Agarose native gel electrophoresis for characterization of antibodies
Cynthia Li a , Teruo Akuta b , Masataka Nakagawa b , Tomomi Sato b , Takashi Shibata b , Toshiaki Maruyama c , C.J. Okumura c , Yasunori Kurosawa c , Tsutomu Arakawa a, ⁎ a Alliance Protein Laboratories, a Division of KBI Biopharma, 6042 Cornerstone Court West, San Diego, CA 92121, United States of America b Research and Development Division, Kyokuto Pharmaceutical Industrial Co., Ltd., 3333-26, Aza-Asayama, Kamitezuna Takahagi-shi, Ibaraki 318-0004, Japan c Abwiz Bio, Inc., 9823 Pacific Heights BLVD, San Diego, CA 92121, United States of America article info abstract Article history: Received 21 January 2020 Received in revised form 14 February 2020 Accepted 16 February 2020 Available online 19 February 2020
This study was conducted to evaluate applicability of the previously reported native agarose gel electrophoresis to the analysis of various monoclonal and polyclonal antibodies. Experiments were carried to test the electrophoresis system for characterization of different monoclonal antibodies and animal serum, analysis of expressed antibodies in cell culture and evaluation of antibody stability. An attempt to optimize the electrophoretic condition was made by adjusting the electrode buffer concentration, electrophoretic run time and agarose concentration. © 2020 Elsevier B.V. All rights reserved.
Keywords: Agarose Native electrophoresis Monoclonal antibody Histidine Expression Basic protein 1. Introduction We have previously reported [1] that native electrophoresis based on agarose gels works well for both acidic and basic globular proteins using a buffer mixture containing equal concentrations of basic histidine (His) and acidic 2-morpholinoethanesulfonic acid (MES) with resultant pH of 6.1 [2,3]. Although the method appeared to be promising, it was not clear how applicable it would be for various protein samples. Since native gel electrophoresis can separate proteins in the native state, there are a number of potential applications, including the analysis of protein complex, non-covalent aggregation, protein-ligand interaction and different conformational states of a particular protein. Native gel electrophoresis can offer a simple and easy measurement for such analyses. Here, we have investigated its application for characterization of different monoclonal antibodies and analysis of animal serum, expression and purification of recombinant monoclonal antibodies and heat aggregation of antibodies and are investigating its broader applications. Monoclonal (mAb) and polyclonal antibodies are among the most important groups of proteins for therapeutic and biochemical applications. Because of their therapeutic importance, mAbs are analyzed by various chromatographic and electrophoretic techniques, including size exclusion chromatography [4–8], reverse-phase HPLC (high performance liquid chromatography) [9], capillary electrophoresis [10,11], ion-exchange and hydrophobic interaction chromatography [7,10,12], SDS-PAGE (sodium dodecylsulfate polyacrylamide gel electrophoresis) [13] and native gel electrophoresis [14]. As a simple way to analyze mAb aggregation, Pathak et al. [14] made an elaborative work to develop native-PAGE that work well for basic proteins. Here we applied our agarose native gel electrophoresis for characterization of antibodies. The mAbs used here are reagent antibodies developed against phospho-tyrosine of short peptides and used to characterize phosphorylation of various kinases. We believe that this agarose native gel electrophoresis offers a simple characterization of reagent mAbs for those who may have no access to chromatographic equipment. 