Stable G-quadruplex DNA structures promote replication-dependent genome instability.

G-quadruplex (G4)-prone structures are abundant in mammalian genomes, where they have been shown to influence DNA replication, transcription, and genome stability. In this article, we constructed cells with a single ectopic homopurine/homopyrimidine repeat tract derived from the polycystic kidney disease type 1 (PKD1) locus, which is capable of forming triplex (H3) and G4 DNA structures. We show that ligand stabilization of these G4 structures results in deletions of the G4 consensus sequence, as well as kilobase deletions spanning the G4 and ectopic sites. Furthermore, we show that DNA double-strand breaks at the ectopic site are dependent on the nuclease Mus81. Hypermutagenesis during sister chromatid repair extends several kilobases from the G4 site and breaks at the G4 site resulting in microhomology-mediated translocations. To determine whether H3 or G4 structures are responsible for homopurine/homopyrimidine tract instability, we derived constructs and cell lines from the PKD1 repeat, which can only form H3 or G4 structures. Under normal growth conditions, we found that G4 cell lines lost the G4 consensus sequence early during clonal outgrowth, whereas H3 cells showed DNA instability early during outgrowth but only lost reporter gene expression after prolonged growth. Thus, both the H3 and G4 non-B conformation DNAs exhibit genomic instability, but they respond differently to endogenous replication stress. Our results show that the outcomes of replication-dependent double-strand breaks at non-B-DNAs model the instability observed in microhomology-mediated break-induced replication (BIR). Marked variability in the frequency of mutagenesis during BIR suggests possible dynamic heterogeneity in the BIR replisome.

Replication stress at microsatellites causes DNA double-strand breaks and break-induced replication.

Short tandemly repeated DNA sequences, termed microsatellites, are abundant in the human genome. These microsatellites exhibit length instability and susceptibility to DNA double-strand breaks (DSBs) due to their tendency to form stable non-B DNA structures. Replication-dependent microsatellite DSBs are linked to genome instability signatures in human developmental diseases and cancers. To probe the causes and consequences of microsatellite DSBs, we designed a dual-fluorescence reporter system to detect DSBs at expanded (CTG/CAG) n and polypurine/polypyrimidine (Pu/Py) mirror repeat structures alongside the c-myc replication origin integrated at a single ectopic chromosomal site. Restriction cleavage near the (CTG/CAG)100 microsatellite leads to homology-directed single-strand annealing between flanking AluY elements and reporter gene deletion that can be detected by flow cytometry. However, in the absence of restriction cleavage, endogenous and exogenous replication stressors induce DSBs at the (CTG/CAG)100 and Pu/Py microsatellites. DSBs map to a narrow region at the downstream edge of the (CTG)100 lagging-strand template. (CTG/CAG) n chromosome fragility is repeat length-dependent, whereas instability at the (Pu/Py) microsatellites depends on replication polarity. Strikingly, restriction-generated DSBs and replication-dependent DSBs are not repaired by the same mechanism. Knockdown of DNA damage response proteins increases (Rad18, polymerase (Pol) η, Pol κ) or decreases (Mus81) the sensitivity of the (CTG/CAG)100 microsatellites to replication stress. Replication stress and DSBs at the ectopic (CTG/CAG)100 microsatellite lead to break-induced replication and high-frequency mutagenesis at a flanking thymidine kinase gene. Our results show that non-B structure-prone microsatellites are susceptible to replication-dependent DSBs that cause genome instability.

Deficiency of the Fanconi anemia E2 ubiqitin conjugase UBE2T only partially abrogates Alu-mediated recombination in a new model of homology dependent recombination.

The primary function of the UBE2T ubiquitin conjugase is in the monoubiquitination of the FANCI-FANCD2 heterodimer, a central step in the Fanconi anemia (FA) pathway. Genetic inactivation of UBE2T is responsible for the phenotypes of FANCT patients; however, a FANCT patient carrying a maternal duplication and a paternal deletion in the UBE2T loci displayed normal peripheral blood counts and UBE2T protein levels in B-lymphoblast cell lines. To test whether reversion by recombination between UBE2T AluYa5 elements could have occurred in the patient's hematopoietic stem cells despite the defects in homologous recombination (HR) in FA cells, we constructed HeLa cell lines containing the UBE2T AluYa5 elements and neighboring intervening sequences flanked by fluorescent reporter genes. Introduction of a DNA double strand break in the model UBE2T locus in vivo promoted single strand annealing (SSA) between proximal Alu elements and deletion of the intervening color marker gene, recapitulating the reversion of the UBE2T duplication in the FA patient. To test whether UBE2T null cells retain HR activity, the UBE2T genes were knocked out in HeLa cells and U2OS cells. CRISPR/Cas9-mediated genetic knockout of UBE2T only partially reduced HR, demonstrating that UBE2T-independent pathways can compensate for the recombination defect in UBE2T/FANCT null cells.

