SEARCH

SEARCH BY CITATION

Keywords:

  • UBR ubiquitin ligase;
  • N-end rule pathway;
  • N-degron;
  • Rec8;
  • Sz. pombe

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results and discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

The N-end rule pathway degrades proteins bearing a destabilization-inducing amino acid at the N-terminus. In this proteolytic system, Ubr ubiquitin ligases recognize and ubiquitylate substrates intended for degradation. Schizosaccharomyces pombe has two similar Ubr proteins, Ubr1 and Ubr11. Both proteins have unique roles in various cellular processes, although the ubr1∆ strain shows more severe defects. However, their involvement in the N-end rule pathway is unclear, and even the N-end rule pathway-dependent proteolytic activity has not been demonstrated in Sz. pombe. Here, we show that: (a) Sz. pombe has the N-end rule pathway in which only Ubr11, but not Ubr1, is responsible; and (b) the C-terminal fragment of the meiotic cohesin Rec8 (denoted as Rec8c) generated by separase-mediated cleavage is an endogenous substrate of the N-end rule pathway. Forced overexpression of stable Rec8c was deleterious in mitosis and caused a loss of the mini-chromosome. In unperturbed mitosis without overexpression, the rate of mini-chromosome loss was five-fold higher in the ubr11∆ strain. Since Rec8 is normally produced in meiosis, we examined whether meiosis and sporulation were affected in the ubr11∆ strain. In unperturbed meiosis, chromosome segregation occurred almost normally and viable spores were produced in the ubr11∆ cells, irrespective of the presence of undegraded endogenous Rec8c peptides. Copyright © 2012 John Wiley & Sons, Ltd.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results and discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Ubiquitin is a small protein that regulates various biological processes by covalently attaching to proteins (Hershko and Ciechanover, 1998). Ubiquitin ligases discern specific peptide sequences, or degrons, in the substrate and attach a ubiquitin molecule to the substrate. The RING-finger motif protein Ubr1 was first identified as a ubiquitin ligase in a specialized proteolytic mode known as the ‘Arg/N-end rule pathway’ in Saccharomyces cerevisiae (Bartel et al., 1990; Varshavsky, 2011). A unique feature of the N-end rule pathway is that the very N-terminal amino acid of the substrate protein determines its stability (Bachmair et al., 1986). The first peptide identified as a substrate for the Arg/N-end rule pathway in eukaryotes was eK, a ~40-residue Escherichia coli Lac repressor-derived sequence (Bachmair and Varshavsky, 1989). Characterization of eK and other model tetrapeptides revealed that the N-degron, an essential sequence for Ubr1-dependent ubiquitylation, can be divided into at least two sub-elements: (a) the absolute and crucial N-terminal amino acid itself, together with a minor contribution from the penultimate amino acid at the second position; and (b) a second sequence with no characteristic features except for the presence of a ubiquitin-accepting lysine (Bachmair and Varshavsky, 1989; Choi et al., 2010; Suzuki and Varshavsky, 1999).

Since the degron is portable, intrinsically stable proteins can be converted to unstable proteins by fusing them to the N-degron sequence. A conditional N-degron system was developed using a heat-labile dihydrofolate reductase mutant protein (DHFRts) as a degradation-inducing signal (Dohmen et al., 1994). In this temperature-sensitive degron (td) system, the DHFRts containing an arginine at the N-terminus (ArgDHFRts) is fused to a protein intended for proteolytic inactivation. The ArgDHFRts-fused protein is functional at the permissive temperature, but is rapidly degraded when the temperature is raised to 36 °C. This heat-induced degron has been successfully used in Sz. pombe (Rajagopalan et al., 2004; Kearsey and Gregan, 2009) as well as in S. cerevisiae, in which it was originally developed.

