Streamlined Method Development for Efficient and Reliable Lipid Nanoparticle Analysis

Messenger RNA (mRNA)-based therapeutics rely on lipid nanoparticles (LNPs) for efficient intracellular delivery. Accurate analysis of LNP components is critical to ensure formulation quality. In this article, a fast, selective, and quantitative liquid chromatography (LC) method was developed for LNP analysis, using evaporative light scattering for sensitive detection of the analytes of interest. Automated screening and systematic optimization of chromatographic parameters, including mobile phase composition, gradient conditions, and column temperature, enabled precise separation and accurate quantification of components. Design space visualization further streamlined the identification of conditions meeting multiple criteria for resolution, sensitivity, and efficiency. This robust method supports the rapid development of mRNA therapeutics and advances pharmaceutical research.

Messenger RNA (mRNA)-based therapeutics have revolutionized modern medicine, offering groundbreaking solutions for treating infectious diseases, genetic disorders, and cancer. By instructing cells to produce specific proteins, mRNA therapies enable targeted disease intervention at the molecular level. However, mRNA’s inherent instability and vulnerability to enzymatic degradation present significant delivery challenges. Lipid nanoparticles (LNPs) have emerged as the most effective delivery system for mRNA therapeutics, providing structural stability, efficient intracellular transport, and controlled release (1,2).

LNPs are composed of four key lipid components: cholesterol, ionizable lipids, PEG-modified lipids, and neutral phospholipids (see Figure 1). Each component plays a distinct role in stabilizing the nanoparticle structure and ensuring effective delivery. Accurate quantification of these components is critical to ensure the safety, efficacy, and consistency of LNP formulations.

This article introduces a liquid chromatography (LC) method for LNP analysis, integrating automated screening, systematic optimization, and evaporative light scattering detection (ELSD), thereby addressing the challenge of lipid components lacking UV absorbance. By optimizing mobile phase composition, gradient conditions, and column parameters, this method delivers precise separation, accurate quantification, and robust analytical performance, supporting the future development of mRNA therapeutics (3).

Method Development

Automated Mobile Phase and Column Screening

Efficient separation of LNP components requires careful selection of mobile and stationary phases to investigate retention behavior and resolution. The initial screening experiments were designed to evaluate the impact of mobile phase composition and column type on the separation of cholesterol, ionizable lipids (heptadecan-9-yl 8-((2-hydroxyethyl)(6-oxo-6-(undecyloxy)hexyl)amino)octanoate [SM-102]), PEG-modified lipids (1,2-dimyristoyl-sn-glycero-3-methoxypolyethylene glycol), and neutral phospholipids (distearoylphosphatidylcholine [DSPC]).

The screening was performed using a Nexera Method Scouting UHPLC system (Shimadzu) equipped with an ELSD-LTIII detector and Shim-pack Scepter Phenyl-120, C8-120, and C4-300 separation columns (all Shimadzu, all 100 mm x 3.0 mm, 1.9-μm) to test a variety of selectivities. The mobile phases consisted of 0.1% formic acid in water as mobile phase A and varying mixtures of acetonitrile (ACN) and methanol (MeOH) as mobile phase B. In a generic gradient (60–95%B in 5 min) at a flow rate of 0.6 mL/min, MeOH ratios were systematically varied from 0% to 90% in 10% increments, resulting in 10 distinct mobile phase compositions. The use of dedicated method development software (LabSolutions MD, Shimadzu) allowed for automated method and batch generation for the screening experiment, including 30 different analytical conditions (10 mobile phases × 3 columns). Switching valves enabled alternation between columns and mobile phases, eliminating manual intervention and reducing potential errors. Mobile phase B composition was adjusted using an automated blending function, ensuring accurate solvent generation while minimizing manual preparation.

The chromatograms were analyzed for key parameters, including retention time, peak shape, and resolution. Special attention was given to the separation of SM-102 and DMG-PEG, as the critical peak pair.

As can be seen in Figure 2a–c, the screening results revealed significant differences in separation performance across the tested columns and mobile phases. The phenyl-120 column demonstrated superior separation when using MeOH ratios of 60%, 70%, and 80%, producing sharp peaks and minimal tailing for SM-102 and DMG-PEG. In contrast, the C8-120 column showed poor peak shapes for DMG-PEG across all MeOH ratios, while the C4-300 column exhibited tailing in SM-102 peaks under similar conditions.

