Regulating levels of centromeric histone H3 (CenH3) variant is crucial for genome stability. Interaction of Psh1, an E3 ligase, with the C terminus of Cse4 has been shown to contribute to its proteolysis. Here, we demonstrate a role for ubiquitination of the N terminus of Cse4 in regulating Cse4 proteolysis for faithful chromosome segregation and a role for Doa1 in ubiquitination of Cse4.
CENTROMERIC histone H3 (CenH3), an evolutionarily conserved histone H3 variant, is essential for chromosome segregation (Stoler et al. 1995; Meluh et al. 1998; Blower and Karpen 2001; Maddox et al. 2012). Mis-localization and overexpression of CenH3 has been observed in cancers and associated with aneuploidy in Drosophila melanogaster (Tomonaga et al. 2003; Heun et al. 2006; Moreno-Moreno et al. 2006). Studies from budding yeast and fruit flies have shown that proteolysis of CenH3 plays an important role in preventing its mis-localization (Collins et al. 2004; Moreno-Moreno et al. 2006, 2011). Psh1, an E3 ligase, interacts with Cse4, the budding yeast CenH3, and mediates its ubiquitination for proteolysis (Hewawasam et al. 2010; Ranjitkar et al. 2010). Despite ubiquitination of Cse4 by Psh1, Cse4 is only partially stabilized when overexpressed in psh1Δ strains (Hewawasam et al. 2010; Ranjitkar et al. 2010), suggesting that additional factors regulate Cse4 protein stability.
In this study, we investigated the role of Cse4 domains in directing Cse4 proteolysis and used a genome-wide screen to identify pathways that regulate Cse4 stability. Since subcellular levels of Cse4 are stringently regulated, strains overexpressing CSE4 or cse4 mutants have been used to identify proteolytic pathways and factors such as Psh1 that mediate its degradation (Hewawasam et al. 2010; Ranjitkar et al. 2010). We initiated our studies with a mutant cse416KR (16KR) and 16KR fusion mutants in which lysines (K) are mutated to arginines (R) (Figure 1, A–C; Supporting Information, Table S1 and File S1). Overexpression of 16KR leads to defects in chromosome segregation, and this correlates with enrichment in chromatin and increased protein stability (Collins et al. 2004; Au et al. 2008) . Our results show that the stability of CK (fusion of Cse4 N terminus amino acids 1–135 and 16KR C terminus) is low (t1/2 = 26 ± 6 min) and is comparable to that of Cse4 (t1/2 = 34 ± 3 min), indicating that the lack of ubiquitination of the C terminus due to K-to-R mutations does not increase protein stability. In contrast, the stability of KC (a fusion of 16KR N terminus and Cse4 C terminus) is high (t1/2 > 120 min) and similar to 16KR (Figure 1, B and C), showing that K-to-R mutations in the N terminus of Cse4 increase protein stability. Consistent with these results are previous observations showing that Cse4 lacking the N terminus (Δ129) is fivefold more stable than full-length Cse4 (Morey et al. 2004). Increased Cse4 stability correlates with chromosome loss, as KC and 16KR strains exhibit fourfold higher chromosome loss than strains expressing CK or wild-type Cse4 (Figure 1G). Based on these results, we propose that the N terminus of Cse4 facilitates its proteolysis.
We next determined whether the N terminus of Cse4 promotes turnover of a heterologous protein, the histone fold domain of histone H3 (H3 HFD). We compared the stabilities of H3, H3 HFD, CH (Cse4 N terminus fused to H3 HFD), and KH (16KR N terminus fused to H3 HFD) (Figure 1D). Consistent with our hypothesis, the half-life of CH (t1/2 = 31 ± 5 min) is less than that of H3 HFD (t1/2 = 119 ± 20 min), H3 (t1/2 = 49 min), or KH (t1/2 > 120 min) (Figure 1, E and F). The faster turnover of CH shows that the N terminus of Cse4 mediates proteolysis independently of the Cse4 C terminus.
