ratios (such as RP-HPLC fractions). More complex mixtures should be separated (at least partially) and run either in a batch mode (fractions) or by using a small aliquot on-line using LC-ES-MS (18, 19, 31). The latter has the advantage of providing retention time information as well as molecular weights for each component.

C Glycopeptides

The degree of elaboration of glycosylation on a given glycopeptide will determine its extent of hydrophilicity as compared with the same un- or deglycosylated peptide sequence. Very hydrophilic substances are usually not sputtered with good sensitivity using LSIMS (3). However, ES-MS and LC-ES-MS work particularly well as noted earlier (see Section IIIC). Many relatively simple examples have been reported (3-5, 11, 18, 19, 23, 31).

However, N-linked glycosylation at even a single asparagine site can be very complex (extreme microheterogeneity). ES-MS is ideal for determination of the global qualitative and semi-quantitative degree of glycoform heterogeneity of such systems. A case in point involves the tryptic glycopeptide (residues 111-185) containing asn¹⁸¹ in the hamster scrapie glycoprotein (32). The calculated mass of this sequence is 8608.6 Da. without the incremental mass of the glycosylation. This glycopeptide was not observed in attempted LSIMS analysis. However, it was readily observed by electrospray carried out in the batch inlet mode as shown in Figure 6a. While the mass resolution and mass assignment were not optimal in the recording of the raw data from very limited quantities of this glycopeptide, the number of components and their proper mass assignments were clearly determined through use of a deconvolution (image enhancement) algorithm called Maximum Entropy (33) as shown in Figure 6b. To establish that the over 30 glycoforms detected were not artifacts, PNGase F was used to deglycosylate this glycopeptide. The ES-MS of the resulting asp¹⁸¹-peptide is shown in Figure 7. The measured molecular weight corresponds to that of the expected deglycosylated peptide within experimental error.

Figure 6a Electrospray spectrum of the MH10+10 species from lys C asn¹⁸¹-glycopeptide isolated from the scrapie glycoprotein by RP-HPLC.

Maximum Entropy deconvolution result of electrospray spectrum shown in 6a.

Figure 7 Electrospray spectrum of the PNGase F peptide product (asp¹⁸¹-111→185) of the scrapie N-linked site at asn¹⁸¹.

D Sequence and Structure Elucidation

As indicated above, these soft ionization methods do not usually provide sufficient fragmentation in a given mass spectrum for structure elucidation purposes without vibrational activation. In addition, even low energy collisional activation is not usually sufficient for unambiguous sequence determination or the characterization of true unknowns (34). However, low energy activation may be adequate for certain types of problems such as the location of amino acid substitutions in protein genetic variants (35), verification of recombinant protein expression (18), quality control of batches of recombinant protein preparations (19), and partial sequence analysis in conjunction with Edman degradation on the same sample (36). We have seen that both matrix assisted laser desorption (maLD) and electrospray techniques can provide direct measurement of the molecular weight and purity or heterogeneity of a protein. Despite the fact that electrospray presently offers the highest accuracy of mass measurement available on intact proteins (for quadrupole instruments, ±.01%), even this level of accuracy is not sufficient to determine some detailed features, such as whether a single disulfide bond is oxidized or reduced (± 2 Daltons) even in an intact small protein such as shark liver fatty acid binding protein (37), or whether a single asparagine is deamidated to aspartic acid (38). The latter was shown to be the case unambiguously by high energy CID sequence analysis of a tryptic peptide in myoglobin being used as a known mass standard for maLD mass scale calibration. Of course, higher levels of mass measurement accuracy are readily achieved using double focusing sector instrumentation with electrospray ionization (22, 24).

Finally, we turn to a brief discussion of the power and versatility of high sensitivity four sector tandem mass spectrometry as a general tool for sequence and structure determination. The first example concerns recent collaborative work from our laboratory on determination of the primary sequence of Galβ1-3(4)GlcNAc α2,3-sialyltransferase, the membrane bound enzyme which catalyzes the final step in the biosynthesis in the sialyl Lewis × antigen, the tetrasaccharide moiety recognized by the E-selectin receptor. This enzyme was first purified from rat liver almost a decade ago by our collaborators, Paulsen and co-workers, and since then several attempts were made to obtain sufficient sequence for cloning, as well as raise antibodies to it.

