Dissecting Differentiation And Maturation Processes Of Human Terminal Erythroblast In Single-Cell Resolution

Objective:To explore a continuous gene expression trends during human terminal erythroid differentiation in single cell resolution with single-cell transcriptome sequencing;to subdivide erythroblasts into finer clusters based on gene expression characteristics of each cell.To profile gene expression network and to identify erythroid regulators during human terminal erythroid differentiation phase.Methods:We collected 3 normal adult donators bone marrow and 1 healthy neonate cord blood as control,and then isolated terminal erythroblasts(CD235a+cells)with Ficoll density gradient centrifugation following by CD235a MicroBeads selected.We captured single cell and constructed sequencing library with 10 X genomics system,and generated data on illumina X ten platform.We applied CellRanger to identified cell clusters after sequencing data was normalization,and discovered signature gene set for each cluster.With Monocle,we reconstructed cell differentiation trajectory and analyzed gene expression patterns during erythroblasts differentiate.We then identified cell differentiation stages based on their signature genes using superCT,Finally,we applied CellRounter to identify terminal erythroid differentiation regulators.We selected AKAP8L,TERF21P and RNF10 to further validated their functional in regulate terminal erythroid differentiation at different stages.Results:After rigorous quality control,8,668 cells and 1.3×104 genes were retained for the subsequent scRNA-seq analysis.Based on gene expression characteristic in individual cells,we applied CellRanger to identify cells clusters,with a continuous arc of 6 clusters as well as 1 separate cluster.When mapped these clusters to cell differentiation stages of ProE,BasoE,PolyE and OrthoE identified by superCT,we observed that 68.7%of cells were OrthoE,but only 1.3%of ProE and 5.0%of BasoE cells.Combined with clusters and differentiation stages,we noted that(1)ProE and BasoE cells were too few to separate from each other,so we did not over-interpreted their feature;(2)PolyE cells can be subdivided into 3 clusters,named as transit-PolyE,early-PolyE and late-PolyE;(3)OrthoE could be divided into two stages:the early-OrthoE and late OrthoE stage.Meanwhile,the number of genes expression decreased when the amounts of hemoglobin genes were accumulated as erythroblasts differentiate.In addition,we noted that both HBB and HBG1/G2 are highly expressed in UCB but only high expressed of HBB in BM.We assembled genes into 3 panels according to their expression trend across the differentiation trajectory.Panel 1 encompasses a set of genes whose expression is down regulated.Panel 2 is a set of genes that highly express at early differentiation stage and persistance until PolyE stage.Panel 3 encompasses a set of genes that are first up regulation at early differentiation stage followed by down regulated.We identified 253 regulators,including 108 of positive regulators and 145 of negative regulators,which participated to regulate terminal erythroid differentiate.The shRNA-mediated knockdown of AKAP8L,TERF2IP and RNF10 using CD34+cell induced differentiation system in vitro validated that these regulators repress erythropoiesis in various stages.Conclusions:1.Based on the conventional stages of terminal erythroid differentiation,PolyE can be further divided into three stages and OrthoE can be subdivided into two stages,and the gene express features and correspondent biological functions of new cell stages had been expounded.more stages based on single cell transcriptome sequencing.2.Fetal terminal erythroblast can highly express both ?-and?-globin mRNA in a single cell.3.We had profiled a continuous gene expression profile during human terminal erythroblast differentiation in vivo.4.We had explored a series of erythroid regulators and their regulation network.