Optimization of Universal Detection Methods for Lipid Nanoparticles

Importance of the topic

Lipid nanoparticles (LNPs) are a central delivery platform for nucleic acid therapeutics, including mRNA vaccines. Accurate quantitation of the four principal lipid components—ionizable lipid (SM-102 in this study), phospholipid (DSPC), PEG-lipid (DSPE‑PEG2000) and cholesterol—is essential for controlling product potency, stability and regulatory compliance. Because these lipids lack chromophores, universal detectors such as evaporative light scattering (ELSD/ELS) and charged aerosol detection (CAD) are commonly used; optimizing and comparing these detectors is important for routine LNP analysis in QC and research laboratories.

Objectives and overview of the study

The study aimed to develop and optimize an LC method for typical mRNA LNP formulations and to directly compare ELSD and CAD for sensitivity (LOD/LOQ), linearity and precision. Two representative LNP sample formulations (spiked model mixtures) and multi-level standards (LOQ to 300 µg/g) containing cholesterol, DSPC and DSPE‑PEG2000 were measured with equivalent chromatographic conditions to evaluate detector performance and determine best practice for quantitation of LNP lipids.

Methodology and instrumentation

• Chromatographic system: ACQUITY Premier UPLC System.
• Column: CORTECS Phenyl, 90 Å, 1.6 µm, 2.1 x 50 mm (p/n 186008379).
• Mobile phases: A = 10 mM ammonium acetate in 90/10 methanol/water; B = 10 mM ammonium acetate in 90/10 acetonitrile/water.
• Flow rate: 0.400 mL/min; injection volume: 5 µL; sample temperature: 12 °C; column temperature: 30 °C.
• Detectors compared: evaporative light scattering detection (ELSD/ELS) and charged aerosol detection (CAD). Data acquisition used ~2 Hz where applicable.
• Standards and samples: up to eight standard levels from LOQ to 300 µg/g; two spiked LNP sample formulations (Sample 1 ≈ 50:38.5:10:1.5 by composition; Sample 2 ≈ 49.5:38:9.5:3; total concentration basis 1000 µg/g in examples).
• Calibration approach: log–log linear curves (recommended for ELSD/CAD) due to limited linear dynamic ranges.

Main results and discussion

• Sensitivity and LOQ: CAD demonstrated superior sensitivity versus ELSD. Determined LOQs were:
  • ELSD: cholesterol = 10 µg/g, DSPC = 10 µg/g, DSPE‑PEG2000 = 15 µg/g.
  • CAD: cholesterol = 5 µg/g, DSPC = 5 µg/g, DSPE‑PEG2000 = 10 µg/g.
• Linearity: Calibration curves (LOQ to 300 µg/g) were fitted using log–log linear regression. Cholesterol exhibited comparable R² values between detectors, while DSPC and DSPE‑PEG2000 showed improved R² on CAD versus ELSD, indicating a more consistent response for these species with CAD at low-to-mid concentration ranges.
• Precision: Area %RSD for LOQ standards and for two spiked samples were consistently lower on CAD than on ELSD. In LOQ assessments, no CAD peak exceeded ~2.3% RSD, and sample repeatability was generally better with CAD, supporting improved quantitative precision.
• Detection physics and implications: CAD measures charge-related aerosol response that tends to be less dependent on particle-size effects than light scattering efficiency in ELSD, which helps explain the improved consistency and sensitivity of CAD at low concentration levels for certain lipids.
• Practical chromatographic performance: Cholesterol produced the largest absolute detector signal among the tested lipids; relative detector performance differences were most notable for DSPC and DSPE‑PEG2000.

Benefits and practical applications of the method

• CAD-based quantitation provides improved sensitivity and precision for LNP lipid components, making it preferable for assays that must measure low-level lipids or require tight repeatability (QC release, stability studies, formulation screening).
• The optimized UPLC method and column selection enable fast separations compatible with universal detectors, supporting higher throughput in routine workflows.
• Log–log calibration is recommended for ELSD and CAD because of limited linear dynamic ranges; this supports accurate quantitation across the relevant concentration span for formulation and batch testing.

Instrumentation

• UPLC system: ACQUITY Premier UPLC (Waters).
• Detectors: Evaporative light scattering detector (ELSD/ELS) and Charged aerosol detector (CAD).
• Analytical column: CORTECS Phenyl, 90 Å, 1.6 µm, 2.1 x 50 mm.
• Consumables and mobile phases: 10 mM ammonium acetate buffers in methanol/water and acetonitrile/water (90/10 compositions) were used as mobile phases A and B.

Future trends and potential uses

• Detector evolution: Continued refinement of aerosol and light-scattering detectors will further reduce size- and chemistry-dependent biases; CAD improvements are likely to enhance low-level quantitation for diverse lipid chemistries.
• Standardization and method transfer: Development of harmonized CAD-based LNP assays and inter-laboratory studies will support regulatory acceptance and method transfer across industry labs.
• Coupling to orthogonal techniques: Combining universal detection with mass-spectrometric or evaporative-based orthogonal approaches could improve lipid identification, impurity profiling and aggregate/oligomer assessments.
• Automation and high-throughput workflows: Faster columns and optimized gradients paired with CAD can accelerate formulation screening and QC release testing for large-scale LNP manufacturing.

Conclusion

For quantitative analysis of LNP components without chromophores, universal detectors are essential. This study demonstrates that CAD outperforms ELSD in sensitivity, linearity for certain lipids and precision, especially at low concentrations. CAD is therefore recommended as the preferred detector for routine quantitation of LNP lipid constituents when higher sensitivity and repeatability are required. The use of log–log calibration and the described UPLC conditions provides a practical, robust workflow for laboratories analyzing LNP formulations.

References

• Steere A, Wong N, Simeone J, Hong P. Optimization of Universal Detection Methods for Lipid Nanoparticles. Waters Corporation; 2026. Poster 720009299EN.