One of the ways a cancer-causing gene works up enough power to turn a normal cell into a cancer cell is by copying itself over and over, like a Xerox machine.
Scientists have long noticed that when cancer-causing genes do that, they also scoop up some extra DNA into their copies.
But it has remained unclear whether the additional DNA helps drive cancer or is just along for the ride.
Using human glioblastoma brain tumor samples, researchers at University of California San Diego School of Medicine and Case Western Reserve University School of Medicine have now determined that all of that extra DNA is critical for maintaining a cancer-causing gene’s activation, and ultimately supporting a cancer cell’s ability to survive.
Comparing those findings to a public database of patient tumor genetics, they also discovered that even if two different tumor types are driven by the same cancer-causing gene, the extra DNA may differ.
The study, published November 21, 2019 in Cell, could explain why drugs will often work for some cancer types but not others.
“We’ve been targeting the cancer-causing gene for therapy, but it turns out we should also think about targeting the switches that are carried along with it,” said co-senior author Peter Scacheri, PhD, Gertrude Donnelly Hess Professor of Oncology at Case Western Reserve University School of Medicine and member of the Case Comprehensive Cancer Center.
When the human genome was first fully sequenced, many people were surprised to find it contained far fewer genes — segments of DNA that encode proteins — than expected. It turns out that the remainder of human DNA in the genome, the non-coding regions, play important roles in regulating and enhancing the protein-coding genes — turning them “on” and “off,” for example.
In this study, the researchers focused on one example cancer-causing gene, EGFR, which is particularly active in glioblastoma, an aggressive form of brain cancer, and other cancers. When copies of EGFR pile up in tumors, they tend to be in the form of circular DNA, separate from the chromosome.
“In 2004, I was the lead on the first clinical trial to test a small molecule inhibitor of EGFR in glioblastoma,” said co-senior author Jeremy Rich, MD, professor of medicine at UC San Diego School of Medicine and director of neuro-oncology and director of the Brain Tumor Institute at UC San Diego Health. “But it didn’t work. And here we are 15 years later, still trying to understand why brain tumors don’t respond to inhibitors of what seems to be one of the most important genes to make this cancer grow.”
The team took a closer look at the extra DNA surrounding EGFR circles in 9 of 44 different glioblastoma tumor samples donated by patients undergoing surgery. They discovered that the circles contained as many as 20 to 50 enhancers and other regulatory elements. Some of the regulatory elements had been adjacent to EGFR in the genome, but others were pulled in from other regions of the genome.
Glioblastoma tumor cells with extrachromosomal EGFR gene amplification (red) indicated by white arrows. The image is credited to Case Western Reserve.
To determine the role each regulatory element plays, the researchers silenced them one at a time. They concluded that nearly every single regulatory element contributed to tumor growth.
“It looks like the cancer-causing gene grabs as many switches it can get its hands on … co-opting their normal activity to maximize its own expression,” Scacheri said.
First author Andrew Morton, a graduate student in Scacheri’s lab, then searched a public database of patient tumor genetic information — more than 4,500 records covering nine different cancer types. He found that the team’s observation was not limited to EGFR and glioblastoma.
Enhancers were amplified alongside cancer-causing genes in many tumors, most notably the MYC gene in medulloblastoma and MYCN in neuroblastoma and Wilms tumors.
“People thought that the high copy number alone explained the high activity levels of cancer-causing genes, but that’s because people weren’t really looking at the enhancers,” Morton said. “The field has been really gene-centric up to this point, and now we’re taking a broader view.”
Next, the researchers want to know if the diversity in regulatory elements across cancer types could also be helping tumors evolve and resist chemotherapy.
They also hope to find a class of therapeutic drugs that inhibit these regulatory elements, providing another way to put the brakes on cancer-causing genes.
“This isn’t just a laboratory phenomenon, it’s information I need to better treat my patients,” said Rich, who is also a faculty member in the Sanford Consortium for Regenerative Medicine and Sanford Stem Cell Clinical Center at UC San Diego Health.