Canonical Ubr ubiquitin ligases are widely conserved in eukaryotes, but not all of these Ubr proteins recognize a destabilizing amino acid at the N-terminus (Tasaki et al., 2005, 2009). Irrespective of whether N-end rule or non-N-end rule degradation operates, the canonical UBR ubiquitin ligases play physiological roles in various cellular events (An et al., 2006, 2010; Heck et al., 2010; Khosrow-Khavar et al., 2012; Kwon et al., 2001, 2003; Nillegoda et al., 2010; Tasaki et al., 2007; Turner et al., 2000; Wang et al., 2004; Xia et al., 2008). Importantly, loss of human UBR1 is a cause of the hereditary disorder Johanson–Blizzard syndrome (Zenker et al., 2005). In addition, Ubr1 assures the fidelity of chromosome segregation (Rao et al., 2001). In anaphase, separase cleaves the Scc1 cohesin subunit to liberate the cohered sister chromosomes. The resultant C-terminal polypeptide of Scc1 must be degraded by the Arg/N-end rule pathway to prevent the loss of chromosomes in budding yeast.

Sz. pombe has two canonical Ubr RING-finger proteins, Ubr1 and Ubr11 (Kitamura et al., 2001). We have previously shown that Ubr11 is required for the utilization of extracellular oligopeptides (Kitamura et al., 2012). Conversely, only Ubr1 regulates the oxidative stress response and drug resistance by controlling the levels of active Pap1, a bZIP transcription factor (Kitamura et al., 2011). Ubr1 is also involved in various processes in Sz. pombe, such as nuclear localization of the proteasome (Takeda and Yanagida, 2005), invasive growth (Dodgson et al., 2009), stability of the Sre1 transcription factor under aerobic conditions (Lee et al., 2011) and meiosis and regulation of cell morphology (our unpublished observations). In spite of these pleiotropic physiological functions, it is still ambiguous whether Ubr1 has a role in proteolysis via the N-end rule pathway in Sz. pombe. It was reported that Ubr11 is required for the induced temperature-sensitive lethality of the ArgDHFRts-bir1 strain (Rajagopalan et al., 2004). Although this may assume the involvement of Ubr11 in the N-end rule pathway, dependency of the lethality of the ArgDHFRts-bir1 strain on the nature of the N-terminal amino acid of DHFR is undetermined. In addition, even the N-end rule pathway-dependent proteolytic activity has not been proven in Sz. pombe.

In this study, we definitely demonstrated that Sz. pombe has the N-end rule pathway in which only Ubr11 is required but Ubr1 is dispensable. We also show that the meiotic cohesin Rec8-derived C-terminal polypeptide (Asp385–Ile561, hereafter denoted as Rec8c), a product of separase-mediated cleavage, is an endogenous substrate of the Ubr11-dependent Arg/N-end rule pathway. The data for the role of the Ubr11/N-end rule pathway in mitosis and meiosis, especially the effect of the presence of stable Rec8c polypeptide, are presented.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results and discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Yeast strains and culture conditions

Sz. pombe strains used in this study are listed in the Supporting information (Table S1). Media and other general yeast methods have been previously described (Forsburg and Rhind, 2006). To repress expression from the nmt promoter, 15–30 µ m thiamine was added to the minimal medium EMM2 (Basi et al., 1993; Maundrell, 1993). For liquid culture, cells were grown in thiamine-free EMM2 to express the intended protein for 18–20 h, as induction from the nmt promoter requires 12–16 h. To induce synchronous meiosis, temperature-sensitive pat1 cells (Iino and Yamamoto, 1985), pregrown in normal EMM2, were cultured in EMM lacking NH4Cl for 15 h at 25 °C, then shifted to 34 °C. NH4Cl (0.25 g/l) was added upon temperature upshift.

