Using Parallel Accumulation Serial Fragmentation (PASEF) to speed up untargeted 4D lipidomics LC-MS/MS workflows

Význam tématu

In-depth lipid profiling is increasingly applied in clinical research on diseases such as cancer and neurological disorders. Lipidomics offers insights into biomarker discovery, disease mechanism studies and patient stratification. High sample throughput without sacrificing data quality is critical for large cohort studies, demanding advanced acquisition strategies capable of rapid, high-sensitivity measurements.

Cíle a přehled studie / článku

The primary objective was to demonstrate how PASEF on a timsTOF Pro platform accelerates untargeted 4D lipidomics workflows. Two major goals were addressed:

• Maximize lipid identifications in an in-depth profiling regime with a 20 min gradient.
• Assess throughput gains using shortened LC gradients of 11 and 6 min while preserving identification confidence.

Použitá instrumentace

The analytical setup comprised:

• Bruker Elute UHPLC coupled to timsTOF Pro.
• Bruker intensity C18 column (100×2.1 mm, 1.9 µm).
• Apollo II ESI source, positive mode, 4500 V capillary voltage.
• timsTOF Pro operated in PASEF MS/MS mode with 100 ms ramp time.
• MetaboScape 5.0 with T-ReX 4D and CCSPredict for data processing.

Použitá metodika

SRM 1950 human plasma lipids were extracted and dissolved in MeOH/DCM. A constant injection of 0.5 µL extract was applied in five replicates. The reversed-phase gradient varied from 6 to 20 min to test throughput versus depth. MS data acquisition covered m/z 100-1500 for precursors and 300-1500 for fragments. Four-dimensional feature detection integrated m/z, retention time, mobility and intensity. Blank subtraction and LipidBlast matching enabled putative annotation. CCS values were both measured and machine-predicted to reinforce identification confidence.

Hlavní výsledky a diskuse

A 20 min gradient analysis yielded 392 lipid identifications across 286 sum compositions, surpassing interlaboratory benchmarks. When gradients shortened to 11 min and 6 min, identified lipids remained at 87 % and 82 % of the 20 min depth, respectively (>200 lipids in 6 min runs). PASEF achieved up to 102 precursor selections within 0.1 min, underlining its high MS/MS throughput. Mobility separation resolved co-eluting isobaric species (e.g. PC 34 :2e vs PE 36 :2) with only 36 mDa mass difference, yielding clean fragment spectra and CCS deviations below 0.5 % against predictions. A spike-in experiment with lipid standards demonstrated that 6 min data suffice for reliable group differentiation by PCA and statistical testing.

Přínosy a praktické využití metody

PASEF enables rapid 4D lipid profiling without compromising selectivity or confidence. The combination of ion mobility separation and high-speed MS/MS:

• Increases sample throughput by nearly fourfold.
• Separates isobaric and isomeric lipids.
• Provides accurate, reproducible CCS metrics for confirmatory identification.
• Facilitates large-scale clinical lipidomics studies with automated deep profiling or high-turnover screening.

Budoucí trendy a možnosti využití

Future developments may integrate machine-learning-driven CCS prediction and expanded in silico spectral libraries to enhance annotation coverage. Coupling PASEF with targeted lipid quantitation and automated data QC will broaden applications in biomarker validation and personalized medicine. Integration with multiomics platforms could further elucidate lipid pathway regulation in health and disease.

Závěr

This work confirms that parallel accumulation serial fragmentation on timsTOF Pro dramatically accelerates untargeted lipidomics while maintaining high data quality. Even with sub-10 min gradients, the four-dimensional approach ensures robust separation, precise fragmentation and reliable identification. PASEF stands as a versatile solution for both comprehensive lipid discovery and high-throughput clinical profiling.

Reference

1. Meier F et al J Proteome Res 14 5378–5387 2015
2. Shevchenko A et al J Lipid Res 49 1137–1146 2008
3. Bowden JA et al J Lipid Res 58 2275–2288 2017
4. Zhou et al standard CCS prediction approach
5. LipidBlast spectral library