Cross-Validation of Lipid Structure Assignment Using Orthogonal Ion Activation Modalities on the Same Mass Spectrometer.

The onset and progression of cancer is associated with changes in the composition of the lipidome. Therefore, better understanding of the molecular mechanisms of these disease states requires detailed structural characterization of the individual lipids within the complex cellular milieu. Recently, changes in the unsaturation profile of membrane lipids have been observed in cancer cells and tissues, but assigning the position(s) of carbon-carbon double bonds in fatty acyl chains carried by membrane phospholipids, including the resolution of lipid regioisomers, has proven analytically challenging. Conventional tandem mass spectrometry approaches based on collision-induced dissociation of ionized glycerophospholipids do not yield spectra that are indicative of the location(s) of carbon-carbon double bonds. Ozone-induced dissociation (OzID) and ultraviolet photodissociation (UVPD) have emerged as alternative ion activation modalities wherein diagnostic product ions can enable de novo assignment of position(s) of unsaturation based on predictable fragmentation behaviors. Here, for the first time, OzID and UVPD (193 nm) mass spectra are acquired on the same mass spectrometer to evaluate the relative performance of the two modalities for lipid identification and to interrogate the respective fragmentation pathways under comparable conditions. Based on investigations of lipid standards, fragmentation rules for each technique are expanded to increase confidence in structural assignments and exclude potential false positives. Parallel application of both methods to unsaturated phosphatidylcholines extracted from isogenic colorectal cancer cell lines provides high confidence in the assignment of multiple double bond isomers in these samples and cross-validates relative changes in isomer abundance.

Electronic and Steric Effects on the Reactivity of Seleniranium Ions with Alkenes in the Gas Phase.

Gas phase ion-molecule reactions between seleniranium ions, R-c-SeCH2CH2+, and cis-cyclooctene were used to probe electronic and steric effects of substituents on kinetics and branching ratios. The second-order rate coefficients increased in the order p-OMeC6H4 < C6H5 < p-BrC6H4 < p-CF3C6H4 < m-NO2C6H4, giving a Hammett plot with R2 = 0.98 and ρ = +1.66. The two main pathways include direct transfer of the selenium moiety to the incoming alkene (π-ligand exchange) and the less favored ring-opening by attack at an iranium carbon to give a cis-bicyclic selenonium ion as supported by density functional theory (DFT) calculations. Branching ratios of each pathway indicated that electron-withdrawing groups directed more attack at carbon than selenium in agreement with previous solution-phase results. Increased steric bulk on selenium was investigated by changing the R group from a methyl to t-butyl, which not only shut down π-ligand exchange but also significantly reduced the overall reactivity. Finally, the reactivity of the iranium ion derived from Se-methylselenocysteine was investigated and shown to react faster and favor π-ligand exchange as the leaving group was changed from ethene to acrylic acid.

Experimental and theoretical investigations into the mechanisms of haliranium ion π-ligand exchange reactions with cyclic alkenes in the gas phase.

Haliranium ions are intermediates often involved in complex cyclisations, where their structure allows for control over stereospecific outcomes. Extending previous studies into their structure and reactivity in the gas phase, this work focuses on the bimolecular reactivity of ethyl bromiranium and iodiranium ions with cyclic alkenes. The products observed via mass spectrometry were broadly attributed to either addition by cyclohexene at the iranium carbon or attack at the heteroatom to undergo associative π-ligand exchange. The model proposed was supported by both kinetic experiments and DFT calculations, where the rate of parent ion consumption proceeded at the collision rate (Br: k2 = 1.25 × 10-9 and I: k2 = 1.28 × 10-9 cm3 molecule-1 s-1) with the subsequent partitioning dependent on the relative stability of the initial intermediates and the relatively large barriers present in the addition pathway. Exploration of the effect of cycloalkene ring strain on the iodiranium ion reactivity was conducted with a series of crossover experiments with 50 : 50 mixtures of either cyclohexene or cis-cyclooctene and styrene, where the outcomes were dependent on the competing ring strain relief gained by reaction with each neutral. The nature of the exchange transition state was determined to be pseudocoarctate following both natural bond orbital (NBO) and anisotropy of the induced current density (ACID) analysis.

