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
. 2023 Aug;620(7972):218-225.
doi: 10.1038/s41586-023-06327-7. Epub 2023 Jul 12.

Retrotransposons hijack alt-EJ for DNA replication and eccDNA biogenesis

Fu Yang #  1 , Weijia Su #  1 , Oliver W Chung  1 , Lauren Tracy  1 , Lu Wang  1   2 , Dale A Ramsden  3 , Z Z Zhao Zhang  4   5
Affiliations

Retrotransposons hijack alt-EJ for DNA replication and eccDNA biogenesis

Fu Yang et al. Nature. 2023 Aug.

Abstract

Retrotransposons are highly enriched in the animal genome1-3. The activation of retrotransposons can rewrite host DNA information and fundamentally impact host biology1-3. Although developmental activation of retrotransposons can offer benefits for the host, such as against virus infection, uncontrolled activation promotes disease or potentially drives ageing1-5. After activation, retrotransposons use their mRNA as templates to synthesize double-stranded DNA for making new insertions in the host genome1-3,6. Although the reverse transcriptase that they encode can synthesize the first-strand DNA1-3,6, how the second-strand DNA is generated remains largely unclear. Here we report that retrotransposons hijack the alternative end-joining (alt-EJ) DNA repair process of the host for a circularization step to synthesize their second-strand DNA. We used Nanopore sequencing to examine the fates of replicated retrotransposon DNA, and found that 10% of them achieve new insertions, whereas 90% exist as extrachromosomal circular DNA (eccDNA). Using eccDNA production as a readout, further genetic screens identified factors from alt-EJ as essential for retrotransposon replication. alt-EJ drives the second-strand synthesis of the long terminal repeat retrotransposon DNA through a circularization process and is therefore necessary for eccDNA production and new insertions. Together, our study reveals that alt-EJ is essential in driving the propagation of parasitic genomic retroelements. Our study uncovers a conserved function of this understudied DNA repair process, and provides a new perspective to understand-and potentially control-the retrotransposon life cycle.

PubMed Disclaimer

Conflict of interest statement

COMPETING INTERESTS

Z.Z, F.Y., and W.S. are co-inventors on a US provisional patent application filed by Duke University related to this work.

