OUP user menu

The localization change of Ybr078w/Ecm33, a yeast GPI-associated protein, from the plasma membrane to the cell wall, affecting the cellular function

Hiromichi Terashima, Kenji Hamada, Kunio Kitada
DOI: http://dx.doi.org/10.1111/j.1574-6968.2003.tb11515.x 175-180 First published online: 1 January 2003


The YBR078W/ECM33 gene of Saccharomyces cerevisiae encodes a glycosylphosphatidylinositol (GPI)-attached protein and its disruptant strain exhibited a temperature-sensitive (ts) growth defect. A HA-tagged Ybr078w protein, which complemented the ts growth phenotype of the ybr078wΔ strain, was predominantly located on the plasma membrane by GPI anchoring. To examine the requirement of the GPI anchoring on the plasma membrane for the function, the ω-minus region of Ybr078w was replaced with those of Ydr534c/Fit1 and Ynl327w/Egt2, which are known as GPI-dependent cell wall proteins. The replacement induced the change in localization of the mutant proteins from the plasma membrane to the cell wall and the mutant proteins lost the function to complement the ts cell growth defect of the ybr078wΔ strain. In addition, a similar result was obtained in a mutant protein, where the authentic SKKSK sequence at the ω-5 to ω-1 site of Ybr078w was replaced with a synthetic ISSYS sequence. It is concluded that the GPI anchoring on the plasma membrane is required for the Ybr078w function.

  • Ybr078w/Ecm33
  • Protein localization
  • Glycosylphosphatidylinositol-anchored protein
  • Cell wall protein
  • Saccharomyces cerevisiae

1 Introduction

A number of glycosylphosphatidylinositol (GPI)-associated proteins have been identified in Saccharomyces cerevisiae [15]. They have a common structure in the C-terminus, a GPI signal, for an attachment to a GPI. The GPI signal is composed of a ω-site, to which a GPI binds, a spacer of 5–10 amino acids, and a hydrophobic stretch of 10–15 amino acids. A transamidase complex of Gaa1 [6], Gpi8 [7] and Gpi16 [8] recognizes the GPI signal of a protein and cleaves the protein at the ω-site, and then binds the resulting C-terminus of the protein to a GPI. This reaction occurs in the endoplasmic reticulum and the GPI-attached proteins anchoring on membranes of the endoplasmic reticulum are transported to the plasma membrane. During this transportation, the proteins are glycosylated. On the plasma membrane, they are further processed. A portion of the GPI moiety is detached and then the GPI-detached protein is covalently linked to β-glucan of the cell wall [9,10]. This incorporation to the cell wall occurs to various extents among the GPI-attached proteins, resulting in some of the proteins being predominantly located on the plasma membrane and some of the proteins being in the cell wall. Based on the localization preference, the GPI-attached proteins are divided into GPI-anchored plasma membrane proteins (GPI-PMP) and GPI-dependent cell wall proteins (GPI-CWP) [11].

Both GPI-PMP and GPI-CWP exist on the cell surface, but their existing manners are different: GPI-PMP is attached on the plasma membrane by a GPI anchor and GPI-CWP is directly associated with β-glucan of the cell wall components. It is generally known that proper cellular localization of proteins is necessary for their functions. Is it the case for the GPI-PMP and GPI-CWP? Is it necessary for GPI-PMP to be localized on the plasma membrane for its function? In this study, we addressed to this using one of GPI-PMPs, Ybr078w/Ecm33. The YBR078W gene has been identified to encode a GPI-associated protein from sequence analyses [3,4] and a reporter protein fused to its GPI signal has been found to preferentially exist on the plasma membrane in a GPI-anchoring manner [11]. A mutation in the same gene designated ECM33 increases the sensitivity to calcofluor white [12]. A disruptant strain of YBR078W/ECM33 exhibited a temperature-sensitive (ts) growth phenotype [13] and the hypersensitivity to oxidative stress [14]. Using the ts growth defect as an indicator for the Ybr078w function, we examined the function of mutant proteins, whose localization was changed from the plasma membrane to the cell wall.

2 Materials and methods

2.1 Strains and media

A YBR078W/ECM33-disruptant strain 78Wd-5D (Mat a ybr078wΔ::CgHIS3 his3 leu2 trp1 ura3) was used. 78Wd-5D is one of tetrad clones from 78WdD (Mat a/αybr078wΔ::CgHIS3/YBR078W his3/his3 leu2/leu2 trp1/trp1 ura3/ura3, in a back ground of W303-1A/W303-1B). CgHIS3 is the HIS3 gene of Candida glabrata [15]. Yeast selection media SC-Ura and SC-Trp were used for maintaining plasmids in transformant cells [16]. The yeast transformation was carried out using the lithium acetate method [17].

