Cell-free DNA methylation profiling in early-onset preeclampsia using methylated DNA sequencing.

INTRODUCTION

Distinct DNA methylation profiles have been identified in maternal plasma cell-free DNA (cfDNA) in early-onset preeclampsia (EOPE). However, findings across studies often lack concordance and the biological mechanisms driving methylation differences are largely unknown. We aimed to identify EOPE-associated differentially methylated regions (DMRs), investigate changes in tissue-specific cfDNA contributions and validate previously reported EOPE-associated regions.

METHODS

First-trimester cfDNA (n = 35) and cfDNA after diagnosis (n = 21) were collected from EOPE pregnancies and matched-controls. Placental tissues were collected from EOPE (n = 5) and control (n = 7) pregnancies. MeD-seq assays were performed to identify DMRs. Placental-contribution to cfDNA, cfDNA concentration, EOPE-specific placental markers, and methylation levels at previously identified EOPE-associated regions were compared between cases and controls. Additionally, liver-, kidney-, and endothelium-specific hypermethylated regions were identified using publicly available datasets and compared in cfDNA.

RESULTS

We identified one DMR in the first trimester and 249 DMRs in our after diagnosis cohort. Total cfDNA concentration was 20-fold higher after EOPE diagnosis compared to controls (p < 0.001). We identified 606 DMRs in EOPE placentas, which unexpectedly were almost not present in cfDNA after EOPE diagnosis. Also, the placental-, liver-, and kidney-specific contributions were relatively decreased, while endothelium-specific methylation was increased. Only 2.7-8.4 % of previously identified regions showed significant differential methylation in our cohort.

DISCUSSION: Differential tissue contributions to cfDNA likely account for DMRs identified after EOPE diagnosis. Further research is required to clarify the (tissue) origins of identified cfDNA DMRs in more detail. Lack of concordance across studies remains a significant limitation, potentially explained by methodological heterogeneity.