The tumor ecosystem of papillary thyroid carcinoma (PTC) is poorly characterized. Using single-cell RNA sequencing, we profile transcriptomes of 158,577 cells from 11 patients’ paratumors, localized/advanced tumors, initially-treated/recurrent lymph nodes and radioactive iodine (RAI)-refractory distant metastases, covering comprehensive clinical courses of PTC. Our data identifies a “cancer-primed” premalignant thyrocyte population with normal morphology but altered transcriptomes. Along the developmental trajectory, we also discover three phenotypes of malignant thyrocytes (follicular-like, partial-epithelial-mesenchymal-transition-like, dedifferentiation-like), whose composition shapes bulk molecular subtypes, tumor characteristics and RAI responses. Furthermore, we uncover a distinct BRAF-like-B subtype with predominant dedifferentiation-like thyrocytes, enriched cancer-associated fibroblasts, worse prognosis and promising prospect of immunotherapy. Moreover, potential vascular-immune crosstalk in PTC provides theoretical basis for combined anti-angiogenic and immunotherapy. Together, our findings provide insight into the PTC ecosystem that suggests potential prognostic and therapeutic implications.

BackgroundDNA methylation was suggested as the promising biomarker for lung cancer diagnosis. However, it is a great challenge to search for the optimal combination of methylation biomarkers to obtain maximum diagnostic performance.ResultsIn this study, we developed a panel of DNA methylation biomarkers and validated their diagnostic efficiency for non-small cell lung cancer (NSCLC) in a large Chinese Han NSCLC retrospective cohort. Three high-throughput DNA methylation microarray datasets (458 samples) were collected in the discovery stage. After normalization, batch effect elimination and integration, significantly differentially methylated genes and the best combination of the biomarkers were determined by the leave-one-out SVM (support vector machine) feature selection procedure. Then, candidate promoters were examined by the methylation status determined single nucleotide primer extension technique (MSD-SNuPET) in an independent set of 150 pairwise NSCLC/normal tissues. Four statistical models with fivefold cross-validation were used to evaluate the performance of the discriminatory algorithms. The sensitivity, specificity and accuracy were 86.3%, 95.7% and 91%, respectively, in Bayes tree model. The logistic regression model incorporated five gene methylation signatures at AGTR1, GALR1, SLC5A8, ZMYND10 and NTSR1, adjusted for age, sex and smoking, showed robust performances in which the sensitivity, specificity, accuracy, and area under the curve (AUC) were 78%, 97%, 87%, and 0.91, respectively.ConclusionsIn summary, a high-throughput DNA methylation microarray dataset followed by batch effect elimination can be a good strategy to discover optimal DNA methylation diagnostic panels. Methylation profiles of AGTR1, GALR1, SLC5A8, ZMYND10 and NTSR1, could be an effective methylation-based assay for NSCLC diagnosis.Electronic supplementary materialThe online version of this article (doi:10.1186/s13148-014-0035-3) contains supplementary material, which is available to authorized users.

BackgroundDNA methylation has been implicated as a promising biomarker for precise cancer diagnosis. However, limited DNA methylation-based biomarkers have been described in esophageal squamous cell carcinoma (ESCC).MethodsA high-throughput DNA methylation dataset (100 samples) of ESCC from The Cancer Genome Atlas (TCGA) project was analyzed and validated along with another independent dataset (12 samples) from the Gene Expression Omnibus (GEO) database. The methylation status of peripheral blood mononuclear cells and peripheral blood leukocytes from healthy controls was also utilized for biomarker selection. The candidate CpG sites as well as their adjacent regions were further validated in 94 pairs of ESCC tumor and adjacent normal tissues from the Chinese Han population using the targeted bisulfite sequencing method. Logistic regression and several machine learning methods were applied for evaluation of the diagnostic ability of our panel.ResultsIn the discovery stage, five hyper-methylated CpG sites were selected as candidate biomarkers for further analysis as shown below: cg15830431, P = 2.20 × 10−4; cg19396867, P = 3.60 × 10−4; cg20655070, P = 3.60 × 10−4; cg26671652, P = 5.77 × 10−4; and cg27062795, P = 3.60 × 10−4. In the validation stage, the methylation status of both the five CpG sites and their adjacent genomic regions were tested. The diagnostic model based on the combination of these five genomic regions yielded a robust performance (sensitivity = 0.75, specificity = 0.88, AUC = 0.85). Eight statistical models along with five-fold cross-validation were further applied, in which the SVM model reached the best accuracy in both training and test dataset (accuracy = 0.82 and 0.80, respectively). In addition, subgroup analyses revealed a significant difference in diagnostic performance between the alcohol use and non-alcohol use subgroups.ConclusionsMethylation profiles of the five genomic regions covering cg15830431 (STK3), cg19396867, cg20655070, cg26671652 (ZNF418), and cg27062795 (ZNF542) can be used for effective methylation-based testing for ESCC diagnosis.Electronic supplementary materialThe online version of this article (10.1186/s13148-017-0430-7) contains supplementary material, which is available to authorized users.

Epigenetic alteration of tumor suppression gene is one of the most significant indicators in human esophageal squamous cell carcinoma (ESCC). In this study, we identified a novel ESCC hypermethylation biomarker ZNF132 by integrative computational analysis to comprehensive genome-wide DNA methylation microarray dataset. We validated the hypermethylation status of ZNF132 in 91 Chinese Han ESCC patients and adjacent normal tissues with methylation target bisulfite sequencing (MTBS) assay. Meanwhile, ZNF132 gene silencing mediated by hypermethylation was confirmed in both solid tissues and cancer cell lines. What is more, we found that in vitro overexpression of ZNF132 in ESCC cells could significantly reduce the abilities of the cell in growth, migration and invasion, and tumorigenicity of cells in a nude mouse model. We validated the Sp1-binding site in the ZNF132 promoter region with chromatin immunoprecipitation assay and demonstrated that the hypermethylation status could reduce the Sp1 transcript factor activity. Our results suggest that ZNF132 plays an important role in the development of ESCC as a tumor suppressor gene and support the underlying mechanism caused by the DNA hypermethylation-mediated Sp1-binding decay and gene silencing.