PATH-46. DIAGNOSTIC IMPACT OF THE CNS TUMOR METHYLATION PROFILING IN A NEUROPATHOLOGY CONSULT PRACTICE

DNA methylation profiling coupled with the application of CNS tumor methylation classifier has contributed to precise and accurate diagnostics for a range of tumor types involving the central nervous system. The impact and characteristics of methylation profiling on tumor diagnosis has not been fully assessed in the setting of neuropathology consultation practice. A consecutive series of 1,258 surgical neuropathology samples obtained primarily in a consultation practice were profiled over 2-year period and analyzed using the DKFZ/Heidelberg CNS tumor methylation classifier. Among the 1,045 cases received from outside institutions for consultation, the classifier was able to refine a histologically diagnosed entity (e.g. medulloblastoma) in 13.3% (n = 139) cases. A substantially new diagnosis was able to be rendered in an additional 17.9% (n = 187) cases, many of which could be confirmed using orthogonal methods. A “suggestive” (0.30-0.84) classifier score was found in 23% (242) cases and we found that complementary methods (UMAP, t-SNE and nearest-neighbors) were able to resolve this uncertainty in 118 cases. We found tumor purity significantly associated with varied classifier score (p = 1.53e-11). Computational tumor purity adjustment by deconvolution on a subset of gliomas provided a proof-of-concept to resolve diagnostics in the setting of low tumor purity. Overall, this work directly assesses the benefit of methylation classification in a set of diagnostically challenging CNS tumors, addresses tumor purity diminished methylation signal and provides complementary approaches to address diagnostics in cases of low-confidence classifier scores.

DiSNEP: a Disease-Specific gene Network Enhancement to improve Prioritizing candidate disease genes

Biological network-based strategies are useful in prioritizing genes associated with diseases. Several comprehensive human gene networks such as STRING, GIANT and HumanNet were developed and used in network-assisted algorithms to identify disease-associated genes. However, none of these networks are disease-specific and may not accurately reflect gene interactions for a specific disease. Aiming to improve disease gene prioritization using networks, we propose a Disease-Specific Network Enhancement Prioritization (DiSNEP) framework. DiSNEP first enhances a comprehensive gene network specifically for a disease through a diffusion process on a gene–gene similarity matrix derived from disease omics data. The enhanced disease-specific gene network thus better reflects true gene interactions for the disease and may improve prioritizing disease-associated genes subsequently. In simulations, DiSNEP that uses an enhanced disease-specific network prioritizes more true signal genes than comparison methods using a general gene network or without prioritization. Applications to prioritize cancer-associated gene expression and DNA methylation signal genes for five cancer types from The Cancer Genome Atlas (TCGA) project suggest that more prioritized candidate genes by DiSNEP are cancer-related according to the DisGeNET database than those prioritized by the comparison methods, consistently across all five cancer types considered, and for both gene expression and DNA methylation signal genes.

Discrimination of DNA Methylation Signal from Background Variation for Clinical Diagnostics

Advances in the study of human DNA methylation variation offer a new avenue for the translation of epigenetic research results to clinical applications. Although current approaches to methylome analysis have been helpful in revealing an epigenetic influence in major human diseases, this type of analysis has proven inadequate for the translation of these advances to clinical diagnostics. As in any clinical test, the use of a methylation signal for diagnostic purposes requires the estimation of an optimal cutoff value for the signal, which is necessary to discriminate a signal induced by a disease state from natural background variation. To address this issue, we propose the application of a fundamental signal detection theory and machine learning approaches. Simulation studies and tests of two available methylome datasets from autism and leukemia patients demonstrate the feasibility of this approach in clinical diagnostics, providing high discriminatory power for the methylation signal induced by disease, as well as high classification performance. Specifically, the analysis of whole biomarker genomic regions could suffice for a diagnostic, markedly decreasing its cost.

