Beyond Particle Size: Building a Complete Analytical Strategy for LNPs

Key Takeaways
Average particle size, polydispersity, and bulk encapsulation efficiency are essential controls, but they may conceal differences in particle populations and payload distribution.
Lipid and payload assays must account for degradants, matrix effects, deformulation, and sample-preparation conditions that can change recovery or interpretation.
Impurity testing must follow the assembled product because interactions between lipids and mRNA can create species that methods developed for naked mRNA may not detect.
Orthogonal, phase-appropriate methods are most informative when physicochemical and stability-indicating results can be interpreted against functional.

An Analytical Problem with Multiple Layers

Average particle size and encapsulation efficiency provide essential information about a lipid nanoparticle (LNP) product, but they do not describe payload distribution, internal organization, degradation profile, or functional performance.^1–3 Each measurement captures only part of a product whose quality depends on interactions among a fragile messenger RNA (mRNA) payload, multiple lipid components, and the structures they form together.

Regulatory and industry frameworks reflect this complexity. The World Health Organization's (WHO) regulatory considerations for mRNA vaccines, the European Medicines Agency's (EMA) draft quality guideline, and the BioPhorum critical quality attribute framework identify numerous attributes across the mRNA, lipid components, finished nanoparticle, impurities, and biological activity.^4–6 Although vaccine-focused guidance should not be applied indiscriminately to every LNP therapeutic, these documents illustrate the breadth of the analytical task.

Three tensions shape that task. Bulk measurements may conceal particle-level heterogeneity. Routine quality control methods may confirm consistency without explaining internal structure. Physicochemical results may remain within specification even when functional performance changes. The challenge is to determine what each assay reveals, what it cannot resolve, and how its result contributes to an integrated account of product quality.

Particle Size Is Not a Single Answer

Particle size and size distribution are among the most familiar LNP measurements. They support formulation development, process monitoring, comparability, and stability assessment, and they can reveal aggregation or shifts in the particle population. However, a mean diameter is a population statistic, not a complete physical description.

Dynamic light scattering (DLS) commonly provides hydrodynamic size and polydispersity measurements. Its speed and accessibility make it useful for routine analysis, but its output may obscure minority populations and does not show how RNA is distributed among particles.^7,8 Benchmark studies using orthogonal techniques have demonstrated that apparently similar LNP formulations can yield different information depending on the measurement principle employed.

Analytical ultracentrifugation offers an additional view of particle populations based on their sedimentation behavior. Its application to mRNA–LNP quality assessment demonstrates how a separation-based method can complement conventional particle-sizing measurements when a formulation contains distinct populations or when an average size does not adequately explain a change in the product.⁹

The relevant analytical questions extend beyond the mean diameter. Analysts may need to determine the breadth of the distribution, the presence of aggregates or other subpopulations, and whether changes in size accompany changes in loading or stability. A routine sizing method can provide a robust control measurement without capturing every dimension of heterogeneity. Recognizing that boundary prevents a precise numerical result from being given more interpretive weight than it can support.

Morphology Adds Information That Size Cannot Capture

Particles with similar hydrodynamic diameters may differ in shape, internal organization, structural uniformity, or the arrangement of lipid and RNA domains. Morphological methods address these features directly rather than inferring them from the behavior of particles in suspension.

Cryogenic electron microscopy (cryo-EM) can visualize individual particles in a near-native frozen state and reveal structural features that an ensemble size measurement averages together. Studies comparing LNP characterization methods have used imaging alongside other techniques to examine particle shape and structural heterogeneity.^7,8 Such analysis may help determine whether a formulation or process change has altered particle architecture even when average size remains comparable.

More specialized methods can probe molecular organization within the particle. Cryogenic Orbitrap secondary ion mass spectrometry has been used to study molecular orientation and stratification in RNA–LNPs, providing a way to examine the spatial arrangement of components within intact nanoparticles.³ This type of structural analysis is distinct from verifying that the intended lipids are present at the correct overall ratios.

Advanced imaging and molecular mapping may not be practical or necessary for every lot-release program. Their value is greatest when they answer development questions, support formulation selection or comparability, investigate unexpected changes, or establish which structural features should be controlled by more routine methods.

Encapsulation Efficiency Does Not Describe Payload Distribution

Encapsulation efficiency is another essential measurement that can create a misleading sense of completeness. Assays commonly compare accessible RNA with total RNA measured after disruption of the nanoparticle. The resulting percentage estimates how much of the measured RNA is protected within the formulation rather than freely accessible outside it.

That calculation says little about how the encapsulated RNA is distributed across the particle population. Research into the payload distribution and capacity of mRNA–LNPs has demonstrated that bulk measurements can conceal differences among particles, including variation in the amount of RNA associated with individual nanoparticles.¹ A formulation with high overall encapsulation could therefore include a mixture of empty, lightly loaded, and more heavily loaded particles.

