Researchers have established a direct analytical correlation between small-angle X-ray scattering profiles and cryogenic electron microscopy imaging to evaluate the structural properties of liposomes and lipid nanoparticles used in vaccine formulations. Detailed in a study published in AAPS Open, the cross-platform analytical approach demonstrates that laboratory-based small-angle X-ray scattering measurements can rapidly assess critical quality attributes, such as bilayer lamellarity in liposomes and nanoscale morphology in lipid nanoparticles. Advanced biophysical characterization techniques are essential during vaccine candidate development to ensure safety, quality, and efficacy while accelerating development timelines. While cryogenic electron microscopy provides high-resolution, direct visualization of individual nanoparticles preserved in a vitrified, near-native state, the technique suffers from operational constraints, including expensive instrumentation, labor-intensive sample preparation, low sample throughput, and localized sampling limitations that hinder routine monitoring across numerous formulation batches.
To overcome these analytical bottlenecks in biopharmaceutical development, the research team correlated small-angle X-ray scattering with cryogenic electron microscopy to validate X-ray scattering as a rapid, higher-throughput screening tool for vaccine formulation and process development. The investigation evaluated two vital drug delivery architectures: phospholipid liposomes frequently employed as immunological adjuvants to stimulate immune responses, and lipid nanoparticles designed to encapsulate therapeutic and antigenic nucleic acids. By pairing the visual, single-particle assessments provided by cryogenic electron microscopy with the solution-phase, ensemble-averaged conformational data generated by small-angle X-ray scattering, the study demonstrated that scattering profiles alone can reliably resolve subtle morphological changes and guide formulation and manufacturing decisions without necessitating continuous electron microscopy imaging for every intermediate process condition.
The experimental evaluation of liposomal formulations utilized vesicles synthesized from 1,2-dioleoyl-sn-glycero-3-phosphocholine, cholesterol, and active pharmaceutical ingredients using a classic thin-film hydration and extrusion process. Formulations were passed through a 100-nanometer pore size extrusion membrane at an extrusion operating temperature of 50 degrees Celsius. For lipid nanoparticle characterization, formulations comprising an ionizable lipid, cholesterol, 1,2-distearoyl-sn-glycero-3-phosphocholine, and polyethylene glycol-dimyristoylglycerol were assembled with ribonucleic acid cargo. These lipid nanoparticles were synthesized under two distinct mixing methodologies: a conventional T-mix impingement process and automated microfluidic mixing via a commercial microfluidic mixing platform, while strictly maintaining equivalent nitrogen-to-phosphate charge ratios across both pathways. The resulting particles were characterized using dynamic light scattering alongside cryogenic transmission electron microscopy and small-angle X-ray scattering.
Cryogenic electron microscopy of the liposomal formulations was conducted on a transmission electron microscope operating at an accelerating voltage of 120 kilovolts and equipped with a 16-megapixel complementary metal-oxide-semiconductor camera. Liposome sample suspensions adjusted to approximately 3 to 4 milligrams per milliliter of total lipid were applied to glow-discharged copper grids, blotted using an automated vitrification unit, plunged into liquid ethane, and transferred to the microscope using a cryo-transfer holder maintained at temperatures below minus 170 degrees Celsius. Grid imaging was performed under low-dose conditions with an electron beam spot size of 5 and an imaging defocus of approximately minus 1.5 micrometers. For lipid nanoparticles, imaging was performed on an advanced transmission electron microscope operating at 200 kilovolts equipped with a direct electron detector, with image acquisition executed using automated collection software at nominal defocus values ranging between minus 7.0 micrometers and minus 4.5 micrometers.
Quantitative image analysis of the acquired electron micrographs categorized liposomes into discrete lamellarity populations based on visual inspection of at least 200 fully enclosed particles across multiple grid apertures. The liposomes were grouped into categories designating vesicles with one bilayer, greater than one bilayer, greater than two bilayers, and five or more bilayers. Unilamellar vesicles exhibited a single dark outer ring surrounding the aqueous core, whereas multilamellar vesicles displayed multiple concentric dark rings. For lipid nanoparticles, analysts quantified particle morphology by evaluating longitudinal and lateral dimensions to classify populations into particles possessing an aspect ratio close to 1, indicating spherical geometry, and particles exhibiting an aspect ratio greater than 1, reflecting elongated or irregular shapes. Particles produced by one mixing process displayed uniform spherical architecture, whereas particles from an alternate mixing process revealed asymmetric bleb-like compartment protrusions.
