ABSTRACT
The locations of active DNA replication origins in the human genome, and the determinants of origin activation, remain controversial. Additionally, neither the predominant sites of replication termination nor the impact of transcription on replication-fork mobility have been defined. We demonstrate that replication initiation occurs preferentially in the immediate vicinity of the transcription start site of genes occupied by high levels of RNA polymerase II, ensuring co-directional replication of the most highly transcribed genes. Further, we demonstrate that dormant replication origin firing represents the global activation of pre-existing origins. We also show that DNA replication naturally terminates at the polyadenylation site of transcribed genes. During replication stress, termination is redistributed to gene bodies, generating a global reorientation of replication relative to transcription. Our analysis provides a unified model for the coupling of transcription with replication initiation and termination in human cells.
INTRODUCTION
Since the identification of cis-acting sequences responsible for the definition of replication origins in S. cerevisiae (Stinchcomb et al., 1979), significant effort has been applied to identify analogous determinants of replication initiation in human cells (Hyrien, 2015; Prioleau and MacAlpine, 2016). However, although origins have been observed via several independent techniques to be enriched close to transcribed genes and a range of other chromatin features (Besnard et al., 2012; Dellino et al., 2013; Langley et al., 2016; Petryk et al., 2016), a coherent model that encompasses both origin specification and activation has not emerged. Throughout eukaryotes, many more replication origins are licensed by MCM2-7 loading in G1 than are required to complete S-phase (Donovan et al., 1997; Edwards et al., 2002). The pool of excess MCMs is required for survival when dNTPs are depleted by the ribonucleotide reductase inhibitor hydroxyurea (HU) (Ge et al., 2007), and has been proposed to allow the firing of ‘dormant’ replication origins that rescue genome replication after replication fork stalling. The identities of these dormant origins, and how they differ from constitutive origins, have not been defined.
DNA is a one-dimensional template that can be simultaneously acted upon by the replication and transcription machineries. The orientation of essential genes in prokaryotes is biased to avoid head-on collisions between these two processes (Rocha, 2003); among eukaryotes, budding yeast and C. elegans show statistical orientation bias of the most highly transcribed genes to the co-directional orientation (Osmundson et al., 2017; Pourkarimi et al., 2016), and significant co-orientation of transcribed genes has been noted in human cells (Hamperl et al., 2017; Petryk et al., 2016). Head-on replication-transcription conflicts are deleterious in eukaryotes, leading to increased DNA damage (Hamperl et al., 2017) and genomic rearrangements (Tran et al., 2017). The mechanisms by which origin location is specified to co-orient replication with transcription in diverse human cell types, and the impact of co-directional transcription on replication-fork progression through genes, remain speculative.
Here, by using Okazaki fragment sequencing (Ok-seq) to infer the direction of replication-fork movement, we define transcription initiation efficiency and gene length as independent determinants of replication origin location and firing efficiency, and show that origin firing occurs close to the transcription start site (TSS), ensuring co-oriented replication of the most highly transcribed genes. We additionally show that dormant origins correspond to constitutive origins. Further, we determine the existence of widespread replication-fork stalling due to co-oriented replication-transcription conflicts at transcription termination sites (TTS) under unperturbed conditions, and in gene bodies during replication stress.
RESULTS
Using Ok-seq, we and others have reported on replication initiation (McGuffee et al., 2013; Petryk et al., 2016; Pourkarimi et al., 2016), replication elongation (Osmundson et al., 2017) and lagging-strand processing (Osmundson et al., 2017; Smith and Whitehouse, 2012). To investigate replication initiation and termination in untransformed human cells, we performed Ok-seq (Petryk et al., 2016) on hTERT-immortalized RPE-1 cells. Previous genome-wide studies of metazoan replication show limited agreement in origin calls (Hyrien, 2015; Prioleau and MacAlpine, 2016), with the exception that all identify significant enrichment of origins close to transcribed genes. Therefore, instead of aiming to identify all potential sites at which replication can possibly initiate in the human genome, we focused our analysis on the efficiency of replication initiation and termination in transcribed regions.
