Transcription-replication conflicts: How they occur and how they are resolved

Nat Rev Mol Cell Biol Transcription-replication conflicts 1 Transcription--replication conflicts: how they occur and how they are resolved Tatiana García-Muse & Andrés Aguilera Centro Andaluz de Biología Molecular y Medicina Regenerativa-CABIMER, Universidad de Sevilla, Seville, Spain corresponding author: [email protected] Abstract The frequent occurrence in cells of transcription and DNA replication results in many encounters and thus conflicts between the transcription and replication machineries. These conflicts constitute a major intrinsic source of genome instability, which is a hallmark of cancer cells. How the replication machinery progresses through a DNA occupied by an RNA polymerase is an old question. Here we review recent data on the biological relevance of transcription-replication conflicts and the factors and mechanisms involved in either preventing or resolving them in eukaryotes. With this we aim to provide our current view of how transcription can generate obstacles to replication, including torsional stress and non-B DNA structures, and of the different cellular processes that have evolved to solve them. Introduction Genomes are templates for multiple biological processes, including transcription, epigenetic modifications, DNA replication, DNA repair and chromosome segregation. In a number of cases, crosstalk between different processes occurring at the DNA may have a positive effect, as in the case of transcription-coupled repair [G] 1. However, in Nat Rev Mol Cell Biol Transcription-replication conflicts 2 other cases, the co-temporal activity of two cellular machineries at the same genomic region may cause a conflict with negative consequences. This is the case of DNA replication and transcription. Research in the past two decades has provided evidence that transcription and replication conflicts constitute a considerable natural intrinsic source of genome instability, which is a hallmark of cancer cells 2. Given that transcription and replication are two essential processes for cell viability and proliferationfunction and that they occur frequently, a high incidence of encounters between the transcription and replication machineries is to be expected. Although transcription can have a positive role effect on replication initiation through transcription-mediated chromatin changes that may facilitate firing of origins of DNA replication origins 3, collisions are a potential threat to genome integrity and cell viability. How the replication machinery progresses through a double-stranded DNA occupied by an RNA polymerase is an old question. Alberts and colleagues elegantly addressed this question using the T4 bacteriophage system in vitro 4; but our actual knowledge of the RNA polymerase structure and mechanisms of transcription elongation suggests that the factors and mechanisms used by cells to solve such conflicts are more complex than previously foreseen. The relevance of transcription as a source of genome instability, as measured by the rate of point mutations or of recombination and chromosome rearrangements, and the putative mechanisms by which such instability is mediated have been reviewed recently 5-8. Here we review recent data on the factors and mechanisms involved in either preventing or resolving transcription--replication collisions, and on their potential consequences. In particular, we discuss how transcription may hinder the progression of the replication forks itself or how transcription activity generates obstacles to replication, including torsional stress [G] and non-B DNA [G] structures, and the different solutions the cells have evolved to avoid, minimize or resolve these collisions or their consequences. [H1] How do collisions occur? A basic difference between the transcription and replication machineries is that the elongating RNA polymerase holoenzyme, which comprises one polymerase subunit, embraces the double-stranded DNA. The nascent RNA chain is synthesized in the active pocket of the RNA polymerase, where it forms a dynamic 9-11-nt RNA-DNA hybrid (Figure 1A). The elongating DNA polymerase holoenzyme, on the other hand, consists of two polymerase subunits (of DNA pol III in bacteria; DNA pol epsilon and Nat Rev Mol Cell Biol Transcription-replication conflicts 3 delta in eukaryotes), each working on a single-stranded DNA (ssDNA) template (Figure 1B). Furthermore, whereas several active RNA polymerases can simultaneously transcribe the same gene, replisomes [G] move alone and are not followed by a second replication fork. No matter whether collisions are co-directional or in head-on orientation (Figure 2), the replication fork cannot go through an elongating RNA polymerase and so their encounters will cause conflicts 9, 10. Although replication fork progression may be affected by collisions in both orientations, data suggest that the consequences of collisions are more dramatic in the head-on orientation 11, 12. When encounters were promoted in yeast artificial systems in a head-on orientation, replication pause sites were detected by two-dimensional (2D)-gel electrophoresis and hyper-recombination was observed 12. By contrast, co-directional orientation did not lead to replication pauses or high levels of hyper-recombination. This difference can be explained if co-directional encounters may in part be resolved once the RNA polymerase terminates transcription. Cells have developed different strategies to reduce or prevent collisions. In bacteria, there is a genome-wide bias towards co-orientation of replication and transcription 13 and inverting transcriptional units to provoke head-on collisions causes replication impairment, proliferation defects and genome instability 11. In eukaryotes a bias towards co-directional replication and transcription is not obvious, but cells seem to have evolved other strategies to reduce head-on collisions. For example, in the Saccharomyces cerevisiae highly transcribed ribosomal DNA (rDNA) genes, replication fork blocking (RFB) sites exist that block fork progression and prevent harmful encounters with RNA polymerases 14. In the mammalian rDNA loci replication and the transcription seem to be efficiently spatially separated in the nucleoli as a way to avoid collisions 15. In other regions of the genome, transcription and replication seem to be separated temporaly 16. Analysis of nascent mRNAs in genes encoding replication factors revealed that active genes transcribed during early replication are replicated late in S-phase and vice versa 17. Importantly, however, it is not clear