Researchers Decode the “Dual Code” of Cell Identity, Release World’s First Single-Cell 3D Genome Map

An international research team led by the Salk Institute and the Arc Institute has built the world’s first integrated map simultaneously analyzing DNA methylation and 3D genome structure at the level of individual human cells.

The team recently announced in the international journal Science that it had released the “Human Single-Cell 3D Genome and DNA Methylation Atlas,” containing epigenetic information from 86,689 individual cells obtained from 16 human tissues. The study was conducted as part of the National Institutes of Health’s (NIH) 4D Nucleome project.

All cells in the human body have the same DNA sequence, but epigenetic regulation enables them to differentiate into cells with different functions, such as heart and nerve cells. Previous studies could analyze only one of the two systems—chemical tags attached to DNA, known as methylation, or the 3D folding structure of chromatin—making it difficult to understand how the regulatory systems interact.

Using a technology called “snm3C-seq,” the research team overcame the technical limitations of simultaneously capturing methylation and 3D chromatin structure within individual cells. This enabled the team to precisely classify 35 major cell types and 206 detailed cell subtypes.

The map is expected to become a key tool for identifying the causes of disease. The research team focused on the fact that most disease-associated genetic variants are located in noncoding DNA regions that do not produce proteins. Using the map, the team was able to accurately identify the cells in which specific disease-risk variants are expressed.

In practice, variants associated with blood glucose regulation appeared specifically in pancreatic endocrine cells; atrial fibrillation risk regions appeared in cardiac muscle cells; and hair loss-associated variants appeared in dermal fibroblasts. Variants associated with schizophrenia and bipolar disorder were concentrated in specific nerve cells.

Joseph Ecker, a professor at the Salk Institute, explained, “Even the same genetic variant can determine whether disease develops depending on the cell’s epigenetic environment. 3D genome information provides a link between noncoding variants and cell function, dramatically narrowing the scope of research into disease mechanisms.”

In particular, the study revealed for the first time that DNA methylation and 3D chromatin structure do not always change in sync. In skeletal muscle cells, the researchers identified a subpopulation in which the 3D structure had the form of a fully differentiated mature cell, while the DNA methylation state retained characteristics of immature muscle stem cells.

This suggests that when cells differentiate, their 3D structure is reorganized first, followed by changes in DNA methylation patterns. The research team predicted that this finding—that the two types of epigenetic information change at different rates—will alter the criteria used to define cell types in the future.

The research team has made all raw data and an interactive visualization web browser freely available so researchers worldwide can use the extensive dataset. The database could serve as high-quality training data for artificial intelligence (AI) models that predict the functions of genetic variants.

This study is an interim result of the 4D Nucleome project. The project’s ultimate goal is to add “time” as a fourth dimension and determine how epigenetic information changes dynamically during development, aging and disease. The reference map established in this study is expected to advance precision medicine and genetic disease research into a new dimension.