Based on these findings, the phenyl-120 column with 60%, 70%, and 80% MeOH in mobile phase B was selected for further optimization.

Evaluation of IPA Addition to Mobile Phase B

In a follow-up experiment, the effect of isopropanol (IPA) addition to the organic solvent was evaluated to further enhance separation performance. Using the phenyl-120 column and MeOH–ACN ratios of 60:40, 70:30, and 80:20, IPA was added in 10% increments from 0% to 50%, generating 18 different mobile phase compositions. Chromatograms were analyzed, and separation performance was quantified using an "evaluation value," calculated as the product of the number of peaks and the sum of resolution factors.

Figure 3 demonstrates how IPA addition influenced retention and selectivity of the four analytes of interest. The optimal composition, ACN–MeOH–IPA = 27:63:10, provided the best balance of resolution between cholesterol and SM-102, peak height for SM-102, and retention time for DSPC in all the analytical conditions tested, as can be seen in Figure 4.

Optimization of Separation Parameters

Building on these screening results, further optimization of the separation parameters was carried out to improve separation performance and ensure robustness. In this next step, key parameters such as initial gradient concentration, gradient time, and column oven temperature were systematically varied to explore their influence on resolution, peak width, and retention behavior.

Three levels of initial gradient concentration (50%, 60%, and 70%), gradient time (4, 5, 6 min), and column oven temperature (30 °C, 40 °C, 50 °C) were evaluated using the software. To streamline the search for optimal separation conditions, the software allowed for visualization of a design space. This approach mapped resolution, peak height, and retention time across the tested parameter ranges, color-coding the method performance for each parameter. Higher initial gradient concentrations and longer gradient times improved resolution, while higher column oven temperatures reduced peak widths. An overlay of the different design spaces highlighted the optimal method conditions: an initial gradient concentration of 70%, gradient time of 4 min, and column oven temperature of 50 °C. These conditions provided resolution ≥1.5 between cholesterol and SM-102, peak height ≥80,000 for SM-102, and elution time ≤6 min for DSPC (Figure 5).

Results and Discussion

Sample Analysis

The optimized conditions were applied to the analysis of an mRNA-LNP formulation. The sample was diluted 1:10 in ethanol prior to analysis. Calibration curves for each lipid were prepared using standard solutions, with concentration ranges tailored to the expected levels in the sample. For cholesterol, concentrations ranged from 6–30 mg/L, 20–100 mg/L for SM-102, 9–45 mg/L for DMG-PEG, and 7–35 mg/L for DSPC. Calibration curves for all lipids demonstrated excellent linearity, with coefficients of determination (r²) exceeding 0.998. Repeatability was confirmed with relative standard deviations (RSDs) for retention time below 0.13% and peak area RSDs below 3.82% across seven replicate analyses. The quantified concentrations of the lipids were consistent with expected values: cholesterol (24.3 mg/L), SM-102 (54.6 mg/L), DMG-PEG (16.6 mg/L), and DSPC (17.5 mg/L). Sample chromatograms, as shown in Figure 6, revealed well-separated peaks for all components, with resolution, peak height, and retention times meeting the predefined criteria.

Conclusion

A robust and precise LC–ELSD method for quantifying LNP components was developed, enabling reliable characterization of mRNA-LNP formulations for pharmaceutical applications. The use of dedicated method development software was instrumental, automating screening, parameter optimization, and design space visualization to achieve efficient and accurate separation. This systematic approach streamlines workflows, driving innovation in pharmaceutical research and accelerating the development of mRNA therapeutics.

Acknowledgment

We gratefully acknowledge Dr. Naohiro Tsuyama from Hiroshima University PSI GMP Center for kindly providing the sample.

References

(1) Hou, X.; Zaks, T.; Langer, R.; Dong, Y. Lipid Nanoparticles for mRNA Delivery. Nat. Rev. Mater. 2021, 6 (12), 1078–1094. DOI: 10.1038/s41578-021-00358-0

(2) Sahin, U.; Karikó, K.; Türeci, Ö. mRNA-Based Therapeutics—Developing a New Class of Drugs. Nat. Rev. Drug Discov. 2021, 20 (8), 573–574. DOI: 10.1038/nrd4278

(3) Fujisaki, S. Efficient Method Development of Lipid Nanoparticle Components Using ELSD, Shimadzu Corporation Application News 01-00927-EN, 2015.