In order to fit inside a microscopic cell, human DNA — a strand of code that measures nearly two meters in length — must be compacted into a 10 micrometer-diameter nucleus. Chromatin, the raw material for making chromosomes, makes this possible by creating a 3D jigsaw puzzle for the DNA.

Understanding the unique gene regulatory mechanisms at play in cells that are diseased versus those that are healthy requires understanding the spatial organization of this chromatin, and molecular biologists use a set of techniques known as Chromosome Conformation Capture (3C) to do so. 

A related approach known as Hi-C combines 3C and next-generation sequencing to comprehensively detect genome-wide chromatin interactions in the nucleus. The chromatin is crosslinked, digested with enzymes and then ligated (joined together) so fragments that are spatially close will have a higher probability of being ligated into the same fragment. These ligation fragments are then enriched and sequenced. The data are then presented as a two-dimensional matrix, recording the pairwise contact frequency between genomic region bins across the whole genome.

Historically, however, bulk Hi-C methods have fallen short in resolving cell-type-specific chromatin interactions. Single-cell Hi-C methods have emerged as an alternative, but these approaches have their own limitations: They are either too low-throughput to capture the diversity of cells in a sample, or too labor-intensive, costly or difficult to integrate into the pipeline of a typical lab setting.

Our new method, Droplet Hi-C, is poised to address the remaining limitations of single-cell sequencing, offering an unprecedented view into the chromatin architecture of individual cells. Droplet Hi-C encapsulates individual cells in droplets by way of a commercialized microfluidic device (10x Genomics) to enable high-throughput single-cell chromatin profiling. It can process tens of thousands of cells in a single experiment, dramatically reducing the cost and hands-on time compared to existing techniques. This scalability makes it possible to profile chromatin architecture in heterogeneous tissues, offering unprecedented insights into how gene regulation varies across cell types and leading, in our case, to the discovery of dynamics in “extra-chromosomal” DNA.

Droplet Hi-C in Action: Mapping the Mouse Cortex

To demonstrate the power of Droplet Hi-C, we turned to the mouse brain. The adult mouse cortex is a highly heterogeneous tissue composed of diverse neuronal and glial cell types, each with its own gene regulatory landscape. Using Droplet Hi-C, we were able to map the chromatin architecture of major cortical cell types, revealing how their chromatin organization correlates with epigenetic modification and gene expression patterns. Additionally, we developed an algorithm to identify chromatin hubs in single cells. Around 5% genomic bins were identified as chromatin hubs, with most being cell type-specific and enriched at super-enhancers and marker genes from the corresponding cell type.

This application of Droplet Hi-C opens the door to new discoveries in neurobiology, where understanding the relationship between chromatin architecture and brain function is crucial. From uncovering the molecular basis of neurological disorders to exploring how chromatin changes during brain development and aging, Droplet Hi-C provides a powerful tool for studying the brain at the single-cell level.

Decoding Cancer: Detecting Oncogenic Events with Droplet Hi-C

Beyond the brain, Droplet Hi-C has significant implications for cancer research. Tumors are notoriously heterogeneous, composed of a mix of cancerous, stromal, and immune cells. To effectively target cancer, we need to understand the chromatin organization in cancer cells, particularly how chromatin architecture drives oncogenic gene expression.

Using Droplet Hi-C, we profiled chromatin architecture in human glioblastoma cells, colorectal cancer cells, and blood cancer cells. In addition to mapping chromatin interactions, we were able to detect copy number variations (CNVs) and large-scale structural variations (SVs), such as duplications, and translocations, that are characteristic of cancer genomes. One of the most exciting findings from our study was the detection of extrachromosomal DNA (ecDNA) in cancer cells. EcDNA, which often harbors oncogenes, plays a key role in cancer progression and drug resistance. By analyzing the chromatin interactions of ecDNA at single-cell resolution, we gained new insights into how to distinguish ecDNA from homogeneously staining regions (HSR) by sequencing data.

Importantly, we also used Droplet Hi-C to track changes in chromatin structure during drug treatment. For example, in glioblastoma cells treated with a tyrosine kinase inhibitor, we observed the disappearance of ecEGFR (an EGFR oncogenic ecDNA) and the emergence of new ecDNA elements carrying other oncogenes. This dynamic evolution of ecDNA during therapy highlights the importance of chromatin profiling in understanding cancer drug resistance mechanisms and developing more effective treatments.

Toward Multi-Modal Profiling: Combining Chromatin and Transcriptome Data

One of the most promising uses of Droplet Hi-C is its ability to perform joint profiling of chromatin architecture and the transcriptome in single cells. By capturing both the 3D genome structure and gene expression data from the same cell, we can directly explore the relationship between chromatin organization and gene regulation. We modified the Droplet Hi-C protocol and created Paired Hi-C, which simultaneously profile RNA and Hi-C from the same single cell or nucleus.

This multimodal approach is particularly valuable for studying how changes in chromatin architecture lead to changes in gene expression in diseases such as cancer. For instance, in our glioblastoma studies, we found that trans-interaction patterns of ecDNA harboring MYC gene before and after drug treatment were associated with distinct transcriptional responses. By integrating transcriptome data with chromatin contact maps, we could pinpoint how ecDNA chromatin interactions drive oncogenic gene expression, providing new targets for cancer therapies.

Looking Ahead: The Future of Chromatin Research with Droplet Hi-C

With its superior throughput, reduced cost, and ability to analyze chromatin architecture in heterogeneous tissues, Droplet Hi-C represents a major leap forward in single-cell genomics. The potential applications of this technology are vast, from basic research into gene regulation and development to clinical applications in cancer diagnosis and treatment. In the future, we envision Droplet Hi-C being used to profile chromatin architecture in patient samples, providing personalized insights into disease mechanisms and treatment responses.

As we continue to refine Droplet Hi-C and expand its applications, we are excited to see how this technology will transform the field of chromatin research. Whether used for studies of brain development, cancer evolution, or gene regulation in other tissues, Droplet Hi-C offers a scalable, powerful tool for uncovering the hidden architecture of the genome.