G4 motifs affect origin positioning and efficiency in two vertebrate replicators

Abstract

DNA replication ensures the accurate duplication of the genome at each cell cycle. It begins at specific sites called replication origins. Genome-wide studies in vertebrates have recently identified a consensus G-rich motif potentially able to form G-quadruplexes (G4) in most replication origins. However, there is no experimental evidence to demonstrate that G4 are actually required for replication initiation. We show here, with two model origins, that G4 motifs are required for replication initiation. Two G4 motifs cooperate in one of our model origins. The other contains only one critical G4, and its orientation determines the precise position of the replication start site. Point mutations affecting the stability of this G4 in vitro also impair origin function. Finally, this G4 is not sufficient for origin activity and must cooperate with a 200-bp cis-regulatory element. In conclusion, our study strongly supports the predicted essential role of G4 in replication initiation.

Figure 1. SNS profiles at the β^A and med14 loci

1. Relative enrichment in SNS along the endogenous β^A globin promoter and gene (adapted from Dazy et al, 2006). Each dot corresponds to a primer pair used for qPCR. The dashed rectangle represents the β^A origin inserted at an ectopic position (1.1-kb fragment) and the black arrow, the transcription start site. The red hexagons represent the positions of the G4 motifs along the locus. G4 motifs in the 5′ to 3′ orientation are above the line, and G4 motifs on the opposite strand are shown below the line.

2. Relative enrichment in SNS along the med14 promoter and gene. Each dot corresponds to a primer pair used for qPCR. The black arrow represents the transcription start site, and the red hexagons, the positions of the G4 motifs along the locus. All the G4 motifs are on the same strand.

Figure 2. The G4 motifs are necessary for origin function of med14 and β^A origins and have additive effects on med14

A Diagram of the WT and modified allele of the med14 origin with its G4 motifs #4 and #5. The green lines indicate the primer pairs used for qPCR analyses of SNS. Primer pair WT is specific for the WT allele; primer pair “modified” is specific for the modified allele; the bkgd primer pair binds 12 kb away from the site of insertion and is designed to detect background. B Relative SNS enrichment of the WT allele and the modified alleles. Error bars show the standard deviation of SNS enrichment for qPCR duplicates in two independent clones. C The endogenous β^A globin locus. Black lines represent the amplicons used for qPCR analyses of SNS at the endogenous locus. For each preparation, SNS enrichment within the endogenous β^A (positive control) was checked; SNS enrichment was arbitrarily normalized with respect to the endogenous ρ origin of the β-globin locus, the value for which was set at 100%. D Transgenes inserted at the ectopic position on chromosome 1. The transgenes contain the β^A origin promoter (white rectangle), the IL2R sequence (gray rectangle), and the SV40 polyA signal, flanked by two copies of the USF-binding site (2XFIV) (black rectangles). The red line and the red cross represent the G4 motif and G4 motif deletion, respectively. The black lines indicate the amplicons used for qPCR analyses of SNS; Primer pairs 1 and 2 are specific for the transgenes; the bkgd amplicon is located 5 kb away from the site of insertion and is designed to detect background. E, F Relative enrichment in SNS of the WT transgene (E) or the transgene lacking the G4 motif (F). The SP1-binding site is underlined in (E). Error bars show the standard deviation of SNS enrichment for qPCR duplicates in two independent clones.

Figure 3. Cis modules cooperate with G4 for origin activity

A–E Relative enrichment in SNS of transgene clones from which cis modules were deleted. The β^A promoter was cut into four modules: modules 1–3 are located 5′ to the G4 and module 4 is located 3′ to the G4. For each construct, deleted modules are indicated by lattice patterned rectangles. Results are shown for transgene lacking module 1 (A), transgene lacking modules 1 and 2 (B), transgene lacking modules 1, 2, and 3 (C), transgene lacking module 3 (D), and transgene lacking module 4 (E). Error bars indicate the standard deviation for SNS enrichment in two independent clones and qPCR duplicates. The black lines indicate the amplicons used for qPCR analyses of SNS.

Figure 4. G4 orientation determines the replication start site

A–C Relative enrichment in SNS of the transgene with the inverted G4 (A), the transgene lacking the 5′ sequences (B) and the transgene with the inverted G4 lacking the 5′ sequences (C). Error bars indicate the standard deviation of SNS enrichment for qPCR duplicates in two independent clones. The black lines indicate the amplicons used for qPCR analyses of SNS.