Plasmid constructions

  • XaaDHFRts–HA–Mcm4. An original plasmid encoding Ub–ArgDHFRts–HA–Mcm4N (with promoter and N-terminal sequences from the genomic mcm4 gene) was obtained from Dr Kearsey (Lindner et al., 2002). This plasmid encodes heat-labile dihydrofolate reductase (ArgDHFRts), which induces rapid degradation of Mcm4 at 36 °C. The arginine at the N-terminus of ArgDHFR was changed to aspartate, leucine or methionine by inverse PCR, using appropriate primers and KOD Plus polymerase (Toyobo). Each plasmid was cut at the SpeI site within the coding region of Mcm4N, and integrated into the genomic mcm4 locus of the host strain. The resultant XaaDHFRts–HA–Mcm4 protein was detected using anti-HA antibody.
  • XaaRec8c–FGFP. First, the open reading frame encoding monoubiquitin was amplified by PCR and cloned into the pCR4Blunt–TOPO vector (Invitrogen). Then, part of the Sz. pombe rec8+ gene corresponding to the separase-generated C-terminal polypeptide (Asp385–Ile561, denoted as Rec8c) was amplified by PCR and cloned behind the last codon of ubiquitin in the same reading frame to generate the plasmid pCR(Ub–AspRec8c). For subsequent excision, an NdeI site was introduced in the PCR primers, one at the initiator ATG codon of the ubiquitin gene and the other after the Rec8c coding region. The resultant ubiquitin–Rec8c fusion gene was excised and inserted into the NdeI site between 6His and Flag tags in the pDUAL–HFG vectors (Pnmt1 or weaker Pnmt4 promoter versions; Matsuyama et al., 2004) to generate pDUAL(Ub–AspRec8c–FGFP). In Sz. pombe cells, once the 6Hisubiquitin–AspRec8c–Flag–GFP (denoted as Ub–AspRec8c–FGFP) fusion protein is produced from this construct, the ubiquitin moiety is immediately cleaved off by abundantly expressed deubiquitylases (Figure 1A). As a result, the first aspartate residue of Rec8c is exposed as the novel N-terminus. This first aspartic acid was mutated (to alanine, arginine, cysteine, histidine, leucine, methionine or tryptophan) by inverse PCR, using the pCR(Ub–AspRec8c) plasmid. After sequence verification, the corresponding fragment was liberated by NdeI digestion, and inserted into the pDUAL–HFG vector, as described above. All plasmids were integrated at the leu1-32 locus of the host Sz. pombe strains.
  • AspRec8c–FGFP–Mei2SATA. A minimum mei2 region (429–733) was amplified by PCR, using p81–mei2SATA as the template (Watanabe et al., 1997), digested with BamHI and BglII and inserted into the BglII site of the above-described pDUAL(Ub–AspRec8c–FGFP) plasmid. As a result, Mei2SATA was expressed from the nmt1 promoter as a fusion protein, AspRec8c–FGFP–Mei2SATA.
image

Figure 1. Sz. pombe has the Arg/N-end rule pathway. (A) Schematic illustration of the XaaDHFRts–HA–Mcm4 (left) and XaaRec8c–FGFP (right) substrate proteins. Both substrates were produced as an N-terminal ubiquitin-fusion protein. Ubiquitin is co-translationally cleaved and the Xaa (Arg or Met in DHFRts, Asp or Met in Rec8c) is exposed as a new N-terminus. (B) Stabilities of the temperature-sensitive ArgDHFRts- and MetDHFRts-tagged Mcm4. Both proteins were expressed from the endogenous mcm4 promoter. Extracts were prepared from cells incubated at 36 °C for the indicated durations, and the DHFRts–HA–Mcm4 protein was detected using an anti-HA antibody. Cdc2, control for protein loading levels. (C) Expression levels of the AspRec8c–FGFP and MetRec8c–FGFP. Both proteins were expressed from the nmt4 promoter, and their expression levels were compared by immunoblotting with an anti-GFP antibody. (D) AspRec8c–FGFP, but not MetRec8c–FGFP, is unstable. Expression was shut off by the addition of thiamine and cycloheximide (+B1/CHX). AspRec8c–FGFP (lanes 1–3) was expressed from the stronger nmt1 promoter to increase its expression to levels comparable to that of the Met version from the weaker nmt4 promoter (lanes 4 and 5)

Download figure to PowerPoint

Protein analysis

Preparation of total cellular protein extracts and immunoblotting were performed as described previously (Kitamura et al., 2011). Anti-GFP (GF200; Nacalai Tesque), anti-HA (3 F10; Roche Diagnostics) and anti-PSTAIRE (sc-53; Santa Cruz Biotechnology) were used as primary antibodies.

