分享是一种态度
文章信息
文章题目:Multimodally profiling memory T cells from a tuberculosis cohort identifies cell state associations with demographics, environment and disease
作者:哈佛医学院 Soumya Raychaudhuri 团队 首先感觉他们实验室主页做的很不错:https://immunogenomics.hms.harvard.edu/lab
链接:https://www.nature.com/articles/s41590-021-00933-1
发表期刊:Nature Immunology (2021-05-24)
数据是公开的:GSE158769 ,采用CITE-seq
代码语言:javascript复制Platforms GPL11154 Illumina HiSeq 2000 (Homo sapiens)
Samples GSM4810298 Memory_T_cell_CITE-seq
GSE158769_exprs_norm.tsv.gz 3.0 Gb (ftp)(http) TSV
GSE158769_exprs_raw.tsv.gz 747.0 Mb (ftp)(http) TSV
GSE158769_meta_data.txt.gz 19.8 Mb (ftp)(http) TXT
Genotype data:phs002025.v1.p1
分析代码:https://github.com/immunogenomics/TB_Tcell_CITEseq
还发开了一个网页:https://immunogenomics.io/tbru/
摘要
- 分析了来自秘鲁的259个肺结核(tuberculosis , TB)病人的500,089个记忆T细胞,得到31种细胞状态
- 在排除年龄、性别、季节、遗传等因素影响后,发现了一种多能型type 17 helper T (TH17) 细胞状态,在结核杆菌发展成肺结核病的过程中丰度和功能都有所下降
方法
数据准备
- PBMC sample preparation
- Flow cytometry of total PBMCs
- CITE-seq of isolated memory T cells
- Bulk RNA sequencing
- Genotyping and genetic data processing
数据分析
- scRNA比对 定量
- scRNA数据拆分
- scRNA数据质控
- 降维
- 聚类 注释
- bulk RNA-seq 比对 定量
- 利用MASC(modeling of associations of single cells )评价不同细胞状态与疾病的关系
先用CITE-seq测了>500,000个T细胞
- 在秘鲁首都利马进行了一次大型的流行病学调查(n= 18,544),从其中招募了264个人,采样了外周血单核细胞进行CITE-seq测定,同时还加入了基因表型数据和蛋白数据
- 图b是进行了6次QC,每一次过滤掉一些细胞,最后得到了500K
- 大部分样本的细胞数量在2k左右
得到记忆T细胞的31种不同状态
利用canonical correlation analysis (CCA) 进行降维,选择top20 canonical variates (CVs)进行批次校正和聚类,根据基因和蛋白marker得到31个细胞状态
- 23/31 were CD4 ; five were CD8 ; one (C-24) was a mixture; Two clusters (C-30 and C-31) were CD4−CD8−
- C-20:contained a subset of CD4−CD8− and CD8 cells expressing innate-like T cell markers, including ZBTB16 and CD161 and CD26 surface proteins
- one central memory cluster (C-25) and distinct GZMK (C-28) and GZMB (C-29) effector subsets, reflecting different cytotoxicities
- high expression of HLA-DR and CD38 surface protein and proliferation-associated MKI67 (C-15 and C-27) represent chronically activated cells
- TH17 cells (C-12, CCR6 and RORC)
- type 1 helper T cells (TH1; C-17, CXCR3 and IFNG and TBX21)
- heterogeneous continuum of intermediate TH1/TH17 states (C-13, C-16 and C-19) with varying degrees of CXCR3, CCR6, CCR5 and CD161 surface protein expression and RORC and TBX21 expression
- CD161 subset of type 2 helper T (TH2) cells (C-14), described as pathogenic, with higher expression of allergy-associated HPGDS and IL17RB
- a subset of FOXP3 Treg cells (C-5) expressing higher CCR6 surface protein and CTLA4 and RORC than other Treg cells
记忆T细胞的状态因人口特征和环境而异
发现:年龄因素的影响最大;性别因素与T细胞状态也高度相关;type 2 helper T (TH2) 细胞状态在冬天(季节因素)搜集的样本中最为丰富
之后发现一个感兴趣的C-12状态
探索每个细胞状态和肺结核病情进展的关系,同时校正其他可能存在影响的因素(年龄、性别、季节、遗传以及其他技术因素),发现:
病人在肺结核发展时期,C-12显著降低了20%,并且在年龄因素和男性因素下也会下降,而季节因素(冬季)会上升
又找到了C-12 surface markers:CD4 CD26 CD161 CCR6 ,这些marker已经被证明了以下功能:
- IL-17-producing TH17 state
- CD26 is a co-stimulatory molecule that promotes cytotoxicity
- CD161 is associated with innate-like function
- CCR6 is a homing marker that directs migration to inflamed sites
那么C-12的下降是疾病的原因还是结果呢?
利用外部数据进行验证,发现这种变化可能早于疾病发展
- predicted C-12 abundance was 9% lower in cases before progression compared to latent controls
往期回顾
单细胞混样测序至少可以区分性别
NC单细胞文章复现(五):tSNE
OSCA单细胞数据分析笔记9—Clustering
scPhere——用地球仪来展示降维结果
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