2. Materials and methods Different monoclonal antibodies were generated against short peptides containing phospho-tyrosine and provided by Abwiz Bio, Inc. and Kyokuto Pharmaceutical Industrial. Monoclonal antibodies were expressed in HEK293 cells. The HEK293 cells were cultured in suspensions in newly developed HEK293 medium (Kyokuto Pharmaceutical Industrial, Tokyo, Japan) or Free style F17™ expression medium (Thermo Fisher Scientific, Waltham, MA, USA) supplemented with 4 mM L-glutamine, 0.1% Kolliphor P188 (Merck, Darmstadt, Germany) in a humidified atmosphere of 95% air and 5% CO2 at 37 °C on a shaker running at 130 rpm. Transfection of HEK293 cells was performed according to the published guideline [15] and our previous report [16], using pTT5 vectors (the National research council of Canada (NRC), International Journal of Biological Macromolecules 151 (2020) 885–890 ⁎ Corresponding author at: Alliance Protein Laboratories, a Division of KBI Biopharma, 6042 Cornerstone Court West, Suite A, San Diego, CA 921210, United States of America. E-mail address: tarakawa@kbibiopharma.com (T. Arakawa). https://doi.org/10.1016/j.ijbiomac.2020.02.185 0141-8130/© 2020 Elsevier B.V. All rights reserved. Contents lists available at ScienceDirect International Journal of Biological Macromolecules journal homepage: http://www.elsevier.com/locate/ijbiomac Montreal, Canada) harboring L and H chains and Transporter 5™ Transfection Reagent (Polysciences, Inc. Warrington, PA, USA). The transfected cells were cultured in a 25 mL or a 300 mL Nalgene PETG Erlenmeyer flask with plain bottom (Thermo Fisher Scientific). The supernatants were harvested 3–7 days after the transfection and subjected to purification using Protein A/G column. The bound antibodies were eluted with a commercial IgG Elution Buffer, pH 2.8 (Thermo Scientific), whose low pH may cause partial denaturation. To reduce timedependent low pH denaturation, the eluted fractions were immediately neutralized by the addition of 1 M pH 8.5 Tris (Bioscience). Goat, sheep, rabbit serum and polyclonal IgG were obtained from Japan Lamb (Hiroshima, Japan). Mouse serum was purchased from FUJIFILM Wako Pure Chamical Corporation (Tokyo, Japan). Mouse and human poly IgG were purchased from Thermo Fisher Scientific (USA). Human plasma was from Kohjin Bio (Saitama, Japan). Agarose-based native gel electrophoresis was run at room temperature at 100 V using His/MES buffer (pH 6.1) as previously reported [1]. Briefly, agarose powder was suspended in His/MES buffer and heated for complete dissolution. The dissolved solution was gelled in a horizontal tray or a vertical gel cassette (Invitrogen). While still hot, appropriate comb was inserted. Reducing and non-reducing SDS-PAGE on a 10–20% gel (SuperSep, FUJIFILM Wako Pure Chemical) were performed with the pre-stained protein marker Precision Plys Protein Dual Color Standards (BIO-RAD, USA). Samples were mixed with His/MES buffer containing methyl green and glycerol (or sucrose) for native gel electrophoresis or 2× reducing or non-reducing sample buffer for SDS-PAGE. Protein was detected by Coomassie Brilliant Blue dye (Quick-CBB PLUS, FUJIFILM Wako Pure Chemical, Japan). 3. Results and discussion Previously, we have established a general protocol for running native electrophoresis using agarose and His/MES buffer at pH 6.1. We were curious about whether or not the established protocol can evaluate different monoclonal antibodies and their expression, purification and stability. 3.1. Different mAbs First, we have generated different monoclonal antibodies of IgG1 class to be used to detect phosphorylated tyrosine of target kinases. Some of them were generated and purified by Abwiz Bio, Inc. Their molecular weights and isoelectric points were calculated from the gene and hence amino acid sequence and are summarized in Table 1. These values will be significantly different from the actual values due to post-translational modifications, in particular glycosylation and resultant sialylation. Nevertheless, it may worth mentioning that while the calculated values are similar for these mAbs used here, a noticeable difference is the lower