Identification of aberrantly methylated differentially expressed genes in glioblastoma multiforme and their association with patient survival

Methods in DNA methylation array dataset analysis: A review

Understanding the intricate relationships between gene expression levels and epigenetic modifications in a genome is crucial to comprehending the pathogenic mechanisms of many diseases. With the advancement of DNA Methylome Profiling techniques, the emphasis on identifying Differentially Methylated Regions (DMRs/DMGs) has become crucial for biomarker discovery, offering new insights into the etiology of illnesses. This review surveys the current state of computational tools/algorithms for the analysis of microarray-based DNA methylation profiling datasets, focusing on key concepts underlying the diagnostic/prognostic CpG site extraction. It addresses methodological frameworks, algorithms, and pipelines employed by various authors, serving as a roadmap to address challenges and understand changing trends in the methodologies for analyzing array-based DNA methylation profiling datasets derived from diseased genomes. Additionally, it highlights the importance of integrating gene expression and methylation datasets for accurate biomarker identification, explores prognostic prediction models, and discusses molecular subtyping for disease classification. The review also emphasizes the contributions of machine learning, neural networks, and data mining to enhance diagnostic workflow development, thereby improving accuracy, precision, and robustness.

Identification of genomic biomarkers and their pathway crosstalks for deciphering mechanistic links in glioblastoma

Glioblastoma is a grade IV pernicious neoplasm occurring in the supratentorial region of brain. As its causes are largely unknown, it is essential to understand its dynamics at the molecular level. This necessitates the identification of better diagnostic and prognostic molecular candidates. Blood-based liquid biopsies are emerging as a novel tool for cancer biomarker discovery, guiding the treatment and improving its early detection based on their tumour origin. There exist previous studies focusing on the identification of tumour-based biomarkers for glioblastoma. However, these biomarkers inadequately represent the underlying pathological state and incompletely illustrate the tumour because of non-recursive nature of this approach to monitor the disease. Also, contrary to the tumour biopsies, liquid biopsies are non-invasive and can be performed at any interval during the disease span to surveil the disease. Therefore, in this study, a unique dataset of blood-based liquid biopsies obtained primarily from tumour-educated blood platelets (TEP) is utilised. This RNA-seq data from ArrayExpress is acquired comprising human cohort with 39 glioblastoma subjects and 43 healthy subjects. Canonical and machine learning approaches are applied for identification of the genomic biomarkers for glioblastoma and their crosstalks. In our study, 97 genes appeared enriched in 7 oncogenic pathways (RAF-MAPK, P53, PRC2-EZH2, YAP conserved, MEK-MAPK, ErbB2 and STK33 signalling pathways) using GSEA, out of which 17 have been identified participating actively in crosstalks. Using PCA, 42 genes are found enriched in 7 pathways (cytoplasmic ribosomal proteins, translation factors, electron transport chain, ribosome, Huntington's disease, primary immunodeficiency pathways, and interferon type I signalling pathway) harbouring tumour when altered, out of which 25 actively participate in crosstalks. All the 14 pathways foster well-known cancer hallmarks and the identified DEGs can serve as genomic biomarkers, not only for the diagnosis and prognosis of Glioblastoma but also in providing a molecular foothold for oncogenic decision making in order to fathom the disease dynamics. Moreover, SNP analysis for the identified DEGs is performed to investigate their roles in disease dynamics in an elaborated manner. These results suggest that TEPs are capable of providing disease insights just like tumour cells with an advantage of being extracted anytime during the course of disease in order to monitor it.

The Profiles of Tet-Mediated DNA Hydroxymethylation in Human Gliomas

Gliomas are the most common primary malignant intracranial brain tumors. Their proliferative and invasive behavior is controlled by various epigenetic mechanisms. 5-hydroxymethylcytosine (5-hmC) is one of the epigenetic DNA modifications that employs ten-eleven translocation (TET) enzymes to its oxidation. Previous studies demonstrated altered expression of 5-hmC across gliomagenesis. However, its contribution to the initiation and progression of human gliomas still remains unknown. To characterize the expression profiles of 5-hmC and TET in human glioma samples we used the EpiJET 5-hmC and 5-mC Analysis Kit, quantitative real-time PCR, and Western blot analysis. A continuous decline of 5-hmC levels was observed in solid tissue across glioma grades. However, in glioblastoma (GBM), we documented uncommon heterogeneity in 5-hmC expression. Further analysis showed that the levels of TET proteins, but not their transcripts, may influence the 5-hmC abundance in GBM. Early tumor-related biomarkers may also be provided by the study of aberrant DNA hydroxymethylation in the blood of glioma patients. Therefore, we explored the patterns of TET transcripts in plasma samples and we found that their profiles were variously regulated, with significant value for TET2. The results of our study confirmed that DNA hydroxymethylation is an important mechanism involved in the pathogenesis of gliomas, with particular reference to glioblastoma. Heterogeneity of 5-hmC and TET proteins expression across GBM may provide novel insight into define subtype-specific patterns of hydroxymethylome, and thus help to interpret the heterogeneous outcomes of patients with the same disease.