These attempts were unsuccessful due to the small amounts of impure protein which were readily obtainable.

Twelve micrograms (300 pmoles) of this 48 kDa protein were isolated from the membranes of 2 kg of rat liver. The active protein was stored in 350 ml of 30 mM sodium cacodylate (pH 6.5), 100 mM sodium chloride, 0.1% Triton® CF 54, and 50% glycerol, as described elsewhere (8). As discussed above, from a mass spectrometric point of view this nonvolatile buffer and detergent combination could have thwarted the successful sequence analysis of this protein preparation. To overcome these problems we pursued the strategy outlined briefly below. We chose to reduce and carboxymethylate this protein in a mixture of this buffer and additional Tris and guanidine hydrochloride prior to dialysis of the carboxymethylated protein against 4 1 of 50 mM N-ethylmorpholine acetate. SDS was added to the dialysis wells late in the dialysis to make up a total of 0.1%. After dialysis, rigorous removal of SDS and remaining traces of Triton® CF 54 was carried out by acetone precipitation of the protein. The protein pellet was then taken up in a buffer of Tris, 2 M urea and calcium chloride and digested with trypsin for 18 h. Separation of the digest was carried out using a reverse phase microbore HPLC and some 50 fractions were collected. Microbore HPLC is advantageous for providing small solvent volumes for these picomole levels of peptide, since the fractions must be kept in solution (not blown down to dryness) to avoid serious adsorptive losses. Thus, microbore HPLC provides appropriate concentrations of peptides in the fractions for mass spectral analysis and sequencing. The HPLC chromatogram of this digest is shown in Figure 8. Upon measurement of the molecular weights of these fractions by liquid SIMS, it was clear that most of these fractions were mixtures of peptides, such as was the case for fraction 22 shown in Figure 9. This fraction contained 5 peptides, two of which were sequenced by high energy CID analysis without further separation. For example, selection of one of the protonated molecular ions at 1283.6 and its collisional activation provided the CID spectrum shown in Figure 10. Interpretation of this CID spectrum established the sequence LTTALDSLHC*R. The second component selected, protonated molecular weight 996.4, was determined to be VSASDGFWK (data not shown). From this single experiment, the sequence of 99 of a total of 374 amino acids of this protein were obtained (39). The enzyme was then cloned using PCR methodology. Finally, the knowledge of the molecular weights of other components in various HPLC fractions was used to verify the correctness of the final cDNA-deduced protein sequence.

Figure 8 HPLC chromatogram of the tryptic digest of the Galβ1,3(4)GlcNAc α2,3-sialyltransferase. Sixty fractions were collected and subjected to mass spectrometric analysis. Fractions labeled yielded sequences by CID analysis, or the eluting peptides were later identified by molecular weight based on the cDNA-deduced amino acid sequence. (Reproduced with permission from J. Biol. Chem.)

Figure 9 LSIMS spectrum of fraction #22 from the tryptic digest of the Galβ1,3(4)GlcNAc α2,3 sialyltransferase. The spectrum was recorded in the positive mode, using glycerol:thioglycerol 1:1 acidified with 1% TFA as matrix.

Figure 10 CID spectrum of peptide T-22B from the Galβ1,3(4)GlcNAc α2,3 sialyltransferase. The peptide sequence is Leu-Thr-Pro-Ala-Leu-Asp-Ser-Leu-His-Cys*-Arg, MH+ = 1283.6. Cys* represents carboxymethyl cysteine. Ions with charge retention at the N-terminus are labeled as a, b, c ions and the C-terminal analogs are designated as x, y, z. The a and x ions are products of a cleavage between the α carbon and the carbonyl group. Ions y and b are formed when the peptide bond is cleaved. Ions c and z are present due to the cleavage between the amino group and the α carbon. The numbering of these fragments is always initiated at the appropriate peptide terminus. The side chain fragmentation occurs between the β and γ carbons of the amino acids, yielding the so-called d (N-terminal) and w(C-terminal ions. C-terminal vi ions are formed via α,β bond cleavage from a number of ions, for example, from yi ions. The immonium ions and the side chain losses from the protonated molecular ion are indicated by the one letter code of the corresponding amino acids. Internal fragment ions are labeled similarly. The nominal masses of all sequence ions observed are shown together with the sequence deduced above the spectrum.

