Set1 regulates telomere function via H3K4 methylation–dependent and -independent pathways and calibrates the abundance of telomere maintenance factors

Abstract

Set1 is an H3K4 methyltransferase that comprises the catalytic subunit of the COMPASS complex and has been implicated in transcription, DNA repair, cell cycle control, and numerous other genomic functions. Set1 also promotes proper telomere maintenance, as cells lacking Set1 have short telomeres and disrupted subtelomeric gene repression; however, the precise role for Set1 in these processes has not been fully defined. In this study, we have tested mutants of Set1 and the COMPASS complex that differentially alter H3K4 methylation status, and we have attempted to separate catalytic and noncatalytic functions of Set1. Our data reveal that Set1-dependent subtelomeric gene repression relies on its catalytic activity toward H3K4, whereas telomere length is regulated by Set1 catalytic activity but likely independent of the H3K4 substrate. Furthermore, we uncover a role for Set1 in calibrating the abundance of critical telomere maintenance proteins, including components of the telomerase holoenzyme and members of the telomere capping CST (Cdc13-Stn1-Ten1) complex, through both transcriptional and posttranscriptional pathways. Altogether, our data provide new insights into the H3K4 methylation–dependent and -independent roles for Set1 in telomere maintenance in yeast and shed light on possible roles for Set1-related methyltransferases in other systems.

INTRODUCTION

As the catalytic component of the COMPASS complex, the methyltransferase Set1 deposits mono-, di-, and trimethylation onto histone H3 lysine 4 (Miller et al., 2001; Roguev et al., 2001; Krogan et al., 2002; Nagy et al., 2002). In yeast, H3K4me3, me2, and me1 are found at gene promoters, gene bodies, and 3′ regions of actively transcribed genes, respectively (Bernstein et al., 2002; Santos-Rosa et al., 2002; Pokholok et al., 2005). Interestingly, in the absence of Set1, genome-wide gene expression analyses have shown that the majority of differentially expressed genes are up-regulated (Venkatasubrahmanyam et al., 2007; Lenstra et al., 2011; Margaritis et al., 2012; Weiner et al., 2012; Jaiswal et al., 2017; Jezek et al., 2017a), indicating that a primary outcome of depleted Set1 and H3K4 methylation is gene derepression or activation. The genes that rely on Set1 for repression are predominantly lowly expressed and are found in genomic regions with limited transcriptional activity, such as Ty elements, ribosomal DNA, meiotic differentiation genes, and telomeres (Nislow et al., 1997; Briggs et al., 2001; Bryk et al., 2002; Berretta et al., 2008; Castelnuovo et al., 2014; Jaiswal et al., 2017).

One of the earliest described phenotypes in yeast cells lacking Set1 was the loss of subtelomeric gene silencing (Nislow et al., 1997; Corda et al., 1999; Krogan et al., 2002). set1Δ cells also have short telomeres and disrupted telomere clustering at the nuclear periphery (Nislow et al., 1997; Krogan et al., 2002; Schneider et al., 2005; Trelles-Sticken et al., 2005). Given the low levels of H3K4 methylation at telomeric regions, it was proposed that Set1 indirectly regulates subtelomeric gene silencing, and likely telomere length, through euchromatic H3K4me3, which is thought to prevent redistribution of the silencing protein Sir2, the histone deacetylase (HDAC) of the SIR complex, from subtelomeres to internal chromosomal sites (Santos-Rosa et al., 2004; Venkatasubrahmanyam et al., 2007). However, subsequent chromatin immunoprecipitation and genetic studies indicate that subtelomeric gene repression by Set1 predominantly occurs through a SIR-independent mechanism (Ng et al., 2003; Leung et al., 2011; Margaritis et al., 2012; Jezek et al., 2017a). These observations are consistent with recent understanding of SIR-mediated silencing at subtelomeres, which revealed largely noncontiguous subtelomeric regions reliant on the SIR complex for silencing, whereas Set1 appears to promote silencing in larger, contiguous regions and at a higher number of subtelomeres (Ellahi et al., 2015; Jezek et al., 2017a). These observations indicate that the precise role for Set1 in maintaining functional telomere-linked properties such as gene repression, telomere length, and nuclear position remains obscure. Interestingly, in previous work, our lab has shown that the set1Δ transcriptome is highly similar to mutants lacking telomere maintenance factors, particularly the telomerase subunit, Est3 (Jezek et al., 2017a). This further suggests that Set1 may have a more direct role in telomere regulation than previously appreciated.

Although most known functions of Set1 have been attributed to methylation of H3K4, there is growing evidence for H3K4me-independent functions for Set1. For example, two other substrates have been described for Set1: the kinetochore protein, Dam1 (Zhang et al., 2005), and another histone modification at H3K37, where Set1 promotes methylation with Set2 (Santos-Rosa et al., 2021). Although not extensively documented for yeast Set1, some orthologues of Set1 have been demonstrated to perform noncatalytic functions as well (Kirmizis et al., 2007; Dorighi et al., 2017; Rickels et al., 2017, 2020). The discovery of these additional substrates and functions implicates Set1 in genome regulation roles separate from its H3K4 methyltransferase activity.

In wild-type cells, multiple protein complexes bind telomere ends and coordinate telomere capping and elongation. These include the Cdc13-Stn1-Ten1 (CST) capping complex and telomerase, composed of the catalytic subunit Est2, regulatory subunits Est1 and Est3, and the RNA component TLC1 (Wellinger and Zakian, 2012; Kupiec, 2014; Jezek and Green, 2019). These and other proteins are critical to telomere health, and numerous mechanisms appear to control their precise abundance in cells. This includes cell cycle–regulated transcription of genes encoding telomerase and CST subunits, ribosomal frameshifting and nonsense-mediated decay to control translation, and posttranslational regulation including control of protein degradation via phosphorylation (Morris and Lundblad, 1997; Enomoto et al., 2004; Larose et al., 2007; Tuzon et al., 2011; Advani et al., 2013; Holstein et al., 2014; Liu et al., 2014; Tucey and Lundblad, 2014; Cesena et al., 2017; Gopalakrishnan et al., 2017; Mersaoui and Wellinger, 2019; Garcia et al., 2020). These studies have revealed that changes in their stoichiometry alter telomere length and stability, indicating that a precise and well-defined balance of telomerase and CST components is critical to telomere maintenance.

In this study, we assayed mutations in Set1, the COMPASS complex, and at H3K4 to assess the specific contribution of H3K4 methylation by Set1 to subtelomeric gene repression and telomere length maintenance. Our data indicate that the catalytic core of Set1 is required for its role in both gene repression and telomere length; however, there are likely H3K4 methyl–independent mechanisms required for the regulation of telomere length. Furthermore, our data identify a role for Set1 and H3K4 methylation in modulating the abundance of telomere maintenance proteins, including components of telomerase and the CST complex, through both transcriptional and posttranscriptional pathways. Altogether, these data provide new insights into how Set1 contributes to telomere maintenance and are likely applicable to related proteins in other organisms.

