Parallelised live microscopy for high-throughput behavioural phenotyping in malaria research

许多更小、更简单的显微镜通常可以看到不止一台大型、昂贵的机器。我们建议创新一种研究质量的、全自动的显微镜设计，这种设计可以为特定的实验量身定做，并可以很容易地复制来并行执行许多实验。有了计算机视觉来控制和分析这些实验，设备和工作人员时间的瓶颈被消除了，跟上新的基因技术成为可能--即使是以前耗时的研究，例如测量疟原虫对红细胞的入侵。我们将开发显微镜和计算机视觉技术，在我们的疟疾实验室和我们的合作者的实验室中展示它们的有效性，并发布允许其他人复制我们的进步的开源设计。高效地筛选数百种不同的突变株的能力将导致对许多疾病的更深入了解，最终创建新的药物发现目标，并可能导致针对疟疾等疾病的新疫苗。由于CRISPR-Cas9协议，基因编辑正处于一场革命之中，细胞表型需要跟上步伐。特别是在疟疾方面，我们的合作者现在正在使用CRISPR扩大恶性疟原虫的基因敲除规模，预计今年将生产200个基因敲除品系，未来五年将生产1000个。虽然菌株的产生规模如此之大，但表型分析却并非如此，即我们无法确定这些基因的功能-我们需要强大且廉价的可扩展表型分析，这涉及到活细胞成像，特别是宿主/病原体入侵的表型分析。通过目前的方法和技术进行这些分析是不可想象的。必须开发和部署新的、更加自动化和负担得起的成像方法。这将使我们能够系统地筛选GM系的几种表型，包括裂殖子数量、胞质分裂、出口和侵袭。我们在这里谈到疟疾的情况，但指出在许多其他传染病中可以确定非常相似的挑战和目标。

Many smaller, simpler microscopes can often see more than one large, expensive machine. We propose to innovate a research-quality, fully automated microscope design that can be tailored to a particular experiment and easily replicated to perform many experiments in parallel. With computer vision to control and analyse these experiments, the bottlenecks of equipment and staff time are removed, and it becomes possible to keep pace with new genetic technologies - even for previously time-consuming studies, for example measuring the invasion of red blood cells by plasmodium parasites. We will develop the microscopy and computer vision technologies, demonstrate their efficacy in our malaria lab and those of our collaborators, and release open-source designs that allow others to replicate our progress. The ability to screen hundreds of different mutant strains efficiently will lead to a deeper understanding of many diseases, ultimately creating new drug discovery targets and potentially leading to new vaccines for conditions like malaria. Gene editing is in the midst of a revolution thanks to CRISPR-Cas9 protocols, and cell phenotyping needs to keep pace. Specifically in malaria, our collaborators are now scaling up knockouts in P. falciparum using CRISPR, and expect to make 200 knockout lines this year, and 1000 in the next five years . While strain generation is scaling so dramatically , phenotyping is not, i.e. we cannot determine the function of these genes - we need robust and cheap scalable phenotyping assays, which involve live cell imaging, and specifically of host/pathogen invasions. It is not conceivable to perform these assays through current methods and technologies. New, much more automated and affordable approaches to imaging have to be developed and deployed. This would then allow us to systematically screen GM lines for several phenotypes, including merozoite number, cytokinesis, egress and invasion. We address here the case of malaria, but point out that very similar challenges and objectives can be identified in many other infectious diseases.