Single-cell Chromatin Conformations mapped by Droplet Hi-C: Part 2

January 12, 2025 · By Stuart P. Atkinson

Multi-panel figure describing a single-cell experimental workflow in mice — Experimental design of Paired-Tag in normal and diabeteic mice, from Wang et al.

1. Introduction: A focus on β Cells in Type 2 Diabetes

In the third in this series of blogs, we reported on the description of Droplet Hi-C by researchers from the laboratory of Bing Ren (University of California, San Diego) (Chang et al.). This scalable, accessible, droplet-based single-cell Hi-C technique represents an optimized method for high-throughput, single-cell chromatin conformation profiling in droplets, which combines in situ chromosomal conformation capture with a commercially available microfluidic platform. Droplet Hi-C simultaneously captures the 3D structure from tens of thousands of individual cells in a single experiment, which the authors initially employed to map cell-type-specific chromatin architecture in the mouse cortex and, as such, addressed critical gaps in chromatin analysis of heterogeneous tissues.

RNA velocity analysis of beta-cell stress in low-fat to high-fat diets — Pseudotime analysis of RNA uncovers distinct sets of beta cell, healthy (beta-hi) and stressed (beta-lo); the latter of which is more prevalent in high-fat diet conditions

2. The Importance of Chromatin Organization to Normal Development and Tumorigenesis

In the fourth of this series of blogs, Epigenome Technologies continues the coverage of Chang et al. by describing how Droplet Hi-C can also evaluate aberrant chromatin structures in cancer cells and map the chromatin interactome of extrachromosomal (ec)DNAs at single-cell resolution, given their relationship with cancer initiation/progression (Aaltonen et al.). ecDNAs – characterized by amplification and a circular structure (Wu et al.) – are prevalent in human cancers and generally harbor oncogenes. While bulk Hi-C analysis can detect structural variations and ecDNAs in tumor tissues (Dixon et al. and Harewood et al.), we lack single-cell resolution and a detailed picture of treatment-associated evolution. Investigating chromatin conformation in primary tissues and tumor biopsies via bulk Hi-C techniques also suffers from challenges associated with sample heterogeneity, a lack of cell-type resolution, and the complexity of data analysis.

We also report how Chang et al. refined Droplet Hi-C to create Paired Hi-C to support simultaneous Hi-C and transcriptomic profiling in single cells and enable the study of gene expression in relation to 3D genome structure. Paired Hi-C supports multimodal epigenetic analysis in single cells; meanwhile, parallel analysis of individual cells for RNA expression and DNA from targeted tagmentation by sequencing or “Paired-Tag” from Epigenome Technologies can generate joint epigenetic and gene expression profiles at the single-cell resolution and detect histone modifications and RNA transcripts in individual nuclei with an efficiency comparable to single-nucleus RNA-seq/ChIP–seq assays.

Multi-panel figure showing single-cell UMAPs built from histone marks, and corresponding IGV tracks — Clustering of pancreatic cells by histone modifications alone shows high correspondence with RNA-based annotations, and concordance between active marks at marker genes and associated cell types.

3. Detecting Chromosomal Aberrations and Uncovers ecDNA Heterogeneity and Evolution

An initial application of Droplet Hi-C to colorectal cancer cell lines revealed elevated copies of the MYC locus (resides on ecDNA), as expected, and distinct structural variations, including a duplication on chromosome 6

Droplet Hi-C distinguished ecDNAs from homogeneously staining regions (HSRs, sites of multiple repeats of a DNA segment) and suggested that ecDNA tends to aggregate into hubs (Hung et al.)
These findings aided the development of a multivariate logistic regression model to predict ecDNA (distinguished from HSRs) based on differences in chromosomal interaction patterns

The authors employed Droplet Hi-C to profile glioblastoma cell lines with/without treatment, which induces drug resistance via the loss of epidermal growth factor receptor (EGFR) ecDNA (Nathanson et al.), highlighting treatment-induced alterations in chromatin architecture at the single-cell level
Non-treated glioblastoma cells displayed elevated interchromosomal contacts at the EGFR and MYC loci; however, interchromosomal contacts at the EGFR locus diminished but increased at the MYC locus after treatment, while they emerged at a new locus encompassing the MDM2i loci