RESULTS

The Set1 SET domain maintains telomere length and subtelomeric gene repression

In the absence of Set1, cells display a number of telomere-related defects (Nislow et al., 1997; Corda et al., 1999; Krogan et al., 2002; Schneider et al., 2005; Trelles-Sticken et al., 2005; Jezek et al., 2017a), including loss of subtelomeric gene repression as demonstrated by derepression of TEL07L-adjacent genes COS12 and YGL262W (Figure 1A) and short telomeres (Figure 1B). In investigating other telomere-related phenotypes of set1Δ cells, we also observed a faster rate of senescence in cells lacking Set1 and the telomerase component TLC1, as compared with tlc1Δ single mutants (Figure 1C). These data show that the telomeres in set1Δ tlc1Δ cells are shortened critically to reach a crisis point, prompting cellular senescence. The set1Δ tlc1Δ cells are able to recover, however, and form survivors through a telomerase-independent mechanism, albeit at a slower rate than tlc1Δ cells (Figure 1C). We also found that loss of Set1 exacerbates the growth defect of cells carrying a temperature-sensitive allele of CDC13, which encodes a member of the CST telomere capping complex (Figure 1D). While cdc13-1 cells show almost no growth at 30°C and some growth at 26°C, the set1Δ cdc13-1 mutants grow much more poorly at 26 and 27°C compared with either single mutant, indicating that Set1 functions in a pathway related to that of Cdc13. This is consistent with our previous observations indicating a broader role for Set1 in telomere maintenance pathways (Jezek et al., 2017a), rather than a targeted function in controlling subtelomeric gene silencing.

FIGURE 1:

Loss of Set1 is linked to telomere defects. (A) RT-qPCR analysis of TEL07L genes (COS12, YGL262W, and ADH4) in wild-type (WT; yEG230) and set1Δ (yEG232) cells. ADH4 is farther from the telomere and not subject to repression by Set1. Expression normalized to the control gene TFC1 and relative to WT is shown. Error bars represent SEM for a minimum of three biological replicates. Significance was evaluated using two-way ANOVA and Tukey’s multiple comparisons test. P values are indicated as follows: *** < 0.001. (B) Southern blot showing terminal telomere fragment molecular weight in WT (yEG230) and set1Δ (yEG232) cells. The terminal telomere fragments were recognized with probe 5′-biotin-CACACCCACACCCACACC-3′. (C) Senescence rate of WT, set1Δ, tlc1Δ, and set1Δ tlc1Δ cells, which originated from tetrad dissection of yEG880. At each time point, population doublings were calculated based on the cell density. Three biological replicates are shown. (D) Fivefold serial dilutions of saturated cultures of the indicated strains were spotted on YPD and incubated at different temperatures before imaging. WT (DLY3001), set1Δ (yEG1296), cdc13-1 (DLY4557/yEG1297), and set1Δ cdc13-1 (yEG1298). All yeast strains described in this figure are derived from the BY4741 background, except in panel D, which are W303.

To better define the molecular contribution of Set1 in preventing telomere-related defects, we monitored subtelomeric silencing and telomere length in mutants specifically lacking H3K4 methylation by carrying an allele expressing the H3K4R mutant as the sole copy of an H3-encoding gene. These cells showed derepression of subtelomeric genes similar to that of set1Δ cells (Figure 2A) and also a similar shortening in telomere length (Figure 2B). We combined the H3K4R mutant with set1Δ and also observed that the subtelomeric gene repression and telomere length were reduced to similar extents in the double mutant compared with both single mutants (Figure 2, A and B). While this suggests an epistatic interaction between SET1 and H3K4, as is expected based on their biochemical relationship, previous work has demonstrated that the Set1 protein is destabilized in the absence of H3K4 methylation in these mutants due to an autoregulatory feedback mechanism (Soares et al., 2014). In addition, the decreased dosage of histones in these strains is likely to modify set1Δ phenotypes, as demonstrated previously (de La Roche Saint-André and Géli, 2021), indicating that it is challenging to conclusively determine the relationship between Set1 and H3K4 methylation using these strains. We therefore used another approach to address whether the methyltransferase activity of Set1 is required for subtelomeric gene repression and telomere length maintenance by generating multiple catalytically inactive mutations of Set1. We used cells expressing FLAG-SET1 from a plasmid under the control of the SET1 promoter in set1Δ cells and generated the H1017L, C1019A, and G951S mutants (Nislow et al., 1997; Schlichter and Cairns, 2005; Ramakrishnan et al., 2016; Cruz et al., 2018). While we did observe complete loss of H3K4 methylation in these mutants, we also found that their steady-state protein levels were much lower than wild-type FLAG-Set1 (Figure 2D), as previously shown for some catalytically inactive mutants (Soares et al., 2014). As expected, the mutants all showed derepression of subtelomeric genes at TEL07L and TEL09R (Figure 2E and Supplemental Figure S1, A and B), largely similar to set1Δ cells carrying an empty vector. In addition, these strains showed short telomeres (Figure 2F), similar to or even shorter than complete loss of Set1. While these data may indicate that Set1 catalytic activity is the primary requirement to support telomere maintenance, the decreased stability of Set1 with these mutations leaves open the possibility that alternate, noncatalytic functions for Set1 may be required. Furthermore, these data underscore the importance of evaluating protein expression of methyltransferases when using mutants to distinguish catalytic and noncatalytic roles.

FIGURE 2:

Analysis of catalytic requirements of Set1 in subtelomeric gene repression and telomere length control. (A) RT-qPCR analysis of TEL07L genes in WT H3K4K (yEG108), set1Δ H3K4K (yEG1342), H3K4R (yEG109), and set1Δ H3K4R (yEG1343) cells, derivatives of the histone shuffle strain WZY42, which is the S288C background (Zhang et al., 1998). Expression was normalized to the control gene TFC1 and is shown relative to WT. Error bars represent SEM for a minimum of three biological replicates. (B) Southern blot showing terminal telomere fragments in yeast strains described for panel A. (C, D) Western blots depicting FLAG-Set1 levels (C) and H3K4 methylation (D) in set1Δ (yEG232) cells expressing FLAG-SET1 (yEG738), with catalytic mutations G951S (yEG984), H1017L (yEG885), and C1019A (yEG886) or an empty vector (EV; yEG647). In panel C, the abundance of each Set1 mutant relative to that of wild type was determined relative to total protein by image analysis (see Materials and Methods) and is indicated in text below the gel. (E) RT-qPCR analysis of TEL07L genes in set1Δ cells carrying an empty vector (EV) (yEG647) or expressing FLAG-Set1 (WT; yEG738) or FLAG-Set1 G951S (yEG984). Expression was normalized to the control gene TFC1 and is shown relative to WT. Error bars represent SEM for a minimum of three biological replicates. (F) Southern blot showing terminal telomere fragment molecular weight in set1Δ cells with empty vector (EV; yEG647), WT FLAG-Set1 (yEG738), or FLAG-Set1 catalytic mutants G951S (yEG984), H1017L (yEG885), and C1019A (yEG886). For gene expression analysis, significance was evaluated using two-way ANOVA and Tukey’s multiple comparisons test. P values are indicated as follows: * < 0.05, *** < 0.001. All yeast strains described in this figure are derived from the S288C/BY4741 backgrounds.

