Balancing speed and depth: Doubling throughput in single-cell proteomics using Orbitrap Astral Zoom MS

Significance of the topic

Single-cell proteomics by LC–MS is critical for resolving cellular heterogeneity, defining cell state and function, and enabling high-content studies in biology and medicine. Improving instrument throughput while retaining proteome depth is essential for scaling studies to hundreds-to-thousands of single cells. This work demonstrates an optimized workflow that leverages the Orbitrap Astral Zoom mass spectrometer and fast UHPLC separations to approximately double throughput with only a modest reduction in identifications, preserving quantitative precision suitable for comparative and high-throughput single-cell studies.

Objectives and study overview

• Evaluate LC–MS acquisition strategies that maximize samples per day (SPD) for low-input and single-cell proteomics without major loss of proteome coverage.
• Optimize DIA acquisition parameters (isolation width, injection time) and FAIMS settings to balance scan speed and depth.
• Benchmark quantitative accuracy and precision using a controlled three-proteome (human/yeast/E. coli) mix and dilution series of HeLa digest, and evaluate performance on real individual HeLa single cells.

Methodology

• Samples: Bulk HeLa digest, HeLa single cells prepared with the cellenONE workflow (cell diameters 15–27 µm), and a three-proteome (HYE) mix with defined compositions (Mix A: 65% human / 30% yeast / 5% E. coli; Mix B: 65% human / 15% yeast / 20% E. coli).
• Chromatography: Thermo Scientific Vanquish Neo UHPLC in direct injection mode; main column used was the IonOpticks Aurora Rapid 8×75 XT (8 cm) for rapid methods and a 25 cm Aurora column for longer gradients (lower SPD).
• Throughput design: Methods targeted 50, 80, 100 and 160 SPD. Short-gradient settings were developed for 100–160 SPD and somewhat longer settings for 50–80 SPD.
• MS acquisition: Orbitrap Astral Zoom mass spectrometer in Low Input application mode using DIA. Key MS settings: FAIMS Pro Duo interface with compensation voltage (CV) -48, carrier gas 3.5 L/min. For 100–160 SPD the preferred DIA setting was 10 Th windows with 20 ms injection time; for 50–80 SPD a 20 Th window with 40 ms injection time was retained from prior optimization.
• FAIMS optimization: Outer electrode temperature varied from 100 ºC down to 80 ºC in 5 ºC steps; 80 ºC provided the best peptide-level identifications.
• Dilution series and sensitivity testing: HeLa digest dilution series from 10 pg to 1,000 pg tested primarily with the 160 SPD method.
• Data processing: Biognosys Spectronaut 20.5 directDIA workflow (library-free) against UniProt databases. Default settings except Quantitation set to MS1 and Post analysis > Use All MS-Level Quantities unchecked. For HYE mixes searches used human, yeast, and E. coli FASTA files; normalization used human FASTA.

Instrumentation used

• Liquid chromatography: Thermo Scientific Vanquish Neo UHPLC system.
• Columns: IonOpticks Aurora Rapid 8×75 XT (8 cm) and Aurora 25×75 XT C18 (25 cm) columns.
• Mass spectrometer: Thermo Scientific Orbitrap Astral Zoom mass spectrometer.
• Ion mobility interface: Thermo Scientific FAIMS Pro Duo.
• Data analysis software: Biognosys Spectronaut 20.5 (directDIA, library-free).

Key results and discussion

• Throughput vs. depth: Doubling throughput from 50 to 100 SPD reduced protein-group identifications by ~7.7%; increasing from 80 to 160 SPD reduced identifications by ~9.7%. Thus, roughly a 2× gain in throughput can be achieved with only a ~8–10% loss in identified protein groups.
• Sensitivity and dynamic range: Using the optimized 160 SPD method, >3,000 protein groups were identified from ~10 pg of HeLa digest (library-free), >5,500 protein groups from 250 pg, and ~6,300 protein groups at 1 ng. For 250 pg on column, the 100 SPD method yielded slightly higher identifications (≈6,000+ protein groups) compared with 160 SPD (≈5,600+ protein groups).
• Single-cell performance: Real individual HeLa single cells analyzed library-free produced deep proteomes. Average identifications were ~5,808 protein groups and ~48,910 peptides for the 100 SPD method, and ~5,240 protein groups for the 160 SPD method; even at the highest throughput (160 SPD) >5,200 protein groups per single cell were reported.
• Quantitative precision: Across dilution series and throughput conditions, quantification remained robust. More than 80% of protein groups had coefficients of variation (CVs) below 20%, and median/mean CVs were generally under ~8–8.5% for most conditions.
• Accuracy benchmark (three-proteome mix): For a 250 pg load analyzed across throughputs, quantitative accuracy matched expected ratios closely: within ~2% for E. coli and ~5% for yeast relative to expected values.
• FAIMS effect: Lowering outer electrode temperature improved peptide-level identifications while protein-group numbers remained largely stable; 80 ºC was optimal in these tests.
• MS parameter optimization: Fast DIA settings (10 Th windows, 20 ms injection) maximized peptide IDs at the highest throughputs (100–160 SPD), while wider windows and longer injection times were retained for lower-throughput methods to preserve depth.

Benefits and practical applications

• The workflow enables high-throughput single-cell proteomic experiments with preserved depth and quantitative reliability, making large-cohort single-cell proteomics more tractable.
• Library-free directDIA processing simplifies analysis and avoids requirements for prior spectral libraries or booster channels, improving flexibility for exploratory studies.
• Optimized FAIMS and fast DIA schemes deliver enhanced peptide sampling and maintain precision—beneficial for differential abundance and cell population studies.
• High sensitivity (identifications from 10 pg loads) opens applications in low-input samples such as rare cell populations, microdissected tissue regions, or clinical material with limited availability.

Future trends and potential uses

• Further method automation and integration with single-cell isolation platforms will increase throughput and reproducibility for large-scale studies.
• Combining optimized fast-DIA approaches with improved sample-preparation chemistries and minimal-loss consumables will push sensitivity lower and raise identification counts at high SPD.
• Hybrid strategies that combine library-free and targeted library-assisted DIA may further improve depth for specific protein panels while retaining throughput.
• Continued tuning of FAIMS parameters, windowing schemes and real-time acquisition strategies (e.g., intelligent DIA) could increase identifications per unit time without sacrificing quantitative performance.
• Applications expand to population-scale single-cell proteomics, rapid phenotyping pipelines, and translational studies where throughput and depth are both critical.

Conclusions

The study demonstrates that the Orbitrap Astral Zoom MS paired with fast UHPLC separations and FAIMS can double sample throughput (e.g., from ~50–80 SPD to ~100–160 SPD) with only modest reductions (~7–10%) in proteome coverage. The optimized 160 SPD workflow identifies thousands of proteins even at picogram loadings, and library-free directDIA quantification remains precise and accurate. These advances enable robust, scalable single-cell and low-input proteomics experiments with a favorable balance between speed and depth.

References

1. Tabiwang N. Arrey et al. Enhanced sensitivity of the Orbitrap Astral Zoom mass spectrometer for deeper proteome coverage in single-cell proteomics. Technical Note 004019.
2. Haoran Huang et al. Orbitrap Astral mass spectrometer allows comprehensive proteome coverage at the single-cell level. Technical Note 003350.
3. Hoch, D.G.; Belford, M.; Heil, L.R.; et al. Low-resolution FAIMS for increased peptide coverage in low-load and single-cell proteomics. Scientific Reports 16, 14454 (2026).