Distinguishing aggregated active pharmaceutical ingredients (APIs) from other particle types is critical to evaluating a protein product’s stability, but standard characterization tools struggle to discriminate proteins from nonproteins with similar sizes, shapes, and morphologies. In a 25 June 2020 “Ask the Expert†webcast, Bernardo Cordovez (founder and chief scientific officer of Halo Labs) introduced his company’s Aura subvisible-particle analyzer. He explained how the device combines an innovative microscopy technique with sophisticated imaging software to characterize subvisible particles more quickly and accurately than incumbent technologies can.
Cordovez’s Presentation
Cordovez observed that flow imaging requires large sample volumes and cannot distinguish proteins from plastics or degraded polysorbates. Raman and Fourier-transform infrared (FTIR) spectroscopy can do so but require considerable time to acquire spectra and significant expertise to determine which signals to isolate and analyze. That relegates spectroscopy to quality-control applications, when it is least advantageous to change a process.
The Aura system uses “fluorescent membrane microscopy†to image and measure particles. First, a 96-well plate is fed into the system to capture a brightfield-microscopy “background†of each well membrane’s topography. Samples are loaded, filtered through a polycarbonate plate perforated with 400-nm pores, and imaged again. Then proprietary software aligns the images and removes the background, enabling detection and measurement of particles between 1 µm and 5 mm with high refractive-index contrast. Such contrast is critical, Cordovez noted, because proteins and aqueous suspensions can have similar optical properties.
Unlike spectroscopy, the Aura system is high throughput. It requires little sample (>5 µL), evaluates 96 samples in under an hour, and integrates easily with automation tools. In minutes, Aura software collates measurements of particle shape, size, morphology, and intensity to generate particle-, sample-, and experimental-level assessments. The system also analyzes entire samples, such that all particles become visible. By contrast, a flow imager reveals ~20% of a sample’s particles.
A fluorescent-labeling workflow enables the Aura instrument to identify particles. After performing the brightfield steps, an analyst can apply 40 µL of 5 mM thioflavin T (ThT) dye to a well plate, suction out residue, and return that plate to the system for imaging. Cordovez noted that ThT is useful because it binds to protein amyloid structures — but not to plastic, stainless-steel, and polysorbate particles.
During a controlled experiment, the Aura system easily distinguished between rotated human IgG and an ethylene tetrafluoroethylene (ETFE) protein-aggregate mimic. Halo analysts loaded a 96-well plate with 24 samples of IgG and ETFE and another 24 wells with an IgG–ETFE mixture. The remaining wells served as controls. Knowing that the IgG and ETFE samples contained ~140,000 particles/mL and ~50,000 particles/mL, respectively, analysts determined that membrane-topography imaging successfully quantitated the particles: The system detected ~185,000 of the ~190,000 particles/mL in the mixed samples.
Subsequent fluorescence imaging also separated the IgG and ETFE easily. The proteins fluoresced upon ThT binding, whereas ETFE particles did not stain and remained dark. IgG and ETFE also showed distinctive side-angle scattering signatures. Based on measurements of such properties, Halo analysts determined that 15.5% of measured particles in the IgG–ETFE blend were nonproteinaceous, which was consistent with readings from the bright-field quantitation step.
Although IgG and ETFE appear to be identical using flow-image microscopy, the Aura system can distinguish between even highly similar particles. It can render data in several ways to offer high-level insights about samples, and it features a second channel that can apply dyes for identification of lipids and protein monomers. Now, Halo Labs is exploring multichannel fluorescence and side-scattering techniques to enhance particle identification further.
Questions and Answers
What dyes besides ThT can Aura systems accommodate? Defaults for the system’s second channel include TMA-DPH and Alexa Fluor 488, BODIPY, and DiI dyes (Thermo Fisher Scientific), which enable distinction of lipids and protein monomers. Further customization is possible based on customer needs.
How long do fluorescence measurements take? A sample can be analyzed for fluorescence in 30 seconds. By contrast, Raman spectroscopy requires several minutes per sample.
How can users leverage the device’s “expression engine?†An Aura system stores all measurements. The expression engine automatically thresholds particles that meet specific criteria.
What role might side-scatter (SIMI) measurement play alongside fluorescence analysis? SIMI assesses a particle’s topography, so it can help confirm whether a particle is a protein.
Watch the full webcast and learn more about the Aura subvisible particle analyzer now.