Set1 contains two RNA recognition motifs (RRM1 and RRM2) near the N-terminus that predominantly bind mRNAs, as well as a small number of noncoding RNAs (Trésaugues et al., 2006; Luciano et al., 2017; Sayou et al., 2017), and the catalytic core of Set1 is composed of the N-SET, SET, and post-SET domains near the C-terminus of the protein (Figure 3A). To determine whether noncatalytic functions of Set1 contribute to its role at telomeres, we generated deletions of RRM1, RRM2, and a region spanning all of RRM1 and RRM2 (RRM1-2) in the FLAG-SET1 construct (Figure 3A). Previous work (Schlichter and Cairns, 2005; Trésaugues et al., 2006; Luciano et al., 2017) demonstrated that the absence of the RRM domains reduces H3K4me3 levels. Therefore, we also generated the reported hyperactive methylation mutant G990E (Schlichter and Cairns, 2005) to help overcome this deficiency and potentially separate H3K4 methyltransferase activity from RNA binding activity. In addition, we tested the H422A point mutation, which is reported to block RNA binding by RRM2 yet maintain H3K4 methylation status. Immunoblotting of the FLAG-Set1 ΔRRM1, ΔRRM2, and ΔRRM1-2 proteins expressed from their endogenous promoter, as well as the H422A point mutant, showed expression similar to that of wild-type FLAG-Set1 for all mutant proteins (Figure 3B and Supplemental Figure S2A). H3K4 methylation levels were also comparable to that in cells with FLAG-Set1, although some degree of H3K4me3 was lost in these mutants, even in combination with the G990E allele (Figure 3C and Supplemental Figure S2B). This is likely due to the reported reduction in Set1 chromatin binding when the RRMs are absent (Sayou et al., 2017).

FIGURE 3:

The catalytic core of Set1, but not the RRM domains, is required for subtelomeric gene repression and telomere length maintenance. (A) Schematic depicting wild-type Set1 domain structure (top) and domain deletions (ΔRRM1, ΔRRM2, ΔRRM1-2) and C-terminal truncation (Δ1-761) used in expression vectors with an N-terminal FLAG tag in set1Δ cells. (B) Western blots of yeast whole cell extracts probed with anti-FLAG showing levels of wild-type (yEG738) and mutant FLAG-Set1 expressed in set1Δ cells (EV, yEG647; G990E, yEG740; H422A, yEG741; ΔRRM1, yEG746; ΔRRM2, yEG747; ΔRRM1-2, yEG748; Δ1-761, yEG1102) and (C) H3K4me3 and H3K4me2 levels for each strain expressing FLAG-Set1 variants. Total protein stain, anti-H3, or anti-H4 was used as loading controls. Asterisk (*) indicates a nonspecific band detected by the anti-FLAG antibody. (D) RT-qPCR analysis of TEL07L genes in set1Δ cells carrying an empty vector (EV) or expressing FLAG-Set1 (WT) or FLAG-Set1 mutants (strains as described in panel B). Expression was normalized to the control gene TFC1 and is shown relative to WT. Error bars represent SEM for a minimum of three biological replicates. Significance was evaluated using two-way ANOVA and Dunnett’s multiple comparisons test. P values are indicated as follows: *** < 0.001. (E) Southern blot showing terminal telomere fragment molecular weight in set1Δ cells with empty vector (EV) or FLAG-Set1 mutants (strains as described in panel B). In all gel image panels, the dashed line separates samples run on the same gel in which intervening lanes were removed for clarity. All yeast strains described in this figure are derived from the BY4741 background.

Analysis of subtelomeric gene expression using reverse transcription–quantitative PCR (RT-qPCR) showed that the RNA binding activity of Set1 is not the primary contributor to gene repression, as the mutants largely maintained repression of TEL07L and TEL09R subtelomeric genes, with only mild derepression observed in ΔRRM1 and ΔRRM2 strains (Figure 3D). Loss of RNA binding by Set1 also appears to minimally affect telomere length, as the H422A mutant and RRM-deleted proteins (ΔRRM1, ΔRRM2, and ΔRRM1-2) displayed telomere lengths more similar to those of wild-type than set1Δ cells (Figure 3E). Similar assays using the FLAG-Set1 RRM-deleted protein combined with the hyperactive G990E mutation were performed; however, there was no change in phenotype in the presence of this mutation (Supplemental Figure S2, C and D).

We next tested whether FLAG-Set1 containing only the catalytic core, residues 762–1081 (Δ1-761) (Kim et al., 2013), was sufficient to promote subtelomeric gene repression and telomere length maintenance. As demonstrated in previous work (Kim et al., 2013), FLAG-Set1 Δ1-761 is stable, highly expressed, and maintains wild-type H3K4me levels (Figure 3, B and C). Cells expressing FLAG-Set1 Δ1-761 closely mimicked the subtelomeric gene expression of cells containing full-length FLAG-Set1 (Figure 3D) and largely rescued the defect in telomere length, with only a modest decrease in length compared with wild-type cells (Figure 3E). These data indicate that the catalytic core of Set1 is sufficient to maintain subtelomeric gene repression and mostly maintains telomere length as well. These results are consistent with previous work indicating that overexpression of the SET domain of Set1 rescues the telomere position effect phenotype of set1Δ cells (Nislow et al., 1997; Corda et al., 1999).

Set1 has also been implicated in DNA damage response (DDR) pathways (Corda et al., 1999; Faucher and Wellinger, 2010). To determine whether Set1’s role in these pathways is also primarily dependent on the SET domain, we tested growth of strains expressing mutant FLAG-Set1 in the presence of hydroxyurea (HU) and caffeine (Supplemental Figure S3). Cells lacking Set1 (empty vector) grew poorly in the presence of either HU or caffeine, whereas the expression of FLAG-Set1 was able to rescue this growth defect. The growth of RNA binding mutants of FLAG-Set1 (ΔRRM1, ΔRRM2, ΔRRM1-2, and H422A) were similar to that of wild type, whereas the catalytic mutants H1017L, C1019A, and G951S were strongly inhibited in the presence of HU or caffeine (Supplemental Figure S3). Expression of the FLAG-Set1 Δ1-761 catalytic core exhibited growth similar to that of full-length FLAG-Set1, indicating that the catalytic core is also sufficient for Set1’s role in the DDR pathway (de La Roche Saint-André and Géli, 2021).

