DNA methylation aging and transcriptomic studies in horses

Cytosine methylation patterns have not yet been thoroughly studied in horses. Here, we profile n = 333 samples from 42 horse tissue types at loci that are highly conserved between mammalian species using a custom array (HorvathMammalMethylChip40). Using the blood and liver tissues from horses, we develop five epigenetic aging clocks: a multi-tissue clock, a blood clock, a liver clock and two dual-species clocks that apply to both horses and humans. In addition, using blood methylation data from three additional equid species (plains zebra, Grevy’s zebras and Somali asses), we develop another clock that applies across all equid species. Castration does not significantly impact the epigenetic aging rate of blood or liver samples from horses. Methylation and RNA data from the same tissues define the relationship between methylation and RNA expression across horse tissues. We expect that the multi-tissue atlas will become a valuable resource.

DNA Methylation Data-Based Classification and Identification of Prognostic Signature of Children With Wilms Tumor

Background: As an epigenetic alteration, DNA methylation plays an important role in early Wilms tumorigenesis and is possibly used as marker to improve the diagnosis and classification of tumor heterogeneity.Methods: Methylation data, RNA-sequencing (RNA-seq) data, and corresponding clinical information were downloaded from the Therapeutically Applicable Research to Generate Effective Treatments (TARGET) database. The prognostic values of DNA methylation subtypes in Wilms tumor were identified.Results: Four prognostic subtypes of Wilms tumor patients were identified by consensus cluster analysis performed on 312 independent prognostic CpG sites. Cluster one showed the best prognosis, whereas Cluster two represented the worst prognosis. Unique CpG sites identified in Cluster one that were not identified in other subtypes were assessed to construct a prognostic signature. The prognostic methylation risk score was closely related to prognosis, and the area under the curve (AUC) was 0.802. Furthermore, the risk score based on prognostic signature was identified as an independent prognostic factor for Wilms tumor in univariate and multivariate Cox regression analyses. Finally, the abundance of B cell infiltration was higher in the low-risk group than in the high-risk group, based on the methylation data.Conclusion: Collectively, we divided Wilms tumor cases into four prognostic subtypes, which could efficiently identify high-risk Wilms tumor patients. Prognostic methylation risk scores that were significantly associated with the adverse clinical outcomes were established, and this prognostic signature was able to predict the prognosis of Wilms tumor in children, which may be useful in guiding clinicians in therapeutic decision-making. Further independent studies are needed to validate and advance this hypothesis.

Tissue- and ethnicity-independent hypervariable DNA methylation states show evidence of establishment in the early human embryo

We analysed DNA methylation data from 30 datasets comprising 3,474 individuals, 19 tissues and 8 ethnicities at CpGs covered by the Illumina450K array. We identified 4,143 hypervariable CpGs ('hvCpGs') with methylation in the top 5% most variable sites across multiple tissues and ethnicities. hvCpG methylation was influenced but not determined by genetic variation, and was not linked to probe reliability, epigenetic drift, age, sex or cell heterogeneity effects. hvCpG methylation tended to covary across tissues derived from different germ-layers and hvCpGs were enriched for associations with periconceptional environment, proximity to ERV1 and ERVK retrovirus elements and parent-of-origin-specific methylation. They also showed distinctive methylation signatures in monozygotic twins. Together, these properties position hvCpGs as strong candidates for studying how stochastic and/or environmentally influenced DNA methylation states which are established in the early embryo and maintained stably thereafter can influence life-long health and disease.

