When you’re looking to access the latest and greatest publications, your VPN is the proxy you need. But now, there’s a new proxy in town to help with your quest for neurodevelopmental epigenetic information: Placenta and the cell free fetal DNA (cffDNA) that originates from it.
A dynamic duo of papers from the lab of Janine LaSalle (UC Davis Genome Center) and their talented team of collaborators profiled two different mammalian models of risk factors for neurodevelopmental disorders (NDDs), such as autism spectrum disorders (ASDs). In order to assay genome-wide DNA methylation profiles, whole-genome bisulfite sequencing (WGBS) data was prepared with the Accel-NGS Methyl-Seq DNA Library Kit and analyzed with CpG_Me and DMRichR. This approach let them uncover methylation profiles in placenta and cffDNA that relate to brain and behavior.
Placenta and Brain Share Cellular Signaling Differences
PCBs (polychlorinated biphenyls) are persistent organic pollutants, and as the name suggests, they persist in our environment and cause developmental neurotoxicity, which makes them an environmental risk factor for NDDs. In a collaboration with neurotoxicologists from the lab of Pamela Lein, the team examined the genome-wide DNA methylation profiles of embryonic placenta and brain from a mouse model of prenatal PCB exposure. This human-relevant exposure model was based on PCB levels found in the serum of mothers who had children with ASDs. This approach let them uncover common connections between the differentially methylated regions (DMRs) from matched placenta and brain:
- An enrichment for altered methylation in regions of bivalent chromatin and genes with functions related to neurodevelopment and cellular signaling
- Specifically, the overlapping genes have functional enrichments for Slit/Robo signaling and Wnt signaling
- The pathways and top gene mappings provide direct evidence that the previously known neuronal PCB-signaling cascades, which involves the Ryanodine receptor and Wnt signaling, leave a methylation profile in brain and placenta
- There’s also an enrichment related to mouse models of neurodevelopmental disorders (NDDs) as well as placenta and brain from human NDDs
Driving this common connection home is the observation from a permutation analysis that the genomic coordinates of the placenta and brain DMRs significantly overlap. Together, these findings uncover a methylationship between placenta and brain from the same animal.
Multi-omic Correlations Between cffDNA and Brain
Now that we’ve seen the signal between placenta and brain, the search for a proximal placental proxy went distal. This time, the team and their collaborators from the California National Primate Research Center and MIND Institute turned their analyses towards maternal obesity, since it associates with an increased risk for children with ASD. They examined offspring from a natural rhesus macaque maternal obesity model, with both obesity intervention (pharmacological and dietary) and lean matched controls. This involved profiling cffDNA methylomes across all trimesters of pregnancy and matched infant brain region methylomes (hippocampus, prefrontal cortex, and hypothalamus) at 6 months. Then, they tested for correlations with immunological and metabolomic assays of maternal blood at matched pregnancy timepoints, two tests of NDD-relevant infant behavior, and infant brain metabolomic and lipidomic assays. Here’s what they found:
- cffDNA and brain DMRs overlap with each other and human NDD DMRs, while also being enriched for within genes with neurodevelopmental functions and the motifs of developmental transcription factors that are methylation-sensitive
- There’s a large block of maternal obesity associated differential methylation that overlaps mir-663 and correlates with maternal immune and metabolic markers, infant behavior, and levels of a lipid in infant prefrontal cortex
- Weighted gene co-methylation analysis (WGCNA) uncovered a module of DMRs with a DUX4 hub region in infant hippocampus that correlated with maternal obesity and:
- Maternal metabolites, cytokines, and DUX4 cffDNA methylation levels across pregnancy
- Infant behavior and hippocampal metabolites and lipids, with genes within this module having known direct effects on the correlated phenotypes (i.e ERBB2 and recognition memory, ASNS and asparagine, and FFAR4 and fatty acids)
Overall, these findings demonstrate that cffDNA methylation profiles reflect brain methylation and relevant behavioral and molecular traits. They also show that obesity interventions modify this profile and warrant further research.
A Step Towards Non-invasive Diagnostics
Taken together, these finding demonstrate the potential of placenta, and cffDNA by proxy, as sample sources that can be non-invasively assayed to provide profiles with some key commonalities to brain DNA methylation. While we can’t yet say whether these relationships will be found in humans, the findings suggest it’s time to start looking.