Revolutionary multi-omics analysis revealing prognostic signature of thyroid cancer and subsequent in vitro validation of SNAI1 in mediating thyroid cancer progression through EMT

Abstract

Thyroid carcinoma (TC), the most commonly diagnosed malignancy of the endocrine system, has witnessed a significant rise in incidence over the past few decades. The integration of scRNA-seq with other sequencing approaches offers researchers a distinct perspective to explore mechanisms underlying TC progression. Therefore, it is crucial to develop a prognostic model for TC patients by utilizing a multi-omics approach. We acquired and processed transcriptomic data from the TCGA-THCA dataset, including mRNA expression profiles, lncRNA expression profiles, miRNA expression profiles, methylation chip data, gene mutation data, and clinical data. We constructed a tumor-related risk model using machine learning methods and developed a consensus machine learning-driven signature (CMLS) for accurate and stable prediction of TC patient outcomes. 2 strains of undifferentiated TC cell lines and 1 strain of PTC cell line were utilized for in vitro validation. mRNA, protein levels of hub genes, epithelial-mesenchymal transition (EMT)-associated phenotypes were detected by a series of in vitro experiments. We identified 3 molecular subtypes of TC based on integrated multi-omics clustering algorithms, which were associated with overall survival and displayed distinct molecular features. We developed a CMLS based on 28 hub genes to predict patient outcomes, and demonstrated that CMLS outperformed other prognostic models. TC patients of relatively lower CMLS score had significantly higher levels of T cells, B cells, and macrophages, indicating an immune-activated state. Fibroblasts were predominantly enriched in the high CMLS group, along with markers associated with immune suppression and evasion. We identified several drugs that could be suitable for patients with high CMLS, including Staurosporine_1034, Rapamycin_1084, gemcitabine, and topotecan. SNAI1 was elevated in both undifferentiated TC cell lines, comparing to PTC cells. Knockdown of SNAI1 reduced the cell proliferation and EMT phenotypes of undifferentiated TC cells. Our findings highlight the importance of multi-omics analysis in understanding the molecular subtypes and immune characteristics of TC, and provide a novel prognostic model and potential therapeutic targets for this disease. Moreover, we identified SNAI1 in mediating TC progression through EMT in vitro.

Keywords: EMT; Multi-omics analysis; Prognostic model; Proliferation; SNAI1; Thyroid cancer.

Fig. 1

Prognostic molecular subtypes of THCA based on multi-omics consensus analysis. A Consensus heatmap of the integrated subtypes based on marker genes, including mRNA, lncRNA, miRNA, DNA CpG methylation sites, and mutated genes. B Clustering of THCA patients using 10 state-of-the-art multi-omics clustering methods. C Consensus clustering matrices of the three novel prognostic subtypes based on the 10 algorithms. D Survival analysis curves of the three subtypes in the cohort of THCA patients, shown in the form of KM curves

Fig. 2

Division and molecular features of integrated consensus molecular subtypes in THCA. A Enrichment heatmap of hallmark cancer-related features in the three subtypes. B Activity profiles of 23 TFs (top) and potential regulatory factors associated with chromatin remodeling in the three subtypes (bottom). C Immune features in the TCGA cohort. The top annotations in the heatmap show immune enrichment scores of tumor-infiltrating lymphocytes, stromal enrichment scores, and MeTIL. The top panel shows the expression of typical immune checkpoint genes, and the bottom panel shows the enrichment levels of 22 TME-related immune cells. D Validation of CS in the NTP algorithm in the TCGA cohort. E Consistency of CS with NTP in the TCGA cohort. F Consistency of CS with PAM in the TCGA cohort

Fig. 3

Development and validation of CMLS prognostic model in TCGA cohort. A Generation of 101 combinations of machine learning algorithms using a comprehensive computational framework. Calculation of C-index using the TCGA-Train, TCGA-Test, and TCGA-Entire cohorts, and ranking based on the average C-index of the validation set. B The identified 28 hub genes selected by the Ridge algorithm. C Univariate Cox regression analysis of the identified 28 hub genes in the TC training cohort. D–F Survival analysis of TC patients of different levels of CMLS scores in the TCGA-Train (D), TCGA-Test (E), and TCGA-Entire cohorts (F)