Methylation pattern of nc886 in non-human mammals

Background: In humans, the nc886 locus is a polymorphically imprinted metastable epiallele. Periconceptional conditions have an effect on the methylation status of nc886, and further, this methylation status is associated with health outcomes in later life, in line with the Developmental Origins of Health and Disease (DOHaD) hypothesis. Animal models would offer opportunities to study the associations between periconceptional conditions, nc886 methylation status and metabolic phenotypes further. Thus, we set out to investigate the methylation pattern of the nc886 locus in non-human mammals. Data: We obtained DNA methylation data from the data repository GEO for mammals, whose nc886 gene included all three major parts of nc886 and had sequency similarity of over 80% with the human nc886. Our final sample set consisted of DNA methylation data from humans, chimpanzees, bonobos, gorillas, orangutangs, baboons, macaques, vervets, marmosets and guinea pigs. Results: In human data sets the methylation pattern of nc886 locus followed the expected bimodal distribution, indicative of polymorphic imprinting. In great apes, we identified a unimodal DNA methylation pattern with 50% methylation level in all individuals and in all subspecies. In Old World monkeys, the between individual variation was greater and methylation on average was close to 60%. In guinea pigs the region around the nc886 homologue was non-methylated. Results obtained from the sequence comparison of the CTCF binding sites flanking the nc886 gene support the results on the DNA methylation data. Conclusions: Our results indicate that unlike in humans, nc886 is not a polymorphically imprinted metastable epiallele in non-human primates or in guinea pigs, thus implying that animal models are not applicable for nc886 research. The obtained data suggests that the nc886 region may be classically imprinted in great apes, and potentially also in Old World monkeys, but not in guinea pigs.

Links of Sleep Duration with Biomarkers of Accelerated Aging: the Baltimore Longitudinal Study of Aging

Sleep disorders and sleep deprivation have been linked to markers of biological aging, including methylation change and increases in white blood cell and neutrophil counts. However, little is known regarding the association of sleep duration with biological markers of aging. We investigated links of self-reported sleep duration with biological aging markers in 615 participants in the Baltimore Longitudinal Study of Aging (BLSA) aged ≥50 years (mean = 71.0 ± 11.2, 49.6% women, 68.8% white) with data on self-reported sleep duration in hours (i.e., ≤6 (n=131), >6 to 7 (n=234), >7 (n=250)), demographics, and genetic and methylation data (mDNA). Our aging biomarker outcomes were four epigenetic clocks (Horvath, Hannum, PhenoAge, and GrimAge), mDNA-estimated PAI1, and estimated granulocyte count. After adjustment for age, sex, and race, compared to those sleeping ≤6 hours, those reporting >7 hours of sleep had faster biological aging according to Hannum age-acceleration, PhenoAge, GrimAge, mDNA-estimated PAI1, and granulocyte count. In addition, sleep duration interacted with age, such that compared to individuals reporting ≤6 hours of sleep, individuals reporting >6 to 7 hours showed lower GrimAge with increasing age, and with sex, such that males with longer sleep duration (>6 to 7 and >7 hours) showed a lower granulocyte count compared to females. Findings suggest that both short and long sleep duration are associated with and may contribute to accelerated aging. Prospective studies in larger samples are needed to examine whether changes in sleep duration precede changes in aging biomarkers.

Epigenome-wide association studies: current knowledge, strategies and recommendations

The aetiology and pathophysiology of complex diseases are driven by the interaction between genetic and environmental factors. The variability in risk and outcomes in these diseases are incompletely explained by genetics or environmental risk factors individually. Therefore, researchers are now exploring the epigenome, a biological interface at which genetics and the environment can interact. There is a growing body of evidence supporting the role of epigenetic mechanisms in complex disease pathophysiology. Epigenome-wide association studies (EWASes) investigate the association between a phenotype and epigenetic variants, most commonly DNA methylation. The decreasing cost of measuring epigenome-wide methylation and the increasing accessibility of bioinformatic pipelines have contributed to the rise in EWASes published in recent years. Here, we review the current literature on these EWASes and provide further recommendations and strategies for successfully conducting them. We have constrained our review to studies using methylation data as this is the most studied epigenetic mechanism; microarray-based data as whole-genome bisulphite sequencing remains prohibitively expensive for most laboratories; and blood-based studies due to the non-invasiveness of peripheral blood collection and availability of archived DNA, as well as the accessibility of publicly available blood-cell-based methylation data. Further, we address multiple novel areas of EWAS analysis that have not been covered in previous reviews: (1) longitudinal study designs, (2) the chip analysis methylation pipeline (ChAMP), (3) differentially methylated region (DMR) identification paradigms, (4) methylation quantitative trait loci (methQTL) analysis, (5) methylation age analysis and (6) identifying cell-specific differential methylation from mixed cell data using statistical deconvolution.

