Ion Mobility-Enabled Metabolite Identification of Tienilic Acid and Tienilic Acid Isomer

Library Number:
PSTR135022460
Author(s):
Lauren Mullin, Giorgis Isaac, Ian Wilson, Gordon Murray, Nathan Anderson, Robert Plumb
Source:
Waters, Imperial College
Content Type:
Posters
Content Subtype:
ASMS
Related Products:
 
 
ACQUITY UPLC I-Class System

Novel aspect:

Implementation of IMS-HRMS for TA and TAI metabolite identification and assessment of CCS modelling techniques for metabolite identification support.

Introduction:

Tienilic acid (TA) is a uricosuric diuretic found to induce immune-mediated hepatotoxicity in patients, while its 3-thiophene isomer (TAI) exhibits direct hepatotoxic effects1. Metabolite identification is a critical step in understanding these differential mechanisms of toxicity. High resolution mass spectrometry (HRMS) is a powerful tool to elucidate metabolite structures. Recent advances in coupling ion mobility separation (IMS) provide further means of metabolite identification and specification. The aim of this study is to demonstrate this approach in the analysis of urine from rats treated with TA and TAI and collected at three different time points. IMS-derived collision-cross section (CCS) experimental values are also compared with theoretical values obtained through machine-learning models, providing an avenue for further metabolite identification support.

Methods:

Rat urine was collected at 2hr, 6hr and 24hr for TA, TAI and blank vehicle treated animals. Data was acquired on a travelling-wave ion mobility (TW-IM) QTof following separation via UPLC using 0.1% formic acid in water (A) and acetonitrile (B) on an HSS T3 (100 x 2.1 mm 1.8µm) UPLC column. Data was acquired using ESI+ over a mass range of 50-1200m/z and a capillary voltage of 1.00kV. Alternating low (6eV) and elevated (35-55eV ramp) collision energy states were used during the run time. Data was processed in commercially available metabolite identification software and CCS values were calculated for all identified compounds. CCS modelling was performed using machine learning methods based on proposed metabolite molecular properties.

Preliminary data:

Metabolite identification of urine samples indicate hydroxylation for both TA and TAI, though occurring in higher relative abundance for TA at all three time points. Structural confirmation of the site of metabolism was provided in the elevated collision energy spectra through the presence of a +16Da shift in the thiophene group product ion as compared to the parent substrate. In general, a greater number of metabolites were identified in TA treated animals than those with TAI, and included reduction followed by glucuronidation, oxidation followed by acetylation, and glycine conjugation. Identifications were supported by the presence of common fragments resulting from loss of the thiophene and carboxylic groups, as well as expected isotope distribution patterns of the double chlorinated TA and TAI molecules and the previously mentioned product ions.

Use of IMS provided a means to separate co-eluting metabolites by drift-aligning spectra, improving spectral clarity and aiding in compound identification in this dataset. In an early assessment of a recently developed machine-learning based CCS modelling program, a subset of the identified metabolite structures’ CCS theoretical values were modelled and found to be all within 2% error of experimental CCS values. Further exploration of CCS modelling utility in identification support will also be presented.


Title Format File Size
Download PDF PDF 889.41kB