Matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS) of high mass proteins (> 100,000 Da), directly deposited on polyethylene membranes, is demonstrated. The spectral quality obtained, using standard sample preparation conditions, is equal or superior to that obtained with metal sample stages. Compared to the use of poly(vinylidene difluoride) transfer membranes, this material allows the acquisition of excellent quality MALDI mass spectra from high-mass proteins with a standard UV laser. This gain in capability is not at the expense of either mass accuracy or signal reproducibility; both approach that obtained with standard sample preparations on stainless steel. In addition, for the applications shown, the use of PE as a sample support reduces the severe ion suppression effects typically observed in the MALDI analysis of high-mass mixtures. This permits more precise mass measurements to be made via the use of internal calibration and is illustrated by the mass measurement of a chimeric mouse/human antibody (MW approximately 150,000 Da) by coaddition of bovine albumin dimer (MW approximately 130,000 Da).

No abstract

Monitoring the stability of immunoglobulin G (IgG) type antibodies is a crucial analytical issue spanning a wide variety of immunological/biotechnological studies, which includes the analysis of conjugated IgG's for drug delivery. Capillary electrophoresis (CE) has proven valuable for the analysis of proteins and has the potential to separate and detect native antibody components. An ideal complement to CE, which is capable of providing the desired detection specificity to provide species identification information, is matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS). Utilizing these two techniques we have developed an antibody examination procedure and monitored the degradation of an internalizing chimeric (human/mouse) monoclonal antibody (BR96). Electropherograms of the antibody after up to 166 h of thermal stress are presented; MALDI mass spectra of the stressed antibody were acquired at the same time points. At 166 h, the percentage of ionization carried by the intact antibody molecular ions M+, M2+, etc., had clearly decreased, while that due to additional ion species had significantly increased. Ions corresponding in mass to loss of one light chain, loss of an Fab arm to yield an Fab/c type fragment, and formation of separated heavy-chain and light-chain moieties were observed. Several of these fragments result from simple disulfide linkage disruption. In addition, species less in mass than common antibody subunits were also observed, demonstrating peptide as well as disulfide bond cleavage. The observation that a small number of well-defined species were formed during the study suggests that the cleavage induced by thermal stress is very site-specific within the IgG.

No abstract

This paper reports the design of an on-line semi-preparative LC-SPE-NMR system and its use in the structural analysis of mixture components at the 0.02-1% level. The combination provides at least a five fold mass sensitivity increase over that obtained from typical analytical LC-SPE systems and a >30-fold total NMR sensitivity enhancement over analysis by LC-NMR. This is accomplished by using a novel on-line device to store, dilute (1-100-fold) and deliver (at an optimized flow-rate) the isolated component of interest to an SPE trap unit. The SPE unit consists of two cartridges connected in parallel to increase the overall SPE capacity and also to decrease the flow-rate through each trap for enhanced trapping efficiency. As the coupling of semi-preparative LC with NMR (through SPE) is well matched in terms of optimal mass loading for both techniques, only one LC-SPE cycle is required to enrich a 50 microg ml(-1) component (1% in a 5 mg ml(-1) mixture) for the acquisition of heteronuclear (1)H-(13)C NMR data using a conventional NMR flow probe. Furthermore, analytes at the 0.02% level (approximately 1 microg ml(-1)) can be studied using 2D (1)H NMR techniques if peak cuts from replicate sample injections (> or =3) are accumulated into the storage/dilution unit and the resulting solution processed by just one SPE trap and elute cycle.

No abstract

Collision-induced dissociation of peptide ions yields "sequence" ions arising from dissociations of the peptide backbone. Recently, Biemann and his collaborators have elucidated fragment ions involving cleavage of all or part of the side-chains, and have characterized them as remote-site fragmentations of the type investigated for other molecular species by Gross et ul. The present work reports results of experiments conducted using a tandem, hybrid mass spectrometer, and devoted to investigating whether remote-site fragmentations (including side-chain cleavages for peptide ions) can be observed for collision energies substantially lower than the keV range used previously. It was found that (laboratory-frame) collision energies of at least 200 eV, and preferably greater, are required for the formation of such fragments. At collision energies in this range the transmission efficiency of the qQ assembly is much lower than for the more usual range of a few tens of eV; this drop in transmission efficiency becomes increasingly severe with increasing mass of the precursor.Biemann and his colIaborators'-4 have elucidated collision-induced dissociation (CID) reactions of protonated peptide ions which involve losses of all or part of the side-chains of the amino acids. These fragmentations give rise to peaks in CID spectra which are not interpretable in terms of the established mechanisms5 which involve rupture of the peptide backbone and give rise to so-called "sequence-ions". Even more important than the increased confidence in sequence assignment thus aff~rdedl-~ by a more complete understanding of a given CID spectrum, is the side-chain specificity of these new reactions.Unlike the tandem double-focusing instrument6 used by Biemann et al. in the original in which the peptide ions are collisionally activated at energies in the keV range, laboratory collision energies (El&) in the RF-only quadrupole gas cell of a hybrid or of a triplequadrupole instrument are typically limited to a few tens of electron-volts. The limitations of this lowenergy collision process, in inducing even the peptidebackbone cleavages5 for larger peptides (molecular weights above loo0 Da or so), have been investigated recently (part 1 of this s e r i e~) .~ The disa pointingly energy deposition, insufficient to overcome the kinetic shifts appropriate to the timescale of the quadrupole reaction cell; these kinetic shifts increase with increasing molecular complexity (number of vibrational modes) and thus with the molecular weight of the peptide precursor ion. The absence of reports of these side-chain cleavages of peptide ions induced at low collision energies, noted previ~usly,~ is explicable in terms of this limited energy deposition' together with the characterization of these processes4 as special cases of the remote charge-site decompositions described by Gross and his collaborators.8-16 The energetics involved in this new class low collisional efficiencies were attributed ip to limited Author to whom correspondence should be addressed.of CID ...

Collisional activation (CA) of large ions at kiloelectronvolt energies is accompanied by unexpectedly large losses of translational energy, which vary with the nature of the collision gas. Previous investigations have concentrated upon subsequent fragmentations occurring within a time window covering a few fis immediately following collision, using massanalyzed ion kinetic energy spectrometry. In the present work, survivor ions were selected for specified values of translational energy loss, and their internal energy contents assessed via their subsequent unimolecular fragmentation reactions within a later time window. Beam collimation was also applied when circumstances permitted to impose angular selection, thus minimizing cross talk between effects of collisional scattering and energy dispersion. It was shown that internal excitation of the reactant ion can account for only a small fraction of the observed loss of translational energy. The recoil energy of the target is thus the principal sink for the translational energy loss, since the latter was always chosen to be less than the lowest excitation energy of the target. This conclusion is shown to be consistent with theoretical models of the CA process. The practical implications of these conclusions for CA of large ions at kiloelectronvolt energies are discussed.