Subtelomeric gene repression, but not telomere length maintenance, depends on H3K4 methylation

In addition to Set1, COMPASS complex subunits and the Rad6-Bre1 H2B ubiquitin ligase complex promote H3K4 methylation: loss of COMPASS subunits Spp1 and Sdc1 leads to a reduction in H3K4me3 and H3K4me2, respectively (Dehé et al., 2006), and rad6Δ mutants have no H3K4me due to the requirement for H2BK123Ub (Sun and Allis, 2002; Dehé et al., 2005). We and others have previously shown that subtelomeric gene repression of TEL07L adjacent genes COS12 and YGL262W is reduced in spp1Δ and sdc1Δ mutants (Figure 4A; Miller et al., 2001; Jaiswal et al., 2017; Jezek et al., 2017a). However, there is less derepression in these cells compared with those lacking Set1, indicating a partial requirement for these COMPASS complex components in subtelomeric gene repression. We also monitored subtelomeric gene expression in rad6Δ cells, which have no H3K4 methylation, and observed that this mutant closely mimics the derepression observed in set1Δ cells (Figure 4A). In Southern blots of terminal telomere fragments, both spp1Δ and sdc1Δ mutants showed little defect in telomere length compared with isogenic wild-type cells (Figure 4B). Similarly, there is some defect in telomere length in the rad6Δ mutant compared with wild type (Figure 4B; Leung et al., 2011; Wu et al., 2017), despite the clear loss of subtelomeric gene repression in these cells. These data support the possibility that defects in telomere length maintenance and subtelomeric gene expression can be separated in some mutants lacking H3K4 methylation. Furthermore, while our data indicate a requirement for the catalytic core of Set1 in telomere maintenance phenotypes, there appears to be only a partial requirement for H3K4 methylation, particularly for telomere length, which shows regulation independent of Spp1, Sdc1, and Rad6.

FIGURE 4:

The requirement for other COMPASS components and Rad6 in subtelomeric gene repression and telomere length. (A) RT-qPCR analysis of TEL07L genes in WT (yEG230), set1Δ (yEG230), spp1Δ (yEG100), sdc1Δ (yEG110), and rad6Δ (yEG623). Expression was normalized to the control gene TFC1 and is shown relative to WT. Error bars represent SEM for a minimum of three biological replicates. Significance was evaluated using two-way ANOVA and Dunnett’s multiple comparisons test. P values are indicated as follows: * < 0.05, *** < 0.001. (B) Southern blot showing terminal telomere fragment molecular weight in WT (yEG230), set1Δ (yEG232), WT (yEG001), spp1Δ (yEG100), sdc1Δ (yEG110), and rad6Δ (yEG623). All yeast strains described in this figure are derived from the BY4741 background.

The abundance of telomere maintenance factors is altered upon loss of Set1

Next, we sought to better define the potential roles for Set1 in telomere length maintenance and subtelomeric gene repression. We and others have previously suggested that Set1 and H3K4 methylation likely support telomere maintenance through additional mechanisms beyond regulating repressive chromatin factors at subtelomeres (Margaritis et al., 2012; Jezek et al., 2017a). This observation is also supported by our finding that the transcriptome of set1Δ mutants is highly correlated with the transcriptome of yeast cells lacking components of the telomerase holoenzyme, such as est1Δ, est3Δ, and tlc1Δ, as well as mutants in other telomere maintenance pathways (Jezek et al., 2017a).

To determine whether any telomere maintenance factors are misregulated in the absence of Set1, we assayed mRNA levels of a subset of relevant genes in set1Δ cells using RT-qPCR (Figure 5A). In these mutants, members of the CST capping complex (Cdc13-Stn1-Ten1) showed altered steady-state mRNA levels, with higher levels of CDC13 and STN1 mRNAs and lower levels of TEN1 mRNA. In addition, the mRNAs coding for telomerase subunits EST1 and EST2 and the telomerase RNA, TLC1, all showed increased abundance in the absence of Set1, whereas the EST3 mRNA is twofold less abundant. The mRNA abundances of additional telomere maintenance regulators, including those coding for the transcription factor Rap1 and members of the RFA, yKu, and Rif1/2 complexes, were also tested; however, there were few statistically significant changes in mRNA levels in set1Δ cells compared with wild type (Figure 5A). The changes in gene expression observed by RT-qPCR were similar to that observed in our previous RNA-seq analysis of set1Δ cells (Martín et al., 2014; Jezek et al., 2017a), summarized in the heatmap in Figure 5B. Although the same strain background was used in each experiment, YKU70 and YKU80 showed more change in expression in set1Δ cells by RNA-seq analysis rather than RT-qPCR; however, the reason for this discrepancy is not known.

FIGURE 5:

Genes encoding telomere maintenance factors show altered abundance levels in set1Δ cells. (A) RT-qPCR analysis of genes encoding telomere regulators in WT (yEG230) and set1Δ (yEG232). Expression was normalized to the control gene TFC1 and is shown relative to WT. Error bars represent SEM for a minimum of three biological replicates. Significance was evaluated using two-way ANOVA and Sidak’s multiple comparisons test. (B) Heatmap depicting the log2 fold change of the same telomere maintenance factors shown in A based on previously published RNA-seq data (Jezek et al., 2017a), with the exception of TLC1, which was not present in the RNA-seq data. (C) RT-qPCR analysis of genes encoding telomere regulators in WT (yEG230), set1Δ (yEG232), spp1Δ (yEG100), sdc1Δ (yEG110), and rad6Δ (yEG623). Expression was normalized to the control gene TFC1 and is shown relative to WT. Error bars represent SEM for a minimum of three biological replicates. Significance was evaluated using one-way ANOVA and Tukey’s multiple comparisons test for individual genes. (D–H) Quantification of protein abundances of Cdc13-MYC, Stn1-MYC, Ten1-MYC, Est1-MYC, and Est3-MYC in wild-type and set1Δ cells. Intensity of MYC signal was determined by Western blot and normalized using a total protein stain as a loading control for a minimum of three biological replicates. Error bars represent SEM. Significance was evaluated using an unpaired t test. Representative images of the immunoblots are depicted below each quantification. (I) RT-qPCR analysis of EST3 in WT H3K4K (yEG108), set1Δ H3K4K (yEG1342), H3K4R (yEG109), and set1Δ H3K4R (yEG1343). Data presented as described for panel A. For all panels, P values are indicated as follows: * < 0.05, ** < 0.01, *** < 0.001. All yeast strains described in this figure are derived from the BY4741 background, except strains containing STN1-MYC used in panel D and histone mutant strains in panel I (see Supplemental Table S1).