The last example concerns our collaborative work on the site of attachment and structural nature of the scrapie glycoprotein phosphatidylinositol glycan membrane anchor. Our investigation of the hamster scrapie glycoprotein is an interesting example of a challenging structural problem exacerbated by the availability of only low nanomole to subnanomole quantities of material at any given time. The complexity of glycosylation at asn¹⁸¹ was discussed above. This discussion will summarize briefly our work on the GPI anchor, since all techniques of current mass spectrometry were helpful in complementary ways in addressing this structural problem.

Since the protein preparation PrP 27-30 used in many of these studies is sensitive to phospholipase C, it was employed to remove the glycerol acyl ether from the phosphatidylinositol glycan attached to the protein C-terminus. Upon subsequent digestion of the resulting glycoprotein with endoproteinase lys C, the C-terminal peptide-GPI anchor was isolated by reverse phase HPLC. Treatment of this fraction with 50% HF at 4° permitted isolation of the peptido-ethanolamine as established by measurement of its protonated molecular weight by LSIMS as 1374. This figure was consistent with predicted lys C cleavage at lys²²⁰ in the cDNA sequence, and the known phosphodiester selective cleavage properties of cold aqueous HF. This experiment established the GPI attachment site at ser²³¹ (40). Another experiment was then carried out to obtain the very hydrophilic non-retained fraction eluting in the HPLC void volume. This fraction was subjected to permethylation to increase its hydrophobicity for LSIMS analysis. The permethylated product was extracted with chloroform and subjected to LSIMS. Three components were detected at 1312, 1557, and 1761 Da. In order to obtain structural information on these components, they were analyzed by selection of the ¹²C isobar of each component in MS-I of a four sector tandem instrument. High energy CID spectra of each component from this mixture were recorded (41) and are shown in Figure 11. Interpretation of these CID spectra provided the sequence and branching of the glycans shown in Figure 12. However, since it is known that HF treatment might degrade unsuspected features of the native structure, further experiments using the native lys C-derived C-terminal peptido-glycan were performed upon availability of ES-MS capability in our laboratory. From the ES measurement of the molecular weights in this preparation, larger components were detected which corresponded to one additional phosphoethanolamine moiety and one sialic acid residue. In order to place the attachment of the sialic acid residue, another more careful aqueous HF, permethylation was carried out. This experiment resulted in detection of two additional higher mass components in LSIMS at 1918.8 and 2122.9 Da. High energy CID analysis was then employed to determine that the sialic acid residue was attached to the branched portion of the anchor glycan shown in Figure 13 (11). Finally, maLD experiments were performed on various preparations of the scrapie glycoprotein to obtain a global view of the complexity and heterogeneity of the protein preparation (42), after enzymatic removal of the two heterogeneous N-linked carbohydrates. These analyses (shown in Figure 14) have been challenges even for maLD, but our present data was obtained by solubilization of these membrane proteins in hexafluoroisopropanol before mixing with chloroform/methane/water containing the maLD matrix, 4-hydroxy-α-cyanocinnamic acid.

Figure 11 High energy CID spectra of the permethylated glycan species of m/z 1312, 1557, and 1761 (nominal masses rounded down to integer values), in the mass range > 500 Da. (Reproduced with permission from ref. 17.)

Figure 12 Structures of the glycans corresponding to the permethylated ions of m/z 1312 (1), 1557 (2a and 2b), and 1761 (3). The formation of the ions is illustrated. (Reproduced with permission from ref. 17.)

Figure 13 GPI anchor glycoforms. (Reproduced with permission from ref. 11.)

Figure 14 LDMS of PrPSc and PrpC

It is clear that without judicious use of microchemical manipulations and all of the new ionization and instrumental techniques in the arsenal, problems such as the GPI anchored scrapie glycoprotein would not be structurally tractable.

