In fission yeast, the ER-residing molecular chaperone calnexin is normally essential for viability. However, a specific mutant of calnexin that is devoid of chaperone function (Δhcd_Cnx1p) induces an epigenetic state that allows growth of Schizosaccharomyces pombe without calnexin. This calnexin-independent (Cin) state was previously shown to be mediated via a non-chromosomal element exhibiting some prion-like features. Here, we report the identification of a gene whose overexpression induces the appearance of stable Cin cells. This gene, here named cif1+ for calnexin-independence factor 1, encodes an uncharacterized nucleolar protein. The Cin cells arising from cif1+ overexpression (Cincif1 cells) are genetically and phenotypically distinct from the previously characterized CinΔhcd_cnx1 cells, which spontaneously appear in the presence of the Δhcd_Cnx1p mutant. Moreover, cif1+ is not required for the induction or maintenance of the CinΔhcd_cnx1 state. These observations argue for different pathways of induction and/or maintenance of the state of calnexin independence. Nucleolar localization of Cif1p is required to induce the Cincif1 state, thus suggesting an unexpected interaction between the vital cellular role of calnexin and a function of the nucleolus.

Introduction

The endoplasmic reticulum (ER) has several vital functions including biosynthesis of organelles, lipid donation to other organelles, Ca2+ homeostasis, as well as folding and quality control of membrane-bound and secreted proteins (Kleizen and Braakman, 2004). The ER lumen is a highly crowded environment with a concentration of resident and nascent proteins of about 100 mg/ml (Kleizen and Braakman, 2004). Despite the elevated levels of unstructured polypeptides within the ER lumen, most of the proteins attain their native conformation and continue on the secretory pathway. The prevention of aggregation and the correct folding of nascent proteins in the ER lumen are mediated by the concerted action of foldases and molecular chaperones (reviewed by Ellgaard and Helenius, 2003; Fewell et al., 2001; Hebert and Molinari, 2007; Kleizen and Braakman, 2004).

Calnexin is a molecular chaperone in the ER that plays key roles in protein folding and quality control (Caramelo and Parodi, 2008; Hebert and Molinari, 2007; Williams, 2006). Calnexin has been implicated in several genetic diseases caused by inherited protein folding defects (Amaral, 2004; Chevet et al., 1999; Kuznetsov and Nigam, 1998; Ni and Lee, 2007). Mechanistically, calnexin interacts with nascent polypeptides as a lectin that binds N-linked glycan, and/or via protein-protein contacts (Bedard et al., 2005; Caramelo and Parodi, 2008; Helenius and Aebi, 2004; Marechal et al., 2004; Moremen and Molinari, 2006; Thammavongsa et al., 2005; Williams, 2006).

Structurally, calnexin is a type-I transmembrane protein. X-ray crystallography studies revealed that the lumenal portion of calnexin consists of a globular domain and an extended hairpin-like arm (Schrag et al., 2001). The globular domain contains a Ca2+-binding site and is responsible for the lectin activity (Schrag et al., 2001; Leach et al., 2002). The hairpin protrusion, also called the P-domain, contains a highly conserved central domain, hcd (Elagoz et al., 1999; Jannatipour and Rokeach, 1995). The hairpin structure is involved in protein-protein interactions with unfolded proteins and with other ER chaperones (Leach et al., 2002; Marechal et al., 2004; Pollock et al., 2004; Williams, 2006).

Although calnexin was first identified as a molecular chaperone, mounting evidence indicates the involvement of calnexin in other cellular process. For instance, it was shown that one of the main roles of calnexin is the retention of irremediably misfolded protein in the ER and their targeting for degradation (Bedard et al., 2005). Most recently, several reports demonstrated that calnexin is involved in apoptosis induced by ER stress (Delom et al., 2006; Delom et al., 2007; Guerin et al., 2008; Takizawa et al., 2004; Tomassini et al., 2004; Torgler et al., 1997; Zuppini et al., 2002). Intriguingly, the knockout of the gene encoding calnexin in different organisms results in diverse phenotypes (Denzel et al., 2002; Lee et al., 2005; Muller-Taubenberger et al., 2001). These phenotypes, however, have not unveiled comprehensive aspects of the novel cellular roles of calnexin because the organisms studied also encode calreticulin, a paralog that is a luminal ER protein and whose functions overlap with calnexin (Anelli and Sitia, 2008).

To address the cellular roles of calnexin, we have investigated the consequence of its deletion in the genetically tractable Schizosaccharomyces pombe (Elagoz et al., 1999; Jannatipour and Rokeach, 1995). Table 1 lists the S. pombe strains used in this study. The calnexin homolog of S. pombe, Cnx1p, displays a high degree of sequence identity and similarity with its mammalian counterparts. Importantly, S. pombe does not encode a calreticulin ortholog, thus allowing direct analysis of the cellular roles of calnexin. We and another group have demonstrated that, under normal conditions, calnexin (Cnx1p) is essential for viability (Jannatipour and Rokeach, 1995; Parlati et al., 1995). Interestingly, certain calnexin mutants devoid of chaperone function, such as mini_Cnx1p (with only 52 amino acids in the lumen), do complement the calnexin knockout (Elagoz et al., 1999; Marechal et al., 2004). These observations demonstrate that the essential function of calnexin in S. pombe is not its chaperone activity but another undefined cellular role.

Table 1.

Yeast strains used for this study

Strain Genotype Source
SP247  h+ade6-M210 ura4-D18 leu1-32 his3D1  (Burke and Gould, 1994)  
SP248  h- ade6-M216 ura4-D18 leu-32 his3D1  (Burke and Gould, 1994)  
SP556  h+ade6-M216 ura4-D18 leu1-32  P. Nurse Lab  
SP3220 (Cdn)   SP248 Δcnx1::his3 + pREP41cnx1+  (Elagoz et al., 1999)  
SP6089   SP248 Δcnx1::his3 + pREP42cnx1+  (Elagoz et al., 1999)  
SP8433  h+ade6M210 ura4D18 leu1-32 Δcnx1::his3 + pREP42cnx1+  This study  
SP7188 (CinΔhed_cnx1)   SP248 Δcnx1::his3 [cif] (derivative of SP3222)   (Collin et al., 2004)  
SP8599   SP6089 + pREP3X   This study  
SP8603   SP6089 + pREP3Xegfp  This study  
SP8520   SP6089 + pREP3Xcif1+  This study  
SP8537   SP6089 + pREP41Xcif1+  This study  
SP8680 (Cincif1)   SP248 Δcnx1::his3 [cif] (derivative of SP8520)   This study  
SP8711   SP8680 + pREP41Xcif1  This study  
SP17814   SP6089 + pREP1cif1KRKR27-30/AAAA  This study  
SP7202 (CinΔhed_cnx1 + pcnx1)   SP7188 + pREP41cnx1  (Collin et al., 2004)  
SP8699 (Cincif1 + pcnx1)   SP8680 + pREP41cnx1  This study  
SP8887  h+/h+ade6-M210/ade6-M216 ura4-D18/ura4-D18 leu1-32/leu1-32 cif1::kanMX4/cif1+  Bioneer  
SP8912  h+ade6-M210 ura4-D18 leu1-32 his3D1 cif1::kanMX4  This study  
SP8952  h+ade6-M210 ura4-D18 leu1-32 his3D1 cnx1::his3 cif1::kanMX4 pREP42/cnx1+  This study  
SP8991   SP8952 + pREP41/Δhcd_cnx1  This study  
SP8997   SP8952 + pREP3X/cif1+  This study  
SP10370   SP556 + pREP41X/fib1-mRFP + pREP42cif1venus  This study  
SP10372   SP7188 + pREP41X/fib1-mRFP + pREP42cif1venus  This study  
SP10376   SP8680 + pREP41X/fib1-mRFP + pREP42cif1venus  This study  
SP10373   SP556 + pREP41X/fib1-mRFP + pREP42cif1KRKR27-30/AAAAvenus  This study  
SP10374   SP7188 + pREP41X/fib1-mRFP + pREP42cif1KRKR27-30/AAAAvenus  This study  
SP10375   SP8680 + pREP41X/fib1-mRFP + pREP42cif1KRKR27-30/AAAAvenus  This study  
Strain Genotype Source
SP247  h+ade6-M210 ura4-D18 leu1-32 his3D1  (Burke and Gould, 1994)  
SP248  h- ade6-M216 ura4-D18 leu-32 his3D1  (Burke and Gould, 1994)  
SP556  h+ade6-M216 ura4-D18 leu1-32  P. Nurse Lab  
SP3220 (Cdn)   SP248 Δcnx1::his3 + pREP41cnx1+  (Elagoz et al., 1999)  
SP6089   SP248 Δcnx1::his3 + pREP42cnx1+  (Elagoz et al., 1999)  
SP8433  h+ade6M210 ura4D18 leu1-32 Δcnx1::his3 + pREP42cnx1+  This study  
SP7188 (CinΔhed_cnx1)   SP248 Δcnx1::his3 [cif] (derivative of SP3222)   (Collin et al., 2004)  
SP8599   SP6089 + pREP3X   This study  
SP8603   SP6089 + pREP3Xegfp  This study  
SP8520   SP6089 + pREP3Xcif1+  This study  
SP8537   SP6089 + pREP41Xcif1+  This study  
SP8680 (Cincif1)   SP248 Δcnx1::his3 [cif] (derivative of SP8520)   This study  
SP8711   SP8680 + pREP41Xcif1  This study  
SP17814   SP6089 + pREP1cif1KRKR27-30/AAAA  This study  
SP7202 (CinΔhed_cnx1 + pcnx1)   SP7188 + pREP41cnx1  (Collin et al., 2004)  
SP8699 (Cincif1 + pcnx1)   SP8680 + pREP41cnx1  This study  
SP8887  h+/h+ade6-M210/ade6-M216 ura4-D18/ura4-D18 leu1-32/leu1-32 cif1::kanMX4/cif1+  Bioneer  
SP8912  h+ade6-M210 ura4-D18 leu1-32 his3D1 cif1::kanMX4  This study  
SP8952  h+ade6-M210 ura4-D18 leu1-32 his3D1 cnx1::his3 cif1::kanMX4 pREP42/cnx1+  This study  
SP8991   SP8952 + pREP41/Δhcd_cnx1  This study  
SP8997   SP8952 + pREP3X/cif1+  This study  
SP10370   SP556 + pREP41X/fib1-mRFP + pREP42cif1venus  This study  
SP10372   SP7188 + pREP41X/fib1-mRFP + pREP42cif1venus  This study  
SP10376   SP8680 + pREP41X/fib1-mRFP + pREP42cif1venus  This study  
SP10373   SP556 + pREP41X/fib1-mRFP + pREP42cif1KRKR27-30/AAAAvenus  This study  
SP10374   SP7188 + pREP41X/fib1-mRFP + pREP42cif1KRKR27-30/AAAAvenus  This study  
SP10375   SP8680 + pREP41X/fib1-mRFP + pREP42cif1KRKR27-30/AAAAvenus  This study  

