Single-cell 3D Genome Atlases Describing Cancer Progression Reveal Stage-specific Alterations and Novel Biomarkers

By Stuart P. Atkinson

image 0
Chromatin tracing revealed that as healthy lung cells evolve into adenomas and then invasive tumors, their 3D genomes become progressively more open and variable. Measurements of distances between chromatin regions across chromosome 13—and similar patterns observed genome-wide—show increasing decompaction and heterogeneity, highlighting stepwise reorganization of genome architecture during cancer development.

The Unknown Evolution and Diversification of the 3D Genome Architecture During Tumor Progression

A range of studies have provided robust evidence for ongoing alterations to three-dimensional (3D) genome organization during normal human development and when comparing cancerous to healthy tissues over time (Yu & Ren) ; indeed, alterations to nuclear size, shape, and chromatin texture can aid cancer diagnosis and grading (Zink, Fischer, & Nickerson). Even given this knowledge, we lack a deep understanding of the evolution and diversification of the 3D genome during cancer progression, as most related research has focused on a single time point.

Researchers from the laboratories of Mandar Deepak Muzumdar and Siyuan Wang (Yale University) recently applied a technique known as chromatin tracing to directly visualize 3D genome organization in cells from tissue samples and then create single-cell 3D genome atlases of lung adenocarcinoma and pancreatic ductal adenocarcinoma progression in mice. Chromatin tracing - an imaging-based 3D genomics method pioneered in their laboratories (Hu & Wang and Wang et al.) enables the direct capture of 3D genome organization data from individual cells within the native tissue environment during cancer progression. This approach offers distinct advantages over related techniques, such as high-throughput chromosome conformation capture, that rely on the population averaging of cells and the indirect inference of altered genome organization in averaged cancer cells.

As reported in their recent Nature Genetics study, (Liu, Jin, and Agabiti et al.), the resultant data enabled a detailed description of stage-specific alterations in 3D genome compaction, heterogeneity, and compartmentalization as lung adenocarcinoma and pancreatic ductal adenocarcinoma progress from normal to preinvasive to invasive disease forms and the identification of clinically-relevant diagnostic, prognostic, and therapeutic biomarkers.

Paired-Tag technology from Epigenome Technologies enables the simultaneous profiling of transcriptomics and epigenetics in single cells. Could the integration of this approach into this fascinating new study have yielded additional insights and added depth to these single-cell atlases of cancer progression? Paired-Tag technology from Epigenome Technologies generates joint epigenetic and transcriptomic profiles at single-cell resolution and detects histone modifications and RNA transcripts in individual nuclei with comparable efficiency to single-nucleus RNA-seq/ChIP-seq assays while avoiding the need for cell sorting. Could Paired-Tag have allowed the direct mapping of 3D genome architectures to transcriptionally and epigenetically defined cell states?

image 1
Analysis of chromatin demixing and interchromosomal distances revealed that 3D genome organization changes in characteristic ways across disease stages. Each cancer state—from normal lung cells to adenomas and invasive tumors—shows a distinct 3D genomic profile, allowing researchers to discriminate between stages and link specific structural patterns with cancer progression.

Single-cell 3D Genome Cancer Cell Atlases: Defining Stage-specific Alterations and Novel Biomarkers

In brief, this exciting study generated comprehensive 3D genome organization atlases in mouse models of lung and pancreatic cancer progression and introduced a pipeline to support the measurement and interpretation of biological heterogeneity during disease progression in complex tissues. Analysis of these model systems revealed a gradual increase in transcriptional heterogeneity, accompanied by stage-specific alterations in 3D genome compaction, heterogeneity, and compartmentalization during progression from normal tissues to preinvasive tumors and ultimately to invasive tumors. Of note, the observation of non-monotonic 3D genome organization evolution suggested the existence of a structural bottleneck in single cells during early tumor progression . Comparison of 3D genome organization across tumors and mice suggested a link between increased compaction/decreased heterogeneity of chromatin conformations in pre-tumorigenic cells with phenotypic convergence driven by the Kras oncogene.

Analysis of these atlases also revealed that single-cell 3D genome organization could distinguish and predict cancer states with high accuracy, despite the data's inherent variability. Furthermore, the authors revealed that high-dimensional 3D genome organization data, when combined with transcriptomic data, identified genes that govern prognosis and delineate cancer cell dependency. Finally, they underscored the potential of their 3D genome organization data to identify novel regulators; for example, this study reported that Rnf2 - a component of the polycomb repressive complex 1, which catalyzes mono-ubiquitination of lysine 119 of histone H2A (Wang et al.) - plays a non-canonical role in the regulation of 3D genome organization during tumor evolution.

image 2
The study identified Rnf2, a component of the polycomb repressive complex 1, as a partial regulator of 3D genome changes occurring during the transition from adenoma to invasive lung adenocarcinoma (LUAD). Correlation analyses showed that loss of Rnf2 mimics many of the 3D structural shifts seen in tumor progression, linking Rnf2 activity to large-scale chromatin compartment changes. Supporting epigenetic data revealed that Rnf2 binds preferentially to both active and repressed genomic regions, underscoring its complex, non-canonical role in reshaping genome architecture during cancer evolution.

Toward the Development of Improved Single-cell 3D Genome Atlases of Cancer Cells

While these findings and the developed methodologies now offer a means to apply single-cell 3D genome organization data to identify potential diagnostic, prognostic, and therapeutic cancer biomarkers, the authors note the need for future studies that could significantly advance the development of improved single-cell 3D genome atlases of cancer cells. Such studies may involve the combination of genome-wide chromatin tracing with spatial transcriptomics, epigenomics, and proteomics in the same single cancer cells within native tissues, thereby directly mapping 3D genome architectures to transcriptionally and epigenetically defined cell states. Furthermore, they believe that the broad application of single-cell 3D genome mapping to cancer patient clinical samples may hold promise for advancing cancer diagnosis, subtype classification, treatment response prediction, and target discovery.

Importantly, the additional integration of simultaneous profiling of transcriptomics and epigenetics in single cells, afforded by applying Paired-Tag technology from Epigenome Technologies, could provide a more detailed description of the progression of lung adenocarcinoma and pancreatic ductal adenocarcinoma, and directly map 3D genome architectures to transcriptionally and epigenetically defined cell states.