isoelectric point of mAb-1. Ten mAbs (labeled mAb-1 to mAb-10) raised against different antigens were subjected to SDS-PAGE analysis. Fig. 1A shows non-reducing SDS-PAGE profile of mAb-1 to mAb-10. The mAb-1 (lane-1) to mAb-3 (lane-3) have a similar size consistent with the similar sequence molecular weights, while reducing SDS-PAGE (Fig. 1B) shows slightly different sizes of the heavy chain (lane-1 to lane-3). Fig. 1A also shows a similar size of mAb-4 (lane-4) to mAb-9 (lane-9), consistent with their similar sequence molecular weight. The mAb-10 (lane-10) showed a larger size, perhaps reflecting its slightly higher sequence molecular weight. Although most of mAbs showed a sharp band, mAb-5 showed smearing (lane-5). Interestingly, while mAb-4 to mAb-10 appeared to be similar in heavy chain size (Fig. 1B), the light chain appeared to show differences in size: clearly the mAb-7 and mAb-10 have a larger size of light Table 1 Molecular weight and isoelectric point. mAb Molecular weight Isoelectric point mAb-1 145,000 6.4 mAb-2 146,000 8.1 mAb-3 146,000 8.2 mAb-4 140,000 8.1 mAb-5 140,000 8.8 mAb-6 141, 000 7.8 mAb-7 142,000 8.1 mAb-8 141,000 8.6 mAb-9 142,000 8.6 mAb-10 144,000 7.6 mAb-11 141,000 7.9 mAb-12 142,000 8.6 mAb-13 142,000 7.9 Fig. 1. SDS-PAGE of 10 mAbs Load, 2 μg. A. Non-reducing SDS-PAGE. B. Reducing SDSPAGE. Fig. 2. Native gel electrophoresis of 10 mAbs on 1% agarose flat-gel in 0.1 M His/0.1 M MES. Run time, 1 h at 80 V. Load, 10–20 μg. 886 C. Li et al. / International Journal of Biological Macromolecules 151 (2020) 885–890 chain. The mAb-5 showed smearing in reducing SDS-PAGE as well (lane-5). Fig. 2 shows the results of horizontal agarose gel electrophoresis with 0.1 M His/0.1 M MES for 3 commercial proteins, i.e., BSA (laneB), chymotrypsin (lane-C) and lysozyme (lane-L), as controls and 10 mAb samples (mAb-1 to mAb-10). In this figure, the top side is the anode and the bottom side is the cathode. At the buffer pH of 6.1, negatively charged BSA migrated toward the anode as expected, while positively charged lysozyme and chymotrypsin migrated toward the cathode. More positively charged and smaller lysozyme migrated faster than chymotrypsin. These 10 mAbs (i.e., mAb-1 to mAb-10) all migrated toward the cathode, indicating that they are positively charged at pH 6.1, meaning that they have an isoelectric point above pH 6.1, consistent with their calculated isoelectric points (Table 1). These mAbs, except mAb-5 (lane-5) and mAb-8 (lane-8), showed a sharp band, consistent with the SDS-PAGE results (Fig. 1). The mAb-1, 2 and 3 (lane-1 to lane-3) showed different mobilities (faster in this order), suggesting that their charged state or the structure in the native state may be different, as they show a similar size in the presence of SDS and the sequence molecular weight (Fig. 1). The observed low mobility of mAb-1 appears to be consistent with its lower isoelectric point of 6.4. The mAb-4 (lane-4) to mAb-7 (lane-7) showed a similar mobility, consistent with their similar size in the presence of SDS. The mAb-10 (lane10) showed the slowest mobility, perhaps reflecting its larger size under non-reducing (Fig. 1A, lane-10) and reducing (Fig. 1B, light chain) conditions, although a possibility of different charged state or structure cannot be excluded. In fact, the isoelectric point of mAb-10 is lower than the mAb-2 to mAb-9. The mAb-5 (lane-5) showed smearing in the native gel (Fig. 2) as in SDS-PAGE (Fig. 1), suggesting aggregation. Since smearing was less severe in SDS-PAGE, aggregates appeared to be dissociable in the presence of SDS. The observed smearing appears to correlate with its low yield (data not shown). On the contrary, the mAb-8 showed smearing only in the native gel, indicating that aggregation can be fully dissociated by SDS. Thus, the observed results demonstrate that the native electrophoresis can distinguish different charged states of antibodies and