2.2 Plasmids

pEαGALHA-CT78 was constructed by inserting the most C-terminal region of 40 amino acids of Ybr078w/Ecm33 just after a reporter gene encoding a secretion signal peptide, α-galactosidase and a hemagglutinin (HA) epitope on pEαGALHA [4]. For determination of the ω-site of Ybr078w/Ecm33, the G residue at amino acid sequence position 406 of Ybr078w was substituted with either D, E, N, Q or T by using a polymerase chain reaction (PCR)-based site-directed mutagenesis method. The plasmids containing the mutant genes are designated pEαGALHA-CT78ωmX, where X represents a substituted amino acid residue.

p78w was constructed by inserting the entire gene of YBR078W/ECM33 into YCplac22 [18]. The 3xHA epitope sequence was inserted in frame into the 25th amino acid position of the Ybr078w sequence, to generate p78wHA. The G residue at the putative ω-site was substituted with Q, E or a terminal codon (TAA) on p78wHA-ωmQ, p78wHA-ωmE or p78wHA-ωmStop, respectively.

The ω-minus region of 13 amino acids of Ybr078w/Ecm33 on p78wHA was replaced with the corresponding regions of Ydr534c/Fit1 [19], Ynl327w/Egt2 [20] and Ymr307w/Gas1 [21,22], generating plasmids p78wHA-ωm534c, p78wHA-ωm327w and p78wHA-ωm307w, respectively. The sequence of five amino acids in the authentic ω-minus region, SKKSK, of Ybr078w/Ecm33 on p78wHA was exchanged with a synthetic sequence of SSSSS or ISSYS, generating plasmids p78wHA-ωmSSSSS and p78wHA-ωmISSYS, respectively.

2.3 GPI attachment

All the methods including isolation of membrane proteins, phase separation of proteins with Triton X-114, treatment with phosphatidylinositol-specific phospholipase (PI-PLC) (Oxford GlycoSciences), sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE), and Western blotting were previously described [4]. An anti-HA monoclonal antibody (12CA5, Roche Diagnostics) was used for detection of reporter proteins on Western blots.

2.4 Cellular fractionation

According to the method previously described [23], yeast transformant cells were separated from the culture medium (culture broth fraction) and washed with a buffer (wash-1, wash-2 fractions) and with dithiothreitol (DTT) (DTT wash fraction). The washed cells were broken with glass beads to generate crude cell lysate (crude lysate fraction). The lysate was fractionated into soluble, membrane and cell wall fractions. A part of the cell wall fraction was treated with laminarinase (laminarinase-treated cell wall fraction). An equivalent amount of each fraction sample was analyzed with SDS–PAGE and Western blotting.

2.5 Cell wall association

Cell walls were isolated and treated with laminarinase as described previously [24]. The treatment of the cell walls with quantazyme was carried out according to the manufacturer's instructions. 3 µl of quantazyme (20 U µl−1, Quantum Biotechnologies) was added to cell walls in 10 µl buffer (10 mM sodium phosphate, pH 7.5, 150 mM NaCl, 5 mM ethylenediamine tetraacetic acid (EDTA), 2 mM phenylmethylsulfonyl fluoride (PMSF)) and incubated for 2 h at 25°C. After centrifugation of the reaction solution, the supernatant was analyzed by SDS–PAGE followed by Western blotting.

2.6 Cell growth

Overnight culture (10–20 µl) of yeast transformant cells grown in the selection medium was inoculated into the 5 ml fresh selection medium and incubated at indicated temperatures. OD600 was monitored every 20 min by using a BioPhotoRecorder (TN-1506, Advantec).

3 Results

3.1 Temperature-sensitive growth of ybr078w/ecm33Δ cells

A deleted strain of YBR078W/ECM33 has shown a temperature-sensitive (ts) growth phenotype in a previous report [13]. We disrupted one of the two YBR078W genes in a diploid strain using the CgHIS3 gene and isolated tetrads on sporulation of the diploid. The ts growth phenotype segregated 2:2 in all tetrads examined and all spores displaying the ts growth phenotype were His prototrophic. One of the resulting ybr078wΔ strains, designated 78Wd-5D, grew normally at 25°C, but did not grow at 37°C (Fig. 1). This growth defect was rescued by a plasmid containing either the YBR078W gene or an HA-tagged YBR078W gene (Fig. 1). The YBR078W-HA gene has an insertion of the HA epitope sequence at the site corresponding to the 25th amino acid sequence position.

Figure 1

Effect of ω-site mutations on cell growth. The ybr078w/ecm33Δ strain, 78Wd-5D, was transformed with plasmids containing ω-site mutated genes and examined for cell growth at 37°C.

