Characterization of end4+, a gene required for endocytosis in Schizosaccharomyces pombe



To understand endocytic trafficking in Schizosaccharomyces pombe, we constructed an end4 disruption mutant. The end4+ gene encodes a protein homologous to Sla2p/End4p, which is essential for the assembly and function of the cytoskeleton and endocytosis in Saccharomyces cerevisiae. We characterized the fission yeast mutant end4Δ as well as ypt7Δ, which is deficient in vacuolar fusion and, hence, endocytosis. The delivery of FM4-64 to the vacuolar membrane, accumulation of Lucifer yellow CH and internalization of plasma membrane protein Map3–GFP were inhibited in the end4 mutant. Deletion of end4 resulted in pleiotropic phenotypes consistent with F-actin depolarization, including high temperature sensitivity, abnormal morphology and mating defects. Extensive missorting of carboxypeptidase Y was detected in the ypt7 mutant; however, little missorting was detected in the end4 mutant. These results indicate that End4p is essential for the internalization process and Ypt7p affects endocytosis at a post-internalization step after the intersection of the endocytic and the vacuolar protein-sorting pathways in fission yeast. Copyright © 2004 John Wiley & Sons, Ltd.


Endocytosis is the process of membrane trafficking from the plasma membrane through membranous compartments to lysosomes. Roles for endocytosis in cell physiology include the uptake of nutrients, the removal and degradation of plasma membrane proteins, the regulation of signalling pathways, the delivery of proteins involved in these processes and the recycling of other proteins to the Golgi for re-use. In the first step of endocytosis, portions of the plasma membrane are invaginated, and plasma membrane proteins, lipids and extracellular components, including fluid, low-molecular-weight compounds, proteins and particles, are taken up.

In Saccharomyces cerevisiae, the endocytic pathway was first shown using Lucifer yellow CH (LY), a low-molecular-weight fluorescent compound (Riezman, 1985). As a result of fluid-phase endocytosis, the dye accumulates in the vacuole. This accumulation depends entirely on endocytic membrane trafficking, since LY is highly soluble and membrane-impermeable (Riezman, 1985). A large number of plasma membrane proteins have been shown to be transported via the endocytic pathway to be degraded in the vacuole. When nutrient conditions change, permeases, which are no longer needed, are endocytosed and degraded (Krampe et al., 1998; Springael and André, 1998; Galan et al., 1996; Beck et al., 1999; Horak and Wolf, 1997; Gitan and Eide, 2000). Endocytosis can also mediate the downregulation of pheromone receptors in response to ligand binding (Dunn and Hicke, 2001; Roth and Davis, 1996).

Several screens have been performed in budding yeast to identify endocytosis-defective mutants known as end, dim, sop and svl mutants (Raths et al., 1993; Munn and Riezman, 1994; Wendland et al., 1996; Luo and Chang, 1997; Zheng et al., 1998, Wiederkeh et al., 2001). Some of the END gene products are cytoskeleton-associated proteins, while others are regulators or components of the actin cytoskeleton (Munn, 2001; Wendland et al., 1998). end4 was first identified in a screen for mutants unable to internalize α-factor and LY (Raths et al., 1993); End4p/Sla2p is essential for endocytosis, the assembly and function of the cytoskeleton (Holtzman et al., 1993; Raths et al., 1993; Na et al., 1995), and vesicle transport from the Golgi (Mulholland et al., 1997; Blader et al., 1999). end6 is allelic to RVS161, which encodes one of the homologues of human amphiphysin in budding yeast, and interacts and cooperates with RVS167, another homologue of amphiphysin (Munn et al., 1995; Navarro et al., 1997; Lombardi and Riezman, 2001). In addition, specific sterols and sphingolipids are also required for endocytic internalization. end8 is identical to lcb1, which encodes a subunit of serine palmitoyl transferase, involved in the first step in the synthesis of sphingolipid (Zanolari et al., 2000). end11 is identical to erg2, which encodes C-8 sterol isomerase, involved in the synthesis of ergosterol (Munn et al., 1999).