aggregation tendency in the native state. The same samples were subjected to 1% agarose vertical gel with the 0.1 M His/0.1 M MES, pH 6.1 system. Fig. 3 shows the pattern for chymotrypsin (lane-C) and lysozyme (lane-L), which were identical to the results with the flat-bed mode: note that BSA that would migrate upward toward the anode on the vertical gel was not included. The pattern for mAb-1, 2, 3 and 4 (lane-1, 2, 3, 4) was identical to that in the flat-bed results, indicating consistency between flat-bed and vertical modes under identical conditions. The mobility and pattern were similar for mAb-6 between the flat-bed (Fig. 2, lane-6) and vertical (Fig. 3, lane-6) modes. The pattern and mobility for mAb-10 (lane-10) were also similar. The smearing pattern of mAb-5 (lane-5) was also similar to the result in Fig. 2 (lane-5), although smearing was seen to be more severe in the vertical mode. While the mAb-7 appeared to be homogeneous in flat-bed (lane-7), it showed some smearing on vertical mode (lane7). Smearing nature of the mAb-8 (Fig. 2, lane-8) was more severe in vertical gel (lane-8), demonstrating heterogeneous nature of this sample. The most striking difference was observed for mAb-9 between flat-bed and vertical modes (lane-9). Although the reason for such a large difference is not clear, there appears to be fundamental difference between flat-bed and vertical modes, perhaps due to different gel thickness (1.5 mm for the vertical gel and 5 mm for the flat-bed gel) and the way the gel contacts with the electrophoresis buffer (only top and bottom portions exposed to the buffer for the vertical gel and the entire gel submerged into the buffer for the flat-bed). Fig. 3. Native gel electrophoresis of 10 mAbs on 1% agarose vertical gel in 0.1 M His/0.1 M MES. Run time, 75 min at 80 V. Load, 2.5–4 μg. Fig. 4. Native gel electrophoresis of 10 mAbs on 3% agarose vertical gel in 0.1 M His/0.1 M MES. Run time, 1.5 h at 80 V. Load, 2.5–4 μg. Fig. 5. Native gel electrophoresis of 12 mAbs on 1% agarose flat-gel in 0.1 M His/0.1 M MES. Run time, 70 min at 80 V. Fig. 6. Native gel electrophoresis of 11 mAbs on1% agarose vertical gel in 0.2 M His/0.2 M MES. Run time, 1 h at 80 V. Load, 2.5–4 μg. C. Li et al. / International Journal of Biological Macromolecules 151 (2020) 885–890 887 Agarose concentration was increased to 3% on vertical gel with an identical 0.1 M His/0.1 M MES system: note that 3% agarose generates high viscosity that traps air bubbles. It is extremely hard to poor the melted viscous agarose solution into the mini-gel cassette (1.5 mm space). Fig. 4 shows the electrophoretic pattern of the same samples used in Fig. 3. In overall, the pattern was nearly identical with 1% agarose, while the bands appeared to be slightly sharper with the 3% gel. The observed similarity and difference between 1% flat-bed and vertical gels was reproduced with 3% agarose vertical gel. The mAb-5, 8 and 9 that exhibited smearing were also smearing on the 3% agarose gel (lane-5, 8 and 9 in Fig. 4). Fig. 5 shows the results of more mAbs on 1% flat-bed agarose gel in 0.1 M His/ 0.1 M MES. All mAbs migrated in the same direction as chymotrypsin (lane-C), demonstrating their isoelectric points above 6.1. The mAb-1, 2, 3 and 4 (lane-1 to 4) showed a similar pattern to Fig. 2. Two batches of mAb-5, i.e., mAb-5.1 (lane-5) and mAb-5.2 (lane-6), showed smearing. It appeared that this particular mAb may not be stable during production, purification or storage. Two batches of mAb-10, i.e., mAb-10.1 (lane-7) and mAb-10.2 (lane-8), were similar to the pattern in Fig. 4 (lane-10). The mAb-8 (lane-10) showed smearing as in Fig. 2 and Fig. 3 (lane-8), indicating its potential instability. The mAb-9 (lane-9) was similar to the band in Fig. 2 (lane-9). Two mAbs, mAb-6 (lane-12) and mAb-7 (lane-11), appeared to be homogeneous as in Fig. 4. A vertical gel in 0.1 M His/0.1 M MES showed a similar pattern (data not shown). Fig. 6 shows the results with a vertical 1% agarose gel in 0.2 M His/ 0.2 M MES buffer system at pH 6.1, where chymotrypsin (lane-C) and lysozyme (lane-L) behaved accordingly. The mAb-1, 2, 3 and 4 (lane-1 to 4) behaved similarly to the flat-bed gel (Fig. 5). The mAb-5.2 (lane5) showed smearing as in Fig. 5 (lane-6). The mAb-10.1 (lane-6), mAB-10.2 (lane-7) and mAb-9 (lane-8) behaved similarly to Fig. 5. The mAb-8 was similar in the flat-bed mode (Fig. 5, lane-10) and the vertical mode (Fig. 6, lane-9). The mAb-7 (lane-10) and mAb-6 (lane11) behaved similarly to Fig. 5. Thus, a higher buffer concentration does appear to slightly affect the electrophoretic pattern. 3.2. Purification fractions Fractions from Protein-A/G purification were analyzed by 1% flatbed electrophoresis in 0.1 M His/0.1 M MES. Fig. 7 shows fractions Fig. 7. Native gel electrophoresis of Protein A/G purification fractions for mAb-11 and mAb-12 on 1% agarose flat-gel in 0.1 M His/0.1 M MES CM, conditioned medium. Run time, 1.5 h at 80 V. Fig. 8. Gel electrophoresis for the analysis of expression of mAb-13 CM, conditioned medium (10 μl loaded). A. Native gel electrophoresis on 1% agarose flat-gel in 0.1 M His/0.1 M MES Run time, 40 min at 100 V. B. SDS-polyacrylamide gel electrophoresis on 10–20 gradient gel Run time, 50 min at 30 mA. 888 C. Li et al. / International Journal of Biological Macromolecules 151 (2020) 885–890 from 2 different mAb samples, i.e., mAb-11 and mAb-12. Lane-B shows the mobility of BSA, indicating that the top side is the anode. As shown in lane-1 (conditioned medium culture supernatant) and 2 (concentrated supernatant), the mAb-11 migrated toward the cathode, indicating that it is positively charged at pH 6.1. Lane-3 shows combined fractions of flow-through and wash fractions, indicating that most impurities were either acidic or neutral, clearly distinguishing from the basic mAb band. Lane-4 to 6 show the eluted fractions, where the majority of eluted mAb-11 was in the first fraction. A similar observation was made for mAb-12, although its expression was much lower. As shown in lane-7 to 9, most cell culture impurities were either acidic or neutral as is the case for mAb-11. The majority of mAb-12 was in the first fraction and appeared to show smearing (lane-10), suggesting that this mAb-12 may be unstable during purification. Fig. 8 shows the results for rabbit mAb-13 expressed in HEK293 cells on 1% flat-bed gel. Lane-d0 to d7 (Fig. 8A) show the HEK293 cell culture medium as a function of cell culture time. Lane-ST shows the purified mAb-13, migrating toward the cathode, indicating that its pI is above 6.1. The observed smearing of the purified mAb-13 is most likely due to the formulation components in the sample interfering with the band sharpness. As shown in lane-d3, the mAb expression was clearly observed after transfection and its expression increased with culture time. It is also clear that most impurities, e.g., host cell proteins, were acidic and increased with cell culture time. This clear distinction between the mAb and the impurities makes the native gel system attractive in analyzing the purity of recombinant mAbs expressed in HEK293 cells during production and purification. Such distinction cannot be obtained by a conventional SDS gel electrophoresis. Fig. 8B shows the SDS gel analysis under reducing and non-reducing condition for the same samples, where the heavy (H) and light (L) chains and the intact mAb-13 were marked. Impurity bands occur above and below the bands corresponding to the mAb-13. There are well-established methods using expensive HPLC columns to quantify host cell proteins essential for production of therapeutic and diagnostic antibodies [17–20]. However, loading such crude materials, containing, e.g., lipids, pigments, sugars, nucleic acids and large product aggregates, onto analytical columns may damage them and therefore