SOFTWARE TOOLS FOR AUTOMATED LC/MS ANALYSIS OF CRITICAL QUALITY ATTRIBUTES OF mRNA MOLECULES

Significance of the Topic

The analytical characterization of messenger RNA (mRNA) critical quality attributes (CQAs) is essential for the development, quality control and regulatory release of mRNA therapeutics and vaccines. Reliable measurement of 5 capping efficiency, poly(A) tail length/heterogeneity and sequence integrity is required to ensure product potency, stability and safety. Integrating high-resolution LC-MS with tailored enzymatic digestions and specialized informatics streamlines comprehensive CQA assessment in a single workflow, reducing time-to-result and minimizing manual interpretation variability.

Goals and Study Overview

This study demonstrates an integrated LC-MS workflow for automated assessment of three mRNA CQAs: (1) 5 capping efficiency and cap isoforms; (2) poly(A) tail average length and heterogeneity; (3) sequence integrity (mapping). The approach combines multi-enzyme digestions, high-resolution data-independent acquisition (MSE) on a Xevo MRT Q-Tof instrument and dedicated software tools (SYNTHETIC Library, MAP Sequence, INTACT Mass) to provide near-complete sequence coverage and automated assignment of digestion products and isomers.

Methodology and Workflow

- Sample: custom in vitro transcribed firefly luciferase (Fluc) mRNA, Cap-1, ~1,919 nt, with poly(A) tail.
- Enzymatic digestion: four endonucleases were used (RNase T1, hRNase4, RapiZyme MC1, RapiZyme Cusativin) with specified cleavage specificities to generate overlapping oligonucleotide fragments. Digestions were performed after thermal denaturation, with enzyme-specific conditions (e.g., RNase T1: 37C, 15 min; hRNase4: 37C, 90 min).
- LC-MS acquisition: UHPLC separations (ACQUITY Premier OST column, 2.1 x 150 mm, 1.7 m, 300 ) at 60C, 0.3 mL/min, gradient 0 50% B over 90 min; 5 L injections. Mobile phases: 10 mM n-dipropylamine (DPA) and 40 mM HFIP in water (A, pH 8.6) and same buffer in 50% MeOH (B).
- Mass spectrometry: Xevo MRT Q-Tof operated in MSE (DIA) mode, 0.5 s scans, m/z 400 4000, low-energy CE 6 V, high-energy CE ramp 30 55 V. Reported MS resolution approximated ~100,000 and mass accuracy typically sub-ppm for many measurements; poly(A) species measured with <10 ppm accuracy.
- Informatics: waters_connect platform with SYNTHETIC Library App (sequence and modifications database), MAP Sequence App (automatic precursor and fragment assignment, isomer discrimination using elevated-energy fragments; processing tolerances e.g., 5 ppm), and INTACT Mass App (deconvolution and poly(A) profiling). Combined results from multiple digests were merged to maximize sequence coverage.

Used Instrumentation

- UHPLC: ACQUITY Premier Binary System with TUV detector (260 nm), ACQUITY Premier OST Column (1.7 m, 300 , 2.1 x 150 mm).
- Mass spectrometer: Xevo MRT (multi-reflecting time-of-flight) Q-Tof.
- Consumables: QuanRecovery MaxPeak 300 L vials; enzymes: RNase T1 (Aspergillus oryzae), hRNase4 (NEB), RapiZyme MC1 and Cusativin (Waters RapiZyme series).
- Software: waters_connect Informatics Platform v4.1.0.17; SYNTHETIC Library App v2.0.0; MAP Sequence App v2.0.0; INTACT Mass App v1.9.

