High throughput targeted proteomic workflows with Adaptive Retention Time on the Orbitrap Astral Zoom MS

Adaptive Retention Time Correction on the Orbitrap Astral Zoom MS: High‑Throughput Targeted Proteomics

Importance of the topic

Rapid, reliable targeted proteomics is critical for validating biomarkers and supporting large cohort studies. As throughput increases and chromatographic peaks narrow, small retention time (RT) shifts from matrix effects, column degradation or system changes substantially reduce the efficiency of scheduled targeted acquisition. Adaptive retention time (aRT) alignment provides a live correction strategy that maintains narrow acquisition windows and high target capacity, enabling robust targeted assays at very high throughputs (e.g., 300 samples per day).

Objectives and overview of the study

The study evaluates the performance of aRT implemented on the Thermo Scientific Orbitrap Astral Zoom MS to:

• Demonstrate robustness of aRT against common RT shift sources (plasma matrix effects, column/sample line replacement, severely degraded column, deliberate flow‑rate induced shifts).
• Assess target recovery and peak truncation for large assays (up to ~2,715 targets) with high‑throughput gradients.
• Show feasibility of combining targeted acquisition with discovery (Hybrid DIA) while preserving DIA depth.

Methodology and data analysis

Experimental design

Reference libraries and RT targets were generated from DIA reference runs analyzed primarily with Spectronaut and Skyline. Targets were selected across the RT range to form small and large assays. Reference acquisitions used MS1 and/or wide‑window DIA to create the aRT spectral map. During targeted runs, periodic alignment scans (MS1 or wide window DIA) were cross‑correlated in real time with the reference to compute instantaneous RT shifts and adapt target acquisition windows accordingly.

Samples and LC‑MS setup

• Samples: Thermo Scientific Pierce HeLa digest (200 ng injections) and individual/pooled human plasma.
• LC: Thermo Scientific Vanquish Neo UHPLC with 15 cm PepMap analytical columns (150 µm ID).
• Throughputs tested: 60 and 300 samples per day (SPD) using fast gradients.
• MS: Orbitrap Astral Zoom MS configured for aRT alignment scans, targeted MS2 scans and narrow‑window DIA in hybrid workflows.

Key algorithmic approach

Alignment scans are acquired periodically and spectrally cross‑correlated with the reference acquisition across small time offsets to estimate RT shift. The estimated shift updates target windows in real time, enabling much narrower scheduling windows than static methods.

Used instrumentation

• Orbitrap Astral Zoom mass spectrometer (Thermo Scientific)
• Vanquish Neo UHPLC system (Thermo Scientific)
• PepMap analytical columns, 15 cm × 150 µm ID (Thermo Scientific)
• Data analysis software: Spectronaut (Biognosys) and Skyline

Main results and discussion

Robustness to RT shifts

aRT effectively tracked and corrected RT shifts induced by plasma matrix effects, routine column or sample line replacement, and severely degraded columns. It also corrected deliberate shifts introduced by flow‑rate variation.

Target recovery and peak truncation for large assays

• Assays containing up to 2,715 targets achieved ≥98% target recovery in tested conditions; a small fraction (~2%) of missing targets were attributed to upstream RT assignment or library artifacts.
• Median peak truncation remained low across assay sizes: 0.4% for 979 targets, 1.1% for 1,900 targets, and 0.7% for 2,715 targets.
• Considering targets with <10% peak truncation, capture rates were 98% (979 targets), 93.3% (1,900 targets) and 96.2% (2,715 targets).

Narrow acquisition windows and throughput

Using aRT allowed reliable use of target acquisition windows <15 seconds at 300 SPD; without aRT, static windows would need to be substantially wider (e.g., ~30 s) to maintain comparable recovery, reducing the number of concurrent targets per run.

Hybrid DIA — combining targeting with discovery

Hybrid workflows combining aRT MS1 alignment scans, narrow‑window DIA and targeted MS2 acquisition were evaluated. Incorporating approximately 1,000 targeted precursors (max inject time 5 ms) produced near‑complete target recovery while minimally impacting discovery performance: <2% decrease in protein group identifications and ~13% decrease in peptide identifications compared with narrow‑window DIA alone. This demonstrates the practical potential to target many peptides of interest (including lower abundance species or PTM‑bearing peptides) while preserving broad DIA‑based discovery.

Benefits and practical applications

• Greatly increased target capacity per run due to narrower windows and live RT correction, enabling large‑scale targeted assays suitable for cohort studies and clinical verification workflows.
• Robust operation across routine sources of RT variability (matrix, column wear, system maintenance), reducing the need for wide static windows or heavy labeled triggers.
• Effective combination of targeted and discovery acquisition modes (Hybrid DIA) allows prioritized monitoring of peptides of interest with minor compromise to proteome coverage.

Limitations and considerations

• Performance depends on quality of the reference acquisition and correctness of initial RT assignments; library artifacts can account for a portion of missed targets.
• Initial evaluations focused on relatively abundant targets; performance for very low abundance or heavily modified peptides in complex mixtures may require further optimization (e.g., injection times, enrichment strategies).
• Implementation requires instrument and software support for real‑time alignment scans and fast cycle times (Astral MS design supports this).

Future trends and potential uses

• Scaling to routine clinical and large‑cohort studies where throughput and reproducibility are critical.
• Integration with machine‑learning RT prediction and improved spectral matching to further reduce missing targets and minimize required reference runs.
• Extending hybrid approaches to targeted PTM monitoring (phosphopeptides, glycopeptides) in unenriched samples, or to quantify low‑abundance biomarkers by combining targeted MS2 with DIA context scans.
• Enhanced multiplexing by combining aRT with shorter gradients, higher MS2 parallelization and optimized scheduling algorithms to increase targets per run without sacrificing discovery depth.

Conclusion

Adaptive RT alignment on the Orbitrap Astral Zoom MS enables robust, large‑scale targeted proteomics at high throughput. aRT maintains high target recovery and low peak truncation for thousands of targets, permits target windows below 15 seconds at 300 SPD, and supports hybrid DIA strategies that combine targeted monitoring of hundreds to thousands of peptides with discovery acquisition while incurring minimal loss of proteome coverage. These capabilities facilitate translation of discovery proteomics into high‑throughput, reproducible targeted assays suitable for biomarker validation and large cohort studies.

References

1. Remes PM, et al. Highly Multiplex Targeted Proteomics Enabled by Real‑Time Chromatographic Alignment. Analytical Chemistry. 2020.

Conflict of interest note

Most authors are employees of Thermo Fisher Scientific, manufacturer of the instrumentation used in these experiments.