Figures

Extended Data Fig. 1 |
Extended Data Fig. 1 |. The engineered HMS-Beagle reporter dominantly forms eccDNA.
a, Schematic design of the HMS-Beagle reporter. An eGFP reporter is inserted into the 3’ UTR of HMS-Beagle sequence in an antisense direction. b, Fly cross scheme to collect samples for measuring the potential integration and eccDNA events from transposon-silenced and transposon-activated flies. c, Integrative Genomics Viewer (IGV) alignments showing reads mapped to HMS-Beagle reporter locus in the genome from embryos laid by transposon-silenced females. Individual purple horizonal bar represents a unique Nanopore read containing eGFP sequence. All of the reads contained at least one of the LTRs and extended to the adjacent region, indicating they were aligned to the original genomic locus of the reporter. d, The distribution of new integrations from engineered HMS-Beagle reporter on Drosophila genome. Each triangle represents a new integration event.
Extended Data Fig. 2 |
Extended Data Fig. 2 |. PCR based assay to measure HMS-Beagle eccDNA.
a, schematic of the design of divergent primers to identify retrotransposon eccDNA. b, AFM imaging to visualize the shapes of DNA. Exonuclease digestion significantly enrich eccDNA for detection in panel c. Scale bar, 500 nm. c, the representative gel image showing retrotransposons predominantly form 1-LTR circles. Performing PCR using total DNA as template produced non-specific bands, likely resulting from the nested transposon fragments resided within the linear genome. Using exonuclease to enrich eccDNA generated two PCR products corresponding to 1-LTR and 2-LTR eccDNA respectively. d and e, Sanger sequencing to validate the formation of HMS-Beagle eccDNA. The PCR products for the very right lane of panel c were cloned into plasmid vector and 11 corresponding colonies were sequenced. Ten of the 11 colonies are from 1-LTR eccDNA (d). One colony is from 2-LTR eccDNA (e). Notably, this 2-LTR eccDNA has 34 bp deletion at the end-end junction site, indicating it is formed by the error-prone NHEJ pathway. This conclusion is further supported by Extended Data Fig. 9.
Extended Data Fig. 3 |
Extended Data Fig. 3 |. eccDNA-Seq to provide direct evidence of circle formation.
a, Schematic of the eccDNA-Seq workflow. After extracting total DNA, linear DNA was removed by Plasmid-safe DNase digestion. eccDNA was amplified by Phi29 DNA polymerase through rolling circle amplification. And the sequencing libraries were prepared and sequenced on a Nanopore instrument. b, The proportion of eccDNA-Seq and Genome-Seq reads mapped to mitochondria (black regions), transposons (pink regions), and the rest of the genome (gray regions). All samples were from fly ovaries. The Genome-Seq libraries were made by the tagmentation method and sequenced by the Nanopore platform to capture circular DNA, such as the mitochondrial genome. c, Bar graph showing qPCR results of mitochondrial DNA copies detected by two sets of primers respectively. The relative abundances are normalized to the spike-in plasmid. The mitochondrial DNA copies are essentially unchanged upon transposon activation in Drosophila ovary. The bars report mean ± standard deviation from three biological replicates (n=3). p values were calculated with a two-tailed, two-sample unequal variance t test. d, Circos plots showing the number of the eccDNA-Seq reads for the four classes of HMS-Beagle circles. From the outer layer to inward: 1-LTR full-length circles, 2-LTR full-length circles, 1-LTR-rearranged circle, and non-LTR rearranged circles.
Extended Data Fig. 4 |
Extended Data Fig. 4 |. eccDNA production from HMS-Beagle requires its mRNA intermediates.
a, RNA-FISH to detect HMS-Beagle mRNA. All flies carrying sh-aub to activate transposons in germline cells. Further introducing sh-white (serving as a control) into the animals does not change transposon activity: HMS-Beagle remains activated. Upon introducing sh-HMS-Beagle construct to silence it, its RNA was undetectable by RNA-FISH. Scale bar, 20 μm, b, Top: primer design to detect HMS-Beagle eccDNA (Extended Data Fig. 2). Bottom: The representative gel image of PCR products showing that HMS-Beagle eccDNA production was abolished when its mRNA production was suppressed by RNAi. Each genotype has three biological replicates.
Extended Data Fig. 5 |
Extended Data Fig. 5 |. Confirmation of the RNAi silencing efficiency in oocytes.
RT-qPCR showing the depletion efficiency of indicated genes by germline-specific RNAi. Relative mRNA levels were normalized to rp49 gene. The bars report mean ± standard deviation from four biological replicates (n=4). p values were calculated with a two-tailed, two-sample unequal variance t test. Silencing Lig3 or Fen1 made flies barely lay eggs/oocytes, impeding a validation of the RNAi silencing efficiency for them.