Psh1 interacts with the CENP-A-targeting domain (CATD) (Black et al. 2007) and ubiquitinates Cse4 at its C terminus in vitro (Hewawasam et al. 2010; Ranjitkar et al. 2010). However, the rapid turnover of both CK and CH, which lack C-terminal lysine residues for ubiquitination or the CATD, respectively, suggests that the N terminus of Cse4 must also contribute to its proteolysis. To investigate the role of Psh1 in N-terminal-mediated proteolysis, we examined the stability of Cse4 mutants in psh1Δ strains. As reported previously, Cse4 is more stable in the psh1Δ strains compared to wild type (t1/2 = 50 vs. 34 min) (Figure 2, A and B) (Hewawasam et al. 2010; Ranjitkar et al. 2010). KC and KH were more stable than CK or CH (t1/2 > 110–120 min vs. 46 min) in the psh1Δ strains, similar to their relative stabilities in the wild-type strain. Given the decreased turnover rate of CK and CH in psh1Δ strains compared to wild-type strains (t1/2 = 46 vs. 26 min and t1/2 = 46 vs. 31 min, respectively) (Figure 1, B and E, and Figure 2A), we conclude that Psh1 only partially contributes to the N-terminal-mediated proteolysis of Cse4 independently of its C terminus. The increased stability of KC in wild-type strains compared to the stability of Cse4 or CK in psh1Δ strains suggests that there are additional pathways regulating Cse4 stability via its N terminus.
Finding a role for the Cse4 N terminus in regulating proteolysis prompted us to identify genes/pathways that mediate Cse4 turnover. A genome-wide screen using a synthetic genetic array (SGA) (Tong et al. 2004; Costanzo et al. 2010) was employed to identify genes essential for growth when CSE4 is overexpressed (GALCSE4). The screen led to identification of genes, including DOA1 and PSH1, involved in ubiquitin-proteasome pathways (W. C. Au and M. A. Basrai, unpublished results). The identification of PSH1 confirms GALCSE4 synthetic lethality in psh1Δ strains (Hewawasam et al. 2010; Ranjitkar et al. 2010). To determine if the N-terminus-mediated proteolysis of Cse4 is ubiquitin-dependent, we pursued further studies of DOA1, which regulates cellular Ub levels (Johnson et al. 1995; Zhao et al. 2009). We validated the SGA results by confirming the lethality conferred by GALCSE4 in psh1Δ and doa1Δ strains (Figure 2C), and the phenotype correlates with greater stability of Cse4 in psh1Δ (t1/2 = 50) and doa1Δ (t1/2 > 120 min) strains (Figure 2, D and E). In contrast to their rapid turnover in psh1Δ, CK and CH are more stable in doa1Δ strains (t1/2 > 120 min) (Figure 2, D and E). Furthermore, endogenously expressed Cse4 is also stabilized in doa1Δ strains (Figure 2F and Table S2), showing that increased stability of Cse4 in these strains is not limited to overexpressed CSE4. Based on these results, we conclude that Psh1 is not the sole regulator of Cse4 proteolysis and that Doa1 facilitates Cse4 N-terminus-dependent proteolysis.
Doa1 regulates Ub levels in yeast cells (Johnson et al. 1995). Since overexpression of UBI4 (encoding Ub) suppresses the phenotypes of doa1Δ (Finley et al. 1987; Lis and Romesberg 2006), we reasoned that it might suppress the lethality of doa1Δ strains overexpressing CSE4. Overexpression of UBI4 suppresses the lethality conferred by GALCSE4 (Figure 3A) and reduces the stability of Cse4 in doa1Δ strains (Figure 3B). Overexpression of UBI4 partially suppresses the GALCSE4-induced lethality of psh1Δ strains and reduces the stability of Cse4 (Figure S1C). These results provide further support for Psh1-independent proteolysis of Cse4 and suggest that Cse4 N-terminus-mediated proteolysis is ubiquitin-dependent.
