Figure 3. Diagram illustrating the dual column operation. Condition indicates pre elution phases such as equilibration and wash. Regen indicates the combination of the post-elution phases: strip, sanitization and re-equilibration. DBC stands for dynamic binding capacity. Grey shade illustrates protein loaded into the column.

Virus Inactivation

The viral inactivation strategy implemented in this study used a plug flow reactor made from a 4.4 cm diameter x 20 cm height SEC (Size Exclusion Chromatography) column with Sephadex G-25 Coarse® resin to achieve a target residence time of 15 minutes. Acidification was performed before the SEC column and neutralization was performed post column. The capture product was titrated to an acidification target of pH 3.5 with an allowable range of ± 0.1. Acid was added in-line with the product line and the acidified product immediately flowed through the flow kit’s static mixer and pH probe. A feedback PID controller is used to continuously adjust the acid pump flow rate to ensure the pH is within the desired range. After the pH is within the acidification target range for an appropriate amount of time, the product is directed through the SEC column which was equilibrated at pH 3.5. Post SEC column, an in-line addition of base is continuously fed to neutralize the acidic product to a target of pH 7.2 with an allowable range of ± 0.2. The viral inactivation procedure during this process began on day 8 of the 14-day run. The start day for this step is determined based on (1) product quality impact from hold time at capture product conditions and (2) system flow rate lower limit to allow VI to start and run for extended duration. For this study VI was run in three sub-batches. For a future optimized process, we aim to operate without pausing.

Anion Exchange Chromatography

The product from VI step flowed through a filter train comprised of two Opticap® Sterile 0.2µm filters on either side of a Mustang Q® Membrane Capsule. This filter train was prepped offline, prior to its connection to the system according to the vendor specifications. This prep work was done the day before the filter train was connected to the system (day 7). The filter train was connected on day 8; this timeline coincided with the start of the viral inactivation step. From here, material was collected in a final 50L product collection bag. This final polishing chromatography step was performed with a single-use membrane in a flowthrough mode. The membrane was not reused. The 0.2µm filter downstream of the Mustang Q membrane ensured that no potential bioburden reached the final product collection bag.

Bioburden Sampling and Testing Strategy

Samples were taken from four points in the process: (1) the cell culture harvest 10L vessel, (2) Protein A product 10L vessel, (3) VI product flow kit vessel and (4) final product collection bag. Samples were taken from sample ports that were specifically installed during set-up and were clamped closed when not in use to prevent potential exposure to the environment. Samples were taken aseptically using 10 mL syringes, about 24 hours apart. Samples were aliquoted into individual, labelled tubes under a biosafety cabinet in order to preserve sterility. Once in the tubes, the samples were stored at 2-8°C until they were ready to be submitted for analysis (see section Analytical Testing
for more details). This sampling method was repeated every day of the run, for each sampling point. The protein A bindable testing was done in order to record the daily titer at each stage. The HCP and DNA concentration assays tested the impurity level at each stage and therefore see how effective each unit operation was at removing the impurities. The SE-HPLC testing was done to determine the monomer purity at each stage and therefore determine daily product quality across the process. The samples for testing bioburden were taken and plated onto culture plates under a biosafety cabinet using L-shaped spreaders. Once done, the plates were stored inside of an incubation chamber running at 35 °C. The plates were monitored every day until 72 hours had passed in order to ensure the sterility of the system. These sampling and bioburden plating methods were repeated every day for each sampling point, through the end of the run.

Analytical Testing

Size exclusion-high performance liquid chromatography (SE-HPLC)
SE-HPLC was performed using a Waters 515 pump, a Waters 2487 Dual Absorbance Detector (Waters) and a Rheodyme 77,251 injector and a TSK Gel G3000 SWXL column (300 mm × 7.8 mm, Tosoh Biosep). The volume of injection was adjusted according to each sample concentration in order to inject 250 µg of mAb1, and separation was performed at a flow rate of 1.0 mL/min. The running buffer was composed of 100mM sodium sulfate, 100mM sodium phosphate dibasic, pH 6.8. UV detection was performed at 280 nm. The results were reported as monomer peak percentage and high molecular weight (HMW) content percentage of the total integrated area.
CHO DNA Real-Time PCR Assay
The assay is performed in a 96-well plate format using a forward and reverse primer set specific for the Chinese hamster short interspersed repetitive element (SINE) DNA sequence. A unique 64 base-pair product is generated from CHO genomic DNA calibrators and test samples within 40 cycles of the PCR. DNA is extracted from test samples and an extracted control using the DNA Extractor Kit® from Wako. Test samples, CHO genomic DNA calibrators and controls are added to a reagent master mix containing primers and SYBR Green. Cycling is performed in three steps as follows: (1) 10 minutes at 95°C for polymerase activation; (2) 40 cycles of 15 seconds at 95°C for denaturation; (3) 40 cycles of 60 seconds at 60°C for annealing of primers and extension. Quantities of DNA in the samples above LLOQ are regressed from an assay specific standard curve generated by the Sequence Detection Systems (SDS) software.
Gyros CHO Host Cell Protein (HCP) Assay
Standard curve, samples and controls are diluted in a proprietary Gyros buffer and loaded into sample microplates. In-house prepared biotinylated sheep anti-CHO HCP capture antibody and fluorescent labeled sheep anti-CHO HCP detection antibodies are diluted and loaded into a reagent microplate. The microplates are loaded on the Gyrolab instrument and automated execution of the assay is initiated. Once all reactions are complete in the CDs are read by the instrument using a fluorescence detector at 647 nm. HCP quantities are regressed from the standard curve generated by the software and corrected for dilution factor. Reported sample concentrations are the mean concentration of the dilutions yielding spike recovery values of 75-125%.
HPLC Protein A bindable Assay for mAb concentration
Protein A bindable HPLC is used to determine IgG concentration utilizing Protein A affinity chromatography separation technique. Test sample is injected onto a PA Immuno-Detection Sensor Cartridge (2.1 mmD x 30 mmL) with PBS buffer as a mobile phase.  The sample flows through the cartridge where the target IgG specifically binds to the Protein A immobilized on the cartridge and non-target components are washed out.  The bound IgG is eluted with 0.1% phosphoric acid in PBS, and detected using UV absorbance at 280 nm.  The IgG content is then quantified using linear regression analysis of the standard curve prepared from the corresponding calibration standards.  Results are reported in µg/mL.