Previous structure-function studies of S. pombe calnexin led us to an unexpected observation. A particular mutant of calnexin that lacks its highly conserved central domain (Δhcd_cnx1) induces a state in which cnx1+ is no longer required for viability (Collin et al., 2004). The sole expression of this mutant at an endogenous level, in presence or not of a wild-type (WT) copy of calnexin, is sufficient to induce the stable state of viability without calnexin in the vast majority of cells. We named this epigenetic state of viability without calnexin Cin (for calnexin-independent). Accordingly, calnexin-less cells were named Cin. By contrast, the state of naïve cells requiring cnx1+ for viability was renamed Cdn (for calnexin-dependent). Extensive characterization of these calnexin-less cells revealed that the Cin state is mediated by a non-chromosomal element that exhibits prion-like features (Collin et al., 2004). Namely, we showed that the Cin state is stable, is dominant in diploids and in merodiploids harboring a cnx1+ plasmid, and is transmitted in a non-Mendelian fashion. Surprisingly, Cin cells are not sensitive to folding stresses, but have cell-wall defects that are corrected by reintroduction of calnexin (Turcotte et al., 2007). These observations indicate that although the Cin state confers viability, it does not compensate for all calnexin functions in S. pombe. As the essential function(s) of S. pombe calnexin remain(s) unknown, the elucidation of the mechanism underlying the Cin state should be instrumental in unraveling novel cellular roles of this ER protein.

So far, the Δhcd_Cnx1p mutant protein is the only known inducer of the Cin state. In order to identify cellular components involved in the triggering and maintenance of the Cin state, we carried out two genetic screens. These screens led to the identification of an uncharacterized gene, which we named cif1+ (for calnexin-independence factor 1). Interestingly, cif1+ codes for a previously uncharacterized nucleolar protein. In this paper, we report that the transient overexpression of cif1+ induces the appearance of a stably inheritable state of calnexin independence. Hence, it provokes the appearance of Cin cells. The Cin cells obtained by cif1+ overexpression exhibited phenotypic differences compared with Cin cells that were induced by the Δhcd_cnx1 calnexin mutant. Moreover, the knockout of cif1+ does not impair the induction or the maintenance of the Cin state induced by the Δhcd_cnx1 mutant. These observations suggest the existence of diverse mechanisms of induction and/or maintenance of the Cin state in these two types of Cin cells. Accordingly, we called them CinΔhcd_cnx1 and Cincif1 to reflect these distinctions. Finally, we showed that nucleolar localization of the calnexin-independence factor Cif1p is required for the induction of the Cin state by cif1+ overexpression.

Results

Identification of cif1+

To identify genes involved in the mechanisms underlying the Cin state, we carried out two unbiased genetic screens scoring for the appearance of Cin cells (see the Materials and Methods). One screen consisted in the overexpression of a S. pombe cDNA library. As the overexpression of a gene creates, in general, a downstream bypass in a given pathway (de la Cruz et al., 1998; Ramer et al., 1992), with this screen we expected to identify candidates having a positive role in the induction and/or maintenance of the Cin state. The second genetic screen consisted of the random mutagenesis by insertion of a ura4+ cassette in the genome. In this case, calnexin-independent cells would appear following the knockout of a gene blocking the induction of the Cin state. Each screen was carried out with cells deleted for calnexin but carrying a plasmid bearing a WT copy of cnx1+ (cnx1Δ + pcnx1+). Following culture in non-selective media, candidate Cin cells were isolated by scoring for cells that could still grow after the loss of the WT calnexin plasmid (supplementary material Fig. S1, and see screens in the Materials and Methods). That the candidate clones were authentic Cin cells was confirmed by western blotting with anti-Cnx1p antibodies (data not shown). The two genetics screens led to the identification of several candidate genes (see the Materials and Methods). The open reading frame (ORF) SPCC364.01 was isolated in multiple Cin clones isolated from both screens and, therefore, we chose to investigate it further. In the ura4+-insertion screen, the cassettes were all integrated at different sites within the promoter region of the gene. The insertion of the ura4+ cassettes did not disrupt the gene but affected its regulation.

Fig. 1.

Primary structure of Cif1p. (A) Predicted amino acid sequence of the S. pombe protein encoded by the open reading frame SPCC364.01, as retrieved from GenBank. The predicted NLS is underlined, and basic residues are in bold. (B) Schematic representation of Cif1p. The NLS is in gray, and the quadruple replacement of KRKR27-30 by alanine residues is indicated.

Fig. 1.

Primary structure of Cif1p. (A) Predicted amino acid sequence of the S. pombe protein encoded by the open reading frame SPCC364.01, as retrieved from GenBank. The predicted NLS is underlined, and basic residues are in bold. (B) Schematic representation of Cif1p. The NLS is in gray, and the quadruple replacement of KRKR27-30 by alanine residues is indicated.

According to the Sanger center database (http://www.genedb.org), this ORF is an uncharacterized gene (for sequence, see Fig. 1A). The predicted protein is rich in basic residues and has an isoelectric point (pI) of 9.6 (http://ca.expasy.org/). Similarity searches revealed a sequence homolog in the closely related fungus Schizosaccharomyces japonicus. No significant sequence alignments with proteins in other organisms were detected. As it induces the Cin state, SPCC364.01 was renamed cif1+.

Overexpression of cif1+ induces a Cin state with distinct propagation properties

To confirm its involvement in the induction of the calnexin-independent state, cif1+ was cloned under the control of the strong nmt1 promoter in the pREP3X overexpressing vector. This construction was transformed into a Δcnx1 + pcnx1+ strain. A plasmid segregation assay was performed to assess whether overexpression of Cif1p induces the appearance of Cin cells. Briefly, cells were cultured in non-selective media for 6 days. Without selective pressure, the cells adopting the Cin epigenetic state lose the calnexin-encoding plasmid because calnexin is not essential for their viability. To discriminate between Cin and Cdn cells, the cultures were plated on non-selective media and the colonies scored on selective plates for the presence or absence of the calnexin-encoding plasmid (for more details, see the Materials and Methods and supplementary material Fig. S1). As shown in Table 2, line 4, overexpression of cif1+ consistently induced the state of viability without calnexin, demonstrated by the presence of calnexin-free cells. Otherwise, an empty plasmid or the overexpression of a non-specific protein like EGFP had no effect on the requirement of calnexin for cell survival, because all cells still contained the calnexin-encoding plasmid (Table 2, lines 2 and 3). The absence or presence of calnexin in these Cin cells was verified by northern and western blotting. The frequency of appearance of calnexin-free cells was about 50%, indicating that the Cin state does not arise from spontaneous mutations or as a result of clonal selection. Interestingly, expression of cif1+ under the control of the more moderate nmt41 promoter did not induce the Cin state (Table 2, lane 5). Thus, these Cin cells most probably arise from the elevated levels of Cif1p, which probably causes a bypass in the pathway inducing the Cin state.

Table 2.

Overexpression of cif1+ induces the Cincif1 state

Constructs Appearance of calnexin-free cells
(1) Δcnx1 + pcnx1+  −  
(2) Δcnx1 + pcnx1+ + pREP3x   −  
(3) Δcnx1 + pcnx1+ + pREP3xegfp  −  
(4) Δcnx1 + pcnx1+ + pREP3xcif1+  +  
(5) Δcnx1 + pcnx1+ + pREP41xcif1+  −  
(6) Cincif1 + pREP42cnx1+  +  
(7) Δcnx1 + pcnx1+ + pREP1cif1KRKR27-30/AAAA  −  
Constructs Appearance of calnexin-free cells
(1) Δcnx1 + pcnx1+  −  
(2) Δcnx1 + pcnx1+ + pREP3x   −  
(3) Δcnx1 + pcnx1+ + pREP3xegfp  −  
(4) Δcnx1 + pcnx1+ + pREP3xcif1+  +  
(5) Δcnx1 + pcnx1+ + pREP41xcif1+  −  
(6) Cincif1 + pREP42cnx1+  +  
(7) Δcnx1 + pcnx1+ + pREP1cif1KRKR27-30/AAAA  −  

Plasmid segregation assays were carried out by culturing Δcnx1 + pcnx1+ cells in non-selective liquid medium and scoring auxotrophies or prototrophies on selective media, as described in the Materials and Methods and supplementary material Fig. S1. − indicates no appearance of calnexin-free cells, and + indicates stable maintenance of the Cin state. The results are representative of three independent experiments.