This distinction may have functional significance. The transfection potency of mRNA-containing LNPs depended on relative loading levels in formulations evaluated by Liao and colleagues.² The result does not establish one universally optimal loading distribution, but it shows why encapsulation percentage and total RNA concentration cannot answer every question about delivery performance.

A more complete characterization strategy may distinguish among total RNA content, free RNA, encapsulation efficiency, and payload distribution. During early development or an investigation, particle-level methods may explain differences that a bulk encapsulation result leaves unresolved. For routine control, development data should establish what the selected measurement can reliably indicate about product consistency.

Lipid Composition Must Be Measured as Both Formula and Structure

Lipid analysis begins with several separate questions: whether the intended components are present, whether their amounts and ratios are correct, whether impurities or degradants have formed, and whether the components are organized as expected within the particle. A method that answers one of these questions does not necessarily answer the others.

A reversed-phase chromatographic method with charged-aerosol detection developed by Kinsey and colleagues measured individual parent lipids and associated impurities and degradants.¹⁰ The method was validated for linearity, accuracy, precision, and specificity in support of process and formulation development. The study also showed that sample-preparation diluent and mobile-phase pH were critical variables for the lipids investigated.

These details matter because the lipid classes in an LNP differ chemically, and their degradants may create additional separation and detection challenges. An assay must demonstrate that it can distinguish parent components from relevant impurities rather than simply produce a total lipid signal.

Composition also differs from structure. A chromatographic method may confirm that the intended lipid ratios are present, while imaging or molecular mapping reveals how those lipids are arranged and associated with the payload. Cryogenic mass spectrometry work reported by Kotowska and colleagues provides an example of examining spatial organization beyond bulk composition.³

Quantitative lipid methods support identity, content, impurity monitoring, stability, and routine control. Structural techniques can help explain how formulation or process changes affect particle architecture. Structural characterization may explain performance differences between lots whose bulk lipid composition appears comparable.

Payload Integrity Cannot Be Separated from Lipid Quality

Payload quality encompasses more than the percentage of apparently full-length mRNA. Relevant attributes include identity, concentration, integrity, capping efficiency, poly(A)-tail characteristics, and the presence of fragments, untailed species, or alternative conformational forms. Encapsulation complicates these analyses because the mRNA must either be extracted from the LNP or measured under conditions that disrupt or deformulate the particle.

Sample preparation is therefore part of the measurement. A systematic evaluation using a formulated mRNA–LNP sample found that denaturant type and concentration, as well as LNP-disruption protocols, could interfere with accurate integrity analysis by capillary gel electrophoresis with laser-induced fluorescence and ion-pair reversed-phase liquid chromatography.¹¹ In that system, optimized electrophoretic conditions included isopropanol precipitation, high urea concentrations, exclusion of formamide from the sample diluent, and a high dye concentration. Those conditions belong to the method studied rather than constituting a universal preparation protocol, but the work demonstrates how the analytical workflow can alter recovery, separation, and interpretation.

Different separation principles also produce different analytical views. Lardellier and colleagues compared a fluorescence-detected fragment analyzer, capillary gel electrophoresis with ultraviolet detection, and ion-pair reversed-phase liquid chromatography for assessing the integrity of mRNA drug substance and drug product.¹² All three monitored integrity, but the chromatographic method provided better resolution, while the automated fragment analyzer supported faster analysis of larger sample sets. Capillary gel electrophoresis with ultraviolet detection also revealed mRNA oligomers that required further characterization.

Integrity by length remains only one part of payload quality. Regulatory and industry frameworks identify the 5' cap and poly(A) tail as separate mRNA attributes requiring appropriate characterization. A DNAzyme-based method reported by Wang and colleagues generated 5' and 3' fragments in a single digestion, enabling concurrent analysis of capping efficiency and poly(A)-tail length.¹³ Polyacrylamide gel electrophoresis and ion-pair reversed-phase liquid chromatography confirmed production of the cleavage fragments, and liquid chromatography-mass spectrometry validated the findings.

The choice of integrity method therefore affects more than convenience or throughput. Extraction, denaturation, separation, and detection determine which species contribute to the reported result. Method development must establish that the assay measures the intended attribute in the formulated product rather than assuming that performance with naked mRNA will transfer unchanged to the LNP matrix.

Impurity Testing Must Follow the Whole Product

The impurity profile of an mRNA–LNP product can originate in RNA synthesis and purification, lipid raw materials, formulation, storage, or interactions between the payload and delivery system. Relevant species may include RNA fragments, untailed RNA, free RNA, residual process materials, lipid impurities, lipid degradants, aggregates, and covalent mRNA-lipid adducts.