Fig. 4

Comparison of CMLS with other prognostic models. A–C Comparison of CMLS with 16 other published TC models in the TCGA-Train, TCGA-Test, and TCGA-Entire cohorts. D Comprehensive column line graph constructed based on CMLS. E Calibration curve of the comprehensive column line graph. F DCA analysis demonstrating the benefit of the comprehensive column line graph in clinical practice for patients. G Comparison of C-index over time between the comprehensive column line graph and CMLS

Fig. 5

Results of immune characteristics related to CMLS. A Distribution of TME immune cell type features between TC patients of relatively higher and lower CMLS score value. B Distribution of immune suppression features between TC patients of relatively higher and lower CMLS score value. C Distribution of immune exclusion features between TC patients of relatively higher and lower CMLS score value. D Distribution of immune therapy biomarkers between TC patients of relatively higher and lower CMLS score value. E Distribution of TMB between TC patients of relatively higher and lower CMLS score value. F The abundance of M0 macrophages in the TC patients of relatively higher and lower CMLS value. G Correlation analysis between M1 macrophages infiltrated abundance and TC patients CMLS score value. H–J Survival analysis of TC patients in the TCGA entire cohort divided by the combination of CMLS score value with TMB (H), M0 macrophages (I), and M1 macrophages (J)

Fig. 6

Results of immune therapy analysis based on CMLS score value. A Survival analysis results of TC patients of different levels of CMLS scores in the IMvigor cohort. B Box plot of CMLS in the Response and NonResponse groups in the IMvigor cohort. C Composition plot of the Response and NonResponse groups in the IMvigor cohort. D Differences in activation levels between TC patients of different levels of CMLS scores at each step of TIP. E Prediction of response to immune therapy in the TC patients of different levels of CMLS scores using the TIDE algorithm. F Differences in CMLS predicted by the TIDE algorithm between the TC patients of different levels of CMLS scores. G Differences in composition predicted by the TIDE algorithm between the TC patients of different levels of CMLS scores. H Survival analysis of TC patients of different levels of CMLS scores in the GSE78220 cohort. I Survival analysis of the high and low CMLS groups in the Braun cohort. J Distribution of CMLS in different immune therapy response groups in the GSE91061 cohort

Fig. 7

Results of drug sensitivity analysis. A GSEA plot of differential genes in the TC patients of relatively higher levels of CMLS scores (based on the hallmark gene set). B Differences in IC50 of Docetaxel_1007 between TC patients of different levels of CMLS scores. C–D Correlation and significance results of Staurosporine_1034 and Rapamycin_1084 in the GDSCv2 database. E, F Correlation and significance results of gemcitabine and topotecan in the CTRP database. G Differential analysis results of drug target genes FLT3 (staurosporine), KDR (rapamycin), and TOP1 (topotecan) between tumor and adjacent tissues. H Paired differential analysis results of drug target genes FLT3 (staurosporine), KDR (rapamycin), and TOP1 (topotecan) between tumor and adjacent tissues

Fig. 8

The knockdown of SNAI1 significantly reduced the TC proliferation and EMT phenotypes. A The mRNA transcript level of CILP in the 2 strains of undifferentiated TC cells (C643, HTH74) and 1 strain of PTC cell line (TPC1). B The mRNA transcript level of SNAI1 in the both strains of undifferentiated TC cells (C643, HTH74) and 1 strain of PTC cell line (TPC1). C The RT-qPCR validation of the knockdown of the target SNAI1 in the C643 cell line. D WB analysis showed up-regulation of EMT markers, including a-SMA and vimentin, in the C643 cell line with SNAI1 knockdown. E The CCK-8 analysis showed distinct reduction in cell proliferation in the C643 cell line with SNAI1 knockdown. F The CFA and transwell analysis results displaying the reduction in the colony formation capability and invasiveness of the C643 cell line with SNAI1 knockdown