3.2 G as the ω-site residue for GPI attachment

Ybr078w has been identified as a potential GPI-anchored protein and the most C-terminal region has the ability to bind a reporter protein to a GPI [3,4]. A sequence analysis suggested the G residue at amino acid sequence position 406 as a putative ω-site residue to which a GPI binds. To experimentally confirm the ω-site, the G residue was substituted with other amino acid residues in the reporter protein and the resulting mutant proteins were examined for GPI attachment (Fig. 2). Reporter proteins with the substitution of G to N or D were detected in a detergent phase of two-phase participation with Triton X-114 and they were significantly released into an aqueous phase by digestion with PI-PLC, indicating GPI anchoring on membranes. Almost no reporter proteins with the substitution of G to E, Q or T were detected in the detergent phase, indicating no GPI anchoring. This residue selectivity for GPI anchoring efficiency was quite similar to those observed in ω-site mutations of the GPI-associated proteins including Gas1 [21], Sed1, Ylr120c, Yir039c and Yap3 [24]. It was concluded that the G residue served as the ω-site residue for GPI attachment.

Figure 2

Analysis of membrane-bound forms of ω-site mutated reporter proteins. The putative ω-site residue G was substituted with N, E, Q, D and T by using PCR-based mutagenesis in the reporter protein on pEαGALHA-CT78. Reporter proteins extracted in a detergent phase by phase separation with Triton X-114 were treated (+) or not treated (−) with PI-PLC and then separated into a detergent phase (D) and an aqueous phase (A). Proteins on the Western blots were detected with anti-HA monoclonal antibody.

3.3 Growth defects by mutations of the ω-site residue

The G residue of Ybr078w-HA was substituted with E, Q or a stop codon and examined for its effects on cell growth. Neither of the genes with the substitutions restored the cell growth at 37°C to the ybr078wΔ strain (Fig. 1). A Western blot analysis revealed that expression levels of these mutant gene products were comparable to that of the wild-type (data not shown). These results indicated that a proper attachment of the Ybr078w protein to a GPI is required for the cell growth at high temperature.

3.4 Replacement of the ω-minus region

Our previous study has divided GPI-attached proteins into two groups based on relative amounts of proteins incorporated into the cell wall: GPI-PMP and GPI-CWP [11]. The ω-minus region, which is upstream from the ω-site, was found to participate in determination of subcellular localization of GPI-associated proteins [24]. To investigate the relation between the localization of Ybr078w and its function, the ω-minus region of 13 amino acids of Ybr078w-HA was replaced with those of other GPI-associated proteins and their effects on cell growth were examined. Ynr307w/Gas1 as a GPI-PMP and Ydr534c/Fit1 and Ynl327w/Egt2 as GPI-CWPs were used for the construction of mutant proteins, 78wHA-ωm307w, 78wHA-ωm534c and 78wHA-ωm327w, respectively (Fig. 3A). The ts growth defect of the ybr078wΔ strain was partially complemented by 78wHA-ωm 307w but not by 78wHA-ωm 534c and 78wHA-ωm 327w (Fig. 3B). Furthermore, two mutant proteins, 78wHA-ωm SSSSS and 78HA-ωm ISSYS, were constructed by exchanging the SKKSK sequence in the ω-minus region of Ybr078w-HA with synthetic sequences SSSSS and ISSYS, respectively, and examined for effects on cell growth. 78wHA-ωm SSSS partially rescued the ts growth defect of the ybr078Δ strain, while 78wHA-ωm ISSYS did not (Fig. 3B).

Figure 3

Effects of the replacement of the ω-minus region on cell growth and subcellular localization. A: Amino acid sequences of ω-minus regions. The ω-minus region of 13 amino acids of Ybr078w-HA was exchanged with those of other GPI-attached proteins or synthetic amino acid sequences. The amino acid sequences which were exchanged are shown. B: Cell growth. The ybr078wΔ strain, 78Wd-5D, was transformed with plasmids containing mutant genes with the replacement in the ω-minus region and examined for cell growth at 36.5°C. C: Subcellular localization of the mutant proteins. Mutant proteins tagged with an HA epitope were expressed in the 78Wd-5D strain at 30°C. The cells were fractionated and HA-tagged proteins were examined with Western blotting using an anti-HA monoclonal antibody. A sample equivalent to the same amount of cells or cell culture media was loaded on an SDS–PAGE gel. Lanes: 1, crude cell lysate fraction; 2, soluble fraction; 3, membrane fraction; 4, cell wall fraction; 5, laminarinase-treated cell wall fraction; 6, dithiothreitol-washed fraction; 7, buffer wash second fraction; 8, buffer wash first fraction; 9, culture broth fraction. D: Proteins associated with the cell wall. Cell walls were isolated from transformant cells and intensively washed with hot SDS. The cell walls were treated with (lane 2) or without (lane 3) laminarinase and with (lane 4) or without (lane 5) quantazyme and the cell wall-associated proteins were analyzed by Western blotting. Lane 1, crude cell lysate.