Additional study co-authors include: Nergiz Dogan-Artun, Princess Margaret Cancer Centre, University Health Network; Zachary J. Faber, Cynthia F. Bartels, Kevin C. Allan, Case Western Reserve University; Graham MacLeod, Stephane Angers, University of Toronto; Megan S. Piazza, Shashirekha Shetty, University Hospitals, Cleveland; Stephen C. Mack, Baylor College of Medicine; Xiuxing Wang, Qiulian Wu, UC San Diego; Ryan C. Gimple, UC San Diego and Case Western Reserve University; Brian P. Rubin, Cleveland Clinic; Peter B. Dirks, The Hospital for Sick Children, Ontario Institute for Cancer Research; Richard C. Sallari, Axiotl, Inc.; Mathieu Lupien, Princess Margaret Cancer Centre, University Health Network, Ontario Institute for Cancer Research, University of Toronto.
The tumor microenvironment (TME) comprises a heterogeneous number and type of cellular and noncellular components that vary in the context of molecular, genomic and epigenomic levels. The genotypic diversity and plasticity within cancer cells are known to be affected by genomic instability and genome alterations.
Besides genomic instability within the chromosomal linear DNA, an extra factor appears in the form of extrachromosomal circular DNAs (eccDNAs; 2–20 kbp) and microDNAs (200–400 bp). This extra heterogeneity within cancer cells in the form of an abundance of eccDNAs adds another dimension to the expression of procancer players, such as oncoproteins, acting as a driver for cancer cell survival and proliferation.
This article reviews research into eccDNAs centering around cancer plasticity and hallmarks, and discusses these facts in light of therapeutics and biomarker development.
Recently, appreciable attention has been drawn to the understanding and harnessing of tumor heterogeneity in the context of preclinical and clinical aspects of cancer biomarkers and therapeutics [1–5]. Tumor heterogeneity is known to be affected by several factors such as molecular, genetic and epigenetic changes, extrachromosomal circular DNAs (eccDNAs), nuclear microDNAs, secreted microDNAs, noncoding miRNAs and environmental pressures [1–9]. It is widely known that genetic aberrations and instabilities at the genome level in several forms drive carcinogenesis, as well as several key tumor hallmarks and adaptations within the tumor microenvironment (TME) [3–9]. Therefore, basic, preclinical and clinical studies are required to seek the change of conducive pro-TME to an anticancer microenvironment, ensuring the success of next-generation therapeutics such as precision and personalized medicine [3–9].
In addition to genomic alterations in linear chromosomal DNA, the contribution of eccDNAs to some extent is envisaged as regards observed tumor heterogeneity and crucial tumor hallmarks such as invasiveness and drug resistance [8–19]. The existence of eccDNAs in normal cells as well as cancer cells was demonstrated a few decades ago. The origin of eccDNAs within the nucleus is facilitated by normal cellular processes, an abnormal DNA repair system and environmental pressures such as stress, carcinogens and pathogens [12–28]. Interestingly, these eccDNAs are reported in several organisms including Drosophila, yeast, plants, animals and humans [8–12,29–34]. However, abundance of these eccDNAs is a matter of investigation across different organisms and between normal and cancer cells. Hence, attempts have been made to understand the mechanisms of formation of eccDNAs in different types of cells such as normal cells, cancer cells, and cancer-associated cells and distinct cellular heterogeneity within a tumor [8–12,29–34].