We next asked if the N terminus of Cse4 is directly ubiquitinated, and we tested the requirement of DOA1 in this process. We performed affinity pull-down using agarose with tandem ubiquitin-binding entities (Ub+) (Hjerpe et al. 2009) (see File S1) with strains expressing Myc-Cse4 or chimeric mutant proteins. Control samples with vector alone, 16KR, or Cse4 with agarose without Ub-binding activity (Ub−) do not show laddering (Figure 3C, lanes 12, 15, 19), but the detection of 16KR in Ub+ suggests that nonmodified (NM) forms of Cse4 (arrow) interact with ubiquitinated proteins bound to Ub+. Consistent with our hypothesis for ubiquitination of the N terminus of Cse4, we observed a laddering pattern for Cse4, CK, and CH strains in the Ub+ samples from wild-type strains, indicative of polyubiquitination of these proteins (Figure 3C, lanes 13, 14, 18). The reduced laddering pattern of CK compared to Cse4 suggests that the lysines in the C terminus are ubiquitinated. Mutation of lysines in the N terminus (KC) show markedly diminished laddering, indicating that the N terminus is ubiquitinated and the interaction of Psh1 with the C terminus of Cse4 is not sufficient for ubiquitination (Figure 3C, lane 16). Further evidence for ubiquitination of the N terminus of Cse4 is the observed laddering pattern of CH and the absence of laddering of H3HFD lacking the N terminus of Cse4 (Figure 3C, lanes 18, 17). Since CH lacks the Psh1-interacting domain, the observed ubiquitination is likely to be Psh1-independent. In contrast to in wild-type strains, ubiquitination of Cse4, CK, CH, and KC was barely detectable in doa1Δ strains (Figure 3C, lanes 20–23). Using HA-tagged Cse4, we observed a sevenfold reduction in high-molecular-weight forms of Cse4 in doa1Δ strains relative to wild type (Figure 3D, lanes 8 and 10, and Figure 3E). That the laddering was largely absent after cycloheximide treatment is consistent with the idea that the high-molecular-weight isoforms represent polyubiquitinated Cse4 that is rapidly degraded (Figure 3D, lane 12). The lack of Cse4 ubiquitination in doa1Δ strains is consistent with increased stability of Cse4 proteins in these strains (Figure 2, D–G).
Consistent with our hypothesis that the N terminus of Cse4 is ubiquitinated, a Cse4 mutant lacking the N terminus (Δ129) does not show laddering in wild-type or doa1Δ strains (Figure 3D, lanes 9 and 11). Overexpression of UBI4 restores laddering of Cse4 in both doa1Δ and psh1Δ strains (Figure S1A, lanes 14 and 14s, and Figure S1B, lanes 9 and 10) but has no effect on the laddering for Δ129-expressing strains (Figure S1A, lane 16). The absence of laddering with Δ129 indicates that the N terminus of Cse4 is required for ubiquitination at the C terminus in vivo. Restoration of laddering similar to that observed for wild type upon overexpression of UBI4 in the psh1Δ mutant provides evidence for Psh1-independent mechanisms of Cse4 ubiquitination. Taken together, these results show that the N terminus of Cse4 is a target of ubiquitination and that DOA1 is required for ubiquitination and proteolysis of Cse4.
In summary, we have shown that the N terminus of Cse4 is a target of ubiquitination and that this modification plays a major role in regulating its proteolytic degradation to ensure faithful chromosome transmission. We propose that ubiquitination of the N and C termini of Cse4 by Psh1 and other Ub ligases in the presence of Doa1 regulates Cse4 proteolysis. Support for our conclusion is derived from observations showing that the N terminus of Cse4 mediates rapid turnover when fused with either mutant C terminus of Cse4 (CK) or a heterologous protein H3 (CH) in a Doa1-dependent manner. Consistent with these results, mutations in the N terminus of Cse4 (KC and KH) makes the protein highly stable. Furthermore, a mutant with a deletion of the N terminus (cse4Δ129) fails to be ubiquitinated. Our observations that overexpressed UBI4 can suppress the lethality conferred by GALCSE4 and increase the ubiquitination and proteolysis of Cse4 in psh1Δ strains provide evidence for additional pathways in addition to Psh1 that can ubiquitinate Cse4. Our results are in agreement with other studies on the role of post-translational modifications of histone N termini in chromatin structure and chromosome segregation (Choy et al. 2011, 2012; Samel et al. 2012). Given the evolutionary conservation of CenH3, future studies to define mechanisms that regulate cellular levels of CenH3 may help us to better understand the link between CenH3 overexpression and its mislocalization observed in cancers.
We thank Sue Biggins and Mitch Smith for strains, Charlie Boone and Michael Costanzo for strains and results of the SGA screen, Richard Gardner for advice on ubiquitin assays, and the Basrai laboratory members for discussions and comments on the manuscript. This work was supported by the National Institutes of Health Intramural Research Program.
Communicating editor: K. M. Arndt
- Received January 29, 2013.
- Accepted March 14, 2013.
- Copyright © 2013 by the Genetics Society of America