Subcellular localization of these mutant proteins was analyzed using a fractionation method [23]. Proteins of 78wHA-ωm 307w and 78wHA-ωm SSSSS exhibited similar patterns in cellular fractions to that of the wild-type Ybr078w-HA (Fig. 3C). On the contrary, the patterns of 78wHA-ωm534c and 78wHA-ωm327w were different from that of Ybr078w-HA. These proteins were detected at the relatively low levels in the membrane fraction and the cell wall fraction and at the relatively high level in the laminarinase-treated cell wall fraction. The protein of 78wHA-ωm ISSYS exhibited a similar pattern to that of Ybr078w-HA except for a little higher amount of the protein in the laminarinase-treated cell wall fraction. These results suggested an increase in the amounts of 78wHA-ωm534w, 78wHA-ωm327w and 78wHA-ωm ISSYS, which were covalently linked to β-glucan of the cell wall. To confirm this, cell walls were isolated, extensively washed with hot SDS, and then proteins associated with the cell wall were analyzed. Laminarinase and quantazyme were used to release the proteins from the cell wall. The amounts of released proteins significantly increased in 78wHA-ωm534c, 78wHA-ωm327w and 78wHA-ωmISSYS compared to that of Ybr078w-HA, but not in 78wHA-ωm307w and 78wHA-ωmSSSS (Fig. 3D). The relative percentages of proteins associated with the cell wall were 10.1, 39.1, 41.3, 9.0, 13.5, and 42.5 for Ybr078w-HA, 78wHA-ωm534c, 78wHA-ωm327w, 78wHA-ωm307w, 78wHA-ωmSSSSS, and 78wHA-ωmISSYS, respectively. These calculations came from the experiment using laminarinase, and the similar results were also obtained when using quantazyme.

4 Discussion

We have shown that Ybr078w predominantly anchored on the plasma membrane with a GPI and that this localization on the plasma membrane is necessary for the cell growth at high temperature. Mutated Ybr078w proteins, where the authentic ω-minus region was replaced with those of Ydr534c/Fit1 and Ynl327w/Egt2, which are known as GPI-CWPs [11,19], failed to complement the ts growth defect of the ybr078wΔ strain. The amounts of these proteins detected in the membrane fraction were significantly reduced and, contrastingly, the amounts of these proteins linked to the cell wall increased. These results indicate that the final localization of these proteins is changed from the plasma membrane to the cell wall. This change in the protein localization probably explains the failure in the complementation of the ts cell growth phenotype. It was supported by an experiment, where the ω-minus region of Ybr078w was replaced with that of Ynr307w/Gas1, which is known as a GPI-PMP. In this replacement, there was no reduction in the amount of proteins located in the membranes and no increase in the amount of proteins associated with the cell wall. The ybr078wΔ cells with the mutant protein grew at high temperature. The growth was, however, somewhat impaired compared with that of the wild-type. This would be explained by a minor structural change of the protein induced by the exchange of amino acid residues at the C-terminus.

In the previous study, we have determined the amino acid residues required for the efficient incorporation of GPI-associated proteins to the cell wall. Proteins with amino acid residues V or Y at the ω-2 site and V or I at the ω-4 or ω-5 site are preferentially incorporated to the cell wall [11]. In this study, this was the case. Ybr078w has the G residue at the ω-site and SKKSK at the position ω-5 to ω-1 in the ω-minus region. Only a small amount of the Ybr078w protein was released from the cell wall by the treatment with laminarinase or quantazyme, indicating the low level of incorporation into the cell wall. The exchange of SKKSK with SSSSS in the Ybr078w protein resulted in no increase of the protein associated with the cell wall, while the exchange of SKKSK with ISSYS resulted in an increase of the protein associated with the cell wall (Fig. 3C and D). Again, this change in localization of the protein was related to the loss of the cell growth ability at high temperature, suggesting the requirement of localization on the plasma membrane for the Ybr078w function. Kre1, a GPI-attached protein without the specific ω-minus sequence, is associated with the cell wall as well as with the plasma membrane. A recent study has revealed that GPI anchoring of Kre1 on the plasma membrane is essential for the binding of the yeast K1 viral toxin [25]. It is possible that other GPI-PMPs also require GPI-mediated anchoring on the plasma membrane for their biological functions.


We thank Noriko Kawamura and Ikuko Matsuo for DNA sequencing.


  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
  23. [23].
  24. [24].
  25. [25].
View Abstract