Fragmentation of leucine enkephalin as a function of laser fluence in a MALDI TOF-TOF

Detailed Characterization of the Postionization Efficiencies in MALDI-2 as a Function of Relevant Input Parameters

A recently introduced technique based on MALDI with laser-induced postionization (PI), also named MALDI-2, increases the ion yields for numerous classes of lipids, metabolites, and carbohydrates in MALDI-MS imaging experiments under certain experimental conditions. Here, we used a semiautomatic LabVIEW-based protocol to investigate and optimize the efficiency of the PI process dependent on four relevant input parameters and a dense parameter grid: pulse energies of the two lasers, delay between the laser pulses, and buffer gas pressure in the ion source. All experiments were conducted with a modified MALDI-2 Synapt G2-S mass spectrometer (Waters) and use of a focal spot size on the sample of 15-17 μm. A wavelength-tunable optical parametric oscillator (OPO) laser served for PI at 260 or 280 nm. The investigated MALDI matrices were: 2,5-dihydroxybenzoic acid (positive ion mode, +), 2,5-dihydroxyacetophenone (+), α-cyano-4-hydroxycinnamic acid (+), norharmane (negative-ion mode, -), and 1,5-diaminonapthalene (-). A porcine brain extract served as lipid standard. In the positive-ion mode, a maximum boost for the generated [M + H]⁺ species was found with a N₂ buffer gas pressure of ∼2 mbar and a delay between the laser emissions of ∼10 μs. Higher optimal delay settings of several 10 μs were registered for the two studied matrices in negative-ion mode. With regard to the laser fluences, best PI efficiencies were reached using maximum available ablation and PI laser pulse energies of up to 25 and 160 μJ, respectively. For analytes not profiting from MALDI-2, best ion signal yields were recorded for ablation laser pulse energies of around 7 μJ, depending on the MALDI matrix. At higher laser pulse energies, sizable fragmentation is observed for these ions. The PI laser pulse energy did not have any influence on the ion signals of these species. For optimal ion yield of all analyte species, best results were obtained with an ablation laser pulse energy of ∼7 μJ and a PI laser pulse energy of ∼160 μJ. Our comprehensive data set provides valuable insight into the mechanisms underlying the MALDI-2 processes and could help to further optimize this emerging technique.

Formation of gas-phase peptide ions and their dissociation in MALDI: insights from kinetic and ion yield studies

Insights on mechanisms for the generation of gas-phase peptide ions and their dissociation in matrix-assisted laser desorption ionization (MALDI) gained from the kinetic and ion yield studies are presented. Even though the time-resolved photodissociation technique was initially used to determine the dissociation kinetics of peptide ions and their effective temperature, it was replaced by a simpler method utilizing dissociation yields from in-source decay (ISD) and post-source decay (PSD). The ion yields for a matrix and a peptide were measured by repeatedly irradiating a region on a sample and collecting ion signals until the sample in the region was completely depleted. Matrix- and peptide-derived gas-phase cations were found to be generated by pre-formed ion emission or by ion-pair emission followed by anion loss, but not by laser-induced ionization. The total number of ions, that is, matrix plus peptide, was found to be equal to the number of ions emitted from a pure matrix. A matrix plume was found to cool as it expanded, from around 800-1,000 K to 400-500 K. Dissociation of peptide ions along b/y channels was found to occur statistically, that is, following RRKM behavior. Small critical energy (E0  = 0.6-0.7 eV) and highly negative critical entropy (ΔS(‡)  = -30 to -25 eu) suggested that the transition structure was stabilized by multiple intramolecular interactions.

A Thermal Mechanism of Ion Formation in MALDI

An important recent discovery concerning the fundamentals of matrix-assisted laser desorption/ionization (MALDI) is that the abundance of each ion appearing in a spectrum is fixed, regardless of the experimental condition, when an effective temperature associated with the spectrum is fixed. We describe this phenomenon and the thermal picture for the ion formation in MALDI derived from it. Accepting that matrix-to-analyte proton transfer is in quasi-equilibrium as supported by experimental data, the above thermal determination occurs because the primary (matrix) ion formation processes are thermally governed. We propose that the abundances of the primary ions are limited by the autoprotolysis-recombination process regardless of how they are initially produced. Finally, we note that primary ion formation, secondary (analyte) ion formation, and their dissociations occur sequentially while the effective temperature of the matrix plume falls steadily due to cooling associated with expansion.

