Year: 2019
Type: article
Abstract: Epigenetic alterations, including changes in DNA methylation, have been identified in cancer and glioblastoma (GBM), where they regulate tumor initiation and progression. While methylation of cytosine to form 5-methylcytosine has been the focus of most research on DNA methylation in cancer, other DNA alterations have also been described. Among these is the recent discovery of the N6-methyladenine (N6-mA) modification in mammals, the role of which remains poorly understood. A recent study published in the journal Cell investigated the expression, regulation, and function of N6-mA in GBM.1 To investigate the role of N6-mA in GBM, the authors first assessed its presence in glioblastoma stem cells (GSC) and primary human tumors using dot blot analysis and a validated N6-mA antibody. They found that N6-mA levels were significantly higher in GSC and GBM tumor tissue than in normal human astrocytes. They then analyzed the genomic localization of N6-mA and its association with its closest genes using DIP-seq and gene ontology enrichment analysis, respectively. They found that N6-mA was enriched at constitutive heterochromatin regions and near genes that regulate neurogenesis and neuronal development pathways. Further analysis of N6-mA localization via ChIP-seq methodology revealed its co-localization with heterochromatic histone modifications and with repressive DNA methylation marks, suggesting that N6-mA acts as a repressive mark. This was confirmed by the demonstration of repression of tumor suppressor genes (CDKN3, RASSF2, and AKAP6) by N6-mA. To determine how N6-mA is regulated in GBM, the authors then performed a series of experiments and identified ALKBH1 as the demethylase that (down)regulates N6-mA in GBM. They showed that ALKBH1 silencing facilitates heterochromatin formation in GBM by regulating N6-mA, which primarily acts through the H3K9me3 histone modification. The authors then investigated genes regulated by ALKBH1 and uncovered hypoxia-induced genes as a major target of ALKBH1 in GBM. Subsequently, the therapeutic potential of inhibiting ALKBH1 with shRNA or CRISPR-Cas9 was assessed in GSCs and in orthotopic GBM xenografts. Inhibition of ALKBH1 led to inhibition of in vitro malignancy parameters and prolongation of animal survival. These data suggest that ALKBH1 is a new therapeutic target in GBM. An apparent contradiction in the presented data is that ALKBH1 inhibition has anti-tumor effects although it leads to upregulation of N6-mA, which is associated with repression of tumor suppressors. To explain this contradiction, the authors speculated that the effects of N6-mA might depend more on its flux and not as much on its absolute levels. In summary, the study demonstrated that N6-mA modifications are present in GBM. These modifications regulate tumor cell growth, are regulated by ALKBH1, and can be potentially targeted for therapy. Although the incidence of primary brain cancer is less compared to other cancer types, the number of patients with brain metastases exceeds 150,000 per year in the United States alone. In the era of personalized medicine, treatment of brain metastases can be tailor made with a proper diagnosis of the primary cancer of origin. Current methods rely mainly on immunohistochemical analyses, which can be subjective and inconclusive. DNA methylation profiling has been shown to be a powerful method for both diagnostic and prognostic classification of cancer. A recent study by Orozco et al. demonstrates the application of DNA methylome for the molecular classification of metastatic brain tumors.1 The authors first established a pipeline after eliminating probes that recognize common SNPs and high GC regions and cause gender bias. Using these filtered probe sets, the authors demonstrate significant DNA methylation differences between primary brain tumors (glioblastomas) and brain metastases. This was confirmed by an analysis of 12 genomic regions for methylation differences using quantitative methylation-specific PCR (qMSP) in an independent cohort of samples. The authors further demonstrate the application of this study by showing differences in methylation signatures in brain metastases samples from differential tissue of origin (breast, skin, and lung) as well as differential methylation patterns that discriminate breast cancer brain metastases based on hormone receptor and HER2 status. These findings were supported by immunohistochemical confirmation. This study not only provides DNA methylation classifiers that can accurately diagnose brain metastases but also paves the way for future clinical application of this methodology using qMSP of tumor tissues. Although immunohistochemistry is a well-established method for intracranial tumor diagnoses, combining methylome assays with immunohistochemistry can lead to more accurate diagnoses and in turn improved therapies in the future for brain metastases. Neurofibromatosis type 1 (NF1) is a common cancer predisposition syndrome that also encompasses low-and high-grade gliomas. NF1 is a negative regulator of RAS; hence, loss of function of NF1 results in an overactive RAS pathway. How NF1-gliomas compare with sporadic gliomas was investigated in a multi-OMICs approach by D’Angelo et al.1 investigating 59 gliomas (35 low- and 24 high-grade) from patients affected with NF1. Peak incidences were observed at 13.5 and 38.8 years. Pediatric NF1-gliomas were more commonly low grade, and the burden of genetic alterations was associated with tumor grade. High-grade NF1-glioma comprised frequent deletions of CDKN2A (58%) and mutations of ATRX (38%) and TP53 (29%); furthermore, mutations affected genes in the PI3-kinase pathway. In contrast, no mutations were identified in characteristic glioma-associated genes such as IDH1/2, the histone H3.3 genes, or TERT. However, gain of TERT copy number and FGF1 was commonly observed together (both 47%) in low-grade NF1-gliomas. Using gene expression profiles the authors identified a subgroup