Mini-chromosome loss assay

The mini-chromosome Ch10–CN2 (Niwa et al., 1989) was used to measure mitotic and meiotic stability. This non-essential 120 kb linear mini-chromosome carries the sup3-5 suppressor gene, which is capable of rescuing the chromosomal ade6-704 nonsense mutation in a host strain. On low adenine medium containing only one-tenth the amount of adenine (YE1/10A or EMM21/10A), cells devoid of the Ch10–CN2 mini-chromosome formed red colonies because of the ade6-704 mutation, whereas the Ch10–CN2-containing cells formed white colonies. To examine the effect of Rec8c overexpression from the nmt1 promoter, cells were cultured in thiamine-free medium for 8 h and plated on EMM21/10ade medium with (OFF) or without (ON) thiamine. To measure stability in meiosis that was not artificially manipulated, Ch10–CN2-containing parental strains pregrown in EMM2 (without adenine) were mated and sporulated. The resultant zygotic spores were dissected using a micromanipulator and grown on rich YE5S. Viability was calculated as the ratio of non-growing spores to the total number of dissected spores. Colonies were replicated on a YE1/10ade plate, and scored for red/white colour. To measure stability in unperturbed mitosis, three white colonies from an EMM2 (without adenine) plate were suspended in water, then spread on YE1/10ade plates. The rate of mitotic loss was calculated from the numbers of white and red colonies after 3–6 days.

Results and discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results and discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Proteolysis via the Arg/N-end rule pathway occurs in Sz. pombe

First, the instability of a heat-labile DHFR-fused protein was examined because the ArgDHFRts–HA–Mcm4 fusion protein has been shown to be degraded at the non-permissive temperature in Sz. pombe (Lindner et al., 2002). To confirm that this degradation follows the N-end rule, a methionine-led MetDHFRts–HA–Mcm4 fusion protein was generated and its stability was compared with that of ArgDHFRts–HA–Mcm4 (Figure 1A, B). When the temperature was raised to 36 °C, the ArgDHFRts–HA–Mcm4 protein rapidly disappeared, as expected (Figure 1B). In marked contrast, MetDHFRts–HA–Mcm4 levels were higher at 28 °C (0 h) and relatively stable at 36 °C. These data formally prove that the Arg/N-end rule degradation pathway operates in Sz. pombe.

The C-terminal fragment of the separase-cleaved mitotic cohesin subunit Scc1 has to be degraded by the Arg/N-end rule pathway in S. cerevisiae (Rao et al., 2001). Rec8 is the counterpart of Scc1 in meiosis, and separase-dependent cleavage of Rec8 is essential for meiosis in Sz. pombe (Kitajima et al., 2003). We examined whether this resultant C-terminal polypeptide of Rec8 (Rec8c) was also a substrate for the Arg/N-end rule pathway. To examine this possibility, the C-terminal polypeptide (Asp385–Ile561), corresponding to the Rec8c fragment after separase-mediated cleavage, was fused to the N-terminus of GFP. The Flag epitope in the vector remains between Rec8c and GFP, and the resultant Rec8c–Flag–GFP fusion protein (denoted as AspRec8c–FGFP; Figure 1A) was examined for its stability. The N-terminal aspartic acid was replaced with methionine as a control, and the expression levels of AspRec8c–FGFP or MetRec8c–FGFP protein were analysed by immunoblotting. The AspRec8c–FGFP protein levels were significantly lower than that of MetRec8c–FGFP (Figure 1C). Furthermore, comparison of the stabilities of these two GFP proteins clearly showed that the AspRec8c-fused GFP was apparently destabilized in an N-terminal aspartate-dependent manner (Figure 1D).

Components of the Arg/N-end rule pathway machinery

In the Arg/N-end rule pathway, Ubr proteins play an essential role as the E3 ubiquitin ligase for substrate recognition and destabilization (Bartel et al., 1990; Tasaki and Kwon, 2007; Tasaki et al., 2012; Varshavsky, 2011). Sz. pombe has two similar Ubr RING proteins, Ubr1 and Ubr11 (Kitamura et al., 2001); however, the role of these Ubr proteins in the N-end rule pathway has not been tested. ArgDHFRts–HA–Mcm4ts was rapidly degraded in the wild-type and ubr1∆ strains at 36 °C, but was stabilized in a ubr11∆ strain (Figure 2A). A Rec8-based substrate was similarly stabilized in ubr11∆ cells. AspRec8c–FGFP was barely detectable in the wild-type and ubr1∆ strains, but easily detected in the ubr11∆ strain (Figure 2B). A further increase in AspRec8c–FGFP levels was not observed in cells simultaneously lacking both Ubr1 and Ubr11 (Figure 2B, C), indicating that only Ubr11, but not Ubr1, plays a role in the Arg/N-end rule pathway. AspRec8c–FGFP also accumulated in the mts2 strain, which harbours a temperature-sensitive mutation in the Rpt2 regulatory subunit of the proteasome, demonstrating that the proteasome is responsible for AspRec8c–FGFP degradation (Figure 2B).