The fission yeast Schizosaccharomyces pombe, taxonomically and evolutionarily distant from the budding yeast (Russell and Nurse, 1986), is genetically and physiologically well characterized. Several factors have been reported to be involved in endocytosis in Sz. pombe, such as a small GTPase Ypt7p (Bone et al., 1998), Nrf1p, which was identified as a negative regulator of the Cdc42p GTPase, and a putative guanine-nucleotide exchange factor Scd1p (Murray and Johnson, 2001). However, little is known about the endocytic pathway in Sz. pombe. Comparison of the S. cerevisiae SLA2/END4 with the fission yeast genome revealed a homologous gene (SPAC688.11), which we designated end4+. To elucidate the components of the endocytic machinery and understand the trafficking from the plasma membrane in fission yeast, we have constructed a disruption mutant end4Δ, and characterized it based on the phenotypes of ypt7Δ cells. Here we provide several lines of evidence supporting the hypothesis that End4p functions in the internalization process of endocytosis, similar to the case in budding yeast. The data lead us to conclude that End4p is essential for fluid-phase endocytosis, vacuolar delivery of FM4-64 and internalization of pheromone receptor Map3p. Furthermore, we also show the participation of Ypt7p in vacuolar protein sorting, of End4p in mating, and of Ypt7p in sporulation. This is the first report of a fission yeast mutant defective at the internalization step of the endocytic pathway.

Materials and methods

Strains, media, and materials

E. coli strain XL1-blue (Stratagene) was used for all cloning procedures. Sz. pombe wild-type strains used in this study were TP4-1D (h+leu1 ura4-D18 ade6-M216 his2), ARC039 (hleu1-32 ura4-C190T) and KJ100-7B (h90leu1-32 ura4-D18). The vps34Δ, cpy1Δ and ypt7Δ mutants were constructed as described previously (Takegawa et al., 1995; Tabuchi et al., 1997; Iwaki et al., 2003). Standard rich medium (YES), synthetic minimal medium (MM) and sporulation medium (ME) for Sz. pombe cells were used as described (Moreno et al., 1991). Nitrogen-free derivative MM-N medium was modified slightly so that it contained only 1% glucose (Isshiki et al., 1992). The S. cerevisiae wild-type SEY6210 (MATα leu2-3,112 ura3-52 his3200 trp1901 lys2-801 suc29) and sla2/end4 strains (Wendland et al., 1996) were obtained from Scott Emr (University of California, San Diego). DNA restriction and modifying enzymes were from either Takara Shuzo (Kyoto, Japan) or New England BioLabs. Expres35S Protein Labelling Mix (NEG-072) for protein labelling was purchased from NEN Life Science Products. FM4-64 and Alexa 488-conjugated phalloidin were from Molecular Probes Inc. All other chemicals were from Sigma Chemical or Wako Pure Chemicals Co. (Osaka, Japan). The plasmid pMS20 (pAL Map3–GFP; map3+ is under the control of its own promoter) was obtained from Dr C. Shimoda (Osaka City University, Osaka, Japan).

Gene disruption

A search of the databases using the amino acid sequence of S. cerevisiae Sla2p/End4p as a query sequence, revealed SPAC688.11 (named end4+) to be highly homologous to SLA2/END4. The end4+ ORF was amplified from chomosomal DNA of Sz. pombe by PCR and the following primers were synthesized: sense, 5′-CAATACTACGCTGTGATACTTTACGTCGCG-3′, and antisense, 5′-CGCAATGTATTTATCATCCCGTATCGACC-3′. A fragment of 4.0 kb was recovered, and cloned into the vector pGEM-T EASY (Promega). A HindIII–HindIII fragment within the cloned end4+ ORF was excised and a 1.6 kb ura4+ cassette was inserted (Grimm et al., 1988). A linearized DNA fragment carrying this disrupted end4+ was used for the transformation of wild-type strains. ura+ transformants were selected and correct integration of the disruption constructs was verified by PCR and enzymatic digestion of PCR products.

Vacuole staining and fluorescence microscopy

To visualize the fission yeast vacuole, the cells were labelled with the lipophilic dye FM4-64 according to the method described by Vida and Emr (1995). Briefly, 1 ml exponentially growing cells in YES medium was harvested by centrifugation and suspended in 0.5 ml YES medium containing 16 nM FM4-64 to be incubated at 27 °C for 30 min with shaking for pulse-labelling. The labelled cells were then washed once with fresh medium and resuspended in 1 ml YES medium without dye to be incubated at 27 °C for 90 min and examined by fluorescence microscopy.

