| Background:Acute B-cell acute lymphoblastic leukemia(B-ALL)is the most common treatable malignant tumor in childhood,but 15%of patients will still have refractory/relapse.Refractory and relapse become B-The leading cause of death in children with ALL.As a heterogeneous malignant tumor,peer-to-peer has made some important progress in existing research,but it is relatively fragmented and lacks integrated single-cell sequencing combined with multi-omics analysis of the molecular mechanism and microenvironment changes of B-ALL.Objective:Construct a mononuclear cell transcription profile of healthy children and B-ALL patients,distinguish between normal B cells and malignant tumor cells,compare the changes of cell subgroups at different stages of treatment,look for refractory/relapse-related subgroups in B-ALL patients,and screen for relevant prognosis The key genes of B-ALL interpret the molecular mechanism and immune microenvironment subgroup changes in the progression of B-ALL in multiple dimensions.Method:We collected bone marrow samples of healthy children and supporting B-ALL patients at different stages of initial diagnosis,after treatment,and recurrence,and recorded clinical data to construct a basic information table for B-ALL children.The bone marrow samples of 1 healthy child and 4 matched B-ALL patients were screened for 10X genomics single-cell RNA sequencing.The single-cell RNA data of 3 healthy children from the GEO database were combined to obtain 8 children(4 healthy controls,4 cases of B-ALL patients)a total of 93620 single-cell transcriptome data.Based on the cluster annotation of healthy controls,we performed cluster annotation and differential gene screening on leukemia sequencing samples.The supporting samples were sequenced to analyze the copy number variation of each chromosome,combined with inferCNV and machine learning modeling to distinguish malignant leukemia cells,and extracted at the same time Complete transcriptome sequencing of the supporting RNA was performed to detect the fusion gene,and the complete genome methylation sequencing of the supporting sample of No.4 B-ALL patient was performed,and multi-omics explained the relevant mechanism of B-ALL refractory relapse.Results:Single-cell RNA sequencing was used to obtain a mononuclear cell transcription map of a healthy child and molecular markers of each subgroup.Based on the obtained 4 normal normal control transcription matrices,four B-ALL transcription maps were constructed respectively.InferCNV analysis found that B-The ALL cell subclusters are in the Pro-B stage.InferCNV combined with machine learning to screen malignant tumor cell subgroups,the analysis results are consistent with the trends of flow cytometry results.There are a variety of cell subclones between B-ALL cells in the initial diagnosis and recurrence stage.The number of immune cells such as T cells in the initial diagnosis stage is low,and the number of immune cells increases significantly after treatment.Gene set enrichment analysis of B-ALL tumor cells and normal B lymphocytes indicated that 5 gene sets are positively correlated with malignant B lymphocytes.Malignant B cells/T lymphocytes of B-ALL patients and healthy control B lymphocytes/T lymphocytes Differential gene analysis showed that RACK1 was highly expressed in B-ALL patients.Whole transcriptome sequencing can find rare fusion genes that cannot be found in routine clinical fusion gene testing packages.The pathways enriched by KEGG in the differentially methylated region of the ALL-4 patient’s whole genome methylation sequencing mainly focus on the cellular metabolic pathways.Conclusion:1)InferCNV combined with machine learning can distinguish leukemia cells in non-specifically sorted B-ALL single-cell data.2)B-ALL cell subclusters are in the Pro B stage.3)B-ALL tumor cells have significant heterogeneity,and non-dominant subgroups at the initial diagnosis period may be related to recurrence.4)The immune microenvironment in the bone marrow also changes during the treatment. |
PDF Full Text Request