The mre11 A470 alleles influence the hereditability and the segregation of telosomes in Saccharomyces cerevisiae

Telomeres, the nucleoprotein complexes at the termini of linear chromosomes, are essential for the processes of end replication, end protection, and chromatin segregation. The Mre11 complex is involved in multiple cellular roles in DNA repair and structure in the regulation and function of telomere size homeostasis. In this study, we characterize yeast telomere chromatin structure, phenotypic heritability, and chromatin segregation in both wild-type [MRE11] and A470 motif alleles. MRE11 strains confer a telomere size of 300 base pairs of G+T irregular simple sequence repeats. This DNA and a portion of subtelomeric DNA is embedded in a telosome: a MNase-resistant non-nucleosomal particle. Chromatin immunoprecipitation shows a three to four-fold lower occupancy of Mre11A470T proteins than wild-type proteins in telosomes. Telosomes containing the Mre11A470T protein confer a greater resistance to MNase digestion than wild-type telosomes. The integration of a wild-type MRE11 allele into an ectopic locus in the genome of an mre11A470T mutant and the introduction of an mre11A470T allele at an ectopic site in a wild-type strain lead to unexpectedly differing results. In each case, the replicated sister chromatids inherit telosomes containing only the protein encoded by the genomic mre11 locus, even in the presence of protein encoded by the opposing ectopic allele. We hypothesize that the telosome segregates by a conservative mechanism. These data support a mechanism for the linkage between sister chromatid replication and maintenance of either identical mutant or identical wild-type telosomes after replication of sister chromatids. These data suggest the presence of an active mechanism for chromatin segregation in yeast.


Introduction
The DNA-RNP structure present at the end of all eukaryotic chromosomes, the telomere, is necessary for end synthesis and protection. Our insight has grown into the polymerases, telomerase regulators, and chromatin components that solve the end-replication problem and contribute to the telomere end-protection [1]. The major mechanism to compensate for replicative attrition is the RNP-reverse transcriptase, telomerase [2]. Telomerase catalyzes the addition of G+T sequences onto pre-existing 3' telomere ends. The activity of several stable In this study, we have used one allele in the A470 motif, mre11A470T, as a marker to separate heritability and telomere chromatin segregation [40]. Telosomes containing the Mre11 or Mre11A470T proteins display strikingly different MNase profiles. A cell population that produces only the mutant allele requires more time of MNase digestion than wild-type cells to cleave telomeres from chromatin. We show here that the association of Mre11A470T with the telosome is less than wild-type association. However, we cannot rule out an additional possibility that Mre11 or Mre11A470T also act in trans to influence telosome structure. We have compared telomere structure and hereditability phenotypes conferred by alleles located at the endogenous site [also called genomic site] with genes integrated at an ectopic site [Fig 1]. We have determined that only the product of the genomic locus is incorporated in the telosome and leads to heritability and conservative telosome segregation.

[A]
Telomere size was determined by Southern blot analysis of XhoI-fractionated yeast DNA from MRE11, mre11A470T, yku70Δ, and mre11Δ cells using a telomere-specific probe. Wild-type strain had telomeres 300 bp in length, while both mre11 mutants had telomeres centered at 150 bps. The yku70Δ telomeres were centered at 200 bp in length. The samples were flanked by λ size markers [NEB].
[B]Telosome formation in MRE11, yku70Δ, and mre11A470T cells. Chromatin isolated from wild-type, yku70Δ, and mre11A470T cells was digested with one unit/100 μl MNase for the times indicated, fractionated, and subject to Southern blot analysis using the A750 probe. The control was performed under the same conditions in the absence of enzyme. The position of 500 bp using the λ size marker is shown on the left and right of the gel. A gel splice is present due to the elimination of any empty lane adjacent to the yku70Δ set of data. The percentage of the signal that represents telomeres is shown below the gel.
The rap1-5 allele of RAP1 was generated as described [42]. Previous studies indicated that the integrated rap1-5 allele has a weak temperature sensitive [ts] phenotype at 37˚C [43]. To test the genetic interaction between MRE11 and RAP1, we constructed the double mutant, rap1-5 mre11A470T. All growth conditions with rap1-5 strains were conducted at 25˚C, the permissive temperature, or 30˚C, the semi-permissive temperature, respectively.

