Clonal selection and optimization of bioreactor products
Clonal selection and optimization of bioreactor products. Most therapeutic monoclonal antibodies (MAbs) are produced by bioengineered mammalian cells that have been cultured in a bioreactor for two to three weeks. The changing redox environment such as temperature may impair the quality of monoclonal antibodies produced (for example, fragmentation, truncation), and proteases, reductases and other chemicals are also released from dead cells. Therefore, it is very valuable to establish an analytical method that can help cell culture to monitor the quality and integrity of IgG products in real time, especially in the process of cell line development and process scale-up.
In the early stages of bioprocessing, the titer of product IgG is usually low, and animal serum is usually contained in the culture medium. This makes it impossible to directly analyze the IgG quality. However, in the past 10 to 20 years, IgG products made with animal-free media and higher titer Chinese hamster ovary (CHO) cells have made it possible to directly analyze IgG at harvest without being affected by the host. The influence of cell protein or medium composition.
Direct analysis of the harvest without IgG purification is a valuable tool for clone selection and bioreactor optimization. This method will save time for sample preparation. It will provide a complete spectrum of IgG fragmentation or aggregation information without changes in product profile due to purification steps (for example, fragmented IgG loss). Here, we have introduced how to use capillary electrophoresis, surface plasmon resonance (SPR) and size exclusion high performance liquid chromatography (SE-HPLC) to monitor the quality of IgG produced during clone selection and purification. The bioreactor has been carried out. Optimization of conditions.
Materials and Method
The bioreactor harvest was provided by the cell culture manufacturing and control (CMC) group. For cSDS, we used the Bioanalyzer 2100 instrument, which is equipped with high-sensitivity protein labeling reagents and high-sensitivity protein chips.
The bioreactor harvest was diluted 20 times, stained with fluorescent dyes, and denatured at 90°C for 5 minutes with or without dithiothreitol (DTT). In order to measure the binding affinity of antigen and antibody, we again used the antigen from Teva Biologics, on the Biacore 3000 instrument, with CM5 sensor chip and amine coupling kit, both from Cytiva (formerly GE).
Finally, for SE-HPLC, we used an Agilent Advance Bio SEC (300Å, 4.6×150 mm, 2.7μm) column on the ultra-high performance liquid chromatography (UPLC system). For conventional SEC analysis, we use 250 mM phosphate buffer and 150 mM NaCl as the mobile phase with a flow rate of 0.3 mL/min. Dilute the sample to 1-2 mg/mL with SEC mobile phase. We recommend the same SEC procedure under low pH conditions, using only 0.1N HCl to adjust the pH of the sample to 3.6, and then dilute it to 1-2 mg/mL with glycine buffer (pH 3.6).
Results and discussion
In the process of clone selection and bioreactor process optimization, our cSDS electrophoresis analysis showed that several monoclonal antibody samples purified by protein A affinity chromatography showed a large amount of fragmentation (data not shown). We further analyzed the same samples (samples with pH 6-7) as above using conventional SE-HPLC, and were surprised to find that there were no product fragments in any samples (Figure 1 below). We believe that if the product fragments elucidated by cSDS electrophoresis analysis are related to IgG through non-covalent interactions, gentle sample handling may not destroy the interactions in SEC.
Therefore, before performing SE-HPLC analysis again, we treated the sample with acid (pH 3.6). The acid treatment mimics MAb purification steps, such as elution of IgG from a protein A column and subsequent virus inactivation. We observed low molecular weight (LMW) fragments of batches 1, 3, 4, and 5 in the obtained SEC chromatogram (see Figure 2 below), which correspond to half-mer IgG and are related to the cSDS electrophoresis display. Even after prolonged incubation at pH 3.6> 24 hours, a complete positive control MAb (batch 2) was found (data not shown).
Bioreactor harvest analysis
It is not clear whether the above-mentioned IgG fragment problem was caused during product purification or has occurred in the bioreactor. Therefore, we turned our attention to the harvest of bioreactors to solve this fragmentation problem. We analyzed the bioreactor harvested samples by unreduced cSDS electrophoresis, and found different fragments corresponding to half IgG molecules, one heavy chain and one light chain according to the molecular weight (Figure 3 below). The positive control bioreactor harvested under the same experimental conditions showed normal conditions. The intact IgG peak was 95% and the percentage of single light chain was low. The 125 kDa peak was characterized by the lack of a light chain IgG product. During the measurement, the medium itself did not emit a signal (data not shown).
IgG fragments are closely related to cell viability. We found that there are obvious fragments in the late stage (the last few days) of the bioreactor operation (see Figure 4 below). When analyzing all bioreactor harvest samples under reducing conditions, only heavy and light chain peaks appeared in cSDS electrophoresis, indicating that the fragmented peaks are part of IgG and may not be derived from host cell proteins (Figure 5 below). However, the baseline increase may come from host cell proteins.
We further analyzed the bioreactor harvest using low pH SE-HPLC as described above to determine whether the SDS and heat treatment required for cSDS electrophoresis analysis can provide a similar explanation for the fragmentation phenomenon. Figure 6 below shows that the low pH (3.6) SEC analysis of the samples harvested from batches 1-5 of the bioreactor clearly showed that hemimer breaks were observed in the samples from batches 1, 3, 4, and 5, while the first The two batches of positive controls were found to be intact.
We found that the pink in purified drugs is related to the degree of IgG fragmentation: the more fragments there are, the darker the pink in purified intermediates and drugs. Some publications have confirmed that the source of this color comes from vitamin B12. The increase in free thiols in fragmented IgG (measured by Ellman analysis) may be the root cause of the pink color in this case.
CHO cells should only secrete well-assembled intact IgG into their culture medium. IgG fragments will be released from dead cells or be reduced by chemicals released by such cells. We did observe a correlation between IgG fragments and low cell viability (data not shown). Some publications indicate that intracellular reductase and substrate/cofactor released from lysed cells may destroy disulfide bonds in IgG molecules, which is the main reason for IgG fragmentation. According to our analysis of the harvest collected from different time points, the reduction in production occurred rapidly and within a short period of time (see Figure 4 above).
We also explored SPR technology (Surface plasmon resonance) to monitor the quality of IgG in the bioreactor harvest. By comparing with the reference standard, we found that the antigen binding affinity of the fragmented IgG sample was reduced by 11%). Although the results are not as profound as those obtained by cSDS electrophoresis and low pH SEC, SPR technology does provide us with another direct quality control of the product IgG function in the bioreactor harvested samples without sample purification.
Quality starts with the bioreactor. We use simple and improved analytical tools to check the harvest of the bioreactor. The results provide useful guidance for clone selection and bioreactor process development. Bioreactor operation is the longest step in the manufacture of therapeutic monoclonal antibody products, with many potential risk factors that may affect product quality (for example, high concentrations of proteases and reduction of chemicals in the later stages of bioreactor operation). Due to concerns about matrix interference between the culture medium and host cell proteins, analytical assays are usually not used for direct IgG analysis of harvested materials from bioreactors. As we have already confirmed, technological advances in bioreactor control and analysis methods have made direct analysis of bioreactor harvests feasible.