Replication initiates in the immediate vicinity of high-conflict transcription start sites
Rightward-moving replication forks generate Okazaki fragments (OFs) that map to the Crick strand, while leftward-moving forks generate Watson-strand fragments (Fig. 1A). Therefore, a replication origin will manifest as an increase in the proportion of OFs mapping to the Crick strand: the efficiency and spatial localization of origin firing will impact the magnitude and gradient of this transition, respectively (Fig. 1B). OFs showed no strand bias around random genomic loci (Fig. S1). Consistent with global origin activity at TSS, meta-analysis of OF Crick strand bias over a 50 kb window surrounding all 17920 annotated TSS showed a symmetrical transition from predominantly leftward-to predominantly rightward-moving forks (Fig. 1C). However, separate analysis of Watson-and Crick-strand genes revealed a profound asymmetry based on gene orientation (Fig. 1D). Therefore, for all subsequent analyses we considered OF strand bias relative to gene orientation, such that transcription occurs from left to right and OF polarity is computationally reversed for Crick-strand genes. All data for TSS analysis were highly reproducible across two biological replicates (Fig. S2).
(A) The relationship between replication direction and Okazaki fragment (OF) strand.
(B) Expected OF distributions around replication origins.
(C) Percent of OFs mapping to the Crick strand across a ±50 kb window around all transcription start sites (TSS) in the human genome. All data in figures 1-3 are from asynchronously dividing hTERT-immortalized RPE-1 cells, and are displayed using 1 kb bins.
(D) Percentage of OFs mapping to the Crick strand around the TSS of Watson (W) or Crick (C) genes.
(E) First derivative of Okazaki fragment strand bias around TSS, oriented such that transcription occurs from left to right. Data are smoothed across two 1 kb bins.
(F) Schematic representation of upstream and TSS-proximal replication initiation inferred from data in Figure 1.
(G) Percentage of replication forks moving left to right around TSS binned by RNA-seq read depth quartile from (Harenza et al., 2017) (panels i-iv).
(H) Percentage of replication forks moving left to right around TSS of actively transcribed genes (FPKM > median, (Harenza et al., 2017)) binned by length according to quartiles for all genes (panels i-iv).
(I) Percentage of replication forks moving left to right around TSS binned by transcriptional volume (FPKM from (Harenza et al., 2017) x gene length).
By analyzing the first derivative of replication direction, we can directly infer the extent of replication initiation at each position relative to the meta-TSS. We observe that, while individually low but cumulatively high levels of initiation occur over a wide region upstream of TSS, initiation is strongly biased to a ~5 kb region immediately adjacent to the TSS itself and is under-represented or absent in gene bodies (Fig. 1E, schematic in Fig. 1F).
To investigate the effect of transcription on TSS-proximal replication initiation, we separated genes into quartiles based on RNA-seq read density (FPKM, Fragments Per Kilobase of transcript per Million mapped reads) across the gene body in RPE-1 cells (Harenza et al., 2017). TSS of genes with higher FPKM showed significantly greater change in strand bias than TSS of weakly or non-transcribed genes (Fig. 1G). However, we also observed a significant length dependence in this effect, such that short genes have weaker origin activity than long genes with equivalent RNA-seq read density (Fig. 1H). Gene length and FPKM are not correlated in RPE-1 cells (r = −0.044): therefore, gene length and transcript number independently correlate with origin firing efficiency. While (assuming equal RNA decay rates) FPKM reports the number of RNA molecules synthesized, the number of RNA polymerases occupying a gene for a given FPKM is linearly related to the length of the gene i.e. a 50 kb gene producing one transcript per minute will be occupied by ten times as many polymerases as a similarly transcribed 5 kb gene. We therefore analyzed OF strand bias around TSS separated by transcriptional volume (FPKM x gene length). Origin activity was strongly correlated with high transcriptional volume (Fig. 1l). All data from Fig. 1 were reproducible in previously published Ok-seq data from HeLa cells (Petryk et al., 2016) (Fig. S3). We note that the distance to the nearest downstream TSS or transcription termination site (TTS) - both of which are intrinsically dependent on gene length - are substantially stronger predictors of origin efficiency than the distance to the nearest upstream TSS or TTS, which are independent of gene length (Fig. S4). Therefore, the effect of gene length on origin efficiency does not simply reflect decreased passive replication from nearby ‘competitor’ origins. We conclude that, in unperturbed human cells, the initiation of DNA replication is strongly biased towards the immediate vicinity of TSS that drive transcription of genes with high RNA polymerase II (RNAP2) occupancy. Thus, while transcription inherently creates conflicts with replication, the coupling of origin firing to RNAP2 occupancy ensures that these conflicts are in the co-directional orientation.
