Full text
| |
| Author(s): |
Maria Prates Rivas
Total Authors: 1
|
| Document type: | Doctoral Thesis |
| Press: | São Paulo. |
| Institution: | Universidade de São Paulo (USP). Instituto de Biociências (IBIOC/SB) |
| Defense date: | 2021-07-30 |
| Examining board members: |
Ana Cristina Victorino Krepischi;
Cecilia Maria Lima da Costa;
Oswaldo Keith Okamoto;
Eduardo Moraes Rego Reis
|
| Advisor: | Ana Cristina Victorino Krepischi |
| Abstract | |
Epigenetic mechanisms in hepatoblastomas: transcriptome analysis and regulation by DNA methylation. Liver tumors in children are rare, representing 1 - 4% of all pediatric solid tumors, with hepatoblastoma being the most frequent. Hepatoblastoma is an embryonic tumor with a very early age of manifestation and characteristic histological presentation that recapitulates different phases of liver organogenesis. Due to its rarity, the molecular mechanisms underlying the development and progression of this tumor type are still poorly explored, and there are few biomarkers for risk stratification. With the lowest rate of somatic mutation among solid tumors, the origin and molecular mechanisms of hepatoblastoma are still discussed in the literature. There is a recognized link between epigenetics, fetal development, and pediatric tumors and, in recent years, the study focused on the epigenetics of hepatoblastomas has gained greater importance as a key mechanism. This study had as its central objective the evaluation of the role of DNA methylation in altering the pattern of gene expression of hepatoblastomas. A total of 30 samples of hepatoblastomas and 11 samples of non-tumor liver tissue were evaluated, derived from collaboration with three pediatric cancer centers in São Paulo (AC Camargo Cancer Center, GRAAC, and ITACI). Liver tumor cell lines and induced pluripotent cells (IPSCs) differentiated into hepatoblasts and hepatocytes-like were also used. Using techniques such as qRT-PCR, western blot, immunohistochemistry, metabolomics, RNASeq, and global level of hydroxymethylation, the objectives were (a) to identify the mechanism of DNA hypomethylation in hepatoblastomas and to determine the level of 5-hydroxymethylcytosine; (b) characterize the samples of hepatoblastoma according to the steps of cell differentiation from hepatocytes; (c) assess the impact of previously identified changes in the tumor methylome on the total transcriptome (coding and non-coding genes). We studied the gene expression pattern of DNA methylation machinery (DNMT1, DNMT3A, DNMT3B, DNMT3L, UHRF1, TET1, TET2, and TET3) in these tumors and investigated the overall level of 5hmC in hepatoblastoma. We detected high expression of epigenetic machinery genes, mainly UHRF1, TET1, and TET2, in association with enrichment of the content of 5hmC. These results indicate the occurrence of active DNA demethylation in hepatoblastomas, mediated by TETs, probably during the early stages of liver development. In addition, lower levels of 5hmC in hepatoblastomas may be associated with reduced overall survival. In addition, we evaluated the use of the gene expression pattern of epigenetic machinery and liver differentiation as pivotal elements for hepatoblastoma stratification. It was possible to stratify tumor samples into three groups, highlighting the importance of 13 (TET1, TET2, TET3, DNMT1, DNMT3A, UHRF1, ALB, CYP3A4, TDO2, UGT1A1, AFP, HNF4A, and FOXA2) genes for the establishment of this classification. The results made it possible to correlate the identified groups with different levels of global DNA methylation, which are in accordance with the phase of liver differentiation in which the tumor cells would meet, which corroborates the hypothesis of blocking the differentiation of the source organ. Thus, we proposed tumor signatures associated with different stages of liver differentiation based on gene expression and DNA methylation level. We evidenced a significant reduction in NNMT gene and protein expression in hepatoblastomas. The higher expression of NNMT was statistically associated with a late diagnosis of hepatoblastoma, a recognized clinical variable with a worse prognosis. The undirected metabolomic analysis detected aberrant lipid metabolism in hepatoblastoma, suggesting that the decrease in NNMT may be associated with the detected reduction in lipid content in tumor samples, possibly related to altered concentrations of metabolic linked to NNMT function or NAD+ 123 precursors. The data presented here showed for the first time that the reduction of NNMT occurs in hepatoblastomas, providing additional support for the theory that the negative regulation of this protein is a general phenomenon in liver cancer. To better understand the altered biological processes, we performed bioinformatics analyzes based on data from RNASeq of normal hepatoblastomas and liver. A group of 1492 genes with differential expression between tumors and control (Differentially Expressed Genes - DEGs) was identified, 1,031 of them regulated positively and 461 with decreased expression. A list of DEGs contains genes encoding proteins and ncRNAs, which are mainly upregulated. Eighty-five of the genes with differential expression are expressed exclusively in our tumors, with approximately 55% of them being ncRNAs. Among the 50 main genes, defined here as those with the greatest expression alteration, 42% correspond to ncRNAs; three (MAGED4, SNO144-16 and RP11-431J24.2) had detectable expression exclusively in tumors and, in addition, 16 of these genes with greater difference in expression between tumors and normal liver had previously been identified as altered in this tumor group (GCK, HMGA2, INS-IGF2, DKK1, CPA6, TTC36, LIX1, PART1, DKK4, CDCA7, GNG4, THRSP, CNDP1, CYP2B7P, ASPG and HAO2). The biological processes were enriched in the sense of negative regulation of metabolism, mainly of amine, nicotinamide and lipids/fatty acids; there is also a change in epigenetic mechanisms, such as DNA methylation, ncRNAs and gene silencing by miRNA; cytotoxicity and signaling pathways, including cytokine-mediated signaling pathway. A miRNA-mRNA interaction network was built, identifying four genes (miR-186, miR-214, miR-377 and miR-494) where each is connected to more than 10 mRNAs. The main biological processes related to this miRNA-mRNA network were also associated with metabolism and oxidation reactions of lipids and carbohydrates. In conclusion, the analysis of RNASeq showed great alteration of metabolic pathways in hepatoblastomas, including lipids, amines and nicotinamides, which resumes previous results of our group. It also revealed major changes in the ncRNA expression network in this type of tumor. (AU) | |
| FAPESP's process: | 16/23462-8 - Epigenetic mechanisms in liver tumors: expression modulation of epigenetic gene regulators and gene expression analysis by RNAseq in hepatoblastoma. |
| Grantee: | Maria Prates Rivas |
| Support Opportunities: | Scholarships in Brazil - Doctorate |