Epigenome-Wide Association Study of Acute Lymphoblastic Leukemia in Children with Down Syndrome

Background: Down syndrome (DS) is associated with an up to 30-fold increased risk of B-cell acute lymphoblastic leukemia (ALL), and DS-ALL patients have worse overall survival and increased long-term treatment-related health conditions compared with non-DS ALL patients. In a recent genome-wide association study of DS-ALL, established ALL genetic risk loci were associated with DS-ALL, with several single nucleotide polymorphisms (SNPs) conferring a larger effect on ALL risk in the context of DS than in euploidy. We performed an epigenome-wide association study (EWAS) to elucidate whether epigenetic differences at birth are associated with risk of subsequent DS-ALL. Methods: The DS-ALL Discovery Study included 147 DS-ALL cases and 198 DS controls from the International Study of Down Syndrome Acute Leukemia, with newborn dried bloodspots (DBS) obtained from California (n=326) and Washington state (n=19) biobanks. The DS-ALL Replication Study included 24 DS-ALL cases and 24 DS controls with newborn DBS from the Michigan Neonatal Biobank. DNA was isolated from DBS, bisulfite converted, and assayed using Illumina Infinium MethylationEPIC Beadchip genome-wide DNA methylation arrays. Raw data were processed using "minfi" and "noob" packages in R. Reference-based deconvolution of blood cell proportions was performed using the Identifying Optimal DNA methylation Libraries (IDOL) algorithm, using DNA methylation data from cord blood reference samples, to estimate proportions of B cells, T cells (CD4+ and CD8+), monocytes, granulocytes, natural killer cells, and nucleated red blood cells. We compared each cell type proportion between DS-ALL cases and DS controls using linear regression adjusting for sex, plate, and principal components (PCs) to account for genetic ancestry. To identify single CpG probes associated with DS-ALL risk, we performed a multiethnic EWAS of DS-ALL in each study using linear regression adjusting for sex, plate, and PCs related to: 1) cell-type proportions and 2) genetic ancestry. Differentially methylated regions (DMRs) were identified using DMRcate and comb-p methods. In the Discovery Study, genome-wide SNP array data were available for 131 cases and 130 controls, and data from targeted sequencing of somatic mutations in exons 2/3 of GATA1 were available for 184/198 DS controls. Results: Deconvolution of blood cell proportions in the DS-ALL Discovery Study showed significantly higher B cell proportions in newborns with DS who later developed ALL (mean=0.0128, sd=0.0151) compared with DS controls (mean=0.00826, sd=0.0115) (P=6.4x10 -4, coefficient=0.0052). A significantly higher B cell proportion at birth was also found in DS-ALL cases in the independent Replication Study (cases mean=0.048, sd=0.024; controls mean=0.039, sd=0.028; P=0.03, coefficient=0.015). In the Discovery Study, the B cell difference remained significant (P=5.8x10 -3) with a similar effect size (coefficient=0.0045) after removal of GATA1 mutation-positive DS controls (n=30). We also investigated whether DS-ALL risk SNPs at ARID5B, IKZF1, GATA3, and CDKN2A may confound the association, but the increased B cell proportions in DS-ALL remained significant and effect estimates slightly increased in SNP genotype-adjusted models (coefficient range:0.0055-0.0059). In the EWAS of DS-ALL, 9 CpGs reached epigenome-wide significance (P<7.67x10 -8), including 2 CpGs overlapping the promoter of the tumor suppressor gene TRIM13, frequently deleted in B-CLL, although none of these showed evidence of association (P<0.05) in the Replication Study. We identified 125 DMRs associated with DS-ALL in the Discovery Study. For 3 DMRs, overlapping genes HOPX, SMIM24, and PPP1R10, all implicated in normal and leukemic stem cell function, there were multiple significant CpGs in the Replication Study (P<0.05) all with effects in the same direction as the Discovery Study DMRs. Conclusions: Increased B cell proportions in newborns with DS may be a risk factor for development of DS-ALL in childhood. This finding, based on DNA methylation data, requires confirmation using conventional cell count measures, and should be explored as a novel biomarker for ALL risk in the non-DS population. Single CpGs and DMRs associated with DS-ALL risk in our Discovery Study require further investigation, including in additional ALL case-control studies in DS and non-DS populations. Disclosures Ma: Celgene/Bristol Myers Squibb: Consultancy, Research Funding.