Interaction between DUE-B and Treslin is required to load Cdc45 on chromatin in human cells.

A key step in the initiation of eukaryotic DNA replication is the binding of the activator protein Cdc45 to promote MCM helicase unwinding of the origin template. We show here that the c-myc origin DNA unwinding element-binding protein, DUE-B, interacts in HeLa cells with the replication initiation protein Treslin to allow Cdc45 loading onto chromatin. The chromatin loading of DUE-B and Treslin are mutually dependent, and the DUE-B-Treslin interaction is cell cycle-regulated to peak as cells exit G1 phase prior to the initiation of replication. The conserved C-terminal domain of DUE-B is required for its binding to TopBP1, Treslin, Cdc45, and the MCM2-7 complex, as well as for the efficient loading of Treslin, Cdc45, and TopBP1 on chromatin. These results suggest that DUE-B acts to identify origins by MCM binding and serves as a node for replication protein recruitment and Cdc45 transfer to the prereplication complex.

Break-induced replication links microsatellite expansion to complex genome rearrangements.

The instability of microsatellite DNA repeats is responsible for at least 40 neurodegenerative diseases. Recently, Mirkin and co-workers presented a novel mechanism for microsatellite expansions based on break-induced replication (BIR) at sites of microsatellite-induced replication stalling and fork collapse. The BIR model aims to explain single-step, large expansions of CAG/CTG trinucleotide repeats in dividing cells. BIR has been characterized extensively in Saccharomyces cerevisiae as a mechanism to repair broken DNA replication forks (single-ended DSBs) and degraded telomeric DNA. However, the structural footprints of BIR-like DSB repair have been recognized in human genomic instability and tied to the etiology of diverse developmental diseases; thus, the implications of the paper by Kim et al. (Kim JC, Harris ST, Dinter T, Shah KA, et al., Nat Struct Mol Biol 24: 55-60) extend beyond trinucleotide repeat expansion in yeast and microsatellite instability in human neurological disorders. Significantly, insight into BIR-like repair can explain certain pathways of complex genome rearrangements (CGRs) initiated at non-B form microsatellite DNA in human cancers.

FANCJ is essential to maintain microsatellite structure genome-wide during replication stress.

Microsatellite DNAs that form non-B structures are implicated in replication fork stalling, DNA double strand breaks (DSBs) and human disease. Fanconi anemia (FA) is an inherited disorder in which mutations in at least nineteen genes are responsible for the phenotypes of genome instability and cancer predisposition. FA pathway proteins are active in the resolution of non-B DNA structures including interstrand crosslinks, G quadruplexes and DNA triplexes. In FANCJ helicase depleted cells, we show that hydroxyurea or aphidicolin treatment leads to loss of microsatellite polymerase chain reaction signals and to chromosome recombination at an ectopic hairpin forming CTG/CAG repeat in the HeLa genome. Moreover, diverse endogenous microsatellite signals were also lost upon replication stress after FANCJ depletion, and in FANCJ null patient cells. The phenotype of microsatellite signal instability is specific for FANCJ apart from the intact FA pathway, and is consistent with DSBs at microsatellites genome-wide in FANCJ depleted cells following replication stress.

AluY-mediated germline deletion, duplication and somatic stem cell reversion in UBE2T defines a new subtype of Fanconi anemia.

Fanconi anemia (FA) is a rare inherited disorder clinically characterized by congenital malformations, progressive bone marrow failure and cancer susceptibility. At the cellular level, FA is associated with hypersensitivity to DNA-crosslinking genotoxins. Eight of 17 known FA genes assemble the FA E3 ligase complex, which catalyzes monoubiquitination of FANCD2 and is essential for replicative DNA crosslink repair. Here, we identify the first FA patient with biallelic germline mutations in the ubiquitin E2 conjugase UBE2T. Both mutations were aluY-mediated: a paternal deletion and maternal duplication of exons 2-6. These loss-of-function mutations in UBE2T induced a cellular phenotype similar to biallelic defects in early FA genes with the absence of FANCD2 monoubiquitination. The maternal duplication produced a mutant mRNA that could encode a functional protein but was degraded by nonsense-mediated mRNA decay. In the patient's hematopoietic stem cells, the maternal allele with the duplication of exons 2-6 spontaneously reverted to a wild-type allele by monoallelic recombination at the duplicated aluY repeat, thereby preventing bone marrow failure. Analysis of germline DNA of 814 normal individuals and 850 breast cancer patients for deletion or duplication of UBE2T exons 2-6 identified the deletion in only two controls, suggesting aluY-mediated recombinations within the UBE2T locus are rare and not associated with an increased breast cancer risk. Finally, a loss-of-function germline mutation in UBE2T was detected in a high-risk breast cancer patient with wild-type BRCA1/2. Cumulatively, we identified UBE2T as a bona fide FA gene (FANCT) that also may be a rare cancer susceptibility gene.