Main Results and Discussion

- Capping efficiency: The Fluc mRNA sample was found to be 100% capped (no cap truncation impurities detected). Two Cap-1 isoforms were observed: m7GpppAm (62.7%) and m7GpppGm (37.3%). The co-occurrence of both methylated Am and Gm Cap-1 species on the same sample is noted as unexpected for a homogeneous IVT product and highlights the assays sensitivity to cap heterogeneity.
- Poly(A) tail: INTACT Mass deconvolution detected ten polyadenosine species spanning 98-mer to 107-mer (reported species C(A)97 to C(A)106). Weighted by relative abundances, the average poly(A) length was 102.3 nt. The most abundant species was the 102-mer. Mass accuracy for poly(A) species was reported <10 ppm and the most abundant charge state showed isotopic resolution (e.g., [M-34H]34).
- Sequence integrity / coverage: Combining data from four enzymatic digests produced near-complete unambiguous sequence coverage of 93.7% (assuming up to four missed cleavages in searches). MAP Sequence App used both precursor and elevated-energy fragmentation ions to assign digestion products automatically and to distinguish structural isomers. Example: a 12-mer isomer pair was automatically confirmed/rejected based on fragment ion coverage criteria (100% vs 50%).
- Sensitivity and accuracy: The Xevo MRT Q-Tof MSE workflow provided excellent mass resolution and sensitivity, enabling detection of low-abundance cap isoforms and poly(A) variants and sub-ppm typical mass accuracy for precursors and fragments in many cases.

Benefits and Practical Applications

- Single-assay multiplexing: A single LC-MS run (same LC-MS method and MSE acquisition) provided data to assess capping, poly(A) tail distribution and sequence mapping, consolidating workflows and saving time and material.
- Automated assignment and isomer resolution: The combination of high-energy fragmentation data and MAP Sequence algorithms reduced manual interpretation; structural isomers and isobaric digestion products were differentiated algorithmically according to user-specified acceptance criteria.
- Multi-enzyme strategy: Using complementary RNases increased unique and unambiguous sequence coverage, reducing reliance on targeted MS/MS for ambiguous regions and enhancing confidence in sequence integrity assessments.
- Fit-for-purpose QC: The workflow is directly applicable to development and QC laboratories for characterization of mRNA therapeutics, enabling monitoring of manufacturing quality attributes and lot release testing.

Future Trends and Potential Applications

- Increased automation and throughput: Further integration of sample handling, automated digestion optimization and batch processing in informatics will accelerate routine QC adoption.
- AI-assisted interpretation: Machine learning models could enhance low-abundance feature assignment, predict likely modification patterns and prioritize ambiguous spectra for manual review.
- Broader CQA integration: Combining this bottom-up oligonucleotide mapping with intact mass, top-down sequencing and orthogonal approaches (e.g., cap-specific enzymatic assays, NGS for sequence confirmation) will strengthen product characterization packages.
- Improved enzymes and chemistries: Development of additional sequence-specific or modification-tolerant RNases will improve coverage of difficult regions and modified nucleotides.
- Regulatory alignment: Standardized informatics reporting and audit-ready workflows will enable smoother translation into GMP/QC environments and regulatory submissions.

Conclusion

The study demonstrates a robust, high-resolution LC-MS(MSE) workflow with multi-enzyme digestion and dedicated informatics to evaluate three key mRNA CQAs in a single analytical pipeline. The combination of the Xevo MRT Q-Tof instrument and SYNTHETIC Library / MAP Sequence / INTACT Mass applications delivered near-complete sequence coverage (93.7%), precise poly(A) profiling (average 102.3 nt) and sensitive detection of cap isoforms (100% capping; 62.7% m7GpppAm, 37.3% m7GpppGm). This platform offers a practical and automatable solution for mRNA characterization in research and quality environments, while future developments in automation, AI and enzyme chemistries will further increase throughput and analytical confidence.

References

1. Waters application note P/N 720008539EN, Tunable Digestion of RNA Using RapiZyme RNases to Confirm Sequence and Map Modifications, 2024.
2. Waters application note P/N 720008553EN, RNA Digestion Product Mapping Using an Integrated UPLC-MS and Informatics Workflow, 2024.
3. Waters application note P/N 720009171EN, Sequence Mapping of mRNA Digests Using the Xevo MRT Mass Spectrometer and waters_connect MAP Sequence 2.0 Application, 2025.