Profiling DNA circles, one cell at a time

SciLifeLab researchers from Karolinska Institutet have developed a new method capable of profiling extrachromosomal circular DNAs (eccDNAs) at single-cell resolution. With this method, they dissected the diversity and cell type specificity of eccDNAs within different cell populations. The results shed light on how these circular DNA elements are generated and how they can influence tumorigenesis. The study is published in Nature Communications.

For most cancer types, oncogenes play a big role as the driving force of tumor initiation and progression. One mechanism by which oncogenes exert their tumorigenic potential is by being amplified inside the cell nucleus, so that many more copies of the oncogene mRNA can be produced. Oncogene amplification in cancer cells is often achieved by packaging many copies of the same oncogene inside so-called extra-chromosomal DNAs or eccDNAs, a special class of circular DNAs that can accumulate in hundreds of copies inside cancer cell nuclei, contributing to dysregulating physiologic gene expression programs.

Using bulk genomic methods such as whole-genome sequencing (WGS) or Circle-seq, one can get an average portrait of the eccDNA repertoire of a specific cell or tumor type. However, eccDNAs are highly dynamic and heterogeneous, thus bulk assays cannot fully capture the diversity and complexity of eccDNAs in single cells.

Single-cell view of eccDNAs: a mix of stochasticity and specificity

To study eccDNAs in tumors at single-cell resolution, a team of researchers from Karolinska Institutet and SciLifeLab, led by Magda Bienko and Nicola Crosetto, developed a new methodology, named scCircle-seq, which enables sequencing eccDNAs in single cells with high sensitivity and precision.

“We were amazed to find that most of the eccDNAs detected by scCircle-seq seem to form stochastically along the genome and are unique to a given cell”, says the study’s lead author, Jinxin Phaedo Chen, who is a PhD student in the Bienko-Crosetto Lab.

Using the newly developed method, the researchers demonstrated that most eccDNAs (90%) largely vary even between cells of the same type and are stochastically inherited during cell division. However, the researchers also found that the genome-wide landscape of eccDNAs differs between cells of different origin and can be used to accurately cluster cells of the same origin together.

The researchers also found that eccDNAs are preferentially produced from chromatin regions enriched in specific histone marks and are induced during replication stress conditions. Furthermore, examination of the relationship between eccDNA copy number and expression of the genes contained in eccDNAs from the same cell revealed no remarkable correlation, except for few oncogenes, including MYC, that were contained within a large recurrent eccDNA in a colorectal cancer cell line.

They were also able to demonstrate that their method can be applied to detect eccDNAs in single tumor nuclei extracted from tumor biopsies, hence paving the way to studying these important genetic elements directly in patient samples.

“Our data indicate that, although most individual eccDNAs are stochastic in terms of size, they tend to form more in specific genomic regions,” says Magda Bienko and continues “once eccDNAs containing genes that provide a massive proliferative advantage, such as MYC, are formed, they can be detected across many cells”.

“I believe that scCircle-seq is a powerful addition to the rapidly expanding single-cell sequencing toolkit that can be used to unravel the complexity of tumors”, says Nicola Crosetto. “I envision that, thanks to its sensitivity, our method could also be harnessed to detect eccDNAs in cell-free plasma DNA, paving the way to future studies assessing the potential of eccDNAs as biomarkers for early disease detection”, Nicola concludes.

The research was supported by grants from the Swedish Cancer Society, the Strategic Research Programme in Cancer at Karolinska Institutet (now Cancer Research KI), the Ragnar Soderberg Foundation, and the Swedish Research Council.


Last updated: 2024-03-06

Content Responsible: victor kuismin(