V Conclusions and Outlook

Mass spectrometry is destined to become a commonplace set of tools alongside HPLC in the biomedical research laboratory. It has a variety of advantages not previously experienced by the biological research community, such as the ability to produce high quality, detailed molecular information at high absolute sensitivity (comparable to the best biological-based assays). Mass spectrometry can routinely visualize the entire composition of the sample (including things you don’t care to know about), and easily handle heterogeneous substances (e.g. glycosylation, phosphorylation, etc.) and mixtures of substances. Its main disadvantage at present is the unusual care required to assure suitable sample preparation, particularly in connection with problems where mass spectrometry has not been previously employed.

While the investment in sufficient instrumentation and skilled practitioners of the art may be on the high side for enjoyment of the full power of the arsenal, this situation is beginning to change rapidly. Hence, it should be anticipated that growing reliance on the daily use of these versatile and powerful new tools (so clearly in evidence at this Sixth Symposium) will become widespread in the biomedical community and the biotechnology industry in the near future.

Acknowledgments

I wish to acknowledge my co-workers and collaborators who contributed to various examples discussed in this chapter. Financial support was provided by the Biomedical Research Technology Program of the National Center for Research Resources of the National Institutes of Health (Grant RR01614), the Biological Instrumentation Program of the National Science Foundation (Grant DIR-8700766), National Institute of Diabetes and Digestive and Kidney Diseases (Grant AM 27643), and the NIEHS Superfund Program (Martyn Smith, UCB) Grant ES04705.

REFERENCES

1. Fenn, J.B., Mann, M., Meng, C.K., Wong, S.F. Science. 1989; 246:64.

2. Hillenkamp, F., Karas, M., Beavis, R.C., Chait, B.T. Anal. Chem. 1991; 193:1193A.

3. Burlingame A.L., McCloskey J.A., eds. Biological Mass Spectrometry. Elsevier Science Publishers: Amsterdam, 1990:700.

4. McCloskey, J.A. Methods in Enzymology. 1990; 960.

5. Carr, S.A., Hemling, M.E., Bean, M.F., Roberts, G.D. Anal. Chem. 1991; 63:2802–2824.

6. Burlingame, A.L., Baillie, T.A., Russell, D.H. Anal. Chem. 1992; 64:467R–502R.

7. See listing in NIH Publication No. 92-1430, Directory, Resources for Biomedical Technology, which may be obtained by calling 301-251-4970.

8. Wen, D. X., Livingston, D. D., Medzihradszky, K., Kelm, S., Burlingame, A. L., Paulson, J. C., J. Biol. Chem. in press 1992; 64:20

9. Baldwin, M.A., Wang, R., Pan, K.-M., Hecker, R., Stahl, N., Chait, B.T., Prusiner, S.B. This volume 1992;

10. Unpublished results, this laboratory

11. Stahl, N., Baldwin, M.A., Hecker, R., Pan, K.-M., Burlingame, A.L., Prusiner, S.B. Biochemistry. 1992; 31:5043.

12. For a recent discussion of these differences see ref. 6 478R-480R references therein

13. Biemann, K., Scoble, H.A. Science. 1987; 237:992.

14. Kaur, S., Hollander, D., Haas, R., Burlingame, A.L. J. Biol. Chem. 1989; 264:16981.

15. Buechter, D.D., Medzihradszky, K.F., Burlingame, A.L., Kenyon, G.L. J. Biol. Chem. 1992; 267:2173.

16. Ojingwa, J., Ding, A., McDonagh, A. F., Burlingame, A. L., Benet, L. Z. Proc. Natl. Acad. Sci. USA. submitted. 1992.

17. Baldwin, M.A., Stahl, N., Reinders, L., Gibson, B.W., Prusiner, S.B., Burlingame, A.L. Anal. Biochem. 1990; 191:174.

18. Hemling, M.E., Roberts, G.D., Johnson, W., Carr, S.A., Covey, T.R. Biomed. Env. Mass Spectrom. 1990; 19:677.

19. Ling, V., Guzzetta, A. W., Canova-Davis, E., Stults, J. T., Hancock, W. S., Covey, T. R. Shushan B.I. Anal Chem. 1991; 63:2909.

