Expression of two additional, distinct G4 DNA binding proteins rescues the cdc13-1 temperature-sensitive phenotype. (a) The G4 DNA-binding Sgs1 RQC domain, overexpressed from a plasmid and under the control of the GAL1 promoter (pRQC), or vector control were transformed into the indicated strain backgrounds and tested in spot assays. (b) A 5× HA-tagged Sgs1 RQC domain is enriched at telomeres of cdc13-1 mutants. The primer sets used were directly adjacent to the telomeres (subtelomeric Y’ repeat) or a control, centromere-proximal primer set (to a portion of SWC4, which has no QFP). The telomeric/centromeric ratio is listed above bars. Error bars are ± 1 s.d., and essentially the same result was obtained in three independent ChIP experiments. Mock IP indicates that cdc13-1 sgs1Δ cells with the RQC-expressing vector were ChIPed without antibody. (c) GAL-induced overexpression of the G4 DNA binding single-chain antibody (scFv), HF1 or empty vector in cdc13-1 mutants. For each assay, serial dilutions of cells were spotted on selective medium containing 2% galactose and grown at the indicated temperatures for 4 d. (d) The 13xMyc-tagged HF1 scFv is enriched at telomeres of cdc13-1 mutants. DNA was amplified as described in (b). Error bars are ± 1 s.d., and essentially the same result was obtained in three independent ChIP experiments.

Telomere-proximal single-stranded (ss) DNA accumulation at NPT (37 °C) is attenuated by two G4 DNA-stabilizing mechanisms. (a,b) ssDNA measurements in the context of STM1 overexpression (a) or sgs1 deletion (b) in cdc13-1 rad24Δ rad52Δ backgrounds. Subtelomeric G-strand ssDNA was probed with a complementary end-labeled ssDNA probe against the Y’ element and quantified by normalizing hybridization signals of native to denatured samples. Each sample was spotted in triplicate, and standard errors are shown. Each graph is one representative experiment. The apparent difference in the timing of ssDNA accumulation between the experiments reflects the time points examined rather than true experimental variability. (c,d) Attenuation of Rad53 phosphorylation by Stm1 overexpression or sgs1 deletion in cdc13-1 rad24Δ rad52Δ cells incubated at NPT (37 °C). Treatment of wild-type cells with 0.033% methyl methanesulfonate (MMS) provides a positive control for Rad53 phosphorylation (4d, lane 1).

G4 DNA-binding proteins and G4 DNA-forming sequences cooperate to inhibit end-resection by Exo1 in vitro. (a) Synthetic substrates having 3′ overhangs that form (bottom) or do not form (top) G-quadruplex DNA. Each contains 51 bp of identical duplex DNA. The 5′ end of each top strand was biotinylated (B) to inhibit resection from the incorrect end, and the terminal base of each bottom strand was labeled with α-³²P-dCTP. We illustrate a parallel intramolecular quadruplex (bottom), although this may not be the actual or only structure formed by this overhang. (b) Analysis of overhang substrates by CD shows a mixture of parallel and anti-parallel G4 DNA (GGG overhang) and a possible GA homoduplex (GAG overhang) (see Supplementary Fig. 6) under the exact conditions used for the Exo1 assay. (c) Thermal difference spectra of GGG and GAG yield a G-quadruplex-specific signal for only the GGG overhang. (d,e) Both Stm1 (d) and HF1 (e) cooperate with G4 DNA overhang to partially rescue Exo1 5′→3′ resection. Quantification of the fraction of remaining full-length sequence; shown are the averages of two (d) or three (e) independent experiments with standard errors and P-values from unpaired two-tailed Student’s t-tests. (f) Quantification of protection by T4 gene 32 protein (T4g32p). An average of two experiments is shown. P > 0.7 for all concentrations of T4g32p.

Overexpression of the G4 DNA binding protein Stm1 rescues growth defects caused by telomere uncapping and is independent of RAD52-dependent homologous recombination. (a) Growth of cdc13-1 mutants carrying pSTM1 or vector alone at permissive temperature (PT, 22 °C) or at semipermissive temperature (SPT, 30 °C). (b) pSTM1 overexpression rescues impaired growth caused by telomere uncapping in stn1-154 mutants at SPT. (c) Growth of cdc13-1 rad52Δ mutants carrying pSTM1 or vector. For each assay, serial dilutions of cells were spotted on selective medium and grown for 3 d (b,c) or 4 d (a). (d) Top: map of a typical telomere containing two tandem subtelomeric Y’ elements, separated by interstitial telomere repeats. Bottom: telomere Southern blots of samples grown in liquid culture at SPT for 2 d. Type I and type II survivors of telomerase inactivation are shown for comparison. The different sizes of internal Y’ fragments are due to short and long forms of Y’. Lanes 1–5 and 6–9 are sections from the same Southern blot.

Loss of the SGS1 activities associated with G4 DNA binding and unwinding rescues cdc13-1 temperature sensitivity and is independent of RAD52-dependent homologous recombination. (a) Growth of cdc13-1 rad24Δ rad52Δ, either SGS1 or sgs1Δ, at permissive temperature versus SPT. (b) Map of specific Sgs1 domains mutated or deleted in this study. (c) Deletion of the G-quadruplex binding RQC domain alone is sufficient to rescue the cdc13-1 growth defect. (d) Loss of Sgs1 helicase activity (K706A point mutation, denoted sgs1-hd) is sufficient to rescue cdc13-1 temperature sensitivity. (e) Sgs1 is the sole component of the Sgs1–Top3–Rmi1 complex that, when lost, is sufficient to rescue the cdc13-1 temperature sensitive phenotype. For each assay, serial dilutions of cells were spotted on YPAD (a,e) or selective medium (c,d) and grown at the indicated temperatures for 3 d (a,c,d) or 4 d (e).

Diminished rescue of cdc13-1 growth at SPT by sgs1 deletion in cells with mutant telomerase RNA templates that decrease QFP at telomeres. (a) cdc13-1 rad24Δ rad52Δ tlc1Δ cells, either SGS1 or sgs1Δ, and with plasmid-borne wild-type TLC1 or the uCu or CuA mutant tlc1 alleles, were spotted on SC-HIS medium and grown for 4 d. The sequence of the TLC1 template, including the mutated region (bold), is as follows: 3′-CACACACCCACACCAC-5′. (b) CD spectra of representative tlc1 mutant telomere sequences and their corrected counterparts (see Table 1). (c) Thermal difference spectra (TDS) of mutant versus corrected sequences, shown as the difference between molar extinction coefficients. G4 DNA yields a negative peak at 295 nm and positive peaks at 273 and 242 nm. All four corrected sequences showed evidence of G4 DNA formation by means of CD (parallel quadruplex for 1-corr and mixed parallel and antiparallel for 2, 3 and 4-corr) and TDS. (d) Thermal denaturation of oligonucleotides was followed by CD absorbance at 263 nm (for oligonucleotides 1, 1-corrected (corr), 2 and 2-corr) or 290 nm (for oligonucleotides 3, 3-corr, 4 and 4-corr).

Model of telomere capping in which G4 DNA cooperates with G-quadruplex binding proteins (G4BP) to inhibit exonucleolytic degradation and checkpoint activation caused by loss of Cdc13 function.