Extended Data Fig. 6 |
Extended Data Fig. 6 |. HMS-Beagle mRNA remains unchanged upon depletion of the components from alt-EJ process.
Transposon activation was achieved by silencing Aub in germline cells. The Y-axis is normalized reads count.
Extended Data Fig. 7 |
Extended Data Fig. 7 |. Immunoprecipitation assay to measure the accumulation of HMS-Beagle single-stranded DNA upon alt-EJ suppression.
The bars report mean ± standard deviation from four biological replicates (n=4). P values were calculated with a two-tailed, two-sample unequal variance t test. Although Mab3034 antibodies used in this experiment have 10-fold higher affinity for single-stranded DNA than double-stranded DNA, they still can bind HMS-Beagle genomic double-stranded DNA across all samples. This would mask the difference of single-stranded DNA across samples and lead to underestimation of the amount of accumulated single-stranded DNA upon alt-EJ inhibition.
Extended Data Fig. 8 |
Extended Data Fig. 8 |. IAP needs its reverse transcriptase, but not integrase, activity for eccDNA biogenesis.
a, Sanger sequencing to validate the IAP reverse transcriptase mutant. b, Sanger sequencing to validate the IAP integrase mutant. c, PCR based assay to measure the production of IAP eccDNA. The very left lane was the condition without introducing IAP plasmid. d-e, Either immunoblotting (d and e) or RT-qPCR (f) to test the silencing efficiency of CRISPRi on depleting the alt-EJ factors. For each gene, two gRNAs were designed. NT (non-targeting) is a random gRNA without a targeting site. For RT-qPCR, relative mRNA levels were normalized to the RR18S gene. The bars report mean ± standard deviation from four biological replicates (n=4). Statistical significance were calculated with a two-tailed, two-sample unequal variance t test. The p value for gRNA-1 is 0.0011, while for gRNA-2 is 0.0014.
Extended Data Fig. 9 |
Extended Data Fig. 9 |. NHEJ pathway is essential for 2-LTR eccDNA biogenesis.
a, Mutating Lig4 abolishes 2-LTR eccDNA production for mdg4 retrotransposon. b, Silencing Ku80 or Lig4 by RNAi reduces mdg4 2-LTR eccDNA formation. c, Sanger sequencing to validate lig4 mutation of the 293T cells. d, Sanger sequencing to validate XRCC4 mutation of the 293T cells. e, mutating either Lig4 or XRCC4 abolishes 2-LTR eccDNA production for IAP retrotransposon.
Extended Data Fig. 10 |
Extended Data Fig. 10 |. Detailed model of the replication cycle of LTR-retrotransposons supported by our study.
Our data support alt-EJ factors mediate a circularization step for retrotransposon 2-nd strand DNA synthesis. While this step can generate full-length linear double-stranded DNA for integration, it appears to dominantly produce 1-LTR eccDNA.
Fig. 1 |
Fig. 1 |. HMS-Beagle predominantly produces eccDNA upon activation.
a, Table to summarize the outcomes of replicated HMS-Beagle DNA detected in Drosophila oocytes. HMS-Beagle activation is achieved by suppressing Aub and Ago3 during oogenesis. b, Workflow to characterize the integration and circularization (eccDNA) events from an engineered HMS-Beagle reporter. The Nanopore sequencing reads were classified as integration by having flanking sequences mapped to the genome or as eccDNA by containing end-to-end junction sites. The eccDNA was further classified into four categories based on their structures. The numbers within parentheses denote the number of reads identified for each type of event. c, eccDNA reads from engineered HMS-Beagle reporter. Each circle represents a read: the solid part represents the sequenced region, and the dashed line represents the gap filled computationally. Salmon: reads supporting 1-LTR full-length circles; gold: one read supporting 2-LTR full-length circle; purple: reads supporting 1-LTR rearranged circles, likely resulting from autointegration; dark green: eccDNA reads do not contain intact LTR. d, Circos plot showing eccDNA reads from endogenous HMS-Beagle. Color scores indicate the mapping coverage throughout the full-length HMS-Beagle consensus. Outer layer: eccDNA reads from HMS-Beagle silenced oocytes. Inner layer: eccDNA reads from HMS-Beagle activated oocytes. Data for this figure were generated from flies carrying sh-aub and sh-ago3 to trigger transposon activation.
Fig. 2 |
Fig. 2 |. Factors from the alt-EJ process drive 1-LTR full-length eccDNA formation.