Importantly, the continuous overexpression of Cif1p was not required for the maintenance of the Cin state, because cells without calnexin could also lose the plasmid that overexpresses Cif1p (data not shown). This observation suggests that cif1+ is found upstream of the inheritance machinery of the Cin state. Finally, the Cin state obtained by overexpression of Cif1p is dominant over calnexin in haploid cells. After re-introduction of pcnx1+, the plasmid was readily lost when not selected for in the plasmid segregation assay (Table 2, line 6).

We previously observed that the Cin state induced by the calnexin mutant Δhcd_Cnx1p (CinΔhcd_cnx1) was dominant in diploid cells (Collin et al., 2004), indicating that the CinΔhcd_cnx1 state stably supersedes the essentiality of calnexin. To further examine the Cin state induced by cif1+ overexpression (Cincif1), Cincif1 cells were mated to cells that were dependent on calnexin for survival, the Cdn strain (Δcnx1 + pcnx1+). Following this, the ability of diploids to lose the calnexin plasmid was assessed by the plasmid segregation assay. Intriguingly, only five of nine diploids tested could stably live without calnexin (Table 3, line 2). This result contrasts with those obtained from diploids formed with CinΔhcd_cnx1 cells, in which all diploids analyzed retained this ability (Table 3, line 1) (Collin et al., 2004). The partial dominance of Cincif1 in diploid cells suggests that this epigenetic state is not stable enough to be maintained in all the diploids. Because Cincif1 haploids remain stably in the calnexin-independent state after re-introduction of the pcnx1+ plasmid but are unstable in diploids, a possibility is that the Cincif1 state is lost in certain cases during mating. Alternatively, the Cincif1 state could be less stable in the 2n background, perhaps owing to a dosage effect of certain gene(s).

Table 3.

Cincif1 is partially dominant in diploid cells

Diploid Total cells Cin cells %of Cin
Δcif1 Δcnx1 pREP42cnx1+ × cif1+ Δcnx1 (CinΔhed_enx1)   7   7   100  
Δcif1 Δcnx1 pREP42cnx1+ × cif1+ Δcnx1 (Cincif1)   9   5   56  
Diploid Total cells Cin cells %of Cin
Δcif1 Δcnx1 pREP42cnx1+ × cif1+ Δcnx1 (CinΔhed_enx1)   7   7   100  
Δcif1 Δcnx1 pREP42cnx1+ × cif1+ Δcnx1 (Cincif1)   9   5   56  

Δcnx1 + pcnx1+ cells were mated with CinΔhed_cnx1 or Cincif1 cells, and submitted to a plasmid segregation assay. For each of these independent diploids, over 200 colonies were streaked to examine the absence of calnexin (Cin state).

In the previous characterization of the CinΔhcd_cnx1 state, we showed that it was transmitted in a non-Mendelian fashion to the meiotic progeny (Collin et al., 2004). To further our characterization, random spore analysis was performed on different diploids constructed by mating of CinΔhcd_cnx1 and Cincif1cnx1 Cin) with the calnexin-dependant Cdn cells containing a genomic disruption of cnx1+ and a calnexin-encoding plasmid (Δcnx1 pcnx1+). As previously observed (Collin et al., 2004), 85% of the meiotic progeny of a CinΔhcd_cnx1 × Cdn diploid were in the state of calnexin independence. By contrast, no Cin cells were obtained in the analysis of more than 250 spores from a Cincif1 × Cdn diploid (not shown). These results suggest either that Cincif1 spores do not form colonies because of a defect in germination, or that the Cincif1 state cannot be propagated to the meiotic progeny. Currently, we cannot distinguish between these two possibilities. In any case, these observations indicate that the Cincif1 and the CinΔhcd_cnx1 states exhibit distinct propagation characteristics, and are therefore different.

Cincif1 cells display phenotypic differences with CinΔhcd_cnx1 cells

We have previously observed that, although the CinΔhcd_cnx1 state allows viability, it does not complement all the functions mediated by calnexin: CinΔhcd_cnx1 cells display cell-wall defects and in stationary phase they exhibit increased caspase activation (Turcotte et al., 2007). Because the CinΔhcd_cnx1 and Cincif1 states showed different properties in their propagation, it was of interest to investigate whether they exhibit additional phenotypic distinctions. Calcofluor is a fluorescent dye that binds polysaccharides concentrating in the septum of dividing cells, and it inhibits the growth of mutants that are defective in cell-wall morphogenesis (Collin et al., 2004; Elagoz et al., 1999; Hampsey, 1997; Rose and Fink, 1987; Turcotte et al., 2007). Interestingly, in microscopy analysis with Calcofluor, CinΔhcd_cnx1 cells grown at 30°C showed fluorescence staining of the septa, whereas some Cincif1 cells exhibited Calcofluor-labeling on the tip instead of the septum (Fig. 2A). When cultured at 37°C, both types of Cin cells were smaller, irregularly shaped and accumulated cell-wall material highly stained with Calcofluor white in the form of large vesicles (Fig. 2A). These observations demonstrate that, like CinΔhcd_cnx1, Cincif1 cells exhibit temperature-dependent defects in the synthesis and/or assembly of cell-wall components. Re-introduction of a plasmid encoding calnexin (pcnx1+) rescued the defects observed at 30°C for the two types Cin of cells, but a clear difference was still observable at 37°C (Fig. 2A) (panels CinΔhcd_cnx1 + pcnx1+ and Cincif1 + pcnx1+). In fact, although CinΔhcd_cnx1 + pcnx1+ displayed WT morphology, Cincif1 + pcnx1+ cells were larger and sometimes had multiple septa. Thus, the adaptative changes in Cincif1 cells that allow them to live in the absence of calnexin cause phenotypic characteristics that cannot be reverted when calnexin is restored.

To further characterize the Cincif1 cells, the sensitivity to hygromycin B was examined by spotting the different strains on media containing this antibiotic. Hygromycin B lowers the viability of S. pombe cells with defects in the early stages of glycoprotein biosynthesis (Turcotte et al., 2007). As previously described, CinΔhcd_cnx1 cells showed sensitivity to hygromycin B, and this phenotype was rescued by reintroduction of cnx1+ on a plasmid (Fig. 2B) (Turcotte et al., 2007). Cincif1 cells were more sensitive to hygromycin B than CinΔhcd_cnx1 cells. Interestingly, reintroduction of a calnexin-encoding plasmid into Cincif1 cells only partially rescued this cell-wall phenotype. SDS is a detergent that disturbs the membrane and exacerbates cell-wall defects (Turcotte et al., 2007). At 30°C, the Cincif1 cells were more sensitive to SDS than the CinΔhcd_cnx1 strain (Fig. 2C). Intriguingly, the differences between the two types of Cin cells were less pronounced at 37°C, possibly because the sum of both SDS and temperature stresses is too severe for both types of Cin cells to cope with. Here again, reintroduction of cnx1+ rescued the sensitivity to SDS of the CinΔhcd_cnx1 strain, but the phenotype of the Cincif1 cells was not completely reversed. Thus, these observations indicate that cell-wall formation is more affected in Cincif1 cells than in CinΔhcd_cnx1 or WT cells.

Caspase activation is a typical marker of induction of programmed cell death (PCD) (Kaufmann et al., 2008; Madeo et al., 2002). In yeast, metacaspase activation can be readily assessed by fluorescence-activated sorting (FACS) of cells stained with the fluorescent probe FITC-VAD-fmk (Herker et al., 2004; Madeo et al., 2002; Roux et al., 2006). As previously reported, CinΔhcd_cnx1 cells display higher levels of metacaspase activation than WT cells in stationary phase, whether they contain a cnx1+ plasmid or not (Turcotte et al., 2007). As shown in Fig. 2D, the Cincif1 strain exhibited more elevated levels of metacaspase activation than CinΔhcd_cnx1 cells in both the exponential and the stationary phase. In contrast to CinΔhcd_cnx1 cells, reintroduction of calnexin into Cincif1 cells reduced the FITC staining to levels that were comparable with those of CinΔhcd_cnx1 + pcnx1 cells, thus suggesting that, in part, this metacaspase phenotype is related to the absence of calnexin in Cincif1 cells. As the Cincif1 state is less stable than the CinΔhcd_cnx1 state, a possibility is that a certain percent of the population could revert to Cdn cells without calnexin, giving rise to non-viable cells rescued by reintroduction of calnexin. Taken together, the distinct cell-wall and metacaspase phenotypes further indicate that the Cincif1 and CinΔhcd_cnx1 states are different.

cif1+ is not essential for vegetative growth of S. pombe

To determine whether Cif1p is an essential protein, several clones from a cif1+ × cif1::kanMX4 diploid strain were sporulated and the meiotic progeny analyzed. The diploids were constructed with two different ade6 alleles (ade6M210 and ade6M216); streaking on low adenine media showed that the spores segregate in a 2:2 ratio. The kanMX4 cassette confers resistance to the G418 antibiotic, which allows the scoring of cif1::kanMX4. Restreaking of the germinated spores on YE + G418 (150 μg/ml) showed a 2:2 segregation of Δcif1 and cif1+ (Fig. 3A). All the spores from the five dissected asci grew at the same rate, showing that deletion of cif1+ does not cause gross growth defects (Fig. 3A). Deletion of cif1+ in G418-resistant haploids was confirmed by northern blotting (Fig. 3B). Thus, these results show that cif1+ is not required for vegetative growth of S. pombe, because cif1::kanMX4 cells exhibited normal morphology and growth rates (Fig. 3A and not shown).

Fig. 2.