Media
Strains in the absence of selection were grown in YPD media or YPD with adenine supplementation [YPAD], to a general facilitator of growth [44]. Strains containing the Kan R cassette were grown in YPD + gentamycin. Strains auxotrophic for any metabolic marker was grown on synthetic minimal media containing the missing amino acid at standard concentrations [44].

Rad5 function
We tested the possible involvement of the rad5-535 allele, present in the W303 genetic background, in the generation of mre11A470T phenotypes. Ubiquitin is a major regulator of repair occurring during transcription [transcription-coupled repair]. Our concern derives from our use of another major repair complex, Mre11/Rad50/Xrs2, that is responsible the response to double-strand DNA damage repair and to some cases of residual transcriptional-coupled repair [45]. Whether there is a relationship between ubiquitin in transcription coupled repair and/or double strand break repair remains unknown. We therefore took a conservative approach and tested the MRE11 and mre11A470T strains for their dependence on the RAD5 gene. To test the possibility that rad5-535 influences the mre11A470T allele, the telomere size of a rad5: Kan R mre11A470T double mutant was compared to single gene mutations. The short telomere length and heritability phenotypes of mre11A470T were not influenced by the rad5-535 allele, indicating that the Rad5 is not involved in the telomeric Mre11A470T function. Wild-type strains containing only the rad5-535 allele also had no influence on the telomere size phenotype and heritability [S1 Fig].

Ectopic integration
The EcoR1/XbaI containing flanking DNA from MRE11 and mre11A470T cells were cloned into pRS306 [46], creating pRS306-8 and pRS306-9, respectively. The URA3 gene in pRS306 was cleaved by NcoI and transformed into strains carrying either MRE11 or mre11A407T loci at the endogenous "genomic" locus [on chromosome XIII]. Integration resulted in Ura3 + at the site of recombination at the ectopic site [also called an "ectopic allele" [e]] on chromosome V. In contrast, all endogenous genomic loci were Ura3in the absence of the ectopic allele. All transformants contained a single copy of either MRE11 or mre11A470T. Phenotypes were specified by the allele at the ectopic and genomic loci as e-MRE11 g-MRE11[AJL551], and e-mre11A470T g-MRE11 [AJL552], e-mre11A470T g-mre11A470T [AJL 553], and e-MRE11 g-MRE11 [AJL 554] strains.

Ectopic assays for additional A470 motif alleles
The additional A470 motif alleles used in this study, mre11V471A, mre11E479A, mre11K480A and mreK482T, were placed at the endogenous site by two step gene replacement of W303 with the pRS306-cloned alleles. These strains were integrated by homologous recombination of the MRE11 gene into the ectopic site [ura3] to form e-MRE11 g-mre11A471T, e-MRE11 g-mre11E479A, e-MRE11 g-mre11K480A, and e-MRE11 g-mreK482T [designated AJL555-AJL558, respectively].
After integration of each A470 motif allele, we routinely retested for the presence of the mutation. To that end, we used the Mre11 forward primer, [5'-CAAACGTATAGATAGATA TACCCAAT-3'], and the Mre11 reverse primer, [5'-GCTCCTCTCAAAATGGCATACCTT G-3'] for PCR amplification and DNA sequencing.

Measurement of growth rates and viability
We measured the growth rates of independent MRE11, mre11A470T, rap1-5, and mre11A470T rap1-5 strains at three temperatures [23˚C, 30˚C, and 37˚C] in three trials. Cells were grown overnight at 30˚C and inoculated into 5 ml of YPAD media at the differing temperatures. The Cellometer Vision fluorescent counting system [Nexelcom Inc.] was used to visualize both viable and inviable cells. Propidium iodide is capable of entering and staining only inviable cells. Thus, viability is given as [[total number of cells-the number of propidium iodide-stained cells] /total cells] x100. Cell growth was measured in the absence of propidium iodide.

Telosome analysis
We analyzed telosome structure as described [26,47]. Specifically, we aliquoted nuclei into 100 microliters [μl] MNase buffer with 1 mM CaCl 2 . A predetermined amount of MNase, based on preliminary studies, was incubated with each strain tested at 30˚C as a function of time [0, 30,60,100,140,190, and 240 seconds [secs], and the reaction was terminated. Negative controls were treated identically, but lacked MNase. The chromatin fraction of each culture was isolated and analyzed by Southern blot analysis using the 750 bp fragment, purified from plasmid Ap135 [A750, kind gift of Dr. Alison Bartuch], as the telomere probe. One of three size markers [the lambda size ladder, 1 [kilo-base] kb ladder and the 100 bp ladder [NEB] flanked each of the gels. Since chromatin experiments display subtle variations among independent experiments, we interpret the results of each experiment with all samples conducted in parallel on a quantitative basis, rather than the mean of data generated in differing experiments. Nonetheless, the pattern of size distributions is qualitatively reproducible between experiments.