Fluid-phase endocytosis was microscopically observed after cells were treated with LY (Sigma). The procedure for staining with LY was described by Murray and Johnson (2001).

Stained cells were observed under a fluorescence microscope (Model BX-60; Olympus, Tokyo, Japan) using a U-MGFPHQ filter set (Olympus) for LY, GFP and Alexa 488-conjugated phalloidin and a U-MWIG filter set (Olympus) for FM4-64. Images were captured with a Sensys Cooled CCD camera using MetaMorph (Roper Scientific, San Diego, CA) and were saved as Adobe Photoshop files on a Macintosh G4 computer.

Actin visualization

A portion (1.5 ml) of the exponentially growing culture was fixed by the addition of formaldehyde to 5% for 30 min. The cells were washed twice in PBS (150 mM NaCl, 40 mM K2HPO4 and 10 mM KH2PO4) and resuspended in 200 µl PBS containing 0.1% (w/v) saponin. They were then incubated for at least 2 h at room temperature with 0.5 U Alexa 488-conjugated phalloidin. The cells were washed thee times with PBS and observed under a fluorescent microscope.

Pulse-chase analysis and immunoblot analysis of the Sz. pombe CPY

For analyses of CPY processing, cells were pulse-labelled with Expres35S Protein Labelling Mix (NEN) for 15 min at 30 °C, and chased at the same temperature for given periods. Immunoprecipitation of CPY was performed using rabbit polyclonal antibody against Sz. pombe Cpy1p, as described previously (Tabuchi et al., 1997). Immunoblot analysis of CPY was performed by replica-plating freshly grown spots onto a nitrocellulose filter for 2-day growth, as previously described (Cheng et al., 2002).

Assay for mating efficiency

For the assay of mating efficiency, Sz. pombe cells were grown on YES medium, and then transferred to MEA plates. They were then incubated for 3 days, and the cultures were examined under a microscope, in order to count the number of zygotes and spores. The mating efficiency (ME) was calculated using the following formula [ME = (2Z + 2A + 0.5S), divided by (H + 2Z + 2A + 0.5S), where H is the number of haploid cells, Z is the number of zygotes, A is the number of asci, and S is the number of free spores, n > 300] (Nakamichi et al., 2002).


Disruption of the fission yeast end4+ causes pleiotropic phenotypes

Many yeast endocytosis-defective mutants share similar phenotypes including temperature-sensitive growth, actin delocalization and random budding in diploid mutants, and deletion of several of these genes results in lethal phenotypes (Wendland et al., 1998). The budding yeast SLA2/END4 was originally characterized as end4, a mutation which causes temperature-sensitive growth defects related to a defect in α-factor and LY internalization (Raths et al., 1993) and the null mutant is viable (Holtzman et al., 1993). Sla2p/End4p is assumed to be an adaptor that links actin to clathin and endocytosis, and regulates both endocytosis and actin cytoskeletal dynamics (Baggett et al., 2003).

Figure 1A shows a comparison of the Sz. pombe End4p and the S. cerevisiae Sla2p/End4p. In the COOH-terminal region, End4p and Sla2p/End4p contain an I/LWEQ module that may bind to F-actin in vitro (McCann and Craig, 1997, 1999). The conserved C-terminal regions were also homologous to mouse and nematode talin. This talin homology domain is required for endocytosis at elevated temperatures in vivo (Baggett et al., 2003). End4p and Sla2p/End4p also share the epsin NH2-terminal homology (ENTH) domain in the NH2-terminus (Kay et al., 1999) and the coiled-coil region in the middle of the protein. The ENTH domain is a phospholipid-binding module (Itoh et al., 2001; De Camilli et al., 2002), and possibly interacts with clathin (Wendland et al., 1999). In S. cerevisiae, the central coiled-coil domain mediates interaction with Sla1p (Gourlay et al., 2003), another component of the endocytic machinery, with Sla2p/End4p (Yang et al., 1999), and with clathin light chain (Henry et al., 2002). End4p shows 40.7% overall identity with Sla2p/End4p of budding yeast. The high degree of homology and conserved domains suggest that the function of Sla2/End4p is conserved between S. cerevisiae and Sz. pombe.

Figure 1.