whether the RNA and DNA polymerases ever actually make contact. It is plausible that before the physical connection occurs, transcriptionand replication-mediated changes in chromatin and DNA structures attenuate the progression of the polymerases. To fully understand conflicts we therefore need to identify the elements and conditions that affect their occurrence. [H1] Cis-elements affecting collisions Nat Rev Mol Cell Biol Transcription-replication conflicts 4 The transcription machinery may constitute a natural obstacle to replication fork progression, but this interference can be direct or indirect, since the transcription process may also generate structural features that have the capacity to hinder replication fork progression, like changes in DNA supercoiling or secondary DNA structures such as hairpins[G], triplex DNA (H-DNA) [G], G-quadruplexes[G] or RNADNA hybrids. [H3] DNA supercoiling Transcription and replication require the unwinding of the DNA molecule. This unwinding leads to positive and negative supercoiling ahead and behind the RNA polymerase, respectively (Figure 3A). The resulting torsional stress [G] is relieved by DNA topoisomerases, which are of Type I or Type II depending on whether they catalyse breakage of one or both DNA strands, respectively. In budding yeast topoisomerase mutants accumulate supercoiling and this torsional stress prevents both transcription and replication of the highly transcribed rDNA 18, suggesting that supercoiling can cause transcription and replication block. Later studies in yeast and human cells have shown that both topoisomerase 1 (human TOP1, yeast Top1) and topoisomerase 2 (human TOP2, yeast Top2) are crucial to prevent transcriptionreplication collisions 19, 20, indicating that unresolved torsional stress can attenuate the progression of both DNA and RNA polymerases and promote transcription-replication conflicts. Genome-wide analysis of Top1 and Top2 distribution in replicating budding yeast cells revealed an association of these enzymes with moving replication forks 19, 21. Moreover, top1 top2 double mutant cells also accumulate DNA damage19. Consistently, analyses of replication by DNA combing [G] have revealed that replication forks are slower in yeast and human Top1-deficient cells 20. Interestingly, in TOP1-depleted human cells there was an increase in fork stalling that correlated with the accumulation of γ-H2AX foci [G] in S-phase and that was suppressed by inhibition of transcription elongation with cordycepin. Therefore, TOP1 activity can prevent transcription-replication conflicts and their harmful consequences 20. Based on these results we suggest that DNA supercoiling is transiently accumulating between the advancing transcription and replication machineries and may be important in the control of their collisions and their adverse effects (Figure 3A). In theory, this phenomenon should be exacerbated in the case of the head-on orientation. However, convergent transcription, which would create the same topological constraint as a head-on collisions, does not pose a major threat to genome integrity and transcription in budding yeast 22. Alternatively, inefficient resolution of the negative supercoiling accumulating behind the elongating RNA polymerase may also Nat Rev Mol Cell Biol Transcription-replication conflicts 5 facilitate local melting of the DNA duplex and, consequently, formation of non-B DNA structures that can block replication fork progression (see below). Indeed, in yeast, divergent transcription was found to enhance chromosome rearrangements 23. [H3] non-B DNA structures and RNA-DNA hybrids Some DNA sequences, especially repetitive sequences, can assume non-B DNA structures such as hairpins, triplex DNA (H-DNA) or G-quadruplexes (also termed Gquartets). Such non-B DNA structures have the capacity to stall replication forks and have been correlated with hotspots of DNA double-strand breaks (DSBs) and chromosomal deletions, translocations and other rearrangements 24. These secondary DNA structures are believed to form preferentially at the ssDNA that is exposed during DNA replication, but they can also be formed during transcription, favoured by the negative supercoiling that is transiently accumulated behind the elongating RNA polymerase (Figure 3). A good example of the putative relevance of non-B structures is provided by G-quadruplexes, which consist of four repeats of at least three guanines that can form four strand-interactions (Figure 3B). G-quadruplexes can form during lagging-strand replication as shown at telomeres 25, as well as during transcription: human cells treated with the G-quadruplex ligand pyridostatin show a tight correlation between pyridostatin binding and γH2AX foci [G] formation, which is reduced by treatment with the transcription inhibitor 5,6-dichloro-1-β-D-ribofuranosylbenzimidazole (DRB) 26. The idea that G-quadruplex formation can be potentiated behind an elongating RNA Polymerase II (Pol II) has been indirectly inferred in yeast expressing the murine G-quadruplex-prone Sμ Ig switch region, which stimulates recombination in combination with conditions of high transcription levels27. The activity of Top1 in these conditions suppresses G-quadruplex-associated recombination, consistent with negative supercoiling enhancing G-quadruplex accumulation 28. It appears that the genomic instability is higher when the orientation of the G-rich strand of the Sµ sequence, with respect to transcription, leaves the G-rich strand in the non-transcribed strand, suggesting that the ssDNA that allows quadruplex formation originates from transcription. Additional support to the idea that non-B DNA structure may contribute to transcription-mediated replication fork stalling comes from studies in mutants of the budding yeast DNA helicase Pif1 (petite integration frequency) and the fission yeast Pfh1 (Pif1-homolog), which unwind G-quadruplexes in vitro. Absence of Pif1 or Pfh1 attenuates or halts replication in regions of high G-quadruplex density and in RNA Pol IIand Pol III-highly transcribed genes 29, 30. Nat Rev Mol Cell Biol Transcription-replication conflicts 6 Another type of transcription-mediated structures able to cause fork stalling that can strongly contribute to transcription-replication conflicts are co-transcriptional RNADNA hybrids (also known as R loops when formed outside of the transcription bubble; Figure 3C). Although they are natural intermediates in class switching recombination and in initiation of mitochondrial DNA replication, RNA-DNA hybrids are formed in conditions that prevent the proper formation of the ribonucleoprotein