| Background and purpose:Thyroid cancer is the most common endocrine malignancy that originates from the follicular epithelium or parafollicular cells of the thyroid gland.In recent years,the incidence of thyroid cancer has increased rapidly worldwide,and the overall incidence of thyroid cancer in China ranks 8th among all malignant tumors.Thyroid cancer can be divided into various subtypes according to the degree of differentiation,among which the least differentiated is anaplastic thyroid cancer（ATC）and the most differentiated is papillary thyroid cancer（PTC）.There are significant differences in their histological patterns,clinical manifestations,treatment methods and prognosis.PTC is the most common,accounting for about 85%-90%of all thyroid cancers,but it is less malignant and has a better prognosis,with a 10-year disease-specific survival rate of more than 90%,whereas ATC accounts for only 2-3%of all thyroid cancers,but it is extremely malignant,with a one-year survival rate of less than 20%,and it causes about 50%of all thyroid cancer-related deaths.To date,there has been much controversy regarding the origin of ATC development.At the pathological level,the specific biomarker TTF-1（galectin-3）of PTC cells can be detected in ATC tissue using immunohistochemistry,suggesting that PTC cells may develop into ATC through dedifferentiation.At the DNA level,although both ATC and PTC patients have a high frequency of BRAFV600E mutations,somatic mutations located in the TP53 and TERT promoters are specifically present in the tumors of ATC patients,suggesting that PTC may progress to ATC through accumulation of mutations.Some studies have speculated,based on clinical phenomena and tumor heterogeneity analysis,that thyroid follicular cells may first develop into PTC and then further dedifferentiate into ATC when cell plasticity changes.Another view,based on genomic evolutionary analysis of somatic mutations,suggests that PTC and ATC separated early in the evolutionary process of thyroid follicular cells and developed through different molecular pathways.However,none of the above evidence can explain and demonstrate the origin and progression pathway of thyroid cancer subtype ATC at the cellular level,as well as the molecular mechanisms and key molecular events driving ATC progression,let alone understanding the different states of the immune microenvironment in PTC and ATC and their interactions with tumor cells.With the continuous development and maturation of single-cell RNA sequencing（scRNA-seq）platforms,we can accurately identify and characterize the status of individual cells,making it possible to explore the tumor heterogeneity and tumor microenvironment of thyroid cancer subtypes at the single-cell level,which is crucial for exploring tumor-related mechanisms including tumor cell plasticity.In addition,bioinformatics analysis allows the construction of pathway models of thyroid tumor cell progression based on scRNA-seq data,revealing the gene regulatory network of ATC development and finding the drivers that determine cell fate.These results,combined with traditional high-throughput sequencing at the DNA level and transcriptome detection at the RNA level,will enable a systematic and comprehensive grasp of ATC development patterns under different molecular events at the multi-omics level,providing a theoretical basis for individualized prevention and treatment of ATC.In this project,we have mapped the single cell transcriptome of thyroid cancer subtypes PTC and ATC for the first time to characterize the cell types and status of thyroid cancer tissues.and explored the effects of tumor heterogeneity,tumor microenvironment and copy number variation on the different clinical phenotypes of PTC and ATC.We also combined genomic and transcriptomic information at the overall tumor level to depict the molecular evolutionary trajectory from normal thyroid follicular cells to PTC and ATC,to identify the drivers and mechanisms of action that determine the progressive fate of thyroid follicular cells,and to experimentally validate them at the cell line level,aiming to provide a theoretical basis for candidate markers and drug targets for the clinical diagnosis and treatment of thyroid cancer patients at the multi-omics level.Methods:The aim of this project is to conduct basic research from key clinical questions and return the findings to the idea of clinical application.We collected seven clinical tissue samples from six thyroid cancer patients,including three ATC tissues,three PTC tissues and one PTC paraneoplastic tissue（NOM）,successively from the Department of Thyroid Surgery,West China Hospital between March 2019 and July 2019,and performed histopathological characterization of the samples by immunohistochemistry,while collecting and collating detailed clinical data of the patients.Fresh thyroid tissues were lysed and purified using collagenase and other enzymes to prepare cell suspensions.The scRNA-seq technology of the 10×Genomics platform combined with bioinformatics analysis was used to map single cells of PTC and ATC tissues for the first time,and the accuracy of the fractionation was verified using immunohistochemistry.We analyzed the