Ask the Expert Liposome and Viral Vector Characterization: Use of Electron Microscopy and Image Analysis

View PDF

BPI-AtE-logowith Dr. Josefina Nilsson

For this webcast, Josefina Nilsson (EM Services business unit head) discussed Vironova’s work, including case studies. She focused on characterization of drug and gene delivery platforms with electron microscopy and image analysis, specifically for systems that use viral vectors or liposomes. Along with two colleagues — Gustaf Kylberg (image analysis expert) and Mathieu Colomb-Delsuc (electron microscopist) — she then answered questions from the audience.

Nilsson’s Presentation
Structural characterization provides important insights into the quality of development and manufacturing practices, including encapsulating systems. Important characteristics include morphology, integrity, sample purity, and the ratio between filled and empty particles. Our analysis can measure unique parameters specific to certain delivery systems such as lipid-based and viral vectors. Using electron microscopy and image analysis, you can gain objective, repeatable, and quantitative structural information cost efficiently.

We start by preparing samples and acquiring images using electron microscopy. Then we use our in-house developed software for image analysis. The program automatically detects a large number of particles to reach statistical significance and classifies each type within the obtained image. It can generate a report of the analysis that includes representative images with a quantitative analysis, statistics, and summary.

Case Studies: Our first study looks at samples using adenoviruses with different formulations. One formulation had adenovirus particles with good integrity, nice morphology, and particles acting as a single unit. However, when the formulation was changed, those particles started to break down and aggregate.

Another study involved samples using an adenoassociated virus (AAV) viral vector taken during purification both before and after sample filtration. Although the images were quite “noisy,” our software was able to detect and categorize different particles: AAVs, aggregates, and protein-based impurities. In this study, filtration increased aggregation from stress on the samples rather than removing aggregates.

Additionally, our analysis can measure the size and shape of particles, classify them, and differentiate between filled and empty particles using cryoelectron microscopy. For lipid-based systems, we can perform lamellarity analysis and measure the thickness of lipid bilayers. The resulting information can help us determine the optimal storage conditions for such particles to maintain their dimensions.

Electron microscopy and image analysis can provide all the structural information needed for effective and high-quality research and development and manufacturing of gene and drug delivery systems.

Questions and Answers
Do you prepare GMP reports of that analysis as part of the service? Yes.

How do you ensure that images are representative? We always try to make the image as representative as possible by exploring the whole area of a grid and looking at areas with different ice thickness. What we provide to the customers is representative of what was in the sample.

How can you ensure that liposome morphology is representative and not affected by preparation? We always acquire a set of images from different areas of the grid so it is representative of the sample. We treat samples by “flash freezing” in a liquid ethane bath, where it goes from room temperature to cryogenic temperatures within milliseconds. This ensures that we don’t affect liposome morphology or integrity.

How many particles do you measure? That depends on how many we have on-grid. With sufficient concentration, we analyze at least 1,500 particles to be sure that we get a representative population and statistically relevant data.

Can you measure polyethelene glycol (PEG) layers of liposomes using your electron microscopy techniques? We cannot measure the PEG layer because it is not dense enough to be seen by electron microscopy. The layer is composed mainly of oxygen, hydrogen, and carbon, which doesn’t give you enough contrast against the carbon support.

How do you discriminate between filled and empty vial particle vectors (e.g., packaging of AAV particles)? We produce a radial density profile. Empty particles appear as open circles; filled particles look like discs. By looking at the radius of each particle, we can see that the profile differs depending on whether it is a disc or circle, so we can discriminate between empty and filled.

Is it possible to accurately gauge particles or liposome concentration with your software — and if so, has it been compared with alternate techniques? No, electron microscopy is not suitable to gather quantitative data when it comes to number or concentration of particles. When we talk about quantization, it is a relative analysis such as a ratio between filled versus empty particles.

More Online
The full presentation of this webcast can be found on the BioProcess International website at the link below.

Listen online at www.bioprocessintl.com/Vironova-ATE-2016.