| Primary liver cancer(PLC)is one of the most common malignant tumors worldwide.In China,hepatocellular carcinoma(HCC)accounts for 93.0%of all PLC cases,with approximately 90%of HCC cases caused by hepatitis B virus(HBV)infection.With advancements in surgical techniques and medication treatments,the 5–year survival rate after curative resection of HCC has exceeded 30%.However,tumor metastasis and postoperative recurrence remain major threats to the long–term prognosis of HCC patients.Currently identified biomarkers of HCC,such as alpha–fetoprotein(AFP),are proteins and their modifications that are expressed during the embryonic stage,closed after birth,and re–expressed in cancer,suggesting that the development direction of HCC is exactly opposite to that of embryonic development,belonging to retro–development.Chronic unresolving inflammation is maintained by immune imbalances caused by the interaction of genetic susceptibility and environmental exposures(e.g.,HBV infection).Chronic unresolving inflammation promotes somatic mutation by imbalance mutage–causing factors and mutational repair factors.Most cells with somatic mutations are eliminated in the competition for survival in the inflammatory microenvironment,and only a small fraction of mutant cells that alter survival signaling pathways and exhibit"stemness"characteristics are able to survive.HCC cells with key somatic mutations interact with the tumor microenvironment(TME),and then reversely develop into cancer stem cells(CSCs).CSCs are defined as a poorly differentiated subpopulation of cancer cells that can grow,regenerate,and invade tissues,often exhibiting characteristics similar to normal stem cells.CSCs drive intra–tumoral heterogeneity and the malignant progression of HCC,and are the root of tumor recurrence,metastasis,and drug resistance.However,the key mechanisms of HCC reverse development have not been elucidated.Which key somatic mutations accumulate continuously during the process of HCC retro–development,what characteristics do tumor cells accumulate in the late stage of HCC development possess,how they interact with the TME to further promote the retro–development of tumor cells,these questions remain unresolved.Clarifying these issues is of great significance for improving the survival of HCC patients and reducing the mortality rate caused by HCC.Part Ⅰ Construction of Liver Organoid Model and Screening of Genes Related to HCC Retro–developmentObjective:Previous studies based on liver cancer cell lines are limited by the homogeneity of cell lines,unable to simulate the real genetic characteristics of HCC,and unable to conduct continuous research on the evolution and development process of HCC.By constructing liver organoid model,genes related to HCC retro–development are screened.Method:(1)Human embryonic stem cell–induced liver organoids were constructed,and the ability of liver organoids to simulate liver development in vitro was identified through immunofluorescent staining of various markers;(2)HCC organoids were constructed from clinically collected surgically resected HBV–infected HCC samples,and the consistency of HCC organoids with cancer tissues was verified through immunohistochemistry and other methods;(3)based on the construction of HCC organoids,key genes related to HCC retro–development were screened in combination with public databases such as TCGA,GEO,GTEx;(4)the biological functions of key genes are studied using cell models and HCC organoids.Results:(1)Through immunofluorescent staining of different stage samples of human embryonic stem cell–induced liver organoids,it was found that SOX2 is highly expressed in hepatic progenitor cell spheres,AFP,and HNF4 are highly expressed in immature liver organoids,and ALB and CYP3A4 are highly expressed in mature liver organoids,indicating that liver organoids can simulate different stages of normal liver development in vitro;(2)a total of 86 fresh resected HBV–related HCC samples were included,only 18cases successfully constructed HCC organoids,with a success rate of about 20.9%,which was related to tumor diameter(P=0.004),peripheral blood alpha–fetoprotein(AFP)(P=0.044),AFP L3(P=0.047),microvascular invasion(P=0.044),TNM staging(P=0.005),BCLC staging(P=0.047).HCC organoids were consistent with cancer tissue in immunohistochemical staining;(3)D–amino acid oxidase(DAO)was lowly expressed in cancer tissues and HCC organoids with strong stemness,and analysis using TCGA database found that the expression of DAO was negatively correlated with the HCC reverse development gene set,and negatively correlated with stemness markers such as CD44,Ep CAM,and CD133(Prominin–1),and epithelial–mesenchymal transition(EMT)related molecules(e.g.,Vimentin,Twist1,SNAI1).However,immunofluorescent staining of human embryonic stem cell–induced liver organoid samples did not find that DAO expression increased with the development of liver organoids;(4)overexpression of DAO inhibited the proliferation,migration,and invasion of liver cancer cell lines;DAO oxidative metabolism of D–amino