Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2022 May 6;14(9):2310.
doi: 10.3390/cancers14092310.

Extrachromosomal Circular DNA from TCGA Tumors Is Generated from Common Genomic Loci, Is Characterized by Self-Homology and DNA Motifs near Circle Breakpoints

Philip D Tatman  1 , Joshua C Black  1
Affiliations

Extrachromosomal Circular DNA from TCGA Tumors Is Generated from Common Genomic Loci, Is Characterized by Self-Homology and DNA Motifs near Circle Breakpoints

Philip D Tatman et al. Cancers (Basel). .

Abstract

Extrachromosomal circular DNA has emerged as a frequent genomic alteration in tumors. High numbers of circular DNAs correspond to poor prognosis suggesting an important function in tumor biology. However, despite mounting evidence supporting the importance of circular DNA, little is known about their production, maintenance, or selection. To provide insight into these processes, we analyzed circular DNA elements computationally identified in 355 TCGA tumors spanning 22 tumor types. Circular DNAs originated from common genomic loci irrespective of cancer type. Genes found in circularized genomic regions were more likely to be expressed and were enriched in cancer-related pathways. Finally, in support of a model for circle generation through either a homology or microhomology-mediated process, circles exhibit homology near their breakpoint. These breakpoints are also enriched in specific DNA motifs. Our analysis supports a model where gene-containing circles emerge from common, highly transcribed regions through a homology-mediated process.

Keywords: TCGA; circular DNA; ecDNA; eccDNA; extrachromosomal DNA; extrachromosomal circular DNA; homologous recombination; microhomology mediated recombination.

PubMed Disclaimer

Conflict of interest statement

The authors declare no conflict of interest.