distribution and interaction of tumor microenvironment components in NOM,PTC and ATC tissues to reveal the factors associated with the degree of malignancy and differentiation of thyroid cancer at the level of tumor microenvironment and tumor heterogeneity.To further answer the controversial clinical questions regarding the origin and progression pathways of ATC cells,we incorporated follicular epithelial cells and tumor cells and constructed the evolutionary trajectory of tumor cells using proposed time series analysis software.Genes with progressively higher or lower expression during tumor cell evolution were screened as candidate markers,and functional and pathway enrichment analyses were performed.Based on the set of genes associated with thyroid differentiation,the thyroid differentiation score of each cell on the enriched proposed timeline was analyzed using ssGSEA.The copy number amplification and deletion regions of PTC and ATC tumor cells compared to normal follicular epithelial cells were explored by copy number variation analysis and validated with whole exome sequencing data for multi-omics.In addition,we constructed a gene co-expression network using SCENIC analysis to search for transcription factors associated with driving ATC progression.Combining the results of these multiple analyses,we identified candidate driver genes and enrichment pathway markers associated with ATC tumor cell progression.To further validate the candidate gene markers,we determined the expression levels of the genes in NOM,PTC and ATC tissues using immunohistochemical staining.Combined with ChIP-seq data analysis,we calculated the set of target genes corresponding to the candidate driver genes and the proportion of different regions of the gene in which their binding sites were located,screened for a subset of target genes that were significantly highly expressed in ATC,and verified whether the candidate pathway markers were also significantly enriched.Combined with multiple published transcriptomic data,the differential enrichment of candidate genes and pathway markers in ATC samples was verified at a multi-omics level.In the TCGA database,PTC samples were divided into high and low groups according to the expression level of candidate driver genes to verify the differential enrichment of candidate pathway markers and to analyze the impact of the expression level of this gene on the overall survival and disease specific survival of PTC patients.In addition,we further validated the analysis results at the cell line level.We set PTC cell line TPC-1 as the control group and TPC-1 overexpressed with the candidate driver gene as the experimental group to detect the effects of the candidate driver gene on the proliferation and migration levels of thyroid tumor cells.Combined with protein immunoblotting assay,the effect of the gene on epithelial-mesenchymal transition（EMT）of PTC tumor cells was detected.Results:We characterized the cell types and features of ATC/PTC tumor tissues at the single-cell level and revealed the presence of tumor heterogeneity;immune microenvironmental analysis indicated that tumor tissues exhibited an immunosuppressive state;we also identified independent evolutionary models for ATC and PTC and that ATC may have evolved from a partial PTC subpopulation;furthermore,the accumulation of copy number variation may be a necessary condition for the evolution of ATC In addition,the accumulation of copy number variants may be a necessary condition for the evolution of ATC;through integration analysis,we determined that the transcription factor CREB3L1 regulates the progression of PTC subpopulations to ATC through the activation of EMT and MTORC1;the results of cell line level experiments suggest that CREB3L1 affects the dedifferentiation of thyroid cancer through the regulation of migration and EMT levels.1.Tumor heterogeneity exists in thyroid tumor tissueUsing scRNA-seq and bioinformatics analysis,we have for the first time mapped single-cell profiles of thyroid cancer subtypes PTC and ATC tumor tissues as well as normal follicular epithelial cells.After data quality control,we obtained a total of46,205 cells from 7 thyroid tissue samples and classified the cells into 16 cell subpopulations based on the set of differentially expressed marker genes with corresponding cell type annotation,including normal follicular epithelial cells,parafollicular cells,endothelial cells,ATC tumor cells,2 PTC tumor cell subpopulations,2 fibroblast cell subpopulations and 8 immune-related cell populations（B cells,T cells and macrophages,etc.）