Protein phosphatase 2A and Cdc7 kinase regulate the DNA unwinding element-binding protein in replication initiation.

The DNA unwinding element (DUE)-binding protein (DUE-B) binds to replication origins coordinately with the minichromosome maintenance (MCM) helicase and the helicase activator Cdc45 in vivo, and loads Cdc45 onto chromatin in Xenopus egg extracts. Human DUE-B also retains the aminoacyl-tRNA proofreading function of its shorter orthologs in lower organisms. Here we report that phosphorylation of the DUE-B unstructured C-terminal domain unique to higher organisms regulates DUE-B intermolecular binding. Gel filtration analyses show that unphosphorylated DUE-B forms multiple high molecular weight (HMW) complexes. Several aminoacyl-tRNA synthetases and Mcm2-7 proteins were identified by mass spectrometry of the HMW complexes. Aminoacyl-tRNA synthetase binding is RNase A sensitive, whereas interaction with Mcm2-7 is nuclease resistant. Unphosphorylated DUE-B HMW complex formation is decreased by PP2A inhibition or direct DUE-B phosphorylation, and increased by inhibition of Cdc7. These results indicate that the state of DUE-B phosphorylation is maintained by the equilibrium between Cdc7-dependent phosphorylation and PP2A-dependent dephosphorylation, each previously shown to regulate replication initiation. Alanine mutation of the DUE-B C-terminal phosphorylation target sites increases MCM binding but blocks Cdc45 loading in vivo and inhibits cell division. In egg extracts alanine mutation of the DUE-B C-terminal phosphorylation sites blocks Cdc45 loading and inhibits DNA replication. The effects of DUE-B C-terminal phosphorylation reveal a novel S phase kinase regulatory mechanism for Cdc45 loading and MCM helicase activation.

Activation of a human chromosomal replication origin by protein tethering.

The specification of mammalian chromosomal replication origins is incompletely understood. To analyze the assembly and activation of prereplicative complexes (pre-RCs), we tested the effects of tethered binding of chromatin acetyltransferases and replication proteins on chromosomal c-myc origin deletion mutants containing a GAL4-binding cassette. GAL4(DBD) (DNA binding domain) fusions with Orc2, Cdt1, E2F1 or HBO1 coordinated the recruitment of the Mcm7 helicase subunit, the DNA unwinding element (DUE)-binding protein DUE-B and the minichromosome maintenance (MCM) helicase activator Cdc45 to the replicator, and restored origin activity. In contrast, replication protein binding and origin activity were not stimulated by fusion protein binding in the absence of flanking c-myc DNA. Substitution of the GAL4-binding site for the c-myc replicator DUE allowed Orc2 and Mcm7 binding, but eliminated origin activity, indicating that the DUE is essential for pre-RC activation. Additionally, tethering of DUE-B was not sufficient to recruit Cdc45 or activate pre-RCs formed in the absence of a DUE. These results show directly in a chromosomal background that chromatin acetylation, Orc2 or Cdt1 suffice to recruit all downstream replication initiation activities to a prospective origin, and that chromosomal origin activity requires singular DNA sequences.

Microsatellite break-induced replication generates highly mutagenized extrachromosomal circular DNAs.

Extrachromosomal circular DNAs (eccDNAs) are produced from all regions of the eucaryotic genome. We used inverse PCR of non-B microsatellites capable of forming hairpin, triplex, quadruplex and AT-rich structures integrated at a common ectopic chromosomal site to show that these non-B DNAs generate highly mutagenized eccDNAs by replication-dependent mechanisms. Mutagenesis occurs within the non-B DNAs and extends several kilobases bidirectionally into flanking and nonallelic DNA. Each non-B DNA exhibits a different pattern of mutagenesis, while sister clones containing the same non-B DNA also display distinct patterns of recombination, microhomology-mediated template switching and base substitutions. Mutations include mismatches, short duplications, long nontemplated insertions, large deletions and template switches to sister chromatids and nonallelic chromosomes. Drug-induced replication stress or the depletion of DNA repair factors Rad51, the COPS2 signalosome subunit or POLη change the pattern of template switching and alter the eccDNA mutagenic profiles. We propose an asynchronous capture model based on break-induced replication from microsatellite-induced DNA double strand breaks to account for the generation and circularization of mutagenized eccDNAs and the appearance of genomic homologous recombination deficiency (HRD) scars. These results may help to explain the appearance of tumor eccDNAS and their roles in neoantigen production, oncogenesis and resistance to chemotherapy.

