Study On The Pathogenic Mechanism Of Cavernous Malformation Based On Single-cell Sequencing Technolog

Background:Cavernous malformations(CMs),also known as cavernomas,are the second most common hemorrhagic vascular abnormality in the central nervous system(after intracranial aneurysms),a type of low-flow vascular malformation with a high prevalence rate of 0.16%to 0.4%.Common symptoms include headache,epilepsy or focal neurological deficits caused by hemorrhage,which can be acute or progressive.CMs in the spinal cord and deep brain have a high disability rate,and brainstem CMs have a high mortality rate,bringing a heavy burden to patients and society.Therefore,it is of great significance to deeply and comprehensively analyze the molecular mechanism of CMs for precise treatment.Objective:The goal of this study was to create a complete cellular map of CMs in the central nervous system using single-cell sequencing,analyze the pathogenic mechanisms of CMs from multiple cell types,and investigate the similarities and differences between CMs with different gene mutation types so as to identify potential therapeutic targets for precise CM treatment.Methods:This study included 12 patients with CMs who underwent surgical resection at Xuanwu Hospital,Capital Medical University,from May 2017 to January 2021.The surgically resected CMs samples were first subjected to whole exome sequencing to determine the gene mutation types of CMs patients,and then single-cell sequencing analysis was performed.The control samples for single-cell sequencing were temporal lobe or frontal lobe tissues obtained from epilepsy patients.Single-cell sequencing was performed using the 10x Genomics platform.After sequencing data was downloaded,Cell Ranger tool was used for preprocessing,and the results obtained after preprocessing were directly read into the Seurat package of R language for data quality control,dimensionality reduction and clustering.The cell differentiation trajectory was constructed using the Monocle 2 package of R language.Cell-cell communication analysis was performed using the CellPhoneDB package of Python and the CellChat package of R language.Single-cell transcription factor regulatory network analysis was performed using the SCENIC tool of R language to predict the potential regulatory factors of cell subclusters.The potential therapeutic targets identified by single-cell sequencing analysis were validated using immunohistochemistry and immunofluorescence assays in CMs samples and control samples.The SPI1 and TWIST1 transcription factors analyzed by SCENIC tool were transfected with plasmids into human umbilical vein endothelial cells and human aortic smooth muscle cells,respectively,and the expression of the marker genes was detected by real-time quantitative PCR.Two-month-old C57BL/6J mice were injected with tissuespecific viruses and then verified the results after separating primary mouse brain cells by magnetic beads.Results:According to the results of whole exome sequencing,the genotypes of 12 CMs patients were divided into two categories:PIK3CA mutations carriers,and non-PIK3CA mutations carriers.PIK3CA mutations carriers could be further divided into 3 subcategories:only PIK3CA mutation group,double somatic mutations(MAP3K3 or CCM1 plus PIK3CA mutations)group,and germline mutation plus PIK3CA mutation group;and non-PIK3CA mutation carriers only carried MAP3K3 mutations.Single-cell atlas of CMs and control samples were constructed by annotating the marker genes of subclusters.Cell communication analysis revealed enhanced communication interactions between endothelial cells,mural cells,and fibroblasts within and between the three cell types in CMs lesions.The main signaling pathways that play a key role in endothelial homeostasis and CMs progression,including the PI3K-AKT signaling pathway,vascular regulation,and response to hypoxia,were shared in CMs of all mutant groups.Reclustering of endothelial cells revealed four subgroups,EC1 subcluster highly expressed extracellular matrix proteins;EC2 subcluster highly expressed MFSD2A,which had normal endothelial cell function in maintaining the integrity of blood brain barrier and had the highest proportion in the control group;EC3 subcluster characterized by immune-related genes,suggesting that this subcluster was mainly related to immune response;EC4 subcluster highly expressed PLVAP.The PLVAP positive EC4 subcluster was unique in CMs lesions,and immunofluorescence staining further confirmed the presence of PLVAP positive EC4 subcluster in CMs lesion samples.Endothelial-to-mesenchymal transition(EndMT)cell type was identified at the single-cell level and confirmed in human CMs lesion samples by immunofluorescence staining.Pseudotemporal trajectory analysis showed that endothelial cells could differentiate into EndMT cells and then transformed into fibroblasts.EndMT was an intermediate transition state,mainly in an immune-activated state.SCENIC analysis predicted that SPI1 was a potential transcription factor that specifically regulated EndMT cells,and experiments in cell and animal models have proved that SPI1 was essential for the process of EndMT.Reclustering of smooth muscle cells revealed four subclusters:pericytes,fibroblast-like smooth muscle cells subcluster SMC1,typical smooth muscle cells subcluster SMC2 and immune-related smooth muscle cells subcluster SMC3.There was a phenotypic switching from typical smooth muscle cells(SMC2 subcluster)to fibroblast-like smooth muscle cells(SMC1 subcluster).SCENIC analysis identified that TWIST1 was a potential regulator of smooth muscle cells subcluster switching,and experiments in cell and animal models have proved that TWIST1 was essential for the phenotypic switching.Analysis of immune cells showed immune activation of microglia and astrocytes in CMs lesion groups.Conclusion:This study constructed a comprehensive single-cell transcriptome atlas from 12 samples of CM and 3 samples of control.A new EC subpopulation marked with PLVAP was identified in lesions uniquely.Its marker genes were also upregulated in the ECs of Pdcd10-KO mice.EndMT cells were identified for the first time in CMs at single-cell levels,which was accompanied by strong immune activation.Transcription factor SPI1 was predicted and verified to be a novel key driver of EndMT.A specific fibroblast-like phenotype was more prevalent in lesion smooth muscle cells,hinting the new mechanism for vessel reconstructions and repairs in CMs,and we found that TWIST1 could induce SMC phenotypic switching.Overall,our findings provided a comprehensive transcriptomic landscape of human CM at single-cell resolution.Through the data mining and experimental verification of single-cell sequencing,the potential therapeutic targets of CMs were proposed,which laid the foundation for subsequent research and clinical treatment of CMs.