Prediction Of Lung Adenocarcinoma Prognosis Through Topologically Associating Domain Boundary Profiling

Lung cancer is a kind of cancer with high invasiveness,high metastasis rate and high morbidity,and its mortality rate is the highest in the world.Lung Adenocarcinoma(LUAD)is the most frequently diagnosed subtype of lung cancer.There are no obvious symptoms in the early stage of LUAD,and tumor invasion and metastasis have occurred by the time of diagnosis.Further research into the biology of LUAD could lead to more effective and targeted drugs.3D chromosome structure plays an indispensable role in gene expression and regulation,cell diversity and identity,tissue development.Using various 3C based techniques,it has been found that chromosomes are organized in cells into a dynamic but non-random multi-level structure.TAD is a structural unit on the megabase scale,characterized by a higher level of interaction within the region than outside the region.In this study,we established a prognostic model for LUAD by integrating HI-C data from lung cell lines and high-throughput data sets from LUAD patients.This model screened 34 prognostic boundary features(PRBFs)and used their expression levels and risk coefficients to construct the risk scoring system for LUAD patients.The concept of the interaction frequency weighted DLR(WDLR)was proposed to characterize the chromatin structure contraction level in a given region to evaluate the chromatin 3D structure conformation in the region near prognostic boundary genes.Subsequently,prognostic boundary genes were analyzed at the level of protein interaction networks and transcription factors to explore the reliability of our prediction and explore the potential role of genes in LUAD.We found that:1.Based on the PRBFs risk score scheme,LUAD patients can be divided into highrisk group and low-risk group,and the prognosis of the two groups is significantly different.This means that PRBFs can be used as a prognostic marker for LUAD.2.In the process of cell cancerization,the three-dimensional structure conformation near the prognostic boundary genes showed two trends of tightness and looseness,indicating that the local structure of prognostic boundary genes was heterogeneous and dynamic.And changes in local three-dimensional spatial structure in LUAD genome can be reflected by interactions between boundary genes and surrounding regions.3.Prognostic boundary genes with different 3D structural conformation changes correspond to different methylation levels and gene expression levels respectively.Genes that become more tightly structured during carcinogenesis correspond to higher promoter methylation levels and lower expression levels.4.Prognostic boundary genes have a very high degree of connectivity in PPI network and are in the core position.Changes of boundary genes have a more profound impact on the network.5.Transcription factors EHF,SP1 and PAX5 can regulate more than 30 prognostic boundary genes,all of which have been proved to play important roles in cancer.11 prognostic boundary genes have been reported in cancer studies.This study revealed the role of TAD boundary in the diagnosis and treatment of LUAD,and the selected boundary features could be used as candidate diagnostic markers of LUAD,thus proving that TAD boundary can be used as another method to explore the prognosis of LUAD.