Modulation of constitutive origin efficiency during replication stress
In response to replication stress, origin firing increases: this is generally interpreted as the activation of dormant origins (Blow et al., 2011). However, firing events during growth in hydroxyurea (HU) have previously been reported to occur in the vicinity of transcribed genes (Karnani and Dutta, 2011). To analyze origin firing efficiency at TSS in the context of increased or decreased dormant origin usage, we treated cells with 0.2 mM HU for 4h before OF collection, and additionally depleted either of the Fanconi Anemia (FA) effector proteins, (Ceccaldi et al., 2016; Michl et al., 2016) FANCD2 or FANCI, by RNAi. This HU treatment regime has previously been shown to decrease average inter-origin distance without activating the checkpoint (Ge et al., 2007): we previously showed that knockdown of FANCI increases inter-origin distance in HU, reflecting reduced origin firing, while knockdown of FANCD2 weakly stimulates origin firing under the same conditions (Chen et al., 2015). In the absence of HU, knockdown of either FANCI or FANCD2 have little or no measureable effect on origin firing, cell cycle progression, or cell doubling (Chen et al., 2015). OF distributions around TSS showed a greater change in strand bias adjacent to high-volume TSS for HU-treated cells relative to untreated (Fig. 2A), consistent with further increased firing efficiency of efficient origins. The rapid decrease in strand bias downstream of the TSS can be explained by increased intragenic fork stalling and is addressed below (Fig. 3). Knockdown of FANCI in HU-treated cells reduced strand bias around high-volume TSS, while knockdown of FANCD2 very slightly increased it (Fig. 2B & S2C). Both short and long genes were affected (Fig. S5). Therefore, under conditions of globally increased or decreased origin firing, we find that the firing efficiency of the most active origins is the most strongly affected. Our data suggest that intragenic replication origins are not required for widespread dormant origin firing, and that dormant origins are in fact constitutive origins with altered firing efficiency.
(A) Percentage of replication forks moving left to right around TSS binned by transcriptional volume, for cells grown in the absence (black) or presence (red) of 0.2 mM hydroxyurea (HU) for 4h before OF collection.
(B) Percentage of replication forks moving left to right around TSS binned by transcriptional volume, for cells treated with siRNAs against FANCD2 (green), FANCI (blue), or mock-treated (black), grown in 0.2 mM hydroxyurea for 4h before OF collection.
Replication terminates at transcription termination sites
We reasoned that the loss of OF strand bias observed at a gene-length-dependent distance downstream of the TSS (Fig. 1H) might result from replication termination at or near the 3’ ends of transcribed genes. To test this, we queried OF distributions around the annotated gene 3’ end (hereafter referred to as the TTS). The TTS corresponds to the site of pre-mRNA cleavage and polyadenylation, and may not directly reflect the site of RNAP2 dissociation for genes that rely on Xrn2-mediated transcriptional termination (Proudfoot, 2016). The number of RNA polymerases terminating transcription should equal the number initiating at the TSS; therefore, we analyzed OF strand bias around the TTS of genes separated by quartiles based on FPKM, which serves as a proxy for initiation frequency. Consistent with widespread replication termination at the TTS of transcribed genes, we observed a dramatic, transcription-dependent reduction in replication forks moving in the direction of transcription through the gene body, occurring precisely at the TTS (Fig. 3A). This termination signal was surrounded by an increase in left-to-right polarity, consistent with diffuse origin firing occurring a variable distance away at nearby TSS (Fig. 3B).
(A) Percentage of replication forks moving left to right around transcription termination sites (TTS) binned by RNA-seq read depth quartile from (Harenza et al., 2017) (panels i-iv).
(B) Schematic representation of replication termination at TTS, with initiation at proximal TSS up-and downstream.
(C) DRIP-seq signal from (14) around TTS of actively transcribed genes (FPKM from (18) > median) separated into high (purple) and low (brown) DRIP bins based on DRIP-seq signal (Hamperl et al., 2017) ±10kb from the TTS.
(D) Length distribution of actively transcribed high-DRIP vs low-DRIP genes
(E)Percentage of replication forks moving left to right around TTS of actively transcribed (FPKM from (Harenza et al., 2017) > median) high-DRIP vs low-DRIP genes.