Promoter Demethylation Upregulates STEAP1 Gene Expression in Human Prostate Cancer: In Vitro and In Silico Analysis

The Six Transmembrane Epithelial Antigen of the Prostate (STEAP1) is an oncogene overexpressed in several human tumors, particularly in prostate cancer (PCa). However, the mechanisms involved in its overexpression remain unknown. It is well known that epigenetic modifications may result in abnormal gene expression patterns, contributing to tumor initiation and progression. Therefore, this study aimed to analyze the methylation pattern of the STEAP1 gene in PCa versus non-neoplastic cells. Bisulfite amplicon sequencing of the CpG island at the STEAP1 gene promoter showed a higher methylation level in non-neoplastic PNT1A prostate cells than in human PCa samples. Bioinformatic analysis of the GEO datasets also showed the STEAP1 gene promoter as being demethylated in human PCa, and a negative association with STEAP1 mRNA expression was observed. These results are supported by the treatment of non-neoplastic PNT1A cells with DNMT and HDAC inhibitors, which induced a significant increase in STEAP1 mRNA expression. In addition, the involvement of HDAC in the regulation of STEAP1 mRNA expression was corroborated by a negative association between STEAP1 mRNA expression and HDAC4,5,7 and 9 in human PCa. In conclusion, our work indicates that STEAP1 overexpression in PCa can be driven by the hypomethylation of STEAP1 gene promoter.

Scouting for common genes in the heterogenous hypoxic tumor microenvironment and their validation in glioblastoma

Investigating the therapeutic and prognostic potential of genes in the heterogeneous hypoxic niche of glioblastoma. We have analyzed RNA expression of U87MG cells cultured in hypoxia compared to normoxia. Common differentially expressed genes (DEGs) from GSE45301 and GSE18494 and their functional enrichment was performed using MetaScape and PANTHER. Hub genes and their ontology were identified using MCode cytoHubba and ClueGO and validated with GlioVis, Oncomine, HPA and PrognoScan. Using the GEO2R analysis of GSE45301 and GSE18494 datasets, we have found a total of 246 common DEGs (180 upregulated and 66 downregulated) and identified 2 significant modules involved in ribosome biogenesis and TNF signaling. Meta-analysis of key genes of each module in cytoHubba identified 17 hub genes (ATF3, BYSL, DUSP1, EGFR, JUN, ETS1, LYAR, NIP7, NOLC1, NOP2, NOP56, PNO1, RRS1, TNFAIP3, TNFRSF1B, UTP15, VEGFA). Of the 17 hub genes, ATF3, BYSL, EGFR, JUN, NIP7, NOLC1, PNO1, RRS1, TNFAIP3 and VEGFA were identified as hypoxia signatures associated with poor prognosis in Glioma. Ribosome biogenesis emerged as a vital contender of possible therapeutic potential with BYSL, NIP7, NOLC1, PNO1 and RRS1 showing prognostic value.

SUPPLEMENTARY INFORMATION

The online version contains supplementary material available at 10.1007/s13205-021-02987-2.

Biomarker Screening and Prognostic Significance Analysis for Renal Cell Carcinoma

Background

Studies report that conventional treatment of clear cell renal cell carcinoma (ccRCC) is effective, but several advanced patients present with poor prognosis. The current study explored potential new tumor markers and therapeutic targets in advanced ccRCC.

