Evaluation of transcription factor knockout impact on paclitaxel response for Triple Negative Breast Cancer

Data and code related to Zenodo repository: 10.5281/zenodo.11238552 Two experimental formats included: 'fixed' prefix: data from terminal time point of siRNA screen applied to HCC1143, HCC1806, and MDA-MB-468 Triple Negative Breast Cancer cell lines. 'live' prefix: data from live-cell imaging of cell cycle reporter (HDHB-mClover/NLS-mCherry) HCC1143 Triple Negative Breast Cancer cell line. Note: 'live' level 1 data is available upon request ([e-mail], [e-mail]). Experimental goal: Evaluate whether siRNA knockdown of transcription factors elevated during paclitaxel response impact cell count, cell morphology or cycling dynamics. Methods: siRNA Knockdown: Cells were plated in 90ul of serum free media per well of a 96 well plate. 24 hours later, siRNA knockdown mixture was prepared using a cell-line optimized concentration of Lipofectamine RNAiMAX (cat 13778075-075, Invitrogen) and siRNA (Horizon Discovery ON-TARGETplus) following RNAiMAX recommended protocol. The final concentration of siRNA per well was 1pmol and the final volume of RNAiMAX per well was 75nL for HCC1143, and 37.5nL for HCC1806 or MDA-MB-468 in 100uL of cell containing volume. 24 hours after siRNA transfection cells were treated with an addition of 100uL complete media containing either DMSO vehicle control or paclitaxel. Fixed-cell assays: Cells were plated at 3000 cells in 100ul of complete media per well in a 96 well plate (#08-772-225, FisherScientific). After 24 hours, an additional 100ul of either vehicle (0.1% DMSO) or paclitaxel containing complete media was added. After 72 hours cells were fixed with 4% Formaldehyde (#28908, ThermoFisher Scientific) for 15 minutes at room temperature, then permeabilized with 0.3% Triton X-100 (#X100-100ML, Sigma Aldrich) for 10 minutes at room temperature, then washed twice with PBS. Fixed cells were then stained with 0.5ug/mL DAPI (4083S, Cell Signaling Technology) in PBS for 15 minutes at room temperature. Following DAPI staining, wells were washed once with PBS, then stained with 1:20,000 HCS CellMask Green in PBS (H32714, Invitrogen) for 15 minutes at room temperature. Wells were washed twice with room temperature PBS and then 4 fields of view per well imaged on an InCell 6000 (GE Healthcare). Images were segmented with two custom Cellpose models to segment the nucleus (using parameters: diameter = 50, chan = DAPI, chan2 = Cellmask Orange) and cytoplasm (using parameters: diameter = 90, chan = Cellmask Orange, chan2 = DAPI). Image quantification was performed in R (v4.3.1) using EBImage (v4.42.0), and cells were annotated based on the number of distinct nuclei segmented within each cytoplasmic mask. HDHB reporter live-cell assays: siRNA knockdown and drug treatment was performed as described above, and then the plate was loaded on an Incucyte S3 (Sartorious) and cells imaged every 15 minutes for 72 hours post drug treatment. At each timepoint 4 fields of view were captured at 20x magnificantion in each well using the phase, red and green channels. A