Telosome quantification
Each lane of one set of gels was scanned by the CLIQ software [Total Lab] and the fragments quantified using the "rubber band" background setting per manufacturer's instructions. High and low signals [in pixels] should be proportional to their abundance, since tract sizes are not the major variable in determining the increases in hybridization intensity [unlike measurements of telomere tract size]. The total pixels remain relatively constant [+/-15%]. The intensity of telomere pixels/ the sum of all fragments gave rise to the % of the signal within the telomere size range. In some cases, the distribution varied slightly, and a broader range of a specific fragment was converted after increased treatment to a single species.

Telomeric DNA isolated from telosomes
We isolated DNA from the chromatin fraction by phenol: chloroform extraction and ethanol precipitation. Samples were digested with XhoI to determine telomere size after Southern blot analysis using the A750 as a probe. We determined the telomere size using the Y' end-point at 870 bp from the XhoI site to terminus of Y' [48].

Cycloheximide [CHX] protein stability assays
The CHX protein stability assay was performed in MRE11 and mre11A470T strains. In the first day, the cells were pre-inoculated overnight in YPD and re-inoculated in YPD to newly prepared YPD media after overnight growth. The cells were grown to mid-log phase, and then 100 μg of CHX was added, incubated with CHX for 0, 15, 30, 60, 90, 120, 160, 150, 180, and 240 minutes [min] at 30˚C. The proteins were extracted by the TCA method [49]. For Mre11 protein detection, a diluted [

Chromatin immunoprecipitation [ChIP] methodology
We followed the procedure of Hecht et al [31] using the following modifications: Cultures were scaled to 20 ml, and cells were incubated with 1% formaldehyde for 15 min at 37˚C before quenching. Cells were washed twice with cold PBS and suspended in collection of the lysate. Cells were sonicated three times for four secs at a setting of five [Dismembrator Model 100, Fisher]. The pellets were then centrifuged, and the supernatant containing the extract was collected. The immunoprecipitation used the ChIP-grade anti-Mre11 antibody [1/150] incubated overnight at 4˚C on a rotator. We pre-equilibrated the protein [A/G] agarose beads [Santa-Cruz; sc2003] for two hours [hrs] at 4˚C. We subsequently washed the beads twice with 500μl lysis buffer at room temperature Ã . The fragments were eluted from beads with 50 μl 100 mM NaHCO 3 , 1% SDS at 65˚C for 15 min. Centrifugation separated the beads from the eluate. 50 μl TE and 100 micrograms [μg] proteinase K were added, and precipitated by ethanol. The final DNA was dissolved in 30 μl TE and subjected to PCR using primers flanking a 200 bp sequence to amplify the X/ telomeric border using the X forward primer 5'-GGAGCAACTTGCGTGAATCGAAGA-3' and the Tel reverse primer : CTGTCGATGATGCTGCTAAACTG.

