Recruitment and sample collection
The donors (N = 10) and recipients from TTTS pregnancies (N = 10) and matched controls (N = 10) were recruited from the Shengjing Hospital of China Medical University from June 2018 to June 2019. TTTS diagnosis and staging were determined with ultrasound criteria according to Quintero [11, 12]. TTTS cases were selected from cases of inevitable abortion before 24 weeks. In the control group, all were singletons who were spontaneously aborted before 24 weeks. There was no significant difference in the time to termination of pregnancy between TTTS and matched controls (P >0.05). The gravidae approved the use of hippocampus tissue for scientific purposes by signing informed consent forms. The study was approved by the Human Ethics Committee of the Shengjing Hospital of China Medical University (2018PS360K).
Tissue processing and DNA extraction
After the termination of pregnancy, the brain tissues were immediately removed, and the hippocampus tissues were dissected on ice and washed using ice-cold phosphate-buffered saline (PBS). Then, the hippocampus tissues were snap-frozen in liquid nitrogen for 10 min and stored at −80 ℃ until processing. Genomic DNA was extracted from hippocampal tissues using the AllPrep DNA/RNA MiniKit (80204, Qiagen, Hilden, Germany) according to the manufacturer’s protocol.
Illumina HumanMethylation850K bead chip
Genomic DNA from each sample was prepared for sodium bisulfite conversion using the EZ DNA methylation Gold Kit (Zymo Research, Irvine, CA, USA). We further assessed genome-wide DNAm based on Illumina Infinium HumanMethylation850K BeadChip (Illumina Inc, San Diego, CA, USA) microarray. The array data were analyzed using ChAMP package in R for deriving the methylation level . We have denoted the methylation status of all probes as the β value, which is the ratio of the methylated probe intensity to the overall probe intensity.
Differential methylated CpG sites analysis
The Limma package was used to identify differential methylated CpG sites . We used |Δβ| ≥0.20 and adjusted P value ≤0.05 as differentially methylated site cut-off points. The CpG sites with Δβ ≥0.20 were considered hypermethylated, and the CpG sites with Δβ ≤−0.20 were considered hypomethylated.
Function enrichment analysis
WEB-based Gene Set Analysis Toolkit (WebGestalt) is a functional enrichment analysis web tool, which has on average 26,000 unique users from 144 countries and territories per year according to Google Analytics. A total of three well-established and complementary methods for enrichment analysis in the WebGestalt: Over-Representation Analysis (ORA), Gene Set Enrichment Analysis (GSEA), and Network Topology-based Analysis (NTA) . We performed the gene ontology (GO) analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis for the identified differential methylation probes (DMP) based on the WebGestalt. P value < 0.05 was used as statistical significance.
Epigenetic Dissection of Intra-Sample Heterogeneity
The EpiDISH (Epigenetic Dissection of Intra-Sample Heterogeneity) package provides tools to infer the fractions of a priori known cell subtypes present in samples. The EpiDISH package provides a function called CellDMC, which allows the identification of differentially methylated cell types in epigenome-wide association studies (EWAS). The EpiDISH package contains four references: two whole blood subtype references, one generic epithelial reference of total immune cells, and one breast tissue reference. [16-18].
Co-methylation network analysis
To analyze the intercorrelation among the identified DMPs, we performed the co-methylation network analysis using the WGCNA package . WGCNA is a network analysis tool that uses hierarchical clustering of correlated methylation states between DMPs to construct weighted co-methylation modules. WGCNA can transform the adjacency matrix into a topological overlap matrix. DMPs can be divided into different co-methylation modules based on the topological overlap matrix dissimilarity measurements. Here, we set the soft-threshold power as 16 to identify key co-methylation modules. The modules with the highest correlation with immune infiltration were selected for further analysis.
Protein-protein interaction network analysis
We constructed the protein-protein interaction (PPI) network using the Search Tool for the Retrieval of Interacting Genes (STRING) database. We upload the differential methylation genes to the STRING database and defined the cut-off as the interaction score of 0.4 (median confidence) . The PPI network was visualized using Cytoscape software . Molecular Complex Detection (MCODE) analysis was used to screen the key gene modules within the PPI network. A k score ≥3 was selected as the cut-off to define the key gene module. We further explored the expression level of the hub differential methylation genes among the three groups.