Computational Study of the Donor-Acceptor Interactions Underlying the Variable Oxygen Probe.

The variable oxygen probe (VOP) is a crystallographic technique that has been used to explore the relative donor abilities of various filled orbitals ranging from vicinal lone pairs to polarized heteroatom-carbon bonds, remote π functionalities, and strained carbon-carbon (CC) bonds. In this study, the donor-acceptor interactions which underlie the VOP have been explored in the gas phase using density functional theory on the model systems 1-13 with natural bond orbital analysis of the various donor-acceptor interactions involving both neutral and charged σ* antibonding orbitals as the acceptor probes. Updated values for the VOP slopes of 1-13 were shown to relate qualitatively with the sum of all significant donor-acceptor interactions present in these derivatives. Application of the VOP to calculated structures of 1-13 with various -OR substituents revealed a similar relationship between the C-OR bond distance to pKa (ROH). However, the VOP slopes in the gas phase were significantly smaller in magnitude than those obtained from crystal structural data, likely due to the valence form (C+-OR) being disfavored in the former, highlighting the advantage of the VOP as an experimental technique to discriminate donor ability more effectively than calculated structures.

Experimental and DFT Studies on the Identity Exchange Reactions between Phenyl Chalcogen Iranium Ions and Alkenes.

The gas-phase ion-molecule identity exchange reactions of phenyl chalcogen iranium ions with alkenes have been examined experimentally in a linear ion trap mass spectrometer by isotope labeling experiments. The nature of both the alkene and the chalcogen play crucial roles, with the bimolecular rates for π-ligand exchange following the order: [PhTe(c-C6H10)]+ + c-C6D10 > [PhTe(C2D4)]+ + C2H4 > [PhSe(c-C6H10)]+ + c-C6D10, with no reaction being observed for [PhSe(C2D4)]+ + C2H4, [PhS(C2D4)]+ + C2H4, and [PhS(c-C6H10)]+ + c-C6D10. The experimental results correlate with RRKM modeling and density functional theory (DFT) calculations, which also demonstrates that these reactions proceed via associative mechanisms. Natural bond orbital (NBO) analysis reveals a shift in the association complexes from a σ-hole interaction to ones mirroring the π-p+ and n-π* at the transition state in accordance with the rates of reaction.

Reactions of Thiiranium and Sulfonium Ions with Alkenes in the Gas Phase.

Ion-molecule reactions between thiiranium ion 11 (m/z 213) and cyclohexene and cis-cyclooctene resulted in the formation of addition products 17a and 17b (m/z 295 and m/z 323, respectively) via an electrophilic addition pathway. Associative π-ligand exchange involving direct transfer of the PhS+ moiety, which has been observed for analogous seleniranium ions in the gas phase, did not occur despite previous solution experiments suggesting it as a valid pathway. DFT calculations at the M06-2X/def2-TZVP level of theory showed high barriers for the exchange reaction, while the addition pathway was more plausible. Further support for this pathway was provided with Hammett plots showing the rate of reaction to increase as the benzylic position of thiiranium ion derivatives became more electrophilic (ρ = +1.69; R2 = 0.974). The more reactive isomeric sulfonium ion 22 was discounted as being responsible for the observed reactivity with infrared spectroscopy and DFT calculations suggesting little possibility for isomerization. To further explore the differences in reactivity, thiiranium ion 25 and sulfonium ion 27 were formed independently, with the latter ion reacting over 260 times faster toward cis-cyclooctene than the thiiranium ion rationalized by calculations suggesting a barrierless pathway for sulfonium ion 27 to react with the cycloalkene.