Search

Liquid chromatography-mass spectrometry (LC-MS) is, due to its high sensitivity and selectivity, currently the method of choice in (bio)analytical studies involving the (comprehensive) profiling of metabolites in body fluids. However, as closely related isomers are often hard to distinguish on the basis of LC-MS(MS) and identification is often dependent on the availability of reference standards, the identification of the chemical structures of detected mass spectral features remains the primary limitation. Infrared ion spectroscopy (IRIS) aids identification of MS-detected ions by providing an infrared (IR) spectrum containing structural information for a detected MS-feature. Moreover, IR spectra can be routinely and reliably predicted for many types of molecular structures using quantum-chemical calculations, potentially avoiding the need for reference standards. In this work, we demonstrate a workflow for reference-free metabolite identification that combines experiments based on high-pressure liquid chromatography (HPLC), MS and IRIS with quantum-chemical calculations that efficiently generate IR spectra and give the potential to enable reference-standard free metabolite identification. Additionally, a scoring procedure is employed which shows the potential for automated structure assignment of unknowns. Via a simple, illustrative example where we identify lysine in the plasma of a hyperlysinemia patient, we show that this approach allows the efficient assignment of a database-derived molecular structure to an unknown.

The role of specific amino acid side-chains in the fragmentation chemistry of gaseous protonated peptides resulting from collisional activation remains incompletely understood. For small peptides containing asparagine and glutamine, a dominant fragmentation channel induced by collisional activations is, in addition to deamidation, the loss of neutral water. Identifying the product ion structures from H2O-loss from four protonated dipeptides containing Asn or Gln using infrared ion spectroscopy, mechanistic details of the dissociation reactions are revealed. Several sequential dissociation reactions have also been investigated and provide additional insights into the fragmentation chemistry. While water loss can in principle occur from the C-terminus, the side chain or the amide bond carbonyl oxygen, in most cases the C-terminus was found to detach H2O, leading to a b2-sequence ion with an oxazolone structure for AlaGln, and bifurcating mechanisms leading to both oxazolone and diketopiperazine species for AlaAsn and AsnAla. In contrast, GlnAla expels water from the amide side chain leading to an imino-substituted prolinyl structure.

Complexes of divalent metal ions (Cu2+ and Ni2+) with histidine tripeptide complexes (HAA, AHA and AAH) are interesting gas-phase models for some of the most widely observed patterns of metal ion binding to peptides and proteins. Gas-phase structures were characterized using infrared multiple photon dissociation (IRMPD) spectroscopy in ion-trapping mass spectrometers, along with density functional theory (DFT) computations. Ground states are square-planar with two deprotonated amide nitrogens bound to the metal ion via a double iminol rearrangement (IM binding mode), but contrary to expectations based on solution behavior, the histidine imidazole group is not bound to the metal, but instead is hydrogen bonded remote from the metal ion. The alternative “charge-solvated” (CS) binding mode (amide carbonyl oxygens binding the metal ion) lies higher in energy, but in many cases was observed to be present as conformationally unrelaxed ions with an abundance (relative to the IM conformation) that was dependent on instrument configuration and source and trap conditions. Taking advantage of the ability to form and trap both IM and CS conformations for a few of the complexes, the infrared spectroscopy of both conformations was explored in the fingerprint (1000–1800 cm−1) and hydrogen-stretching (3200–3800 cm−1) regions. For the fingerprint region, agreement is very good between the observed IRMPD spectra and the spectra predicted by DFT calculations at the B3LYP/6–311 + +g(d,p) level. Agreement in the H-stretching region is not perfect, but the characteristic IM and CS spectral patterns are evident.

Infrared Multiple Photon Dissociation (IRMPD) Spectroscopy has been used to generate gas phase vibrational spectra for six protonated tetracycline derivatives: chlortetracycline, meclocycline, minocycline, doxycycline, tetracycline, and metacycline. The IRMPD spectrum for protonated chlortetracycline is compared here to theoretical gas phase vibrational spectra obtained from geometry optimization and frequency electronic structure calculations using the B3LYP hybrid density functional method with the 6-311 + G(d,p) basis set. The most energetically favorable protonation sites, tautomer structures, and hydrogen bond orientations were determined. Protonation at the dimethyl amine functional group and an extensive hydrogen bonding network lead to the most energetically favourable structures. Calculated spectra directly resembled the IRMPD spectrum, supporting the conclusion that the probable gas phase structure of protonated chlortetracycline has been determined.