Cincif1 cells display phenotypic differences with CinΔhcd_cnx1 cells. (A) Calcofluor staining of Cin cells. Log phase cells (OD595 0.8) were cultured at 30°C or 37°C, and incubated with 20 μg/ml Calcofluor white for 5 minutes. Nomarski-interference images of the same fields of cells stained with Calcofluor are shown above. (B) Sensitivity to cell-wall stress. Exponentially growing cells were adjusted at OD595 0.8, serially diluted (10-1 to 10-4), and spotted on MM plates with or without the indicated chemicals. Plates were incubated at 30°C for 7 days. Results are representative of three independent assays. (C) SDS sensitivity of Cin cells. A volume of 450 μl of a culture at exponential phase (OD595 0.5) was added to 13 ml MM+0.7% agar, and poured on plates containing MM+2% agar. A circle of 3M paper of 5 mm diameter was placed in the center of each plate and 10 μl 10% SDS was added on the circle. Petri dishes were incubated for 7 days at 30°C or at 37°C. The ratio of halos was calculated with respect to WT at 30°C. Results are the average of three independent assays. (D) Metacaspase activity of Cin cells. Metacaspase activity was evaluated by the CaspACE FITC-VAD-fmk fluorescence assay on exponential-phase and 4-day stationary-phase cultures. Stained cells were counted by flow cytometry analysis. Samples of 10,000 cells were analyzed from at least three independent cultures for each strain. Results are the mean of these assays. In (C) and (D), the statistical significance of differences in the results was evaluated by a Student's t-test, pairwise calculated with the lanes indicated. **P<0.01, *P<0.05. Error bars: s.e.m.

Fig. 2.

Cincif1 cells display phenotypic differences with CinΔhcd_cnx1 cells. (A) Calcofluor staining of Cin cells. Log phase cells (OD595 0.8) were cultured at 30°C or 37°C, and incubated with 20 μg/ml Calcofluor white for 5 minutes. Nomarski-interference images of the same fields of cells stained with Calcofluor are shown above. (B) Sensitivity to cell-wall stress. Exponentially growing cells were adjusted at OD595 0.8, serially diluted (10-1 to 10-4), and spotted on MM plates with or without the indicated chemicals. Plates were incubated at 30°C for 7 days. Results are representative of three independent assays. (C) SDS sensitivity of Cin cells. A volume of 450 μl of a culture at exponential phase (OD595 0.5) was added to 13 ml MM+0.7% agar, and poured on plates containing MM+2% agar. A circle of 3M paper of 5 mm diameter was placed in the center of each plate and 10 μl 10% SDS was added on the circle. Petri dishes were incubated for 7 days at 30°C or at 37°C. The ratio of halos was calculated with respect to WT at 30°C. Results are the average of three independent assays. (D) Metacaspase activity of Cin cells. Metacaspase activity was evaluated by the CaspACE FITC-VAD-fmk fluorescence assay on exponential-phase and 4-day stationary-phase cultures. Stained cells were counted by flow cytometry analysis. Samples of 10,000 cells were analyzed from at least three independent cultures for each strain. Results are the mean of these assays. In (C) and (D), the statistical significance of differences in the results was evaluated by a Student's t-test, pairwise calculated with the lanes indicated. **P<0.01, *P<0.05. Error bars: s.e.m.

In order to examine whether the CinΔhcd_cnx1 or the Cincif1 states affect the levels of Cif1p, cell extracts from these strains were fractionated by SDS-PAGE and immunoblotted with polyclonal antibodies elicited against recombinant Cif1p. As shown in Fig. 3C, western blotting with anti-Cif1p showed a major band around 37 kDa in all the strains tested, except for the Δcif1 cells. The levels of Cif1p in WT and both Cin strains were similar, showing that the continuous overexpression of cif1+ is not necessary for the maintenance of either of the Cin states.

Cif1p is not required for induction of the CinΔhcd_cnx1 state or for the maintenance of either Cin state

Because overexpression of cif1+ stably induces the Cincif1 state, we asked whether this gene was required for the induction of the CinΔhcd_cnx1 state. To this end, we constructed a test strain Δcif1 Δcnx1 in which the calnexin knockout is complemented with a plasmid coding for cnx1+. This strain was submitted to a plasmid segregation assay. Deletion of cif1+ did not provoke the spontaneous appearance of Cin cells (Table 4, line 1). Thus, it is unlikely that Cif1p acts as a negative regulator of the Cin state because in such case its deletion would have caused the spontaneous appearance of cells viable without calnexin.

Table 4.

cif1+ is not required for the induction of the CinΔhcd_cnx1 state

Constructs Appearance of calnexin-free cells
(1) Δcif1 Δcnx1 pREP42cnx1+  −  
(2) Δcif1 Δcnx1 pcnx1 pREP41 Δhcd_cnx1  +  
(3) Δcif1 Δcnx1 pcnx1 pREP3xcif1+  +  
Constructs Appearance of calnexin-free cells
(1) Δcif1 Δcnx1 pREP42cnx1+  −  
(2) Δcif1 Δcnx1 pcnx1 pREP41 Δhcd_cnx1  +  
(3) Δcif1 Δcnx1 pcnx1 pREP3xcif1+  +  

Plasmid segregation assays were carried out by growing Δcnx1 Δcif1 + pcnx1+ cells in non-selective medium and scoring auxotrophies or prototropies on selective media, as described in the Materials and Methods. The results are representative of three independent experiments.

To assess for the induction of the CinΔhcd_cnx1 state in a cif1 disruption background, the Δcif1 Δcnx1 + pcnx1+ strain was transformed with plasmids expressing Δhcd_Cnx1p at normal levels. If cif1+ was required for the induction of the CinΔhcd_cnx1 state, the Δcif1 Δcnx1 double knockout strain should not give rise to calnexin-less cells. As shown in Table 4 line 2, expression of Δhcd_Cnx1p in the Δcif1 background led to the appearance of calnexin-less cells, showing that cif1+ is neither required for the induction of the CinΔhcd_cnx1 state nor for the maintenance of this state in the mitotic progeny. Next, to assess whether Cif1p is required for the maintenance of the Cincif1 state, we transformed a plasmid overexpressing cif1+ into the test strain Δcif1 Δcnx1 + pcnx1+. Interestingly, the Cincif1 state was induced in the Δcif1 background and it was stably maintained in the mitotic progeny even in the absence of the cif1+-overexpression plasmid (not shown). Thus, once it is triggered by cif1+ overexpression, the Cincif1 state can be maintained without the continuous presence of Cif1p in the cell. These observations demonstrate that the mechanism of maintenance of the Cincif1 state is downstream and independent of cif1+.

Fig. 3.

cif1+ is unessential for vegetative growth. (A) Tetrad analysis of a cif1+ × cif1::kanMX4cif1) diploid. Dissection was performed on MM+AULH, and colonies from germinated spores were streaked on YE+150 μg/ml G418 to verify the presence of the cif1::kanMX4 marker, and on MM+low adenine+ULH to verify the segregation of the ade6 marker. All the tetrads showed 2:2 segregation of both cif1 and ade6 alleles. (B) Northern blot analysis of Cin and Δcif1 cells. Samples of 5 μg of total RNA prepared from exponential phase cells were loaded on a 1.2% agarose-formaldehyde gel. RNA was probed with a 32P-labeled DNA probe encompassing the entire cif1+ or cnx1+ coding sequence. Photograph of the ethidium-bromide (EtBr) staining shows the 18S and 25S rRNAs as loading control. Cells harboring calnexin on a plasmid have higher levels of cnx1+ RNA; however, these cells contain the same amount of Cnx1p, thus confirming that the levels of plasmid-encoded Cnx1p are equivalent to genomic expression. (C) Western blotting of Cin and Δcif1 cells. Cell extracts were made as described in the Materials and Methods, and 10 μg of material was loaded for fractionation on a 10% SDS-PAGE. Proteins were transferred to a nitrocellulose membrane; Western blotting was carried out using rabbit polyclonal anti-Cif1p serum at a 1:5000 dilution, and with anti-Cnx1p polyclonal antibodies at a 1:35,000 dilution. As a loading control, the same membrane was immunoblotted with and anti-tubulin monoclonal antibodies at a 1:5000 dilution.

Fig. 3.

cif1+ is unessential for vegetative growth. (A) Tetrad analysis of a cif1+ × cif1::kanMX4cif1) diploid. Dissection was performed on MM+AULH, and colonies from germinated spores were streaked on YE+150 μg/ml G418 to verify the presence of the cif1::kanMX4 marker, and on MM+low adenine+ULH to verify the segregation of the ade6 marker. All the tetrads showed 2:2 segregation of both cif1 and ade6 alleles. (B) Northern blot analysis of Cin and Δcif1 cells. Samples of 5 μg of total RNA prepared from exponential phase cells were loaded on a 1.2% agarose-formaldehyde gel. RNA was probed with a 32P-labeled DNA probe encompassing the entire cif1+ or cnx1+ coding sequence. Photograph of the ethidium-bromide (EtBr) staining shows the 18S and 25S rRNAs as loading control. Cells harboring calnexin on a plasmid have higher levels of cnx1+ RNA; however, these cells contain the same amount of Cnx1p, thus confirming that the levels of plasmid-encoded Cnx1p are equivalent to genomic expression. (C) Western blotting of Cin and Δcif1 cells. Cell extracts were made as described in the Materials and Methods, and 10 μg of material was loaded for fractionation on a 10% SDS-PAGE. Proteins were transferred to a nitrocellulose membrane; Western blotting was carried out using rabbit polyclonal anti-Cif1p serum at a 1:5000 dilution, and with anti-Cnx1p polyclonal antibodies at a 1:35,000 dilution. As a loading control, the same membrane was immunoblotted with and anti-tubulin monoclonal antibodies at a 1:5000 dilution.