Small-angle X-ray scattering measurements were performed using an in-house laboratory X-ray instrument equipped with a high-flux, very long focus copper K-alpha microfocus X-ray source and a two-dimensional, vacuum-compatible hybrid photon-counting pixel detector. Solution samples spanning 20 to 30 microliters were loaded into a 1.8-millimeter-diameter quartz capillary using an automated robotic autoloader under an operating chamber vacuum pressure of approximately 0.03 millibar. For liposome suspensions, data collection spanned a scattering vector range of 0.01 to 0.5 inverse angstroms at a constant sample-to-detector distance, requiring total exposure durations between 1 and 2 hours. Lipid nanoparticle measurements were executed across an extended scattering vector range spanning 0.005 to 0.5 inverse angstroms over multiple detector distances with total exposure times of 3 to 4 hours, and automated cosmic radiation rejection, radial integration, and background buffer subtraction were completed using dedicated processing software.
The resulting scattering intensity curves revealed structural signatures that directly mirrored the lamellarity states identified in electron micrographs. For unilamellar liposome preparations containing exclusively a single phospholipid bilayer, the small-angle X-ray scattering profile exhibited a smooth, single broad feature resembling an inverted parabola within the scattering vector range between 0.03 and 0.3 inverse angstroms. In contrast, liposome populations containing multilamellar vesicles produced scattering profiles that developed two pronounced constructive interference peaks. The primary scattering peak emerged at approximately 0.1 inverse angstroms, reflecting constructive interference between stacked phospholipid bilayers, with the intensity and sharpness of this interference peak scaling directly with increasing bilayer lamellarity and ordering within the particle population.
To extract quantitative lamellarity parameters from the scattering curves, the researchers performed numerical modeling using the publicly available scattering analysis software SASVIEW, implementing a specialized stacked lamellar model. By constraining fit variables to physically relevant initial estimates—including lipid head and tail lengths, scattering length densities, and bilayer repeat spacing calculated directly from the first peak position—the software accurately converged on the average bilayer count, designated as N. Representative fits yielded values of N equals 1 with a reduced chi-squared value of 1.39 for unilamellar samples, N equals 2 with a chi-squared of 1.37 for intermediate mixtures, and N equals 3 with a chi-squared of 2.73 for more lamellar samples. For more complex multilamellar preparations, fitting yielded an average bilayer count of N equals 4 with a chi-squared value of 2.91, demonstrating reliable numerical tracking of bilayer stacking.
The study demonstrated the direct practical application of this analytical pairing by tracking the structural evolution of liposomes during membrane extrusion processing without acquiring intermediate electron micrographs. Measurements combining dynamic light scattering and small-angle X-ray scattering revealed that after only two extrusion passes, liposomes possessed an average hydrodynamic diameter of approximately 360 nanometers, a high polydispersity index of approximately 0.8, and a high lamellarity of N equals 5 from scattering model fits. As extrusion processing progressed from 2 to 10 passes, the particle size and polydispersity decreased sharply, with polydispersity reaching 0.133. Additional extrusion cycles up to 30 passes refined the vesicles into monodisperse 100-nanometer particles exhibiting a final lamellarity of N equals 2, confirming that scattering profiles accurately map batch refining during manufacturing.
When applied to lipid nanoparticles, small-angle X-ray scattering profiles successfully distinguished between differing internal structures and particle geometries generated by distinct mixing techniques. Guinier analysis determined a radius of gyration of 188.7 plus or minus 1.2 angstroms for uniformly spherical nanoparticles produced by Process A, compared to a larger radius of gyration of 257 plus or minus 5 angstroms for non-uniform particles produced by Process B. Modeling scattering data with a shape-independent broad peak model revealed a distinctive broad peak at 0.073 inverse angstroms with a correlation length of 41.2 angstroms in Process B particles, which was entirely absent in Process A preparations. Furthermore, calculated Porod exponents were 3.86 for the spherical Process A particles, approaching the ideal theoretical value of 4 for spheres, compared to 3.52 for the irregular Process B particles.