pre-column or pre-treatment was required for removing these impurities. On the contrary, this agarose native gel electrophoresis offers a simple protocol even with crude materials: it compares well to HPLC methods and can be performed rapidly and cost effectively. Next, various animal sera were examined on a 5% agarose flat-bed gel. Fig. 9 shows the results for mouse (lane-1 and 2), goat (lane-3 and 4), sheep (lane-5 and 6), rabbit (lane-7 and 8) and human (lane-9 and 10) serum, where albumin and polyclonal IgG (Poly IgG) were marked. The odd lanes correspond to the serum and the even lanes correspond to the purified polyclonal IgG. In all cases, smearing IgG bands were observed, migrating toward to the cathode, while albumin and other serum proteins occur as acidic proteins, again demonstrating contrasting migration of antibodies and other serum proteins. It should be pointed that the mobility of albumin is species-dependent and the sheep (lane-5) and goat (lane-3) sera appeared to contain a larger quantity of IgG than other species. 3.3. Stability Whether or not we can follow degradation of mAb on the current native gel electrophoresis system was examined. A mAb, mAb-14, and BSA were heat treated. As shown in Fig. 10, both chymotrypsin (laneC) and BSA control (lane-B) showed expected mobility. Lane-B1 and B2 show BSA before and after heating at 60 °C for 2 h. While unheated BSA (lane-B1) showed a single band, the heated BSA exhibited two bands (lane-B2), most likely due to aggregation [1]. Lane-1 and 2 show the mAb-14 stored at 4 and − 20 °C, while lane-3 to 6 show the mAb-14 heated for 24 h at 37 (lane-3), 50 (lane-4), 60 (lane-5) and 70 (lane-6) °C. This particular mAb migrated toward the cathode as in chymotrypsin (lane-C). There appeared to be no changes up to 50 °C heating. After 60 °C heating (lane-5), it showed smearing and no apparent bands were observed after 70 °C heating (lane-6). The observed heat degradation is consistent with the general stability of antibodies as determined by spectroscopic melting analysis or calorimetry. 4. Conclusion Here, we have shown that native agarose gel electrophoresis in 0.1 M His/0.1 M MES (or 0.2 M) buffer can be used to distinguish different mAbs and evaluate their purity. While all mAbs tested here showed migration toward the cathode, a majority of HEK293 cell contaminants were either acidic or neutral, clearly separated from the mAb band. The current agarose gel electrophoresis can also be used to see protein profile in native state of various sera and detect aggregation/ degradation. Fig. 9. Native gel electrophoresis of various sera on 5% agarose flat-gel in 0.1 M his/0.1 M MES. Run time, 2 h at 100 V. Fig. 10. Native gel electrophoresis of heat-treated BSA and mAb-14 on 1% flat-gel in 0.1 M His/0.1 M MES. Run time, 1 h at 80 V. C. Li et al. / International Journal of Biological Macromolecules 151 (2020) 885–890 889 CRediT authorship contribution statement Cynthia Li: Visualization. Teruo Akuta: Writing - review & editing. Masataka Nakagawa: Visualization. Tomomi Sato: Visualization. Takashi Shibata: Visualization. Toshiaki Maruyama: Supervision, Project administration, Visualization. C.J. Okumura: Funding acquisition. Yasunori Kurosawa: Validation, Visualization. Tsutomu Arakawa: Funding acquisition. Yasunori Kurosawa: Conceptualization, Project administration, Writing-original draft, Writing-review & editing. References [1] C. Li, T. Arakawa, Agarose native gel electrophoresis of proteins, Int. J. Biol. Macromol. 140 (2019) 668–671. [2] T. McLellan, Electrophoresis buffers for polyacrylamide gels at various pH, Anal. Biochem. 126 (1982) 94–99. [3] D.P. Goldenberg, Analysis of protein conformation by gel electrophoresis, in Protein Structure A Practical Approach Second Edition, T.E. Creighton, ed., pp187–218, IRL Press, Oxford. [4] R. Bansal, S. Gupta, A.S. Rathore, Analytical platform for measuring aggregation of monoclonal antibody therapeutics, Pharm. Res. 36 (2019)(152-xxx). [5] L.A. Hassan, M.A. Al-Ghobashy, S.S. Abbas, Evaluation of the pattern and kinetics of degradation of adalimumab using a stability-indicating orthogonal testing protocol, Biomed. Chromatogr. 