We also tested whether mRNA abundance of a subset of telomere maintenance factors is altered in the spp1Δ, sdc1Δ, and rad6Δ mutants (Figure 5C). While some changes in mRNA levels were observed in the spp1Δ cells lacking H3K4me3, the sdc1Δ mutant most closely resembled set1Δ cells, indicating that loss of H3K4me2 is likely a main factor in disrupting the levels of mRNAs encoding telomere maintenance proteins in set1Δ cells. The rad6Δ mutant showed the most difference from the profile observed in set1Δ cells. While this may in part be due to loss of H3K4 methylation, it also likely stems from other functions for Rad6, including its role in promoting H3K79 methylation (Lee et al., 2007; Leung et al., 2011; Wu et al., 2017).

Previous reports indicate that the expression levels and stoichiometry of telomere regulators such as CST and telomerase are highly calibrated by the cell (Morris and Lundblad, 1997; Enomoto et al., 2004; Larose et al., 2007; Tuzon et al., 2011; Advani et al., 2013; Holstein et al., 2014; Liu et al., 2014; Tucey and Lundblad, 2014; Cesena et al., 2017; Gopalakrishnan et al., 2017; Mersaoui and Wellinger, 2019; Garcia et al., 2020). Our analysis of mRNA abundance indicates that the levels of these proteins are likely changing in the absence of Set1. To monitor protein abundance of the CST and telomerase complexes, we epitope tagged CST and telomerase subunits in wild-type and set1Δ cells and performed quantitative immunoblotting (Figure 5, D–H). The target protein abundance was monitored using an anti-MYC antibody, and total protein levels were determined by staining the blot with a general protein stain. Cdc13-MYC levels were slightly but significantly reduced in set1Δ relative to wild type, and Stn1-MYC protein levels appear to be more reduced in cells lacking Set1, whereas Ten1 levels do not significantly differ in set1Δ cells (Figure 5, D–H). Interestingly, the altered abundance of Stn1 protein does not parallel the changes in STN1 mRNA expression, suggesting that it may be subject to translational or posttranslational control in the absence of Set1.

The telomerase subunit Est1 showed increased protein abundance in set1Δ cells, consistent with higher levels of EST1 mRNA in the mutant (Figure 5, A and G). Intriguingly, our analysis of Est3-MYC protein levels showed only a minor decrease in Est3-MYC abundance in set1Δ cells compared with wild type (Figure 5H). This was surprising, given that the EST3 mRNA abundance was the most substantially decreased in the absence of Set1 (Figure 5A). It was also decreased in the sdc1Δ and rad6Δ mutants (Figure 5C), and the H3K4R mutant showed similarly reduced levels of EST3 mRNA (Figure 5I), indicating that the abundance of this mRNA is likely dependent on functional Set1 and H3K4 methylation levels.

Regulation of EST3 mRNA in set1Δ mutants

To better understand the relationship between EST3 mRNA and protein levels, we also evaluated the steady-state levels of the mRNA encoding EST3-MYC. This mRNA had equivalent abundances in wild-type and set1Δ cells, unlike the endogenous EST3; however, its levels were substantially decreased compared with the endogenous mRNA (Supplemental Figure S4A). Currently, immunoblotting of the endogenous protein is not feasible, as there is not an available antibody for the recognition of yeast Est3. Previous reports have indicated that epitope tags disrupt the function of Est3 (Tuzon et al., 2011; Tucey and Lundblad, 2014); therefore, we evaluated the function of our Est3-MYC construct in subtelomeric gene repression and telomere length. Wild-type cells expressing Est3-MYC showed a decrease in subtelomeric gene repression compared with wild-type cells expressing the endogenous untagged Est3, an effect that was further exacerbated in combination with set1Δ (Supplemental Figure S4B). We also observed that cells with Est3-MYC have shorter telomeres, approximately equivalent to those of set1Δ cells (Supplemental Figure S4C). However, set1Δ mutants with EST3-MYC show even shorter telomeres than set1Δ alone, indicating that each allele is contributing to the short telomere phenotype through a distinct pathway. These data suggest that the EST3-MYC allele causes a loss-of-function phenotype due to decreased mRNA abundance and disrupted protein function, and it is not likely to accurately reflect endogenous protein levels of Est3.

Despite the complications in accurately evaluating Est3 protein levels, we next addressed whether Set1 may have a role in transcriptional regulation of EST3 via the H3K4me3 or H3K4me2 methyl marks. Given that other mutants that show reduced H3K4 methylation also showed lower levels of EST3 mRNA (Figure 5), we used chromatin immunoprecipitation (chIP) to determine the abundance of H3K4me3 and H3K4me2 at the EST3 promoter, 5′ ORF and 3′ ORF. We compared this to levels at similar regions of EST1, for which the mRNA is increased in the absence of Set1. Both H3K4me3 and H3K4me2 were present to similar extents in the promoters and 5′ ends of EST3 and EST1; however, there was a marked difference at the 3′ end of the ORFs (Figure 6, A and B). Both histone marks, though particularly H3K4me2, showed increased levels at the 3′ end of EST3 compared with EST1. This difference suggests that control of EST3 mRNA abundance may rely on 3′ deposition of H3K4me2 and that EST3 regulation may have a greater dependency on H3K4 methylation than EST1 regulation. This finding is also supported by observations that sdc1Δ mutants, but not spp1Δ mutants, show increased EST3 mRNA (Figure 5C).

FIGURE 6:

Regulation of endogenous EST3 mRNA and epistasis analysis of set1Δ est3Δ mutants. chIP of H3K4me3 (A) and H3K4me2 (B) shown relative to total H3 at EST3 and EST1 gene loci in WT cells (yEG230). Primer sets amplify promoter (P), 5′ coding region (5′), and 3′ coding region (3′). The 5′ coding region of PMA1 serves as a positive control for H3K4me3/H3K4me2 chromatin localization, and the TEL07L region is a negative control. SEM of three biological replicates is shown. (C) Relative abundance of EST3 and STN1 mRNAs in WT (yEG230) and set1Δ (yEG232) cells at 10 and 20 min after thiolutin treatment. EST3 and STN1 levels were normalized to the RNA pol II–independent transcript sCR1. SEM of three biological replicates is shown. (D) RT-qPCR analysis of TEL07L genes in WT (yEG1362), set1Δ (yEG1363), est3Δ (yEG1364), and set1Δ est3Δ (yEG1365). Expression was normalized to the control gene TFC1 and is shown relative to WT. Error bars represent SEM for a minimum of three biological replicates. Significance was evaluated using two-way ANOVA and Tukey’s multiple comparisons test. P values are indicated as follows: * < 0.05, ** < 0.01. (E) Southern blot showing terminal telomere fragment molecular weight in WT (yEG1362), set1Δ (yEG1363), est3Δ (yEG1364), and set1Δ est3Δ (yEG1365). All yeast strains described in this figure are derived from the BY4741 background.