Real-Time PCR ChIP analysis of Mre11 abundance in telomeric chromatin
Cells were grown to OD600 = 0.2 in 20 ml of fresh liquid media to mid-log phase. Real-time qPCR was conducted according to the manufacturer's instructions [Bio-Rad] as summarized below. We cross-linked proteins in vivo using a 1% formaldehyde for 15 min in a shaker at room temperature. The formaldehyde reaction was then quenched with 125 mM glycine. The cells were washed twice with cold PBS and suspended in 400ul lysis buffer. Cells were lysed with glass beads in a bead-beater four times for 20 secs each, transferred to a new tube and sonicated twice for 10 secs. The lysate was collected in a new tube and centrifuged for 10 min at 4˚C. We immunoprecipitated the supernatant overnight at 4˚C with a 1/150 antibody dilution on a rotator. We subsequently added 25μl of a 50% slurry of the pre-equilibrated protein A/G. Agarose beads were incubated for two hrs on a rotator at 4˚C. The beads were pelleted and washed with twice with 500 μl lysis buffer for two times at room temperature, once with 500 μl deoxycholate buffer beads [10 mM Tris ph7.5 0.25 M LiCl and 0.5% sodium deoxycholate and 1 mM EDTA] at room temperature, and once with 500 μl, TE for five min. The fragments were eluted from beads at 65˚C in 50 μl elution buffer for 15 min. The supernatant was collected and transferred to a new tube. The cross-links were reversed after incubating at 65˚C in 50 μl elution buffer overnight.
The beads were removed from the supernatant, and 50 μl TE and 100 μg proteinase K were added and incubated for three hrs at 55˚C. One hundred μl TE and 200 μl PCIA were added, mixed well and centrifuged for five min at room temperature. The supernatant was ethanol precipitated for 30 min at -20˚C, and the precipitate isolated after spinning at 14,000 x g for 30 min at 4˚C. The pellet was dried and dissolved in 300 μl TE. Procedures are accessible to the public by the MIQE convention [50].
We conducted real-time qPCR [Bio-Rad] using the CFX96 Touch™ apparatus and the primers Xforward and Telreverse described above. The data were analyzed using the Bio-Rad CFX Manager™ Software and the ΔΔC value was derived. Cq values from the CFX Manager software were used to determine the ΔCq [Target Cq-reference Cq], ΔCq expression [2 -ΔCq ], mean ΔCq expression [average replicates], ΔCq expression standard deviation], and the ΔΔCq [normalized value to mre11Δ]. The ChIP analysis was carried out as before, and reverse crosslinking performed in a final volume of two microliters of ChIP DNA using two different dilutions of protein in 96 well plates that were subjected to qPCR.

Heritability and telosome structure in mre11A470T cells
The altered telosomes of mre11A470T alleles. MRE11 and mre11A470T telomere tract lengths were measured by gel electrophoresis of XhoI sensitive cleavage Y' elements that are 870 bp proximal to the telomere. The lengths were 300 bp and 150 bp, respectively, in Y' and X [40] classes and packaged in telomere chromatin particles, termed telosomes [Fig 1]. After MNase treatment of chromatin, wild-type cells protect a telosome DNA fragment of 400-600 bp, as reported [26]. In contrast, the mre11A470T allele telosome is more resistant to MNase and protects a more diffuse pattern of subtelomeric sequences [100-900 bp]. Quantifying the signals leads to the conclusion that the mre11A470T telosome is present at maximal abundance after 240 secs of digestion, rather than 140 secs for wild-type telomeres. These data indicate that the telosome extends into subtelomeric sequences.
Micrococcal nuclease resistance phenotype. The mre11A470T allele is more resistant than wild type to MNase. In the mre11A470T allele, extended times of digestion are required to reach the point of telosome release from chromatin. These data imply that the structure or folding of Mre11A470T and associated proteins protect telosomes from digestion. The delay in cleavage of mre11A470T chromatin is not a consequence of short telomeres, since yku70Δ telosomes has a wild-type level of resistance [ Fig 1B]. In contrast, the mre11 null allele confers a variety of chromatin structures from wild type to mutant, and may reflect multiple states of the telosome in the absence Mre11. As expected, DNA isolated from the chromatin fraction gave rise to the initial telomere sizes [ S2 Fig]. The effect of altered telosomes on telomere sequence. A change in the relative position of the telomeres may be the consequence of Mre11A470T-induced changes in the TLC1 telomerase RNA template region. The telomere sequence could be altered or constrained in the mutant strain. We therefore sequenced multiple independent clones of both wild type and mre11A470T strains. However, we find no changes in sequence or sequence structure]. The sequences of both the wild-type, [S3 Fig] and mre11A470T DNA [S4 Fig], vary between subclones, but follow the yeast consensus sequence G 1-3 T. Wild-type telomeres were conserved in the first 150 bp sequences distal to the terminus, but varied significantly in the terminal half due to deletions and sequence mismatches.
ChIP analysis of MRE11 and mre11A470T strains. To determine whether the Mre11 and Mre11A470T proteins are capable of binding to the telosome, we conducted a simple chromatin immunoprecipitation [ChIP] analysis using excess Mre11 and Mre11A470T [41]] [Fig 2A,  top]. Mre11 proteins were present as intact species of known size [80 kD] after Western analysis using the monoclonal antibodies directed against a C-terminal peptide of Mre11 [Abcam]. We crosslinked proteins in strains containing MRE11 or mre11A470T alleles, and we immunoprecipitated cellular extracts with the same anti-Mre11 antibody. After elution, de-crosslinking, and protease treatment, a DNA was released that could support the PCR amplification of a 200 bp product at the junction between the X element and telomeric sequences. This fragment is maintained within the telosome, based on telomere size analysis. The mre11Δ allele, on the other hand, generated no protein species or amplified DNA after ChIP analysis [Fig 2B]. These binding data reinforce that both Mre11 and Mre11A470T proteins are a part of the telosome.
To estimate the relative binding of Mre11 and Mre11A470T to the telosome at X class telomeres, we performed real-time qPCR ChIPs [Fig 2C]. DNA isolated from chromatin that was immunoprecipitated by anti-Mre11 [ABCAM] in MRE11, mre11A470T, or mre11Δ extracts. The immunoprecipitated DNA was subject to real-time PCR to quantify the 200 bp junctional fragment [as a proxy for telomeres] using the same PCR primers as described above. We found a 3-4-fold increased abundance of telosomes containing wild-type Mre11 over Mre11A470T from wild-type and mre11A470T cells.
An alternative explanation for the ChIP data is a four-fold decrease in the stability of the mutant protein.

