(a) Schematic short telomere assay: structures of VII-L end before (parental) and after FLP recombination in experimental (left) and control (right) strains. Recombination generates VII-L telomere with ~100 bp telomeric DNA in experimental strain (short) or ~300 bps (WT) telomere in control strain. Restriction enzyme sites: XhoI (X), StuI (S), EcoRV (V), PstI (P), HindIII (H). (b) Experimental and control strains expressing Mre11-MYC, Rad50-MYC, or Xrs2-MYC were arrested, FLP was induced, and cells were removed from both alpha factor and galactose (0 min), and proceeded through synchronous cell cycle. Filled and open squares indicate, respectively, average fold enrichment of tagged protein ± one standard deviation to short (open) or WT (closed) VII-L telomere compared to binding to non-telomeric ARO1 locus in same sample. Fold enrichment scale is not the same on three panels. (c) Using same samples as in panel b, binding of indicated protein to WT length VI-R telomere was determined in experimental (open triangles) and control (closed triangles) strains. Scale for fold enrichment is different from other panels. Binding of each protein to VI-R telomeres in two strains was not significantly different at any time point (p-values all ≥ 0.12). (d) Xrs2 binds preferentially to short telomeres for at least two cell cycles. Experimental strain expressing Xrs2-MYC was synchronized. Samples were taken over two cell cycles and processed for ChIP. Average levels Xrs2-MYC binding to short VII-L (open squares) and WT length VI-R (open triangles) telomeres are shown.

The control and experimental strains expressing Mec1-HA (a–c) or Tel1-HA (d) were synchronized and processed as described (Fig. 1). (a) Mec1-HA binding to the VII-L telomeres occurred at low levels at 60, 75 and 90 mins. The timing and level of Mec1-HA binding to the short VII telomere (open squares) and the WT length VII-L telomere (closed squares) were indistinguishable at all time points (p-values from 0.19 to 0.96). (b) Mec1-HA binding to the WT length VI-R and XV-L telomeres was indistinguishable from binding to the short VII-L telomere (taken from panel a and shown for comparison) except at 75 and 90 min (p= 0.01 and 0.03). Only data from the experimental strain are shown. (c) The experiment in panel a was carried out in tel1Δ versions of the control and experimental strains expressing Mec1-HA. Binding to the short and WT length VII-L telomeres was indistinguishable from binding to the non-telomeric ARO1 locus in the same samples (to which the values are normalized). (d) Experiments to determine if Tel1-HA binding is affected by absence of Mec1 were done in sml1Δ derivative of the WT and mec1Δ strains53. Tel1-HA binding to the short VII-L telomere or to the WT VI-R telomeres (panel d) was indistinguishable in sml1Δ (MEC1) and mec1Δ sml1Δ (mec1) cells (p-values from 30 to 90 min were 0.08 to 0.84).

(a) Schematic of chromosome VII-L end in DSB strains24. TG80-HO contains 80 bp TG_1–3 on centromere side of HO site; N80-HO contains 80 bp lambda DNA. V (EcoRV) sites (b) Mec1-HA binding to DSBs. For b, c, cells were synchronized and processed as described in (Fig. 1) except an extra sample was taken from G1 arrested cells before addition of galactose. Mec1 binding was determined at N80-HO (closed circles) and TG80-HO (open circles) before and after HO induction. Results are average fold enrichment ± one standard deviation over binding to control site ARO1. (c) Rfa1-MYC binding to DSBs. Rfa1-MYC binding was determined at N80-HO (closed circles) and TG80-HO (open circles) before and after induction of HO. Results are average percent of DNA in input. Rfa1-MYC binding to N80-HO was higher than at TG80-HO break (p-values 0.00037 to 0.017), in all post-galactose samples except the 45 min time point (p= 0.093). (d) Samples used in panel c were examined for Rfa1-MYC binding to internal ARO locus in both TG80 (open triangles) and N80 (closed triangles) strains. The level and timing of Rfa1-MYC binding was equivalent in the two strains. (e) Cdc13-MYC binding to TG80 (open circles) and N80 (closed circles) in first cell cycle after breakage. (f) The same samples in panel e for the N80 strain are shown with expanded scale.