Analysis of dimensionless Kratky plots and normalized pair distance distribution functions provided complementary insights into nanoparticle architecture. Both lipid nanoparticle preparations displayed initial features characteristic of compact globular structures with a peak position near a dimensionless coordinate of 1.73 and a normalized peak height near 1.1. However, the less uniform particles from Process B exhibited a secondary broad peak spanning dimensionless values between 15 and 30, reflecting internal biphasic structural heterogeneity and phase-separated compartments. Pair distance distribution analysis performed using the GNOM algorithm produced a quality fit estimate score of 0.92 for the uniform particles from Process A, indicating an excellent solution, whereas the fitting quality score decreased to 0.66 for the heterogeneous Process B nanoparticles.
The researchers also investigated the sensitivity of small-angle X-ray scattering to variations in encapsulated cargo by comparing lipid nanoparticles formulated using the uniform mixing process with differing ribonucleic acid lengths. Evaluations tested formulations encapsulating short 20-base-pair small interfering ribonucleic acid molecules, 2000-nucleotide single-stranded ribonucleic acid molecules, and 4000-nucleotide single-stranded ribonucleic acid molecules. Scattering profiles were virtually identical at scattering vectors above 0.01 inverse angstroms across all cargo lengths, confirming that internal lipid packaging and nanoscale organization remained unchanged. Observable deviations occurred exclusively in the low-scattering-vector region below 0.01 inverse angstroms, reflecting overall particle size scaling from a radius of gyration of 207 plus or minus 2 angstroms for small interfering ribonucleic acid up to 288 plus or minus 6 angstroms for the 4000-nucleotide cargo.
The study highlights specific physical limitations inherent to each analytical method and emphasizes the necessity of pairing scattering with dynamic light scattering. Small-angle X-ray scattering provides ensemble-averaged structural metrics in solution but cannot resolve rare, discrete sub-populations or detect particle sizes exceeding approximately 120 nanometers without specialized ultra-small-angle configurations. Conversely, cryogenic electron microscopy directly visualizes individual structural aberrations and heterogeneous sub-populations but remains constrained by vitrification grid artifacts, limited field sampling, and intensive sample throughput bottlenecks. By formally establishing rigorous numerical and qualitative correlations between these modalities, the investigators demonstrated that automated laboratory small-angle X-ray scattering can serve as a dependable, rapid screening tool, reserving labor-intensive electron microscopy for targeted validation and substantially accelerating vaccine formulation pipelines.
Subject of Research: Correlation of small-angle X-ray scattering and cryogenic electron microscopy for vaccine nanoparticle characterization
Article Title: Adding to the pharmaceutical analytical toolbox for liposome and lipid nanoparticle characterization: correlating Small Angle X-ray Scattering (SAXS) and Cryo-EM imaging
Article References: Pabit, S., Chang, I. Y.-T., Kotowski, D., Jr., Gupta, S., Kummar, S., Ghodasara, M., Xiong, J., Smith, J., Austin, L., David, C., Boyd, D., Wang, M. M., McNevin, M. J., Mukherjee, M., & Farrell, C. J. (2026). Adding to the pharmaceutical analytical toolbox for liposome and lipid nanoparticle characterization: correlating Small Angle X-ray Scattering (SAXS) and Cryo-EM imaging. AAPS Open, 12(1), Article 36. https://doi.org/10.1186/s41120-026-00179-x
Image Credits: AI Generated
DOI: 10.1186/s41120-026-00179-x
Keywords: Small-Angle X-Ray Scattering, Cryogenic Electron Microscopy, Lipid Nanoparticles, Liposomes, Vaccine Formulation, Biophysical Characterization, Dynamic Light Scattering, Nanomedicine, Adding, pharmaceutical, analytical, toolbox
News Source: Kristina Jarvis. (November 13, 2026). Correlating X-Ray Scattering and Electron Microscopy Accelerates Vaccine Nanoparticle Screening. Scienmag.