Next, we investigated whether cif1+ plays a role in the transmission of the CinΔhcd_cnx1 state to the meiotic progeny. Accordingly, we constructed a diploid by mating the Δcnx1 Δcif1 + pcnx1+ strain with CinΔhcd_cnx1 cells. This diploid was sporulated and the germinated spores were analyzed by restreaking on selective and non-selective medium, to check for the cif1::kanMX4 marker (G418) and for the presence or absence of cnx1+-encoding plasmid (uracil marker). A total of 1387 spores from different diploids were examined, and 50.2% of the Cin cells were Δcif1 (Table 5). These results demonstrate that cif1+ is not required for the propagation of the CinΔhcd_cnx1 state to the mitotic or the meiotic progeny. Furthermore, these observations suggest that cif1+ is upstream of, or in a parallel pathway to, Δhcd_Cnx1p in the mechanism of induction and maintenance of the CinΔhcd_cnx1 state. Because it is not possible to isolate Cincif1 spores (see above), we were not able to assess the importance of cif1+ in the propagation of the Cincif1 state to the meiotic progeny.

Table 5.

cif1+ is not required for the maintenance of either Cin state

No. of spores studied
Diploid Total Cin spores Cin/cif1+ Cin/Δcif1 %of Cin/Δcif
Δcif1 Δcnx1 pREP42cnx1+ × cif1+Δcnx1 (CinΔhcd_cnx)   1387   660   696   50.2  
No. of spores studied
Diploid Total Cin spores Cin/cif1+ Cin/Δcif1 %of Cin/Δcif
Δcif1 Δcnx1 pREP42cnx1+ × cif1+Δcnx1 (CinΔhcd_cnx)   1387   660   696   50.2  

A diploid was obtained by crossing a CinΔhcd_cnx1 strain (cif1+Δcnx1) and a calnexin-dependant (Cdn) strain deleted for cif1+ (Δcif1 Δcnx1+ pcnx1+), and sporulated in low nitrogen media. Asci were digested overnight with lysing enzymes, and plated on low adenine media to distinguish between haploid (red or pink) and diploid (white) cells. Haploids were then restreaked on YE+G418 150 μg/ml to examine the presence of the cif1::kanMX4 marker, and on selective MM to examine the presence of the pcnx1+ plasmid.

Nucleolar localization of Cif1p is required for the induction of the Cincif1 state

A nuclear localization signal (NLS) was predicted in the N-terminal part of the sequence (Fig. 1A,B) (see http://cubic.bioc.columbia.edu/predictNLS/) (Cokol et al., 2000). To examine its subcellular localization, a Cif1p-Venus fusion was constructed using a variant of the yellow fluorescent protein (Venus; see the Materials and Methods) and this construction was transformed into WT, CinΔhcd_cnx1 and Cincif1 cells. Fluorescence microscopy imaging of fixed cells revealed the presence of foci in the nuclei of the three types of cells. The foci colocalized with a fibrillarin-RFP (Fib1-RFP) fusion, which is a nucleolar marker (Henriquez et al., 1990) (Fig. 4A), but not entirely with the fluorescent marker DAPI, which stains chromatin. The overlay clearly indicates that Cif1p is present in the nucleolus of the three different cell types. These observations demonstrate that the localization of Cif1p is not altered in either the CinΔhcd_cnx1 or the Cincif1 state. Moreover, microscopy examination of DAPI and Fib1-RFP fluorescence revealed that CinΔhcd_cnx1 and Cincif1 states do not affect the shape of the nucleus or the nucleolus. This was confirmed by using one additional marker for each compartment, Htt2p for the nucleus and Gar2p for the nucleolus (see supplementary material Fig. S2).

Next we investigated whether the nucleolar localization of Cif1p is important for the induction of the Cincif1 state. Accordingly, we constructed a Cif1p variant in which the KRKR27-30 residues at the beginning of the predicted NLS were substituted with four alanines (Fig. 1B). A Venus fusion of this Cif1p-KRKR27-30/AAAA mutant displayed fluorescence in the cytosol (Fig. 4B). This result demonstrates the importance of the KRKR27-30 residues for proper localization of Cif1p. The background nuclear staining could be due to interactions with other nuclear proteins or to some residual functionality of the NLS. Remarkably, cells overexpressing the Cif1p KRKR27-30/AAAA mutant did not give rise to calnexin-independent cells (Table 2, line 7). Thus, the correct nuclear and/or nucleolar targeting of Cif1p is necessary for the induction of the Cincif1 state. In addition, it is possible that these residues could be also important for interaction with partners necessary for Cincif1 induction. Importantly, this result rules out the possibility that the Cincif1 state is induced by a nonspecific effect of overloading the cell with this protein, because the presence of Cif1p in the cytosol does not induce the state of calnexin independence. Moreover, these observations suggest that at least part of the pathway compensating for the loss of calnexin function in Cincif1 cells involves nuclear and/or nucleolar components.

Fig. 4.

Nucleolar localization of Cif1p is abolished by the KRKR27-30/AAAA mutation. (A) Cif1p localizes to the nucleolus of WT and both types of Cin cells. Exponentially growing cells expressing a Cif1p-Venus and a Fib1p-mRFP fusion were fixed as described in the Materials and Methods. Slides were mounted with a DAPI-containing media (1 μg/ml DAPI, 1 mg/ml p-phenylenediamine, 90% glycerol). Cells were examined by fluorescence microscopy. Venus, mRFP, DAPI, merge and Nomarski show the same fields of cells for each strain. (B) Replacement of KRKR27-30 with alanines abolishes the nucleolar localization of Cif1p. Exponentially growing cells expressing a Cif1p-KRKR27-30/AAAA-Venus fusion and a Fib1p-mRFP fusion were fixed and analyzed by fluorescence microscopy as in panel A.

Fig. 4.

Nucleolar localization of Cif1p is abolished by the KRKR27-30/AAAA mutation. (A) Cif1p localizes to the nucleolus of WT and both types of Cin cells. Exponentially growing cells expressing a Cif1p-Venus and a Fib1p-mRFP fusion were fixed as described in the Materials and Methods. Slides were mounted with a DAPI-containing media (1 μg/ml DAPI, 1 mg/ml p-phenylenediamine, 90% glycerol). Cells were examined by fluorescence microscopy. Venus, mRFP, DAPI, merge and Nomarski show the same fields of cells for each strain. (B) Replacement of KRKR27-30 with alanines abolishes the nucleolar localization of Cif1p. Exponentially growing cells expressing a Cif1p-KRKR27-30/AAAA-Venus fusion and a Fib1p-mRFP fusion were fixed and analyzed by fluorescence microscopy as in panel A.

Discussion

Here, we identify a protein that can induce a stably inheritable state of calnexin independence in S. pombe. The protein was named Cif1p, and its encoding gene, SPCC364.01, was accordingly renamed cif1+. The fact that we found this gene in several clones in two different genetic screens is indicative of its importance in the triggering of the calnexin-independence mechanism.

The Cin cells obtained by overexpression of cif1+ (Cincif1 cells) exhibit different phenotypic features to CinΔhcd_cnx1 cells, which were characterized previously (Collin et al., 2004; Turcotte et al., 2007). Although both types of Cin cells are unaffected by the ER-stress agent DTT (not shown), Cincif1 cells are more affected than are CinΔhcd_cnx1 cells in the synthesis and/or assembly of the cell wall, and exhibit higher levels of metacaspase activity than CinΔhcd_cnx1 cells. As with the CinΔhcd_cnx1 state, the Cincif1 state is dominant over plasmid-encoded calnexin in haploids. However, unlike the CinΔhcd_cnx1 state, the Cincif1 state is partially dominant in diploid cells (five out of nine diploids studied). Also, unlike CinΔhcd_cnx1 cells, Cincif1 × Cdn diploids do not give rise to Cincif1 sporulants. This suggests that, in Cincif1 spores, the Cincif state does not completely replace the function of calnexin, leading to a defect in the germination process. Alternatively, the Cincif1 state might not be stably propagated to the meiotic progeny. Presently, we cannot distinguish between these two possibilities. Importantly, these observations indicate that at least two different epigenetic states of calnexin independence can arise in S. pombe, CinΔhcd_cnx1 and Cincif1.

In our first characterization, we observed that the propagation of the CinΔhcd_cnx1 epigenetic state is mediated via a non-chromosomal element exhibiting some prion-like features (Collin et al., 2004). Thus, it could be postulated that the distinctions between the two Cin states might be caused by different prion strains. Strains were described for the [PSI+] prion of S. cerevisiae that differed in their phenotypic strength and stability (Collinge and Clarke, 2007; King and Diaz-Avalos, 2004; Safar et al., 1998; Tanaka et al., 2004; Uptain et al., 2001). As CinΔhcd_cnx1 and Cincif1 appear to exhibit different phenotypes and transmission properties, they could be a reflection of prion variants. The fungal prions best characterized present typical hallmark features. In vitro, purified prion proteins like Sup35, Rnq1 and Het-s aggregate into amyloidal fibers able to transmit their corresponding [PSI+], [PIN+] and [Het-s] prion phenotypes by transformation into naïve cells (King and Diaz-Avalos, 2004; Maddelein et al., 2002; Patel and Liebman, 2007; Tanaka et al., 2004). In vivo, overexpression of the prion protein induces the prion conformer and the corresponding phenotype, which are dependent on the continued biosynthesis of the prion protein (reviewed by Benkemoun and Saupe, 2006; Tuite and Cox, 2006; Uptain and Lindquist, 2002; Wickner et al., 2007). As overexpression of Cif1p leads to the Cincif1 state, we examined whether this protein exhibits prion properties. In preliminary in vitro studies, we observed that Cif1p produces aggregates in the shape of fibers, as judged by atomic force microscopy (supplementary material Fig. S3). cif1+ is not required for the maintenance of either the CinΔhcd_cnx1 or the Cincif1 state. Therefore, it is unlikely that the maintenance of either Cin epigenetic state would depend on a mechanism based on the inheritance of Cif1p in a prion conformer. However, this does not exclude the possibility that Cif1p could be a prion protein, nor does it rule out the existence of a prion-based propagation mechanism downstream of Cif1p.

