DISMISS: detection of stranded methylation in MeDIP-Seq data

Practical application of a Bayesian network approach to poultry epigenetics and stress

Background

Relationships among genetic or epigenetic features can be explored by learning probabilistic networks and unravelling the dependencies among a set of given genetic/epigenetic features. Bayesian networks (BNs) consist of nodes that represent the variables and arcs that represent the probabilistic relationships between the variables. However, practical guidance on how to make choices among the wide array of possibilities in Bayesian network analysis is limited. Our study aimed to apply a BN approach, while clearly laying out our analysis choices as an example for future researchers, in order to provide further insights into the relationships among epigenetic features and a stressful condition in chickens (Gallus gallus).

Results

Chickens raised under control conditions (n = 22) and chickens exposed to a social isolation protocol (n = 24) were used to identify differentially methylated regions (DMRs). A total of 60 DMRs were selected by a threshold, after bioinformatic pre-processing and analysis. The treatment was included as a binary variable (control = 0; stress = 1). Thereafter, a BN approach was applied: initially, a pre-filtering test was used for identifying pairs of features that must not be included in the process of learning the structure of the network; then, the average probability values for each arc of being part of the network were calculated; and finally, the arcs that were part of the consensus network were selected. The structure of the BN consisted of 47 out of 61 features (60 DMRs and the stressful condition), displaying 43 functional relationships. The stress condition was connected to two DMRs, one of them playing a role in tight and adhesive intracellular junctions in organs such as ovary, intestine, and brain.

Conclusions

We clearly explain our steps in making each analysis choice, from discrete BN models to final generation of a consensus network from multiple model averaging searches. The epigenetic BN unravelled functional relationships among the DMRs, as well as epigenetic features in close association with the stressful condition the chickens were exposed to. The DMRs interacting with the stress condition could be further explored in future studies as possible biomarkers of stress in poultry species.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12859-022-04800-0.

Genome-Wide DNA Methylation and Its Effect on Gene Expression During Subclinical Mastitis in Water Buffalo

Subclinical mastitis (SCM) in buffalo is one of the most challenging paradoxes for the dairy sector with very significant milk production losses and poses an imminent danger to milch animal’s milk-producing ability. We present here the genome-wide methylation specific to SCM in water buffalo and its consequential effect on the gene expression landscape for the first time. Whole-genome DNA methylation profiles from peripheral blood lymphocytes and gene expression profiles from milk somatic cells of healthy and SCM cases were catalogued from the MeDIP-Seq and RNA-Seq data. The average methylation in healthy buffaloes was found to be higher than that in the SCM-infected buffaloes. DNA methylation was abundant in the intergenic region followed by the intronic region in both healthy control and SCM groups. A total of 3,950 differentially methylated regions (DMRs) were identified and annotated to 370 differentially methylated genes (DMGs), most of which were enriched in the promoter region. Several important pathways were activated due to hypomethylation and belonged to the Staphylococcus aureus infection, Th17 cell differentiation, and antigen processing and presentation pathways along with others of defense responses. DNA methylome was compared with transcriptome to understand the regulatory role of DNA methylation on gene expression specific to SCM in buffaloes. A total of 4,778 significant differentially expressed genes (DEGs) were extracted in response to SCM, out of which 67 DMGs were also found to be differentially expressed, suggesting that during SCM, DNA methylation could be one of the epigenetic regulatory mechanisms of gene expression. Genes like CSF2RB, LOC102408349, C3 and PZP like, and CPAMD8 were found to be downregulated in our study, which are known to be involved in the immune response to SCM. Association of DNA methylation with transposable elements, miRNAs, and lncRNAs was also studied. The present study reports a buffalo SCM web resource (BSCM2TDb) available at http://webtom.cabgrid.res.in/BSCM2TDb that catalogues all the mastitis-related information of the analyses results of this study in a single place. This will be of immense use to buffalo researchers to understand the host–pathogen interaction involving SCM, which is required in endeavors of mastitis control and management.

Sperm Methylome Profiling Can Discern Fertility Levels in the Porcine Biomedical Model

A combined Genotyping By Sequencing (GBS) and methylated DNA immunoprecipitation (MeDIP) protocol was used to identify—in parallel—genetic variation (Genomic-Wide Association Studies (GWAS) and epigenetic differences of Differentially Methylated Regions (DMR) in the genome of spermatozoa from the porcine animal model. Breeding boars with good semen quality (n = 11) and specific and well-documented differences in fertility (farrowing rate, FR) and prolificacy (litter size, LS) (n = 7) in artificial insemination programs, using combined FR and LS, were categorized as High Fertile (HF, n = 4) or Low Fertile (LF, n = 3), and boars with Unknown Fertility (UF, n = 4) were tested for eventual epigenetical similarity with those fertility-proven. We identified 165,944 Single Nucleotide Polymorphisms (SNPs) that explained 14–15% of variance among selection lines. Between HF and LF individuals (n = 7, 4 HF and 3 LF), we identified 169 SNPs with p ≤ 0.00015, which explained 58% of the variance. For the epigenetic analyses, we considered fertility and period of ejaculate collection (late-summer and mid-autumn). Approximately three times more DMRs were observed in HF than in LF boars across these periods. Interestingly, UF boars were clearly clustered with one of the other HF or LF groups. The highest differences in DMRs between HF and LF experimental groups across the pig genome were located in the chr 3, 9, 13, and 16, with most DMRs being hypermethylated in LF boars. In both HF and LF boars, DMRs were mostly hypermethylated in late-summer compared to mid-autumn. Three overlaps were detected between SNPs (p ≤ 0.0005, n = 1318) and CpG sites within DMRs. In conclusion, fertility levels in breeding males including FR and LS can be discerned using methylome analyses. The findings in this biomedical animal model ought to be applied besides sire selection for andrological diagnosis of idiopathic sub/infertility.