Microsatellite break-induced replication generates highly mutagenized extrachromosomal circular DNAs.

Extrachromosomal circular DNAs (eccDNAs) are produced from all regions of the eucaryotic genome. In tumors, highly transcribed eccDNAs have been implicated in oncogenesis, neoantigen production and resistance to chemotherapy. Here we show that unstable microsatellites capable of forming hairpin, triplex, quadruplex and AT-rich structures generate eccDNAs when integrated at a common ectopic site in human cells. These non-B DNA prone microsatellites form eccDNAs by replication-dependent mechanisms. The microsatellite-based eccDNAs are highly mutagenized and display template switches to sister chromatids and to nonallelic chromosomal sites. High frequency mutagenesis occurs within the eccDNA microsatellites and extends bidirectionally for several kilobases into flanking DNA and nonallelic DNA. Mutations include mismatches, short duplications, longer nontemplated insertions and large deletions. Template switching leads to recurrent deletions and recombination domains within the eccDNAs. Template switching events are microhomology-mediated, but do not occur at all potential sites of complementarity. Each microsatellite exhibits a distinct pattern of recombination, microhomology choice and base substitution signature. Depletion of Rad51, the COPS2 signalosome subunit or POLη alter the eccDNA mutagenic profiles. We propose an asynchronous capture model based on break-induced replication from microsatellite-induced DNA breaks for the generation and circularization of mutagenized eccDNAs and genomic homologous recombination deficiency (HRD) scars.

Replication fork stalling and checkpoint activation by a PKD1 locus mirror repeat polypurine-polypyrimidine (Pu-Py) tract.

DNA sequences prone to forming noncanonical structures (hairpins, triplexes, G-quadruplexes) cause DNA replication fork stalling, activate DNA damage responses, and represent hotspots of genomic instability associated with human disease. The 88-bp asymmetric polypurine-polypyrimidine (Pu-Py) mirror repeat tract from the human polycystic kidney disease (PKD1) intron 21 forms non-B DNA secondary structures in vitro. We show that the PKD1 mirror repeat also causes orientation-dependent fork stalling during replication in vitro and in vivo. When integrated alongside the c-myc replicator at an ectopic chromosomal site in the HeLa genome, the Pu-Py mirror repeat tract elicits a polar replication fork barrier. Increased replication protein A (RPA), Rad9, and ataxia telangiectasia- and Rad3-related (ATR) checkpoint protein binding near the mirror repeat sequence suggests that the DNA damage response is activated upon replication fork stalling. Moreover, the proximal c-myc origin of replication was not required to cause orientation-dependent checkpoint activation. Cells expressing the replication fork barrier display constitutive Chk1 phosphorylation and continued growth, i.e. checkpoint adaptation. Excision of the Pu-Py mirror repeat tract abrogates the DNA damage response. Adaptation to Chk1 phosphorylation in cells expressing the replication fork barrier may allow the accumulation of mutations that would otherwise be remediated by the DNA damage response.

Suppressors of Break-Induced Replication in Human Cells.

Short tandem DNA repeats are drivers of genome instability. To identify suppressors of break-induced mutagenesis human cells, unbiased genetic screens were conducted using a lentiviral shRNA library. The recipient cells possessed fragile non-B DNA that could induce DNA double-strand breaks (DSBs), integrated at an ectopic chromosomal site adjacent to a thymidine kinase marker gene. Mutagenesis of the thymidine kinase gene rendered cells resistant to the nucleoside analog ganciclovir (GCV). The screen identified genes that have established roles in DNA replication and repair, chromatin modification, responses to ionizing radiation, and genes encoding proteins enriched at replication forks. Novel loci implicated in BIR included olfactory receptors, the G0S2 oncogene/tumor suppressor axis, the EIF3H-METTL3 translational regulator, and the SUDS3 subunit of the Sin3A corepressor. Consistent with a role in suppressing BIR, siRNA knockdown of selected candidates increased the frequency of the GCVr phenotype and increased DNA rearrangements near the ectopic non-B DNA. Inverse PCR and DNA sequence analyses showed that hits identified in the screen increased genome instability. Further analysis quantitated repeat-induced hypermutagenesis at the ectopic site and showed that knockdown of a primary hit, COPS2, induced mutagenic hotspots, remodeled the replication fork, and increased nonallelic chromosome template switches.