image

Figure 2. Components of the Arg/N-end rule pathway in Sz. pombe. (A) ArgDHFRts–HA–Mcm4 was stabilized in the absence of Ubr11. Extracts were prepared from cells cultured at 26 °C (lanes 1, 3 and 5) or at 36 °C for 1 h (lanes 2, 4 and 6). (B) Degradation of AspRec8c–FGFP requires Ubr11 (lane 4), but not Ubr1 (lane 5). Simultaneous loss of Ubr11 and also Ubr1 (lane 6) did not further increase the AspRec8c–FGFP levels more than ubr11∆ alone. AspRec8c–FGFP also accumulated after inactivation of the proteasome in the temperature-sensitive mts2 mutant (lane 7). Lane 1, wild-type (wt) cells expressing GFP only; lane 2 (NC), ubr11∆ cells without AspRec8c–FGFP-expressing plasmid as a negative control. (C) AspRec8c–FGFP was stabilized in the absence of the arginyltransferase Ate1 (lane 5). Lane 1 (PC), extracts of the wild-type strain expressing stable MetRec8c–FGFP as a positive control. (D) Instability of AspRec8c depends on the Ubc2 ubiquitin-conjugating enzyme. A minimum region of the Mei2SATA protein, which induces meiosis and terminates cell cycle progression, was fused to the C-terminus of AspRec8c–FGFP and expressed from the vitamin B1-repressible nmt1 promoter in the host strains shown in parentheses. The resultant AspRec8c–FGFP–Mei2SATA protein inhibited growth only in the ubc2∆ and ubr11∆ strains in the inducing condition (left, –B1) but not in the other 10 ubc strains (see Supporting information, Figure S1A)

Download figure to PowerPoint

In the ubiquitin-dependent eukaryotic Arg/N-end rule pathway, arginine has to be added as a new N-terminus for degradation if the first residue of an N-end rule substrate is acidic (Asp or Glu) (Varshavsky, 2011). The arginyl-tRNA-transferase Ate1 catalyses this reaction (Balzi et al., 1990). To examine whether the N-terminal aspartic acid of the Rec8c peptide has to be modified for degradation in the N-end rule pathway in Sz. pombe, a gene-encoding possible Ate1 homologue, which we named ate1+, was deleted from the Sz. pombe genome. The level of AspRec8c–FGFP accumulation in the ate1∆ strain was similar to that in ubr11∆ cells (Figure 2C), demonstrating that in Sz. pombe Ate1 is essential for the degradation of N-end rule substrates bearing an acidic amino acid at the N-terminus, as in other organisms.

In general, the E2/ubiquitin-conjugation enzyme (Ubc) cooperates with ubiquitin ligase in the ubiquitylation reaction cascade. To identify which Ubc enzyme among the 11 ubiquitin-specific Ubc proteins is required for the N-end rule pathway in Sz. pombe, a screening substrate was devised. In growing cells, activity of Mei2, a key inducer of meiosis, is repressed by Pat1-mediated phosphorylation. Mei2 becomes constitutively active if the two phosphorylation sites are mutated to unphosphorylatable alanine (S438A T527A, referred to as Mei2SATA). Forced expression of the Mei2SATA protein induces meiosis irrespective of the intracellular condition and nutritional status, thus preventing further cell cycle progression (Watanabe et al., 1997). The minimum essential region of Mei2SATA (429–733) was fused to the C-terminus of AspRec8c–FGFP, and the resultant fusion protein (denoted as AspRec8cG–Mei2SATA) was expressed from the repressible nmt1 promoter. Expression of AspRec8cG–Mei2SATA had no effect on cell growth in the wild-type strain. In contrast, the same fusion protein strongly inhibited cell growth in the ubr11∆ strain (Figure 2D), possibly due to AspRec8c stabilization and inhibition of cell proliferation by the fused Mei2SATA in the absence of Ubr11. AspRec8cG–Mei2SATA induction inhibited cell growth only in the ubc2∆ mutant among the 11 ubc mutants examined (Figure 2D; see also Supporting information, Figure S1A). Although ubc2∆ cells inherently exhibit weak and delayed cell growth, GFP expression did not have such an inhibitory effect. In conclusion, only Ubc2 (a Rad6 homologue, also known as Rhp6) is indispensable, and none of the other 10 Ubc proteins (Ubc1, 4, 6, 7, 8, 11, 13–16) are required for the Arg/N-end rule pathway in Sz. pombe. Requirement of the Ubc2 homologues is consistent with the findings in S. cerevisiae and mammals (Varshavsky, 2011).

