Iodine-Assisted Laser Desorption/Ionization Time of Flight Mass Spectrometry

Sajid Bashir1,2, Sai Raghuveer Chava3,4,*, Jingbo L Liu2,4, and Peter J Derrick 1,5,#

Keywords: MALDI; α-cyano-4-hydroxycinnamic acid; Desorption; Ionization; Mass Spectrum

Introduction

Ultraviolet Matrix-assisted laser-desorption/ionization (UV MALDI summarized in Figure 1) has become an important analytical technique in the characterization of a range of compounds. Under MALDI conditions, a UV absorbing molecule (the ‘matrix’) is pre-mixed with the analyte under investigation. The binary mixture is deposited onto a stainless-steel slide and allowed to dry. Once dry, a UV laser (l=337nm) is fired, and ions are generated. On average, the penetration depth of a typical nitrogen laser is estimated to be about 85 nm, thus the excitation profile is essentially two-dimensional [1]. Many problems remain with MALDI. These can be grouped into problems with (a) poor desorption and (b) poor ionization collectively leading to low signal intensity for the analyte.

Results

The matrix alpha cyano-cinnamic acid (aCCa) to substance P without and without 100 parts-per-million mass percent iodine (dissolved in methanol) was shown under identical conditions, except laser power (LP) was adjusted to yield at least three times signal-to-noise ratio, to determine the limit of detection (LOD) is summarised in Figure 3. Each sample was deposited ten times and the signal standard deviation is also shown for each test-system. In all but two cases the signal intensity was superior for the system supplemented with iodine, and where the non-iodine system yield greater peptide intensities, the variation in the peptide signal was greater across the ten samples, although the average was greater. It appears that in some instances the ‘sweet spot’, an area of optimal matrix-to-analyte (around 1 or 10 nM analyte) were achieved, although with iodine more uniform results were obtained.

% (aCCaI +transferin digest)

mass-to-charge Figure 4A: Iodine assisted laser desorption/ionization time-of-flight mass spectrometry of transferrin digest. The insert is the low matrix mass region.

Human transferrin, a 698 amino acid iron carrier protein with a theoretical pI of 6.81, an average mass of 77063.89 and monoisotopic mass of 77013.63 Da was digested using trypsin at 37 oC with ammonium bicarbonate buffer and the hypothetical digests at 1500 Da summarised in Table 1 was likewise analyzed using iodine (Figure 4A) and standard aCCa matrix only (Figure 4B). All of the fragments listed in Table 1 were identified using aCCaI system but not with aCCa alone. In the aCCa, approximately 73% of the predicted peaks were observed. To system was deposited onto a copper grid (G) and the tryptic mixture was deposited, onto the grid. The grid was washed with alkali salts of sodium, potassium, and acetic acid before the addition of the digest and substance P (in trifluoroacetic acid, 0.01%) was added as an external calibration standard, the results are shown in Figure 4C.

aCCa + transferrin digest

12000 low-mass region

mass-to-charge Figure 4B: Matrix-assisted laser desorption/ionization time-of-flight mass spectrometry of transferrin digest. The insert is the low matrix mass region.

Click to enlarge

Figure 4C: Iodine-assisted laser desorption /ionization time-of-flight mass spectrometry of transferrin deposited into a metal copper grid (G) corresponding to Figure 2C and substance P as external calibrant.

One-hundred-fold increase in peak intensity was observed which dwarfed the protein digest mixture showing the protonated and alkali species for substance P much greater than when used without a grid system. No copper peaks were observed, and only 11% of the expected protein digest peaks were observed.

The proton affinity of alpha cyano cinnamic acid was calculated to 180 kcal/mol with an error of 3 kcal/mol and that of the amino acid glycine was 210 kcal/mol with an error of 3 kcal/mol. This suggests that proton transfer from the ground state of the matrix to the analyte can occur where the proton affinity of the acceptor is greater than the donor. Other peptides are likely to have proton affinities greater than 210 kcal/mol, suggesting that the excited state matrix species may be involved in addition to possible alkali metal attachment. Therefore, the parameter that may define ion yield is not the PA value but relative species that are capable of charge transfer and this tendency pay depend upon the analyte affinities, which may be related to isoelectric point (pI) values. To test this hypothesis, a less acidic protein was analyzed, sperm whale myoglobin, theoretical pI: 8.70, average mass 17331.09, and monoisotopic mass of 17320.25 was evaluated and shown in Figure 4D and Table 2. All except one fragment sequence was observed using aCCaI and nine sequences were obtained from aCCa alone (data not shown), suggesting that assisted iodine desorption is efficient with both acidic and basic protein fragments. To investigate further, an eleven amino acid wide mass band (WB) peptide mixture was analyzed using aCCa alone and all peaks were observed. The peak intensity for the protonated species was summed and normalized to 100%. The WB mix spectra are shown in Figure 4E. The same peptides were also detected using aCCaI, however, the summed peak intensities were 373% versus 100% suggesting that under near-identical conditions, greater peak intensities could be obtained using iodine supplementation.