20. Carr, S.A., Huddleston, M.J., Bean, M.F. Proceedings of the 40th Annual Conference on Mass Spectrometry and Allied Topics. 1992; 973–974.

21. Schindler, P.A., Hall, S. C. Collet, X., Fielding, C.J., and Burlingame, A.L. (in preparation).

22. Wada, Y., Tamura, J., Musselman, B.D., Kassel, D.B., Sakurai, T., Matsuo, T. Rapid Commun. Mass Spectrom. 1992; 6:9–13.

23. Maltby, D., Hall, S.C., Medzihradszky, K., Burlingame, A.L. Proceedings of the 40th Annual Conference on Mass Spectrometry and Allied Topics. 1992; 24–25.

24. Musselman, B.D., Tamura, J., Cody, R.B. Abstracts, Kyoto ’92 International Conference on Biological Mass Spectrometry. 1992; 97.

25. Poulter, L., Green, B.N., Burlingame, A.L.Burlingame A.L., McCloskey J.A., eds. Biological Mass Spectrometry. Elsevier: Amsterdam, 1990:119–128.

26. Bitsch, F. (1991). Ph. D. Thesis, University of Strasbourg.

27. Beavis, R.C., Aduru, S., Chait, B.T. Proceedings of the 38th Annual Conference on Mass Spectrometry and Allied Topics. 1990; 281–282.

28. Annan, R.S., Köchling, H.S., Hill, J.A., Biemann, K. Rapid Commun. Mass Spectrom. 1992; 6:298–302.

29. Falick, A.M., Schindler, P.A., van, Dorsselaer. Proceedings of the 40th Annual Conference on Mass Spectrometry and Allied Topics 1992; 1625–1626

30. Beavis, R.C., Chait, B.T. Proc. Natl. Acad. Sci. USA. 1990; 87:6873–6877.

31. Medzihradszky, K. F., Settineri, C. A., Maltby, D. A., and Burlingame, A. L. (1992). This volume.

32. Stahl, N., Baldwin, M.A., Teplow, D.B., Hood, C, Gibson, B.W., Burlingame, A.L., and Prusiner, S.B. Biochemistry, submitted.

33. Max. Entropy Software, Fisons Instruments, UK.

34. See discussion and references therein in reference6, pp.478R-480R

35. Witkowska, H.E., Bitsch, F., and Shackleton, C.H. L. (1992). Hemoglobin, submitted.

36. Hunt, D.F., Michel, H., Dickinson, T.A., Shabanowitz, J., Cox, A.L., Sakaguchi, K., Appella, E., Grey, H.M., Sette, A. Science. 1992; 256:1817–1820.

37. Medzihradszky, K.F., Gibson, B.W., Kaur, S., Yu, Z., Medzihradszky, D., Burlingame, A.L., Bass, N. Eur.J. Biochem. 1992; 203:327–339.

38. Zaia, J., Annan, R.S., Biemann, K. Rapid Commun. Mass Spectrom. 1992; 6:32.

39. Burlingame, A. L., Medzihradszky, K. F., Wen, D. X., Livingston, B. D., Kelm, S., Paulson, J. C. "Abstracts, LVII Cold Spring Harbor Symposium on Quantitative Biology:. The Cell Surface. 1992; 32:32. [May 27-June 3].

40. Stahl, N., Baldwin, M.A., Burlingame, A.L., Prusiner, S.B. Biochemistry. 1990; 29:8879–8884.

41. Baldwin, M.A., Stahl, N., Burlingame, A.L., Prusiner, S.B. In Methods: A Companion to Methods in Enzymology. 1990; 1:306–314.

42. Baldwin, M.A., Pan, K-M., Hecker, R., Wang, R., Beavis, R., Chait, B., Stahl, N., Burlingame, A.L., Prusiner, S.B. Proceedings of the 40th Annual Conference on Mass Spectrometry and Allied Topics. 1992; 1593–1594.

