Unlocking the immunopeptidome - High-sensitivity and accurate mass spectrometry with the Orbitrap Astral Zoom mass spectrometer for superior antigen characterization

Significance of the topic

The immunopeptidome — the set of peptides presented by MHC/HLA molecules on the cell surface — is central to antigen presentation and T-cell surveillance. Direct identification of these peptides by liquid chromatography–mass spectrometry (LC–MS) is indispensable for discovery of tumor-specific antigens, neoantigens and other clinically relevant epitopes. Analytical challenges include low peptide abundance, lack of predictable proteolytic termini, and limited sample amounts from clinical biopsies. High-sensitivity, wide dynamic-range MS workflows are therefore critical to advance translational and basic immunopeptidomics.

Objectives and overview of the study

This technical note assessed depth and spectral quality of HLA Class I peptide identification using data-dependent acquisition (DDA) on the Orbitrap Astral Zoom mass spectrometer. The study aimed to demonstrate sensitivity across orders of magnitude of sample input (equivalent of 1e5 to 1e8 cells), to evaluate the contribution of FAIMS gas-phase fractionation for singly versus multiply charged peptides, and to characterize spectral quality and quantitative linearity relevant for neoantigen discovery and de novo sequencing.

Methodology

Sample preparation: HLA Class I complexes were immunoaffinity-enriched from IM-9 human B-lymphocyte cells (starting from ~1e8 cells) using the pan-class I antibody (W6/32). Bound peptides were eluted with 30% acetonitrile, dried and reconstituted in 3% ACN/5% formic acid. To simulate varying input, aliquots representing cell-equivalents of 1e5, 1e6, 1e7 and 1e8 were analyzed.

LC–MS and acquisition: Nano‑UHPLC separations used an IonOpticks Aurora Ultimate XT 25 cm C18 column on a Vanquish Neo UHPLC. The Orbitrap Astral Zoom MS was operated in a Low Input application mode with ion pre‑accumulation (bent trap) to favor single‑ion detection. A three‑CV FAIMS strategy (−20 V, −45 V and −60 V) was employed to include both singly charged species (favored at −20 V) and multiply charged peptides (favored at −45/−60 V). DDA cycle time and MS2 settings were tuned for high-resolution full scans (120,000) and Astral MS2 acquisition. AGC modulation adjusted injection times across different loadings.

Data analysis: PEAKS Studio 12 (DeepNovo Peptidome workflow) performed combined database search (UniProt human, no-enzyme setting) and de novo sequencing. Search parameters included variable Met oxidation and N‑terminal acetylation, precursor tolerance 10 ppm, fragment tolerance 0.02 Da, and peptide‑level FDR at 1%. MHCMotifDecon 1.2 was used for motif and allele prediction.

Instrumentation

• Orbitrap Astral Zoom mass spectrometer
• Thermo Scientific FAIMS Pro Duo interface
• Thermo Scientific EasySpray ion source
• Thermo Scientific Vanquish Neo UHPLC system
• IonOpticks Aurora Ultimate XT 25 × 75 C18 column and heater
• AssayMap Bravo automated immunoaffinity system (Agilent) used for peptide enrichment
• PEAKS Studio 12 and MHCMotifDecon 1.2 for data processing and motif analysis

Main results and discussion

Depth of coverage: Average peptide identifications (n=3) increased with input: ~9,006 peptides at 1e5, ~16,574 at 1e6, ~26,971 at 1e7 and ~27,553 at 1e8 cell equivalents, demonstrating the workflow can recover tens of thousands of immunopeptides from low to high inputs.

Charge-state capture and FAIMS impact: The fraction of singly charged peptides varied with loading (about 13%–47% depending on sample amount). A small subset (~4.8%) were observed as both singly and multiply charged. Using three FAIMS compensation voltages enabled inclusion of singly charged species (−20 V) and enriched sampling of multiply charged peptides (−45/−60 V), expanding overall coverage.

Peptide characteristics: Length distribution matched Class I expectations, with most peptides of 8–9 amino acids and 9‑mer motifs displaying canonical HLA‑I binding patterns. Allele deconvolution recovered multiple Class I alleles; HLA‑A*02:05 was among the most frequent in the dataset.

Spectral quality and quantitative performance: High MS2 spectral quality was maintained across the 1e5–1e8 range. Representative spectra showed consistent b/y ion series down to ~150 m/z across loads. PEAKS MS2 correlation scores were predominantly >0.7, supporting confident identifications and suitability for de novo sequencing. Quantitative linearity across loads was excellent (R2 ≈ 0.9997), and observed dynamic range extended to about seven orders of magnitude, enabling detection of peptides spanning from very low to very high cellular copy numbers.

Method enablers: The Low Input application mode plus ion pre‑accumulation increased sensitivity for single‑ion detection. AGC ensured consistent ion populations per scan across input levels, maintaining spectral quality without manual tuning. FAIMS provided gas‑phase fractionation that improved capture of singly charged peptides that are common in immunopeptidomics due to non‑basic termini.

Benefits and practical applications

• Clinical applicability: The demonstrated sensitivity supports analysis of limited clinical biopsy material and low‑abundance antigen detection relevant to personalized cancer immunotherapy.
• Neoantigen discovery: High spectral quality and de novo capability increase chances of finding noncanonical peptides and mutated neoantigens not present in databases.
• Scalability: The method tolerates a wide range of sample loadings (1e5–1e8 cell equivalents) without compromise in identification quality.
• Improved confidence: High MS2 correlation and consistent fragmentation patterns facilitate robust peptide assignment and downstream motif/allele analyses.

Future trends and applications

• Integration with DIA and hybrid acquisition strategies could further increase coverage and quantitative reproducibility at scale.
• Enhanced informatics — machine learning prediction of spectra and improved de novo algorithms — will improve identification of noncanonical and post‑translationally modified HLA peptides.
• Single‑cell and micro‑scale immunopeptidomics workflows are likely to benefit from continued sensitivity gains and optimized sample preparation.
• Broader clinical deployment will require standardized pipelines for enrichment, LC–MS acquisition and rigorous FDR control across cohorts.

Conclusion

The Orbitrap Astral Zoom MS, combined with FAIMS Pro Duo and a Vanquish Neo UHPLC, delivers high sensitivity, wide dynamic range and excellent spectral quality for immunopeptidomics. The Low Input application mode with ion pre‑accumulation and FAIMS gas‑phase fractionation enables comprehensive HLA‑I peptide coverage from very low to high sample inputs, supporting neoantigen discovery and translational immunopeptidomics workflows.

References

1. Kaabinejadian S, Barra C, Alvarez B, Yari H, Hildebrand W, Nielsen M. Accurate MHC motif deconvolution of immunopeptidomics data reveals high relevant contribution of DRB3, 4 and 5 to the total DR immunopeptidome. Front Immunol. 2022 Jan 26. DOI: 10.3389/fimmu.2022.835454