Bulk RNA sequencing of A375, BT549, SKBR3 and 315A cell lines [cell line bulk RNA-seq]. Bulk RNA sequencing of A375, BT549, SKBR3 and 315A cell lines [cell line bulk RNA-seq]

We first validated the robustness and high throughput capability of the GTO protocol in multiple cell lines. The key step to success of GTO is determining the optimum number of cycles for first amplification of cDNA. This step needs to be result in a balance between cDNA and gDNA reads so that the gene expression profiles generated are good while not comprimising on the CNA profile quality. This validation was done in SKBR3, A375 and BT549 cell lines. They were then compared to 315A control diploid cell line also processed by the GTO method. The scRNAseq and scDNAseq data generated by GTO was alsocompared to bulk RNAseq and DNAseq data generated in triplicates for each cell line by traditional methods. After the method was tested in cell lines, we next applied the method to mouse models to understand tumor biology. The mouse model we used here is an orthotopic transplantation model of pancreatic cancer in syngenic mice. EGFP positive KPC 1199 tumor cells were injected into the pancreas of adult wild-type mice to generate primary tumors in pancreas and metastases in liver, spleen, peritoneum and kidneys. These primary tumors and mets were then extracted and dissociated into single cell suspensions. The single cell suspensions were analyzed by flow cytometry with antibodies against CD45, Epcam and for endogeneous GFP to isolate single stromal and tumor cells. These cells were then processed by GTP protocol to analyze gDNA by CNV profiles and RNA through read counts from the same cell. To validate our data and also have controls, we generated scRNAseq gene expression data and scDNAseq CNA profiles from the 2D cells at the same passage number at which they were injected into the mice. The scRNAseq data was generated using Fluidigm C1 machine on theie medium IFC for mRNA seq according to user manual instructions. The scDNAseq data for CNA profiles was generated by sorting single nuclei into PCR plates and then using SeqXE kit from Sigma Aldrich for Whole genome amplification using their instructions again. The data analysis was done similar to scRNAseq or scDNAseq data from GTO. Overall design: In this study we describe a new method for GTO (Genomics Transcriptomics One-Tube) that allows for simeltaneous sequencing of RNA and DNA from a single cell in the same tube. This method relies on dual amplification. The first amplification is done immediately after the mRNA is transcribed to cDNA using SMART-Seq2 protocol (Picelli et al., 2014) to increase the level of cDNA to that comparable to gDNA. The second amplification with universal primers from SeqXE kit for whole genome amplification (from Sigma Aldrich) amplifies the cDNA and gDNA fragments to amounts required for library preparation. The reads generated from this mixed library are separated bioinformatically to exonic and nonexonic reads depending on where they align in the reference genome to generate the RNAseq and DNAseq data respectively. We present data generated from single in-vitro cells of multiple cell lines (A375, BT549, SKBR3, 315A) and also from single in-vivo cells isolated from tumors of an orthotopic tranplantation mouse model of pancreatic cancer (using KPC1199-EGFP cell line).

我们首先在多种细胞系中验证了GTO实验方案的稳定性与高通量性能。GTO实验成功的关键步骤，在于确定cDNA（互补脱氧核糖核酸，complementary DNA）首轮扩增的最优循环次数。该步骤需实现cDNA与gDNA（基因组DNA，genomic DNA）测序reads的平衡，以在保证生成的基因表达谱质量的同时，不损害CNA（拷贝数变异，Copy Number Alteration）谱的质量。该验证在SKBR3、A375及BT549细胞系中完成，随后将其与同样经GTO方法处理的315A对照二倍体细胞系进行比对。由GTO方法获得的scRNAseq（单细胞RNA测序，single-cell RNA sequencing）与scDNAseq（单细胞DNA测序，single-cell DNA sequencing）数据，与各细胞系经传统方法三次生物学重复获得的bulk RNAseq（批量RNA测序，bulk RNA sequencing）及DNAseq数据进行比对。

在细胞系中完成方法验证后，我们接下来将其应用于小鼠模型，以探究肿瘤生物学特性。本研究使用的小鼠模型为同基因小鼠胰腺癌原位移植模型。将EGFP（增强型绿色荧光蛋白，Enhanced Green Fluorescent Protein）阳性的KPC 1199肿瘤细胞注射至成年野生型小鼠的胰腺，以在胰腺内形成原发肿瘤，并在肝脏、脾脏、腹膜及肾脏中形成转移灶。随后提取这些原发肿瘤与转移灶，并解离为单细胞悬液。通过流式细胞术，使用针对CD45、Epcam的抗体及内源性GFP标记，对单细胞悬液进行分析，以分离单个基质细胞与肿瘤细胞。随后将这些细胞经GTO实验方案处理，通过CNV（拷贝数变异，Copy Number Variation）谱分析gDNA，并通过同一细胞的测序reads计数分析RNA。

为验证数据并设置对照，我们从与注射至小鼠体内的同一传代的2D培养细胞，生成了scRNAseq基因表达数据及scDNAseq CNA谱。scRNAseq数据按照用户手册说明，使用Fluidigm C1设备及其Medium IFC芯片进行mRNA测序。针对CNA谱的scDNAseq数据，通过将单个细胞核分选至PCR板，再使用Sigma Aldrich旗下的SeqXE试剂盒进行全基因组扩增，并遵循其说明书操作。数据分析方式与GTO方法获得的scRNAseq及scDNAseq数据一致。

实验整体设计：本研究介绍了一种新型GTO（Genomics Transcriptomics One-Tube，基因组转录组单管法）方法，可在同一管内对单个细胞的RNA与DNA同时进行测序。该方法依赖双重扩增：首轮扩增在mRNA反转录为cDNA后立即启动，采用SMART-Seq2实验方案（Picelli等，2014），以将cDNA的丰度提升至与gDNA相当的水平；第二轮扩增使用来自Sigma Aldrich的SeqXE试剂盒的通用引物，对cDNA与gDNA片段进行扩增，直至达到文库制备所需的起始量。通过生物信息学方法，根据测序reads在参考基因组上的比对位置，可将混合文库产生的reads分为外显子reads与非外显子reads，分别生成RNAseq与DNAseq数据。本研究提供了多种细胞系（A375、BT549、SKBR3、315A）的单细胞体外数据，以及从胰腺癌原位移植小鼠模型的肿瘤组织中分离的单细胞体内数据（使用KPC1199-EGFP细胞系）。