particle, as shown in yeast and human cells 31, 32. Evidence from yeast to mammalian cells suggest that RNA-DNA hybrids can form naturally and may constitute an important transcription intermediate that can provoke replication fork stalling at telomeres, the rDNA regions, CpG islands [G] and other sites at specific Pol II-transcribed genes, including 3’-end regions. RNA-DNA hybrids have been thoroughly and extensively reviewed recently 3336 and will not be discussed further here. It is important to note, however, that an enrichment of sequences with high probability of forming non-B DNA structures or RNA-DNA hybrids and to undergo transcription-replication conflicts are observed at some fragile sites (Box 1). [H1] Mechanisms for preventing conflicts Owing to the impact of transcription-replication collisions on genomic stability and thereby potentially on disease, cells have evolved mechanisms to prevent such encounters. The factors that minimize collisions include the transcription machinery itself, as well as factors that help or facilitate replication progression through transcribed DNA. [H3] The RNA polymerase transcription apparatus Some clues to understand how the RNA polymerase directly contribute to transcriptionreplication conflicts are starting to emerge from the analysis of several RNA polymerase mutants in bacteria and yeast, but we are still far from having a complete view. A critical step in transcription that seems to be relevant to conflicts is RNA polymerase ‘backtracking’, which refers to the process by which the RNA polymerase reverses its progression to enzymatically remove the last incorporated ribonucleotide. This allows restarting transcription elongation following a pause provoked by hindrances during transcription elongation or as part of a regulatory process to coordinate the different steps of transcription and RNA processing 37. A backtracked RNA polymerase is able to block replication progression in Escherichia coli. Using specific promoter sequences that allow modulation of the polymerase activity it was shown that a permanently-arrested elongating polymerase causes DSBs 38, 39. Such Nat Rev Mol Cell Biol Transcription-replication conflicts 7 breaks were inferred to be replication-dependent since treatment with the replication inhibitor hydroxyurea prior to transcription activation avoided their appearance. The clash between the replisome and the backtracked RNA polymerase was interpreted to be responsible for the formation of DSBs 39. Consistent with this view, the E. coli transcription elongation factors GreA and GreB , which promote the release of backtracked and stalled RNA polymerases, seem to reduce the consequences of conflicts (Figure 4A). In the absence of GreA and GreB and under substantial transcription activity induced by starvation, replication progression was completely blocked 40. A similar role was proposed for the yeast transcription elongation factor TFIIS 41, but it remains to be seen if this activity has any effect on putative collisions. Direct involvement of the transcription apparatus in modulating transcriptionreplication conflicts was demonstrated in recent studies utilizing RNA polymerase mutants that compromise the stability of transcription complexes. These RNA polymerase mutants were shown to suppress growth defects of E. coli cells lacking factors that help resolve collisions such as the DNA helicases Rep (Replicase), UvrD (uv resistant protein D) and DinG (damage-inducible protein G) 42. These results suggest that less stable transcription complexes may not compromise replication progression since they do not seem to form strong replication obstacles 42. Also, several yeast RNA Pol II mutants with transcription elongation defects exhibited replication impairment, inferred by 2D-gels, bromodeoxyuridine [G] incorporation by DNA polymerases or by altered distribution of Rrm3 (rDNA recombination mutation protein 3), which is a replicative helicase required for replication progression through DNA obstacles 43. It is likely that following a collision the RNA Pol II is released from the DNA to allow passage of the replisome, as is the case in bacteria 38. Interestingly, one of these yeast RNA Pol II mutants, the yeast rpb1-1 mutant (of the largest RNA Pol II subunit), has tighter attachment to chromatin than wild-type RNA Pol II, as determined by chromatin immunoprecipitation, supporting the idea that RNA Pol II mutants with increased attachment to chromatin could aggravate the consequences of a transcription-replication encounter 43. These results suggest that the transcription machinery, and RNA Pol II itself, may participate in managing transcription-replication conflicts through the feasibility of their eviction from DNA following a collision. The recent observation that PAF1C (RNA polymerase II-associated factor 1 complex) triggers RNA Pol II degradation at sites of collisions 44 supports this view. [H3] Replication fork barriers Nat Rev Mol Cell Biol Transcription-replication conflicts 8 Replication forks have to deal along their path with non-nucleosomal protein–DNA complexes that assemble at genes and regulatory elements. In bacteria the barrier formed by the transcription complex is able to pause replication forks, and resumption of replication requires specific DNA helicases 45, 46. In yeast, different regions that impede replication fork progression in vivo have been identified, the most representative being the fork barrier found in the 35S rRNA gene in the rDNA 47. The rDNA region provides the best model to study the impact of replication stress generated by transcription owing to its high transcription rate and high density of replication origins. The replication barrier consists of DNA replication fork-blocking protein Fob1 bound to the specific RFB sequence, which prevents head-on collisions between RNA and DNA polymerases (Figure 4B). Interestingly, replication fork progression through the RFB–Fob1 complex requires the helicase Rrm3. Deletion of the rrm3 gene (rrm3∆) increases replication pauses at rDNA, resulting in breakage and accumulation of excised rDNA circles 14, 48. Although Rrm3 could be seen therefore as a complementary factor acting in trans to promote replication fork passage through protein barriers, fork pausing in rrm3∆ mutants is also increased in other pause sites such as in tRNA genes or telomeres 49, 50. Notably, though, other pause sites are found at highly transcribed genes, but are not exacerbated in rrm3∆ mutants, suggesting that other factors may have roles in the prevention or resolution of collisions 9. [H3] The RecQL5 DNA helicase Human RecQ-like