acids produced hydrogen peroxide,inducing HCC cell apoptosis;DAO enhanced the sensitivity of sorafenib–resistant HCC organoids to sorafenib;overexpression of DAO did not reduce the expression of stemness markers(CD44,CD133,Ep CAM)and EMT–related molecules in liver cancer cell lines and HCC organoids.Conclusion:(1)Human embryonic stem cell–induced liver organoids can simulate different stages of normal liver development in vitro;(2)the success or failure of HCC organoid construction is related to the degree of stemness of cancer tissues,HCC organoids retain the immunohistochemical characteristics of cancer tissues,but only tumor cells with a lower degree of differentiation can be cultured in vitro for a long time;(3)D–amino acid oxidase is negatively correlated with the HCC retro–development gene set,and negatively correlated with stemness markers and EMT–related molecules;(4)overexpression of DAO inhibits the proliferation,migration,and invasion of liver cancer cells,enhances the sensitivity of sorafenib–resistant HCC organoids to sorafenib;DAO oxidative metabolism of D–amino acids,produces hydrogen peroxide,induces cell apoptosis.Part Ⅱ Screening of Key Somatic Mutation Gene in HCC OrganoidsObjective:In the process of constructing HCC organoids,cell subpopulations with a lower degree of differentiation have a stronger survival ability,thus they can be continuously cultured in vitro.The key somatic mutations accumulated in HCC organoids are the main driving force in the process of HCC retro–development.By performing multi–omics sequencing analysis on HCC organoids,the key somatic mutations accumulated in the process of HCC evolution and development are clarified.Method:For HBV-associated HCC,HBV insertion mutations are characteristic variants in such patients,therefore,HBV capture sequencing,whole–exome sequencing,and transcriptome sequencing were performed on 20 HCC organoids,and the key somatic mutation situations in HCC organoids were analyzed:(1)combining HBV capture sequencing and transcriptome sequencing,HBV integration mutation patterns in HCC organoids were analyzed;(2)combining HBV capture sequencing results of 50 HCC cancer tissues and paired cancer–adjacent tissues,key HBV integration mutations retained in HCC organoids were screened;(3)combining somatic mutation situations of HBV–related HCC in TCGA database,high–frequency somatic mutation genes retained in HCC organoids were screened;(4)according to the DAO expression,20 HCC organoids were divided into two groups,and the somatic mutation between the two groups was analyzed.Results:(1)From cancer–adjacent tissues to cancer tissues and then to HCC organoids,HBV integration mutations at the DNA level gradually accumulate;HBV integration at the DNA level in HCC organoids mainly occurs in introns and intergenic regions,exonic regions only account for 1.85%,and the HBx region is the main insertion fragment of HBV integration;(2)a total of 5 genes(CTNND2,TERT,PREX2,SORCS2,and TRAPPC9)repeatedly occur HBV integration mutations in cancer tissues,cancer–adjacent tissues,and HCC organoids.MLL4 only repeatedly occurs HBV integration mutations in cancer tissues and HCC organoids,while FN1 only repeatedly occurs HBV integration mutations in cancer–adjacent tissues;in HCC organoids,there are 14 genes that can detect HBV integration mutations at both DNA and RNA levels,but the host genomic breakpoints,HBV insertion sequences,and directions of HBV integration mutations at DNA and RNA levels are inconsistent;(3)in TCGA database HBV–related liver cancer tissues,there are 80 genes with a mutation rate of more than 5%,among which 57(71.3%)also occur exonic mutations in more than 5%of HCC organoids;(4)at the DNA level,the number of HBV insertion mutations in HCC organoids with low DAO expression was higher than that in the DAO high–expression group.The somatic mutation genes unique to the low–DAO expression group were mainly enriched in the PI3K/Akt,AMPK and Rap1signaling pathways,while the somatic mutation genes unique to the DAO high–expression group were mainly enriched in the TGF–βsignaling pathway.Conclusion:(1)HBV integration mutations at the DNA level gradually accumulate in the process of HCC development;HBV integration mainly occurs in introns,and the HBx region is the main insertion fragment of HBV;(2)high–frequency HBV insertion mutation genes such as TERT,MLL4,and CTNND2 accumulate in HCC organoids;HBV insertion at the RNA level is mostly caused by direct attacks by HBV rather than transcription from DNA level HBV insertion;(3)71.3%of high–frequency somatic mutation genes in cancer tissues are retained in HCC organoids.