Post-transcriptional gene silencing triggers dispensable DNA methylation in gene body in Arabidopsis

Spontaneous post-transcriptional silencing of sense transgenes (S-PTGS) is established in each generation and is accompanied by DNA methylation, but the pathway of PTGS-dependent DNA methylation is unknown and so is its role. Here we show that CHH and CHG methylation coincides spatially and temporally with RDR6-dependent products derived from the central and 3′ regions of the coding sequence, and requires the components of the RNA-directed DNA methylation (RdDM) pathway NRPE1, DRD1 and DRM2, but not CLSY1, NRPD1, RDR2 or DCL3, suggesting that RDR6-dependent products, namely long dsRNAs and/or siRNAs, trigger PTGS-dependent DNA methylation. Nevertheless, none of these RdDM components are required to establish S-PTGS or produce a systemic silencing signal. Moreover, preventing de novo DNA methylation in non-silenced transgenic tissues grafted onto homologous silenced tissues does not inhibit the triggering of PTGS. Overall, these data indicate that gene body DNA methylation is a consequence, not a cause, of PTGS, and rule out the hypothesis that a PTGS-associated DNA methylation signal is transmitted independent of a PTGS signal.

DNA methylation as a marker for prenatal smoke exposure in adults

Prenatal cigarette smoke is an environmental stressor that has a profound effect on DNA methylation in the exposed offspring. We have previously shown that some of these effects persist throughout childhood and into adolescence. Of interest is whether these signals persist into adulthood.We conducted an analysis to investigate associations between reported maternal smoking in pregnancy and DNA methylation in peripheral blood of women in the Avon Longitudinal Study of Parents and Children (ALSPAC) (n=754; mean age 30 years). We observed associations at 15 CpG sites in 11 gene regions, MYO1G, FRMD4A, CYP1A1, CNTNAP2, ARL4C, AHRR, TIFAB, MDM4, AX748264, DRD1, FTO (FDR < 5%). All but two of these CpG sites have previously been identified in relation to prenatal smoke exposure in the offspring at birth and the majority showed persistent hypermethylation among the offspring of smokers.We confirmed that most of these associations were not driven by own smoking and that they were still present 18 years later (N = 656; mean age 48 years). In addition, we replicated findings of a persistent methylation signal related to prenatal smoke exposure in peripheral blood among men in the ALSPAC cohort (N = 230; mean age 53 years). For both participant groups, there was a strong signal of association above that expected by chance at CpG sites previously associated with prenatal smoke exposure in newborns (Wilcoxon rank sum p-value < 2.2 × 10−4). Furthermore, we found that a prenatal smoking score, derived by combining methylation values at these CpG sites, could predict whether the mothers of the ALSPAC women smoked during pregnancy with an AUC 0.69 (95% 0.67, 0.73).

Mbd2 Contributes to DNA Methylation-Directed Repression of the Xist Gene

ABSTRACT Transcription of the Xist gene triggers X chromosome inactivation in cis and is therefore silenced on the X chromosome that remains active. DNA methylation contributes to this silencing, but the mechanism is unknown. As methylated DNA binding proteins (MBPs) are potential mediators of gene silencing by DNA methylation, we asked whether MBP-deficient cell lines could maintain Xist repression. The absence of Mbd2 caused significant low-level reactivation of Xist, but silencing was restored by exogenous Mbd2. In contrast, deficiencies of Mbd1, MeCP2, and Kaiso had no detectable effect, indicating that MBPs are not functionally redundant at this locus. Xist repression in Mbd2-null cells was hypersensitive to the histone deacetylase inhibitor trichostatin A and to depletion of the DNA methyltransferase Dnmt1. These synergies implicate Mbd2 as a mediator of the DNA methylation signal at this locus. The presence of redundant mechanisms to enforce repression at Xist and other loci is compatible with the hypothesis that “stacking” of imperfect repressive tendencies may be an evolutionary strategy to ensure leakproof gene silencing.