Mass spectrometers, instruments that use electric and/or magnetic fields to measure a gas-phase ion's mass-to-charge ratio (m/z), are used in a wide variety of applications—with the field having a reputation for providing good sensitivity and high-informing power. Protein analysis (proteomics) is a relatively recent affair for the field and was enabled in the late 1980s with the advent of biomolecule ionization methods such as electrospray ionization (ESI) and matrix-assisted laser desorption ionization (MALDI). Today, the area of protein analysis garners considerable attention from many in the mass spectrometry (MS) field; given the myriad of possible protein forms and their broad dynamic range (abundance) in the cell, the analytical challenge is paramount. Here we discuss a developing technology—ion/ion chemical reactions—that promises to transform how we think about and conduct protein sequence analysis via MS.

Mass Spectrometry-Based Proteomics Overview

Perhaps the most common and effective mass spectrometry-based protein sequencing methods have been those based on the bottom-up strategy (1,2,3). Here a protein or an entire proteome is enzymatically digested with trypsin (approximately 40 peptides/protein), chromatographically separated, and interrogated with tandem mass spectrometry [MS/MS; the process of ion fragmentation with subsequent mass-to-charge ratio (m/z) measurement]. Electrospray ionization (ESI) is the typical interface and is used to convert condensed phase ions eluting from the high-performance liquid chromatography (HPLC) column to multiply protonated molecules in the gas phase. The mass spectrometer first records the m/z of each peptide ion and then selects the ions individually to obtain sequence information via MS/MS. Collision of the peptide cation with rare gas atoms is the most widely used method to induce peptide fragmentation (collisional-activated dissociation; CAD). If the deposited energy is distributed randomly, a homologous series of backbone fragment m/z peaks is produced. Peptide identification, however, can only be achieved if a complete or nearly complete distribution of backbone cleavages is present. CAD often fails in this regard when the peptide contains: (i) certain amino acids, particularly those that inhibit random protonation along the peptide backbone; (ii) a posttranslational modification (PTM) that dissociates by a lower energy pathway (than that involved in cleavage of the amide linkage); or (iii) more than approximately 15 amino acids (i.e., large peptides) (4). Note that CAD has been used to characterize peptides with PTMs and even whole proteins; however, the amount of information obtained from CAD fragmentation, and the probability of success, falls off rapidly as a function of these parameters.

In a very real sense, these limitations have narrowly defined the field of MS-based protein measurement [i.e., mandating the use of trypsin to render a collection of peptides of suitable size and sequence (no internal basic residues) for successful MS/MS]. This requirement poses three significant limitations. First, making an already complex mixture even more complex is a considerable disadvantage. Second, the view that one gene equals one protein is simply not accurate for the study of higher eukaryotes; the majority (up to three-fourths) of human proteins are alternatively spliced, hence the detection of tryptic peptides will rarely reveal this diversity (5,6,7). This is mainly because a detected peptide could come from any form of the protein, spliced or nonspliced. And third, after transcription, messenger RNA (mRNA) editing, and translation, a further level of complexity is added in the form of protein PTM.

Provided that bottom-up analysis does actually generate a suitable MS/ MS spectrum for the identification of a PTM-containing peptide, assigning biological meaning and relevance can be difficult. First, PTMs on multi-domain proteins and among components of protein-protein machines work in concert; to determine their biological relevance, these patterns must be detected within the context of one another (across the whole protein). Second, analysis of short peptide sequences is unlikely to reveal isoform identity (8). Different variants perform different functions and will be modified in distinct ways. New technologies that are capable of sequencing large peptides and whole proteins will be required to acknowledge this diversity in addition to these important, yet unmonitored (at the protein level), biological events.

Large-Molecule MS

For the reasons presented above, large peptide (i.e., 3–10 kDa) and whole protein sequencing recently has become an area of interest in the field of proteomics. Kelleher, McLafferty, and others have pursued direct dissociation of intact proteins either by CAD or electron capture dissociation (ECD) (9). Most often this experiment utilizes the high resolving power of a Fourier transform-ion cyclotron resonance-MS (FTICR-MS) but not always. Whatever the case, the intact protein mass is measured while the product ions (MS/ MS) are used for sequencing and locating sites of modification. When the process is implemented successfully, the entire protein is characterized—a feat rarely achieved with the bottom-up strategy. Unfortunately, mainstream implementation of this top-down type analysis has been challenging for the following reasons: (i) ESI-charge envelopes reduce sensitivity; (ii) precursor m/z peaks can overlap and result in mixed MS/MS spectra (in complex mixtures); (iii) tandem mass spectral complexity can be high, even for the highest resolving power instrument; (iv) CAD is less effective with increasing precursor mass; and (v) coupling ECD with chromatography has been challenging. Doubtless, the ability to routinely characterize large molecules—either whole proteins or large peptides—via MS would significantly increase the informing power of a proteomics experiment (vide supra). Evolution in this field, however, will require the development of new tools and technologies. Here, we discuss a developing area—ion/ion chemical reactions, that, when compiled into sequences, show great promise to propel evolution in the emergent field of large-molecule MS.