(F)(A) Percentage of replication forks moving left to right around (TTS) binned by RNA-seq read depth quartile from (Harenza et al., 2017) (panels i-iv), for cells grown in the absence (black) or presence (red) of 0.2 mM hydroxyurea (HU) for 4h before OF collection.
(G)Schematic representation of the change in replication termination observed under replication stress.
Because RNA-DNA hybrids (R-loops) have been proposed to modulate both replication-fork stalling and DNA damage, we analyzed genes separately based on their likely levels of R-loop formation at TTS in RPE cells. Most differences in R-loop formation between cell types and even species can be attributed to differences in transcription (Sanz et al., 2016): therefore, we separated genes with FPKM above the median level in RPE cells into two bins based on the DRIP-seq signal observed ±10 kb from their TTS in HeLa cells (Hamperl et al., 2017), to obtain high-and low-DRIP TTS gene sets (Fig. 3C). We note that the genes in our high-DRIP set are significantly shorter than those in our low-DRIP set (Fig. 3D), possibly reflecting some contribution of signal from TSS-proximal R-loops in short genes. Analysis of OF strand bias around the TTS of high-and low-DRIP genes indicated equivalent replication termination regardless of the propensity to form R-loops (Fig. 3E). The upward gradient observed upstream of the TTS for high-DRIP genes is due to origin firing at TSS, which are closer to TTS due to the length difference observed in Fig. 3D. Data from Fig. 3 A-E were reproducible in previously published Ok-seq data from HeLa cells (Petryk et al., 2016) (Fig. S6), and are therefore not attributable to differences in R-loop formation between cell types. Thus, we conclude that replication naturally terminates at the 3’ end of transcribed genes, and that termination is independent of R-loop formation. Replication origins are located at variable distances from TTS: therefore, this effect must be driven by replication-fork stalling or arrest as opposed to simply by replication initiation kinetics (McGuffee et al., 2013).
We next analyzed replication-fork mobility around TTS in the context of growth in 0.2 mM HU (Fig. 3F, schematic in Fig. 3G). All data for TTS were again reproducible across replicate datasets (Fig. S7). We observed both an increase in the overall level of gene-associated replication termination, and a change in the predominant location of termination from the TTS to the gene body. Consistent with this observation, analysis of OF strand bias around TSS indicates that the proportion of replication forks moving co-directionally through gene bodies decreases more rapidly in the presence than the absence of 0.2 mM HU (Fig. 2A), but is unaffected by knockdown of FANCI or FANCD2 (Fig. 2B). Thus, replication termination is significantly altered by 0.2 mM HU: we propose that replication forks normally stall upon encountering RNAP2 paused at the TTS, but that intragenic replication-transcription conflicts become the predominant cause of replication-fork stalling under replication stress.
DISCUSSION
Co-orientation of replication and transcription in multicellular eukaryotes
Our data suggest that DNA replication initiation is inherently coupled to transcription initiation, such that replication preferentially initiates immediately adjacent to the TSS of genes with high RNAP2 occupancy during S-phase (model, Fig. S8). Spatially limiting initiation in transcribed genomic regions to the TSS of the most highly transcribed genes ensures that these show the strongest bias towards co-directional replication in any proliferative cell type, regardless of its transcriptional profile. Activation of additional genes, for example during differentiation or oncogenesis (Macheret and Halazonetis, 2018), would concurrently change the replication profile of the cell to mitigate conflicts with transcription. Many unicellular organisms, including prokaryotes and S. cerevisiae define replication origins via the use of cis-acting sequences: in these organisms, the orientation of conflict-prone genes has been biased by evolution (Merrikh, 2017; Osmundson et al., 2017; Paul et al., 2013), presumably as a result of damage induced by head-on collisions followed by re-orientation (Srivatsan et al., 2010). However, the use of such cis-acting sequences in an organism with many transcriptionally distinct proliferative cell types would be problematic because it would enforce deleterious head-on replication-transcription conflicts in a cell-type-specific fashion. Thus, while unicellular organisms can co-orient transcription with replication by genome evolution, the alternative strategy - to co-orient replication with transcription - is more robust for organisms with many distinct transcriptomes.