ATP-dependent DNA helicase Q5 (RECQL5), a member of the RecQ family [G] of DNA helicases, is to date the protein with the best-characterised active role in preventing transcription-replication collisions. RECQL5 forms a stable complex with RNA Pol II and several in vivo and in vitro studies indicated it has a negative regulatory role in transcription elongation 51. ChIP-seq analysis with an RNA Pol II antibody revealed that transcription up-regulation in cells lacking RECQL5 increases transcription pausing, arrest and backtracking, suggesting that uncontrolled and high transcription rates lead to transcriptional stress 52. RECQL5 associates with the replicative DNA sliding clamp PCNA (proliferating cell nuclear antigen), which suggests that RECQL5 is also involved in replication. Accordingly, RECQL5-deficient cells fail to incorporate bromodeoxyuridine in conditions of replication stress and rapidly accumulate DNA damage, effects which can be alleviated by fully arresting replication with the replication inhibitor aphidicolin 53. Importantly, spontaneous DSBs accumulate in RECQL5-depleted cells during replication, but only in association with RNA Pol II transcription, as the spontaneous breaks are located in transcribed genes Nat Rev Mol Cell Biol Transcription-replication conflicts 9 and transcription inhibition eliminated their appearance 54. Furthermore, RECQL5 has been shown to have a role in suppressing genome rearrangements associated preferentially with common fragile sites (Box 1) and transcribed genes 52. These data have led to the proposal that RECQL5 prevents transcription-replication collisions. Recently, RECQL5 was also ascribed a role in preventing the formation of RNA-DNA hybrids. Apparently, RECQ5 promotes TOP1 SUMOylation by facilitating the interaction with the PIAS1-SRSF1 E3 ligase complex. This modification is necessary for the binding of TOP1 to RNA Pol II and for the efficient recruitment of mRNA processing factors to transcriptionally active sites, thereby reducing the formation of RNA-DNA hybrids, as inferred by the increased levels of such hybrids in cells defective in RECQ5-dependent TOP1 SUMOylation 55. Therefore, RecQL5 may maintain genome integrity by actively limiting the occurrence of transcription-replication conflicts and/or by reducing the accumulation of non-B DNA structures generated during transcription that could enhance replication blockage (Figure 3). [H3] Chromatin remodelling In eukaryotes, transcription and replication occur in the context of highly structured chromatin. Following replication the chromatin state is maintained by coupling the deposition of recycled parental histones with newly-synthesized histones on the duplicated DNA, which is carried out by histone chaperones and chromatin remodelling complexes. Even though not much work has been done on the role of chromatin remodelling in diminishing transcription-replication conflicts, evidence exists about the importance of such factors on collisions. This has been clearly shown for the histone chaperone FACT (facilitates chromatin transcription) complex, which was initially found to be required for chromatin remodelling during transcription but is also involved in DNA replication 56-58. Yeast and human cells lacking FACT complex activity have high levels of transcription–replication collisions, exhibiting fork progression impairment that correlates with increased genomic instability. However, when transcription was inhibited with cordycepin in FACT-depleted human cells, the rate of fork progression was restored to normal 59. Therefore, chromatin-reorganizing factors, such as FACT, can prevent collisions by promoting the replication fork progression through transcribed regions (Figure 4C). It would certainly be interesting to see whether this view would also apply to the INO80 remodelling complex since it was shown in E. colibudding yeast that Ino80 (inositol requiring mutant 80) cooperates with the transcription factor PAF1C to trigger RNA Pol II degradation at sites of collisions 44. Nat Rev Mol Cell Biol Transcription-replication conflicts 16 This works shows that co-directional conflicts at highly transcribed rRNA genes can stall replication in vivo in B. subtilis. 11. Srivatsan, A., Tehranchi, A., MacAlpine, D.M. & Wang, J.D. Co-orientation of replication and transcription preserves genome integrity. PLoS Genet. 6, e1000810 (2010). 12. Prado, F. & Aguilera, A. Impairment of replication fork progression mediates RNA pol II transcription-associated recombination. EMBO J. 24, 1267-1276 (2005). This study demonstrates that head-on transcription and replication encounters cause replication fork impairment that leads to genome instability. 13. Merrikh, H., Zhang, Y., Grossman, A.D. & Wang, J.D. Replication-transcription conflicts in bacteria. Nature Rev. Microbiol. 10, 449-458 (2012). 14. Ivessa, A.S., Zhou, J.Q. & Zakian, V.A. The Saccharomyces Pif1p DNA helicase and the highly related Rrm3p have opposite effects on replication fork progression in ribosomal DNA. Cell 100, 479-489 (2000). 15. Smirnov, E. et al. Separation of replication and transcription domains in nucleoli. J. Struct. Biol. 188, 259-266 (2014). 16. Gilbert, D.M. Replication timing and transcriptional control: beyond cause and effect. Curr. Opin. Cell Biol. 14, 377-383 (2002). 17. Meryet-Figuiere, M. et al. Temporal separation of replication and transcription during S-phase progression. Cell Cycle 13, 3241-3248 (2014). 18. Brill, S.J., DiNardo, S., Voelkel-Meiman, K. & Sternglanz, R. Need for DNA topoisomerase activity as a swivel for DNA replication for transcription of ribosomal RNA. Nature 326, 414-416 (1987). 19. Bermejo, R. et al. Top1and Top2-mediated topological transitions at replication forks ensure fork progression and stability and prevent DNA damage checkpoint activation. Genes Dev. 21, 1921-1936 (2007). 20. Tuduri, S. et al. Topoisomerase I suppresses genomic instability by preventing interference between replication and transcription. Nature Cell Biol. 11, 13151324 (2009). This genome-wide analysis of TOP1 deficient cells shows correlation between fork stalling and DNA breaks at transcribed genes, suggesting that TOP1 reduces R-loops-dependent transcription–replication conflicts. 21. Bermejo, R. et al. Genome-organizing factors Top2 and Hmo1 prevent chromosome fragility at sites of S phase transcription. Cell 138, 870-884 (2009). 22. García-Rubio, M.L. & Aguilera, A. Topological constraints impair RNA polymerase II transcription and causes instability of plasmid-borne convergent genes. Nucleic Acids Res. 40, 1050-1064 (2012). 