Virus is used for gene therapy. 3D illustration
Featured Story

Do You Know Your AAV Capsid Ratio?

Gene therapies correct genetic disorders by replacing missing or defective genes through transplantation of a normal gene into patient cells. Adeno-associated virus (AAV) vectors are commonly used as the delivery system for these therapies. A major challenge in manufacturing AAV vectors is the presence of impurities (i.e., empty or partial capsids), which may affect the efficacy and safety of the vector product (i.e., full capsid). It is essential to determine the ratio of these impurities with that of the vector product, since this is a critical quality requirement for the AAV production process.

Confident Quality Control with Transmission Electron Microscopy

While orthogonal approaches such as capillary isoelectric focusing (cIEF), ion exchange chromatography (IEC), and analytical ultracentrifugation (AUC) can determine the ratio of full to partial or empty capsids, only transmission electron microscopy (TEM) provides a visual morphologic assessment (i.e., images), allowing you to identify impurities with greater clarity.

Get Started

When combined with a deep learning algorithm that is quality-controlled for accuracy by trained image analysis scientists and pathologists, TEM achieves both qualitative and quantitative assessment of AAV capsids. In addition, impurities and aggregation can be evaluated and quantified. TEM-based impurity testing is far superior to other methods, which deliver ratios without visualization of the capsids or evaluation of aggregation and impurities.

Electron Microscopy Capabilities

Our electron microscopy facility houses three transmission electron microscopes (Hitachi H7600, JEOL JEM-1011, and JEOL JEM-1400+) and a JEOL IT500HRLA scanning electron microscope equipped with an energy-dispersive X-ray spectrometer (EDS).

Microscope Operating Voltage (kV) Resolution
Hitachi H760 80 0.35 nm at 120 kV
JEOL JEM-1011 80 0.45 nm
JEOL JEM-1400+ 80 0.38 nm
JEOL IT500HRLA 0.5 to 30 1.5 nm at 30 kV (high vacuum mode)
1.8 nm at 15 kV (low vacuum mode)

Each TEM instrument is each equipped with a GLP-validated 16-megapixel mid-mount camera (AMT XR-16-Active Vu). The camera is Peltier cooled, and provides 8 frames per second at 4.4 binning (full field). All studies are performed by according to GLP by a group of 8 electron microscopists who have more than 125 years of combined experience.


Use of TEM and Deep Learning for Classification of AAV Capsids



Imaging and Analysis of Capsids

Capsids can be prepared for cryo-EM (by plunge-freezing unfixed samples) or for routine TEM (fixed or unfixed). For routine negative stain TEM, BSL-1 samples can be prepared without fixation while BSL-2 samples must be inactivated with a fixative containing glutaraldehyde. We currently offer negative staining, imaging, and determination of the empty/partial to full capsid ratio using an artificial intelligence-based image analysis platform (Visiopharm). The capsid suspension is loaded onto a formvar-coated grid, stained with a heavy metal (i.e., uranyl acetate or phosphotungstic acid), and imaged by TEM.

The digital images (magnification ~210,000x) are imported into our GLP-validated Visiopharm platform, which is trained to identify capsids based on the amount of stain within the structure. Electron-lucent (unstained) homogenous capsids are classified as full, those with an electron-dense core (stained) are classified as empty, and capsids containing non-homogeneous staining within the core are classified as partially full.

The figure below demonstrates classification of viral capsids as empty (red), full (green), and partially full (blue) capsids, which are used to automatically provide a ratio of empty to full capsids. 

The figure below demonstrates classification of viral capsids as empty (red), full (green), and partially full (blue) capsids, which are used to automatically provide a ratio of empty:full capsids.

In contrast to other providers of this service, our study packages include algorithm optimization on each set of samples and quality control of algorithm performance on all TEM images. Furthermore, all images with the algorithm overlay are archived with the study. This results in increased accuracy and reliability of the resulting data.

Get Started