Does a charge tag really provide a fixed charge?

In biomolecule mass spectrometry, peptides are usually ionized by an acidic solvent or matrix. It is generally accepted that the mobilization of attached protons weakens backbone bonds and facilitates peptide fragmentation. While this is often advantageous, it can lead to a multiplicity of fragment ions (both N-terminal an, bn and C-terminal yn ions). Formation of too many types of fragment ions can overly complicate spectra. Although very basic arginines tend to sequester protons and to some extent simplify spectra, charge tags, small molecules containing a permanent charge (for example, a quaternary ammonium ion), have been proposed as an even better solution to the problem. Charge-tagged peptides have a localized positive charge that is assumed to be absolutely fixed. Therefore, in MS/MS experiments, only Nterminal or C-terminal fragment ions should be formed through charge-remote mechanisms depending on the position of the tag. Such a simplified series of contiguous fragment ions would provide complete sequencing information and would be ideal for spectral interpretation. Sensitivity should also be greatly improved after derivatization, since charge-tagged peptides are already ionized and need not be protonated. However, the lack of a mobile proton may reduce fragmentation efficiency which would be a potential disadvantage. Most of the charge tags that have been developed label the N-terminus. Two popular examples are trimethylammonium butyric acid (TMAB) and tris(2,4,6-trimethoxyphenyl) phosphonium acetic acid (TMPP-Ac). Herein, we introduce a (4-trimethylammoniumbutyryl (TMAB)) positive charge onto the N-terminus of several peptides. It has been reported that this modification improves peptide detection sensitivity, and leads to the production of Nterminal fragment ions in MALDI-PSD mass spectra. However, we now demonstrate that the positive charge does not always remain fixed at the N-terminus of the peptide. Figure 1a is the ESI mass spectrum of charge-tagged Fibrinopeptide A. The presence of only a single peak associated with the doubly charged, derivatized peptide (from the charge tag and one proton) indicates that the reaction between the labeling reagent (TMAB-NHS; NHS= N-hydroxysuccinimide) and peptides was complete. Singly-charged ions associated with Fibrinopeptide A (ADSGEGDFLAEGGGVR) and its derivative were isolated and fragmented in a MALDI TOF/TOF apparatus. Results are shown in Figure 1b and 1c. As anticipated, yn ions are the dominant features in Figure 1b because of the presence of C-terminal arginine residue. We had expected to observe N-terminal fragment ions from the TMABderivatized peptide, but the spectrum is surprisingly more complex (Figure 1c). In this case, Bn represents a TMABlabeled b ion (Bn= bn+ 127). An* and Bn* represent An and Bn ions that have lost trimethylamine (An*=An 59 and Bn*=Bn 59). Hines et al. as well as Che and Fricker. have observed the loss of trimethylamine from this charge tag. Remarkably, although the charge is supposed to be Nterminal, signals from immonium ions, y ions and even some ions 14 Da heavier than regular y ions (y+ 14 ions) were still observed in the spectrum. These ions must be protonated, despite there being no mobile proton in the charge-tagged peptide. Observation of y+ 14 ions is particularly remarkable. We observed similar results with derivatized peptides GSFGSAIR and FVDGSIR, indicating that these chargetagged peptides routinely undergo unexpected complex fragmentation processes. Figure 1. a) ESI-MS spectrum of doubly charged labeled Fibrinopeptide A (TMAB-ADSGEGDFLAEGGGVR), m/z 833.09 Da; MALDI-TOF/ TOF CID (1 kV) and spectra of singly charged b) Fibrinopeptide A and c) the TMAB derivative. See text for details.