(A) Comparison of fission yeast End4p and budding yeast Sla2p/End4p. The ENTH domain, central coiled-coil region and C-terminal domain containing I/LWEQ motifs are indicated. (B) Schematic representation of the end4+ gene disruption. The large open arrow shows the end4+ ORF. B, BglII; H, HindIII; X, XhoI. (C) Phenotypes of end4Δ and ypt7Δ cells. Wild-type (WT), end4Δ and ypt7Δ cells were grown at 30 °C for 3 days on YES plates containing 0.3 M CaCl2 or supplemented with 50 mM MOPS (pH 7.0) and adjusted to pH 7.0 with NaOH

To examine the role of End4p in fission yeast in vivo, an end4 disruption mutant (end4Δ) was constructed (Figure 1B). The phenotype of end4Δ cells was compared to that of ypt7Δ cells, which was reported to be defective in endocytosis and vacuolar fusion (Bone et al., 1998; Murray and Johnson, 2001). Colonies of wild-type (ARC039), ypt7Δ and end4Δ cells were streaked onto YES plates and incubated at 30 °C and 37 °C for 3 days. Deletion of either end4 or ypt7 makes cells temperature-sensitive for growth (Figure 1C). end4Δ and ypt7Δ cells also showed Ca2 + sensitivity, but sensitivity to neutral pH was only seen in end4Δ cells. These results are similar to what is observed in S. cerevisiae sla2/end4 strains; mutation or deletion of SLA2/END4 renders cells unable to grow at 34 °C, which may result from an aberrant actin cytoskeleton (Holtzman et al., 1993).

The end4Δ mutant has a unique cell shape and depolarized actin cytoskeleton

As determined by Nomarski optics (Figures 2–6), the shape of end4Δ cells was apparently abnormal, similar to the case in budding yeast (Holtzmann et al., 1993). This observation demonstrates that loss of End4p function in the fission yeast had a significant effect on cell morphology, suggesting a correlation between fission yeast End4p and the actin cytoskeleton.

Figure 2.

Mislocalization and disorganization of actin patches in end4Δ cells. Wild-type, end4Δ and ypt7Δ cells were grown in YES medium at 27 °C and then at 37 °C for 3 h. Cells were fixed and stained with Alexa 488-phalloidin

Figure 3.

Time course of FM4-64 internalization via the endocytic pathway in end4Δ cells. Living wild-type cells (A) and end4Δ cells (B) were labelled with 32 nM FM4-64 on ice for 30 min. Cells were washed with ice-cold medium, resuspended in fresh YES medium, and then incubated at 30 °C for the periods indicated

Figure 4.

Vacuole responses of Sz. pombe. (A) Vacuoles of wild-type, ypt7Δ or end4Δ cells were stained with FM4-64 as described in Materials and methods. (B) Cells were then transferred to water for 2.5 h, and observed under a fluorescence microscope

Figure 5.

(A) Accumulation of LY is reduced in the end4 disruption mutant. Wild-type (WT), end4Δ and ypt7Δ cells were incubated for 60 min at 28 °C in YES medium containing LY (5 mg/ml). The washed cells were viewed with a fluorescence microscope (right) and Nomarski optics (left). (B) S. cerevisiae end4 mutant is sensitive to AZC. Wild-type SEY6210 and end4 mutant cells were streaked on SC medium supplemented with 0.1 mg/ml AZC, then incubated at 30 °C for 3 days. (C) Sz. pombe end4Δ and ypt7Δ cells are sensitive to AZC. Wild-type, end4Δ, ypt7Δ and ppr1Δ cells were streaked on SC medium containing 0.3 mg/ml AZC, and then incubated at 30 °C for 3 days

Figure 6.

Translocation of Map3–GFP is slowed in end4Δ cells. Exponentially growing cells harboring pMS20 in MM-Leu medium were shifted to the nitrogen-free liquid medium MM-N. Cells were withdrawn from the culture at the time points indicated after the shift and observed under a fluorescence microscope. (A) Wild-type, (B) ypt7Δ and (C) end4Δ cells

Upon a shift to 37 °C, the actin cytoskeleton of wild-type cells shows two identifiable structures (Figure 2). Actin cables run in the longitudinal direction of the cell from one end to the other, while cortical patches are highly polarized, being concentrated at both growing ends. The ypt7Δ cells displayed comparatively normal actin structures under these conditions. In end4Δ cells, a severe defect was observed. Actin cables were present in these cells; however, they appeared to be randomly oriented (Figure 2). The cortical actin structures were delocalized. Even at the permissive temperature, depolarized actin patches were observed in end4Δ cells (data not shown).