A simple mechanism to explain the increased replication origin activity of the genes occupied by the most RNAP2 is that chromatin accessibility independently determines both MCM2-7 loading and the recruitment of replisome components to these licensed replication origins. We note that chromatin accessibility, as opposed to RNA-seq read density, has previously been reported as the best correlate of replication timing (Hansen et al., 2009). Under conditions of global origin activation or repression, a proportionally similar activation/repression of all origins will lead to the greatest change in origin efficiency occurring at origins with intrinsically high firing efficiencies (Fig. 2). Thus, dormant origin firing globally maintains co-orientation of replication and transcription, and can simply be considered as a change in average origin efficiency as opposed to the specific regulation of distinct classes of origin.
There are fewer actively transcribed genes than the number of origins required to replicate the human genome. Inactive regions of the genome lack both spatially localized regions of highly accessible chromatin and a requirement to co-orient replication with transcription: therefore, specifically localized replication origins would serve no biological purpose in these regions. Transient chromatin opening, likely driven by low-levels of pervasive transcription factor binding and/or transcription, could ensure a sufficient density of MCM2-7 loading and activation to support genome duplication in these regions through a distribution of origin firing events that largely disfavors initiation at specific individual sites. We speculate that the poor overlap between genome-wide origin mapping studies (Prioleau and MacAlpine, 2016) is due to a lack of localized origin firing outside highly transcribed regions as opposed to an intrinsic flaw with any origin detection approach.
Intuitively, replication-fork initiation at TSS followed by stalling or arrest at TTS might impair the replication of intragenic regions flanked by convergent, highly transcribed genes. However, the ability of both RNA polymerases (Gros et al., 2014; Gros et al., 2015) and the replicative helicase (Douglas et al., 2018) to push loaded MCM2-7 double hexamers would lead to a gradual redistribution of un-fired origins towards these intragenic regions, where they could serve as a pool of licensed ‘rescue’ origins.
Re-orientation of replication-transcription conflicts during replication stress
Although DNA replication and transcription are, on average, co-oriented throughout the actively transcribed regions of the genome, we observe a robust replication termination signal at the TTS of transcribed genes (Fig. 3). Therefore, co-oriented collisions between the replication and transcription machineries are highly prevalent, and RNAP paused at the 3’ ends of genes acts as a physical barrier to even co-directional replication-fork progression. Replication-fork stalling due to co-oriented collisions with RNA polymerase at gene 3’ ends has previously been observed in prokaryotes (Mirkin et al., 2006) but not, to our knowledge, in eukaryotes. Unlike in prokaryotes (Lang et al., 2017), but similarly to replication-fork arrest at tRNA genes in yeast (Osmundson et al., 2017), R-loops do not appear to impede replisome mobility despite their impact on the ultimate outcome of collisions (Tran et al., 2017; Hamperl et al., 2017; Paulsen et al., 2009; Stirling et al., 2011). We note that independent screens in HeLa cells (Paulsen et al., 2009) and S. cerevisiae (Stirling et al., 2011) identified RNA splicing and 3’ end formation factors as preventative against DNA damage. It will be interesting to determine the relative contributions of R-loop modulation and replisome mobility to the increased damage when RNA processing is impaired.
The locations of replication origins are minimally affected by hydroxyurea-mediated dNTP depletion, but the progression of replication forks through genes is severely impeded (Fig. 2A, 3F). We propose that, during replication stress, co-directional replication-transcription conflicts within the gene body become more frequent due to reduced replisome speed, leading to intragenic replisome stalling. Under these conditions, replication of gene 3’ ends must be rescued by a head-on fork. Such a global re-orientation of replication-transcription conflicts towards the head-on orientation would greatly increase DNA damage and ATR signaling, which are strongly orientation-dependent (Hamperl et al., 2017). Therefore, we propose that conflict re-orientation, as opposed to unscheduled dormant origin firing or impaired replication fork movement per se due to replication-transcription conflicts, underlies the toxic effects of replication stress and the transcription-dependent fragility of long genes (Helmrich et al., 2011).
DATA AVAILABILITY
A token to access the raw and processed data is available upon request from the corresponding authors.
AUTHOR CONTRIBUTIONS
Y-H. C., M.K. and P.T. performed research, S.K. and D.J.S. analyzed data., D.F., T.T.H. and D.J.S. conceived and supervised the study. All authors interpreted data. D.J.S. wrote the manuscript with input from all authors.
COMPETING INTERESTS
No competing interests exist.