Phenotype and structural characteristics of alleles at ectopic and genomic positions
Epistatic relationship between MRE11 and mre11A470T alleles in heritability and chromatin structure. In considering the ectopic/genomic system, we first had to resolve whether the allele at an ectopic site is silenced. Several lines of evidence argue against this possibility. First, a TAP-tag derivative of both wild type and mre11A470T was integrated at an ectopic site and produced proteins that are also stable in a CHX assay [ S5B Fig]. Second, the allele at the ectopic site can be expressed, since the wild-type allele at the URA3 locus on chromosome IX can complement an mre11: Kan R allele at the genomic site [see 4, right]. Third, based on realtime qPCR analysis, reverse transcription of genes located at both genomic and ectopic sites have a three-to four-fold increase over either strain carrying the genomic allele alone [ S6 Fig]. These data suggest that the genes at ectopic and genomic sites are both transcriptionally active. Hence, the ectopic sites produce transcripts and intact tagged proteins.
Strains carrying the wild-type gene at the endogenous genomic locus [g] were transformed and integrated with the mre11A470T gene at an ectopic [e] ura3 locus. Conversely, strains carrying the mre11A470T allele at the genomic locus were integrated with MRE11 at the ectopic ura3. The only phenotypes observed were the result of the gene location at the endogenous locus [Fig 3].
Both the telomere tract length and MNase digestion pattern of chromatin were analyzed in e-MRE11 g-MRE11 and e-mre11A470T g-MRE11 strains. In the opposing experiment, strains were analyzed in e-mre11A470T g-MRE11 and e-mre11A470T g-mre11A470T. We will term this type of strain as "ectopic/genomic" to reflect the positions of the two alleles. The heritability of short mutant telomeres in e-MRE11 g-mre11A470T and g-mre11A470T strains persisted over 100 generations [the longest period tested]. In contrast, the wild-type telomere lengths persisted in e-mre11A470T g-MRE11 and g-MRE11 strains [Fig 4].
We found similar MNase patterns and spacing [diffusion of chromatin] isolated from e-MRE11 g-mre11A470T, and e-mre11A470T g-mre11A470T, and isolated from e-mre11A470T g-MRE11 and e-MRE11 g-MRE11 cells sub-cultured for the same number of generations [ Fig  5]. This is similar to single mutants mre11A470T [maximal extent: 100 bp-900 bp] and yku70Δ [maximal extent 100 bp-900 bp] telosomes [Fig 1B]. In one case, an mre11A470T gene that integrated into an ectopic site in a MRE11 strain [Fig 5, [e-mre11A470T gMRE11; 240 min]] produced a "diffuse" wild-type telosome [maximal extent: 200 bp-1000 bp]. Although insufficient to make a statistically sound conclusion, the diffusion phenotype is biased towards short telomeres. These diffuse telomeres are likely to represent a broader range of MNase cleavage sites that are less defined to a specified size range. Thus, in most aspects, the final phenotype is close to the genomic locus, regardless of the presence of opposing alleles at ectopic sites. As a control, e-MRE11 complemented a null allele.
Assays of additional A470 alleles reveal their dominant epistasis when the A470T allele is at the genomic site and the MRE11 allele is at the ectopic site. Each A470 motif allele confers either shorter telomeres [V471A, E479A], that [similar to the mre11A470T allele,] [Fig 6] or larger intermediate telomere sizes [K480A, A482T] that have equivalent telomere sizes in the presence or absence of the wild-type allele. These data demonstrate that a common trait of A470 alleles confine the phenotypic expression to the genomic loci.

