Metacore

Enrichment analyses for gene ontologies were performed using MetaCore^TM (Clarivate Analytics) on the SNPs that compose the fasting insulin rPRS. Furthermore, gene-based enrichment analyses were performed in FUMA¹ (MacArthur et al., 2017 (link); Watanabe et al., 2017 (link); Aguet et al., 2019 (link)) after mapping the SNPs composing the fasting insulin rPRS to genes with the biomaRt package in R (Durinck et al., 2005 (link), Durinck et al., 2009 (link)). We also used GeneMANIA (Warde-Farley et al., 2010 (link)) to determine if the genes were part of a network. Specifically, the gene list derived from the fasting insulin rPRS is entered in GeneMANIA. GeneMANIA then extracts linked mRNA expression data from the Gene Expression Omnibus (GEO) and connects co-expressed data to form functional association networks. The node sizes represent gene scores indicating the number of paths that start at a given gene node and end up in one of the query genes.

Statistical significance of nonparametric data was analyzed by Mann–Whitney U-Test. Data represent mean ±1 SD. Calculations were performed using Minitab® v16 (Minitab Inc. USA). Network and Pathway data analysis was performed using Key Pathway Advisor and Metacore^TM (Clarivate Analytics, UK).

The 2417 and 3619 unique genes mapped from the maternal EPDS-vCpGs bearing female fetus and EPDS-vCpGs from female fetus-facing placenta tissues were respectively imported into MetaCore^TM (v21.1.70400; Clarivate Analytics) for pathway enrichment and transcription factor analyses. 25,641 and 25, 281 unique genes mapped from the vCpGs of both maternal blood and placenta sources were used respectively as the reference list. P-values for transcription factor analyses were determined using hypergeometric intersection.
Each input gene mapped from maternal EPDS-vCpGs bearing female fetuses was also queried against previously associated phenotypes using FUMA (Watanabe et al., 2017 (link)). FUMA was also used to examine the up-regulated expression of these genes mapped from EPDS-vCpGs, relative to genes mapped from vCpGs, in specific tissues based on GTex v8 RNA-seq data (GTEx Consortium and Ardlie, 2015 (link)). Bonferroni corrected p-values were provided for the up-regulated differentially expressed gene sets from FUMA.

Biological interpretation of genes that comprised our genetic score was performed using enrichment analysis using MetaCore^TM (Clarivate Analytics). The enrichment identifies statistically significant pathway maps and gene ontology processes associated with this list of genes after false discovery rate (FDR) correction, to summarize the most enriched and pertinent biology associated with the set of genes under investigation (Huang et al., 2009 (link)). We also performed enrichment analysis to identify genes differentially expressed at different developmental phases, via functional mapping of genetic and expression using the FUMA tool (Watanabe et al., 2017 (link)).

HT29-MTX cells were cultured in 24-well Transwell® (Corning, Corning, NY, USA), and incubated with treatments of MRx0518_LV, MRx0518_HK and MRx0518_SN at a MOI of 100:1 (or equivalent) for 3 h at 37 °C under anaerobic conditions. Cells were washed and lysed, and RNA was isolated from lysate using an RNeasy Mini Kit (Qiagen, Hilden, Germany). RNA was converted to cDNA using a GeneChip™ High Throughput WT PLUS Kit, which was then hybridized to a GeneChip™ Human Transcriptome Array 2.0. Microarray chips were washed and stained using a GeneChip™ Fluidics Station 450 instrument and the GeneChip™ Expression Wash, Stain and Scan kit, and then scanned using a GeneChip™ Scanner 3000 instrument (Thermo Fischer Scientific). Data analysis was carried out using Transcriptome Analysis Console 4.0 software (Thermo Fischer Scientific). Data were normalized using the Robust Multiarray Average algorithm, and fold changes were calculated using the normalized log₂-transformed values of treated cells relative to respective controls. Data were filtered using cut-offs of p < 0.05, fold change of <−1.5 and ≥1.5, and the presence of a gene symbol and coding variants. Pathway analysis was carried out using MetaCore™ (Clarivate Analytics, Philadelphia, PA, USA).