Image-Seq: A high-density microfluidic trap array for single cell transcriptome analysis coupled with image based phenotyping

摘要 能够以大规模并行的方式将单细胞RNA测序和基于图像的表型分析相结合 将使单细胞的功能和基因表达之间建立直接的相关性。 然而，现有技术的根本局限性阻碍了这一目标的实现， 高通量自动化平台。在这里，我们建议通过创建一个 高密度陷阱阵列，其包含在陷阱阵列中的已知地址处印刷的独特DNA条形码。的 我们提出的技术，我们称之为Image-Seq，将涉及以高密度组织单个细胞， 阵列，在多个波长下以高分辨率对每个细胞/孔进行成像，最后制备单细胞用于 使用本地打印的DNA条形码作为可在NGS数据集中追踪的细胞标识符的RNA-seq 回到特定的活细胞图像。为了实现这一目标，杜克团队将与应用微阵列公司合作， 将在独特的空间地址打印一系列DNA条形码。目标1将重点展示DNA 条形码可以高保真打印并用作细胞mRNA逆转录的模板。 目标2将展示实现单细胞的高通量捕获、自动成像和 直接在微流控芯片内裂解单细胞。目标3将证明有能力获得两个活的 细胞图像和每个单细胞在混合的人类和小鼠细胞系中的转录组谱，以及在 分离的组织样本该项目在基础研究和后续研究中有许多潜在的应用- 临床应用。其中一个应用是更好地利用有限的样本， 可以从组织中获得数千个解离的单细胞，这是目前通过其他方法不可能的。 单细胞分析工作流程。另一个潜在的应用是药物敏感性测试，这涉及时间- lapse成像来量化单个细胞的生长速率，并将这些与这些细胞的基因表达分析进行比较。 同样的细胞，以牵连的信号通路所调用的耐药细胞。

ABSTRACT The ability to combine single cell RNA sequencing and image-based phenotyping in a massively parallel format would enable direct correlations to be made between the function and gene expression of single cells. However, the fundamental limitations of existing technologies have prevented the realization of this goal in a high-throughput automated platform. Here we propose to solve this systems design challenge by creating a high-density trap array that contains unique DNA barcodes printed at known addresses in the trap array. The proposed technology, which we refer to as Image-Seq, will involve organizing single cells in a high-density array, imaging each cell/well at high resolution at multiple wavelengths, and finally preparing the single cells for RNA-seq using the locally printed DNA barcodes as cellular identifiers that can be traced in NGS datasets back to specific live cell images. To achieve this goal, the Duke team will partner with Applied Microarray, who will print an array of DNA barcodes at unique spatial addresses. Aim 1 will focus on demonstrating that DNA barcodes can be printed with high fidelity and used as templates for reverse transcription of cellular mRNA. Aim 2 will demonstrate the ability to achieve high throughput trapping of single cells, automated imaging, and lysis of single cells directly inside the microfluidic chips. Aim 3 will demonstrate the ability to obtain both a live cell image and a transcriptome profile of each single cell in mixed human and mouse cell lines and also in dissociated tissue samples. This project has many potential applications both in basic research and in follow- on clinical applications. One application is in making better use of limited samples where only hundreds to thousands of dissociated single cells can be obtained from a tissue, which is not currently possible by other single cell analysis workflows. Another potential application is in drug sensitivity testing, which involves time- lapse imaging to quantify single cell growth rates and compare these to the gene expression analysis of those same cells in order to implicate the signaling pathways invoked by drug resistant cells.