Unlocking the archived proteome: High-throughput, deep FFPE proteome profiling using the Orbitrap Astral mass spectrometer

Importance of the topic

FFPE (formalin-fixed, paraffin-embedded) tissue archives represent an indispensable clinical resource for retrospective studies, biomarker discovery and translational oncology because they preserve tissue morphology and are accompanied by extensive clinical annotation. However, chemical crosslinking from formalin and paraffin embedding complicates protein extraction and digestion, historically limiting throughput, sensitivity and reproducibility of proteomic analyses. Streamlined, high-throughput workflows that recover deep proteome coverage from FFPE sections while remaining robust and scalable are therefore critical to enable large-cohort proteomic studies directly from clinically relevant material.

Goals and overview of the study

The technical note demonstrates an end-to-end, rapid FFPE proteomics workflow built around the Orbitrap Astral mass spectrometer and the OptiSpray ion source coupled to two LC platforms (Thermo Vanquish Neo UHPLC and the Evosep Eno). The objectives were to: (1) simplify and shorten FFPE sample preparation, (2) enable sensitive peptide detection across low-to-moderate peptide loads (20–200 ng), (3) explore trade-offs between proteome depth and acquisition speed (60, 180, 500 samples per day, SPD), and (4) confirm that biological differences between tumor and normal lung tissues can be robustly captured at high throughput.

Methodology

The workflow integrates rapid deparaffinization (xylene/ethanol washes), formalin crosslink reversal and lysis, and a streamlined digestion/cleanup using the EasyPep Mini kit. Protein and peptide amounts were quantified by Rapid Gold BCA and a fluorometric peptide assay prior to LC–MS/MS. Three LC throughput modes were evaluated: 60 SPD (~20 min gradient on an OptiSpray µPAC Neo 50 cm nano cartridge), 180 SPD (~6.8 min gradient on an OptiSpray µPAC Neo high-throughput cartridge), and 500 SPD (~2.3 min gradient using Evosep Eno with an Evosep EV1182 column). MS acquisition used data-independent acquisition (DIA) with directDIA processing in Spectronaut and Pulsar as the search engine. Data were filtered at 1% precursor and protein-group FDR and analyzed in R for downstream visualization and statistics.

Used instrumentation

• Mass spectrometer: Orbitrap Astral Mass Spectrometer with Thermo OptiSpray ion source.
• UHPLC systems: Thermo Vanquish Neo UHPLC (direct injection for 60 and 180 SPD) and Evosep Eno (for 500 SPD).
• Columns/cartridges: OptiSpray µPAC Neo 50 cm nano cartridge; OptiSpray µPAC Neo high-throughput cartridge; Evosep EV1182 Performance Column; Evosep Evotip Pure tips for the Evosep workflow.
• Sample prep: Thermo EasyPep Mini MS Sample Prep Kit; detergents including n-dodecyl β-D-maltoside (DDM).
• Consumables and reagents: LC–MS grade water, acetonitrile, formic acid; Pierce Rapid Gold BCA and Pierce fluorometric peptide assay kits.
• Software: Biognosys Spectronaut (directDIA mode), Pulsar search engine, RStudio (R) for downstream analysis.

Main results and discussion

• Turnaround time: The complete FFPE-to-data workflow can be completed in under one day (≈3 h deparaffinization/lysis, 3–4 h digestion + cleanup with drying time, <25 min LC–MS run per sample, ~20 min automated data processing).
• Proteome depth: At the deepest setting (60 SPD, 20 min gradient) ~8,600 protein groups and ≈105,000 peptides were identified from 200 ng on-column. Proteome coverage scaled down with higher throughput: at 180 SPD (≈7 min) ~6,100 protein groups and ~62,000 peptides (~70% of 60 SPD protein IDs) and at 500 SPD (≈2 min) the method still recovered several thousand proteins (~45% of the 60 SPD core proteome depending on load).
• Low-input performance: Even at low on-column loads (20 ng) the workflow delivered robust identifications (several thousand protein groups and tens of thousands of peptides), indicating effective peptide recovery and sensitivity of the Astral platform coupled with optimized LC.
• Quantitative reproducibility: Triplicate CVs at the protein level were low across SPDs and sample types: median CVs were ~5–6% for 60 SPD, ~7–10% for 180 and 500 SPD, and generally remained within 15% at the highest throughput, supporting reliable quantitative comparisons in large cohorts.
• Biological concordance: Fold-change measurements comparing lung tumor vs. matched normal tissues were highly correlated across acquisition modes (Pearson r = 0.93 for 180 vs. 60 SPD; r = 0.87 for 500 vs. 60 SPD), indicating that accelerated acquisition maintains biological signal fidelity.
• Differential expression and pathways: Using 200 ng and 60 SPD, ~1,500 protein groups were significantly regulated (|log2FC| ≥1, p ≤0.05). Upregulated tumor pathways included neutrophil extracellular trap formation, transcriptional dysregulation in cancer, glycolysis/gluconeogenesis, HIF-1 signaling, ECM–receptor interaction and central carbon metabolism. Regulated proteins were distributed across abundance ranks, showing that the workflow captures both high- and mid/low-abundance disease-relevant changes.

Benefits and practical applications

• Scalability: The protocol affords flexible trade-offs between depth and throughput to suit discovery experiments, cohort-wide profiling or targeted biomarker studies.
• Compatibility with archived clinical samples: Enables direct interrogation of FFPE tissue banks, facilitating retrospective studies tied to clinical metadata.
• Operational robustness: Reduced manual steps, automated ion source operation and short gradients minimize instrument contamination and increase laboratory throughput and reproducibility.
• Low sample requirement: Reliable performance at tens of nanograms of input opens applications for precious or microdissected clinical specimens.

Future trends and potential applications

• Wider adoption of ultra-high-speed DIA on platforms like the Orbitrap Astral will accelerate population-scale proteomics from clinical archives, enabling larger retrospective biomarker discovery and validation studies.
• Integration with spatial and single-cell sample-prep techniques could extend depth while preserving tissue context, allowing combined morphological and proteomic mapping of tumor microenvironments.
• Further improvements in sample processing chemistries and automated front-end systems are likely to reduce batch effects and increase throughput without sacrificing sensitivity, facilitating routine incorporation into translational pipelines.
• Combining rapid FFPE proteomics with phosphoproteomics, targeted assays and machine-learning-driven data analysis will expand mechanistic insights and clinical utility.

Conclusion

The reported workflow demonstrates that with optimized FFPE sample preparation, fast LC methods and the high acquisition speed and sensitivity of the Orbitrap Astral, it is possible to achieve deep and reproducible proteome coverage from archived FFPE lung tissues in under a day. Throughput can be increased up to ~8-fold with only moderate losses in depth while preserving quantitative precision and biologically meaningful differential expression, enabling large-scale translational proteomic studies directly from clinical archives.

Reference

1. Haines M., et al. High-Throughput Proteomic and Phosphoproteomic Analysis of Formalin-Fixed Paraffin-Embedded Tissues. Molecular & Cellular Proteomics. 2025.
2. Zhu Y., et al. High-throughput proteomic analysis of FFPE tissue samples facilitates tumor stratification. Molecular Oncology. 2019;13:2305–2328.
3. Wang Y., et al. Effects of tumor metabolic microenvironment on regulatory T cells. Molecular Cancer. 2018.
4. Liu Y., Vandekeere A., Xu M., Fendt S.M., Altea-Manzano P. Metabolite-derived protein modifications modulating oncogenic signaling. Frontiers in Oncology. 2022.