Figure 5. Potential G4 probability of the β^A G4#1 motif and of its mutants

The 17-mer β^A G-rich motif can, in principle, fold into 10 intramolecular G4, formed by the stacking of three G-quartets. All 10 possibilities are shown. The runs of guanine residues involved in the G4 are highlighted in green. For each mutant, we defined the potential G4 (PG4) probability as the number of possible G4 divided by the number of possible G4 for the WT sequence (i.e., 10). According to the position of the single point mutation, the PG4 probability is 0 (for m3, m7, m11, m15, m16, and m17), 0.1 (for m2 and m12), 0.3 (for m6 and m8), 0.4 (for m1 and m13), 0.6 (for m4, m5, m9, and m10), and 1 (for m14). The studied point mutations are indicated by a red square and cover most of the different PG4 probabilities.

Figure 6. Single-base point mutations within the β^A G4 motif affect origin activity to different extents

A–F Relative enrichment in SNS of G4 motif mutants m12 (A), m16 (B), m6 (C), m14 (D), m4 (E), and m9 (F). The mutated base is indicated in red. The SP1-binding site is underlined. Error bars indicate the standard deviation of SNS enrichment for qPCR duplicates in two independent clones.

Figure 7. β^A origin activity is correlated with the stability of its G4

1. Plot of the PG4 probability (as defined in Fig 5) of the 17-mer G-rich sequences against origin activity (relative enrichment in SNS as defined in Fig 6). The linear correlation is given by the following equation: PG4 P = 0.024 + 0.01 relative SNS enrichment, R = 0.926.

2. Plot of the percentage of oligonucleotides bound to hRPA, at concentrations of 2 nM oligonucleotide and 50 nM hRPA, as determined by electrophoretic mobility shift assays (Supplementary Fig S7) against origin activity (relative enrichment in SNS as defined in Fig 6). The binding reactions were carried out in the presence of 100 mM KCl, at 20°C. As a reference value, the percentage of oligonucleotides bound to hRPA was about 90%, for a WT 17-mer bearing several 7-deaza-guanines impairing G4 folding (Supplementary Fig S7). A linear fitting results in a correlation coefficient R = 0.672.

Figure 8. Comparison of the extent of the replication timing shift of the two different groups of constructs

1. Representation of the transgenes inserted at the ectopic position on chromosome 1. The transgenes contain the β^A origin promoter (white rectangle), the IL2R sequence (gray rectangle), and the SV40 polyA signal, flanked by two copies of the USF-binding site (2XFIV) (black rectangles). The green lines correspond to primers used for timing analyses; (With) is specific for the transgene; (Without) is specific to the unmodified allele; (Both) binds 5 kb away from the insertion.

2. Typical examples of timing analysis for transgenes displaying strong (WT #1) and low (ΔG clone #1) SNS enrichment. Blue, red, green, and purple bars represent sorted S-phase fractions from early to late stages of replication, S1, S2, S3, and S4, respectively. The early control consists of primer pairs binding within the endogenous β^A globin locus.

3. The distribution of the −ΔL + ΔE values from 14 replication timing assays is shown. For clones with a high degree of SNS enrichment, n = 6, corresponding to two independent clones of three constructs (WT, m14, and minvG4), and for clones with a low degree of SNS enrichment, n = 8, corresponding to two independent clones of four constructs (ΔG, m12, m16, and m6). Rectangles represent the quartiles (25% of the –ΔL + ΔE values below and above the median value, in red), and the deviations show the smallest and greatest –ΔL + ΔE values. A P = 0.0007 was obtained in Wilcoxon's nonparametric two-tailed test, with α = 5%.

Figure 9. Models describing how G4 might cooperate with a cis module to regulate the replication start site

The orange box represents the cis element; the black arrows represent normal replication fork progression. A, B The orientation of the G4 motif determines the positioning of the replication start site (green arrow) and that the position of the G4 motif with respect to its cooperating cis element is flexible. C, D Two proposed models explaining how bimodal origins might be generated. One cis element surrounded by two G4 motifs in opposite orientations can induce two start sites (C). Two cis elements that cooperate with two G4 motifs in opposite orientations are required (D).