Accelerating metabolite identification with mass spectrometry

Significance of the topic

Metabolite identification (MetID) is a cornerstone of small-molecule drug discovery and development because metabolic transformation governs exposure, efficacy, and safety. Confident structural assignment and quantitation of metabolites — including low-abundance and reactive species — enables early risk assessment, design optimization, and regulatory compliance. High-resolution accurate-mass (HRAM) LC-MS workflows coupled with automated data processing reduce time-to-insight and help prevent late-stage surprises that can derail clinical programs.

Objectives and overview of the study

This whitepaper demonstrates integrated HRAM LC-MS approaches to accelerate and increase confidence in MetID. The goals are to (1) showcase automated processing for isotope-labeled glutathione (GSH) adduct discovery, (2) demonstrate sensitivity gains and matrix-interference reduction using intelligent acquisition, and (3) illustrate multi-stage fragmentation (MSn) for complex bi- functional molecules such as PROTACs. Case studies use model compounds (clozapine, nefazodone, montelukast, timolol, and PROTACs MZ1/dBET1) to exemplify the workflows.

Methodology

Sample preparation and incubation:

• In vitro incubations in rat or human liver microsomes or S9 fractions with appropriate cofactors (e.g., NADPH) to generate Phase I/II metabolites.
• GSH trapping for reactive metabolite capture using a 1:1 mix of unlabeled and stable isotope-labeled GSH ([13C2,15N]) to create diagnostic isotopic patterns for adducts.
• Quench, protein precipitation and, in some cases, spiking into plasma to mimic in vivo matrix complexity.

Chromatography and mass spectrometry:

• Reversed-phase UHPLC separations (Vanquish family) with columns tuned for selectivity (Accucore C30, Hypersil GOLD C18).
• HRAM MS acquisition using Orbitrap platforms: Exploris 240 for high-resolution full-scan and HCD-MS2, Ascend and IQ-X Tribrid systems for MSn-enabled experiments with ion trap fragmentation when needed.
• Acquisition modes: full-scan HRAM, Data-Dependent Acquisition (DDA), and intelligent background exclusion workflows (AcquireX) with rapid polarity switching where beneficial.

Data processing and annotation:

• Automated, node-based processing in Compound Discoverer to combine targeted and untargeted analyses in a single workflow.
• Pattern Scoring to detect isotope-labeled GSH incorporation, Search Neutral Losses to flag GSH-specific fragments, and FISh Scoring for in silico fragmentation matching and structural annotation.
• Use of spectral libraries (mzCloud), formula prediction, database searching, and in-silico predictions to increase annotation confidence.

Used instrumentation

• Thermo Scientific Vanquish Horizon / Flex / Vanquish UHPLC systems.
• Thermo Scientific Orbitrap Exploris 240 mass spectrometer.
• Thermo Scientific Orbitrap Ascend BioPharma Tribrid mass spectrometer (for MSn).
• Thermo Scientific Orbitrap IQ-X Tribrid mass spectrometer.
• Columns: Thermo Scientific Accucore C30 and Hypersil GOLD C18 Selectivity.
• Software: Compound Discoverer, TraceFinder, mzCloud, and AcquireX intelligent acquisition.

Main results and discussion

Detection and annotation of GSH adducts:

• Using clozapine as a model, the Exploris 240 + Compound Discoverer workflow identified nine GSH-bound reactive metabolites by combining isotope-label pattern detection, neutral loss searching, and FISh-based fragmentation matching. Sub-ppm mass accuracy and isotopic fine structure allowed confident elemental formula assignment and structural localization (e.g., hydroxylation site assignment).

Improved sensitivity and metabolite discovery using AcquireX:

• Intelligent background exclusion (AcquireX) increased metabolite identification rates by ~30–50% versus conventional DDA across tested compounds — example counts: nefazodone 10 (DDA) vs 21 (AcquireX); montelukast 4 vs 11; timolol 5 vs 7.
• Polarity-switching full-scan HRAM data with sub-ppm mass errors and isotopic fine structure enabled assignment of challenging metabolites (e.g., sulfur-containing timolol metabolite C13H24O5N4S at m/z 349.1541 / 347.1395 in opposite polarities) with robust elemental composition confirmation.

MSn for structurally complex PROTACs:

• For PROTAC model compounds, the Ascend Tribrid with MSn (ion-trap-based stepwise fragmentation) combined with AcquireX produced comprehensive fragmentation trees that localized transformation sites and resolved isomeric metabolites. The study reported 24 metabolites for MZ1 and 12 for dBET1, with MS3 spectra pinpointing oxidation sites in either the ligand or linker regions and distinguishing hydrolysis/cleavage products.

Automation and throughput gains:

• Compound Discoverer’s integrated workflows reduced manual curation by automating pattern detection, neutral loss searching, in-silico fragmentation matching (FISh), and statistical prioritization, enabling faster triage of biologically relevant features from complex datasets.

Benefits and practical applications

• Higher confidence in structural assignments through combined orthogonal evidence: accurate mass, retention time, isotopic fine structure, isotope-label patterns, and MSn fragmentation.
• Better capture of reactive and low-abundance metabolites (e.g., GSH adducts) that are critical for early safety assessment.
• Reduced matrix interference and increased MS/MS triggering for drug-related features using AcquireX, improving discovery in plasma or microsomal matrices.
• Scalable workflows applicable across discovery-to-ADME stages: targeted quantitation of known metabolites and untargeted profiling for unknowns in a single acquisition/processing pipeline.

Future trends and opportunities

• Deeper integration of intelligent acquisition with real-time decision algorithms and machine learning to prioritize biologically relevant precursors and reduce false positives.
• Expanded use of isotope-labeling strategies (stable isotopes beyond GSH) for metabolic flux analysis and absolute quantitation in MetID workflows.
• Wider adoption of MSn-capable HRAM platforms in routine MetID to resolve isomers and complex biotransformations (notably for modalities such as PROTACs and other heterobifunctional molecules).
• Automated annotation improvements via larger, curated spectral libraries and enhanced in-silico fragmentation tools to accelerate dereplication and putative identification.
• Regulatory harmonization toward acceptance of HRAM/MSn-derived structural evidence for metabolite safety assessments as these technologies mature and become standardized.

Conclusion

Integrated HRAM LC-MS workflows that combine high mass accuracy, intelligent acquisition (AcquireX), MSn fragmentation, and automated node-based data processing (Compound Discoverer) materially improve the speed, sensitivity, and confidence of metabolite identification. These capabilities address key MetID pain points — low-abundance/reactive species detection, matrix interference, and structural localization in complex molecules — thereby accelerating lead optimization and reducing downstream safety risk.

Reference

• White paper: Increased confidence in drug metabolite identification through intelligent data acquisition strategies and multiple fragmentation techniques on the Orbitrap Tribrid MS platform.
• Case study: A streamlined solution for confident detection and identification of isotope-labeled glutathione-trapped reactive drug metabolites using the Orbitrap Exploris 240 MS and the Compound Discoverer software.