23. Pannunzio, N.R. & Lieber, M.R. Dissecting the roles of divergent and convergent transcription in chromosome instability. Cell Rep. 14, 1025-1031 (2016). 24. Zhao, J., Bacolla, A., Wang, G. & Vasquez, K.M. Non-B DNA structure-induced genetic instability and evolution. Cell Mol. Life Sci. 67, 43-62 (2010). 25. Audry, J. et al. RPA prevents G-rich structure formation at lagging-strand telomeres to allow maintenance of chromosome ends. EMBO J. 34, 1942-1958 (2015). 26. Rodriguez, R. et al. Small-molecule-induced DNA damage identifies alternative DNA structures in human genes. Nature Chem. Biol. 8, 301-310 (2012). Nat Rev Mol Cell Biol Transcription-replication conflicts 17 27. Kim, N. & Jinks-Robertson, S. Guanine repeat-containing sequences confer transcription-dependent instability in an orientation-specific manner in yeast. DNA Repair (Amst) 10, 953-960 (2011). 28. Yadav, P. et al. Topoisomerase I plays a critical role in suppressing genome instability at a highly transcribed G-quadruplex-forming sequence. PLoS Genet. 10, e1004839 (2014). 29. Sabouri, N., McDonald, K.R., Webb, C.J., Cristea, I.M. & Zakian, V.A. DNA replication through hard-to-replicate sites, including both highly transcribed RNA Pol II and Pol III genes, requires the S. pombe Pfh1 helicase. Genes Dev. 26, 581-593 (2012). 30. Paeschke, K., Capra, J.A. & Zakian, V.A. DNA replication through G-quadruplex motifs is promoted by the Saccharomyces cerevisiae Pif1 DNA helicase. Cell 145, 678-691 (2011). 31. Huertas, P. & Aguilera, A. Cotranscriptionally formed DNA:RNA hybrids mediate transcription elongation impairment and transcription-associated recombination. Mol. Cell 12, 711-721 (2003). 32. Li, X. & Manley, J.L. Inactivation of the SR protein splicing factor ASF/SF2 results in genomic instability. Cell 122, 365-378 (2005). 33. Santos-Pereira, J.M. & Aguilera, A. R loops: new modulators of genome dynamics and function. Nature Rev. Genet. 16, 583-597 (2015). 34. Skourti-Stathaki, K. & Proudfoot, N.J. A double-edged sword: R loops as threats to genome integrity and powerful regulators of gene expression. Genes Dev. 28, 1384-1396 (2014). 35. Sollier, J. & Cimprich, K.A. Breaking bad: R-loops and genome integrity. Trends Cell Biol. 25, 514-522 (2015). 36. Costantino, L. & Koshland, D. The Yin and Yang of R-loop biology. Curr. Opin. Cell Biol. 34, 39-45 (2015). 37. Nudler, E. RNA polymerase backtracking in gene regulation and genome instability. Cell 149, 1438-1445 (2012). 38. Pomerantz, R.T. & O'Donnell, M. Direct restart of a replication fork stalled by a head-on RNA polymerase. Science 327, 590-592 (2010). 39. Dutta, D., Shatalin, K., Epshtein, V., Gottesman, M.E. & Nudler, E. Linking RNA polymerase backtracking to genome instability in E. coli. Cell 146, 533-543 (2011). 40. Tehranchi, A.K. et al. The transcription factor DksA prevents conflicts between DNA replication and transcription machinery. Cell 141, 595-605 (2010). This work illustrates the function of RNA polymerases, transcription factors and DNA repair proteins in preventing and/or resolving transcription–replication conflicts. 41. Dutta, A. et al. Ccr4-Not and TFIIS function cooperatively to rescue arrested RNA polymerase II. Mol. Cell Biol. 35, 1915-1925 (2015). 42. Baharoglu, Z., Lestini, R., Duigou, S. & Michel, B. RNA polymerase mutations that facilitate replication progression in the rep uvrD recF mutant lacking two accessory replicative helicases. Mol. Microbiol. 77, 324-336 (2010). 43. Felipe-Abrio, I., Lafuente-Barquero, J., García-Rubio, M.L. & Aguilera, A. RNA polymerase II contributes to preventing transcription-mediated replication fork stalls. EMBO J. 34, 236-250 (2015). 44. Poli, J. et al. Mec1, INO80, and the PAF1 complex cooperate to limit transcription replication conflicts through RNAPII removal during replication stress. Genes Dev. 30, 337-354 (2016). 45. Merrikh, C.N., Brewer, B.J. & Merrikh, H. The B. subtilis accessory helicase PcrA facilitates DNA replication through transcription units. PLoS Genet. 11, e1005289 (2015). Nat Rev Mol Cell Biol Transcription-replication conflicts 18 46. Gupta, M.K. et al. Protein-DNA complexes are the primary sources of replication fork pausing in Escherichia coli. Proc. Natl. Acad. Sci. 110, 72527257 (2013). 47. Brewer, B.J. & Fangman, W.L. A replication fork barrier at the 3' end of yeast ribosomal RNA genes. Cell 55, 637-43 (1988). 48. Torres, J.Z., Bessler, J.B. & Zakian, V.A. Local chromatin structure at the ribosomal DNA causes replication fork pausing and genome instability in the absence of the S. cerevisiae DNA helicase Rrm3p. Genes Dev. 18, 498-503 (2004). 49. Ivessa, A.S. et al. The Saccharomyces cerevisiae helicase Rrm3p facilitates replication past nonhistone protein-DNA complexes. Mol. Cell 12, 1525-1536 (2003). 50. Ivessa, A.S., Zhou, J.-Q., Schulz, V.P., Monson, E.K. & Zakian, V.A. Saccharomyces Rrm3p, a 5' to 3' DNA helicase that promotes replication fork progression through telomeric and subtelomeric DNA. Genes Dev. 16, 13831396 (2002). 51. Popuri, V., Tadokoro, T., Croteau, D.L. & Bohr, V.A. Human RECQL5: guarding the crossroads of DNA replication and transcription and providing backup capability. Crit. Rev. Biochem. Mol. Biol. 48, 289-299 (2013). 52. Saponaro, M. et al. RECQL5 controls transcript elongation and suppresses genome instability associated with transcription stress. Cell 157, 1037-1049 (2014). This study shows that RECQL5 attenuates transcription elongation and suppresses genome rearrangements at common fragile sites, indicating that RECQL5 prevents transcription-replication conflicts. 53. Hu, Y., Lu, X., Zhou, G., Barnes, E.L. & Luo, G. Recql5 plays an important role in DNA replication and cell survival after camptothecin treatment. Mol. Biol. Cell 20, 114-123 (2009). 54. Li, M., Xu, X. & Liu, Y. The SET2-RPB1 interaction domain of human RECQ5 is important for transcription-associated genome stability. Mol. Cell. Biol. 31, 2090-2099 (2011). 55. Li, M., Pokharel, S., Wang, J.-T., Xu, X. & Liu, Y. RECQ5-dependent SUMOylation of DNA topoisomerase I prevents transcription-associated genome instability. Nature Commun. 6, 6720 (2015). 56. Orphanides, G., Wu, W.H., Lane, W.S., Hampsey, M. & Reinberg, D. The chromatin-specific transcription elongation factor FACT comprises human SPT16 and SSRP1 proteins. Nature 400, 284-288 (1999). 57. Foltman, M. et al. Eukaryotic replisome components cooperate to process histones during chromosome replication. Cell Rep. 3, 892-904 (2013). 58. Abe, T. et al. The histone chaperone facilitates chromatin transcription (FACT) protein maintains normal replication fork rates. J. Biol. Chem. 286, 3050430512 (2011). 