We also evaluated whether Set1 may regulate posttranscriptional processing or stability of the EST3 mRNA. In this assay, we compared EST3 abundance to that of STN1 because its protein levels appeared decreased by immunoblot despite increased abundance of its mRNA (Figure 5, C and H). Wild-type and set1Δ cells were treated with the transcription inhibitor thiolutin, and the mRNA abundance of each transcript was measured by RT-qPCR and compared with that of the RNA pol III–controlled transcript sCR1. In this assay, we observed more rapid turnover of the EST3 transcript in set1Δ cells compared with wild type (Figure 6C; P value = 0.013 from two-way analysis of variance [ANOVA]), indicating a shorter half-life for the EST3 mRNA upon loss of Set1. This differed from STN1, which showed similar levels in both wild-type and set1Δ cells following thiolutin treatment (P value = 0.976 from two-way ANOVA), suggesting that each of these transcripts is subject to different regulatory mechanisms dependent on Set1.

Overlapping functions of Set1 and Est3 in telomere maintenance phenotypes

To investigate the dependence of telomere maintenance defects in set1Δ on EST3 expression, we generated set1Δ est3Δ mutants following sporulation of double heterozygous diploids and evaluated subtelomeric gene repression and telomere length. As expected, both set1Δ and est3Δ single mutants show increased subtelomeric gene expression (Figure 6D); however, there is a synergistic increase in expression in set1Δ est3Δ double mutants. These data suggest that both Set1 and Est3 are separately required for subtelomeric gene repression and loss of each of their activities disrupts multiple pathways normally repressing genes within subtelomeric chromatin. Both set1Δ and est3Δ single mutants also exhibited shorter telomeres (Figure 6E); however, est3Δ mutants were much shorter than set1Δ mutants. Interestingly, the set1Δ est3Δ double mutants showed telomere lengths similar to those of est3Δ single mutants, indicating that Set1 and Est3 may contribute to the same pathway required for maintaining appropriate telomere lengths. These data suggest that one factor that may contribute to short telomeres in set1Δ cells is reduced EST3 mRNA abundance; however, this is not likely a key contributor to the misregulation of subtelomeric gene repression in the absence of Set1.

DISCUSSION

The lysine methyltransferase Set1 has long been implicated in the regulation of subtelomeric gene repression and telomere length maintenance (Nislow et al., 1997; Corda et al., 1999; Krogan et al., 2002; Schneider et al., 2005; Trelles-Sticken et al., 2005; Jezek et al., 2017a), yet its specific role in these processes has not been completely defined. Based on the association of Set1 and H3K4 methylation with active transcription, the euchromatic distribution of H3K4me3 has been implicated in the maintenance of repressive chromatin at subtelomeres through largely indirect mechanisms (Santos-Rosa et al., 2004; Venkatasubrahmanyam et al., 2007). However, other observations suggest that there are likely additional mechanisms that underlie Set1’s role in telomere maintenance pathways. For example, loss of COMPASS components and other factors that deplete H3K4 methyl levels do not always display phenotypes similar to those of set1Δ cells (Leung et al., 2011; Wu et al., 2017). In addition, we previously reported that the transcriptome profile of set1Δ mutants is most highly correlated with mutants in genes directly linked to telomere maintenance, such as those encoding telomerase subunits (Jezek et al., 2017a). Given these observations, we sought to more systematically define the role for Set1 and H3K4 methylation in two metrics of telomere health: chromatin-mediated gene repression and telomere length.

The Set1 SET domain is required for telomere health through H3K4 methylation–dependent and -independent pathways

Previous studies using reporter assays have suggested that the SET domain of Set1 is primarily responsible for maintaining subtelomeric gene repression (Nislow et al., 1997; Corda et al., 1999); however, the specific contribution of Set1 catalytic methyltransferase activity at H3K4 has been difficult to assess. We used a series of genetic mutations, including the H3K4R point mutation, COMPASS mutants, and mutations in Set1, to quantitatively compare subtelomeric gene expression and simultaneously evaluate telomere length. In a set1Δ background, we expressed FLAG-SET1 from its endogenous promoter on a plasmid and generated multiple point mutations known to abrogate catalytic activity. As loss of catalytic activity is reported to destabilize the Set1 protein (Soares et al., 2014), we monitored abundance of these mutants to clarify whether or not these specific mutations promote Set1 instability. In all mutants tested, we observed decreased levels of FLAG-Set1, indicating that use of these mutants does not provide clean separation of Set1 catalytic and noncatalytic functions.

In addition to catalytic activity localized within the SET domain, Set1 contains two RRMs, which bind a diversity of RNAs (Schlichter and Cairns, 2005; Luciano et al., 2017; Sayou et al., 2017), though whether they contribute to Set1-dependent telomere maintenance had not been directly investigated. Previous studies have also differed on the extent to which H3K4 methylation is affected by loss of the RRM domains (Schlichter and Cairns, 2005; Trésaugues et al., 2006; Luciano et al., 2017). However, our immunoblotting shows equivalent expression of all Set1 mutants tested with moderate loss of H3K4me3 in the RRM deletion mutants, although not in the H422A point mutation that disrupts RNA binding (Trésaugues et al., 2006). This mutant analysis showed that the catalytic core of Set1 is predominantly responsible for maintaining subtelomeric gene repression and telomere length. Interestingly, while the expression of the Set1 catalytic core (Set1 Δ1-761) restores H3K4 methylation levels, the distribution of H3K4me3 and H3K4me2 in the genome is disrupted (Soares et al., 2017). This suggests that specific placement of the H3K4 methyl marks is not required for Set1-dependent telomere maintenance, or that an alternate function of the catalytic core of Set1 may be required. In addition, our analysis of gene expression and telomere length in COMPASS mutants spp1Δ and sdc1Δ, as well as rad6Δ, showed that subtelomeric gene expression changes correlated with the relative loss of H3K4 methyl species in these mutants. However, this was not the case for telomere length, which showed little change from wild type in all of these mutants. These findings indicate that telomere length maintenance and repression of subtelomeric chromatin are separable processes that show differential requirements for H3K4 methylation.

In addition to catalyzing H3K4 methylation, the SET domain of Set1 has additional functions. Set1 methylates the kinetochore-associated protein Dam1 (Zhang et al., 2005), and more recently, H3K37me1 has been reported to be targeted by Set1 together with the H3K36 methyltransferase Set2 (Santos-Rosa et al., 2021). In a noncatalytic role, the SET domain of Set1 physically interacts with the DNA damage protein Mec3 (Corda et al., 1999), which may also contribute to its function at telomeres. Furthermore, orthologues of Set1 have been shown to perform functions independent of the catalytic methyltransferase activity (Kirmizis et al., 2007; Dorighi et al., 2017; Rickels et al., 2017, 2020). Further identification of mutants that separate functions of the SET domain is needed to better define the specific contributions of catalytic and noncatalytic activities to Set1’s role in telomere maintenance, as well as its other biological roles.