Suppression of unimolecular decay of laser desorbed peptide and protein ions by entrainment in rarefied supersonic gas jets under weak electric fields

Unimolecular decay of sample ions imposes a limit on the usable laser fluence in matrix-assisted laser desorption/ionization (MALDI) ion sources. Traditionally, some modest degree of collisional sample ion cooling has been achieved by connecting MALDI ion sources directly to gas-filled radio frequency (RF) multipoles. It was also discovered in the early 1990s that gas-filled RF multipoles exhibit increased ion transmission efficiency due to collisional ion focusing effects. This unexpected experimental finding was later supported by elementary Monte Carlo simulations. Both experiments and simulations assumed a resting background gas with typical pressures of the order of 1 Pa. However, considerable additional improvements can be achieved if laser desorbed sample ions are introduced immediately after desorption, still within the ion source, in an axisymmetric rarefied supersonic gas jet with peak pressure of the order of 100 Pa and flow velocities >300 m/s, and under weak electric fields. We describe here the design principle and report performance data of an ion source coined "MALDI-2," which incorporates elements of both rarefied aerodynamics and particle optics. Such a design allows superb suppression of metastable fragmentation due to rapid collisional cooling in <10 μs and nearly perfect injection efficiency into the attached RF ion guide, as numerous experiments have confirmed.

Degree of ionization in MALDI of peptides: thermal explanation for the gas-phase ion formation

Degree of ionization (DI) in matrix-assisted laser desorption ionization (MALDI) was measured for five peptides using α-cyano-4-hydroxycinnanmic acid (CHCA) as the matrix. DIs were low 10(-4) for peptides and 10(-7) for CHCA. Total number of ions (i.e., peptide plus matrix) was the same regardless of peptides and their concentration, setting the number of gas-phase ions generated from a pure matrix as the upper limit to that of peptide ions. Positively charged cluster ions were too weak to support the ion formation via such ions. The total number of gas-phase ions generated by MALDI, and that from pure CHCA, was unaffected by the laser pulse energy, invalidating laser-induced ionization of matrix molecules as the mechanism for the primary ion formation. Instead, the excitation of matrix by laser is simply a way of supplying thermal energy to the sample. Accepting strong Coulomb attraction felt by cations in a solid sample, we propose three hypotheses for gas-phase peptide ion formation. In Hypothesis 1, they originate from the dielectrically screened peptide ions in the sample. In Hypothesis 2, the preformed peptide ions are released as part of neutral ion pairs, which generate gas-phase peptide ions via reaction with matrix-derived cations. In Hypothesis 3, neutral peptides released by ablation get protonated via reaction with matrix-derived cations.

Occurrence of C-terminal residue exclusion in peptide fragmentation by ESI and MALDI tandem mass spectrometry

By screening a data set of 392 synthetic peptides MS/MS spectra, we found that a known C-terminal rearrangement was unexpectedly frequently occurring from monoprotonated molecular ions in both ESI and MALDI tandem mass spectrometry upon low and high energy collision activated dissociations with QqTOF and TOF/TOF mass analyzer configuration, respectively. Any residue localized at the C-terminal carboxylic acid end, even a basic one, was lost, provided that a basic amino acid such arginine and to a lesser extent histidine and lysine was present in the sequence leading to a fragment ion, usually depicted as (b(n-1) + H(2)O) ion, corresponding to a shortened non-scrambled peptide chain. Far from being an epiphenomenon, such a residue exclusion from the peptide chain C-terminal extremity gave a fragment ion that was the base peak of the MS/MS spectrum in certain cases. Within the frame of the mobile proton model, the ionizing proton being sequestered onto the basic amino acid side chain, it is known that the charge directed fragmentation mechanism involved the C-terminal carboxylic acid function forming an anhydride intermediate structure. The same mechanism was also demonstrated from cationized peptides. To confirm such assessment, we have prepared some of the peptides that displayed such C-terminal residue exclusion as a C-terminal backbone amide. As expected in this peptide amide series, the production of truncated chains was completely suppressed. Besides, multiply charged molecular ions of all peptides recorded in ESI mass spectrometry did not undergo such fragmentation validating that any mobile ionizing proton will prevent such a competitive C-terminal backbone rearrangement. Among all well-known nondirect sequence fragment ions issued from non specific loss of neutral molecules (mainly H(2)O and NH(3)) and multiple backbone amide ruptures (b-type internal ions), the described C-terminal residue exclusion is highly identifiable giving raise to a single fragment ion in the high mass range of the MS/MS spectra. The mass difference between this signal and the protonated molecular ion corresponds to the mass of the C-terminal residue. It allowed a straightforward identification of the amino acid positioned at this extremity. It must be emphasized that a neutral residue loss can be misattributed to the formation of a y(m-1) ion, i.e., to the loss of the N-terminal residue following the a(1)-y(m-1) fragmentation channel. Extreme caution must be adopted when reading the direct sequence ion on the positive ion MS/MS spectra of singly charged peptides not to mix up the attribution of the N- and C-terminal amino acids. Although such peculiar fragmentation behavior is of obvious interest for de novo peptide sequencing, it can also be exploited in proteomics, especially for studies involving digestion protocols carried out with proteolytic enzymes other than trypsin (Lys-N, Glu-C, and Asp-N) that produce arginine-containing peptides.

