Smc5/6 Is a Telomere-Associated Complex that Regulates Sir4 Binding and TPE

Abstract

SMC proteins constitute the core members of the Smc5/6, cohesin and condensin complexes. We demonstrate that Smc5/6 is present at telomeres throughout the cell cycle and its association with chromosome ends is dependent on Nse3, a subcomponent of the complex. Cells harboring a temperature sensitive mutant, nse3-1, are defective in Smc5/6 localization to telomeres and have slightly shorter telomeres. Nse3 interacts physically and genetically with two Rap1-binding factors, Rif2 and Sir4. Reduction in telomere-associated Smc5/6 leads to defects in telomere clustering, dispersion of the silencing factor, Sir4, and a loss in transcriptional repression for sub-telomeric genes and non-coding telomeric repeat-containing RNA (TERRA). SIR4 recovery at telomeres is reduced in cells lacking Smc5/6 functionality and vice versa. However, nse3-1/ sir4 Δ double mutants show additive defects for telomere shortening and TPE indicating the contribution of Smc5/6 to telomere homeostasis is only in partial overlap with SIR factor silencing. These findings support a role for Smc5/6 in telomere maintenance that is separate from its canonical role(s) in HR-mediated events during replication and telomere elongation.

Fig 1. Smc5/6 is a telomere binding complex.

(A) A schematic representation of the Smc5/6 complex showing the location of Nse3 as part of a trimeric sub-complex located at the head region where Smc5 and Smc6 meet. (B) Chromatin immunoprecipitation (ChIP) followed by qPCR was performed on Smc6^FLAG (JC1594) at the indicated time points after release from α-factor. The fold enrichment at three native subtelomeres (Tel1L, Tel6R and Tel15L) compared to a control (ctrl) late replicating region on Chromosome V (469104–469177) is reported with the mean ± SD for n≥3 experiments performed in technical duplicate. (*) Indicates a statistically significant level of enrichment compared to the ctrl with p values < .05 by a two-tailed t-test. Smc6^FLAG enrichment at Tel1L is higher at 0 and 15 minutes after release, but with p values = 0.08 and p = 0.06 respectively. The lower panels show flow cytometry on ChIP samples with an asynchronous culture shown in black at the 0 time point. (C) Drop assay of exponentially growing wild type (JC470) and nse3-1 (JC3607) cells that were grown for 48 hours at the indicated temperatures on YPAD and 1:5 serial dilutions. (D) Schematic diagram of Nse3. “MHD” represents Melanoma Homology Domain in Nse3 protein. Seven amino acid substitutions in Nse3-1 are shown in red. (E) Chromatin immunoprecipitation (ChIP) on Smc6^FLAG in wild type (JC1594), nse3-1 (JC2630), mms21-11 (JC2075) and the non-tagged (nt) control strains for wild type (JC470), nse3-1 (JC3607), and mms21-11 (JC1879) in asynchronous cultures. The fold enrichment levels are relative to the late-replicating control region on Chr V for n = 3 experiments with the mean ± SD. All primers are listed in S2 Table. Enrichment levels for wild type and mutant cells with p values < .05 from a two-tailed t-test are indicated by (*). (F) Telomere length was determined as previously described [15]. Southern blot analysis was performed on 1μg XhoI-digested genomic DNA hybridized with a radiolabeled poly (GT/CA) probe in wild type (JC471), nse3-1 (JC3032), mms21-11 (JC1981), and smc6-9 (JC1358).In higher eukaryotes, telomeres are challenged by the continuous loss of DNA due to the end replication problem. However, in Saccharomyces cerevisiae, telomere length is maintained by the continued expression of telomerase, an enzyme containing a RNA subunit that serves as a template for de novo telomere synthesis [16]. After the 3’ end is extended by telomerase, the replicative DNA polymerase fills in the complementary strand. Both telomerase extension and semiconservative replication at telomeres are included in the final events of S phase (for review see [17]). In the absence of telomerase activity, telomeres shorten extensively, leading to senescence, however a small percentage of cells survive by extending their telomeres through the HR dependent alternative lengthening of telomeres (ALT) pathway [–21].

Fig 2. Smc5/6 is critical for telomere clustering and Sir4 binding to telomeres.

(A) Rap1-GFP foci in WT (JC1822) and nse3-1 (JC3041) cells counted as a measurement for telomere clustering with representative merged images of GFP and DIC channels. (B-C) The number of GFP-Rap1 foci was determined for cells within G1 (unbudded) or S (small budded cells) phases in at least 100 cells for each cell cycle stage, and (D) compared with mms21-11 (JC1827) and smc6-9 (JC2710). (E-F) Western blot analysis and immunofluorescence staining using α-Myc antibody (green in IF) to detect Sir4^Myc in WT (JC3433), nse3-1 (JC3452), mms21-11 (JC3597), and smc6-9 (JC2907) cells with DAPI staining shown in blue. (G) ChIP was performed on Sir4^Myc as in Fig 1E from asynchronous cultures and in more than one isogenic strain if available. The fold enrichment for each strain is calculated for n≥3 experiments with the mean ± SD at three native subtelomeres (Tel1L, Tel6R and Tel15L). The p values < 0.05 from a two-tailed t-test are indicated by (*) for wild type (JC2671 and JC3433), nse3-1 (JC3452 and JC3849), mms21-11 (JC3597), and smc6-9 (JC2907 and JC3087) and non-tagged (nt) control strains included wild type (JC470), nse3-1 (JC3607), mms21-11 (JC1879), and smc6-9 (JC1358).

