Screening Metabolomics of Human Urine Using a Waters™ High Resolution QTof Mass Spectrometer and MARS Data Processing Software

Importance of the Topic

Metabolomic profiling of human urine offers deep insights into physiological and environmental influences on health. Rapid, high-resolution screening methods are key to detecting subtle metabolic changes associated with disease, treatment effects, lifestyle and exposure.

Study Objectives and Overview

This work demonstrates an end-to-end workflow for untargeted urine metabolomics. Three pooled urine samples representing non-smokers, passive nicotine exposure, and active smokers (NIST SRM 3671) were analyzed to evaluate reproducibility, sensitivity, and the ability to distinguish lifestyle groups.

Methodology and Instrumentation

Sample Preparation:

• Combined 200 µL aliquots into a QC pool; spiked with a diluted LCMS Reference Standard.
• Diluted 10 µL sample + QC with water and reference mix; vortexed; five replicates injected.

LC–MS Conditions:

• ACQUITY I-Class Premier FTN with BEH Amide 1.0×50 mm, 1.7 µm at 45 °C.
• Mobile phases: water + 0.1% FA (A) and acetonitrile + 0.1% FA (B); 1 µL injection, 0.6 mL/min gradient.
• Xevo G3 QTof in positive mode; 50–1200 Da, MSE continuum, 10 Hz, ±5 ppm mass accuracy.

Data Processing:

• Direct import from waters_connect into MARS™ software.
• Peak detection: baseline correction, smoothing, alignment, deisotoping, gap filling.
• Feature matrix generation (m/z@RT) and multivariate analysis (PCA, PLS-DA).
• Compound identification via HMDB and custom libraries with MS/MS and optional CCS matching.
• Trend analysis, quantification tools and pathway mapping.

Main Findings and Discussion

The combined HILIC-UPLC/QTof platform delivered stable retention times and consistent mass accuracy within ±5 ppm over five orders of magnitude dynamic range. Unsupervised PCA clearly separated non-smokers, passive and active smokers, driven by exogenous markers.

Approximately 6000 features were putatively annotated; high-confidence identifications (score ≥80) were color-coded for reliability. In silico drug metabolite prediction identified cocaine metabolites such as benzoylecgonine, which could be excluded to refine group separation. Trend clustering revealed metabolites and endogenous compounds with distinct level-dependent abundance patterns.

Pathway mapping linked identified compounds to human metabolic networks, exemplified by tyramine in the tyrosine metabolism map.

Benefits and Practical Applications of the Method

• Robust, high-throughput acquisition for large cohort studies.
• Customizable workflows tailored to discovery goals.
• Integrated multivariate statistics, identification, quantification and pathway analysis.
• Direct data import from Waters platforms without file export overhead.

Future Trends and Opportunities

Emerging QA/QC guidelines (mQACC) will drive standardization in untargeted metabolomics. Integration of collision cross-section data and expanded spectral libraries (MoNa, MiMe) will improve identification confidence. Advances in machine learning and automated pathway inference could further enhance biomarker discovery and personalized medicine applications.

Conclusion

The workflow combining HILIC UPLC, high-resolution QTof mass spectrometry and MARS data processing delivers rapid, reliable urine metabolomic screening. It provides robust reproducibility, sensitive detection, and comprehensive data analysis from peak picking through pathway profiling, suited for large-scale discovery studies.

Reference

1. Waters Application Note. Reproducibility and Robustness of the Xevo G3 QTof. 2024.
2. Waters Application Brief. Improved Transmission on Xevo G3 QTof. 2022.
3. Waters Application Brief. Sensitivity and Linear Dynamic Range of the Xevo G3 QTof. 2023.
4. Beger RD, et al. Metabolomics QA/QC Consortium Best Practices. Metabolomics. 2019;15(1):4.
5. Evans AM, et al. QA/QC Practices in LC-MS Untargeted Metabolomics. Metabolomics. 2020;16(10):113.
6. Goracci L, et al. MARS Software for Untargeted LC-MS Metabolomics. Anal. Chem. 2024;96(4):1468–1477.
7. Wishart DS, et al. HMDB 5.0: Human Metabolome Database. Nucleic Acids Res. 2022;50(D1):D622–D631.
8. MassBank of North America (MoNA). Fiehn Lab, UC Davis.
9. Wishart DS, et al. MiMeDB: Human Microbial Metabolome Database. Nucleic Acids Res. 2023;51(D1):D611–D620.