Expansion cooling in the matrix plume is under-recognized in MALDI mass spectrometry

Time-of-flight (TOF) mass spectra for a peptide (Y(6)) were obtained by utilizing matrix-assisted infrared laser desorption ionization (IR-MALDI) with glycerol as the matrix and by ultraviolet MALDI with α-cyano-4-hydroxycinnamic acid (CHCA), sinapinic acid (SA), and 2,5-dihydroxybenzoic acid (DHB). Collisional activation during ion extraction and exothermicity in the gas-phase proton transfer were found to be unimportant as the driving forces for in-source (ISD) and post-source (PSD) decays, indicating that the thermal energy acquired during photo-ablation is responsible for their occurrence. The temperatures of [Y(6) + H](+) in the 'early' and 'late' matrix plumes were estimated by the kinetic analysis of the ISD and PSD yields, respectively. The order of the temperatures was glycerol < DHB ≈ SA < CHCA in the early plume and glycerol < DHB < SA < CHCA in the late plume. For each matrix, the temperature in the late plume was lower than in the early plume by 300-400 K, which was attributed to expansion cooling. The model (thermalization followed by expansion cooling) proposed to explain the occurrence of both rapid ISD and slow PSD is not only in sharp contrast with but also mutually exclusive with the prevailing explanation that the exothermicity in proton transfer and in-plume collisional activation are the driving forces for ion fragmentation in MALDI. The model also explains why MALDI is more successful for mass spectrometry of labile molecules than other desorption techniques that do not utilize a matrix. Factors affecting the plume temperature are also discussed.

Leucine enkephalin--a mass spectrometry standard

The present article reviews the mass spectrometric fragmentation processes and fragmentation energetics of leucine enkephalin, a commonly used peptide, which has been studied in detail and has often been used as a standard or reference compound to test novel instrumentation, new methodologies, or to tune instruments. The main purpose of the article is to facilitate its use as a reference material; therefore, all available mass spectrometry-related information on leucine enkephalin has been critically reviewed and summarized. The fragmentation mechanism of leucine enkephalin is typical for a small peptide; but is understood far better than that of most other compounds. Because ion ratios in the MS/MS spectra indicate the degree of excitation, leucine enkephalin is often used as a thermometer molecule in electrospray or matrix-assisted laser desorption ionization (ESI or MALDI). Other parameters described for leucine enkephalin include collisional cross-section and energy transfer; proton affinity and gas-phase basicity; radiative cooling rate; and vibrational frequencies. The lowest-energy fragmentation channel of leucine enkephalin is the MH(+)  → b(4) process. All available data for this process have been re-evaluated. It was found that, although the published E(a) values were significantly different, the corresponding Gibbs free energy change showed good agreement (1.32 ± 0.07 eV) in various studies. Temperature- and energy-dependent rate constants were re-evaluated with an Arrhenius plot. The plot showed good linear correlation among all data (R(2)  = 0.97), spanned over a 9 orders of magnitude range in the rate constants and yielded 1.14 eV activation energy and 10(11.0)  sec(-1) pre-exponential factor. Accuracy (including random and systematic errors, with a 95% confidence interval) is ±0.05 eV and 10(±0.5)  sec(-1), respectively. The activation entropy at 470 K that corresponds to this reaction is -38.1 ± 9.6 J mol(-1)  K(-1). We believe that these re-evaluated values are by far the most accurate activation parameters available at present for a protonated peptide and can be considered as "consensus" values; results on other processes might be compared to this reference value.