Rap1 and Mre11 interaction
Rap1 and Mre11 mutant proteins confer synthetically lethal at 37˚C. We have shown that Mre11 can be defined both as part MNase nuclease sensitivity and by ChIP analysis. The telosome has previously been characterized as a component of the telomeric chromatin [26,47,52]. Although we do not know the relative degree of association of each component, we will assume they are generally part of the same telosome particle. We therefore tested the genetic or physical association of Rap1 and Mre11 at a basic level. To this end, we examined the growth and viability characteristics of a double mutant containing the mre11A470T and a weak temperature sensitive [ts] allele of RAP1, rap1-5. The rap1-5 was chosen, in place of more severe rap1 alleles, since rap1-5 defects would allow changes in double mutants to be observed before lethality. MRE11, RAP1, MRE11 rap1-5, mre11A470T RAP1, and mre11A470T rap1-5 cells were assayed for both cell growth viability and cell growth rate. Viability [as measured by Telosome heritability and chromatin structure in yeast the % of viable cells] was tested for cells at 23˚C, 30˚C and 37˚C, reflecting the permissive, semi-permissive and restrictive temperatures of rap1-5. Viability was not altered at 23˚C and 30˚C in any strain, while mre11A470T grown at 37˚C decreased viability to 90%, and the double mutant to 60% [Fig 7]. [The rap1-5 gene did not confer a viability phenotype over this To check whether the wild-type gene at an ectopic site still behaves phenotypically as a wild type, despite an mre11Δ null allele, genomic DNA was extracted by phenol: chloroform: isoamyl alcohol extraction, precipitated with ethanol, digested with XhoI and subjected to Southern Analysis using the A750 probe. The e-MRE11 allele complements the short telomere phenotype of a mre11Δ strain, indicating that the ectopic site is expressed.
https://doi.org/10.1371/journal.pone.0183549.g004 Chromatin characterization from ectopic/genomic strains. Telosome formation of e-MRE11 g-MRE11, e-mre11A470T, g-mre11A470T, e-MRE11g-mre11A470T, e-mre11A470T g-MRE11, and the controls MRE11 and mre11A470T were prepared as described in 2. The chromatin was extracted and digested by one unit /μl MNase at the designated time points. The gel was subjected to Southern blot analysis using A750 as the probe. The percentage of the signal that represents telomeres is shown below the gel lanes.
https://doi.org/10.1371/journal.pone.0183549.g005 short time interval]. This non-epistatic relationship between the viability conferred by the phenotypes of mre11A470T and rap1-5 suggests that Rap1 and Mre11 act independently on the telomere through different pathways or form a toxic complex that decreases viability.
Rap1 and Mre11 pathways influence viability [Fig 7] and growth rate [Fig 8] in distinct fashions. Growth rates were more significantly affected. At all temperatures, mre11A470T rap1-5 and mre11A470T cell growth rates were reduced, with the double mutant having the most severe phenotype at each temperature [Fig 8]. The rap1-5 strain showed no change in viability at any temperature within this time frame [Fig 7]. Only the mre11A470T and mre11A470T rap1-5 strains were decreased in viability, albeit after differing times. Yet, cells stop growth before viability is lost, suggesting a window in which other events after growth cessation can still influence viability. Given the proteins involved, it is likely that either the lack of viability or the length of the cell cycle is the consequence of telomere-mediated genomic instability.
Chromatin structure of rap1-5 mre1A470T telomeres. To correlate growth rate with the telosome structure of the double mutant, we conducted a Southern blot of the products of MNase digestion in wild-type, mre11A470T, rap1-5 and mre11A470T rap1-5 cells grown at the semi-permissive temperature, 30˚C [Fig 9]. Both the mre11A470T rap1-5 and mre11A470T chromatin are relatively resistant to MNase until 100-240 secs, when both rapidly form telosomes. Wild-type and rap1-5 cells start to form telosomes at an earlier time than mre11A470T. Hence, rap1-5 does not facilitate more resistance to MNase. This is consistent with the small decrease in growth rate at 30˚C in mre11A470T cells and the double mutants. We are somewhat disadvantaged by the inability to analyze chromatin structure at 37˚C, given the high levels of unviability. This difference in sensitivity of different samples to reaction conditions of a single experiment is less than the difference in sensitivity to inherent variability between experiments. Nonetheless, the same qualitative pattern is maintained between identical experimental trials. The epistatic dominance of the A470 motif alleles at the genomic locus to wild-type. Several additional A470 motif alleles, mre11V471A, mre11E479A, mre11K480A and mre11K482T in the presence of MRE11 at the ectopic ura3 locus. A Southern blot analysis of XhoI-digested genomic DNA, defining the Y' class of telomeres, was used to determine telomere size. A470 motif mutant strains could not be complemented by e-MRE11, as is the case for the mre11A470T allele. Wild-type strains were used as controls. Arrows point to wildtype and mutant strains that were used as controls. https://doi.org/10.1371/journal.pone.0183549.g006

