Cell type-specific DNA methylation in neonatal cord tissue and cord blood: a 850K-reference panel and comparison of cell types

ABSTRACT

Accounting for cellular heterogeneity is essential in neonatal epigenome-wide association studies (EWAS) performed on heterogeneous tissues, such as umbilical cord tissue (CT) or cord blood (CB). Using a reference-panel-based statistical approach, the cell type composition of heterogeneous tissues can be estimated by comparison of whole tissue DNA methylation profiles with cell type-specific DNA methylation signatures. Currently, there is no adequate DNA methylation reference panel for CT, and existing CB panels have been generated on lower coverage Infinium HumanMethylation450 arrays. In this study, we generate a reference panel for CT and improve available CB panels by using the higher coverage Infinium MethylationEPIC arrays. We performed DNA methylation profiling of 9 cell types isolated from CT and CB samples from 14 neonates. In addition to these cell types, we profiled DNA methylation of unfractionated CT and CB. Cell type composition of these unfractionated tissue samples, as estimated by our reference panels, was in agreement with that obtained by flow cytometry. Expectedly, DNA methylation profiles from CT and CB were distinct, reflecting their mesenchymal and hematopoietic stem cell origins. Variable CpGs from both unfractionated CT and its isolated cell types were more likely to be located in open seas and intronic regions than those in CB. Cell type specific CpGs in CT were enriched in intercellular matrix pathways, while those from CB were enriched in immune-related pathways. This study provides an open source reference panel for estimation and adjustment of cellular heterogeneity in CT and CB, and broadens the scope of tissue utilization assessed in future neonatal EWAS studies.

KEYWORDS:

• Cellular heterogeneity
• neonatal epigenome wide association studies
• tissue-specificity
• DNA methylation
• reference panel

Disclosure statement

YSC has received reimbursement for speaking at conferences sponsored by companies selling nutritional products. He is part of an academic consortium that has received research funding from Abbott Nutrition, Nestec and Danone. The other authors declare no competing interests.

Supplementary material

Supplementary data for this article can be accessed here.

Additional information

Funding

This work was supported by the A*STAR Strategic Positioning Funds (SPF2013/002) available to NK. XL is supported by Duke-NUS block fund (R-913-200-127-263) and Ministry of Education, Singapore Academic Research grant Tier 2 (MOE2018-T2-1-046).