High-throughput Robotic Microinjection System and Its Application in Constructing Gene-edited Macrophages with Enhanced Tumor-killing Ability
DescriptionRobotic microinjection has high accuracy at the single-cell level; thus, it could avoid immunogenicity, chemical toxicity, and high mortality. However, the process has the disadvantage of low throughput compared with other transfection methods. The throughput of most existing robotic microinjection systems ranges from several hundred to one thousand cells within a production cycle, far less than the number required for clinical use. Low throughput remains the bottleneck problem for successful implementation of robotic microinjection in practical applications.In this work, we will develop new high-throughput robotic microinjection technologies that will hit a record of injecting 4000 cells per hour, with a corresponding high success rate and survival rate for macrophages (e.g., > 80%). This high throughput could be achieved using a microinjection strategy of coordinated dual micromanipulator system (here denoted a dual module system) and deep learning-enhanced cell detection and segmentation. The flux of 4000 cells per hour is estimated on the basis of the time interval between two coordinated injections in the system, the time required for intelligent processing steps, and our previous research result on the single-module injection system (1500 cells per hour). Macrophages are important immune effector cells, and therefore they will be used as exemplary cells for our cell transfection studies by microinjection. We will conduct this research in three aspects. First, we will develop dual-module robotic microinjection system that uses two coordinated micromanipulators on the same cell processing platform to inject cells on the basis of injection sequence optimization to increase productivity. The two micromanipulators insert cells synergistically in an optimal injection path leading to substantially increased throughput. Second, we will develop an attentive single shot multi-box detector to detect cells and a U-Net framework to segment cells, which could considerably improve the operation efficiency and subsequently increase the throughput. The use of deep learning algorithms for cell detection and segmentation will avoid not only the monotonous preparation for cell staining but also the issues associated with dye toxicity and potential interference with injection samples. Third, we will apply the developed robotic microinjection system to inject gene-editing plasmids into macrophages and construct immune-enhanced cell lines with tumor-killing ability. We will study in-vitro and in-vivo tumor models to verify the antitumor effects of LR4-overexpressed macrophages.The successful realization of high-throughput injection will make robotic microinjection a competitive tool in many biomedical applications, such as plasmid DNA transfection, and promote the precise cell therapy in Hong Kong.
|Effective start/end date||1/01/22 → …|