Discussion
Several lines of evidence indicate that telosome structure, hereditability and segregation are aberrant in the presence of the mre11A470T allele. First, an extended time of MNase cleavage is required to generate the Mre11A470T telosome. Real-time qPCR has shown a four-fold greater abundance of Mre11 than Mre11A470T in unsynchronized cells. This difference is not due to variations in protein stability.
Second, the mutant may not just alter structure, but could also facilitate the influence of a chaperone or regulatory RNA, such as hsp90 [53]] and TERRA [ [54,55]], respectively. TERRA in yeast is currently an experimental 'black box' that may be the key to understand higher-level G-4DNA/telomeric DNA/ complexes [56]. Third, we have previously shown that the mre11A470T allele favors recombination over telomerase activity at chromosomal telomeres, based on the inability to heal a telomere seed sequence [41]. Indeed, ctf18 mre11A470T cells can confer a complete loss in telomerase activity in vivo [41]. This finding is consistent with the bypass of telomerase senescence in mre11A470T tlc1Δ, the apparent consequence of elevated telomere recombination [35]. One speculative possibility is that mre11A470T requires a cofactor to act directly on recombination. Alternatively, the structures of 3' termini may be a deciding factor in the activation of telomerase through the binding of different factors that influence heritability, telosome segregation, and genomic stability.
Finally, since Mre11 is dimeric, the conservative segregation pattern of telosome observed in mre11A470T cells must be related to the Mre11 and Mre11A470T homodimer and heterodimer stability. Alternatively, the absence of a chromatin modification may prevent the incorporation of heterodimers. This possibility has precedent from positive and negative effects on transcription elicited by histone modification within a microenvironment (e.g., [36,37]).

The heterodimer paradox
The lack of a phenotype generated from the ectopic locus is paradoxical since both genomic and ectopic alleles are expressed. MRE11 at an ectopic site can complement genomic mre11Δ alleles, the combination of the genomic and ectopic loci increases the steady state level of transcription, and a Tap-tagged MRE11 or mre11A470T can be expressed at the ectopic site. Proteins encoded by MRE11 or mre11A470T at the genomic and ectopic sites are also equally capable of translation [ S5 Fig]. For example, heterodimers must be different, depending upon the genomic allele, in order to produce only two possible phenotypes from all allelic combinations. Specifically, the e-mre11A470T g-MRE11 (1) and e-MRE11 g-mre11A470T (2) should both produce a mre11A470T/ MRE11 heterodimer. If so, they have distinct phenotypes. Possibly, Mre11/Rad50 interaction, 3' end formation, or heterodimer instability interfere with the production of heterodimers.

Conservative segregation, the heterodimer problem and heritability
A conservative segregation pattern is a general requirement for telomeric chromatin, regardless of the specific mechanisms involved. There is precedence for conservative segregation or spreading of heterochromatin that is caused by specific base modifications [57].
In order to explain our results, we must assume that heterodimers are dysfunctional in an activity or in stability. Alternatively, there may be a strong selection for homodimer formation on telosomes. Telomeres have non-redundant mechanism of telomere cohesion, and may be specific to sister chromatid cohesion. Thus, separation occurs first at G2. Possibly, alignment of sister chromatids contributes for cohesin-mediated transient sister chromatid separation in G2 [Fig 10]. Alternatively, a closer proximity of the sister alleles may help to maintain the preexisting state.

