Introduction Low-grade brain gliomas (LGG) are slow-growing tumors originating from glial cells. LGG are the most frequent brain neoplasms in children and young adults. According to WHO classification the most frequent histotypes are pilocytic astrocytoma (PA) (WHO grade I), ganglioglioma (WHO grade I) pleomorphic xantoastrocytoma (WHO grade II) and diffuse gliomas (WHO grade II). PAs most frequently occur in the cerebellum, but are also found in other areas of the infratentorial (FCP) region like, brain stem, and fourth ventricle and in areas of the supratentorial (SVT) region like optic chiasm, diencephalon, third ventricle, and cerebral. Despite excellent overall survival of the patients, tumor recurrence is common. This is intrinsically linked to surgical accessibility, with complete resection associ ated with significantly longer progression-free survival than subtotal, partial or no resection. As such, PA is often considered as a chronic disease, and both the tumor and treatments can cause significant morbidity. The identification of clinically relevant molecules could allow to improve the diagnosis and to develop innovative therapies against specific molecular targets. Materials and Methods Genome-wide methylation at 27,578 CpG sites (spanning 14,495 genes) was performed in 20 PAs and on 4 non-tumoral brain tissue using the Illumina Infinium HumanMethylation27 (27K) assay. The clustering and subgroups classification were conducted using unsupervised hierarchical clustering provided by R function “hClust”. Pathway enrichment analyses were performed using the web-based tool ToppGene Suite. Gene expression validation was performed by Real-Time qPCR. Results and Discussion We found relevant differences in DNA methylation profile between tumoral and non-tumoral samples. Such differences correspond to pathways involved in carcinogenesis as expected. We then performed a differential methylation analysis on two different subgroups obtained considering two distinct features: tumor localization (SVT vs FCP) and age of onset (≤ 3 yo vs > 3 yo). The unsupervised hierarchical clustering shows how the relevant differential methylation regions (adjusted p-value < 0.01) are able to identify in each subgroup two distinct clusters that reflect quite substantially the subgroup categories. We then restricted the pathway enrichment analysis to the promoter CpG Island loci, more likely to be involved in the gene expression regulation, to highlight biological relevant pathways impacted by DNA methylation. We found that the most implicated ones (adjusted p-value < 0.05) in both subgroups are Cadherin signaling and Wnt signaling pathways. We performed an enrichment and validation of 27K CpG Island significant results from our study, using a 450K dataset, from Lambert et al, focusing on SVT vs FCP subgroup. 62% of CpG islands were confirmed as having a differential methylation “direction” concordant to our 27K results. The validation by Real-Time qPCR of two selected genes hypermethylated in the SVT group shows a downregulation for both genes. From the comparison to the UCSC gene expression level observed in the normal tissue from the same localizations, we identified two classes of biomarkers: tumoral and topographical biomarkers. Conclusions Our results indicate that a differential methylation analysis is able to identify alterations which may affect gene pathways important for the tumor development and progression, and to provide useful biomarkers distinguishing clinical relevant tumor categories.
Scheda prodotto non validato
Attenzione! I dati visualizzati non sono stati sottoposti a validazione da parte dell'ateneo
|Titolo:||Searching for methylome alterations in human low-grade gliomas as potential diagnostic and prognostic epigenetic biomarkers|
|Data di pubblicazione:||2016|