.Among them,PTC tumor cells were divided into two subpopulations based on differentially expressed genes,indicating the heterogeneity of PTC tumor cells.We distinguished them from PTC tumor cells based on marker genes specifically expressed in different tumor cells,such as PBK expression in ATC tumor cells,and verified by immunohistochemistry that the gene was only expressed in ATC tissue samples,demonstrating the accuracy of our cell subpopulation and cell type annotation.Cell ratio analysis revealed that PTC and ATC tissues had fewer normal follicular cells compared to NOM,but contained a high proportion of PTC and ATC tumor cells,respectively,demonstrating the accuracy of our sampling as well as analysis.We also found that the proportions of fibroblasts and immune cells were also increased in PTC and ATC tissues and validated these results using published transcriptomic data combined with reverse convolution analysis.2.Thyroid tumor tissues present immunosuppressed stateBased on the tumor-associated macrophage gene set,combined with ssGSEA enrichment score analysis,we found that macrophage subpopulation 1 tended to be more similar to pro-inflammatory M1-type macrophages,while macrophage subpopulation 2 tended to be more similar to M2-type macrophages exhibiting pro-oncogenic features.Proportional analysis revealed that the pro-inflammatory macrophage subpopulation 2 was significantly increased in ATC tissues,which may explain the higher malignancy and worse clinical prognosis of ATC.We also found that ATC tissues contained a higher proportion of FAP-positive fibroblasts,which can promote an immunosuppressive state in the tumor microenvironment.In addition,a higher proportion of NK and T cells in thyroid cancer tissues expressed marker genes for T-cell depletion,regulatory T cells,and tumor-associated myeloid cells,which showed a gradual increase in expression from normal thyroid follicular cells to PTC cells to ATC cells,indicating that the tumor microenvironment in ATC is significantly immunosuppressed.We calculated the enrichment scores of immune cell deficiency and immune regulation based on the relevant gene sets and using ssGSEA analysis,which again confirmed the immunosuppressive status of ATC.Based on the analysis of the interaction between tumor cells and immune cells,we found that ATC cells had higher frequency interactions with immune cells compared to PTC and follicular epithelial cells.Further analysis showed that ATC cells could directly inhibit the killing effect of NK and T cells through ligand receptor interactions of PD-1:PD-L1,PD-1:PD-L2,and PD-1:FAM3C,and this result explained the mechanism of action of the tumor microenvironment of ATC tissues presenting an immunosuppressive state.In addition,the ligand HLA-E of PTC cells strongly interacts with the receptor KLRC1 of NK cells,which encodes the best-known NK inhibitory receptor NKG2A,suggesting that PTC cells can directly inhibit NK cells through the HLA-E:KLRC1 interaction.3.Determine the independent evolutionary model of ATC/PTC and the subpopulation of ATC derived from PTCBy subpopulation analysis,we found that PTC tumor cells were further divided into 10 cell subpopulations.Among them,subpopulation 10,which was in a more distant position,was basically derived from ATC tissue,indicating that there might be significant differences between PTC cells of ATC tissue origin（aPTC）and PTC cells of PTC tissue origin（pPTC）.Using the proposed time series analysis,it was found that follicular epithelial cells progressed through different pathways to become PTC and ATC tumor cells,respectively.Notably,aPTC cells were located on the evolutionary trajectory of ATC,suggesting that ATC may have progressed from a small number of PTC cells.Based on the relevant gene set and combined with ssGSEA analysis,we found that the thyroid differentiation scores gradually decreased and the ERKS signaling pathway gradually activated during the proposed timeline of follicular epithelial cell evolution into ATC tumor cells.On the trajectory of PTC to ATC transition,enrichment analysis of genes with progressively higher expression revealed specific activation of EMT and MTORC1 signaling pathways.On the opposite trajectory,the expression levels of some tumor-associated genes were progressively enhanced and the p53 pathway was specifically activated.The above results suggest that driving the progression of PTC and ATC depends on the specific activation of different signaling pathways and cellular states,and that ATC may be further progressed from aPTC cells.4.The accumulation of copy number variation is necessary for the evolution of PTC subpopulations into ATCsCopy number variation analysis of tumor cells showed that ATC cells had a higher copy number variation load compared to aPTC and pPTC.p TC tissue-derived PTC1and PTC2 had similar copy number variation levels,suggesting that their origins should be the same.In addition,there is a large amount of specific copy number variation in ATC cells compared to aPTC,except in partially identical regions,suggesting that the accumulation of amplifications or deletions of these chromosomes drives the progression from aPTC to ATC.We validated the results of the single-cell-level analysis using whole-exome sequencing data,particularly in ATC tumor samples,where large segments of copy number amplification were detected within the 41.1 to47.9 megabase region of chromosome 11.Genes within this region were specifically highly expressed in ATC cells and these genes were progressively enriched during the progression from follicular epithelial cells to ATC cells in the proposed timeline.The above results suggest that the accumulation of copy number variants may be necessary for the evolution of ATC tumors.5.Identification of CREB3L1 regulates aPTC progression to ATC