Methods

Biomarker gene expression of ccRCC was retrieved from GEO database and the Cancer Genome Atlas Kidney Clear Cell Carcinoma (TCGA-KIRC) database. Gene ontology (GO) analysis and protein–protein interaction (PPI) networks of biomarker genes were constructed using the Database for Annotation, Visualization, and Integrated Discovery (DAVID) tool. Kaplan–Meier analysis and receiver operating characteristic curve (ROC) analysis were performed to explore the prognostic and diagnostic roles of these genes. Gene set enrichment analysis (GSEA) analysis was used to determine hallmark functions of the biomarker genes. qRT-PCR was used to verify the reliability of the analysis results in tumor tissues.

Results

A total of 21 upregulated genes were identified between advanced ccRCC and early ccRCC (grade III+IV vs grade I+II). Gene ontology analysis showed that the 21 upregulated genes were mainly implicated in biological processes including metabolic and lipid transport. The findings showed that 7 out of the 21 genes were significantly upregulated in 72-paired samples retrieved from the TCGA-KIRC. High expression of 5 genes indicated a poor prognosis of overall survival and disease-free survival in KIRC. Three genes effectively distinguished renal cancer tissue and adjacent renal tissues in a total of 533 ccRCC samples. GSEA showed that the 3 biomarkers were significantly enriched in epithelial–mesenchymal transition, G2M checkpoint, and angiogenesis. The results of qRT-PCR showed that STEAP3, IBSP, and AQP9 had a significant identification effect in ccRCC.

Conclusion

The findings showed that 3 biomarkers were significantly upregulated in advanced ccRCC and could be used for diagnosis, prediction, and potential novel therapeutic targets for progression of ccRCC.

Epigenetic regulation of somatostatin and somatostatin receptors in neuroendocrine tumors and other types of cancer

Both somatostatin (SST) and somatostatin receptors (SSTRs) are proteins with important functions in both physiological tissue and in tumors, particularly in neuroendocrine tumors (NETs). NETs are frequently characterized by high SSTRs expression levels. SST analogues (SSAs) that bind and activate SSTR have anti-proliferative and anti-secretory activity, thereby reducing both the growth as well as the hormonal symptoms of NETs. Moreover, the high expression levels of SSTR type-2 (SSTR2) in NETs is a powerful target for therapy with radiolabeled SSAs. Due to the important role of both SST and SSTRs, it is of great importance to elucidate the mechanisms involved in regulating their expression in NETs, as well as in other types of tumors. The field of epigenetics recently gained interest in NET research, highlighting the importance of this process in regulating the expression of gene and protein expression. In this review we will discuss the role of the epigenetic machinery in controlling the expression of both SSTRs and the neuropeptide SST. Particular attention will be given to the epigenetic regulation of these proteins in NETs, whereas the involvement of the epigenetic machinery in other types of cancer will be discussed as well. In addition, we will discuss the possibility to target enzymes involved in the epigenetic machinery to modify the expression of the SST-system, thereby possibly improving therapeutic options.

Performance Comparison of Deep Learning Autoencoders for Cancer Subtype Detection Using Multi-Omics Data

A heterogeneous disease such as cancer is activated through multiple pathways and different perturbations. Depending upon the activated pathway(s), the survival of the patients varies significantly and shows different efficacy to various drugs. Therefore, cancer subtype detection using genomics level data is a significant research problem. Subtype detection is often a complex problem, and in most cases, needs multi-omics data fusion to achieve accurate subtyping. Different data fusion and subtyping approaches have been proposed over the years, such as kernel-based fusion, matrix factorization, and deep learning autoencoders. In this paper, we compared the performance of different deep learning autoencoders for cancer subtype detection. We performed cancer subtype detection on four different cancer types from The Cancer Genome Atlas (TCGA) datasets using four autoencoder implementations. We also predicted the optimal number of subtypes in a cancer type using the silhouette score and found that the detected subtypes exhibit significant differences in survival profiles. Furthermore, we compared the effect of feature selection and similarity measures for subtype detection. For further evaluation, we used the Glioblastoma multiforme (GBM) dataset and identified the differentially expressed genes in each of the subtypes. The results obtained are consistent with other genomic studies and can be corroborated with the involved pathways and biological functions. Thus, it shows that the results from the autoencoders, obtained through the interaction of different datatypes of cancer, can be used for the prediction and characterization of patient subgroups and survival profiles.

