Kitted universal MAM: Automatable Sample Processing for all Stages of Biological Drugs

Significance of the topic

Multiple Attribute Monitoring (MAM) by LC-MS is an increasingly important analytical approach for comprehensive quality control of biological therapeutics. It enables direct, simultaneous monitoring of critical quality attributes (CQAs) such as sequence variants, host-cell proteins, and post-translational modifications (PTMs). Minimizing artifactual modifications introduced during sample handling (notably deamidation and oxidation) and removing interfering contaminants are essential to obtain accurate, regulatory‑grade quantitation for process development, comparability, and release testing of biologics.

Objectives and overview of the study

This work demonstrates an end-to-end, automatable sample preparation workflow for MAM based on the S-Trap sample-processing chemistry integrated with a Tecan Freedom EVO liquid-handling platform and Resolvex A200. Key goals were to: reduce artifactual PTMs introduced during digestion, remove detergents and other contaminants, deliver high reproducibility and throughput (up to 96 samples/run), and produce LC-MS data suitable for robust CQA monitoring across production stages from cell broth to final drug product.

Methodology

• Sample types: NISTmAb RM 8671 model monoclonal antibody; spike-in experiments with heavy stable‑isotope labeled SILuMAB; blank controls and calibration-peptide mixture for retention-time and quantitative checks.
• Automated processing: S-Trap chemistry executed on Tecan Freedom EVO (EVO 150) with 8-channel Air LiHa, RoMa, DiTis and integrated Resolvex A200 for plate handling. Workflow steps included reduction (TCEP), alkylation (IAA), acidification, binding to S-Trap, washes to remove detergents/excipients, digestion (trypsin), elution and drying, and reconstitution with retention-time calibration mix.
• Optimized digestion buffer (referred to as MAM buffer) was developed to limit non-enzymatic deamidation and other chemical artifacts during sample preparation.
• LC-MS acquisitions: data collected on multiple platforms (ZenoTOF, Agilent 6546, Thermo Q Exactive, Bruker timsTOF Pro) to assess sequence coverage, PTM identification and interplatform reproducibility. Data analysis used Protein Metrics Byologic and Spectronaut among other tools.
• Analytical checks: retention time stability, extracted ion chromatogram (XIC) area CVs for selected doubly charged tryptic peptides, blank adjacent wells to monitor cross-contamination, detergent removal tests with Tween-20 and PEG.

Instrumentation used

• Tecan Freedom EVO 150 with 8‑channel Air Liquid Handling Arm (Air LiHa), Robotic Manipulator Arm (RoMa), and conductive disposable tips (DiTis).
• Resolvex A200 plate centrifugation/processing module.
• S-Trap sample‑processing kits (96‑well format).
• Mass spectrometers: ZenoTOF, Agilent 6546, Thermo Q Exactive, Bruker timsTOF Pro.
• Software: Protein Metrics Byologic, Spectronaut for identification and quantitation.

Main results and discussion

• Reduction of artifactual deamidation: the optimized MAM digestion buffer reduced artificial peptide deamidation approximately 4- to 5-fold compared to a standard overnight digestion in 50 mM ammonium bicarbonate (ABC) at 37 °C. Across biological matrices (blood and plasma cells), a majority of peptides showed reduced deamidation when processed with the MAM buffer.
• Contaminant removal: S-Trap processing efficiently removed detergents (example: 1% Tween-20), PEG and other interfering species. Signal comparisons and dilution experiments indicated >10,000–20,000-fold effective reduction of detergent signal after cleanup, supporting low residual interference for LC-MS injection.
• Reproducibility and chromatography: LC runs displayed acceptable retention time shifts and good peak alignment across replicates. Extracted ion chromatogram area CVs for selected doubly charged tryptic peptides were ≤20% overall; calibration-mix peptides showed CV ≈10%, indicating LC‑MS instrumental contribution to variability.
• Proteome and PTM coverage: high sequence coverage (99–100%) of the model mAb was obtained, with expected glycosylation, disulfide mapping, oxidations and deamidations observed. Digestions were reproducible and sensitive enough to identify over 100 host‑cell proteins (HCPs) in a murine cell‑line derived mAb using timsTOF analysis.
• No cross-contamination observed: adjacent blank wells prepared and processed through the automated workflow showed no detectable carryover peptides, validating plate-based automation for high‑throughput runs.

Benefits and practical applications of the method

• Reduced sample‑prep artifacts increases confidence in PTM quantification, a regulatory necessity for biologics characterization and comparability studies.
• Automated S-Trap workflow minimizes operator errors and variability, enabling consistent high‑throughput processing (up to 96 samples per run) for process development, stability studies, biosimilar comparison and release testing.
• Effective removal of detergents, surfactants and excipients improves LC-MS sensitivity and reduces analytical failures from matrix interferences.
• Compatibility with multiple MS platforms and data-analysis pipelines allows flexible deployment across analytical labs and QC environments.

Future trends and potential applications

• Broader deployment of automated MAM workflows in regulated QC, with formal method qualification/validation for batch release and stability testing.
• Further refinement of digestion chemistries to suppress additional artifactual PTMs and to support labile modifications (e.g., phosphorylation, glycan microheterogeneity) with enhanced fidelity.
• Integration with laboratory information management systems (LIMS) and automated data processing pipelines (including AI/ML approaches) to accelerate CQA reporting and to enable real‑time process monitoring.
• Adaptation for diverse biologic formats (bispecifics, ADCs, fusion proteins) and direct coupling to upstream process analytics for in‑process control.
• Continued cross‑platform benchmarking to harmonize MAM results across MS instrument types and software tools.

Conclusion

The automated S-Trap based MAM workflow implemented on a Tecan Freedom EVO/Resolvex A200 platform provides a robust, high‑throughput approach for preparing biologic samples for LC‑MS analysis. It substantially reduces artifactual deamidation, removes detergents and other contaminants, delivers high sequence coverage and reproducible peptide quantification, and prevents cross‑contamination in plate formats. These attributes make the workflow well suited for routine application in biologics characterization, biosimilarity assessments and regulated testing environments.

References

• Tajiri‑Tsukada et al. Bioengineered. 2020. DOI: 10.1080/21655979.2020.1814683 (cited as context for deamidation concerns).
• Authors and affiliations of the study: John D. Laycock, Heather A. Eastwood, Manuel Bauer, Barry Boyes, Ben Libert, Marshall Bern, Lawrie Veale, Janet Tam, Stefan Loroch, Brett Phinney, Michael Krawitzky, John P. Wilson; organizations include Tecan, Advanced Materials Technology, Protein Metrics, Ruhr-University Bochum, UC Davis, Bruker Daltonics and ProtiFi LLC.