The end4Δ mutant shows a delay of FM4-64 delivery to the vacuoles

FM4-64 is a lipophilic styryl dye and used as a marker for the endocytic pathway and vacuoles in budding yeast (Vida and Emr, 1995). To examine the kinetics of FM4-64 transport to the vacuoles, exponentially growing end4Δ cells were labelled with FM4-64 on ice and chased for given periods at 30 °C. After a 30 min chase, FM4-64 signals were not found in plasma membranes in wild-type cells, as shown in Figure 3A. In end4Δ cells, FM4-64 stained primarily the cell perimeter, consistent with a possible insertion into the plasma membrane after the chase for 30 min (Figure 3B). After a 60 min chase, the plasma membrane signal of FM4-64 decreased, while vacuolar membrane staining relatively increased. From this kinetics analysis, it was suggested that End4p is involved in membrane internalization, although FM4-64 could stain the vacuolar membrane in end4Δ cells after a long period of incubation.

The Sz. pombe End4p is essential for fluid-phase endocytosis but not for vacuolar fusion

Isolated vacuoles from S. cerevisiae can undergo fusion in vitro (Conradt et al., 1992; Mayer et al., 1996). Ypt7p is required for this process (Haas et al., 1995). In normal growth media, Sz. pombe has numerous small vacuoles (Figure 4A). When the cells were transferred to water, a smaller number of much larger vacuoles was apparent as a result of vacuolar fusion (Figure 4B) (Bone et al., 1998). Consistent with a previous report (Bone et al., 1998), ypt7Δ cells had smaller vacuoles than wild-type cells and showed no obvious change in vacuolar morphology when shifted to water. Vacuoles of end4Δ cells showed normal fusion in response to hypotonic stress, indicating that End4p is not required for vacuole fusion in vivo.

In Sz. pombe, ypt7Δ cells show a defect in fluid-phase endocytosis (Murray and Johnson, 2001). To determine whether end4Δ cells also have a defect in fluid-phase endocytosis, the accumulation of LY was observed. Wild-type, end4Δ and ypt7Δ cells were incubated in LY for 1 h at 28 °C. Aliquots were washed extensively to remove excess LY and the cells were viewed under Nomarski and fluorescence optics (Figure 5A). LY accumulated in the vacuoles of the wild-type strain as expected. On the other hand, end4Δ and ypt7Δ cells showed almost no accumulation of LY. These results show that End4p is essential for fluid-phase endocytosis in fission yeast.

The endocytosis-defective mutants are sensitive to the toxic compound AZC

A toxic four-membered ring analogue of L-proline, L-azetidine-2-carboxylic acid (AZC), is incorporated into proteins in competition with L-proline and causes the synthesis of misfolded proteins, thereby inhibiting cell growth. AZC enters budding yeast cells primarily though the function of the general amino acid permease, Gap1p (Hoshikawa et al., 2003). We postulated that amino acid transporters would be stable and active on the plasma membrane in endocytosis-defective mutants, leading to AZC sensitivity. Consistent with this hypothesis, AZC stopped the proliferation of the budding yeast end4/sla2 mutant (Figure 5B).

The amino acid permease responsible for L-proline uptake is unidentified in fission yeast. However, Sz. pombe has one copy of the N-acetyltransferase gene, ppr1+, which detoxifies AZC, and gene disruption of ppr1 makes fission yeast cells sensitive to AZC (Nomura et al., 2003). Sz. pombe wild-type, end4Δ, ypt7Δ and ppr1Δ cells were investigated on SC plates containing 0.3 mg/ml AZC (Figure 5C). Wild-type cells showed AZC-resistant growth but end4Δ, ypt7Δ, and ppr1Δ cells could not grow on AZC-containing plates. These results suggest that the sensitivity to AZC of yeasts is a potential indicator of endocytic defects.

