Digital PCR Assay Utilizing In-Droplet Methylation-Sensitive Digestion for Estimation of Fetal cfDNA From Plasma

Abstract

Objective: Recent guidelines suggest that non-invasive prenatal screening (NIPS) should be offered to all patients with singleton and twin pregnancies. Accurate determination of fetal fraction in cell-free DNA (cfDNA) is vital for reliable NIPS outcomes. We propose a methylation-based approach using droplet digital PCR (ddPCR) and methylation-sensitive restriction enzyme (MSRE) digestion for fetal fraction quantification as an affordable and fast solution.

Method: Following biomarker discovery using early pregnancy placental genomic DNA (gDNA) and cfDNA from non-pregnant female individuals, we designed assays targeting MSRE-compatible regions based on contrasting methylation patterns between maternal and fetal cfDNA. We established a proof-of-concept ddPCR workflow on the Bio-Rad Droplet Digital PCR QX600 instrument.

Results: Testing the fetal fraction assay multiplex on 137 prospective clinical samples demonstrated high concordance with NGS results for both female and male pregnancies as well as with chromosome Y-based calculations for samples with a male fetus. Reproducibility analysis indicated lower variability compared to previously reported NGS performance.

Conclusion: This study showcases the potential of this novel, 6-color, high-multiplex methylation ddPCR panel for accurate measurement of fetal fraction in cfDNA samples. It presents opportunities to integrate such methodology as a standalone measurement to assess the quality of samples undergoing NIPS.

FIGURE 1

Discovery of fetal and maternal biomarkers using whole genome methylation sequencing of placenta tissue and non‐pregnant maternal cfDNA. (A) Genome‐wide methylation frequencies reveal similarities between samples of the same group. UMAP projections of samples based on whole genome methylation frequency profiles separate samples by DNA source (gDNA and cfDNA) and sample source (placenta vs. plasma). (B) Significant differentially methylated CpG loci were identified distinguishing early pregnancy (EP) placenta tissue samples from non‐pregnant (NP) female cfDNA samples. A Volcano plot was made using 2500 randomly selected CpG loci, sites were colored if the average methylation frequency difference between sample groups was > 0.5 or less than −0.5 and the Bonferroni adjusted p‐value was < 0.05. The colored points refer to significantly differentially methylated sites, where “fetal‐diff” loci are significantly hypermethylated in placenta tissue compared to non‐pregnant cfDNA, and vice versa for “maternal‐diff” sites. (C) Differentially methylated loci distinguish sample origins. A heatmap was made using 2500 randomly selected CpG loci. The non‐pregnant plasma samples (NP), pregnant plasma samples, and early pregnancy (EP) placenta sample groups and fetal genders were annotated by color on the left. Individual sample identifiers (S#) were included on the right with each sample group. A color heat scale was used to represent the methylation frequency of CpG sites, blue for hypo‐methylated and red for hyper‐methylated status. (D) Biomarkers were selected from differentially methylated sites and labeled as fetal if they were hypermethylated (≥ 0.8) in early pregnancy placenta and hypomethylated (≤ 0.2) in non‐pregnant samples and vice versa for maternal. Boxplots of methylation frequency distributions for both fetal biomarkers and maternal biomarkers were made for all samples and separated into panels for each sample group.

FIGURE 2

In silico validation of differential methylation analysis. (A) Chromosomal z‐scores showed elevated methylation frequencies at fetal‐specific differential sites (fetal‐diff) on chromosomes 18 and 21 in the trisomy 18 (T18) and trisomy 21 (T21) cfDNA samples, respectively. Z‐score distributions are displayed for each chromosome in every sample, representing the proportion of CpG base sequencing counts methylated at fetal‐diff sites, normalized to the distributions observed in four euploid pregnancies. Chromosomes 18 and 21 are highlighted in the respective trisomy 18 and 21 samples. NP corresponds to non‐pregnant cfDNA. (B) Conversely, chromosomal z‐scores revealed decreased methylation frequencies at maternal‐specific differential sites (maternal‐diff) on chromosomes 18 and 21 in the trisomy 18 and trisomy 21 samples, respectively. The z‐score distribution plot for maternal‐diff sites was normalized as described in (A). (C) Cell types relevant to early fetal development were identified by gene set enrichment analysis (GSEA). GSEA was performed on all genes containing significantly differentially methylated fetal or maternal biomarkers at their promoter regions (± 2 kb from the transcription start site). Genes were ranked by the median p‐value of CpGs within the promoter region.

FIGURE 3

Fetal fraction ddPCR assay performance with cfDNA samples. (A) ChrY‐based fetal fraction estimates are consistent with methylation‐based fetal fraction estimates for male fetal samples using ddPCR. The scatterplot compares fetal fraction estimations from clinical samples using multiplex ddPCR assays targeting chrY and methylation biomarkers. The Pearson correlation coefficient (R = 0.64, p = 3e‐7, n = 67) was calculated for chrY and methylation values from male samples. (B) Combined ddPCR fetal fraction readouts aligned with NGS‐based methods. The combined ddPCR readout included chrY and methylation‐based measurement for male fetal samples, and methylation‐based measurement for female samples. The Pearson correlation coefficient (R = 0.70, p = 5e‐18, n = 113) was calculated using both male and female samples. (C) Evaluation of duplicate blood draws demonstrates the reproducibility of the ddPCR‐based assay. Fetal fraction values from the methylation‐based ddPCR assay were compared by analyzing measurements from independently extracted duplicate blood samples and run on different plates. The Pearson correlation coefficient (R = 0.95, p = 4e‐48, n = 126) was calculated using both male and female samples.