Amino acid specificity for N-end rule-dependent proteolysis

One remarkable feature of the N-end rule pathway is its amino acid specificity, which confers protein instability (Bachmair et al., 1986; Varshavsky, 2011). Ubr proteins bind to bulky hydrophobic and basic amino acids at the N-terminus and induce degradation in model substrates harbouring such N-terminal amino acids (Bachmair et al., 1986; Tasaki et al., 2005, 2009). Acidic Asp and Glu are also potent degradation-inducing amino acids. To determine whether these criteria were also applicable in Sz. pombe, the original Asp at the N-terminus of AspRec8c–FGFP was changed to one of following amino acids: basic (Arg, His), or bulky hydrophobic (Trp, Leu). MetRec8c–FGFP was used as a control. These XaaRec8c–FGFP proteins were expressed in the wild-type or ubr11∆ strains, and their protein levels were compared by immunoblotting. In the ubr11∆ strain, all proteins were stabilized and expressed at significant levels comparable to that of the stable MetRec8c–FGFP (Figure 3A). In the wild-type strain, however, the proteins, except for the stable Met version, were unstable and only small amounts of protein were detected. The degree of instability differed between amino acids: the effects of Leu and Trp were weaker than the effects of the basic and acidic residues (lanes 7 and 9). The effect of Ala or Cys at the N-terminus was also tested, and both Ala- and Cys-led proteins were stably expressed even in the wild-type strain (lanes 14 and 15) at levels comparable to that of the stable Met version (lane 13) and the Asp version in the degradation-defective ubr11∆ strain (lane 16).

image

Figure 3. N-terminal amino acid specificity in Arg/N-end rule-dependent instability. (A) XaaRec8c–FGFP proteins harbouring the indicated amino acids at the N-terminus were expressed in the wild-type (wt) and ubr11∆ strains from the nmt4 (lanes 1–11) or nmt1 (lanes 12–16) promoter. Protein levels were monitored by immunoblotting. (B) The N-terminal leucine was less effective for inducing degradation. Strains expressing the indicated XaaDHFRts–HA–Mcm4 were cultured as in Figure 1B

Download figure to PowerPoint

Similarly, LeuDHFRts–HA–Mcm4 was degraded at 36 °C but the extent of degradation was milder than that of the quite unstable Asp versions (Figure 3B). Therefore, the overall feature of N-terminal specificity is conserved in Sz. pombe, although the degree of the instability of each amino acid varies.

Chromosome segregation in the ubr11 mutant

We noticed that the growth of ubr11∆ cells retarded in response to the overexpression of AspRec8c–FGFP from the strongest nmt1 promoter (Figure 4A; see also Supporting information, Figure S1B). AspRec8c–FGFP did not affect cell growth in the wild-type strain, but the stabilized MetRec8c–FGFP induced a similar slow growth (Figure 4A), demonstrating that growth retardation was caused by the stabilization of Rec8c polypeptide itself and not by other factors. It has been reported that expression of a stable version of the separase-generated Scc1 cohesin C-terminal fragment perturbs faithful chromosome segregation during mitosis in budding yeast (Rao et al., 2001). To examine whether a similar defect was induced by the stabilized Rec8 C-terminal peptide, a mini-chromosome loss assay was performed. The chromosomal ade6-704 mutation in the host strain imparts a red colour under adenine-limiting conditions, but the sup3-5 gene on the Ch10–CN2 mini-chromosome suppresses adenine auxotrophy, and thus cells harbouring Ch10–CN2 form white colonies (Niwa et al., 1989). Overexpression of AspRec8c–FGFP in the ubr11∆ ade6-704 strain containing the Ch10–CN2 mini-chromosome resulted in the formation of red colonies, indicating the loss of the mini-chromosome (Figure 4B). AspRec8c–FGFP overexpression had no effect on the wild-type strain; however, mini-chromosome loss occurred upon expression of the stable MetRec8c–FGFP. Therefore, overproduction of the stabilized Rec8c interferes with chromosome separation in mitosis, as observed in S. cerevisiae. In unperturbed mitosis without Rec8c overexpression, the rate of Ch10 mini-chromosome loss was about five-fold higher in ubr11∆ cells (2.34%) than in wild-type cells (0.42%).