Altered abundance of telomere maintenance factors in set1Δ mutants through transcriptional and posttranscriptional regulation

To better define the role of Set1 in telomere maintenance, we assayed steady-state mRNA and protein abundance of a series of telomere regulators in cells lacking Set1, Spp1, Sdc1, or Rad6, with a focus on components of the CST telomere capping complex and the telomerase holoenzyme. At the mRNA level, we observed alterations to the abundance of STN1, TEN1, EST1, EST3, and TLC1. There appeared to be an enhanced dependence on H3K4me2 or H3K4me1 for regulating mRNA abundance, as sdc1Δ mutants most closely mimicked set1Δ cells. Interestingly, in monitoring protein abundance of these factors, we observed that the mRNA and protein levels change in parallel for some factors, such as Est1; however, some proteins displayed changes in abundance that could not be explained by changes at the mRNA level. For example, STN1 transcript levels are increased in set1Δ, while Stn1-MYC was decreased. This is also seen, albeit to a lesser extent, with Cdc13, whose transcript levels increased in set1Δ cells while the protein abundance was slightly reduced. This suggests altered translational or posttranslational regulation of CST components in the absence of Set1. Both Cdc13 and Stn1 are subject to posttranslational regulation: they are phosphorylated by the cyclin-dependent kinase, Cdk1, and this modification serves to stabilize the complex at the telomere (Tseng et al., 2009; Liu et al., 2014; Gopalakrishnan et al., 2017). This phosphorylation, and possibly other regulatory mechanisms including transcriptional control, may be disrupted in set1Δ cells due to their aberrant cell cycle progression (Beilharz et al., 2017). The abundance and stoichiometry of CST complex components is exquisitely regulated in the cell, and previous work has identified multiple mechanisms of regulation and shown that alterations to any of the complex members can impact telomere function and promote feedback regulation of other telomere maintenance factors (Morris and Lundblad, 1997; Enomoto et al., 2004; Larose et al., 2007; Tuzon et al., 2011; Advani et al., 2013; Holstein et al., 2014; Liu et al., 2014; Tucey and Lundblad, 2014; Cesena et al., 2017; Gopalakrishnan et al., 2017; Mersaoui and Wellinger, 2019; Garcia et al., 2020). Our data indicate that Set1 promotes the proper balance of CST components within the cell and this disruption in set1Δ cells likely contributes to impaired telomere function.

In addition to altered abundance of CST factors, we noted differential abundance of EST1 and EST3 mRNAs, each of which encodes components of the telomerase holoenzyme. The increase in EST1 in set1Δ cells was reflected in higher abundance of the Est1-MYC protein. The enrichment of H3K4me3 and H3K4me2 at the EST1 gene suggests that it may be subject to Set1-dependent transcriptional regulation; however, other mechanisms controlling protein abundance may also be disrupted in the absence of Set1. For example, Est1 protein levels are regulated posttranslationally through a proteasome-dependent process involving the ubiquitin ligase Ufd4 and the ubiquitin-binding complex Cdc48-Npl4-Ufd1 (Lin et al., 2015). Further experiments are required to determine the relative contribution of these mechanisms to controlling Est1 abundance.

At the transcript level, EST3 showed the largest change in abundance in set1Δ cells and related mutants. This is consistent with our previous findings that the transcriptome in set1Δ cells is highly correlated with that of est3Δ mutants (Jezek et al., 2017a). Further analysis of the EST3 mRNA half-life following thiolutin treatment indicated more rapid degradation of EST3 mRNA than that of STN1, suggesting that EST3 posttranscriptional regulation is disrupted in the absence of Set1. Interestingly, the EST3 mRNA undergoes programmed ribosomal frameshifting and is regulated by the nonsense mediated decay (NMD) pathway (Morris and Lundblad, 1997; Enomoto et al., 2004). One possibility is that there is increased shunting of the EST3 mRNA to the NMD pathway in set1Δ cells, potentially through misregulated frameshifting or other processing or translation defect.

We investigated Est3 protein abundance in set1Δ cells by integrating a MYC tag at the C-terminus, as there is no available antibody for detecting Est3. It has previously been reported that C-terminal epitope tags can disrupt function of Est3 (Tuzon et al., 2011; Tucey and Lundblad, 2014). Indeed, we observed decreased gene repression and shorter telomeres with the addition of the C-terminal MYC tag, though also a substantial decrease in mRNA abundance. While we could not conclusively determine whether full-length Est3 protein showed decreased abundance in set1Δ cells, epistasis analysis of set1Δ est3Δ double mutants suggested that Set1 and Est3 contribute separately to subtelomeric gene repression. However, set1Δ est3Δ cells have telomere lengths similar to those of est3Δ single mutants, indicating that Set1 and Est3 both contribute to the same pathway to maintain telomere length.

In summary, our data indicate that Set1’s role at telomeres depends on the activity of its catalytic core, although H3K4 methylation–independent functions likely contribute to telomere length maintenance. The role for Set1 in subtelomeric gene repression is most closely correlated with H3K4 methyl status, indicating that this mark is required for gene repression and, further, that misregulated subtelomeric chromatin is not the primary driver of telomere shortening in set1Δ mutants. These differential requirements for H3K4 methylation are also evident in our analysis of mRNA abundance of telomere maintenance factors, with genes such as EST3 likely dependent on H3K4 methyl status for regulation, but not EST1 or STN1. These transcripts show disrupted transcriptional and posttranscriptional regulation in the absence of Set1, as well as possible translational or posttranslational mechanisms directing protein abundance. This likely reflects the multiple mechanisms that precisely calibrate the abundance of these critical telomere regulators. Altogether, this study provides new and unexpected insights into how Set1 promotes telomeric and subtelomeric function and provides the foundation for dissecting the roles of H3K4 methyl species and additional targets of Set1 in telomere maintenance pathways in yeast and other systems.

MATERIALS AND METHODS

Request a protocol through Bio-protocol.