Characterization of Mre11A470T
The mre11A470T allele confers a unique chromatin structure. The telosome that contains the Mre11A470T protein requires longer MNase incubation for the release of the mutant particle from the total chromatin. Also, relative to wild type, the distribution is more diffuse than expected. This distribution suggests conformational changes that result in promiscuous subtelomeric MNase cut sites. However, this argument oversimplifies the mechanism since we have identified an expected wild type phenotype that is diffuse. The source of this property remains unknown. However, one speculative mechanism is a region of lowered stability or altered conformation, such as a fold-back structure [38], formed between an mre11A470T telomere and the subtelomeric region. The fold-back structure may have a lowered fidelity of MNase cut sites, and as a consequence, a broad range of telomere sizes after digestion. Telosome behavior in rap1-5 mre11A470T cells. Cells for telosome analysis were grown at 30˚C. Chromatin was isolated from MRE1I, mre11A470T, rap1-5, and mre11A470T rap1-5 and digested with one unit /100 μl MNase for the indicated time [in seconds] by Southern analysis using telomere sequences as the probe. Both mre11A470T and the double mutant behave similarly at 30˚C. We cannot rule out some subtle difference between these two strains. An asterisk refers to values greater than 0 but having insufficient pixels for quantification. https://doi.org/10.1371/journal.pone.0183549.g009

Rad50/Mre11 interactions
One of the reasons for our interest in the Mre11A470T protein is its proximity to Mre11/ Rad50 interface in eukaryotic crystal structures [39]. Indeed, the "dominant" epistatic pattern is a common feature of the A470 alleles. In each case, a wild-type gene integrated at an ectopic site failed to produce wild-type telomeres. This feature suggests their common involvement of the A470 motif in related functions, possibility regulating the Mre11A470T/Rad50 structure.

Synergistic effects of rap1-5 and mre11A470T
The rap1-5 mre11A470T double mutant is synergistic at all temperatures in growth rate, and at 37˚C in viability. Rap1 and Mre11 appear to act on the same substrate through differing pathways, both of which affect growth rate. Pathways are less clear in the telosome profiles at 30˚C, in which mre11A470T is epistatic to rap1-5. Mre11 may play a role in the equilibrium between Tel1 and Rif1 and in size homeostasis [16,59]. Mre11 may participate in the formation of homeostasis, while Rap1 may play a role in protection against non-homologous end-joining [60,61]. Experiments that examine the dynamics of wild-type strains are needed to resolve this issue.  Exchange between chromatids is shown by the green arrow. A sister chromatid specific mechanism of telomere separation in G2 may be mediated by a unique mechanism that is essential for subsequent separation [58]. The choice of ectopic vs genomic allele may be the consequence of an irreversible dimerization process that is established before integration of the ectopic allele.

Supporting information
https://doi.org/10.1371/journal.pone.0183549.g010 Telosome heritability and chromatin structure in yeast subjected to Southern blot analysis using the A750 probe. Two exposures are shown to visualize all fragments. (TIF) S3 Fig. Telomere sequencing in mre11A470T strains. We sequenced multiple X-class telomeres from independent clones of wild-type or mre11A470T cells using the C-tail method [51] with the X-class primers adjacent to a telomere tract. The X-class telomeres represent a group of 13 telomeres that have identical junctions with the telomere. S3A: Sequence of three clones of a mre11A470T was conducted on independent telomeres derived from a single source. The color code is present simply for convenience and is not meant to indicate causality. To test the transcriptional expression level of overall Mre11 in strains carrying two loci [genomic and ectopic], real time-qPCR of the reverse transcribed RNA was conducted as described in Materials and Methods. The first, second, and third bars refer to WT [g-MRE11] [as a control], e-MRE11 g-MRE11 and e-mre11A470T g-MRE11 strains, respectively. The expression of both e-MRE11 g-MTE11 and e-mre11A470T g-MRE11 were three-to four-fold greater than strains having only a single genomic locus. The Y axis represents the fold increase in transcripts relative to wild type. (TIF)