of low-grade NF1 gliomas that exhibited enhanced immune gene expression. More detailed analysis associated CD8+ T-cell-specific signatures with increased T-cell infiltrates. Along the same lines, expression of neo-antigens was found associated with the high- but not the low immune expresser low-grade NF1 glioma subgroup. However, no association of high immune gene expression was found with any morphologic subtype. Based on these interesting findings, the authors suggested that enhanced immune surveillance of these low-grade NF1 gliomas may contribute to their indolent behavior. To compare the NF1-gliomas to sporadic gliomas, they were classified according to a DNA methylation-based classifier that uses 1233 “glioma-specific” methylation probes established with the TCGA pan-glioma dataset that classifies all gliomas into 6 subclasses: 3 for the IDH mutant gliomas (G-CIMP-low, G-CIMP-high, 1p/19q codel) and 3 for IDH wild-type gliomas (Classic-like, Mesenchymal-like, and LGm6).2 The NF1-gliomas all classified with the heterogeneous group of LGm6 that comprises grade 2, 3, and 4 gliomas, of which some are denoted pilocytic like (PA-like). Based on a more detailed analysis of the whole LGm6 group, ATRX mutations seemed to be associated with worse outcome, regardless of tumor grade. The authors propose that ATRX mutations, or loss of ATRX expression in IDH wild-type gliomas, maybe worthy to be considered as biomarkers in the WHO tumor classification. The rich dataset presented here is published as a resource paper that provides many more interesting insights on the underpinnings of NF1-gliomas. It will be of interest to learn how these well-characterized NF1-gliomas classify according to the recently published methylation classifier3 that allows for discovery of new entities. Chromothripsis and chromoanasynthesis are catastrophic events that lead to clustered genomic rearrangements and affect one or very few chromosomes.1 In humans, chromothripsis was detected across a wide range of tumor entities, whereas chromoanasynthesis was mainly described in the context of congenital diseases.2 In a recent issue of Nature Communications Ratnaparkhe and colleagues3 demonstrated by using whole-genome sequencing frequent complex genomic rearrangements (n=16/26) in brain tumors developing in mice deficient for factors involved in homologous-recombination-repair or non-homologous-end-joining. Catastrophic events were tightly linked to Myc/Mycn amplification, with increased DNA damage and inefficient apoptotic response already observable at early postnatal stages. Inhibition of repair processes and comparison of the mouse tumors with human medulloblastomas (n=68) and glioblastomas (n=32) identified chromothripsis as being associated with MYC/MYCN gains and with DNA repair deficiencies. In analogy to the clinical use of PARP inhibitors in the context of BRCA-deficient breast cancer, the findings of this important paper point toward therapeutic opportunities to target DNA repair defects in tumors with complex genomic rearrangements. Future studies will help in defining the specific subsets of patients who may benefit from such a treatment. The widespread adoption of next-generation sequencing methods has generated a tremendous wealth of molecular information and revealed the substantial plasticity and heterogeneity of transcriptional programs in malignant brain tumors.1,2 Nevertheless, many specimens collected by individual laboratories and stored for years or decades cannot be subjected to conventional RNA sequencing because of their extensive RNA degradation. In a recent study, Chu and colleagues report a new method known as “Chromatin Run-On and Sequencing” (ChRO-seq) that can be used to identify gene expression and transcriptional regulation in long-term archival specimens independent of their RNA integrity.3 This method takes advantage of the ability of RNA polymerase II (RNApol) to remain associated with chromatin recovered from frozen tissue, which facilitates mapping the chromosomal locations where RNA was being transcribed at the moment of tissue collection, even if mature mRNAs and long non-coding RNAs have become completely degraded. The investigators confirmed that ChRO-Seq technology produces maps of gene expression comparable to conventional RNAseq data and can reveal the transcriptional profile of gliomas that have been stored in conventional laboratory freezers for up to 30 years. When applied to a collection of 24 archival glioblastoma specimens, ChRO-Seq technology was able to determine the molecular subgroup of each tumor as well as the differences in gene expression between tumor and normal brain tissue. Furthermore, because RNApol exhibits a characteristic pattern of association with transcriptional regulatory elements such as gene promoters and enhancers, the investigators were able to map those “active” (i.e., occupied by RNApol) regulatory regions of the genome controlling gene expression in each individual tumor. This mapping revealed that glioblastomas exhibit three major programs of transcriptional regulation, matching transcriptional programs observed in stem cells, differentiated support cells (e.g., astrocytes and fibroblasts), and infiltrated immune cells. Surprisingly, most of the active enhancers in the tumors resembled expression patterns from normal brain, with only about 12% of enhancers being unique to glioblastoma and largely resembling expression patterns from developing nervous system. Of the different transcriptional programs, only the immune-related transcription factors and their associated binding sites were predictive of poor prognosis, matching prior observations in the field.4 Overall, ChRO-Seq extends the power of next-generation sequencing to recover transcriptomic information from many cell and tissue specimens that were previously unsuitable for these studies.
Source: Neuro-Oncology
Language: English
Subfield: Molecular Biology
Domain: Life Sciences
Open Access status: bronze
OpenAlex ID: W4233793769