单个癌症的成纤维细胞亚群跟泛癌差异很大吗

Posted on by

交流群有小伙伴希望我们《生信菜鸟团》的单细胞周更专辑作者发现这个新鲜出炉的肝癌单细胞文章，标题是：《CD36+ cancer-associated fibroblasts provide immunosuppressive microenvironment for hepatocellular carcinoma via secretion of macrophage migration inhibitory factor》，简单的看了看里面的图表，就是降维聚类分群后，针对肝癌里面的成纤维细胞亚群进行细分和讲解，数据集链接是；

可以看到是 7 HBV-related HCC tissues and 4 adjacent liver tissues 的总共11个单细胞转录组样品，样品具体信息如下所示：

GSM6127499 Human hepatocellular carcinoma tissues [2a-0706]
GSM6127500 Human hepatocellular carcinoma tissues [1a-0618]
GSM6127501 Human hepatocellular carcinoma tissues [1a-0629]
GSM6127502 Human hepatocellular carcinoma tissues [1a-0706]
GSM6127503 Human hepatocellular carcinoma tissues [1a-0707]
GSM6127504 Human hepatocellular carcinoma tissues [1a-0727]
GSM6127505 Human hepatocellular carcinoma tissues [1a-0831]
GSM6127506 Human adjacent liver tissues [1b-0618]
GSM6127507 Human adjacent liver tissues [2b-0618]
GSM6127508 Human adjacent liver tissues [X-1b]
GSM6127509 Human adjacent liver tissues [B-2b]

但是它给出来矩阵的时候是混合的，其实也很简单就可以读取的，我们《生信菜鸟团》的单细胞周更专辑作者分享过好几次了基础文件读取技巧啦，详见： 读取不同格式的单细胞转录组数据及遇到问题的解决办法

GSE202642_barcodes.tsv.gz 569.6 Kb 
GSE202642_features.tsv.gz 325.6 Kb 
GSE202642_matrix.mtx.gz 666.5 Mb

可以看到作者的第一层次降维聚类分群是9大亚群 ：

  • epithelial and tumor cells (5059 cells, 5.6%, marked with EPCAM, ALDH1A1 and ALB)
  • B cells (1332 cells, 1.5%, marked with MS4A1 and CD79A)
  • T cells (24895 cells, 27.5%, marked with CD3D and CD3E)
  • natural killer (NK) cells (13868 cells, 15.3%, marked with FGFBP2 and FCG3RA)
  • myeloid-derived suppressor cells (MDSCs) (11087 cells, 12.2%, marked with ITGAM and CD33)
  • monocytes or macrophages (9978 cells, 11.0%, marked with CD68, CD163, and CD14)
  • dendritic cells (7784 cells, 8.6%, marked with ITGAX)
  • fibroblasts (2495 cells, 2.8%, marked with ACTA2 and COL1A2)
  • endothelial cells (14074 cells, 15.5%, marked with PECAM1 and vWF;

同样的是肝癌，并没有出现NK细胞占比超50%的情况，但是前面我们分享的 SingleR说是NK细胞你就相信了吗，数据集GSE162616里面的3个样品的单细胞数据，就Cells in clusters 0, 1, 2, 3, 7, 9, 10, 12, and 14 were annotated as NK cells based on SingleR ，也就是说占比超过了50%，要么是数据集GSE162616有问题，要么是这个数据集GSE162616的数据挖掘文章作者处理有问题。

前面的数据集GSE202642的第一层次降维聚类分群是9大亚群可视化如下所示：

9大亚群可视化

文章的分析很简单，，就是降维聚类分群后，针对肝癌里面的成纤维细胞亚群进行细分和讲解：

成纤维细胞亚群进行细分和讲解

可以看到作者把肝癌里面的成纤维细胞亚群分成了5个并且进行了功能性讲解，也就是常规的go或者kegg数据库注释啦， 就有了5个亚群的名字（vCAFs, mCAFs, lpmCAFs (CD36+ CAFs), lpCAFs and apCAFs,），如下所示：

常规的go或者kegg数据库

而且使用了小鼠单细胞数据集去验证了：

  • To further discern fibroblast heterogeneity, we then sorted 1746 CAFs in our scRNA-seq analyses derived from 7 murine HCC tumors and clustered the CAFs into 7 subpopulations

当然了，因为物种不一样，所以肝癌病人的单细胞跟小鼠是有差异，但是绝大部分情况还是类似的：

  • we found that all 5 CAF subtypes identified in human HCC tumors were also represented in mouse HCC tissues
  • the 5 mouse CAF subsets possessed populations and gene signatures similar to those of human CAFs, including vCAFs, mCAFs, CD36+CAFs, lpCAFs and apCAFs

早期就有很多其它癌症也有成纤维单细胞亚群解析

甚至有泛癌层面，比如2022年底的：《Pan-cancer single-cell analysis reveals the heterogeneity and plasticity of cancer-associated fibroblasts in the tumor microenvironment》，就是对 10 种实体癌类型的 226 个样本进行泛癌分析，以单细胞分辨率分析 TME，说明异质癌症相关成纤维细胞 (CAF) 的共性/可塑性：

  • 148 primary tumor, 53 adjacent, and 25 normal samples from 164 donors enrolled in 12 studies
  • 855,271 cells from 226 samples were included and integrated based on a batch effect correction algorithm
  • 34 TME-related clusters (c1-c34, 569,759 cells),
    • fibroblasts (c1-c8, DCN, COL1A1),
    • lymphocytes (c9-c17, CD3D, CD3E),
    • myeloid cells (c18-c24, CD68, CD14),
    • endothelial cells (c25-c28, VWF, PECAM1),
    • plasma cells (c29-c34, IGHG1, JCHAIN)

也有具体的癌症层面的单细胞数据挖掘文章，比如同样是2023发表的数据挖掘文章：《Cancer-associated fibroblast classification in single-cell and spatial proteomics data》，分析了已经发表的来自14名乳腺癌患者肿瘤组织的约11万细胞里面的16,704个基质细胞，分群12数量亚群后合并为10个生物学意义亚群，分别是(9 CAF types and one cluster of pericytes) ，最后的单细胞亚群的各自的基因如下所示：

  • RGS5 for pericytes
  • MCAM for vCAFs
  • MME (together with NDRG1 and ENO1) for tCAFs
  • IDO1 for ifnCAFs
  • HSPH1 for hsp_tCAFs
  • MMP11 (together with COL1A1 and POSTN) for mCAFs
  • PLA2G2A, CDF and CD34 for iCAFs
  • MKI67 for dCAFs
  • CCL21 (and CCL19) for rCAFs
  • CD74 and HLA-DR for apCAFs
  • PDPN and FAP for the general category of activated CAFs

而且也是在多个癌症里面的检验了，可视化如下所示：

单细胞亚群的各自的基因

所以最后有成纤维单细胞亚群规范命名标准吗

可以看到，每个癌症的单细胞数据集分析结果，多个癌症整合的结果，不同的物种，其实最后得到的单细胞亚群非常难以对比。但是如果是我们说b细胞和t细胞，就没有癌症没有物种的考虑，当然了如果是具体到b细胞里面的细分亚群，或者t细胞里面的细分亚群，也会面临同样的问题就是跨越数据集就很难去比较了。

​

 

Comments are closed.