Developmental And Functional Gene Interactions Analysis Of Osteoblast Based On Single-Cell RNA Sequencing

Background: Osteoblasts play an important role in bone mineralization and bone construction.Osteoblasts are involved in biological processes such as calcium deposition,bone matrix formation,and secretion of bone metabolism related proteins.Osteoblasts can differentiate into osteocytes and regulate the dynamic balance of bone formation and bone absorption with osteoclasts.The heterogeneity of cell populations is a common characteristic of various cell types.The specific function of each cell also depends on its unique identity acquired during the molecular coding process of DNA,RNA,and protein transformation.Classical cell types can also be further divided into different cell subsets based on systematic differences in gene expression profiles.This will provide an important foundation for studies of the specific functions of different cell subtypes.In recent years,the progressive single cell RNA sequencing(sc RNA-seq)technology has become a widely used experimental technology in the studies regarding cell heterogeneity by unbiased detection of cell transcription profiles at a single cell resolution level.In addition,various pathophysiological processes in the human body are controlled by the coordinated expression of multiple genes,rather than relying on the expression status of one or several molecules.This issue has not been explored in osteoblast differentiation and function studies at the single cell level.Objective: To solve the above problems,this study was based on sc RNA-seq,combined with multiple gene association analysis algorithms,to explore gene interaction networks such as transcription factor regulation and cell communications that related to osteoblast differentiation or functions.Furthermore,combined with in vivo/ in vitro independent datasets,in vivo/ in vitro molecular biological experiments and bulk RNA sequencing technology,the potential mechanisms of target transcription factors/network members affecting osteoblast differentiation and function were verified.Methods: 1.The research subjects of this study included fresh femoral head samples from a 31-year-old male patient with osteoarthritis and osteoporosis.The patient’s main clinical manifestations included hip joint pain,limited hip joint movement and limited function(For in-house sc RNA-seq experiment);Peripheral blood samples from 944 healthy male subjects(aged 20-64 years)(For in-house RNA-seq experiment),20 postmenopausal and 20 premenopausal low BMD subjects,as well as 20 postmenopausal and 20 premenopausal normal BMD subjects(public dataset GSE56815).2.The femoral head was collected from the 31-yearold male patient during hip replacement surgery.Freshly harvested bone tissue fragments were incubated with type II collagenase.After cell digestion,a part of the cell mixture were collected as microenvironment cells.After negative selection of 7-AAD,ALPL + /CD45/34/31-cells were collected as osteoblasts.3.The two cell suspensions were detected by sc RNA-seq according to the standard process of 10 X Genomics.Then data alignment,filtering,barcode and UMI counting were performed on the sequencing data to obtain the bone microenvironment data set and osteoblast sorting data set.4.Single-cell regulatory network inference and clustering(SCENIC)and cell specific network(CSN)construction algorithms were used to analyze potential functional transcription factors and target genes in various osteoblast subpopulations.5.The transcription factors CREM,FOSL2,FOXC2,RUNX2,and CREB3L1 were experimentally validated from different perspectives using the in vitro bone marrow-derived mesenchymal stem cells(BM-MSC)osteogenic differentiation public dataset,in vivo mouse osteoblast sc RNA-seq public dataset,in vitro MC3T3-E1 cell line osteogenic differentiation induction dataset,and in vivo mouse bone tissue slice experiments.6.The interaction between osteoblastic lineage cells(OBCs)and immune cells in the bone microenvironment was systematically analyzed by using Cellphone DB and Cell Chat,two intercellular communication analysis tools.7.By using algorithms such as Weighted Gene Coexpression Network Analysis(WGCNA),CSN,and the Least Absolute Shrinkage and Selection Operator(LASSO),we analyzed gene associations and their correlation with BMD in the peripheral circulation monocyte expression profile of 944 male and 80 female samples.Results: 1.Transcription factor regulation analysis results:(1)After we got the first sc RNA-seq dataset of human osteoblasts,the heterogeneity of osteoblasts was systematically analyzed based on gene expression and transcription factor regulation analysis,and osteoblasts in vivo were divided into four different subtypes: preosteoblast-S1,preosteoblast-S2,intermediate osteoblast,and mature osteoblast.(2)CREM and FOSL2 regulons have the highest active score in preosteoblast-S1,FOXC2 regulons have the highest active score in intermediate osteoblast,and RUNX2 and CREB3L1 regulons have the highest active score in mature osteoblast.(3)CREM,FOSL2,FOXC2,RUNX2,and CREB3L1 regulons have potential associations with functional state transitions related to immunity,cell proliferation,and differentiation.2.Independent data set and experimental validation results:(1)The expression changes of CREM,FOSL2,FOXC2,RUNX2,and CREB3L1 transcription factors and their target genes during osteogenesis were consistent with the previous analysis results in the validation dataset.(2)Foxc2 gene expressed in osteoblasts in vivo.Functional experiments showed that Foxc2 mainly promoted the early stages of the differentiation process of osteoblasts.3.Bone-immunological interactions in bone microenvironment:(1)The cellular communication network between osteoblastic lineage cells(OBCs)and bone microenvironment immune cells was constructed.(2)Identified ligand and receptor genes that were related to bone function,such as JAG1,NOTCH1/2,NAMPT-INSR,and MDK-SDCs.4.Peripheral circulation related bone-immunological interactions:(1)Identified a gene coexpression module(midnightblue module)which was significantly associated with a monocyte subtype(Mono4).(2)Gene interactions in the core subnetwork of the midnight blue module have a potential protective effect on bone mineral density(BMD)in elderly men.(3)In postmenopausal women,a osteoporosis risk prediction model was established using the midnight blue module core subnetwork gene which could effectively predict BMD levels.Conclusions: 1.This is the first study to reveal the regulatory profile of transcription factors at the single cell level of human osteoblasts in vivo,and to explore the potential role of CREM,FOSL2,FOXC2,RUNX2,and CREB3L1 regulons in the process of osteoblastic differentiation.2.Functional experiments showed that FOXC2 mainly affected the osteogenesis process at the early stage.3.Ligands and receptor genes related to bone function such as JAG1,NOTCH1/2,NAMPT-INSR,and MDK-SDCs were identified in the cellular communication network between OBCs and bone microenvironment immune cells.4.Gene interactions in the core subnetwork of the midnight blue module were associated with the protective function of BMD levels in elderly men.The osteoporosis risk prediction model established in postmenopausal women can effectively predict the BMD level of this population.Our findings provide a framework for understanding gene relationships during osteogenesis at the single-cell level,thereby laying the foundation for exploring characteristic gene functions from a novel gene relationship perspective.