工艺参数对抗CD3×Claudin 18.2双特异性抗体聚体形成的影响

Effect of Process Parameters on the Formation of Anti-CD3 and Claudin 18.2 Bispecific Antibody Aggregates

近年来,随着多种双特异性抗体药物的成功上市和基因工程技术的发展,双特异性抗体已成为抗体领域的研究热点。由于双特异性抗体结构复杂,其在生产过程中往往更容易产生聚体,不仅为下游纯化增加了难度,也提高了实验成本。为减少生产过程中聚体的产生,以表达抗CD3×Claudin 18.2双特异性抗体的CHO-K1细胞为研究对象,在2 L生物反应器上利用试验设计(Design of Experiment,DOE)法探究多种工艺参数对CHO-K1细胞合成抗CD3×Claudin 18.2双特异性抗体过程中聚体形成的影响,建立了工艺参数与聚体含量间的模型,确定了优化的工艺参数控制范围,并在200 L生物反应器中对优化的工艺条件进行了验证。DOE结果显示:pH、补料浓度和碳源浓度对聚体形成影响显著且均呈负效应。当pH为7.0~7.2、补料浓度为5.5%和碳源浓度为8~10 g/L时,200 L生物反应器中的验证结果显示聚体的含量为(27.62±4.55)%,与模型预测的结果具有一致性。

In recent years,with the successful marketing of a variety of bispecific antibody drugs and the development of genetic engineering technology,bispecific antibody has become a popular research topic in the antibody field.Due to the complex structure of bispecific antibodies,aggregates are often more likely to form during the manufacturing process,which not only increases the difficulty of downstream purification,but also increases the cost of experiments.In order to reduce bispecific antibody aggregates in the production process,CHO cells expressing anti-CD3 and Claudin 18.2 bispecific antibodies were used as the research object.Design of experiment(DOE)approach was used in a 2 L bioreactor to investigate the influence of various process parameters on the formation of aggregates during the production of CD3 and Claudin 18.2 bispecific antibodies.The model between process parameters and aggregate content was established,the control range of optimized process parameters was determined,and the optimized process conditions were verified in a 200 L bioreactor.The results of the DOE approach show that pH,feed concentration and carbon source concentration have significant and negative effects on aggregate formation.When the pH is 7.0-7.2,the feed concentration is 5.5%,and the carbon source concentration is 8-10 g/L,the verification results in the 200 L bioreactor show that the aggregate content is(27.62±4.55)%,which is consistent with the results predicted by the model.