Non-treated cells contained ecEGFR/ecMYC; however, treated cells only contained ecMYC/ecMDM2
ecEGFR disappeared rapidly upon treatment, ecMDM2 appeared only in treatment-resistant cells, and treatment induced the rapid transformation of the structure and size of the ecMYC amplicon

IGV tracks showing peak-TSS links — Paired-Tag allows cis-CREs to be linked to TFs via expression/abundance correlations, explicating the combinatorial logic of expression regulation

4. Droplet Hi-C: Detecting Glioblastoma Sample Heterogeneity

When applied to primary glioblastoma samples, Droplet Hi-C effectively detected copy number variations, ecDNAs, structural variations, and chromatin architecture specific to malignant cells

They observed expected chromosomal aberrations (including at the EGFR locus), while the previously described model confirmed ecEGFR enrichment in malignant cells
Droplet Hi-C identified structural variations specific to malignant cells, including the translocation of the tumor suppressor gene IKZF1 to ecEGFR without its promoter, leading to decreased transcription

Chromatin compartment analysis revealed that state-specific gene expression aligned with chromatin architecture in primary glioblastoma samples

Comparing compartment differences between different states identified ~1,800 switched compartments, which significantly associated with differential gene expression and state-specific biological functions
These data underscored insulation score alterations and loss of domain boundaries in malignant cells, which can rewire enhancer-promoter interactions and contribute to glioblastoma progression (Rahme et al.)

Droplet Hi-C analysis of bone marrow mononuclear cells from a patient with myelodysplastic syndrome and secondary acute myeloid leukemia revealed alterations consistent with the patient’s remission after treatment

Changes included the disappearance of ecMYC, a diminished proportion of malignant cells, and the loss of cancer-specific enhancer-promoter interactions

Chromatin velicty analysis for two histone marks in single cells — Joint Chromatin Velocity and RNA Velocity refines the trajectory of beta-cell stress.

5. Paired Hi-C: Simultaneous Transcriptome and Hi-C Profiling in the Same Single Cell

The team modified Droplet Hi-C to work with the 10x Genomics Chromium Single Cell Multiome kit, creating Paired Hi-C, which simultaneously profiles RNA and Hi-C from the same single nuclei at high throughput and efficiency

Applying Paired Hi-C to the adult mouse cortex identified excitatory and inhibitory neurons and non-neuronal types
Compartment scores strongly correlated between Droplet Hi-C and sn-m3C-seq profiles from the same cell types, with concordant changes in compartment and gene expression at cell-type-specific marker genes

Applying Paired Hi-C to post-treatment glioblastoma cells revealed that gene expression correlated with ecDNA copy number, although the relationship between ecDNA structure and gene expression remains complex

They observed different trans-interaction patterns of ecMYC in untreated/treated glioblastoma cells associated with distinct transcriptional responses
Additional analyses highlighted the interplay between ecDNA dynamics, cellular states, and drug resistance

Droplet Hi-C Supports Chromatin Conformation Analysis in Cancer Cells; Paired Hi-C Takes a Step Further

This fourth blog article of the series describes how applying Droplet Hi-C can improve our understanding of tumor evolution and treatment response by analyzing ecDNA dynamics and help explore the regulatory programs driving tumor progression/drug resistance by evaluating structural variations and chromatin architecture. The evolution of Droplet Hi-C into Paired Hi-C permits the definition of links between chromatin reorganization/structural alterations to gene expression, which may allow the correlation of tumor progression/drug resistance with alterations to gene regulatory mechanisms.

Paired-tag represents a platform similar to Paired Hi-C, creating joint epigenetic and gene expression profiles at single-cell resolution and detecting histone modifications and RNA transcripts in individual nuclei. This advance was also developed in the Bing Ren lab; now, Epigenome Technologies provides optimized Paired-Tag kits and services under an exclusive license from the Ludwig Institute for Cancer Research.

For more on how Droplet Hi-C supports single-cell chromatin conformation mapping of cancer cells and for more on the groundbreaking potential of Paired-Hi-C, see Nature Biotechnology, October 2024.