Scaling-up low input spatial proteomics using Evosep Eno on the timsUltraAIP

Significance of the topic

The study addresses a critical bottleneck in spatial and single‑cell proteomics: achieving deep proteome coverage from very low input materials while maintaining high throughput and reproducibility. Combining miniaturized sample handling, rapid standardized chromatography and ultra‑high sensitivity trapped ion mobility mass spectrometry enables scalable analysis of hundreds to thousands of small regions or single cells, which is essential for resolving phenotypic heterogeneity in tissues and for large clinical cohorts.

Study objectives and overview

The work evaluates performance, scalability and reproducibility of an end‑to‑end workflow that couples cellenONE‑based low‑input sample preparation and proteoCHIP capture with the Evosep Eno rapid chromatography system and the timsUltra AIP instrument operated in dia‑PASEF mode. Benchmark samples included K562 peptide dilutions, single FaDu and HeLa cells (and 10‑cell mini‑bulks), and laser microdissected human FFPE tonsil regions (30,000 µm²) representing epithelial, B‑cell and T‑cell niches. Data processing used Spectronaut 21 in directDIA mode; cross‑gradient comparisons applied a downshifted normal distribution imputation when required.

Methodology and instrumentation

Workflow summary:

• Sample collection: single cells dispensed into proteoCHIP EVO96 by cellenONE X1 Neo; laser microdissection and gravity capture used for FFPE regions into proteoCHIPs.
• Sample processing: lysis, reduction/alkylation and enzymatic digestion performed in the cellenONE environment; typical preparation time ~2–2.5 h with heat‑assisted denaturation steps as described.
• Cleanup and loading: peptides cleaned and loaded onto Evotips Pure (low‑input loading from 250 pg up to 100 ng tested).
• Chromatography: Evosep Eno using standardized gradients (500, 300, 200, 100, 60, 30 SPD — samples per day) and Whisper Zoom rapid gradients (120, 80, 40 SPD) optimized for single‑cell sensitivity.
• Mass spectrometry: timsUltra AIP operating in dia‑PASEF acquisition mode for high sensitivity and ion mobility separation.
• Data analysis: Spectronaut 21 directDIA; for combined analyses across gradients a downshifted normal distribution imputation was applied to mitigate missing values.

Main results and discussion

Key findings:

• Sensitivity on diluted K562 digest: at a 250 pg load Evosep Eno + timsUltra AIP identified ~1,600 protein groups (≈8,700 precursors) using the 500 SPD method, increasing to ~3,900 protein groups (≈34,000 precursors) at 60 SPD. At the highest tested load (100 ng) identifications rose to ~5,800 protein groups (~53,000 precursors) at 500 SPD and ~8,400 protein groups (~138,000 precursors) at 60 SPD, demonstrating strong dynamic range across gradient speeds and loads.
• Single‑cell performance: single FaDu and HeLa cells yielded robust protein group identifications across both standard and Whisper Zoom gradients. Whisper Zoom 120–40 SPD delivered improved sensitivity for single cells, while standard gradients (500–60 SPD) provided a practical balance between depth and throughput.
• Spatial FFPE tonsil analysis: laser microdissected 30,000 µm² regions annotated as epithelial (CDH1), B‑cell (CD19) and T‑cell (CD3) niches produced consistent identification of cell type marker proteins across all standard gradient speeds (500–30 SPD). Multivariate analyses (PCA, Pearson correlation) showed reproducible clustering by niche and preserved spatial proteomic signatures even at short run times.
• Reproducibility and scalability: protein identification rates and abundance correlations were maintained across gradients and sample types, indicating that the workflow scales from single‑cell to small tissue regions without major loss of biological information. Data imputation enabled integrated statistical analyses across gradients.

Benefits and practical applications

The described workflow offers several practical advantages:

• High throughput: standardized Evosep Eno gradients (expressed as SPD) enable routine scaling of sample throughput according to experimental needs.
• Low‑input capability: reliable identifications from sub‑nanogram peptide loads and single cells expand applicability to precious clinical and archival FFPE samples.
• Spatial resolution preserved: LMD capture combined with sensitive MS permits detection of niche‑specific marker proteins from small tissue regions, supporting deep visual/spatial proteomics studies.
• Reproducibility: consistent quantitative profiles across gradients facilitate comparative studies and cohort designs.

Future trends and potential applications

Projected developments and opportunities:

• Further miniaturization and automation of sample handling to increase throughput and reduce losses for routine single‑cell and spatial proteomics in large cohorts.
• Tighter integration of imaging data (immunofluorescence or histology) with proteomic readouts to enable multimodal single‑cell spatial omics pipelines.
• Enhanced analytical strategies for cross‑gradient harmonization and missing‑value handling, improving comparability across throughput‑driven study designs.
• Application in translational research and biomarker discovery using archival FFPE tissue microregions and clinical sample sets where input is limiting.

Conclusion

The combination of cellenONE sample preparation, Evosep Eno rapid chromatography and timsUltra AIP dia‑PASEF MS provides a scalable, reproducible and highly sensitive platform for low‑input spatial and single‑cell proteomics. The workflow reliably detects cell‑type marker proteins from small FFPE tissue regions and achieves meaningful proteome depth from sub‑nanogram peptide loads and single cells. This approach enables high‑throughput spatial proteomics experiments suitable for dissecting tissue heterogeneity and for large cohort studies where sample material is limited.

References

1. P. Skowronek, M. Mann et al., Molecular & Cellular Proteomics, 2022, 21(9):100279.
2. Bruker Daltonics, Application Note LCMS‑206, 1815135, 2023.
3. Bruker Daltonics, Application Note LCMS‑213, 1901456, 2023.
4. Bruker Daltonics, Application Note LCMS‑222, 1911577, 2024.
5. Bruker Daltonics, Application Note LCMS‑228, 1914261, 2024.
6. Bruker Daltonics, Application Note LCMS‑233, 1915345, 2025.
7. Bruker Daltonics, Application Note LCMS‑238, 1918674, 2025.
8. Bruker Daltonics, Application Note LCMS‑261, 1929343, 2026.
9. Bruker Daltonics, Application Note LCMS‑262, 1929349, 2026.