METHODS
Cell Culture/EdU labeling
hTERT immortalized RPE-1 cells (ATCC) were grown in Dulbecco’s modified Eagle’s medium: Nutrient Mixture F-12 (DMEM/F-12) media (Life Technologies) supplemented with 10% fetal bovine serum (FBS), 3% sodium bicarbonate and 1% Pen-Strep. In brief, exponentially growing cells of 50-60% confluency were plated in 150mm dishes (approximately 15-20 plates per sample/replicate). siRNA transfections were done according to (Chen et al., 2015). EdU labeling was done for 2 min (4 min for low-dose HU) at 20 μM final concentration. Cells were either untreated or treated with low-dose HU (200 μM final concentration).
Ok-seq
Okazaki fragments were purified and libraries generated essentially as described (Petryk et al., 2016), with the most major modification being the use of gel-purified adaptor duplexes from (Smith and Whitehouse, 2012). Libraries were sequenced using the HiSeq-2500 platform and reads were aligned to the hg19 build of the human genome.
Data analysis
TSS and TTS locations were obtained from the UCSC genome browser http://genome.ucsc.edu - genes devoid of uniquely mapping OF reads on both the Watson and Crick strand through the entire gene body were excluded from analysis. Genes were separated by RNA-seq read density in RPE-1 cells, length or DRIP-seq read density in HeLa cells as described in the main text.
(A) Data were analyzed as in Fig. 1C, using Ok-seq data from HeLa cells (7)
(B) Data were analyzed as in Fig. 1D, using Ok-seq data from HeLa cells (7)
(C) Data were analyzed as in Fig. 1E, using Ok-seq data from HeLa cells (7)
(D) Data were analyzed as in Fig. 1G, using Ok-seq data from HeLa cells (7)
(E) Data were analyzed as in Fig. 1H, using Ok-seq data from HeLa cells (7)
(F) Data were analyzed as in Fig. 1I, using Ok-seq data from HeLa cells (7)
All gene expression data for HeLa cells are from the EBI gene expression atlas https://www.ebi.ac.uk/gxa/home - absent values for FPKM were set to zero.
(A) Percentage of replication forks moving left to right around TSS of actively transcribed genes (FPKM (18) > median), where the TSS of the most proximal upstream gene is under (black) or over (green) 50 kb from the TSS being analyzed.
(B) Percentage of replication forks moving left to right around TSS of actively transcribed genes (FPKM (18) > median), where the TTS of the most proximal upstream gene is under (black) or over (red) 50 kb from the TSS being analyzed.
(C) Percentage of replication forks moving left to right around TSS of actively transcribed genes (FPKM (18) > median), where the TSS of the most proximal downstream gene is under (black) or over (green) 50 kb from the TSS being analyzed.
(D) Percentage of replication forks moving left to right around TSS of actively transcribed genes (FPKM (18) > median), where most proximal downstream TTS is under (black) or over (red) 50 kb from the TSS being analyzed. Note that this TSS-TTS distance is equivalent to the length of the gene.
Percentage of replication forks moving left to right around TSS binned by transcriptional volume (FPKM (18) × gene length) for cells treated with siRNAs against FANCD2 (green), FANCI (blue), or mock-treated (black), grown in 0.2 mM hydroxyurea for 4h before OF collection.
(A) Data were analyzed as in Fig. 3A, using Ok-seq data from HeLa cells (7)
(B) Data were analyzed as in Fig. 3C: DRIP-seq data are from HeLa cells (14)
(C) Data were analyzed as inFig. 3E, using Ok-seq data from HeLa cells (7): : DRIP-seq data are from HeLa cells (14)
ACKNOWLEDGEMENTS
We thank the NYU Genome Technology Center for assistance with TapeStation and sequencing. We thank D. Remus, I. Whitehouse, E. Mazzoni, and H. Klein for helpful discussions. Y-H C was funded in part by the Molecular Oncology and Immunology NCI training program through NYU School of Medicine (5T32CA009161-40). Work in T.H. laboratory is supported by grants from the NIH (ES025166), V Foundation BRCA Research and Basser Innovation Award. Work in D.J.S. laboratory is supported by grants from the NIH (GM127336, GM114340) and the Searle Scholars Program.
Footnotes
One sentence summary: Co-oriented replication-transcription conflicts are prevalent in human cells because initiation of both processes is coupled.