59. Herrera-Moyano, E., Mergui, X., Garcia-Rubio, M.L., Barroso, S. & Aguilera, A. The yeast and human FACT chromatin-reorganizing complexes solve R-loopmediated transcription-replication conflicts. Genes Dev. 28, 735-748 (2014). This work provides compelling evidence of the role of chromatinreorganizing complex FACT in the resolution of transcription-replication conflicts. 60. Castellano-Pozo, M. et al. R loops are linked to histone H3 S10 phosphorylation and chromatin condensation. Mol. Cell 52, 583-590 (2013). Nat Rev Mol Cell Biol Transcription-replication conflicts 19 61. Groh, M., Lufino, M.M., Wade-Martins, R. & Gromak, N. R-loops associated with triplet repeat expansions promote gene silencing in Friedreich ataxia and fragile X syndrome. PLoS Genet. 10, e1004318 (2014). 62. Zaratiegui, M. et al. RNAi promotes heterochromatic silencing through replication-coupled release of RNA Pol II. Nature 479, 135-138 (2011). The authors show that at pericentromeric regions the co-transcriptional RNA interference machinery releases the RNA polymerase to avoid conflicts with the replication machinery. 63. Castel, S.E. et al. Dicer promotes transcription termination at sites of replication stress to maintain genome stability. Cell 159, 572-583 (2014). 64. Lopes, M., Foiani, M. & Sogo, J.M. Multiple mechanisms control chromosome integrity after replication fork uncoupling and restart at irreparable UV lesions. Mol. Cell 21, 15-27 (2006). 65. Heller, R.C. & Marians, K.J. Replication fork reactivation downstream of a blocked nascent leading strand. Nature 439, 557-562 (2006). 66. Deshpande, A.M. & Newlon, C.S. DNA replication fork pause sites dependent on transcription. Science 272, 1030-1033 (1996). 67. Nguyen, V.C. et al. Replication stress checkpoint signaling controls tRNA gene transcription. Nat. Struct. Mol. Biol. 17, 976-981 (2010). 68. Molla-Herman, A., Vallés, A.M., Ganem-Elbaz, C., Antoniewski, C. & Huynh, J.- R. tRNA processing defects induce replication stress and Chk2-dependent disruption of piRNA transcription. EMBO J. (2015). 69. Im, J.S. et al. ATR checkpoint kinase and CRL1betaTRCP collaborate to degrade ASF1a and thus repress genes overlapping with clusters of stalled replication forks. Genes Dev. 28, 875-887 (2014). 70. Yeo, C.Q. et al. p53 Maintains Genomic stability by preventing interference between transcription and replication. Cell Rep. 15, 132-146 (2016). 71. Cabal, G.G. et al. SAGA interacting factors confine sub-diffusion of transcribed genes to the nuclear envelope. Nature 441, 770-773 (2006). 72. Blobel, G. Gene gating: a hypothesis. Proc. Natl. Acad. Sci. USA 82, 8527-8529 (1985). 73. Bermejo, R. et al. The replication checkpoint protects fork stability by releasing transcribed genes from nuclear pores. Cell 146, 233-246 (2011). 74. Rossi, S.E., Ajazi, A., Carotenuto, W., Foiani, M. & Giannattasio, M. Rad53mediated regulation of Rrm3 and Pif1 DNA helicases contributes to prevention of aberrant fork transitions under replication stress. Cell Rep. 13, 80-92 (2015). 75. Bhatia, V. et al. BRCA2 prevents R-loop accumulation and associates with TREX-2 mRNA export factor PCID2. Nature 511, 362-365 (2014). 76. Hill, S.J. et al. Systematic screening reveals a role for BRCA1 in the response to transcription-associated DNA damage. Genes Dev. 28, 1957-1975 (2014). 77. Schlacher, K. et al. Double-strand break repair-independent role for BRCA2 in blocking stalled replication fork degradation by MRE11. Cell 145, 529-542 (2011). 78. Schwab, R.A. et al. The Fanconi anemia pathway maintains genome stability by coordinating replication and transcription. Mol. Cell 60, 351-361 (2015). 79. García-Rubio, M.L. et al. The Fanconi anemia pathway protects genome integrity from R-loops. PLoS Genet. 11, e1005674 (2015). References 78 and 79 provide evidence for a novel role of Fanconi anemia factors in preventing transcription-replication conflicts mediated by Rloops. Nat Rev Mol Cell Biol Transcription-replication conflicts 20 80. Hatchi, E. et al. BRCA1 recruitment to transcriptional pause sites is required for R-loop-driven DNA damage repair. Mol. Cell 57, 636-647 (2015). 81. Yüce, Ö. & West, S.C. Senataxin, defective in the neurodegenerative disorder ataxia with oculomotor apraxia 2, lies at the interface of transcription and the DNA damage response. Mol. Cell. Biol. 33, 406-417 (2013). 82. Alzu, A. et al. Senataxin associates with replication forks to protect fork integrity across RNA-polymerase-II-transcribed genes. Cell 151, 835-846 (2012). 83. Maldonado, E. et al. A human RNA polymerase II complex associated with SRB and DNA-repair proteins. Nature 381, 86-89 (1996). 84. Duch, A. et al. Coordinated control of replication and transcription by a SAPK protects genomic integrity. Nature 493, 116-119 (2013). 85. Alcasabas, A.A. et al. Mrc1 transduces signals of DNA replication stress to activate Rad53. Nature Cell Biol. 3, 958-965 (2001). 86. Glover, T.W., Berger, C., Coyle, J. & Echo, B. DNA polymerase alpha inhibition by aphidicolin induces gaps and breaks at common fragile sites in human chromosomes. Hum. Genet. 67, 136-142 (1984). 87. Zhang, H. & Freudenreich, C.H. An AT-rich sequence in human common fragile site FRA16D causes fork stalling and chromosome breakage in S. cerevisiae. Mol. Cell 27, 367-379 (2007). 88. Gerhardt, J. et al. The DNA replication program is altered at the FMR1 locus in fragile X embryonic stem cells. Mol. Cell 53, 19-31 (2014). 89. Letessier, A. et al. Cell-type-specific replication initiation programs set fragility of the FRA3B fragile site. Nature 470, 120-123 (2011). 90. Ozeri-Galai, E. et al. Failure of origin activation in response to fork stalling leads to chromosomal instability at fragile sites. Mol. Cell 43, 122-131 (2011). 91. Le Tallec, B. et al. Common fragile site profiling in epithelial and erythroid cells reveals that most recurrent cancer deletions lie in fragile sites hosting large genes. Cell Rep. 4, 420-428 (2013). 92. Wilson, T.E. et al. Large transcription units unify copy number variants and common fragile sites arising under replication stress. Genome Res. 25, 189-200 (2015). 93. Helmrich, A., Ballarino, M. & Tora, L. Collisions between replication and transcription complexes cause common fragile site instability at the longest human genes. Mol. Cell 44, 966-977 (2011). This work documents transcription-replication conflicts in very long human genes and shows that the instability of common fragile sites located within those genes is dependent on transcription and R-loop formation. 94. Barlow, J.H. et al. Identification of early replicating fragile sites that contribute to genome instability. Cell 152, 620-632 (2013). 