Confident 4D Annotation of Polar Metabolites Using Standardized Retention Times Combined with TIMS-HRMS Acquisition

Importance of the topic

Application of robust, multi-dimensional identification criteria is essential in non-targeted metabolomics because small polar metabolites are difficult to separate, readily form isomers, and often lack comprehensive spectral libraries. Combining orthogonal properties — accurate mass, chromatographic retention, ion mobility (CCS), isotopic pattern and MS/MS — increases confidence in annotations, supports transferability between labs and platforms, and enables reliable tracking of metabolic dynamics in biological systems such as brown adipocytes undergoing thermogenic activation.

Objectives and overview of the study

This work presents a proof-of-concept 4D annotation workflow that anchors non-targeted discovery to standardized retention time (RT) anchors from a 3-hydrazinylphenylhydrazine (3-NPH) derivatization kit and to experimentally measured collision cross section (CCS) values obtained on TIMS instruments. The goals were to (1) build a transferable 4D Target List of 100 3-NPH–derivatized metabolites (RT, CCS, exact mass, structure), (2) extend coverage to compounds without standards using in-silico derivatization and predicted CCS/MS/MS, and (3) demonstrate the approach by profiling metabolic changes in murine brown adipocytes stimulated with norepinephrine over a 24 h time course.

Methods and workflow

The study workflow combined experimental and in-silico steps:

• Derivatization: Samples and standards were derivatized with 3-NPH (3-hydrazinylphenylhydrazine) using the biocrates MxQuant kit procedure to improve RP-LC retention and ionization of polar metabolites.
• Target List construction: Purified reference standards were analyzed to generate experimental TIMS-CCS values on a timsTOF Pro 2. These CCS values were paired with established RT anchors from the MxQuant UHPLC method to produce a 4D Target List of 100 derivatized metabolites including SMILES/InChI and exact masses.
• In-silico expansion: MetaboScape’s in-silico derivatization generated all theoretical 3-NPH derivatives from structural lists and predicted corresponding CCS and in-silico MS/MS spectra to support annotation of compounds lacking reference standards.
• Data acquisition and processing: Biological extracts (murine brown adipocytes at 0, 4, 24 h after norepinephrine) were analyzed on LC-TIMS-PASEF platforms (timsMetabo, negative mode, VIP HESI) using the MxQuant UHPLC method. MetaboScape 2026b performed mass/mobility calibration, 4D feature extraction, de-adducting, de-isotoping, grouping, automated annotation scoring (AQScore) and in-silico comparisons.

Used instrumentation

• timsTOF Pro 2 (Bruker) for generation of experimental CCS reference values.
• timsMetabo (Bruker) LC-TIMS-PASEF platform with VIP HESI ionization for routine sample acquisition.
• UHPLC system using the biocrates MxQuant kit default method for RT standardization.
• MetaboScape 2026b software for automated calibration, 4D feature extraction, AQScore evaluation, and in-silico derivatization/fragmentation and CCS prediction.

Main results and discussion

• 4D Target List: A curated list of 100 3-NPH–derivatized polar metabolites was produced, each entry containing RT anchors, experimental CCS (from timsTOF Pro 2), exact mass, isotopic pattern criteria and structural identifiers. This enabled multi-criteria matching across platforms.
• Annotation confidence: Annotation leveraged five orthogonal criteria (accurate m/z, RT, isotopic pattern (mSigma), MS/MS similarity, and CCS). The AQScore summarizes these metrics to prioritize high-confidence matches and reduce false positives arising from isomers or derivatization artifacts.
• In-silico support: For analytes lacking authentic standards, in-silico derivatization plus predicted CCS and fragmentation spectra provided complementary evidence, increasing the number of annotated features and bridging gaps in empirical libraries.
• Biological application: When applied to norepinephrine-stimulated brown adipocytes (0, 4, 24 h), the workflow captured dynamic metabolic changes consistent with thermogenic lipolysis. Notable observations included a transient increase at 4 h in medium- and long-chain fatty acids (e.g., stearic, myristic, oleic acids) consistent with lipolytic release, followed by a return toward baseline at 24 h. Central carbon metabolism and amino acid-related metabolites showed time-dependent shifts: lactate, propionate, pyruvate and beta-alanine increased over time, whereas malate, aspartate and glycine decreased at 4 h then largely normalized by 24 h.
• Complementarity: PCA and loading analyses indicated that target-list-driven annotations (empirical RT/CCS matches) and in-silico-derived annotations were complementary, jointly explaining variance in the dataset and enabling more complete biological interpretation.

Benefits and practical applications

• Improved annotation confidence: Integrating RT anchors with experimental CCS and MS/MS reduces ambiguity from isomers and increases trust in identifications for targeted review.
• Transferability: Using standardized RT anchors from an established derivatization kit and platform-specific CCS makes annotation lists more portable between labs using TIMS instruments.
• Expanded coverage: In-silico derivatization enables annotation of compounds without available standards, accelerating discovery while maintaining multi-criteria scoring to control confidence.
• Biological insights: The approach is suitable for time-course and perturbation studies (e.g., drug treatment, stimulation), where reliable annotation of polar metabolites and fatty acids is critical for interpreting metabolic regulation.

Future trends and potential applications

• Scaling CCS libraries: Expanding experimentally derived CCS databases across instrument generations and chemistries will strengthen cross-platform annotation and support community reference resources.
• Improved prediction models: Advances in machine learning for CCS and MS/MS prediction will increase the accuracy of in-silico annotations and decrease reliance on physical standards.
• Standardized RT indexing: Wider adoption of standardized retention time anchors for derivatization kits and columns will improve reproducibility between laboratories and facilitate automated transfer of target lists.
• Clinical and regulatory translation: With robust QA/QC, the 4D approach may be adapted for regulated workflows in clinical metabolomics and biomarker verification, provided derivatization and instrument-specific factors are controlled.
• Automation and integration: Tighter integration of derivatization protocols, instrument control and data-analysis pipelines will speed throughput for large cohort studies and multi-omics workflows.

Conclusion

Combining standardized RT anchors from 3-NPH derivatization kits with experimental TIMS-derived CCS values and multi-criteria scoring (m/z, isotopic pattern, MS/MS, RT, CCS) yields a practical 4D annotation strategy that increases confidence and coverage in non-targeted metabolomics. In-silico derivatization complements empirical libraries by enabling plausible annotations for analytes without standards. The proof-of-concept application to norepinephrine-stimulated brown adipocytes demonstrated biologically meaningful metabolic dynamics associated with thermogenic lipolysis, illustrating the method's utility for time-course and perturbation studies.

References

1. Anagho-Mattanovich M., et al. Multi-omics analysis of thermogenic lipolysis in brown adipocytes. 2025. (Referenced study describing derivatization and biological context).
2. Bruker. TIMS-enabled 4D-Metabolomics workflow for the automated analysis of derivatized analytes; Bruker SN-05.
3. Review: Lipolysis – A highly regulated multi-enzyme complex mediates the catabolism of cellular fat stores. (Review describing lipolysis mechanisms and prior findings consistent with metabolic shifts observed).