Automation of PTM Analysis: Highlights From ASMS 2018

Protein therapeutics are often complex large molecules. However, they are usually not uniform in structure since post-translational modification (PTM) produces a range of products by adding phosphate, sulfate, glycans, ubiquitin, etc. For example, ubiquitin tags the protein for destruction, usually by hydrolysis. There are 200 protein therapeutics recorded.

Drug regulators expect that developers of biotherapeutics will offer several critical quality attributes and map these to the structure of the active pharmaceutical ingredient (API).

A poster at ASMS 2018 by Ben Niu and colleagues at Medimmune (Gaithersburg, MD) reported an automated workflow for accelerated analysis of PTMs of proteins during lead development (Niu, B.; Skilton, S. et al. “A Vendor-Neutral MAM Workflow for Accelerated PTMs Profiling Analysis,” Poster WP 718, ASMS 2018).

The method is useful for developing a quality target product profile (QTPP) and associated control strategy. Automation of the QTPP process is desirable since manual methods are slow, repetitive, tedious, and subject to human variances.

To overcome these difficulties, the authors developed an automated workflow for multiattribute monitoring (MAM) of large data sets recorded from dozens of samples exploring for dozens of CQAs. This workflow is consistent with quality-by-design guidance.

The first step is a proteolytic digestion of the protein. This can be automated or manual sample prep. LC/ESI/MS/MS analysis is next. When formulation and accelerated stability were the issue, the stressed and control samples were run back-to-back. Results were batch-processed to improve validity of the comparison.

A typical result was a table that mapped the particular peptide to occupancy of the target PTM at the site. Listed PTMs included oxidation, high pH, N-succinimide, deamidation, lysine cleavage, and glycosylation. Collection of the data in a concise, consistent format facilitates focusing on the unexpected. This can be examined for correlation or inconsistency with the molecular attributes.

The collected data is examined with a batcher and Byonic viewer (Protein Metrics, San Carlos, CA) to prepare quantitative data for preliminary review. After passing preview and review, the report is prepared in a PDF format.

The authors conclude that automated MAM analysis provides consistent results in a few hours, compared to a similar number of days using reference legacy (manual) technology. Reducing human involvement improves analysis variance while maintaining a low false discovery rate (FDR). Importantly, the automated data processing improves data quality since traceability is easier. This approach will be expanded to other parts of the drug development and registration process.

The authors expect to add new peak detection (NPD), sequence variants, and characterization of host-cell proteins.

Robert L. Stevenson, Ph.D., is Editor Emeritus, American Laboratory/Labcompare; e-mail: [email protected]

Related Products

Comments