Changes in dissociation efficiency and kinetics of peptide ions induced by basic residues and their mechanistic implication

With matrix-assisted laser desorption ionization (MALDI) time-of-flight (TOF) mass spectrometry, total abundance of product ions formed by dissociation inside (in-source decay, ISD) and outside (post-source decay, PSD) the source was measured for peptide ions [Y(5)X + H](+), [XY(5) + H](+), [Y 2)XY(3) + H](+), and [XY(4)X + H](+) (X = tyrosine (Y), histidine (H), lysine (K), and arginine (R) with H for the ionizing proton). α-Cyano-4-hydroxycinammic acid was used as matrix. Product abundance became smaller in the presence of basic residues (H, K, and R), in the order Y > H ≈ K > R. In particular, product abundances in ISD of peptide ions with R were smaller than those with H or K by an order of magnitude, which, in turn, were smaller than that for [Y(6) + H](+) by an order of magnitude. Product abundance was affected by the most basic residue when more than one basic residue was present. A kinetic explanation for the data was attempted under the assumption of quasi-thermal equilibrium for peptide ions in MALDI plume which undergoes expansion cooling. Dramatic disparity in product abundance was found to arise from small difference in critical energy and entropy. Results indicate similar transition structures regardless of basic residues present, where the ionizing proton keeps interacting with a basic site. Further implication of the results on the dissociation mechanism along b-y channels is discussed.

A comparative study of in- and post-source decays of peptide and preformed ions in matrix-assisted laser desorption ionization time-of-flight mass spectrometry: effective temperature and matrix effect

In-source decay (ISD) and post-source decay (PSD) of a peptide ion ([Y(6) + H](+)) and a preformed ion (benzyltriphenylphosphonium, BTPP) generated by matrix-assisted laser desorption ionization (MALDI) were investigated with time-of-flight mass spectrometry. α-Cyano-4-hydroxycinammic acid (CHCA) and 2,5-dihydroxybenzoic acid (DHB) were used as matrices. For both ions, ISD yield was unaffected by delay time, indicating rapid termination of ISD. This was taken as evidence for rapid expansion cooling of hot "early" plume formed in MALDI. CHCA was hotter than DHB for [Y(6) + H](+) while the matrix effect was insignificant for BTPP. The "early" plume temperature estimated utilizing previous kinetic results was 800-900 K, versus 400-500 K for "late" plume. The results support our previous finding that the temperature of peptide ions interrogated by tandem mass spectrometry was lower than most rough estimates of MALDI temperature.

Comparison of LID versus CID activation modes in tandem mass spectrometry of peptides

We report our contribution to the systematic investigation of peptide fragmentations performed on high-performance Tof equipment, operating in MS and MS/MS modes, such as ESI-QqTof and MALDI-Tof/Tof instruments that are commonly available today in proteomic laboratories. Whereas the former analyzer's configuration provides low-energy collision-induced dissociations (CID), the latter allows tunable activation methods of the selected parent ion to induce either metastable laser-induced dissociations (LID) or high-energy CID ('gas on spectra LID'). Fragmentation of the monoprotonated ion of 53 peptides (FW 807-2853 g/mol) was undertaken upon low-energy CID on an ESI-QTof mass spectrometer (Waters) as well as high-energy CID and LID conditions on a MALDI Ultraflex mass spectrometer (Bruker). Systematic comparison of MS/MS spectra provided useful information on the performance of each piece of equipment for efficient peptide sequencing and also insights into the observed fragmentation behaviors.

Resonance excitation and dynamic collision-induced dissociation in quadrupole ion traps using higher-order excitation frequencies

Fragmentation of the pentapeptide leucine enkephalin (YGGFL) is accomplished via higher-order resonances combined with simultaneous analysis of low-mass product ions. Two methods of achieving excitation are explored: (1) 0.5 ms resonant excitation at the omega and at Omega-omega secular frequencies of ion motion (where Omega is the radio-frequency (rf) drive frequency) in a manner similar to both pulsed q collision-induced dissociation (PQD) and high amplitude short time excitation (HASTE), and (2) 0.5 ms pulse of the omega or at Omega-omega excitation frequencies when the secular frequency of the ions is quickly swept across resonance conditions (pulsed q dynamic CID, PqDCID). In both methods of excitation, the rf amplitude on the ring electrode is rapidly decreased after excitation, therefore enabling analysis of low-mass product ions. Maximum fragmentation efficiencies of approximately 20% can be obtained with pulsed CID with both regular and high-order frequency excitation, while pulsed DCID offers maximum efficiencies of approximately 12%. All the excitation methods studied offer increased internal energy depositions when compared to conventional CID, as measured by the a4/b4 product ion ratios of leucine enkephalin. These ratios were as high as 13:1 for pulsed CID and 8:1 for PqDCID. Successful mass analysis of the low-mass ions is observed with both pulsed CID and PqDCID. The combined benefit of high internal energy deposition and wider dynamic mass range offers the possibility of increased sequence coverage and the identification of unique internal fragments or high-energy product ions which may provide complementary information to biological applications of conventional CID. This is the first report on deliberate fragmentation of precursor ions at a higher-order component of the ion secular frequency combined with a successful mass analysis of the low-mass ions through pulsed CID and PqDCID.