cytoplasmic mask was computed from the mean of normalized red/green channel (cellpose parameters: diameter = 57, chan = mean(normalized(red), normalized(green)), and a nuclear mask was computed from the red channel (cellpose parameters: diameter = 30, chan = DAPI) using custom trained Cellpose models. Image quantification was performed in R (v4.3.1) using EBImage (v4.42.0). An additional perinuclear ring mask was computed as the 11 pixel dilation from the nuclear mask, but still bound by the cytoplasmic mask. To determine mClover localization thresholds for cell cycle assignment, 250 cell images were randomly selected and manually assigned to the G1, S/G2 or M cell cycle state based on mClover localization. The mClover intensity ratios were then used to determine thresholds for automated cell cycle phase calling which was applied to the rest of the data set (Supplemental Figure 5A). Mononuclear cells with a Perinuclear:Nuclear mean intensity ratio greater than 0.8 and Nuclear:Cytoplasmic total intensity less than 0.5 were assigned to the S/G2 phase. Mononuclear and Multinuclear cells with a Nuclear:Cytoplasmic total intensity ratio greater than 0.8 and Perinuclear:Nuclear mean intensity ratio less than 0.8 were assigned to the ‘M’ phase. The remainder of mononuclear cells were assigned ‘G1’, and the remainder of multinucleated cells were assigned ‘Multinucleated’. Included files: fixed_level_1-plate_#.zip : Six .zip archives containing the raw images (DAPI/CellMask/Brightfield) from fixed-cell experiments. plate 1: HCC1143 cells treated with plate A schema plate 2: HCC1143 cells treated with plate B schema plate 3: HCC1806 cells treated with plate A schema plate 4: HCC1806 cells treated with plate B schema plate 5: MDA-MB-468 cells treated with plate A schema plate 6: MDA-MB-468 cells treated with plate B schema fixed_level_2: Data quantified from cellpose masks at the single-nuclei level (redundant cytoplasm information) fixed_level_3: Data from 'fixed_level_2.csv' collapsed to the single cell level, including staining intensity and aggregate nuclear information fixed_incell_to_cellpose.rmd: R markdown code for converting original incell files (fixed_level_1) to RGB images for cellpose segmentation fixed_image_quantification.rmd: R markdown code for quantifying images using cellpose segmentation masks and original images (fixed_level_1) fixed_cellpose_models.zip: Archive including cellpose models used for fixed experiment live_level_2: Data quantified from cellpose masks at the single-nuclei level (redundant cytoplasm information) live_level_3: Data from 'live_level_2.csv' collapsed to the single cell level, including staining intensity and aggregate nuclear information live_level_4: Data from 'live_level_3.csv' collapsed to the single condition level summarizing the number, multinucleation status and phase of cells at each time point. live_image_quantification.rmd: R markdown code for quantifying images using cellpose segmentation masks and original images (live_level_1). l ive_incu_archive2rgb.rmd: R markdown code for converting incucyte archive formatted data into RGB images, where the blue channel is the arithmetic mean of the min-max (0-1) normalized red and green channels. live_cellpose_models.zip: Archive including cellpose models used for live experiment.