Putative Epigenetic Biomarkers of Stress in Red Blood Cells of Chickens Reared Across Different Biomes

Production animals are constantly subjected to early adverse environmental conditions that influence the adult phenotype and produce epigenetic effects. CpG dinucleotide methylation in red blood cells (RBC) could be a useful epigenetic biomarker to identify animals subjected to chronic stress in the production environment. Here we compared a reduced fraction of the RBC methylome of chickens exposed to social isolation to non-exposed. These experiments were performed in two different locations: Brazil and Sweden. The aim was to identify stress-associated DNA methylation profiles in RBC across these populations, in spite of the variable conditions to which birds are exposed in each facility and their different lineages. Birds were increasingly exposed to a social isolation treatment, combined with food and water deprivation, at random periods of the day from weeks 1-4 after hatching. We then collected the RBC DNA from individuals and compared a reduced fraction of their methylome between the experimental groups using two bioinformatic approaches to identify differentially methylated regions (DMRs): one using fixed-size windows and another that preselected differential peaks with MACS2. Three levels of significance were used (P ≤ 0.05, P ≤ 0.005, and P ≤ 0.0005) to identify DMRs between experimental groups, which were then used for different analyses. With both of the approaches more DMRs reached the defined significance thresholds in BR individuals compared to SW. However, more DMRs had higher fold change values in SW compared to BR individuals. Interestingly, ChrZ was enriched above expectancy for the presence of DMRs. Additionally, when analyzing the locations of these DMRs in relation to the transcription starting site (TSS), we found three peaks with high DMR presence: 10 kb upstream, the TSS itself, and 20-40 kb downstream. Interestingly, these peaks had DMRs with a high presence (>50%) of specific transcription factor binding sites. Three overlapping DMRs were found between the BR and SW population using the most relaxed p-value (P ≤ 0.05). With the most stringent p-value (P ≤ 0.0005), we found 7 and 4 DMRs between treatments in the BR and SW populations, respectively. This study is the first approximation to identify epigenetic biomarkers of long-term exposure to stress in different lineages of production animals.

Detecting rare asymmetrically methylated cytosines and decoding methylation patterns in the honeybee genome

Context-dependent gene expression in eukaryotes is controlled by several mechanisms including cytosine methylation that primarily occurs in the CG dinucleotides (CpGs). However, less frequent non-CpG asymmetric methylation has been found in various cell types, such as mammalian neurons, and recent results suggest that these sites can repress transcription independently of CpG contexts. In addition, an emerging view is that CpG hemimethylation may arise not only from deregulation of cellular processes but also be a standard feature of the methylome. Here, we have applied a novel approach to examine whether asymmetric CpG methylation is present in a sparsely methylated genome of the honeybee, a social insect with a high level of epigenetically driven phenotypic plasticity. By combining strand-specific ultra-deep amplicon sequencing of illustrator genes with whole-genome methylomics and bioinformatics, we show that rare asymmetrically methylated CpGs can be unambiguously detected in the honeybee genome. Additionally, we confirm differential methylation between two phenotypically and reproductively distinct castes, queens and workers, and offer new insight into the heterogeneity of brain methylation patterns. In particular, we challenge the assumption that symmetrical methylation levels reflect symmetry in the underlying methylation patterns and conclude that hemimethylation may occur more frequently than indicated by methylation levels. Finally, we question the validity of a prior study in which most of cytosine methylation in this species was reported to be asymmetric.

Detecting rare asymmetrically methylated cytosines and decoding methylation patterns in the honeybee genome

Context-dependent gene expression in eukaryotes is controlled by several mechanisms including cytosine methylation that primarily occurs in the CG dinucleotides (CpGs). However, less frequent non-CpG asymmetric methylation has been found in various cell types, such as mammalian neurons, and recent results suggest that these sites can repress transcription independently of CpG contexts. In addition, an emerging view is that CpG hemimethylation may arise not only from deregulation of cellular processes but also be a standard feature of the methylome. Here, we have applied a novel approach to examine whether asymmetric CpG methylation is present in a sparsely methylated genome of the honeybee, a social insect with a high level of epigenetically driven phenotypic plasticity. By combining strand-specific ultra-deep amplicon sequencing of illustrator genes with whole-genome methylomics and bioinformatics, we show that rare asymmetrically methylated CpGs can be unambiguously detected in the honeybee genome. Additionally, we confirm differential methylation between two phenotypically and reproductively distinct castes, queens and workers, and offer new insight into the heterogeneity of brain methylation patterns. In particular, we challenge the assumption that symmetrical methylation levels reflect symmetry in the underlying methylation patterns and conclude that hemimethylation may occur more frequently than indicated by methylation levels. Finally, we question the validity of a prior study in which most of cytosine methylation in this species was reported to be asymmetric.