As the Cincif1 state can be maintained in the Δcif1 context, Cif1p is probably a component of a pathway of induction rather than being part of the propagation machinery of calnexin independence per se. That the CinΔhcd_cnx1 epigenetic state can be triggered in the Δcif1 background suggests that Cif1p acts either upstream of Δhcd_Cnx1p or in a parallel induction pathway. Clarification of this point requires further identification of genes involved in the induction and maintenance of these Cin states, as well as the charting of the interactions between these cellular components. However, the fact that the CinΔhcd_cnx1 and Cincif1 states exhibit distinct phenotypic and genetic characteristics is at this point more supportive of the notion that Cif1p and Δhcd_Cnx1p act in parallel pathways.

We discovered the Cin state because it replaces the essential function of calnexin in the Δcnx1 background and this is monitored as the loss of the cnx1+ plasmid in non-selective media. Certainly, the gene encoding calnexin is not readily lost in a natural setting. In nature, the Cin epigenetic states are probably adaptive mechanisms that provide the cell with additional fitness to withstand stresses normally handled by pathways involving calnexin. In fact, it appears that both Cin states have a functional ER despite the absence of calnexin. Namely, Cin cells do not display sensitivity to the ER stressor DTT, and they do not show increased levels of the typical marker of unfolded protein response BiP (Fig. 3C). To obtain clues about the essential function of calnexin and the cellular roles of the Cin states, we are currently screening metabolic and environmental conditions that could lead to the induction of calnexin independence.

The elucidation of the biological activity of Cif1p might provide indication as to the cellular role of the Cin state, but the function(s) of Cif1p is unknown as yet. Similarity searches revealed an ortholog in the recently annotated genome of S. japonicus, but no other significant sequence homologies were found in other organisms. Its basic character (predicted pI 9.6) and the nucleolar localization of Cif1p suggest that it could be a nucleic-acid-binding protein with a potential role in ribosome biogenesis. The lack of straight sequence homologies in the databases does not rule out the possibility of functional homologs in other organisms such as mammals. Several nucleic-acid-binding proteins exhibit three-dimensional structural and functional conservation without obvious sequence similarities (Fribourg and Conti, 2003; Ma et al., 2005; Meka et al., 2003; Nichols et al., 1999). The identification of Cif1p functional homologs should lead to the discovery of Cin-like mechanisms in other organisms.

Interestingly, replacement of the KRKR27-30 residues at the beginning of the predicted NLS by four alanines disrupted the nucleolar localization of Cif1p by affecting its entry in the nucleus (Fig. 4B; supplementary material Fig. S4). Importantly, this mutation obliterates the ability of Cif1p to induce the Cincif1 state by overexpression. This result indicates that the nuclear and/or nucleolar localization of Cif1p is required for induction of the Cincif1 state. However, we cannot exclude the possibility that mutation of the KRKR27-30 residues could, in addition, directly affect the interaction of Cif1p with some partner in the pathway of induction of the Cincif1 state.

We have previously demonstrated that the essential function of calnexin is contained in its C-terminal domain, a stable portion of calnexin devoid of chaperone activity (Elagoz et al., 1999; Marechal et al., 2004). Recently, we reported that the cytosolic tail of calnexin is involved in mediating an ER-stress-induced apoptotic signal (Guerin et al., 2008). Further characterization of this apoptotic pathway and investigation of the cellular function of Cif1p will reveal whether an apoptotic mechanism and the Cin epigenetic state are functionally linked. As the membrane of the ER abuts that of the nucleus, it is possible that Cif1p and the cytosolic tail of calnexin interact biochemically. This possibility was addressed by immunoprecipitation; however, no interaction between Cif1p and calnexin was observed under the conditions tested (supplementary material Fig. S5).

Overexpression of the nucleolar protein Cif1p induces viability in the absence of calnexin. Thus, part of the Cincif1-state induction pathway takes place in the nucleolus. Further characterization of the function(s) of Cif1p and the identification of its interacting factors will allow us to elucidate the pathway in which this nucleolar protein acts, and will also shed light on how the Cin state replaces the essential function of calnexin. Interestingly, the experiments presented here show a connection between an as-yet-uncharacterized nucleolar activity and the replacement of an essential ER chaperone. This supports the notion that the cellular functions of calnexin are not completely understood. Exploring the functions of calnexin might shed light on the interaction of the ER with other organelles such as the nucleolus.

Ribosome biogenesis is tightly coordinated with cell growth, and so are the functions of the ER, which include membrane synthesis (both lipids and proteins) and protein secretion (Lowe and Barr, 2007; Warner, 1999). Our work suggests a communication between the ER and the nucleolus, hinting that these compartments work together for the proper maintenance of the cell. Investigation of the functions of Cif1p should provide clues about the molecular details of a crosstalk between the ER and the nucleolus during their adjustment of cell growth.

Materials and Methods

S. pombe genetics, strains and media

S. pombe strains used in this study are listed in Table 1. Yeast cells were cultured in Edinburgh minimal medium (MM) supplemented with required nutrients as previously described (Moreno et al., 1991) and with 2% glucose, at 30°C (unless indicated otherwise). S. pombe transformations were performed using the poly(ethylene glycol) (PEG)-lithium acetate procedure (Elbe, 1992). A heterozygous diploid strain knockout for SPCC364.01/cif1+ (cif1::kanMX4) was purchased from Bioneer (Daejeon, Korea). Tetrads were dissected on MM supplemented with adenine (A), uracil (U), leucine (L) and histidine (H) (MM+AULH) with a Nikon Eclipse E400 micromanipulator. S. pombe matings and random spore analyses were carried out as previously described (Collin et al., 2004).

Genetic screens

Insertion mutagenesis of S. pombe was performed as described in (Chua et al., 2000). Briefly, a ura4+ cassette was amplified by PCR with a pair of degenerated oligonucleotides: ura4 for 5′-NNNNNNNNNNNNNNNNNNAGCTTAGCTACAAATCCCACTGGC-3′ and ura4 rev for NNNNNNNNNNNNNNNNNNTGTGATATTGACGAAACTTTTTGACAT-3′. Subsequently, cells were cultured on selective medium (MM+A) to select for transformants, following which the cells were cultured in non-selective medium (MM+AU) to eliminate unstable insertions. Finally, the transformed cells were cultured in plasmid loss medium (MM+AL) to allow for the loss of the pcnx1+ plasmid (see supplementary material Fig. S1). Approximately 4×104 cells were screened for the induction of the Cin state by plating on plasmid loss solid medium (MM+AL) and replica-plating colonies on selective media. Colonies growing on the first medium but not the second (i.e. without the plasmid encoding calnexin) were conserved for further analysis. The absence of Cnx1p was confirmed by western blotting. The locus of insertion of the ura4+ cassette was determined by inverse PCR (Chua et al., 2000). For the overexpression screen, a cDNA S. pombe bank on the pREP3X overexpression vector (a kind gift from Chris Norbury, University of Oxford, Oxford, UK) was transformed. The screening protocol to isolate Cin cells was as the insertion strain. Plasmids were extracted from S. pombe Cin candidate clones and transformed into E. coli, as previously described (Alfa et al., 1993). Following which, the cDNA of the candidate clones were identified by automated DNA sequencing. In addition to cif1+, we isolated several candidate genes involved in different cellular pathways. These included the ER translocon subunit Sec61β and the glycolytic enzymes pyruvate kinase and phosphoglycerate kinase. These candidate genes are currently under investigation.

DNA constructs

The coding sequence of cif1+ was obtained by PCR amplification from genomic DNA of WT S. pombe Sp556. The primers used contained XhoI restriction sites (underlined), and are: forward pc364forA (5′-GATCTCGAATGAGTGAAGAAATTATAACA-3′) and reverse pc364revA (5′-GATCTCGAGTCAAATATCATCTTTTTTGGC-3′). The ORF of cif1+ was cloned into the XhoI site of pREP3X or pREP41X vectors. The correct orientation and sequence was verified by digestion with XmnI/BamHI, PCR amplification and sequencing. A plasmid expressing a fusion of Cif1p with the yellow fluorescent protein (YFP) was constructed with the Venus YFP variant (kindly provided by Atsushi Miyawaki, RIKEN, Japan) (see Nagai et al., 2002). Venus was fused to the C-terminal of Cif1p in pREP42 plasmid, in a sequential cloning. First, cif1+ was cloned NdeI-SmaI (restriction sites are underline) using primers: forward Nde1-364-for (5′-AAACATATGAGTGAAGAAATTATA-3′) and reverse pc364rec, containing a site for SmaI (underline) and a KpnI (bold) (5′-GACCCCGGGCTCGAGTTACATGGCATTCAAGTCCTCTTCAGAAATGAGCTTTTGCTCGGTACCAATATCATCTTT-3′). Venus was amplified with the pair of primers GFPCtermlongfor containing a KpnI site (bold) (5′-AAAGGTACC- ATGGTGAGCAAGGGCGAGGAGCTG-3′) and BamHIvenus_stop containing a BamHI site (underline) (5′-AAAGGATCCTTACTTGTACAGCTCGTC-3′). The amplified fragment was digested with KpnI and BamHI, and inserted into the pREP42 plasmid containing cif1+ also digested with these two enzymes. For Cif1p, the KRKR27-30/AAAA mutant, the codons for amino acids 27-30 (KRKR) were changed with four alanine residues using overlap PCR. First, two fragments of cif1+ containing the quadruple mutation were constructed using for the 5′ part, NdeI-364-for and 364-KRKR/AAAA-rev (5′-CTTTTCAGCTCGCACCTCTGAAGCAGCAGCAGCCGGCTGCTTTTCTCCCACTGATGTTTC-3′), and for the 3′ part, 364-KRKR/AAAA-for (5′-GGGAGAAAAGCAGCCGGCTGCTGC TGCTTCAGAGGTGCGAG-3′) and stop-364-BamHI, containing a BamHI site (underlined) (5′-AAAGGATCCTCAAATATCTTTTTT-3′). To obtain the complete mutant, overlap PCR amplification was performed using the two fragments and the NdeI-364-for and stop-364-BamHI primers. To create a fusion of Cif1p-KRKR/AAAA with Venus, the 3′ primer used contained a KpnI site instead of BamHI, and mutants were cloned in frame with Venus in the pREP42 plasmid. Fibrillarin fused with the monomeric red fluorescent protein (mRFP; Roger Y. Tsien, University of California, San Diego, CA) was constructed in two steps. First, mRFP was amplified using a pair of oligonucleotides: mRFP1-forw2 containing a XhoI restriction site (underlined) and a PmeI restriction site (bold) (5′-CCGCTCGAGGTTTAAACGCCTCCTCCGA-3′), and mRFP1-reverse, containing a BamHI site (underlined) (5′-CGGGATCCTTAGGCGCCGGTGGAGTG-3′). Subsequently, the PCR product was digested with XhoI and BamHI and cloned in pREP41X. Fibrillarin was amplified from genomic DNA with Fib1-XhoI-forward (XhoI site underlined) (5′-AGACTCGAGATGGCATATACACCAGGTTCA-3′) and Fib1-PmeI-no-stop-reverse (PmeI site in bold characters) (5′-AGAGTTTAAACCCTGATGTCTCAAGTATTTTCCTAC-3′). Following which, the PCR product was digested with XhoI and PmeI, and inserted into pREP41X-mRFP previously digested with the same enzymes.