Some impurities emerge only after the components have been combined. Packer and colleagues described the formation of an mRNA-lipid adduct as a mechanism associated with loss of mRNA activity in LNP delivery systems.¹⁴ In the analytical comparison conducted by Lardellier and colleagues, ion-pair reversed-phase liquid chromatography detected mRNA-lipid adducts that the capillary electrophoresis methods did not resolve.¹² An assay panel developed primarily around naked mRNA could therefore overlook species created within the formulated product.

Detection of an additional peak does not, however, establish that the species is harmful. The mRNA oligomers detected by capillary gel electrophoresis in the Lardellier study were isolated and evaluated in vitro. The oligomeric material expressed the target protein similarly to monomeric mRNA, although the authors noted that further study would be needed to establish equivalence in other biological contexts, including potential in vivo differences.

Impurity evaluation consequently requires identification and, where appropriate, investigation of functional relevance. Peak area alone does not establish mechanism or risk. Orthogonal separation, structural or mass analysis, forced-degradation studies, and functional assays can help determine whether a detected species represents degradation, an alternative conformation, an analytical artifact, or a change with biological consequences.

Functional Performance Closes the Analytical Loop

Physicochemical methods establish essential information about identity, composition, integrity, consistency, and stability. They do not independently demonstrate that an LNP delivers intact mRNA and supports expression of the encoded protein. Functional assays provide that complementary evidence.

The connection between analytical attributes and function becomes especially important in stability studies. Kamiya and colleagues evaluated the effects of temperature, cryoprotectants, vibration, light exposure, and syringe aspiration on the physicochemical properties and protein-expression ability of a benchmark mRNA-LNP formulation.¹⁵ Storage at -80 °C without cryoprotectant reduced protein expression in that system and was associated with particle aggregation. Vibration and light also produced changes under some of the tested conditions. Because the study used a particular formulation, the findings support product-specific stress testing rather than universal handling limits.

A three-month heat-stress study by Tong and colleagues examined mRNA degradation, particle size, encapsulation efficiency, and in vitro cell potency, then evaluated correlations between the stability-indicating measurements and functional activity.¹⁶ The relationship between mRNA degradation and potency contained two regions, which the authors interpreted as indicating a critical cut-off associated with mRNA degradation in the product studied. Temperature-dependent behavior also appeared in the relationship between LNP size and potency.

These studies illustrate why stability assessment should cover the payload, particle, and biological response. A decline in potency may occur alongside mRNA degradation, changes in particle size, altered loading, chemical modification, or several changes at once. Correlation does not establish causation by itself, but it can identify measurements that warrant further investigation and may prove useful for monitoring a specific formulation. The potency method, in turn, must respond reliably to product changes relevant to the formulation if it is to serve as a functional anchor for the physicochemical program.

Building an Orthogonal and Phase-Appropriate Analytical Strategy

An effective analytical strategy assigns each method a defined purpose. Routine controls may include particle size and distribution, RNA content, encapsulation efficiency, lipid identity and content, mRNA integrity, selected impurity measurements, and a functional potency or expression assay. Extended characterization may examine morphology, internal structure, particle-level payload distribution, lipid stratification, higher-resolution degradant identification, and relationships among physical, chemical, and functional attributes.

Orthogonality depends on informational value rather than assay count. Two methods based on different principles may corroborate the same attribute, reveal different subpopulations, or expose method-specific bias. The studies of formulated mRNA integrity show how sample preparation, separation mechanism, detection format, resolution, and throughput can change the information obtained from assays nominally measuring the same attribute. Similar reasoning applies to particle sizing, imaging, lipid analysis, and potency.

The analytical package should also evolve with development. Early studies need methods that discriminate among formulations and help identify mechanisms of instability or loss of activity. Process development requires measurements that connect material attributes and process parameters with product quality. Later development places greater emphasis on validated, stability-indicating methods suitable for specification setting, release, and continued control. Comparability exercises and investigations may draw again on advanced structural methods that are unnecessary for routine testing.

Regulatory guidance and published critical quality attribute frameworks can define the territory, but they cannot select the final method panel for a specific product. That selection must reflect the formulation, payload, manufacturing process, stage of development, known degradation pathways, and relationship between measured attributes and function.

From Measured Attributes to Product Understanding

Analytical maturity does not require every available method for every batch. It requires enough complementary evidence to explain what the product is, how consistently it is made, how it changes, and whether those changes affect performance.

For LNP products, that means connecting particle measurements with population structure, encapsulation with payload distribution, lipid content with degradation and organization, and mRNA integrity with the effects of formulation and sample preparation. Stability-indicating results become most useful when they can be interpreted against functional activity.

The strongest analytical package is defined by the coherence of the evidence it produces. Its methods should convert a collection of measurements into a defensible model of product quality and performance.