Current focuses on eccDNAs are also inclined toward understanding ways to mitigate their abundance and tumorigenic impact within the cellular heterogeneity of a tumor. Various potential avenues are suggested in the form of gene therapy, small RNA mimetics to interfere with eccDNAs, natural sources of small RNAs and inhibitors of the DNA repair protein system [35–40]. Furthermore, evidence from eccDNA research is being translated into avenues for development of cancer biomarkers and diagnostic tools [35–40]. This mini-review addresses the recent developments in deciphering a link between eccDNAs and TME that may be translated into future therapeutic and diagnostic avenues.Go to:
Tumor heterogeneity driving tumor hallmarks
Recently, the complex nature of cancer has been attributed to the tumor heterogeneity emanating at different levels including at the molecular, genomic and epigenomic level, and various environmental pressures [1–9,41–44]. In a true sense, tumor heterogeneity that drives growth, proliferation and invasiveness is essentially additive in nature [1,2,9,43,44]. With reference to genetic heterogeneity, research has established that aberrations and variations at the gene level within the linear chromosomal DNAs drive tumor heterogeneity in the form of procancer factors [8,12,13,15–19,25]. However, there is limited evidence demonstrating the implications of increased number of eccDNAs (2–20 kbp size) and also circular microDNAs (200–400 bp) within the nucleus of cancer cells [8–12,15–19,25,33,34]. In brief, heterogeneity displayed by various factors including genetic, epigenetic, eccDNAs, nuclear microDNAs, secreted microDNAs, miRNAs and environmental pressures contributes toward complexity and plasticity within a tumor. A summary of updated tumor hallmarks is presented in Figure 1.
Extrachromosomal circular DNAs
In living organisms, the cellular landscape is known to have distinct genetic components including chromosomal nuclear DNA, coding RNAs, noncoding regulatory RNAs and so on. Besides chromosomal nuclear DNA, eccDNAs are reported in cytoplasmic organelles such as mitochondria in animal cells and both mitochondria and chloroplasts in plant cells [10,11,17,20,29–32].
Besides the mitochondrial and viral origin of eccDNAs, eccDNAs are present in the nucleus of normal and unhealthy cells (including cancer cells), and are clearly linked to coherent cellular signaling and functions [12,14,17–19,24–28]. These specialized eccDNAs are mostly suggested to originate from repetitive genomic sequences such as telomeric DNA or rDNA [1,8,12–19,25,34].
Various organisms including yeast, Caenorhabditis elegans, Drosophila melanogaster, mammals and plants have been shown to harbor eccDNAs in the nucleus and their presence is linked to normal functions and associated phenotypes of these organisms [8,10–20,25,29–34]. Another observation indicates the complexity of small circular DNA in D. melanogaster and also the homology between small circular DNA and middle-repetitive chromosomal DNA. These small circular DNAs are shown to possess mobility which can result in deletions, mutations, chromosomal variations and oncogenesis .
There are contrasting differences in the description of the basis of generation of eccDNAs in eukaryotic cells. One study proposes that in the case of Drosophila, eccDNAs are formed by a preferred homologous recombination process between tandem repeats, and this study further suggested that formation of eccDNAs is independent of any DNA repair process .
Meanwhile, another study on Xenopus suggests that de novo synthesis from naked DNA and telomere DNA of sperm nuclei results in the development of extrachromosomal circular telomeric DNA consisting of vertebrate repeats of telomeric DNA (TTAGGG)n .
Previous research screening yeast genomic DNA indicated that eccDNAs are found in abundance (~23%), which led to the claim that eccDNAs are one of the factors involved in mutation and evolutionary characteristics of the eukaryotic genome .
Interestingly, the contribution of nonrepetitive DNA to the rise of eccDNAs is drawing wide attention in the case of both normal and cancer cells [1,2,5,8,10,12–19,25]. Shoura et al. presented evidence in support of the presence of eccDNAs in C. elegans and human-derived cell lines . Data indicate that eccDNAs can originate from both coding and noncoding regions in response to normal cellular settings.
It is important to mention that in prokaryotic and eukaryotic cells, besides the main genome-contributing chromosome, small circular double-stranded DNA such as eccDNAs originate from repetitive DNA sequences within the genome (2–20 Kbp) and serve various purposes including drug resistance, cellular survival, metabolism, signaling and death [12,14,17–28].