Regulatory Roles of Six-Transmembrane Epithelial Antigen of the Prostate Family Members in the Occurrence and Development of Malignant Tumors

The human six-transmembrane epithelial antigen of the prostate (STEAP) proteins, which include STEAP1-4 and atypical STEAP1B, contain six transmembrane domains and are located in the cell membrane. STEAPs are considered archaeal metal oxidoreductases, based on their heme groups and F420H2:NADP+ oxidoreductase (FNO)-like structures, and play an important role in cell metal metabolism. Interestingly, STEAPs not only participate in biological processes, such as molecular transport, cell cycling, immune response, and intracellular and extracellular activities, but also are closely related to the occurrence and development of several diseases, especially malignant tumors. Up to now, the expression patterns of STEAPs have been found to be diverse in different types of tumors, with controversial participation in different aspects of malignancy, such as cell proliferation, migration, invasion, apoptosis, and therapeutic resistance. It is clinically important to explore the potential roles of STEAPs as new immunotherapeutic targets for the treatment of different malignant tumors. Therefore, this review focuses on the molecular mechanism and function of STEAPs in the occurrence and development of different cancers in order to understand the role of STEAPs in cancer and provide a new theoretical basis for the treatment of diverse cancers.

Mesenchymal Transformation: The Rosetta Stone of Glioblastoma Pathogenesis and Therapy Resistance

Despite decades of research, glioblastoma (GBM) remains invariably fatal among all forms of cancers. The high level of inter- and intratumoral heterogeneity along with its biological location, the brain, are major barriers against effective treatment. Molecular and single cell analysis identifies different molecular subtypes with varying prognosis, while multiple subtypes can reside in the same tumor. Cellular plasticity among different subtypes in response to therapies or during recurrence adds another hurdle in the treatment of GBM. This phenotypic shift is induced and sustained by activation of several pathways within the tumor itself, or microenvironmental factors. In this review, the dynamic nature of cellular shifts in GBM and how the tumor (immune) microenvironment shapes this process leading to therapeutic resistance, while highlighting emerging tools and approaches to study this dynamic double-edged sword are discussed.

Identification of Key Differentially Expressed Transcription Factors in Glioblastoma

Glioblastoma (GBM) is the most frequent malignant brain tumor in adults. Our study focused on uncovering differentially expressed genes (DEGs) and their methylation in order to identify novel diagnostic biomarkers and potential treatment targets. Using GBM RNA-sequencing data from The Cancer Genome Atlas (TCGA) database, DEGs between GBM samples and paracancer tissue samples were analyzed. Enrichment analysis for DEGs and transcription factors (TFs) was performed. A total of 1029 upregulated genes and 1542 downregulated genes were identified, which were associated mainly with multiple tumor-related and immune-related pathways such as cell cycle, mitogen-activated protein kinase signaling pathway, leukocyte transendothelial migration, and autoimmune thyroid disease. These DEGs were enriched for 174 TFs, and six TFs were differentially expressed and identified as key TFs in GBM: HOXA3, EN1, ZIC1, and FOXD3 were upregulated, while HLF and EGR3 were downregulated. A total of 1978 DEGs were involved in the regulatory networks of the six key differentially expressed TFs. High expression of EN1 was associated with shorter overall survival, while high expression of EGR3 was associated with shorter recurrence-free survival. The six TFs were differentially methylated in GBM samples compared with paracancer tissues. Our study identifies numerous DEGs and their associated pathways as potential contributors to GBM, particularly the TFs EN1, EGR3, HOXA3, ZIC1, FOXD3, and HLF. The differential expression of these TFs may be unlikely driven by aberrant methylation. These TFs may be useful as diagnostic markers and treatment targets in GBM, and EN1 and EGR3 may have predictive prognostic value.

Identification of aberrantly methylated differentially expressed genes and associated pathways in endometrial cancer using integrated bioinformatic analysis