image

Figure 4. Effects of the Rec8 C-terminal polypeptide on mitosis and meiosis in the ubr11∆ strain. (A) Overexpression of the stable Rec8c peptides interferes with growth. XaaRec8c–FGFP was induced from the nmt1 promoter in the absence of thiamine (–B1). Cells were cultured for 3 days (see also Supporting information, Figure S1B). (B) Overexpression of the stable Rec8c induces loss of the mini-chromosome. The same strains shown in (A) were cultured in low-adenine medium for 5 days. (C) The endogenous Rec8c polypeptide generated by separase persists during meiosis in the ubr11∆ strain. Synchronous meiosis was induced by the pat1ts mutation in the wild-type and ubr11∆ strains. Levels of the full-length Rec8–GFP protein and the C-terminal Rec8c–GFP polypeptides (indicated by arrow) were analysed at the indicated time points, using anti-GFP antibody. In the right-most lane (PC), an extract expressing stable MetRec8c–FGFP was used as a control. White arrowheads indicate non-specific bands. (D) Loss of mini-chromosome in mitotically growing ubr11∆ strain. After dissection, spores were grown in rich yeast extract medium (with adenine), then replicated onto either low-adenine yeast extract medium (left) or minimal medium without adenine supplementation (right). Nine representative colonies that exhibit various degrees of colour from white to deep red are shown. Cells in deep red colonies were unable to grow in the absence of adenine, indicating a complete loss of the Ch10–CN2 mini-chromosome

Download figure to PowerPoint

Under physiological conditions, Rec8 functions mainly in meiosis. If the separase-generated Rec8c polypeptide is an authentic physiological substrate for Ubr11, this C-terminal fragment should be stabilized during meiosis in the ubr11∆ strain. To address this, full-length Rec8–GFP was expressed from its own promoter by tagging the chromosomal rec8 gene with GFP, and its protein levels during synchronous meiosis were monitored by immunoblotting (Figure 4C). Rec8 was not detected under growing conditions but the full-length proteins were transiently expressed in the early stage of meiosis. In the wild-type strain, separase cleaved Rec8 at the time of chromosome segregation (4 h) and Rec8 disappeared thereafter. Full-length Rec8 also disappeared in the ubr11∆ strain with a ~1 h delay (5 h), but importantly, the separase-generated Rec8c–GFP fusion polypeptide remained for a further 2 h after cleavage. Together, these results formally prove that the separase-generated C-terminal polypeptide of endogenous Rec8 is an authentic physiological substrate for Ubr11 and, therefore, the Arg/N-end rule pathway in Sz. pombe. This C-terminal fragment was finally degraded by an unidentified proteolytic activity other than the N-end rule pathway.

Because the endogenous Rec8c polypeptides remained undegraded for several hours during meiosis in ubr11∆ cells, as described above, it is likely that this remaining Rec8c somehow affects chromosome segregation. In both the wild-type and ubr11∆ strains, first and second meiotic nuclear divisions occurred and four-spored asci of normal appearance were eventually formed. The total viability of the spores was not significantly different between the wild-type and ubr11∆ strains (96.5% and 92.7%, respectively). We also analysed the spores to determine whether they lost the mini-chromosome during meiosis and/or growth after germination of spores in a genetic cross in which both parents of the mating pair contained the Ch10–CN2 mini-chromosome. Because the first residue of the endogenous Rec8 C-terminal polypeptide is aspartic acid and its degradation depends on Ate1 (Figure 2C), spores from the ate1∆ diploid strain were also examined. The proportion of deep red colonies from the tetrads of the four viable spores, an indication of mini-chromosome loss, was not different during zygotic meiosis in the three strains: 22.6% (wild-type strain), 26.1% (ubr11∆) and 21.4% (ate1∆). Therefore, the rate of chromosome loss did not significantly increase in the Arg/N-end rule pathway mutants during meiosis under physiological conditions. However, we noticed that some colonies were not white but exhibited a pale pink colour (Figure 4D). The proportion of pink colonies was significantly higher in ubr11∆ (33.5%) than in the wild-type strain (< 1%). These results suggested that the mini-chromosome was gradually lost from some mitotically growing cells following spore germination, consistent with the five-fold higher mini-chromosome loss rate in mitotically growing ubr11∆ cells.