Currently, approaches such as Circle-Seq, 2D gel electrophoresis and next-generation sequencing are reported to validate the presence of eccDNAs in various eukaryotic cells including yeasts, plant cells and animal cells. Taken together, cutting edge genomics tools clearly support the presence of eccDNAs in various cell types including of the mammalian system and these results advocate further exploration to find clues and linkages for normal developmental processes and disease conditions including tumors.
Mechanisms of formation of eccDNAs
In both normal and cancer cells, formation of eccDNAs is suggested to be both by DNA replication-dependent and -independent means [12,14,17–19,22–28]. Commonly, eccDNAs from organisms ranging from yeasts, to plants and to animals are suggested to originate from long-repetitive, short-repetitive and nonrepetitive sequences of the genome [14,17,33].
In the majority of cases, these eccDNAs are known to contain ribosomal genes, transposon remnants, coding tandemly repeated genes (HXT6/7, ENA1/2/5 and CUP1-1/-2), noncoding chromosomal high-copy tandem repeats, telomeric DNA and oncogenes [12,14,17–19,22–28]. Recently, Shibata et al. convincingly suggested the existence of eccDNA as microDNAs in the range of 200–400 bp . Their data suggest that these eccDNAs can originate from unique nonrepetitive DNA sequence in mouse tissues as well as mouse and human cell lines.
The eccDNAs or ring chromosomes are suggested to cause deletions, mutations, replications, amplifications or translocations of genes; these genetic alterations are attributed to eccDNAs for their uniqueness in changing size, and self-deletion and integration within the genome [12,14,17–19,28,45]. As described earlier, evidence was found on de novo synthesis of eccDNAs in Xenopus embryos, which suggested that eccDNA formation is not a random DNA replication between maternal tandemly repeated multimers and the paternal genomic DNA template, but is a de novo process unique to the preblastula stage of development .
In contrast to the idea of de novo synthesis and a DNA replication-dependent process of formation for eccDNAs, another study suggests that their origin requires excision of chromosomal sequences facilitated by sequence independent enzymes independent of ATP solely with the help of Mg2+ and ameliorated through the double-strand DNA break repair system .
Recently, microDNAs ranging from 200–400 bp have been reported to originate mostly from the nonrepetitive sequences within the genome as a result of heightened abnormalities in the DNA repair system including homologous recombination, nonhomologous recombination and the mismatch repair system [12,14,17–28]. However, little is known about the mechanisms inducing the alterations of copy number of eccDNAs within the nucleus of animal, yeast and plant cells.
Additional views are accumulating to indicate the existence of DNA-independent mechanisms involving the DNA repair system to support the enhanced presence of eccDNAs in cancer cells [12,14,17–19,26–28]. In research supporting the basis of elevated levels of eccDNA in mammalian cells, Cohen et al. showed that extrachromosomal circular major satellites can be a hotspot for the generation of eccDNAs .
Their data also suggest the contribution of DNA replication and DNA ligase IV toward generation of eccDNA in highly proliferating cancer cells. As an additional basis of eccDNA generation, data indicate that defective MSH3 DNA mismatch repair protein can result in decreased levels of eccDNAs emanating from the non-CpG regions during normal cellular physiology .
Another finding reports on the presence of a varied size of cell-free eccDNAs, from 31–19,989 bp, and showed a higher GC content in case of microDNAs less than 500 bp. Furthermore, the formation of these eccDNAs is linked to the non-homologous end joining mechanisms of the DNA repair pathway . Taken together, these eccDNAs are suggested to be present in both normal and abnormal cells. Currently, eccDNAs are being studied for their role in genetic variation, evolution, genomic instability, mutation and tumorigenesis.