Fig 3. Smc5/6 physically associate with Sir4 and is important for TPE.

(A) Co-immunoprecipitation (Co-IP) as described in the materials and methods section was performed in cells carrying Sir4^Myc and Nse3^HA (JC3736) with Nse3^HA (JC2823) as control or (B) Sir4^Myc and Smc6^FLAG (JC3853) with Smc6^FLAG (JC1594) as a control. (C) ChIP was performed on Smc5^FLAG in wild type (JC3728) and sir4Δ (JC3720) and (D) Smc6^FLAG in wild type (JC1594) and sir4Δ (JC3732) and non-tagged (nt) strains in wild type (JC470) and sir4 Δ (JC3737) as described in Fig 1E. The fold enrichment levels are relative to the late-replicating control region on Chr V for n≥3 experiments with the mean ± SD at three native subtelomeres (Tel1L, Tel6R and Tel15L) with p values < .05 from a two-tailed t-test indicated. (E) TPE was determined in strains with the URA3 reporter at the adh4 locus of Chromosome VIIL. Tenfold (1:10) serial dilutions of overnight cultures were spotted onto SC (complete medium) and SC + .1% 5-FOA plates at 25°C and 34°C in wild type (JC1991), sir4Δ (JC3818), nse3-1 (JC3860), mms21-11 (JC1080) and smc6-9 (JC1077) isogenic strains.

Fig 4. The nse3-1 allele exhibits genetic interactions with the loss of SIR4 and RIF2.

(A) TPE was determined in strains with the URA3 reporter at the adh4 locus of Chromosome VIIL as in Fig 3E. Overnight cultures were spotted onto SC (complete medium) and SC + .1% 5-FOA plates at 34°C in wild type (JC1991), sir4Δ (JC3818), nse3-1(JC3860), nse3-1 sir4Δ (JC3870) isogenic strains. (B) Transcription levels in wild type (JC470), nse3-1 (JC3607), sir4Δ (JC3737), and nse3-1 sir4Δ (JC3741), and (C) rif2Δ (JC2992) and nse3-1 rif2Δ (JC3269) at sub-telomeric genes CHA1 and VAC17 on Tel3L and YR043C on Tel9R as described in [62, 63]. Expression values are mRNA levels relative to ACT1 and normalization to wild type cells. Error bars represent ± SD of n = 3 experiments with p values < .05 from a two-tailed t-test indicated by (*). (D) Chromatin immunoprecipitation (ChIP) was performed on Rif2^Myc and showed similar levels of recovery in wild type (JC2380) and nse3-1 (JC3235) mutants. (E) ChIP on Smc6^FLAG in wild type (JC1594), rif1Δ (JC2754) and rif2Δ (JC3074) cells with enrichment levels for untagged strains in wild type and mutants shown in S5C and S5D Fig. The mean ± SD of the fold enrichment at three native subtelomeres (Tel1L, Tel6R and Tel15L) relative to the control (ctrl) late replicating region on Chromosome V (469104–469177) is reported. In rif2Δ mutants the p values < .05 = 0.53 (Tel1L), 0.13 (Tel6R), and 0.15 (Tel15L) indicated that the difference was not significant from wild type. (F) Telomere length was determined as previously described [15]. Southern blot analysis was performed on 1μg XhoI-digested genomic DNA hybridized with a radiolabeled poly (GT/CA) probe in wild type (JC470), nse3-1 (JC3607), rif1Δ (JC3448), nse3-1 rif1Δ (JC3623), rif2Δ (JC2992), nse3-1 rif2Δ (JC3269), rad52Δ (JC1427), nse3-1 rad52Δ (JC3629), rif2Δ rad52Δ (JC3603), and nse3-1 rad52Δ rif2Δ (JC3627) strains.

Fig 5. Increases in TERRA and telomere shortening are additive in nse3-1 sir4Δ double mutant cells.

(A and B) TERRA expression was determined by RT-qPCR for Tel1R and Tel6R, X only telomeres, at 28°C and 34°C in wild type (JC470), nse3-1 (JC3607), rif2Δ (JC2992), nse3-1 rif2Δ (JC3269), sir4Δ (JC3737), nse3-1 sir4Δ (JC3741), and sir4Δ rif2Δ (JC3738). TERRA expression from Y’ telomeres is shown in S8 Fig. Statistical significance with p values < .05 (*) or < .01(**) are reported from a two-tailed t-test. (C) Telomere length was determined as in Fig 1F by Southern blot analysis on 1μg XhoI-digested genomic DNA hybridized with a radiolabeled poly (GT/CA) probe in wild type (JC470), nse3-1 (JC3607), sir4Δ (JC3737), and nse3-1 sir4Δ (JC3741). (D) A model comparing telomere organization in wild type and nse3-1 mutants. The Smc5/6 complex localizes to telomeres but significantly decreases in nse3-1 mutants (Fig 1E). Moreover, nse3-1 alleles exhibit shorter telomeres, reduced telomere clustering, reduced Sir4 binding and defects in TPE. When nse3-1 is combined with the loss of SIR4, the resulting double mutant cells show additive defects in transcriptional repression and telomere shortening.