In conclusion, meiosis proceeded normally and produced mostly viable spores in ubr11∆ cells, regardless of the fact that endogenous Rec8 C-terminal polypeptides remained undegraded during meiosis. A haploid Sz. pombe cell has only three chromosomes. If the chromosomes missegregate in meiosis, most of the resultant spores should be non-viable because they will not receive a proper set of chromosomes. Therefore, missegregation did not occur at significant levels, if any, during meiosis in ubr11∆ cells under physiological conditions. Rad21 cohesin, a fission yeast counterpart of Scc1, also has to be cleaved by separase upon chromosome segregation (Tomonaga et al., 2000). Since Asn180 or Asp232 is exposed as the new N-terminus of the C-terminal polypeptide after cleavage, it is likely that the C-terminal polypeptide of separase-cleaved Rad21 is also a substrate for Ubr11. The five-fold increase in mini-chromosome loss rate observed in ubr11∆ cells may be due to the stabilization of the Rad21 C-terminal polypeptide. However, the growth rates of ubr11∆ and wild-type cells were comparable. Therefore, undegraded C-terminal peptides of both Rad21 and Rec8 do not exert any profound effect at the physiological expression level.

Among two canonical Ubr proteins in Sz. pombe, only Ubr11 but not Ubr1 is required for the Arg/N-end rule pathway, although ubr1∆ cells exhibit severe defects in diverse biological processes. The only apparent defect in the N-end rule pathway-deficient ubr11∆ strain is the inability to utilize extracellular oligopeptides (Kitamura et al., 2012), but the corresponding substrate for Ubr11 is still unknown. Detailed analysis of this regulation is currently under way, and will shed light on the biology of the Ubr N-recognin besides the degradation of potentially deleterious cohesin C-terminal peptides.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results and discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

We would like to thank Dr Kearsey for the DHFR degron plasmid, and Dr Niwa and the National Bio-Resource Project (NBRP) of MEXT, Japan, for the Sz. pombe strains. This study was supported in part by a Grant-in-Aid for Scientific Research (No. 18570006) from the Japan Society for the Promotion of Science (to K.K.).

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results and discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results and discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Supporting information on the internet

The following supporting information may be found in the online version of this article:

FilenameFormatSizeDescription
yea_2936_figS1.epsPS document1703KFigure S1. Degradation of AspRec8c depends on the Ubc2 and Ubr11. (A) The minimum region of Mei2SATA that halts cell cycle progression was expressed as a fusion of AspRec8c with FGFP. The resultant AspRec8c–FGFP–Mei2SATA protein did not interfere with growth in most of the ubiquitin-conjugating enzyme mutants, except for the ubc2∆ mutant (see also Figure 2B). For the temperature-sensitive ubc4 and ubc11/ubcP4 strains, growth was examined at the semi-permissive temperature (33°C). Strains were: wt (KSP2529); ubr11∆ (KSP2530); ubc1∆ (KSP2533); ubc4 (KSP2569); ubc11 (ubcP4) (KSP2570); ubc6∆ (KSP2534); ubc7∆ (KSP2535); ubc8∆ (KSP2536); ubc13∆ (KSP2537); ubc14∆ (KSP2538); ubc15∆ (KSP2568); and ubc16∆ (KSP2539). (B) Expression of AspRec8c–FGFP interfered with growth in the ubr11∆ strain. This inhibitory effect was not observed with GFP expression alone. Note that these strains are prototrophs, therefore the growth retardation was unrelated to the mini-chromosome loss shown in Figure 4A. Strains were: wt (KSP2426); ubr11∆ AspRec8c–FGFP (KSP2429); ubr11∆ GFP (KSP2222a)
yea_2936_tableS1.docWord document46KTable S1. Sz. pombe strains used in this study

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.