Tumor hallmarks & eccDNAs
In nature, the eukaryotic system shows unique developmental patterns, alterations in phenotype due to environmental cues, various types of stress, aging and chemotherapeutic cancer drug resistance. These processes are associated with the high abundance of eccDNAs [4,19,32,45–47]. Recently, it has become apparent that tumor hallmarks are driven by various factors including genetic, epigenetic and environmental pressures. At the genetic level, chromosomal and extrachromosomal components of nuclear and mitochondrial origin have been shown to contribute toward tumor heterogeneity [13,18,19]. These extra players are commonly found in various tissues and cell types, and in both normal and diseased conditions. Due to their highly heterogeneous origins and widespread occurrence in nearly all eukaryotes, eccDNAs are believed to reflect the genome’s plasticity and instability [4,19,37,45–47]. The significance of eccDNAs in eukaryotic DNA amplification associated with cancer development and metastasis has been studied .
Additionally, another experiment provides an insight into another possible approach – that submicroscopic elements comprising human MYC oncogenes that replicate semiconservatively serve as a precursor of double-minutes . Sen et al. reported on the successful lysis of colon carcinoma cells to detect the submicroscopic structures of eccDNAs by alkaline lysis and PCR .
An earlier study suggests the presence of eccDNAs in HeLa S3 cells, which could have originated due to nonhomologous recombination within the nonrepetitive or low-copy DNA sequences . Hence, rejoining fragmented DNA could be a step to generate eccDNAs in the cellular system. It is true that under different circumstances, healthy and unhealthy cells contain and release extracellular free eccDNAs of variable size, sequence complexity, copy number and homology to chromosomal DNA [4,19,37,45–50].
These variations in eccDNAs – specifically with reference to tumor tissues – are attributed to several factors including genotype of the tumor, intratumor heterogeneity and environmental factors including lifestyle, nutrition, stress and so on [4,19,37,45–51].
A significant research analysis in glioblastoma of approximately 198 patients has shown the prevalence of amplification-linked extrachromosomal mutations due to which there are momentous mutations in oncogenes such as PDGFRA or EGFR in glioblastoma due to environmental effects .
A recent report describes the presence of microDNAs, a class of eccDNAs, in 20 independent human lymphoblastoid cell lines in the context of treatment of chemotherapeutic drugs .
In this experiment, their findings confirmed the generation of 190 bp microDNAs from the active regions of the genome and drug treatment demonstrated an increased proportion of microDNA loci in human lymphoblastoid cell lines. Interesting evidence has been found using whole-genome sequencing that detected extrachromosomal DNA in approximately half the human cancers and varied in frequency in each cancer type.
Recent whole-genome sequencing data on 17 different cancer types showed the presence of eccDNAs in nearly half of human cancers . Furthermore, these data indicate that frequency of eccDNAs may vary by tumor type. In this study, the authors supported the role of eccDNAs in the acceleration of tumor progression and malignancy.
Despite valid questions regarding the possible origin and nature of the DNA sequence, nature of function and clinical relevance, the existence of eccDNAs is attracting an appreciable place in the area of cancer biomarkers and therapeutics .
Contribution of eccDNAs in drug resistance
It has been suggested that the eccDNAs derived from a single cancer cell or heterogeneous cancer cells intensify the complexity of the resultant tumor, displaying drug resistance [1,35–40]. One such drug, Methotrexate, an antimetabolite drug, has been shown to induce drug resistance in colon cancer cell lines and the observed drug resistance is linked to the presence of intrachromosomal homogeneously staining regions or double minutes . Another report suggests the role of extrachromosomal plasmid DNA to confer drug resistance in cervical cancer . Overall, the driving forces behind cancer drug resistance, propelled by the presence of eccDNAs, are as yet unclear and in future focused studies are warranted to uncover these issues to aid in cancer therapeutics.
Heather Buschman – UCSD
The image is credited to Case Western Reserve.
Original Research: Closed access
“Functional Enhancers Shape Extrachromosomal Oncogene Amplifications”. Peter Scacheri et al.