Central to the success of most any proteomics experiment is the choice of peptide fragmentation method. There is now a dizzying array of peptide fragmentation options available to proteomics researchers. I briefly outline a few of the more widespread options, discussing general mechanisms of fragmentation and strengths and weaknesses of each.
Collisional activation techniques are by far the most pervasive peptide fragmentation techniques. This is largely due to the (relative) ease of implementation of such techniques on nearly every classification of mass spectrometer available commercially.For nearly a decade after the development of soft ionization techniques (ESI, MALDI) made possible large-scale, tandem-in-time based proteomics experiments, resonant-excitation collision-activated dissociation (CAD/CID) dominated the proteomics landscape. During this process, a supplemental frequency corresponding to the fundamental frequency of the selected peptide precursor is applied between electrodes of the ion trap. This in turn induces energetic collisions with the bas gas and a competitive process, resonant ejection of the ions from the trap; the relative partitioning is dictated by the reduced Mathieu parameter (q). Each collision imparts a small amount of energy to the peptide precursor, which is re-distributed throughout the covalent bonds of the peptide. The peptide internal energy continues to increase until the energy of activation for the weakest covalent (or non-covalent) bond is reached. This is most typically the peptide bond, the cleavage of which generates b- and y-type fragment ions. Strengths of this method include the highest activation efficiency of any of the methods to be discussed (largely because as a resonant-excitation method, product ions are not susceptible to secondary activation) and ease of implementation (available commercially on virtually every ion trapping system). Weaknesses include an incompatibility with isobaric tag-based quantitation (because the low mass cutoff is typically too high to retain reporter ions), limited effectiveness for peptides modified with a labile moiety (e.g. phosphorylation, glycosylation), and limited effectiveness for peptides without a mobile proton (mobile proton theory).
In addition to resonant-excitation collisional activation, so-called ‘beam-type’ activation is commonly employed in proteomics experiments. During this process, ions are shuttled from one region of a mass spectrometer to another. Provided that precursor ions possess sufficient translational kinetic energy, collisions with neutral gas species again deposit energy to the peptide. This type of fragmentation is used on quadrupole-time-of-flight, triple quadrupole, and recently ion trapping instrumentation. In theory, this is a very similar process to CAD (vide supra); in practice, however, the resulting spectra are quite different. This is because each singular collision with a neutral gas molecule is of a much higher energy than the CAD process, resulting in much larger jumps in internal energy for peptides. What this means is that the activation barriers of several covalent bonds are simultaneously exceeded, and any of them may fragment. Ultimately, this is responsible for the main strengths of HCD, namely improved performance for non-tryptic peptides and peptides modified with labile PTMs alike. Another benefit of HCD is compatibility with isobaric tagging techniques. This is because HCD, unlike CAD, is readily performed at low RF amplitudes, which in turn facilitate the retention of low mass reporter ions used for quantitation.Drawbacks of this technique include the requirement for an external RF device to serve as the collision cell; it should be noted, however, that recently it was reported that ESI ion injection optics of stand-alone ion traps can serve this function effectively (McAlister, MCP 2011).
In 1998, Roman Zubarev described a process he termed ‘Electron Capture Dissociation’, ECD. During ECD, peptides are immersed in free electrons at near thermal energy; subsequent capture of an electron by a peptide converts the peptide to a radical species. A radical-induced re-arrangement ultimately fragments the N–Cα bond of the peptide, creating c- and z·– type ions. ECD has several strengths; these include a propensity to cleave inter-residue bonds more or less randomly (although cleavage n-terminal to proline residues is never observed) and virtual indifference to labile post-translational modifications (e.g. phosphorylation, glycosylation). The primary drawback of ECD is that for efficient fragmentation, peptides must be immersed within a dense population of free electrons, a pre-condition not easily met within common, less expensive RF ion traps. While some groups have made progress on ways to implement ECD within RF devices with encouraging results, such implementations remain non-trivial, largely limiting ECD to expensive FT-ICR mass spectrometers.
In 2004, John Syka and Joshua Coon pioneered a way to effect ECD-like fragmentation on RF ion trapping devices using radical reagent anions in lieu of free electrons, termed ‘Electron Transfer Dissociation’, ETD. During ETD, provided that a suitable reagent anion is selected (relatively low electron affinity), the anion transfers an electron to the cation peptide. The mechanism is thought to be virtually identical to ECD, with the caveat that ETD is slightly less exothermic because the electron affinity of the anion must be overcome. ETD is likewise well-suited to the analysis of labile PTMs, particularly phosphorylation and glycosylation. Additionally, ETD is very effective for analysis of peptides possessing high charge density (~ low m/z values). Drawbacks of ETD include diminishing fragmentation efficiency with increasing precursor m/z value due to intra-molecular non-covalent interactions which impair the separation of c- and z·– type ions following electron transfer, a problem that is particularly poignant for doubly protonated precursors. ETD is also not well-suited for use with isobaric tagging techniques (although work from Coon group and McLuckey group showed that it is not altogether impossible).
Photodissociation is a process during which peptide precursors are bombarded with photons of a selected wavelength, most commonly infrared multi-photon dissociation (IRMPD, λ = 10.6 um) and ultraviolet photodissociation (UVPD, λ = 10–400 nm). Peptides subsequently accumulate internal energy until the energy of activation is achieved, leading to dissociation. The process of energy accumulation during photodissociation is distinct from collisional methods in that each time a photon is captured, a discrete amount of energy is transferred to the peptide precursors. A general advantage of photodissociation techniques is that the photon flux, total irradiation period, and photon wavelength can be varied, allowing a high degree of tunability. Also, photo-dissociation can allow for the selective activation of molecules depending upon the presence or absence of chromophoric moieties (e.g. phosphopeptides or disulphide bonds).
A common photodissociation technique is infrared multiphoton dissociation (IRMPD). This technique utilizes IR photons (λ = 10.6 um) which are relatively low energy (~0.12 eV/photon); dissociation of peptides typically occur only after the absorption of several photons, making IRMPD somewhat similar to CAD in that both are ‘slow heating’ techniques. As peptides gain energy via IRMPD, the peptide bond is cleaved, generating primarily b- and y- type product ions. An advantage IRMPD possesses over CAD is that efficient activation is possible at relatively low RF amplitudes, allowing for the retention of lowm/z ions. A disadvantage is that IRMPD is a non-resonant activation method; as such the primary product ions are themselves susceptible to photo-activation, resulting in lower efficiency of activation than CAD. For large-scale, discovery-based proteomics the possibility of ‘over-reacting’ peptide precursors further necessitates more careful consideration of activation conditions. It is, however, important to note that for some applications non-resonance is an advantage, as it allows for potentially richer spectra containing diagnostic side chain losses. A further disadvantage for potential photo-dissociation users is that almost all photodissociation capable mass spectrometers have been, to various degrees, modified in-house. Currently, the only commercially available instruments equipped with photodissociation capabilities are FT-ICR mass spectrometers (IRMPD).
Text contributed by Aaron Ledvina