a, Schematic of the design of divergent primers to identify HMS-Beagle eccDNA. b-d, The representative gel image to show whether components from the HR (b), NHEJ (c), or alt-EJ (d) process are required for the formation of 1-LTR full-length eccDNA. e, Circos plot showing 1-LTR full-length eccDNA reads from endogenous HMS-Beagle. Color scores indicate the mapping coverage throughout the full-length HMS-Beagle consensus. The numbers within parentheses denote the number of reads identified from each genotype. Data for this figure were generated from flies carrying sh-aub to trigger transposon activation.
Fig. 3 |
Fig. 3 |. Blocking alt-EJ process abrogates DNA synthesis and all eccDNA production from HMS-Beagle.
a, A model to depict how alt-EJ-mediated circularization drives DNA synthesis, thus is essential for eccDNA production and mobilization. Step 1: tRNA fragment pairs with the primer binding site (PBS) to initiate the 1st-strand DNA synthesis via reverse transcription to form RNA:DNA hybrid. Step 2: RNase H activity to remove RNA from the RNA:DNA hybrid, but leaving a polypurine tract (PPT). Step 3: PPT initiates the 2nd-strand DNA synthesis for the 3′-LTR. Step 4: alt-EJ-mediated circularization drives the synthesis of the remaining of the 2nd-strand DNA. b, qPCR to quantify the relative abundance of single-stranded HMS-Beagle DNA. The bars report mean ± standard deviation from three biological replicates (n=3). p values were calculated with a two-tailed, two-sample unequal variance t test. c, Circos plot showing 2-LTR full-length eccDNA reads from endogenous HMS-Beagle. Color scores indicate the mapping coverage throughout the full-length HMS-Beagle consensus. The numbers within parentheses denote the number of reads identified from each genotype. These circles are likely generated by joining the 2 LTRs together via NHEJ. d, Circos plot showing rearranged eccDNA reads from endogenous HMS-Beagle. Color scores indicate the mapping coverage throughout the full-length HMS-Beagle consensus. The numbers within parentheses denote the number of reads identified for each genotype. 1-LTR rearranged circles are likely generated by autointegration events: LTR to attack its own interstitial sequences in cis. 0-LTR rearranged circles are possibly the by-products of autointegration. Data for this figure were generated from flies carrying sh-aub to trigger transposon activation.
Fig. 4 |
Fig. 4 |. Blocking alt-EJ process abrogates HMS-Beagle mobilization.
a, Nanopore sequencing to directly examine the biogenesis of full-length double-stranded linear DNA from replicated HMS-Beagle. Red- and blue-colored reads reflect the sequenced plus and minus strands respectively. b, Dot plots to display the new integrations from HMS-Beagle. Each triangle represents an integration event detected by Nanopore genome-seq. The numbers in parentheses present the total amounts of integration events detected. The numbers of integration detected under transposon-silenced condition likely represent the false-positive rates from our methodology. Data for this figure were generated from flies carrying sh-aub to trigger transposon activation.
Fig. 5 |
Fig. 5 |. mdg4 and mammalian IAP form eccDNA via the alt-EJ factors.
a, Schematic design of the divergent primers to detect mdg4 or IAP eccDNA. b, mdg4 retrotransposon produces both 1-LTR and 2-LTR eccDNA at the pupal stage, the time window when mobilization occurs. Performing PCR using total DNA as template produced non-specific bands. Using exonuclease to enrich eccDNA generated two PCR products corresponding to 1-LTR and 2-LTR eccDNA respectively, as confirmed by Sanger sequencing. c, eccDNA production from mdg4 depends on alt-EJ factors. d, Suppressing the factors from alt-EJ repair process blocks IAP eccDNA biogenesis. The PCR products corresponding to 1-LTR and 2-LTR eccDNA were confirmed by Sanger sequencing. NT (non-targeting) is a random gRNA without a targeting site in the human genome.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Wells JN & Feschotte C A Field Guide to Eukaryotic Transposable Elements. Annual review of genetics (2020). 10.1146/annurev-genet-040620-022145 - DOI - PMC - PubMed
    1. Kazazian HH Jr. & Moran JV Mobile DNA in Health and Disease. The New England journal of medicine 377, 361–370 (2017). 10.1056/NEJMra1510092 - DOI - PMC - PubMed
    1. Fueyo R, Judd J, Feschotte C & Wysocka J Roles of transposable elements in the regulation of mammalian transcription. Nature reviews. Molecular cell biology (2022). 10.1038/s41580-022-00457-y - DOI - PMC - PubMed
    1. Frank JA et al. Evolution and antiviral activity of a human protein of retroviral origin. Science 378, 422–428 (2022). 10.1126/science.abq7871 - DOI - PMC - PubMed
    1. Wang L et al. Retrotransposon activation during Drosophila metamorphosis conditions adult antiviral responses. Nature genetics 54, 1933–1945 (2022). 10.1038/s41588-022-01214-9 - DOI - PMC - PubMed