Rfa1-MYC cells were treated as in (Fig. 1) or immuno-precipitated with anti-γ H2Aserum (Panels d–f). Data are average percent immuno-precipitated DNA ± one standard deviation. (a) Rfa1-MYC binding to short and WT length VII-L telomeres. Average enrichment at short VII-L was modestly higher than for WT VII-L but difference was only significant at 37.5 min (p= 0.02; p-values for other time points were 0.08 to 0.42). Rfa1-MYC binding occurred at time of telomere replication (60–75 min). (b) Rfa1-MYC binding to short VII-L telomere (open squares; data are from panel a), WT length VI-R (open triangles) and XV-L (open circles) telomeres in experimental strain was indistinguishable except at early points (p-values for 45–90 min 0.03 to 0.45). Data from control strain are also indistinguishable (p-values 0.17 to 0.98) from binding in experimental strain. (c) Rfa1-MYC binding to ARO1 locus is same in experimental (open triangles) and control (closed triangles) strains (p-values 0.19 to 0.95). (d) H2A phosphorylation was similar at short (open squares) and WT length (closed squares) VII-L telomeres except at 75 min (p= 0.04). (e) γ-H2AX levels at VI-R and XV-L telomeres are constant throughout cell cycle. Binding to telomeres is shown only for the experimental strain but values for both telomeres were indistinguishable in the control versus experimental strains (p= 0.08 to 0.98). (f) γH2A phosphorylation at RPL11A (diamonds) and ARO1 (triangles) in experimental strain.

(a) Schematic of assay. (b–e) Rif2 but not Rif1 content is lower at short versus WT length telomeres. Samples from three (Yku80 and Rap1) or five (Rif1 and Rif2) independent TLC1 or tlc1Δ spore clones were subjected to ChIP after ~25–30 generations of spore outgrowth. Samples expressing MYC-tagged proteins (panels b–e) were immuno-precipitated with an anti-Myc antibody. Samples from an untagged strain were immuno-precipitated with anti-Rap1 antibody (panel d). Purified DNA was analyzed by quantitative PCR to determine the level of protein binding to VI-R and XV-L telomeres. Data are expressed as the average fold enrichment of telomeric sequence over the non-telomeric ARO1 sequence in the same immuno-precipitate. Error bars indicate one standard deviation from average. The differences between binding levels of Yku80 (p= 0.354, VI-R; 0.840, XV-L) and Rif1 (p= 0.731, VI-R; 0.492, XV-L) at WT and short telomeres were not significant. Binding levels of Rap1 (p= 0.026, VI-R; 0.009, XV-L) and Rif2 (p= 1.2×10⁻³, VI-R; 1.3×10⁻⁶, XV-L) were significantly different at WT and short tlc1Δ telomeres.

(a) Schematic of assay. (b–c) After ~30 cell generations, 3–9 independent spores from tlc1Δ RIF1 RIF2, tlc1Δ rif1Δ, and tlc1Δ rif2Δ, all expressing Tel1-HA, were subjected to ChIP with anti-HA antibody. Average length of telomeric DNA in each sample was determined (representative gels for telomere XV-L, panel b). Data are expressed as percent decrease in the average telomere length of the immuno-precipitate compared to length of input. Error bars, one standard deviation from mean. Percent differences correspond to the following average absolute differences in telomere length in input versus Tel1 immuno-precipitated DNA: WT: 38 (VI-R); 76 (XV-L) bps; rif1Δ: 33 bps (VI-R); 78 (XV-L); rif2Δ: 18 (VI-R); 26 (XV-L) bps. (d) Model for Rif1 and Rif2 distribution along telomere. WT length telomere (top) of ~300 bp C_1–3A/TG_1–3 duplex DNA contains ~17 Rap1 binding sites (not all are shown), and a short single strand TG_1–3 tail. Short and WT telomeres contain equal number of heterodimeric Ku complexes and Cdc13-Stn1-Ten1 complexes. For simplicity, we show one Ku and one Cdc13 complex per telomere. Rif1 and Rif2 come to the telomere by interaction with Rap1. As telomeres shorten, they lose Rif2 before Rif1, suggesting that Rif1 is positioned centromere proximal to Rif2. Thus, 300 bp WT and 150 bp short telomere would have ~the same amount of Ku, Cdc13, and Rif1. The shorter telomere has ~50% of Rap1 and Rif2 levels as at WT telomere.