Plasmid segregation experiments

S. pombe strains bearing one or two plasmids were cultured for 7 days at 30°C in 5 ml liquid MM with adenine and appropriate supplements to make a non-selective medium (uracil and/or leucine) (see supplementary material Fig. S1).

Phenotypic assays

The stress-resistance assay, sodium dodecyl sulfate (SDS)-sensitivity assays, and in vivo staining for metacaspase activity were performed as described in (Turcotte et al., 2007).

Northern blotting

Northern blotting was carried out as described previously (Jannatipour and Rokeach, 1995). Both the cif1 and the cnx1 probes encompass the full-length gene. The cif1 probe was isolated by XhoI digestion of the pREP3X/cif1 plasmid, and the cnx1 probe was isolated by a BamHI/NdeI digestion of the pREP42/cnx1+ plasmid.

Immunoblotting

Cell extracts were prepared from exponentially growing cells by bead-beating in lysis buffer (50 mM Tris-HCl pH 7.5, 200 mM NaCl, 10 mM MgCl2, 1 mM EDTA, 1 mM DTT, 5% glycerol, 0.1% Triton X-100). Rabbit polyclonal antibody against Cif1p was used at a 1:5000 dilution. Anti-Cnx1p rabbit serum (Jannatipour and Rokeach, 1995) was used at a 1:35,000 dilution. Anti-BiP rabbit serum (Marechal et al., 2004) was used at a 1:20,000 dilution. For loading controls, mouse monoclonal antibody against α-tubulin (Sigma T-5168) was used a 1:5000 dilution. To raise polyclonal antibodies against Cif1p, the protein was expressed in bacteria with a hexahistidine tag and purified in denaturant conditions. This recombinant protein was used to immunize rabbits for producing polyclonal antibodies, and the serum was further purified with acetone powder of Δcif1 cells (Harlow and Lane, 1988).

Microscopy imaging

Calcofluor-white staining was done as previously described in (Elagoz et al., 1999). Cif1p-Venus fusion imaging was performed as described (Guerin et al., 2008). Microscopy analysis was performed with a fluorescence inverted microscope (Nikon TE2000U). Images were acquired using a motion-picture camera, CCD coolSnapFX M 12 bit (Roper Scientific, Tucson, AZ), and treated with UIC Metamorph software.

We thank Chris Norbury (University of Oxford, Oxford, UK) for the gift of the S. pombe cDNA library in the pREP3X overexpression vector, and Atsushi Miyawaki (RIKEN, Japan) for providing the clone encoding the Venus YFP variant. We thank all members of the Rokeach laboratory for fruitful discussions and critical reading of this manuscript. This work was supported by grants of the Canadian Institutes of Health Research (Grant MOP_62703) and National Science and Engineering Research Council (Grant 171325) to L.A.R. P.B.B. received an NSERC PhD studentship. P.B.B., R.G., and C.T. received scholarships from the Faculté des Études supérieures-Départment of Biochemistry, Université de Montréal.