Elucidation of Covalent Modifications and Noncovalent Associations in Proteins by Electrospray Ionization Mass Spectrometry

Joseph A. Looa¹, Rachel R. Ogorzalek Looa², David R. Goodletta, Richard D. Smitha, Alfred F. Fuciarellib, David L. Springerb, Brian D. Thrallb and Charles G. Edmondsab,     aChemical Sciences Department, Pacific Northwest Laboratory, Richland, WA 99352; bBiology and Chemistry Department, Pacific Northwest Laboratory, Richland, WA 99352

Publisher Summary

This chapter discusses elucidation of covalent modifications and noncovalent associations in proteins by electrospray ionization mass spectrometry (ESI–MS). ESI–MS provides powerful insight into the structure of intact proteins and their enzymatically derived peptides. ESI affords a distribution of multiply charged molecular ions that may be analyzed by, for example, conventional quadrupole mass analyzers. The elucidation of primary structure of unknown proteins is one important area of application. The gentle nature of the ESI process permits noncovalent associations of some proteins to be probed by this technique. Proteins are folded into specific molecular conformations maintained by noncovalent interactions. Proteins interact with other molecules to form complex structures, accounted for by noncovalent interactions within and between macromolecules. ESI–MS has demonstrated the capability for examining a range of protein noncovalent interactions. ESI–MS can be used to study protein dimers and peptide–protein associations. The chapter discusses use of ESI–MS to analyze a variety of noncovalently bound complexes.

I Introduction

Electrospray ionization mass spectrometry (ESI-MS) (1, 2) provides powerful insight into the structure of intact proteins and their enzymatically derived peptides (3, 4). ESI affords a distribution of multiply charged molecular ions which may be analyzed by, for example, conventional quadrupole mass analyzers (M/ΔM ≤ 1000) and mass may be determined with precision (± 0.01%) substantially better than available by conventional biochemical techniques. The elucidation of primary structure of unknown proteins is one important area of application. Additionally, where the primary structure of the analyte is constrained (e.g., among mutant protein species) or where a known structure has been subjected to chemical or natural post-translational transformations, this technique provides an extraordinarily sensitive and selective probe of the nature and extent of these modifications. For example, acrylamide adducted hemoglobin subunits and the products of proteolytic digestion may be directly analyzed by ESI-MS with potential application as a biomarker of occupational exposure. Similarly, the radiation induced modifications and the natural post-translational modifications of histone proteins may be studied.

In contrast, the gentle nature of the ESI process permits noncovalent associations of such species to be probed by this technique. Proteins are folded into specific molecular conformations, maintained by noncovalent interactions. Proteins interact with other molecules to form complex structures, accounted for by noncovalent interactions within and between macromolecules. ESI-MS has demonstrated the capability for examining a range of protein noncovalent interactions. ESI-MS can be used to study protein dimers and peptide-protein associations. In order to better understand the nature of these interactions, we have used ESI-MS to analyze a variety of noncovalently bound complexes.

II Materials and Methods

The ESI source, utilizing a liquid sheath electrode, has been previously described (2). Methanol or distilled deionized water was used as the sheath liquid, with the applied electrospray voltage at +4 kV and +5.5 kV, respectively. Sulfur hexafluoride was used as the ESI source sheath gas and is an effective electron scavenger to suppress corona discharge, especially at the higher ESI voltages necessary for spraying highly aqueous solutions (i.e., water sheath) (5). The single quadrupole mass spectrometer (Extrel components, Pittsburgh, PA; m/z limit 2000) and triple quadrupole mass spectrometer (Sciex TAGA 6000E, Thornhill, Ontario, Canada; m/z limit 1400) utilize a modified atmospheric pressure inlet incorporating a differentially pumped interface, also previously reported (2). Adjustment of the voltage difference between the nozzle and skimmer elements (Δ NS) is used to control droplet desolvation in this region, and at higher levels of Δ NS, collisional dissociation of the molecule.

Reference proteins were purchased from Sigma Chemical (St. Louis, MO) and were used without further purification. Peptide and protein solutions were prepared in distilled deionized water with addition of various amounts of acetic acid (typically 5% v/v). A syringe pump controlled the analyte delivery flow rate to the ESI source at 0.35 to 0.50 μL min−1. A separate syringe pump delivered the sheath liquid flow at a rate of 3 μL min−1.

