Collaborative Research: EAGER/Tools4Cells: Translating single cell data into an ultra-high resolution spatial map using fluorescent marker genes

分子生物学的最新技术进步现在使确定多细胞生物体中每个单独细胞内的信使RNA转录本的整个群体成为可能。这项技术被称为单细胞RNA测序(scRNA-seq)，具有在植物中广泛应用的潜力，以更好地了解植物的发育、进化和逆境响应。在本项目中，将以植物根作为模型器官，构建一个超高分辨率的3D模型，显示该模型中嵌入的单个细胞的基因表达数据。使用这种方法，用户可以在植物器官中定位由荧光标记标记的细胞，并确定标记和未标记细胞中数千个基因的表达水平。该工具还可用于组合来自不同报告基因的荧光图像，以了解标记基因和同一样本中表达的其他基因的基因表达的相似性和差异性。这项工作的更广泛影响包括研究结果的内在价值，这些成果将通过植物细胞图集(PCA)传播给广泛的研究界。这些结果将包括收集图像数据的协议，构建3D图像的计算管道，以及在植物根的概念模型中注释和分配细胞类型的方法。图像分析和机器学习的计算管道将与详细的文件和用户手册以及同行审查的出版物一起存放到公共储存库。将为研究生和博士后提供研究培训，并将通过与弗吉尼亚州立大学的合作，为VSU学生开发高级基因组数据分析培训讲习班。连接单个细胞的空间位置和每个细胞内的基因表达模式是植物细胞生物学研究的前沿。目前可用的scRNA-seq协议不能保存每个细胞的空间位置，而使用嵌入组织的物理切片的空间转录组方法分辨率有限。这个迫切的项目的目标是建立一种新的植物空间转录组分析方法。植物研究界的一个主要资源是过去几十年积累的大量转基因报告基因系(如启动子-GFP系)。该项目将利用这一大型报告基因资源，使用相同的GFP标记系进行成像和scRNA-seq实验的原理证明研究。以植物根分生组织区域为模型系统，我们将生成选定启动子-GFP标记系的scRNA-seq数据，并将应用机器学习模型来准确预测GFP+和GFP-细胞。荧光成像和语义标记将用于合并和建模3D根部图像和GFP表达。最后，将开发一种机器学习方法，将scRNA-seq数据映射到3D根模型。这一新方法的结果将与现有数据进行比较，并将在PLANTA中进行验证。总而言之，这项工作将为开发任何植物物种的3D表达模型提供一种强大的新方法。这种方法的结果可以用来解决与发育过程中的不对称基因表达和根以及植物其他组织或器官的胁迫反应有关的问题。这个项目是由分子和细胞科学部门(细胞动力学和功能计划)和综合组织系统(生理机制和生物力学计划)共同资助的，这两个部门都是生物科学总监。这个奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

Recent technical advances in molecular biology now make it possible to determine the entire population of messenger RNA transcripts within each individual cell of a multicellular organism. This technology, known as single-cell RNA sequencing (scRNA-seq), has the potential to be broadly applied in plants to better understand their development, evolution, and stress responses. In this project, the plant root will be used as a model organ to construct an ultra-high-resolution 3D-model displaying gene expression data from individual cells embedded in this model. Using this method, a user can locate cells labeled by a fluorescent marker in a plant organ and determine the expression levels of thousands of genes in both labeled and unlabeled cells. This tool can also be used to combine fluorescent images from different reporter genes to understand the similarity and differences of gene expression for both the marker gene and other genes expressed in the same sample. The Broader Impacts of the work include the intrinsic merit of the research results, which will be disseminated to the broad research community via the Plant Cell Atlas (PCA). These results will include protocols for collecting image data, a computational pipeline for constructing 3D images, and a method to annotate and assign cell types in a conceptual model of plant roots. The computational pipeline for image analysis and machine learning will be deposited to public repository with detailed documentation and user manuals and peer-reviewed publications. Research training will be provided to graduate students and a post-doc and, through a collaboration with Virginia State University, training workshops will be developed for advanced genomic data analysis for VSU students.Connecting spatial location of individual cells and gene expression patterns within each cell is the frontier of plant cell biology research. Currently available scRNA-seq protocols do not preserve spatial locations of each cell, whereas spatial transcriptome approaches using physical slices of embedded tissues have limited resolution. The goal of this EAGER project is to establish a new approach for spatial transcriptome analysis in plants. One major resource from the plant research community is a large number of transgenic reporter gene lines (e.g. promoter-GFP lines) that have been accumulated for the past several decades. This project will leverage this large reporter gene resource to perform a proof of principle study using the same GFP marker lines for both imaging and scRNA-seq experiments. Using the meristematic region of plant roots as our model system, scRNA-seq data for selected promoter-GFP marker lines will be generated and machine learning models will be applied to accurately predict GFP+ and GFP- cells. Fluorescent imaging and sematic labeling will be used to merge and model 3D root images and GFP expression. Finally, a machine learning method will be developed to map the scRNA-seq data to the 3D root model. Results from this new approach will be compared with existing data and will be validated in planta. Together, this work will provide a powerful new approach to develop 3D expression models for any plant species. Results from this method can be used to address questions related to asymmetrical gene expression in development and stress responses in roots, as well as in other tissues or organs in plants. This project is jointly funded by the Divisions of Molecular and Cellular Sciences (Cellular Dynamics and Function program) together with Integrative Organismal Systems (Physiological Mechanisms and Biomechanics program) , both in the Biological Sciences Directorate.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.