95. Hoffman, E.A., McCulley, A., Haarer, B., Arnak, R. & Feng, W. Break-seq reveals hydroxyurea-induced chromosome fragility as a result of unscheduled conflict between DNA replication and transcription. Genome Res. 25, 402-412 (2015). 96. Halazonetis, T.D., Gorgoulis, V.G. & Bartek, J. An oncogene-induced DNA damage model for cancer development. Science 319, 1352-1355 (2008). 97. Miron, K., Golan-Lev, T., Dvir, R., Ben-David, E. & Kerem, B. Oncogenes create a unique landscape of fragile sites. Nature Commun. 6, 7094 (2015). 98. Bartkova, J. et al. DNA damage response as a candidate anti-cancer barrier in early human tumorigenesis. Nature 434, 864-870 (2005). 99. Valovka, T. et al. Transcriptional control of DNA replication licensing by Myc. Sci. Rep. 3, 3444 (2013). Nat Rev Mol Cell Biol Transcription-replication conflicts 21 100. Srinivasan, S.V., Dominguez-Sola, D., Wang, L.C., Hyrien, O. & Gautier, J. Cdc45 is a critical effector of myc-dependent DNA replication stress. Cell Rep. 3, 1629-1639 (2013). 101. Jones, R.M. et al. Increased replication initiation and conflicts with transcription underlie Cyclin E-induced replication stress. Oncogene 32, 3744-53 (2013). Acknowledgements The authors would like to thank B. Gómez-González for her comments on the manuscript and D. Haun for style supervision. Research in A.A.'s laboratory is funded by grants from the Spanish Ministry of Economy and Competitiveness, the Junta de Andaluc a, the uropean nion , orldwide ancer esearch, and the European Research Council. The authors apologize to those whose work could not be cited owing to space limitations. Competing interests statement The authors declare no competing interests. BOX 1. Fragile sites as hotspots of transcription-replication collisions Fragile sites are genomic regions exhibiting constrictions or gaps in metaphase chromosomes following replication stress. They are categorized into two classes: rare fragile sites are found in <5% of individuals and arise from trinucleotide repeat expansion, and common fragile sites (CFSs), which are found in all individuals and are not associated with repeat expansion 86. Fragile sites are frequently enriched in sequences that can stall DNA replication, such as AT-dinucleotide-rich sequences of high DNA flexibility in CFSs, as shown in yeast at FRA16D 87, or in other replicationattenuating sequences, as shown in human cells expressing the rare fragile site FRAXA at the FMR1 locus 87, 88. Scarcity in origins of DNA replication or inefficient replication activation in large genomic regions may explain the fragility of some CFSs 89, 90. Interestingly, mapping of CFSs in several human cell lines revealed they are located mostly within large genes. Also, there is a high correlation between CFSs and recurrent chromosomal rearrangements observed in cancer cells, and a similar correlation emerged from the analysis of copy number variants, whose hotspots matched CFSs when located in large regions of active transcription in both human and mouse cells 91, 92. These data suggest that concomitant transcription and replication may lead to fragility. Related to this is the observation that RNA-DNA hybrids Nat Rev Mol Cell Biol Transcription-replication conflicts 22 accumulate in the long FHIT, WWOX or IMMP2L genes, which harbour the CFSs FRA3B, FRA16D and FRA7K, respectively 93, as well as in the rare fragile sites FXN and FRAXA 61, among others 2. Recently, chromatin immunoprecipitation analysis with the single-strand DNA binding protein replication protein A (RPA) following replication stress has allowed the identification and mapping of a new class of fragile sites, termed early replication fragile sites (ERFSs), which contrary to CFSs are located near replication origins and within actively transcribed genes, strengthening the possibility that their fragility result from transcription-replication conflicts 94. Similarly, using the Break-seq [G] technique in cells following exposure to the replication-stress agent hydroxyurea, replication-induced double strand breaks were mapped preferentially at genes whose expression is[Au: OK?] induced under replication stress conditions 95. Therefore, genomes contain hotspots for transcription-replication collisions, which can manifest as different forms of fragility. BOX 2. Oncogenes and transcription-replication collisions. An oncogene refers to a gene that when mutated contributes to the development or progression of cancer, whereas the term proto-oncogene is reserved to its wild-type allele. Oncogenes generally regulate cell division, cell differentiation and/or cell death. This is the case of the oncogene c-Myc, which regulates transcription of several genes that control cell growth and cell cycle progression2. As replication stress and genomic instability are hallmarks of cancer cells 96, it seems plausible that oncogenes may increase the rate of transcription-replication conflicts, which will serve as a source of genomic instability. Altered expression of cyclin E or oncogenic Ras induces chromosomal fragility at sites that co-localize with large genes and only partially overlap with the canonical, replication stress-induced fragile sites 97. Oncogene expression can negatively affect replication by promoting replication origin activation, as shown for c-Myc or cyclin E 98, 99. Using DNA combing and cell-free extracts derived from Xenopus laevis eggs it has been shown that c-Myc increases activation of earlyreplicating origins, resulting in elevated fork collapse and subsequent DNA damage accumulation 100. Although such replication fork collapses may occur independently of transcription, the excess of active replication forks may increase the probability of collisions. Consistent with this view, DNA damage resulting from replication impairment by cyclin E overexpression was partially suppressed by the transcription inhibitor cordycepin, suggesting that collisions can indeed contribute to oncogene-induced replication stress101. It would be important to determine the general relevance of this Nat Rev Mol Cell Biol Transcription-replication conflicts 23 phenomenon in cancer cells and to explore the possibility of using transcriptionreplication collisions as a selective target in cancer therapy. Figure legends Figure 1. Transcription and replication. A. A small portion of the double DNA helix is unwound by the RNA polymerase (RNAP) to enable transcription (known as "transcription bubble"). DNA unwinding by the RNAP generates positive and negative supercoiling[G], which is alleviated by topoisomerases. In eukaryotes, transcription also involves chromatin modification and remodelling. The