Unthan Simon, Radek Andreas, Wiechert Wolfgang, Oldiges Marco, Noack Stephan
Institute of Bio- and Geosciences, IBG-1: Biotechnology, Systems Biotechnology, Forschungszentrum Jülich, 52425, Jülich, Germany.
Microb Cell Fact. 2015 Mar 11;14:32. doi: 10.1186/s12934-015-0216-6.
The throughput of cultivation experiments in bioprocess development has drastically increased in recent years due to the availability of sophisticated microliter scale cultivation devices. However, as these devices still require time-consuming manual work, the bottleneck was merely shifted to media preparation, inoculation and finally the analyses of cultivation samples. A first step towards solving these issues was undertaken in our former study by embedding a BioLector in a robotic workstation. This workstation already allowed for the optimization of heterologous protein production processes, but remained limited when aiming for the characterization of small molecule producer strains. In this work, we extended our workstation to a versatile Mini Pilot Plant (MPP) by integrating further robotic workflows and microtiter plate assays that now enable a fast and accurate phenotyping of a broad range of microbial production hosts.
A fully automated harvest procedure was established, which repeatedly samples up to 48 wells from BioLector cultivations in response to individually defined trigger conditions. The samples are automatically clarified by centrifugation and finally frozen for subsequent analyses. Sensitive metabolite assays in 384-well plate scale were integrated on the MPP for the direct determination of substrate uptake (specifically D-glucose and D-xylose) and product formation (specifically amino acids). In a first application, we characterized a set of Corynebacterium glutamicum L-lysine producer strains and could rapidly identify a unique strain showing increased L-lysine titers, which was subsequently confirmed in lab-scale bioreactor experiments. In a second study, we analyzed the substrate uptake kinetics of a previously constructed D-xylose-converting C. glutamicum strain during cultivation on mixed carbon sources in a fully automated experiment.
The presented MPP is designed to face the challenges typically encountered during early-stage bioprocess development. Especially the bottleneck of sample analyses from fast and parallelized microtiter plate cultivations can be solved using cutting-edge robotic automation. As robotic workstations become increasingly attractive for biotechnological research, we expect our setup to become a template for future bioprocess development.
近年来，由于先进的微升规模培养设备的出现，生物工艺开发中培养实验的通量大幅提高。然而，由于这些设备仍需要耗时的手工操作，瓶颈仅仅转移到了培养基制备、接种以及最终培养样品的分析上。在我们之前的研究中，通过将BioLector嵌入机器人工作站，朝着解决这些问题迈出了第一步。这个工作站已经能够优化异源蛋白生产过程，但在对小分子生产菌株进行表征时仍然存在局限性。在这项工作中，我们通过整合更多的机器人工作流程和微孔板分析，将我们的工作站扩展为一个多功能的微型中试工厂（MPP），现在它能够对广泛的微生物生产宿主进行快速准确的表型分析。
建立了一个全自动收获程序，该程序根据单独定义的触发条件，从BioLector培养物中反复对多达48个孔进行采样。样品通过离心自动澄清，最后冷冻以备后续分析。在MPP上集成了384孔板规模的灵敏代谢物分析，用于直接测定底物摄取（特别是D-葡萄糖和D-木糖）和产物形成（特别是氨基酸）。在第一个应用中，我们对一组谷氨酸棒杆菌L-赖氨酸生产菌株进行了表征，并能够快速鉴定出一株显示L-赖氨酸滴度增加的独特菌株，随后在实验室规模的生物反应器实验中得到了证实。在第二项研究中，我们在一个全自动实验中分析了一株先前构建的能够转化D-木糖的谷氨酸棒杆菌菌株在混合碳源培养过程中的底物摄取动力学。
所展示的MPP旨在应对早期生物工艺开发中通常遇到的挑战。特别是，利用前沿的机器人自动化技术，可以解决快速并行微孔板培养中样品分析的瓶颈问题。随着机器人工作站在生物技术研究中变得越来越有吸引力，我们期望我们的设置能够成为未来生物工艺开发中的一个模板。