Replication-dependent instability at (CTG) x (CAG) repeat hairpins in human cells.

Instability of (CTG) x (CAG) microsatellite trinucleotide repeat (TNR) sequences is responsible for more than a dozen neurological or neuromuscular diseases. TNR instability during DNA synthesis is thought to involve slipped-strand or hairpin structures in template or nascent DNA strands, although direct evidence for hairpin formation in human cells is lacking. We have used targeted recombination to create a series of isogenic HeLa cell lines in which (CTG) x (CAG) repeats are replicated from an ectopic copy of the Myc (also known as c-myc) replication origin. In this system, the tendency of chromosomal (CTG) x (CAG) tracts to expand or contract was affected by origin location and the leading or lagging strand replication orientation of the repeats, and instability was enhanced by prolonged cell culture, increased TNR length and replication inhibition. Hairpin cleavage by synthetic zinc finger nucleases in these cells has provided the first direct evidence for the formation of hairpin structures during replication in vivo.

Treslin, DUE-B, and GEMC1 cannot complement Sld3 mutants in fission yeast.

Initiation of DNA replication in eukaryotes is an evolutionarily conserved process that involves two distinct steps: the formation of prereplication complexes at replication origins in G1 and the assembly of preinitiation complexes (pre-ICs) in S phase, which leads to activation of the replication helicase. For the assembly of pre-ICs in yeast, formation of the Sld2-Dpb11-Sld3 complex is a critical event that requires phosphorylation of Sld2 and Sld3 by cyclin-dependent kinase. In mammals, RecQL4 and TopBP1 are excellent ortholog candidates for Sld2 and Dpb11, respectively. In this past year, three TopBP1-interacting proteins Treslin/Ticrr, GEMC1, and DUE-B have been identified in metazoans as possible functional orthologs of the yeast Sld3. To test this hypothesis, we carried out several complementation tests in fission yeast. The proteins were expressed at various levels in the temperature-sensitive sld3-10 mutant and in cells that lack endogenous Sld3. Our result showed that none of these metazoan proteins could rescue growth defect of the sld3 mutants. Although the result may have several interpretations, it is possible that the helicase activation in mammals has diverged in complexity during evolution from that in yeasts and may involve multiple players that interact with TopBP1.

Analysis of Trinucleotide Repeat Stability by Integration at a Chromosomal Ectopic Site.

Expansions of CNG microsatellite tracts are responsible for several neurodegenerative diseases, including myotonic dystrophy type 1, Huntington disease, and spinocerebellar ataxia type 8. Here we show that expanded (CNG)n repeats are susceptible not only to expansions and contractions, but are prone to DNA double strand breaks following replication stress. We describe a general strategy for the construction of clonal cell lines containing CNG repeats of various lengths, in which the microsatellites are integrated using the yeast FLP recombinase at a single ectopic recombination acceptor site in the HeLa genome. We illustrate two types of (CTG/CAG) cell lines, one of which contains dual fluorescent marker genes flanking the (CTG/CAG) repeat, and one which does not. We show that long CNG repeats are prone to DNA double strand breaks (DSBs) upon exposure of these cell lines to prolonged replication stress.

Differential binding of replication proteins across the human c-myc replicator.

The binding of the prereplication complex proteins Orc1, Orc2, Mcm3, Mcm7, and Cdc6 and the novel DNA unwinding element (DUE) binding protein DUE-B to the endogenous human c-myc replicator was studied by chromatin immunoprecipitation. In G(1)-arrested HeLa cells, Mcm3, Mcm7, and DUE-B were prominent near the DUE, while Orc1 and Orc2 were least abundant near the DUE and more abundant at flanking sites. Cdc6 binding mirrored that of Orc2 in G(1)-arrested cells but decreased in asynchronous or M-phase cells. Similarly, the signals from Orc1, Mcm3, and Mcm7 were at background levels in cells arrested in M phase, whereas Orc2 retained the distribution seen in G(1)-phase cells. Previously shown to cause histone hyperacetylation and delocalization of replication initiation, trichostatin A treatment of cells led to a parallel qualitative change in the distribution of Mcm3, but not Orc2, across the c-myc replicator. Orc2, Mcm3, and DUE-B were also bound at an ectopic c-myc replicator, where deletion of sequences essential for origin activity was associated with the loss of DUE-B binding or the alteration of chromatin structure and loss of Mcm3 binding. These results show that proteins implicated in replication initiation are selectively and differentially bound across the c-myc replicator, dependent on discrete structural elements in DNA or chromatin.