References

Alfa, C., Fantes, P., Hyams, J., McLeod, M. and Warbrick, E. (
1993
).
Experiments with Fission Yeast
. Cold Spring Harbor, NY: Cold Spring Harbor Press.
Amaral, M. D. (
2004
). CFTR and chaperones: processing and degradation.
J. Mol. Neurosci.
23
,
41
-48.
Anelli, T. and Sitia, R. (
2008
). Protein quality control in the early secretory pathway.
EMBO J.
27
,
315
-327.
Bedard, K., Szabo, E., Michalak, M. and Opas, M. (
2005
). Cellular functions of endoplasmic reticulum chaperones calreticulin, calnexin, and ERp57.
Int. Rev. Cytol.
245
,
91
-121.
Benkemoun, L. and Saupe, S. J. (
2006
). Prion proteins as genetic material in fungi.
Fungal Genet. Biol.
43
,
789
-803.
Caramelo, J. J. and Parodi, A. J. (
2008
). Getting in and out from calnexin/calreticulin cycles.
J. Biol. Chem.
283
,
10221
-10225.
Chevet, E., Jakob, C. A., Thomas, D. Y. and Bergeron, J. J. (
1999
). Calnexin family members as modulators of genetic diseases.
Semin. Cell Dev. Biol.
10
,
473
-480.
Chua, G., Taricani, L., Stangle, W. and Young, P. G. (
2000
). Insertional mutagenesis based on illegitimate recombination in Schizosaccharomyces pombe.
Nucleic Acids Res.
28
,
E53
.
Cokol, M., Nair, R. and Rost, B. (
2000
). Finding nuclear localization signals.
EMBO Rep.
1
,
411
-415.
Collin, P., Beauregard, P. B., Elagoz, A. and Rokeach, L. A. (
2004
). A non-chromosomal factor allows viability of Schizosaccharomyces pombe lacking the essential chaperone calnexin.
J. Cell Sci.
117
,
907
-918.
Collinge, J. and Clarke, A. R. (
2007
). A general model of prion strains and their pathogenicity.
Science
318
,
930
-936.
de la Cruz, J., Daugeron, M. C. and Linder, P. (
1998
). Smart genetic screens. In
Yeast Gene Analysis
(ed. A. J. P. Brown and M. F. Tuite), pp.
269
-295. San Diego, CA: Academic Press.
Delom, F., Emadali, A., Cocolakis, E., Lebrun, J. J., Nantel, A. and Chevet, E. (
2006
). Calnexin-dependent regulation of tunicamycin-induced apoptosis in breast carcinoma MCF-7 cells.
Cell Death Differ
.
14
,
586
-596.
Delom, F., Fessart, D. and Chevet, E. (
2007
). Regulation of calnexin sub-cellular localization modulates endoplasmic reticulum stress-induced apoptosis in MCF-7 cells.
Apoptosis
12
,
293
-305.
Denzel, A., Molinari, M., Trigueros, C., Martin, J. E., Velmurgan, S., Brown, S., Stamp, G. and Owen, M. J. (
2002
). Early postnatal death and motor disorders in mice congenitally deficient in calnexin expression.
Mol. Cell. Biol.
22
,
7398
-7404.
Elagoz, A., Callejo, M., Armstrong, J. and Rokeach, L. A. (
1999
). Although calnexin is essential in S. pombe, its highly conserved central domain is dispensable for viability.
J. Cell Sci.
112
,
4449
-4460.
Elbe, R. (
1992
). A simple and efficient procedure for transformation of yeasts.
Biotechniques
13
,
18
-19.
Ellgaard, L. and Helenius, A. (
2003
). Quality control in the endoplasmic reticulum.
Nat. Rev. Mol. Cell. Biol.
4
,
181
-191.
Fewell, S. W., Travers, K. J., Weissman, J. S. and Brodsky, J. L. (
2001
). The action of molecular chaperones in the early secretory pathway.
Annu. Rev. Genet.
35
,
149
-191.
Fribourg, S. and Conti, E. (
2003
). Structural similarity in the absence of sequence homology of the messenger RNA export factors Mtr2 and p15.
EMBO Rep.
4
,
699
-703.
Guerin, R., Arseneault, G., Dumont, S. and Rokeach, L. A. (
2008
). Calnexin is involved in apoptosis induced by endoplasmic reticulum stress in the fission yeast.
Mol. Biol. Cell
19
,
4404
-4420.
Hampsey, M. (
1997
). A review of phenotypes in Saccharomyces cerevisiae.
Yeast
13
,
1099
-1133.
Harlow, E. and Lane, D. (
1988
).
Antibodies: A Laboratory Manual.
Cold Spring Harbor, NY: Cold Spring Harbor Press.
Hebert, D. N. and Molinari, M. (
2007
). In and out of the ER: protein folding, quality control, degradation, and related human diseases.
Physiol. Rev.
87
,
1377
-1408.
Helenius, A. and Aebi, M. (
2004
). Roles of N-linked glycans in the endoplasmic reticulum.
Annu. Rev. Biochem.
73
,
1019
-1049.
Henriquez, R., Blobel, G. and Aris, J. P. (
1990
). Isolation and sequencing of NOP1: a yeast gene encoding a nucleolar protein homologous to a human autoimmune antigen.
J. Biol. Chem.
265
,
2209
-2215.
Herker, E., Jungwirth, H., Lehmann, K. A., Maldener, C., Frohlich, K. U., Wissing, S., Buttner, S., Fehr, M., Sigrist, S. and Madeo, F. (
2004
). Chronological aging leads to apoptosis in yeast.
J. Cell Biol.
164
,
501
-507.
Jannatipour, M. and Rokeach, L. A. (
1995
). The Schizosaccharomyces pombe homologue of the chaperone calnexin is essential for viability.
J. Biol. Chem.
270
,
4845
-4853.
Kaufmann, S. H., Lee, S. H., Meng, X. W., Loegering, D. A., Kottke, T. J., Henzing, A. J., Ruchaud, S., Samejima, K. and Earnshaw, W. C. (
2008
). Apoptosis-associated caspase activation assays.
Methods
44
,
262
-272.
King, C. Y. and Diaz-Avalos, R. (
2004
). Protein-only transmission of three yeast prion strains.
Nature
428
,
319
-323.
Kleizen, B. and Braakman, I. (
2004
). Protein folding and quality control in the endoplasmic reticulum.
Curr. Opin. Cell Biol.
16
,
343
-349.
Kuznetsov, G. and Nigam, S. K. (
1998
). Folding of secretory and membrane proteins.
N. Engl. J. Med.
339
,
1688
-1695.
Leach, M. R., Cohen-Doyle, M. F., Thomas, D. Y. and Williams, D. B. (
2002
). Localization of the lectin, ERp57 binding, and polypeptide binding sites of calnexin and calreticulin.
J. Biol. Chem.
277
,
29686
-29697.
Lee, W., Lee, T. H., Park, B. J., Chang, J. W., Yu, J. R., Koo, H. S., Park, H., Yoo, Y. J. and Ahnn, J. (
2005
). Caenorhabditis elegans calnexin is N-glycosylated and required for stress response.
Biochem. Biophys. Res. Commun.
338
,
1018
-1030.
Lowe, M. and Barr, F. A. (
2007
). Inheritance and biogenesis of organelles in the secretory pathway.
Nat. Rev. Mol. Cell. Biol.
8
,
429
-439.
Ma, Y., Dostie, J., Dreyfuss, G. and Van Duyne, G. D. (
2005
). The Gemin6-Gemin7 heterodimer from the survival of motor neurons complex has an Sm protein-like structure.
Structure
13
,
883
-892.
Maddelein, M. L., Dos Reis, S., Duvezin-Caubet, S., Coulary-Salin, B. and Saupe, S. J. (
2002
). Amyloid aggregates of the HET-s prion protein are infectious.
Proc. Natl. Acad. Sci. USA
99
,
7402
-7407.
Madeo, F., Herker, E., Maldener, C., Wissing, S., Lachelt, S., Herlan, M., Fehr, M., Lauber, K., Sigrist, S. J., Wesselborg, S. et al. (
2002
). A caspase-related protease regulates apoptosis in yeast.
Mol. Cell
9
,
911
-917.
Marechal, A., Tanguay, P. L., Callejo, M., Guerin, R., Boileau, G. and Rokeach, L. A. (
2004
). Cell viability and secretion of active proteins in Schizosaccharomyces pombe do not require the chaperone function of calnexin.
Biochem. J.
380
,
441
-448.
Meka, H., Daoust, G., Arnvig, K. B., Werner, F., Brick, P. and Onesti, S. (
2003
). Structural and functional homology between the RNAP(I) subunits A14/A43 and the archaeal RNAP subunits E/F.
Nucleic Acids Res.
31
,
4391
-4400.
Moremen, K. W. and Molinari, M. (
2006
). N-linked glycan recognition and processing: the molecular basis of endoplasmic reticulum quality control.
Curr. Opin. Struct. Biol.
16
,
592
-599.
Moreno, S., Klar, A. and Nurse, P. (
1991
). Molecular genetic analysis of fission yeast Schizosaccharomyces pombe.
Methods Enzymol.
194
,
795
-823.
Muller-Taubenberger, A., Lupas, A. N., Li, H., Ecke, M., Simmeth, E. and Gerisch, G. (
2001
). Calreticulin and calnexin in the endoplasmic reticulum are important for phagocytosis.
EMBO J.
20
,
6772
-6782.
Nagai, T., Ibata, K., Park, E.S., Kubota, M., Mikoshiba, K. and Miyawaki, A. (
2002
). A variant of yellow fluorescent protein with fast and efficient maturation for cell-biological applications.
Nat. Biotechnol.
20
,
87
-90.
Ni, M. and Lee, A. S. (
2007
). ER chaperones in mammalian development and human diseases.
FEBS Lett.
581
,
3641
-3651.
Nichols, M. D., DeAngelis, K., Keck, J. L. and Berger, J. M. (
1999
). Structure and function of an archaeal topoisomerase VI subunit with homology to the meiotic recombination factor Spo11.
EMBO J.
18
,
6177
-6188.
Parlati, F., Dignard, D., Bergeron, J. J. and Thomas, D. Y. (
1995
). The calnexin homologue cnx1+ in Schizosaccharomyces pombe, is an essential gene which can be complemented by its soluble ER domain.
EMBO J.
14
,
3064
-3072.
Patel, B. K. and Liebman, S. W. (
2007
). “Prion-proof” for [PIN+]: infection with in vitro-made amyloid aggregates of Rnq1p-(132-405) induces [PIN+].
J. Mol. Biol.
365
,
773
-782.
Pollock, S., Kozlov, G., Pelletier, M. F., Trempe, J. F., Jansen, G., Sitnikov, D., Bergeron, J. J., Gehring, K., Ekiel, I. and Thomas, D. Y. (
2004
). Specific interaction of ERp57 and calnexin determined by NMR spectroscopy and an ER two-hybrid system.
EMBO J.
23
,
1020
-1029.
Ramer, S. W., Elledge, S. J. and Davis, R. W. (
1992
). Dominant genetics using a yeast genomic library under the control of a strong inducible promoter.
Proc. Natl. Acad. Sci. USA
89
,
11589
-11593.
Rose, M. D. and Fink, G. R. (
1987
). KAR1, a gene required for function of both intranuclear and extranuclear microtubules in yeast.
Cell
48
,
1047
-1060.
Roux, A. E., Quissac, A., Chartrand, P., Ferbeyre, G. and Rokeach, L. A. (
2006
). Regulation of chronological aging in Schizosaccharomyces pombe by the protein kinases Pka1 and Sck2.
Aging Cell
5
,
345
-357.
Safar, J., Wille, H., Itri, V., Groth, D., Serban, H., Torchia, M., Cohen, F. E. and Prusiner, S. B. (
1998
). Eight prion strains have PrP(Sc) molecules with different conformations.
Nat. Med.
4
,
1157
-1165.
Schrag, J. D., Bergeron, J. J., Li, Y., Borisova, S., Hahn, M., Thomas, D. Y. and Cygler, M. (
2001
). The Structure of calnexin, an ER chaperone involved in quality control of protein folding.
Mol. Cell
8
,
633
-644.
Takizawa, T., Tatematsu, C., Watanabe, K., Kato, K. and Nakanishi, Y. (
2004
). Cleavage of calnexin caused by apoptotic stimuli: implication for the regulation of apoptosis.
J. Biochem. (Tokyo)
136
,
399
-405.
Tanaka, M., Chien, P., Naber, N., Cooke, R. and Weissman, J. S. (
2004
). Conformational variations in an infectious protein determine prion strain differences.
Nature
428
,
323
-328.
Thammavongsa, V., Mancino, L. and Raghavan, M. (
2005
). Polypeptide substrate recognition by calnexin requires specific conformations of the calnexin protein.
J. Biol. Chem.
280
,
33497
-33505.
Tomassini, B., Malisan, F., Franchi, L., Nicolo, C., Calvo, G. B., Saito, T. and Testi, R. (
2004
). Calnexin suppresses GD3 synthase-induced apoptosis.
FASEB J.
18
,
1553
-1555.
Torgler, C. N., de Tiani, M., Raven, T., Aubry, J. P., Brown, R. and Meldrum, E. (
1997
). Expression of bak in S. pombe results in a lethality mediated through interaction with the calnexin homologue Cnx1.
Cell Death Differ.
4
,
263
-271.
Tuite, M. F. and Cox, B. S. (
2006
). The [PSI+] prion of yeast: a problem of inheritance.
Methods
39
,
9
-22.
Turcotte, C., Roux, A., Beauregard, P. B., Guerin, R., Senechal, P., Hajjar, F. and Rokeach, L. A. (
2007
). The calnexin-independent state does not compensate for all calnexin functions in Schizosaccharomyces pombe.
FEMS Yeast Res.
7
,
196
-208.
Uptain, S. M. and Lindquist, S. (
2002
). Prions as protein-based genetic elements.
Annu. Rev. Microbiol.
56
,
703
-741.
Uptain, S. M., Sawicki, G. J., Caughey, B. and Lindquist, S. (
2001
). Strains of [PSI(+)] are distinguished by their efficiencies of prion-mediated conformational conversion.
EMBO J.
20
,
6236
-6245.
Warner, J. R. (
1999
). The economics of ribosome biosynthesis in yeast.
Trends Biochem. Sci.
24
,
437
-440.
Wickner, R. B., Edskes, H. K., Shewmaker, F. and Nakayashiki, T. (
2007
). Prions of fungi: inherited structures and biological roles.
Nat. Rev. Microbiol.
5
,
611
-618.
Williams, D. B. (
2006
). Beyond lectins: the calnexin/calreticulin chaperone system of the endoplasmic reticulum.
J. Cell Sci.
119
,
615
-623.
Zuppini, A., Groenendyk, J., Cormack, L. A., Shore, G., Opas, M., Bleackley, R. C. and Michalak, M. (
2002
). Calnexin deficiency and endoplasmic reticulum stress-induced apoptosis.
Biochemistry
41
,
2850
-2858.

Supplementary information