Multi-Omics Analysis Identifies the Heterogeneity and Prognosis in Bladder Cancer

Background The multi-omics integrated analysis can help researchers understand the biological behavior of bladder cancer(BCa) in a more systematic and comprehensive manner, and further provide new clues for finding valuable tumor markers and therapeutic targets. Methods In this study, we applied the DNA methylation data to construct a prognosis classifier and stratified the BCa patients into high- and low-risk subtype. The differences of transcriptome, single nucleotide variants and copy number variations between two subgroups were explored for finding the changes of molecular mechanism. Results With 18 pairs DNA methylation samples, ten differentially methylated positions(DMPs) were identified and applied to evaluate the risk score of each sample. Kaplan-Meier survival analysis displayed that BCa patients with high risk had a poor prognosis than the lower(p<0.0001). In transcriptome analysis, many immune related pathways and biological process changed between high- and low-risk patients. The results also displayed that naive B cells, plasma cells, CD8+ T cells and T cell regulatory(Tregs) infiltrated less in high-risk patients and these patients were less sensitive to immunotherapy and chemotherapy. As for single nucleotide variants, we found that TP53, CDKN1A, STAG2 and other genes were more frequently mutated in high-risk BCa patients. Only copy number variation in high-risk patients were displayed for the limitation of TCGA data. Conclusions The high- and low-risk patients identified by DNA methylation data of bladder cancer were significant different in survival. The comprehensive comparison of multi-omics data between subgroups can help clinicians find the heterogeneity of tumor biological behavior and contribute to precision treatment in bladder cancer.

PATH-18. INTEGRATING TUMOR MICROENVIRONMENT WITH GENOMIC ABERRATIONS AND TUMOR CLASS IN ADULT GLIAL/GLIONEURONAL TUMORS BY DECONVOLUTION OF BULK METHYLATION DATA

It is increasingly recognized that the tumor microenvironment (TME) plays a critical role in the biology of cancer. To better understand the role of non-neoplastic immune cellular components in CNS tumors, we applied a deconvolution approach to bulk DNA methylation array data using methylCIBERSORT on 450k/850k methylation data from a set (n= 4057) of high- and low-grade glial and glioneuronal tumors. Using the cell type proportion data as input, we used dimension reduction (UMAP) to visualize sample-wise patterns of that emerge from the cell type proportion estimations. In glioblastomas (n= 2076) we identified distinct tumor clusters based on immune cell proportion and, interestingly, TME-based cluster groups demonstrated an association with specific genetic alterations such as EGFR amplification and/or CDKN2A/B homozygous deletion. Among 1178 IDH-mutant gliomas, clustering of tumors according to immune cell proportions led to 2 major subgroups, which largely aligned with 1p/19q co-deletion status. Among the non-codeleted tumors (IDH-mutant astrocytomas, N=734), clustering of immune cell decomposition revealed clusters which showed distinct proportions of a key genomic aberration in these tumors (CDKN2A/B loss). To investigate the possible role of monocyte proportion-relative gene expression and promoter methylation of the immune checkpoint PD-L1 and PD-L2 genes, we used a data subset (n=594) samples with matched gene expression profiles. We observed significantly high positive correlations (R=0.54 and 0.68, respectively) between monocyte proportion and expression of PD-L1 and PD-L2, in line with prior reports that monocytic cells can express these immune markers. Consistent with this, we found high negative correlations (R= -0.51 and -0.61, respectively) between monocytes and promoter methylation of PD-L1 and PD-L2, respectively. Overall, the findings highlight specific roles of the TME in biology and classification of adult CNS tumors, where specific immune cell admixtures correlate with tumor types and genomic aberrations.