HPLC separations for the hemoglobin subunits and for the proteolytic digests were on C4 and C18 reversed phase columns, respectively, with aqueous trifluoroacetic acid (TFA)/acetonitrile gradient. Detection and quantification of ¹⁴C acrylamide adducted species were by standard liquid scintillation techniques. V8 proteinase digestion was carried out in ammonium carbonate (pH 7.8) buffered solution with enzyme/substrate ratio of 1/50.

Histone proteins were isolated from fresh chicken erythrocytes (10 ml packed volume). The cells were lysed by freezing and thawing in lysis buffer (0.01 M MOPS, pH 7.2; 0.15 M NaCl; 0.2% Nonidet NP40), washed twice with lysis buffer without detergent, and nuclei were pelleted at 2000 × G for 10 minutes. Proteins were extracted from nuclei in 0.4 N H2SO4 by stirring on ice for 4 hours, and the acid insoluble material was removed by centrifugation. The supernatant was dialyzed against water overnight and concentrated in Centricon-10 cartridges (Amicon). Histones were separated and collected by reversed-phase HPLC with a Biorad RP318 column and a solvent system of acetonitrile and 0.1 % TFA. Purified H2B fraction was mixed with 10x or 100x concentrations of thymine or thymidine in deionized water and irradiated in a ⁶⁰Co gamma source to total doses ranging from 1 to 10 Gy while saturated with nitrous oxide.

III Results and Discussion

A Analysis of Covalent Modifications by ESI-MS

1 Characterization of acrylamide adducted hemoglobin

Hemoglobin (Hb) adducts have been used to monitor occupational exposure to a number of compounds. Until recently the most common procedure to identify and characterize Hb adducts has been to cleave the adduct from the protein and derivatize and characterize it by gas chromatography-mass spectrometry (GC-MS). To extend these approaches, we examined acrylamide adducted Hb using ESI-MS in an attempt to identify the adducting moiety as well as to obtain information on the amino acid residues involved. ¹⁴C acrylamide and purified human Hb (type A0) were incubated under conditions that yielded high adduct levels. ESI-MS analysis of the intact HPLC-separated subunits indicated distinct differences in the ion profiles for the adducted and nonadducted species. Mass spectra of the intact β-subunit demonstrated three distinct mass increments of approximately 71, 102 and 136 Da. ESI-MS of HPLC purified fractions of the V8 proteinase digestion localizes the first of these on the peptide bearing Cys-93 (V91-101). The measured mass increment for this species (MrMeas 71.5 ± 1.0) is consistent with the expected thioether derivative (Mrcalc 71.08). The structure of the second observed adduct is unknown but it is likewise localized on cysteine bearing peptides, including V102-121 which bears Cys-112. Similarly, the structure of the third adducted species is unknown and as it does not appear among the adducted peptides we hypothesize that it is not localized on a particular amino acid residue, e.g., the cysteine bearing peptides. Studies to confirm the cysteine residue as the location of the adducting moiety by tandem mass spectrometry and to further refine the picture of acrylamide adduction in this system are in progress.

2 Characterization of radiation induced modifications of histones

A basic subunit of chromatin structure is the nucleosome consisting of 146 base pairs of DNA wrapped around a central histone protein octamer with two additional histones binding the linker regions. The histones, lysine and arginine rich proteins, participate in this reversible complex with DNA called nucleohistone. Exposure of cellular constituents to free radicals generated with ionizing radiation results in modifications which can play a significant role in altered cellular function. Interaction of hydroxyl radicals with DNA results in modifications to base and deoxyribose moieties. In addition, hydroxyl radicals can result in generation of DNA-protein cross links in nucleohistone. Methodology involving acid hydrolysis of irradiated nulceohistone followed by trimethylsilylation and subsequent GC-MS has revealed the presence of DNA base-amino acid dimers (6). However, sequence information is not provided by this methodology. Application of ESI-MS to the measurement of DNA-protein cross links in nucleohistone provides a useful method to obtain information not available with other mass spectrometric