33 (2019), e4676. [6] J. Liu, T. Eris, C. Li, S. Cao, S. Kuhns, Assessing analytical similarity of proposed Amgen biosimilar ABP 501 to adalimumab, BioDrugs 30 (2016) 321–338. [7] H.L. Svilenow, A. Kulakova, M. Zalar, A.P. Golovanov, P. Harris, G. Winter, Orthogonal techniques to study the effect of pH, sucrose, and arginine salts on monoclonal antibody physical stability and aggregation during long-term storage, J. Pham. Sci. 109 (2020) 584–594. [8] C. Nowak, J.K. Cheung, S.M. Dekkatore, A. Katiyar, R. Bhat, J. Sun, G. Ponniah, A. Neill, B. Mason, A. Beck, H. Liu, Forced degradation of recombinant monoclonal antibodies: a practical guide, MABS 9 (2017) 1217–1230. [9] C. Temporini, R. Colombo, E. Calleri, S. Tengattini, F. Rinaldi, G. Massolini, Chromatographic tools for plant-derived recombinant antibodies purification and characterization, J. Pham. Biomed. Anal. 129 (2019)(xxx-xxx). [10] A. Lechner, J. Giorgetti, R. Gahoual, A. Beck, E. Leize-Wagner, Y.N. Francois, Insights from capillary electrophoresis approaches for characterization of monoclonal antibodies and antibody drug conjugates in the period 2016-2018, J. Chromatogr. B Anal. Biomed. Life Sci. 1122-1123 (2019) 1–17. [11] D. Suba, Z. Urbányi, A. Sargό, Method development and qualification of capillary zone electrophoresis for investigation of therapeutic monoclonal antibody quality, J. Chromatogr. B Anal. Technol. Biomed. Life Sci. 1032 (2016) 224–229. [12] B.L. Duivelshof, S. Fekete, D. Guillarme, V. D’Atri, A generic workflow for the characterization of therapeutic monoclonal antibodies-application to daratumumab, Anal. Bioanal. Chem. 411 (2019) 4615–4627. [13] B. Yang, W. Li, H. Zhao, A. Wang, Y. Lei, Q. Xie, S. Xiong, Discovery and characterization of CHO host cell-protease-induced fragmentation of a recombinant monoclonal antibody during production process development, J. Chromatogr. B Anal. Technol. Life Sci. 1112 (2019) 1–10. [14] M. Pathak, D. Dutta, A. Rathore, Analytical QbD: development of a native gel electrophoresis method for measurement of monoclonal antibody aggregates, Electrophoresis 35 (2014) 2163–2171. [15] Y. Durocher, S. Perret, A. Kamen, High-level and high-throughput recombinant protein production by transient transfection of suspension-growing human 293-EBNA1 cells, Nucleic Acids Res. 30 (2002) E9. [16] H. Shimizu, M. Nakagawa, N. Todaka, K. Imaizumi, Y. Kurosawa, T. Maruyama, C.J. Okumura, T. Shibata, Y. Tanaka, Y. Sato, Y. Ono, T. Akuta, Improving the quality of a recombinant rabbit monoclonal antibody against PLXDC2 by optimizing transient expression conditions and purification method, Protein Expr. Purif. 146 (2018) 27–33. [17] A.L. Tscheliessnig, J. Konrath, R. Bates, A. Jungbauer, Host cell protein analysis in therapeutic protein bioprocessing-methods and applications, Biotechnol. J. 8 (2013) 665–670. [18] J. Zhu-Shimoni, C. Yu, J. Nishihara, R.W. Wong, F. Gunawan, M. Lin, D. Krawitz, P. Liu, W. Sandval, M. Vamderlan, Host cell protein testing by ELISAs and the use of orthogonal methods, Biotechnol. Bioeng. 111 (2014) 2367–2379. [19] H. Falkenberg, D.M. Waldera-Lupa, M. Vanderlaan, T. Schwab, K. Krapfenbauer, J.M. Studts, T. Flad, T. Waerner, Mass spectrometric evaluation of upstream and downstream process influences on host cell protein patterns in biopharmaceutical products, Biotechnol. Prog. 35 (2019), e2788. [20] D.G. Bracewell, R. Francis, C.M. Smales, The future of host cell protein (HCP) identification during process development and manufacturing linked to a risk-based management for their control, Biotechnol. Bioeng. 112 (2015) 1727–1737. 890 C. Li et al. / International Journal of Biological Macromolecules 151 (2020) 885–890