文章信息
文章题目:Cross-tissue organization of the fibroblast lineage 发表在:Nature 日期:12 May 2021 链接:https://www.nature.com/articles/s41586-021-03549-5 作者:Genentech团队
- 基因泰克公司(英语:Genentech Inc.),全名基因工程科技公司(英语:Genetic Engineering Technology),是一家生物科技公司,由创投人罗伯特·史旺森(Robert A. Swanson)与生物化学家赫伯特·博耶(Herbert Boyer)博士于1976年成立,这件事也被视为是生物科技产业的起点。罗氏药厂于2009年3月26日以大约468亿美元完成了基因泰克的收购,并完全拥有此公司。截至2019年2月为止,基因泰克有超过13,697名员工。
- 还发了一个在线工具:https://www.fibroxplorer.com/ FibroXplorer is an interactive data portal that allows users to examine gene expression at the single-cell level in ~230,000 mouse fibroblasts from 17 tissues, 50 datasets and 12 disease states
摘要
- 很多研究都表明:成纤维细胞在单个组织中存在异质性,但是在不同健康和患病器官的组织中并没有证明
- 文章整合了2个物种的17个组织,涵盖11种疾病状态,50个数据集,得到了230,000个细胞scRNA数据
- 比较了人和小鼠的成纤维细胞图谱发现,成纤维细胞的转录状态是保守的
- 文章提出假设,成纤维细胞的异质性是由于稳态组织和疾病组织之间的波动差异导致的
实验设计
重点还是落在小鼠
- Mouse tissue digestion and stromal cell isolation or identification by FACS
- qPCR, RNA extraction and cDNA synthesis
- Mouse scRNA-seq and cell hashing
- Tissue processing for histology
- RNAscope in situ hybridization
- DSS-induced colitis
- Tumour inoculation
- Mouse bulk RNA-seq analysis
- Mouse bulk ATAC–seq analysis
- Human tissue digestion and stromal cell isolation
- Open chromatin regions (OCRs) identification
- Motif enrichment analysis
- ATAC–seq and RNA-seq concordance
- Mouse scRNA-seq meta-analysis
- Human scRNA-seq meta-analysis
- Pseudo-bulk analytical strategy
- Projection of human gene sets onto mouse perturbed-state atlas
首先看稳态小鼠组织的成纤维细胞
先做了bulk RNA和ATAC-Seq:
单细胞图谱构建思路(图a):
得到10个细胞群,marker分别是:Pi16 , Col15a1 , Ccl19 , Coch , Comp , Cxcl12 , Fbln1 , Bmp4 , Npnt and Hhip
- Ccl19 fibroblastic reticular cells (FRCs),
- Coch red pulp fibroblasts,
- Cxcl12 mesenchymal stromal cells and osteolineage cells,
- Fbln1 and Bmp4 intestinal fibroblasts,
- Comp fibroblasts,
- Npnt alveolar fibroblasts
- Hhip peribronchial fibroblasts
几乎所有的组织都有Pi16 and Col15a1 clusters
DEGs in the Pi16 cluster (Pi16, Dpp4 and Ly6c1) suggested an identity similar to adventitial stromal cells Col15a1 cluster exhibited an association with the basement membrane, evidenced by expression of Col4a1, Hspg2 and Col15a1
图d是:Slingshot lineage inference identified trajectories that emerged from the Pi16 cluster, passed through the Col15a1 cluster, and ended at specialized clusters
scRNA和bulkRNA的一致性还是很高的,说明scRNA在提高精度的同时,没有引入其他的技术误差:
结论是:在稳态小鼠组织中,Pi16 and Col15a1 的成纤维细胞亚群普遍存在,当然还有一些特异的亚群,它们之间可能存在进化的关联
用转基因小鼠进行验证
受伤或患病小鼠紊乱组织的成纤维细胞
组织的紊乱包括:感染、损伤、癌症、纤维化、代谢变化和关节炎
探索了17 publicly available scRNA-seq datasets across 13 tissues to generate a perturbed-state fibroblast atlas (n = 99,596 cells
也是得到10个cluster:Pi16 , Col15a1 , Ccl19 , Cxcl12 , Comp , Npnt , Hhip , Adamdec1 , Cxcl5 and Lrrc15
发现:Cxcl12 , Ccl19 , Comp , Npnt , and Hhip 这几个cluster在基因表达和组织分布和稳态组织的结果很相似;Cxcl5 , Adamdec1 and Lrrc15 这几个是紊乱组织特有的,激活了稳态中没有的细胞状态
人的成纤维细胞图谱
假设:小鼠在稳定和紊乱状态的成纤维细胞亚型方面可能与人类表现出一些相似性
使用3个胰腺癌病人的癌和癌旁样本:n = 21,262 cells (图a-c),分成了CAF (c3和c8)和normal fibroblasts
在c8的20个高表达基因中,有12个也在稳态中高表达:Dpt Pi16 (DPT, IGFBP5, IGFBP6, C3, APOD) or Dpt Col15a1 (CXCL12, SMOC2, C7, FBLN5, MFAP4, LUM, FMO2)
接着,拿C3 PDAC CAFs colon fibroblasts lung fibroblasts from individuals with non-small cell lung cancer (NSCLC), IPF or COVID-19 (n = 10,355 cells)进行整合,得到6个cluster
NPNT alveolar fibroblasts derived from individuals with lung disease ADAMDEC1 and CCL19 31 clusters primarily from colitis samples PI16 cluster, in which the human universal signature was most enriched two myofibroblast clusters defined by LRRC15 and COL3A1 expression:
- LRRC15 myofibroblasts were enriched in cells from individuals with pancreatic and lung cancer,
- COL3A1 cluster was enriched in cells from patients with COVID-19
进而验证了小鼠的图谱:LRRC15 myofibroblasts and ADAMDEC1 fibroblasts
另外还发现了COVID-19病人中存在一个独特的myofibroblast细胞群
最后留下疑问:It remains unclear why two universal Dpt fibroblast subtypes exist, though we speculate that this may represent a necessary division of labour within the lineage.
往期回顾
OSCA单细胞数据分析笔记8—Dimensionality reduction
多发性骨髓瘤发展过程中肿瘤和免疫细胞的共同进化
单细胞数据Seurat包的tSNE三维可视化
明码标价之10X技术单细胞(2.5万每个)(标准100G测序数据)
任意细胞亚群的差异分析
如果你对单细胞转录组研究感兴趣,但又不知道如何入门,也许你可以关注一下下面的课程
- 生信爆款入门-2021第4期
- 数据挖掘(GEO,TCGA,单细胞)2021第4期
- 明码标价之共享96线程384G内存服务器