Yeast strains and growth conditions

Yeast strains, plasmids, and primers used in this study are listed in Supplemental Tables S1, S2, and S3, respectively. Gene knockouts, incorporation of epitope tags, and plasmid construction and transformation were performed using standard methods (Longtine et al., 1998; Jaiswal et al., 2017; Jezek et al., 2017b; Tran et al., 2018; Jaiswal et al., 2020). The FLAG-SET1 allele was constructed by targeting of a 2xFLAG-containing PCR cassette upstream of the SET1 coding sequence, as described (Moqtaderi and Struhl, 2008). This sequence was PCR-amplified from genomic DNA, along with the SET1 promoter and 3′UTR, and cloned into pRS316 for generating mutants in a plasmid-based expression system. Amino acid substitutions and domain deletions were made using either Gibson assembly or the Q5 Site-Directed Mutagenesis Kit (NEB). Isogenic single and double mutant strains were generated via haploid mating, sporulation, and tetrad dissection. Genotypes were confirmed using PCR (endogenous mutations) or plasmid sequencing. Standard liquid and solid growth media were used (yeast extract, peptone, dextrose [YPD] and synthetic complete [SC]-dropout). The CDC13 temperature-sensitive mutant strain was a gift from David Lydall (Newcastle University) (Holstein et al., 2014).

RNA extraction and RT-qPCR to measure mRNA abundance

RNA was extracted from 1.5 ml of logarithmically growing cells (OD₆₀₀ ∼0.6–0.8) using the Masterpure Yeast RNA Purification kit (Epicentre). Residual genomic DNA was removed using the Turbo DNA-free kit (Ambion). cDNA was synthesized using the qMax cDNA Synthesis Kit (Accuris) with 1 μg of template RNA. qPCR was performed to assess mRNA transcript levels with qMax Green Low Rox qPCR Mix (Accuris), 0.5 μl of cDNA, and 20 μM forward and reverse gene specific primers. Amplification was performed using the CFX384 Real-time Detection System (Bio-Rad). Each reaction was performed in technical triplicate and a minimum of three biological replicates. mRNA abundance was determined relative to the control gene TFC1 and displayed as fold change (or relative expression) compared with wild type. To measure mRNA degradation, cells were treated with 10 μg/ml thiolutin (ENZO Life Sciences) and harvested at 10 and 20 min after treatment. mRNA abundance was determined relative to the control gene sCR1 and normalized to the zero time point. All qPCR primers used are listed in Supplemental Table S3.

DNA preparation and Southern blots of telomere length

Telomere length was assessed using a nonradioactive Southern blotting protocol (Feng et al., 2013). Briefly, 10 ml of yeast culture was grown overnight at 30°C and harvested. Pellets were resuspended in yeast breaking buffer (10 mM Tris-HCl, pH 8.0, 100 mM NaCl, 1 mM EDTA, pH 8.0, 1% SDS, 2% Triton X-100) and vortexed with acid-washed glass beads (Sigma-Aldrich) with phenol–chloroform–isoamyl alcohol (PCI; 25:24:1) (VWR Scientific). After centrifugation, DNA was isolated from the supernatant via ethanol precipitation and treated with RNaseA. Genomic DNA was digested with XhoI (NEB) and subject to PCI extraction and ethanol precipitation. Prepared DNA (40–60 μg) was loaded onto an 0.8% agarose gel (0.5% tris borate EDTA [TBE]) and run for 25 h at 100 V. Ethidium bromide staining and imaging was used to observe DNA migration. Following denaturation and washing (Feng et al., 2013), DNA was transferred onto an N+hybond membrane (Amersham) before cross-linking. The blot was probed with 15 ng/ml biotinylated telomere probe (5′-biotin-CACACCCACACCCACACC-3′) at 65°C overnight. Nucleic acids on the blot were detected using a Chemiluminescent Nucleic Acid Detection Module Kit (ThermoFisher) per the manufacturer’s instructions and imaged using a LiCor C-Digit chemiluminescence imager.

Yeast senescence assays

To obtain replicatively “young” cells for all appropriate genotypes, tetrad dissections of heterozygous diploid strains were used to generate isogenic haploid strains. Senescence assays were performed as described (Kozak et al., 2010; Lu and Liu, 2010). Briefly, cells were grown on plates for 3 d and single colonies were used to inoculate 10 ml of YPD. The cells were grown overnight with shaking at 30°C. Cell density was calculated using a hemocytometer, and new cultures were diluted to 1 × 10⁴ cells/ml from the overnight cultures. After 24 h, cell density was recorded and new cultures were inoculated at 1 × 10⁴ cells/ml. This procedure was repeated every 24 h for up to 15 d. The number of population doublings per culture was calculated for each time point taken. Experiments were performed in biological triplicate.

Yeast spot assays

Cells were grown at 30°C to saturation in 5 ml of YPD and then diluted to 1.0 OD₆₀₀ units. Tenfold serial dilutions were plated on the appropriate media, including YPD, SC-URA, SC-URA + 100 mM hydroxyurea, or SC-URA + 10 mM caffeine. For temperature sensitivity assays, plates were grown at the indicated temperatures and imaged up to 4 d. For DNA damage sensitivity assays, plates were grown at 30°C and imaged daily for up to 7 d.

Immunoblotting

Lysates for immunoblotting were prepared using either 0.2 M NaOH treatment or bead-beating in lysis buffer (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1 mM EDTA, 1% NP-40) (Jezek et al., 2017a; Tran et al., 2018). Protein concentrations were determined by Bradford assay and normalized to load equivalent amounts on SDS–PAGE. Proteins were transferred to Immobilon-FL polyvinylidene difluoride membrane (Fisher Scientific). Blots were probed overnight with primary antibodies, followed by incubation with the appropriate secondary antibody. Blots were imaged using an Odyssey CLx Scanner and processed using ImageStudio (LiCor). Total Protein Stain (LiCor) was used to determine total protein amounts for quantitation. Antibodies used in this study were c-MYC 9E10 (Novus Biologicals NB600-302SS), monoclonal anti-FLAG M2 (Sigma-Aldrich F1804), H3K4me3 (Active Motif 39159), H3K4me2 (Active Motif 39142), H3 (Abcam ab1791), H4 (Abcam ab31830), IRDye 680RD goat anti-mouse immunoglobulin G (IgG) (LiCor 926-68070), and IRDye 800CW goat anti-rabbit IgG (LiCor 926-32211).

Chromatin immunoprecipitation

chIP was performed as previously described (Jezek et al., 2017a,b; Jethmalani et al., 2021). All primers used for amplification of chromatin regions are listed in Supplemental Table S3. The ratio of percent input for the histone mark (H3K4me3 or H3K4me2) is shown relative to percent input for histone H3.

Abbreviations used:

chIP

chromatin immunoprecipitation

CST

Cdc13-Stn1-Ten1 complex

DDR

DNA damage resposne

EV

empty vector

HDAC

histone deacetylase

HU

hydroxyurea

me

methyl

NMD

nonsense mediated decay

ORF

open reading frame

RRM

RNA recognition motif

RT-qPCR

reverse transcription quantitative polymerase chain reaction

SC

synthetic complete

SDS–PAGE

sodium dodecyl-sulfate polyacrylamide gel electrophoresis

SIR

silent information regulator

TEL

telomere

Ub

ubiqtuin

YPD

yeast extract, peptdone, dextrose.