% (acca+I2/Mb(dig))

mass-to-charge Figure 4D: Iodine-assisted laser desorption /ionization time-of-flight mass spectrometry of myoglobin digest, redeposited onto crushed crystals of aCCa, corresponding to Figure 2B.

The wideband peptide mixture consists of peptides that have a wide variation in theoretical isoelectric points, example includes the peptide bradykinin (10.3), bombesin (7.4), dermorphin (5.7), ELEDOS (where O=Pyrrolysine, 3.9), GGA (5.8), leucine-enkephalin (5.7), methionine-enkephalin (5.6), renin substrate (7.5), somatostatin 28 (10.3), substance P (11.4), and vasopressin (5.8) that span a range of pI from 3.9 (ELEDOS) to 11.4 (substance P).

% ([aCCa+WBMix])

m / z Figure 4E: Matrix-assisted laser desorption /ionization time-of-flight mass spectrometry of wide mass band (WB) peptide mixture deposited into crushed aCCa crystals corresponding to Figure 2B and summed to 100% peak intensities for the protonated species.

AH ANa MH MNa

LP % ([WBMix+H] +)

0 -10 GGA Lenk Menk Eledos DemorpDABradVDRVaspSub P Bomb Somat RSTD

acca+peptide Figure 4F: Matrix-assisted laser desorption /ionization time-of-flight mass spectrometry of wide mass band (WB) peptide mixture deposited into crushed aCCa crystals corresponding to Figure 2B as a function of laser power for the protonated species, avera.

The laser power for the protonated peaks, in general, was higher than the sodiated peaks and also were lower in general for acidic peptides relative to the entire spectra of peptides in the WB mix. The protonated/sodiated peaks for aCCa only or aCCa I (open circle, open square) analytes in protonated or sodiated form were also greater, with the population being split evenly between the two forms. With iodine alone, the protonated peak dominated relative to the sodiated peak (blue line versus blue line in graph 3F).

Conclusion

We have demonstrated that ion intensities could be enhanced by the synthesis of long-chain matric species (computational based on gas-phase proton affinities) or deposition of alpha hydroxycinnamic acid onto structured surfaces or deposition upon crystallized and crushed aCCa as a dual deposition system. Alternatively, by the addition of an ultra-low dose of molecular iodine, the ion intensities were always greater than under matrix-assisted laser desorption/ ionization-time of flight conditions, which is attributed to enhanced gas-phase desorption. As the radical species were not detected, only protonated or sodiated (or to a lesser degree potassiated) pseudo molecular ions, it appears the iodine promoted the desorption step to a greater degree than ionization, but also significant enhancement of the ionization step. Using the pseudo molecular ion peak intensities of the protonated and sodiated species for a wideband peptide mixture and comparison to aCCa only, it is suggested that 2/3 of the enhancement is due to desorption and 1/3 due to enhancement in the ionization step, where the molecular iodine is acting as a photosensitizer by transferring charge to proton and sodium ions rather than forming iodine or matrix adducts. Further analysis of the dependence of laser power suggests that ion transfer occurs between 5-7 eV. The data collectively indicate that iodine desorption involves bulk exciton decay at the surface and thermally assisted H+ diffusion that is trapped at the surface due to surface defects resulting in metastable decay. This would also account for the observation that aCCa deposited onto crushed aCCa crystals resulted in greater ion intensities.

Acknowledgment

The authors are grateful to the National Science Foundation (NSF) Centers of Research Excellence in Science and Technology (NSF CREST, HRD-0734850, and CBET MRI-0821370) at the Texas A&M University-Kingsville. The Robert C Welch (AC-0006) Departmental Grant is also duly acknowledged.

Author Contribution

S. Bashir conducted the research, wrote the first draft. S. R. Chava validated all the structures and J. Liu drew Scheme 1 and 2, assisted with the hypothetical synthesis scheme and statistical analysis on other figures 2B-4.

Tribute

† It is with profound sadness with which SB writes this acknowledgment. There is a word in the Old Testament, salah, meaning to pause, to reflect, and to think. I deeply miss Peter, but I will never forget the important lessons he taught me about being a scientist, teacher, and human being. JL Liu writes I never met Dr. Derrick, but his knowledge to advance science and responsibility to educate the young generation has been an inspiration in my academic career. We shall never forget the sweet, courageous, and dedicated man, whose name will always linger in our thoughts. Thank you, Dr. Peter Derrick, and God Speed. On February 15, 2020, I also lost my father due to SARS2, due to the lockdown I was unable to travel or attend the funeral or meet with my mother or brother and sisters, and six months later at the time of writing this tribute the situation has not changed. This article is in memory of all of us who have lost a loved one due to SAR2.