Figures

Figure 1
Figure 1
DNA circles are found at common loci irrespective of tumor type. (A) Schematic depicting the analysis for each of the 355 TCGA tumor samples. Random circles were constrained to non-centromere locations and forced to generate the same number of and size of real as each individual tumor. (B) Pan-cancer analysis of circles show many tumors have common genomic locations with circles. Top of each chromosome represents the distribution of random circle controls. Below the line represents true circles from TCGA samples. The transition from red to black indicates greater than 75% of samples with a circle in the bin. The coordinates for regions with circles in 191 or more tumors are reported in Table S1. (C) Distribution of the 3 million 1 kb bins with the number of samples with circles in each bin. (D) QQ analysis of the distribution of bins by number of samples. Significance of the overall distribution was assessed by KS test and the deviation of bins with high numbers of samples was assessed by Kurtosis. p-values less than 0.05 were considered significant. An * next to a chromosome indicates a p-values less than 10−16 by KS for that chromosome.
Figure 2
Figure 2
More genes in circles are expressed, and are more highly expressed, than genes located elsewhere in the genome. (A) Schematic showing how genes associated with circles were isolated. No genes that crossed a circle breakpoint were considered associated with a circle. (B) The pan-cancer distribution of the percent of genes expressed in the genome vs expression of genes associated with circles. (C) The distribution of the percentage of genes expressed in the genome vs genes associated with circles for individual cancer types. (D) Schematic showing how genes were categorized to determine the expression differences between genomic genes and genes associated with circles. (E) Pan-cancer analysis of mean gene expression for expressed genes in the genome vs gene associated circles. (F) Individual cancer analysis of mean gene expression for expressed genes in the genome vs genes associated with circles. Significance was determined by two tailed t-test, * indicates a p-value less than 0.05.
Figure 3
Figure 3
The circle-associated transcriptome is enriched with cancer-related pathways. (A) Schematic showing how commonly enriched pathways were defined. Pathway analysis was performed on the highest expressed genes in each tumor, defined as any gene with an FPKM value greater than one standard deviation above the mean. The common, significantly enriched pathways (in black) were then tallied across tumors and displayed as a waterfall plot. (B) Commonly enriched pathways in the circle-associated transcriptome. The percent of tumors with a significant enrichment of a specific pathway is indicated in white lettering inside each bar, while the length of each bar indicates the absolute number of samples with a significant enrichment (see Methods). (C) Schematic showing the isolation of genes commonly found in circles, prior to pathway enrichment. Expressed genes found in circles were isolated for each individual tumor. The number of tumors that expressed a gene located in a circle was totaled for all genes. (D) Pathways common to genes found in circles in greater than 50% of tumors. The length of the bar is the log10(1/p-value). Significance for enrichment was determined using a modified Fisher’s exact text with a p-values less than 0.05 considered significant.
Figure 4
Figure 4
The breakpoints of circular DNA exhibit self-homology. (A) Percent of circles with self-homology compared to random and scrambled controls. (B) Percent of circles with genes with self-homology compared to random and scrambled controls. (C) Percent of circles without genes with self-homology compared to random and scrambled controls. (D) Length of homology in bp for circles compared to random and scrambled controls. (E) Length of homology in bp for circles with genes compared to random and scrambled controls. (F) Length of homology in bp for circles without genes compared to random and scrambled controls. (G) Homology is enriched near the circle breakpoint in all circles. (H) Homology is enriched near the circle breakpoint in circles with genes. (I) Homology is enriched near the circle breakpoint in circles without genes. In = region inside the circle, out = region outside the circle. Significance was determined using an ANOVA, with a post-hoc Tukey test. * indicates p-values less than 0.05.
Figure 5
Figure 5
Circles are self-microhomologous. (A) Percent of circles with self-microhomology compared to random and scrambled controls. (B) Percent of circles with genes with self-microhomology compared to random and scrambled controls. (C) Percent of circles without genes with self-microhomology compared to random and scrambled controls. (D) Length of microhomology in bp for circles compared to random and scrambled controls. (E) Length of microhomology for circles with genes compared to random and scrambled controls. (F) Length of without genes compared to random and scrambled controls. (G) Traces showing the position of microhomology with respect to the circle breakpoint for all circles, random and scrambled controls. (H) Traces showing the position of microhomology with respect to the circle breakpoint for circles with genes, random and scrambled controls. (I) Traces showing the position of microhomology with respect to the circle breakpoint for all circles without genes, random and scrambled controls. Significance was calculated by ANOVA, with a post-hoc Tukey test. * indicates p-values less than 0.05.
Figure 6
Figure 6
DNA binding motifs are enriched at circle breakpoints. (A) Most commonly enriched DNA binding motifs found within ±250 bp of a circle breakpoint for individual tumors. The white lettering shows the percent of tumors with enrichment of the motif while the length of the bar is the absolute number of tumors. (B) Most commonly enriched DNA binding motifs found within ±250 bp of the breakpoint of circles with genes for individual tumors. The white lettering shows the percent of tumors with enrichment of the motif while the length of the bar is the absolute number of tumors. (C) Most commonly enriched DNA binding motifs found within ±250 bp of the breakpoint of circles without a gene for individual tumors. The white lettering shows the percent of tumors with enrichment of the motif while the length of the bar is the absolute number of tumors. (D) Venn diagram depicting the overlap of the total number of enriched motifs found in at least 50% of tumors for circles with genes and circles without genes. (E) Ten most commonly enriched DNA binding motifs in circles with genes that are not enriched in circles without genes. (F) Ten most commonly enriched DNA binding motifs in circles without genes that are not enriched in circles with genes. All values of significance were calculated in the PWMEnrich R package and are based on a simulated modified Fisher’s exact test. Values less than 0.001 were considered significant for this analysis. The complete list of factors with motif enrichments can be found in Tables S4 and S5.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

References

    1. Hotta Y., Bassel A. Molecular size and circularity of DNA in cells of mammals and higher plants. Proc. Natl. Acad. Sci. USA. 1965;53:356–362. doi: 10.1073/pnas.53.2.356. - DOI - PMC - PubMed
    1. Radloff R., Bauer W., Vinograd J. A dye-buoyant-density method for the detection and isolation of closed circular duplex DNA: The closed circular DNA in HeLa cells. Proc. Natl. Acad. Sci. USA. 1967;57:1514–1521. doi: 10.1073/pnas.57.5.1514. - DOI - PMC - PubMed
    1. Zuo S., Yi Y., Wang C., Li X., Zhou M., Peng Q., Zhou J., Yang Y., He Q. Extrachromosomal Circular DNA (eccDNA): From Chaos to Function. Front. Cell Dev. Biol. 2022;9:792555. doi: 10.3389/fcell.2021.792555. - DOI - PMC - PubMed
    1. Ott C.J. Circles with a Point: New Insights into Oncogenic Extrachromosomal DNA. Cancer Cell. 2020;37:145–146. doi: 10.1016/j.ccell.2020.01.008. - DOI - PubMed
    1. Wang T., Zhang H., Zhou Y., Shi J. Extrachromosomal circular DNA: A new potential role in cancer progression. J. Transl. Med. 2021;19:257. doi: 10.1186/s12967-021-02927-x. - DOI - PMC - PubMed

LinkOut - more resources