Endometrial cancer (EC) is a fatal female reproductive tumor. Bioinformatic tools are increasingly developed to screen out molecular targets related to EC. In this study, http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE17025 and http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE40032 were obtained from Gene Expression Omnibus (GEO). “limma” package and Venn diagram tool were used to identify hub genes. FunRich was used for functional analysis. Retrieval of Interacting Genes Database (STRING) was used to analyze protein‐protein interaction (PPI) complex. Cancer Genome Atlas (TCGA), GEPIA, immunohistochemistry staining, and ROC curve analysis were carried out for validation. Univariate and multivariate regression analyses were performed to predict the risk score. Compound muscle action potential (CMap) was used to find potential drugs. GSEA was also done. We retrieved seven oncogenes which were upregulated and hypomethylated and 12 tumor suppressor genes (TSGs) which were downregulated and hypermethylated. The upregulated and hypomethylated genes were strikingly enriched in term “immune response” while the downregulated and hypermethylated genes were mainly focused on term “aromatic compound catabolic process.” TCGA and GEPIA were used to screen out EDNRB, CDO1, NDN, PLCD1, ROR2, ESPL1, PRAME, and PTTG1. Among them, ESPL1 and ROR2 were identified by Cox regression analysis and were used to construct prognostic risk model. The result showed that ESPL1 was a negative independent prognostic factor. Cmap identified aminoglutethimide, luteolin, sulfadimethoxine, and maprotiline had correlation with EC. GSEA results showed that “hedgehog signaling pathway” was enriched. This research inferred potential aberrantly methylated DEGs and dysregulated pathways may participate in EC development and firstly reported eight hub genes, including EDNRB, CDO1, NDN, PLCD1, ROR2, ESPL1, PRAME, and PTTG1 that could be used to predict EC prognosis. Aminoglutethimide and luteolin may be used to fight against EC.

Identification of DNA methylation-driven genes by integrative analysis of DNA methylation and transcriptome data in pancreatic adenocarcinoma

Pancreatic adenocarcinoma (PAAD) is a painful and fatal disease that undoubtedly remains a health care priority and offers significant therapeutic challenges. The significance of epigenetic modifications, including DNA methylation in tumor development, has gained the attention of researchers. Identifying DNA methylation-driven genes and investigating the mechanisms underlying the tumorigenesis of PAAD are of substantial importance for developing methods of physiological evaluation, treatment planning and prognostic prediction for PAAD. In the present study, a comprehensive analysis of DNA methylation and gene expression data from 188 clinical samples was performed to identify DNA methylation-driven genes in PAAD. In addition, the diagnostic and prognostic value of DNA methylation-driven genes was evaluated using receiver operating characteristic curve, survival and recurrence analyses. A total of 7 DNA methylation-driven genes, namely zinc finger protein 208 (ZNF208), eomesodermin (EOMES), prostaglandin D2 receptor (PTGDR), chromosome 12 open reading frame 42 (C12orf42), integrin subunit α 4 (ITGA4), dedicator of cytokinesis 8 and protein phosphatase 1 regulatory inhibitor subunit 14D (PPP1R14D), were identified. All of them may be used to diagnose PAAD with excellent specificity and sensitivity (area under curve, >0.8). Of the 7 DNA methylation-driven genes, 6 were significantly associated with overall survival (OS) and recurrence-free survival (RFS) P<0.05). Among them, ZNF208, EOMES, PTGDR, C12orf42 and ITGA4 were significantly negatively associated with the OS rate and positively associated with the recurrence rate, while PPP1R14D was significantly positively associated with the OS rate and negatively associated with the recurrence rate. The present study provides novel insight into the epigenetic alterations associated with the occurrence and progression of PAAD, thereby increasing the mechanistic understanding of this disease, offering potential novel molecular biomarkers and contributing to the development of therapeutic targets for PAAD.

The Tumor Suppressive Roles and Prognostic Values of STEAP Family Members in Breast Cancer

OBJECTIVE

To investigate the expression patterns and prognostic values of STEAP family members in the occurrence and development of breast cancer.

MATERIALS AND METHODS

The Human Protein Atlas was used to analyze the expression level of STEAPs in human normal tissues and malignant tumors. ONCOMINE datasets were analyzed for the comparison of the STEAPs levels between malignant cancers and corresponding normal tissues. Kaplan-Meier plotter was used to analyze the prognostic value of STEAPs in breast cancer patients.

RESULTS

STEAPs were widely distributed in human normal tissues with diverse levels. Normally, it is predicted that STEAP1 and STEAP2 were involved in the mineral absorption process, while STEAP3 participated in the TP53 signaling pathway and iron apoptosis. The results from ONCOMINE showed downregulation of STEAP1, STEAP2, and STEAP4 in breast cancers. Survival analysis revealed that breast cancer patients with high levels of STEAP1, STEAP2, and STEAP4 had a good prognosis, while those with low expression had high overall mortality.

CONCLUSION

STEAP1, STEAP2, and STEAP4 are predicted to be the potential prognostic biomarkers for breast cancer patients, providing novel therapeutic strategies for them.