progression of RNA polymerase requires the activity of transcription elongation factors. The nascent RNA is co-transcriptionally processed by different factors. B. At the replication fork the DNA helicase minichromosome maintenance complex (MCM) opens the double helix and the NA polymerases Polε and Polδ extend the leading and lagging strand, respectively. Synthesis of each new NA molecule is initiated by the Polα-Primase complex Polα-Pri). Lagging strand synthesis leads to the formation of ssDNA, which is coated with replication protein A (RPA). Fork progression requires the activity of several replication cofactors, including the clamp proliferating cell nuclear antigen (PCNA). DNA unwinding by the replication fork generates positive supercoiling, which is alleviated by topoisomerases. Replication also entails reassembly of recycled and de novo-synthesized nucleosomes at the newly synthesized DNA. Dashed aArrows indicate the direction of fork progression andRNA and DNA polymerases synthesis. Figure 2. Head-on and co-directional transcription-replication collisions. A. Progression in opposite directions of an RNA Polymerase (RNAP) and a replication fork leads to head-on collisions, which induce pausing and blockage of the replication fork and may lead to its collapse and the formation of DNA breaks . B. Progression of an RNA Polymerase and a replication fork in the same direction leads to co-directional collisions if the fork moves faster than the RNA Polymerase. Co-directional collisions can be resolved by displacement of the RNA Polymerase from the DNA. MCM, minichromosome maintenance complex; Polε, DNA polymerase ε; Polδ, DNA polymerase δ. Figure 3. Conditions that affect the occurrence of transcription-replication collisions. A. Convergence of an RNA Polymerase (RNAP) and a replication fork when oriented head-on can lead to the accumulation of positive DNA supercoiling Nat Rev Mol Cell Biol Transcription-replication conflicts 24 between them, which induces pausing of the fork. B. The partial unwinding of DNA by the negative supercoiling generated behind the RNA polymerase can enable the formation of non-B DNA structures, such as G-quadruplexes, which may constitute an obstacle for replication fork progression. C. Other non-B DNA structures include RNADNA hybrids, which also may constitute an obstacle for fork progression. (B and C) Once a stable non-B DNA structure capable of blocking fork progression is cotranscriptionally formed, the direction of transcription or the presence of the RNA Polymerase itself would be in principle irrelevant for the formation of the transcriptionreplication collision. MCM, minichromosome maintenance complex; Polε, DNA polymerase ε; Polδ, DNA polymerase δ. Figure 4. Mechanisms preventing transcription-replication collisions. A. Pausing of RNA polymerase (RNAP) is normally resolved by backtracking, which disengages the 3'-end of the RNA molecule from the active site and leads to back and forth sliding of the RNA polymerase. The GreA and GreB RNA cleavage factors stimulate the removal of the extruded RNA and the reactivation of transcription. Without GreA and GreB the RNA polymerase might stall and become an obstacle for the replication fork, leading to transcription-replication collisions. B. The replication fork barrier (RFB) site is a DNA sequence located near the 3'-end of rRNA genes and prevents transcription– replication conflicts in the budding yeast. DNA replication fork-blocking protein Fob1 is required for RFB activity as without it there is no replication fork arrest at the RFB, resulting in transcription-replication collisions. C. Chromatin remodelling by the FACT complex facilitates transcription as well as replication fork progression. Without FACT altered chromatin reorganization results in transcription-replication collisions. D. At pericentromeric regions co-transcriptional RNAi releases the RNA polymerase thereby allowing completion of DNA replication. Without the RNAi machinery, failure to release the RNA polymerase during S-phase results in transcription–replication collisions. DnaB, DNA replicase B; Polα, DNA polymerase III α; MCM, minichromosome maintenance complex; Polε, NA polymerase ε; Polδ, DNA polymerase δ. Figure 5. Resolving transcription–replication collisions to avoid genome instabilityby the DNA damage response[Au: OK?]. A stalled replication fork can activate the ATR-dependent checkpoint to solve collisions and avoid their consequences. Specific factors are recruited to resolve or prevent the obstacle, including DNA helicases (Rrm3), mRNA processing proteins or chromatin remodelling complexes (FACT), DNA helicases (Rrm3) or mRNA processing proteins [Au: THO is not mentioned in the main text nor explained here, so we removed it, OK?] . In Nat Rev Mol Cell Biol Transcription-replication conflicts 25 addition, the RNA Polpolymerase (RNAP) could be released at transcription termination sites, with the help of BRCA1 and helicase SETX, or the INO80 histone remodelling complex and transcription factor PAF1C complex, as a way to avoid the collisions. Similarly, at tRNAs transcription sites de RNA Pol III is directly evicted during S-phase. Stabilization and resumption of stalled forks at transcribed DNA regions can occur via DNA repair factors, such as those of the Fanconi anemia repair pathway including tumour suppressors BRCA1 and BRCA2. Finally different DNA repair pathways can act at collision sites, if these degenerate into DNA lesions. Although in some examples direct involvement of the ATR-dependent checkpoint has been reported, in other cases is yet unknown. MCM, minichromosome maintenance complex; Polε, NA polymerase ε; Polδ, DNA polymerase δ. Glossary: Break-seq. Technique to map chromosome breaks based on DSB labelling and next generation sequencing. Bromodeoxyuridine. Synthetic analogue of the thymidine nucleoside used to follow DNA synthesis. CpG islands. Chromosomal regions with high density of non-methylated CpG sequences, often located at gene promoters. DNA combing. A method for the analysis of single DNA molecules; used for studying DNA replication. DNA damage response (DDR). Network of DNA damage repair and checkpoint factors that together to deal with DNA lesions. G-quadruplexes. Four repeats of at least three guanines that can interact to form fourstranded DNA structures. Hairpins. DNA structures in which a strand folds on itself and forms intrastrand base pairing. non-B DNA. Any DNA structure that is different from right-handed double helix with 10 nucleotides per turn. Phosphomimetic. Proteins with amino acid substitutions that simulate their phosphorylated state.