through activation of EMT and MTORC1Based on SCENIC analysis,we established a gene co-expression network in thyroid cells to search for driver transcription factors regulating ATC tumor cell evolution and found that only CREB3L1 was significantly enriched in cells of ATC tissue origin and its expression gradually increased from aPTC to ATC.Comprehensive analysis revealed that the vast majority of CREB3L1-positive cells were derived from ATC tumor cells,and the gene was located just within the large amplified region of chromosome 11 associated with ATC as described above.Immunohistochemical staining revealed that CREB3L1 was highly expressed in the ATC component and to a lesser extent in the PTC component.The above results suggest that the transcription factor CREB3L1 may play a crucial role in the progression of ATC.By analyzing chromatin immunoprecipitation sequencing data in public databases,we found that the binding sites of CREB3L1 to target genes were mainly concentrated in the promoter region.A total of 2,222 potential target genes for CREB3L1 were obtained after filtering based on gene expression and ChIP-seq signals,most of which as well as the EMT and MTORC1 signaling pathways were also enriched in ATC cells,indicating that the transcription factor CREB3L1 is a driver gene for these signaling pathways.Combined with multiple analyses of thyroid cancer transcriptome data showed that the transcription factor CREB3L1 as well as genes related to the signaling pathways EMT and MTORC1 were significantly enriched in ATC samples compared to normal controls.The PTC samples in the TCGA database were divided into high and low groups based on CREB3L1 expression levels,and a total of 1,115 differentially expressed genes were screened,while both EMT and MTORC1 signaling pathways were also significantly enriched in the CREB3L1 high expression group.In addition,PTC patients in the CREB3L1 high expression group had significantly worse overall survival rates and disease specific survival rates.Given the similar clinical features of thyroid cancer metastases from ATC and other tumors,we found that CREB3L1 was expressed only in ATC tissues after immunohistochemical staining of thyroid metastases from multiple tumors,suggesting that the expression level of this gene could be used as a genetic marker for clinical diagnosis of thyroid metastases from ATC and other tumors.All these results suggest that the expression level of CREB3L1 may determine the clinical outcome and cell fate of thyroid cancer and promote the progression of ATC.6.Experimental evidence that CREB3L1 affects the dedifferentiation of thyroid cancer by regulating migration and EMT levelsThe results of experimental studies at the cell line level showed that the migration level was significantly increased after overexpression of CREB3L1 in the PTC cell line TPC-1,with no significant change in the proliferation level.The results of protein immunoblotting experiments showed that the expression of EMT-related proteins VIMENTIN,SNAI1 and TWIST1 was enhanced after TPC-1 overexpression of CREB3L1,and the expression of epithelial cell-associated protein CDH1 was decreased,indicating that the level of EMT was significantly increased.The above results indicated that the expression level of CREB3L1 could alter the differentiation process of thyroid cancer by regulating the migration and EMT levels of PTC cells.Conclusion:In this project,we mapped the single-cell transcriptional profiles of thyroid cancer subtypes PTC and ATC for the first time.It was found that there was tumor heterogeneity in PTC cells.Analysis of the tumor microenvironment showed that ATC had more FAP-positive fibroblasts and M2-type macrophages than PTC,leading to a stronger immunosuppressive effect of ATC.Loss-of-function lymphocytes were only present in PTC and ATC tissues,whereas regulatory T cells and suppressor myeloid cells were only present in ATC,suggesting that a stronger immunosuppressive state may be associated with ATC having a more malignant clinical profile.In addition,direct interactions through ligand receptors lead to suppression of immune cells by tumor cells.Follicular cells can progress through different evolutionary trajectories to ATC and PTC,respectively,and ATC may progress from aPTC.progression of ATC is dependent on specific activation of the EMT and MTORC1 signaling pathways,whereas enrichment of the p53 pathway can drive the evolution of PTC.In addition,ATC cells have a higher CNA load and the accumulation of CNA promotes the evolution of aPTC to ATC.We determined that the transcription factor CREB3L1 can drive the progression of aPTC to ATC by activating the EMT and MTORC1 signaling pathways,and experimentally demonstrated that the expression level of CREB3L1 can influence the dedifferentiation of thyroid cancer by regulating the EMT level of PTC cells and the associated migration ability of cells.Furthermore,not only the expression level of CREB3L1 correlates with disease recurrence,overall survival and disease-specific survival of PTC patients,but this gene can also be used as a candidate marker for clinical differentiation of thyroid metastases from ATC and other tumors. |
PDF Full Text Request