Detection of Brain-Derived Cell-Free DNA in Plasma

Abstract

Background: Neuronal loss is a major pathological feature of neurodegenerative diseases. The analysis of plasma cell-free DNA (cfDNA) is an emerging approach to track cell death events in a minimally invasive way and from inaccessible areas of the body, such as the brain. Previous studies showed that DNA methylation (DNAm) profiles can be used to map the tissue of origin of cfDNA and to identify molecules released from the brain upon cell death. The aim of the present study is to contribute to this research field, presenting the development and validation of an assay for the detection of brain-derived cfDNA (bcfDNA). Methods: To identify CpG sites with brain-specific DNAm, we compared brain and non-brain tissues for their chromatin state profiles and genome-wide DNAm data, available in public datasets. The selected target genomic regions were experimentally validated by bisulfite sequencing on DNA extracted from 44 different autoptic tissues, including multiple brain regions. Sequencing data were analysed to identify brain-specific epihaplotypes. The developed assay was tested in plasma cfDNA from patients with immune effector cell-associated neurotoxicity syndrome (ICANS) following chimeric antigen receptor T (CAR-T) therapy. Results: We validated five genomic regions with brain-specific DNAm (four hypomethylated and one hypermethylated in the brain). DNAm analysis of the selected genomic regions in plasma samples from CAR-T patients revealed higher levels of bcfDNA in participants with ongoing neurotoxicity syndrome. Conclusions: We developed an assay for the analysis of bcfDNA in plasma. The assay is a promising tool for the early detection of neuronal loss in neurodegenerative diseases.

Keywords: DNA methylation; cfDNA; epihaplotypes; neurodegeneration; target bisulfite sequencing.

Figure 1

Study design. The flowchart summarizes the design of the study, including the selection and validation of genomic regions with brain-specific DNAm [10].

Figure 2

DNAm profile of brain-specific targets. (A–C) The boxplots report the DNAm values of 3 representative microarray probes resulting from the comparison between brain and non-brain (other) tissues from the GEO repository. (D–F) The boxplots report the DNAm values of the 3 CpG sites depicted in panels (A–C), measured by target bisulfite sequencing in 44 GTEx brain and non-brain tissues. (G–I) The line plots show the mean DNAm levels of the CpG sites included in the 3 representative target regions in brain and non-brain tissues (other). Note that the figure shows 3 out 8 of the selected targets; the remaining targets are reported in Figures S1 and S2.

Figure 3

Epihaplotype frequencies of brain-specific targets. The heatmaps represent the frequencies of epihaplotypes (rows) in the GTEx samples (columns) for targets (A) APC2, (B) PACRG and (C) C10orf90. Heatmap colours go from white (frequency of a certain epihaplotype near to 0% in a certain sample) to dark blue (frequency of a certain epihaplotype near to 100% in a certain sample). The colour bar above the heatmap indicates brain (orange) and non-brain (blue) tissues. For the sake of clarity, the heatmaps include only the 10 most frequent epihaplotypes for each target region. Dendrograms were generated using complete-linkage clustering on Euclidean distances matrix. Note that the figure shows 3 out 8 of the selected targets; the remaining targets are reported in Figure S3.

Figure 4

Analysis of bcfDNA in plasma samples from patients with and without ICANS. (A) Read depth of the 5 target regions analysed through target bisulfite sequencing in the plasma samples from the 5 patients who received CD19 CAR T-cell therapy. (B) Levels of bcfDNA in CAR T patients who developed or did not develop ICANS.