(A) Oligonucleotide sequences. The 39Ceb oligonucleotide mimicks one full CEB1 repeat. 39Cebm is a control sequence with five base substitutions (shown in bold) (B) Melting profiles. Absorbance at 295 nm vs temperature plots for 39Ceb (triangles) and 39Cebm (circles) each at 3 µM strand concentration. Melting experiments are performed in 10 mM lithium pH 7.2 cacodylate buffer supplemented with 0.1 M KCl. (C) Thermal difference spectra. Thermal difference spectra result from the difference between the absorbance recorded at 79±2°C and at 40±2°C in a 10 mM lithium pH 7.2 cacodylate buffer supplemented with 0.1 M KCl. Thermal difference spectra are normalized (TDS_norm = TDS/max(TDS)) over the 220–335 nm wavelength range. Full line: 39Ceb; dotted line: 39Cebm. (D) Circular dichroism spectra. Oligonucleotides were prepared at 140 µM strand concentration and annealed in 1 M NaCl as in helicase experiments, then immediately diluted to 3 µM in a 10 mM lithium cacodylate 1 M NaCl, pH 7.2 buffer. Full line: 39Ceb; dotted line: 39Cebm. Spectra were recorded at three different temperatures: 25°C (squares), 65°C (circles) and 90°C (triangles). (E) Behavior of the 39Ceb and 39Cebm sequences on a non-denaturing gel. Oligonucleotides were prepared at 140 µM strand concentration, annealed in 1 M NaCl as in helicase experiments and loaded on a non-denaturing 15% acrylamide gel supplemented with 20 mM NaCl and run at 21°C. Migration markers are 1: double-stranded DNA (9 and 12 bp) and 2: (dT)₁₅, (dT)₂₁ and (dT)₃₀ oligomers. (F) In vitro unwinding of 2 nM G-quadruplex DNA (G4-CEB1, left) versus 2 nM double stranded DNA oligonucleotide substrate (D20, right) in presence of decreasing amount of Pif1 (WT or helicase-dead pif1-K264A) for 15 minutes at 35°C in presence or absence of ATP. (G) Quantifications of the gels shown in F. (H) Kinetics of unwinding of G-quadruplex DNA (left) versus double stranded DNA oligonucleotide D20 (right) in presence of 100 nM Pif1. (I) Quantifications of the gels shown in H.

In the absence of Rad27, accumulation of unresolved flap structures inside CEB1 during replication generates recombinogenic structures that are repaired by homology-dependent strand displacement and annealing (SDSA). While in the absence of Pif1, the persistence of unprocessed G-quadruplex secondary structures in CEB1, during replication, transcription or other processes, initiates DNA lesions that are also repaired by SDSA, leading to minisatellite rearrangements.

(A) Structure of the genomic locus containing CEB1-1.8. (B) Southern blot analysis of CEB1-1.8 instability in haploid strains: WT (ORT2914), pif1Δ (ORT4843), pif1-K264R (ORT5083-4E). Each lane contains DNA extracted from pools of 12 independent colonies digested by AluI and hybridized with a CEB1-0.6 probe. (C) Structure of CEB1-1.8 rearrangements obtained in pif1Δ haploids. Each of the 42 CEB1-1.8 repeats is represented by a colored box and numbered (top). Nine rearrangements were sequenced and classified in three categories (1 to 3). The name of each rearranged allele is at the left. Hybrid repeats are represented by the two colors corresponding to the fused repeats. The white box in P23 indicates a motif that cannot be attributed to a specific parental motif. (D) CEB1 instability is Rad52 and Rad51 dependent. Southern blot analysis of CEB1-1.8 instability in pif1Δ rad52Δ (ORD7565-2C) and pif1Δ rad51Δ (ORD7574-9B) haploid strains. Same legends as in (B). Additional bands marked with an asterisk are presumably due to partial digestion by AluI.

Comparison of synthetic-CEB1-WT-1.7 (A) and synthetic-CEB1-Gmut-1.7 (B) instability in WT, pif1Δ and rad27Δ haploid strains by Southern blot analysis. In order to increase the number of independent colonies analyzed in pif1Δ and WT strains, colonies are pooled for DNA extraction and the number of colonies analyzed per well is indicated under each gel. DNA is digested by ApaI/SpeI and hybridized with a CEB1-WT or CEB1-Gmut probe. When several rearranged minisatellites migrate at the same size they are considered as clonal and are counted only one time. The frequency of instability for each synthetic minisatellite is reported in Table 3. (C) Frequency of natural and synthetic CEB1 minisatellites according to the size of the alleles in pif1Δ cells.