本数据集关联Zenodo仓库（Zenodo repository）编号10.5281/zenodo.11238552的相关数据与代码。本数据集包含两类实验格式的数据：以'fixed'为前缀的数据来自针对三阴性乳腺癌细胞系HCC1143、HCC1806及MDA-MB-468的siRNA（小干扰RNA）筛选终端时间点样本；以'live'为前缀的数据来自HCC1143三阴性乳腺癌细胞系的细胞周期报告基因（HDHB-mClover/NLS-mCherry）活细胞成像数据。注意：'live'一级数据需通过邮件申请（联系人邮箱：[e-mail]、[e-mail]）。 实验目标：评估紫杉醇响应过程中上调的转录因子的siRNA（小干扰RNA）敲低是否会影响细胞计数、细胞形态或细胞周期动力学。 实验方法： 1. siRNA（小干扰RNA）敲低流程：将细胞以每孔90μL无血清培养基接种于96孔板。24小时后，按照Lipofectamine RNAiMAX（货号13778075-075，Invitrogen）推荐的实验流程，使用适配细胞系的浓度配制siRNA敲低混合液，所用siRNA为Horizon Discovery的ON-TARGETplus系列。每孔siRNA终浓度为1pmol；对于HCC1143细胞系，每孔RNAiMAX终体积为75nL，HCC1806与MDA-MB-468细胞系则为37.5nL，体系总体积为100μL（含细胞）。siRNA转染24小时后，向每孔加入100μL完全培养基，分别添加DMSO溶剂对照或紫杉醇。 2. 固定细胞实验：将3000个细胞以每孔100μL完全培养基接种于96孔板（货号08-772-225，FisherScientific）。24小时后，向每孔加入100μL含0.1% DMSO的溶剂对照或紫杉醇的完全培养基。培养72小时后，使用4%甲醛（货号28908，ThermoFisher Scientific）于室温固定细胞15分钟，随后用0.3% Triton X-100（货号X100-100ML，Sigma Aldrich）于室温透化10分钟，再用磷酸盐缓冲液（PBS）洗涤两次。固定后的细胞用0.5μg/mL DAPI（货号4083S，Cell Signaling Technology）于室温染色15分钟，染色完成后用PBS洗涤一次，再用1:20000比例稀释的HCS CellMask Green（货号H32714，Invitrogen）于室温染色15分钟。随后用室温PBS洗涤孔板两次，使用InCell 6000成像系统（GE Healthcare）对每孔采集4个视野的图像。图像分割采用两个自定义Cellpose模型：细胞核分割参数为直径=50、通道=DAPI、通道2=Cellmask Orange；细胞质分割参数为直径=90、通道=Cellmask Orange、通道2=DAPI。图像定量分析使用R语言（版本4.3.1）结合EBImage包（版本4.42.0）完成，并根据每个细胞质掩码内分割到的独立细胞核数量对细胞进行注释。 3. 活细胞成像实验：siRNA敲低与药物处理流程同上，随后将板放置于Incucyte S3成像系统（Sartorius）中，于药物处理后每15分钟成像一次，连续采集72小时。每个时间点以20倍放大倍率对每孔采集明场、红色通道与绿色通道共4个视野。使用自定义训练的Cellpose模型生成细胞质掩码与细胞核掩码：细胞质掩码由归一化后红色/绿色通道的均值计算得到（Cellpose参数：直径=57、通道=mean(normalized(red), normalized(green))）；细胞核掩码由红色通道计算得到（Cellpose参数：直径=30、通道=DAPI）。图像定量分析使用R语言（版本4.3.1）结合EBImage包（版本4.42.0）完成。额外生成核周环掩码：以细胞核掩码为基础向外扩张11像素，且不超出细胞质掩码范围。为确定mClover定位的细胞周期分配阈值，随机选取250张细胞图像，根据mClover定位手动将其划分为G1、S/G2或M细胞周期状态，随后利用mClover强度比确定自动化细胞周期阶段调用的阈值，并应用于剩余数据集（补充图5A）。具体分类规则如下：单核细胞若核周:细胞核平均强度比大于0.8且细胞核:细胞质总强度比小于0.5，则划分为S/G2期；单核及多核细胞若细胞核:细胞质总强度比大于0.8且核周:细胞核平均强度比小于0.8，则划分为M期；其余单核细胞划分为G1期，其余多核细胞划分为多核细胞。 包含文件说明： - "fixed_level_1-plate_#.zip"：共6个压缩归档文件，包含固定细胞实验的原始图像（DAPI/CellMask/明场）。各板对应样本如下： - 板1：经方案A处理的HCC1143细胞 - 板2：经方案B处理的HCC1143细胞 - 板3：经方案A处理的HCC1806细胞 - 板4：经方案B处理的HCC1806细胞 - 板5：经方案A处理的MDA-MB-468细胞 - 板6：经方案B处理的MDA-MB-468细胞 - "fixed_level_2"：以单细胞核水平量化得到的数据（包含冗余的细胞质信息） - "fixed_level_3"：将"fixed_level_2.csv"的数据聚合至单细胞水平，包含染色强度与汇总细胞核信息 - "fixed_incell_to_cellpose.rmd"：用于将原始InCell文件（fixed_level_1）转换为可供Cellpose分割的RGB图像的R Markdown代码 - "fixed_image_quantification.rmd"：使用Cellpose分割掩码与原始图像（fixed_level_1）完成图像定量分析的R Markdown代码 - "fixed_cellpose_models.zip"：包含固定细胞实验所用Cellpose模型的压缩归档文件 - "live_level_2"：以单细胞核水平量化得到的数据（包含冗余的细胞质信息） - "live_level_3"：将"live_level_2.csv"的数据聚合至单细胞水平，包含染色强度与汇总细胞核信息 - "live_level_4"：将"live_level_3.csv"的数据聚合至单处理条件水平，汇总了各时间点的细胞数量、多核状态与细胞周期阶段 - "live_image_quantification.rmd"：使用Cellpose分割掩码与原始图像（live_level_1）完成图像定量分析的R Markdown代码 - "live_incu_archive2rgb.rmd"：用于将Incucyte归档格式的数据转换为RGB图像的R Markdown代码，其中蓝色通道为最小-最大（0-1）归一化后的红色与绿色通道的算术均值 - "live_cellpose_models.zip"：包含活细胞实验所用Cellpose模型的压缩归档文件