(4)HBV insertion mutations accumulated in HCC organoids with low DAO expression,and somatic mutant genes unique to the DAO low expression group could promote the malignant progression of HCC.Part Ⅲ Clarifying the Characteristics of Key Tumor Cell Subpopulations in the Progression of HCC Evolution and Development and Their Interaction with the TMEObjective:To explore the characteristics of key tumor cell subpopulations accumulated in the process of HCC evolution and development and their interaction with the TME,providing a real evolutionary temporal analysis for the study of HCC retro–development.Method:A total of 30 samples from 14 patients were collected clinically for single–cell nuclear transcriptome sequencing,including 4 cases of primary HCC cancer tissues(PT)and 3 cases of paired cancer–adjacent tissues(PN)(since 1 case of paired cancer–adjacent tissue did not meet the quality control,it was excluded),4 cases of late–stage recurrent HCC cancer tissues(LT)and 4 cases of paired cancer–adjacent tissues(LN),3cases of early–stage recurrent HCC cancer tissues(ET)and 3 cases of paired cancer–adjacent tissues(EN),as well as 3 cases of HCC organoids(OG)and their derived 3 cases of HCC cancer tissues(OT)and 3 cases of paired cancer–adjacent tissues(ON).From cancer–adjacent tissues to cancer tissues and then to HCC organoids as one evolutionary model,from primary HCC to late–stage recurrent HCC and then to early–stage recurrent HCC as another evolutionary model,single–cell nuclear transcriptome sequencing was performed on these two evolutionary models;using R packages Cellranger and Seurat for clustering and cell annotation,using infer CNV to identify tumor cells,using databases such as GO,KEGG,Reactome for functional enrichment analysis of differential genes,using Monocle2 for pseudotime analysis,using Cell Phone DB software for cell communication analysis,GSEA for gene set enrichment analysis.Results:(1)A total of 84768 hepatocytes,45300 epithelial cells,6935 endothelial cells,8702 hepatic stellate cells,17202 macrophages,13107 T cells,and 6756 mature B cells were annotated;(2)in HCC organoids,a subpopulation of macrophages highly expressing RPS5 and RPL18A was retained,with RPS5~+macrophages playing a key role in maintaining the proliferation of HCC organoids.IL7R~+RASGRF2~+T cells,IL2RA~+RTKN2~+T cells,BNC2~+RTN1~+B cells,SNAP25–AS1~+NPSR1–AS1~+B cells,and EBF2~+SEMA5B~+HSCs continuously accumulated from primary HCC to late recurrent HCC and then to early recurrent HCC,indicating these cell subpopulations play crucial roles in the evolution and development of HCC;(3)a total of 26,390 malignant cells within hepatocytes and epithelial cells were identified.Further grouping revealed the gradual accumulation of CYP3A7~+c0 subgroup and the low S100A6–expressing c3 subgroup from primary HCC to late recurrent HCC and then to early recurrent HCC.The low CAPN13–expressing c1 and c5 subgroups were enriched in primary HCC,whereas the high Ep CAM–expressing c2,c7,and c10 subgroups were specifically enriched in HCC organoids.The c0,c3,and c9 were defined as key subgroups in late evolutionary HCC tissues(T1),c1 and c5 as key subgroups in early evolutionary HCC tissues(T0),and c2,c7,and c10 as key subgroups in late evolutionary HCC organoids(O1);(4)HSCs specifically acted on T1 through FGFR1_FGFR2 and TIMP1_FGFR2 ligand–receptor binding,on T0through Fc Rn complex_ALB and FN1_a5b1 complex,and on O1 through NRP1_SEMA3A and FN1_a2b1 complex.Macrophages mainly acted on T1 through CD74_APP and CD46_JAG1 ligand–receptor binding,on T0 through Fc Rn complex_ALB and CMKLR1_RARRES2,and on O1 through NRP1_SEMA3A and NRG2_ERBB4.T1mainly interacted with HSCs through FGFR2_XPR1 and COL18A1_a2b1 complex,T0through Fc Rn complex_ALB and EGFR_COPA,and O1 through TIMP1_FGFR2 and COL4A5_a1b1 complex.Both T1 and T0 mainly acted on macrophages through FN1_a Vb1 complex and EGFR_COPA,while O1 mainly interacted with macrophages through DLK1_NOTCH2 and SEMA3E_PLXND1.Conclusion:(1)Three distinct key tumor cell clusters were identified:T1,selected from an immune–competent in vivo environment,representing late evolutionary HCC tumor subgroups;T0,also selected from an immune–competent in vivo environment,representing early evolutionary HCC tumor subgroups;and O1,selected from an immune–deficient in vitro environment,representing late evolutionary HCC tumor subgroups.Among them,O1 exhibited the highest degree of dedifferentiation and stemness;T0 was in an early evolutionary stage with weaker stemness;T1 contained tumor cell subgroups adapted to tumor microenvironment selection,with varying degrees of stemness;(2)Whether T1,T0,or O1,they mainly communicated with macrophages and hepatic stellate cells.Macrophages and HSCs mainly acted on T0 through Fc Rn complex_ALB,thereby inhibiting its dedifferentiation.T1 interacted with macrophages and HSCs through CD74_APP,FN1_a Vb1 complex,FGFR2_XPR1,and FGFR1_FGFR2,thereby continuously accumulating in the evolution of HCC.O1 mainly interacted with macrophages through DLK1_NOTCH2 and NRP1_SEMA3A,thereby promoting its own dedifferentiation and increased stemness.Taken together,based on the liver organoid model,we screened out the genes and key somatic variants associated with HCC retro–development.We found that the tumor cell clusters enriched in HCC organoids had strong“stemness”,mainly interacting with macrophages through DLK1_NOTCH2 and NRP1_SEMA3A,and then promote their own retro–development. |