Internalization of Map3–GFP is inhibited in end4Δ cells

To analyse the internalization of plasma membrane proteins, we used Map3–GFP as a monitoring probe. Map3p is the mating pheromone M-factor receptor expressed in h+ strains. Expression of map3+ has been shown to increase under conditions of nitrogen starvation (Tanaka et al., 1993), and cell surface Map3–GFP was internalized by endocytosis after the shift to the nitrogen-free medium (Hirota et al., 2001). In the wild-type h+ cells, localization of Map3–GFP on the cell surface was apparent 1 h after the shift to a nitrogen-free medium (Figure 6A). Then, fluorescence accumulated in the cytoplasm and virtually no fluorescence was visible on the cell surface 6 h after the shift. In ypt7Δ cells, Map3–GFP also accumulated in the cytoplasm 6 h after the shift to the nitrogen-free medium (Figure 6B). In contrast, some Map3–GFP, produced in end4Δ cells, remained on the cell surface even 6 h after the shift (Figure 6C). Thus, we conclude that the end4Δ mutant is impaired in its ability to internalize plasma membrane proteins.

Sz. pombe CPY is missorted in ypt7Δ cells but largely not missorted in end4Δ cells

In S. cerevisiae, several screens were designed to identify internalization mutants or to select for mutants that affect this first step of endocytosis (Raths et al., 1993; Munn and Riezman, 1994; Wendland et al., 1996). Independently, screens for vacuolar protein sorting (vps) mutants have isolated strains that affect endocytosis at a post-internalization step (Bryant and Stevens, 1998; Raymond et al., 1992). The biosynthetic pathway to the vacuole and the endocytic pathway intersect at the prevacuolar compartment (PVC). Mutants that have defects in membrane trafficking from the PVC to the vacuole therefore also have an impaired endocytic membrane transport (Piper et al., 1995; Munn and Riezman, 1994; Wichmann et al., 1992; Schimmöller and Riezman, 1993). One of the characteristics of vps mutants is that a fraction of the newly synthesized vacuolar protease carboxypeptidase Y (CPY) is missorted and as a consequence secreted from the yeast cells.

We have previously reported the missorting of Sz. pombe CPY (SpCPY) in fission yeast vps mutants, vps34Δ and vps33Δ (Tabuchi et al., 1997; Iwaki et al., 2003). The secretion of SpCPY from ypt7Δ and end4Δ cells was tested using a colony-blot assay (Figure 7A). The ypt7Δ mutant cells secreted SpCPY; on the other hand, the end4Δ cells did not. To examine the processing of newly synthesized SpCPY in the ypt7Δ and end4Δ mutants, we performed a pulse-chase analysis. During the synthesis, SpCPY undergoes characteristic modifications and changes in its apparent molecular mass; after a 30 min chase, all SpCPY had been transported to the vacuole and matured (Figure 7B). The ypt7Δ mutant showed a severe sorting defect for SpCPY. After the 30 min chase, only a small amount of mCPY was detected. In contrast, the rate of appearance of proCPY and mCPY in the end4Δ mutant was similar to that in the wild-type cells. Taken together, these results indicate that Ypt7p is implicated in the trafficking from PVC to the vacuole and End4p is not required for vacuolar protein transport.

Figure 7.

Disruption of ypt7 results in mislocalization of Sz. pombe CPY protein. (A) Filter immunoblot for the detection of secreted SpCPY. The indicated strains were grown on MM plates in contact with the nitrocellulose filter at 30 °C for 2 days. The filter was processed for immunoblotting using a rabbit polyclonal antibody against Cpy1p. cpy1Δ was used as a negative control and vps34Δ was used as a positive control for Cpy1p missorting. (B) Processing of SpCPY in vivo. Wild-type (WT), end4Δ and ypt7Δ cells were pulse-labelled with Express-35S-label for 15 min at 30 °C and chased for 30 min. The immunoprecipitates were separated on an SDS-10% polyacrylamide gel. The autoradiograms of the fixed dried gels are shown. The positions of proCPY (110 kDa) and mature CPY (mCPY: 32 kDa) are indicated

Endocytosis-defective mutants show severe defects in mating and sporulation

In budding yeast, some of the actin cytoskeletal mutants including the sla2/end4 mutant share common phenotypes: defects in sporulation, endocytosis, and growth on high NaCl-containing media (Whitacre et al., 2001). As described above, the fission yeast end4Δ mutant showed defects in endocytosis and actin polarization, but was tolerant of 0.3 M NaCl (data not shown). To elucidate the influence of gene disruption on sporulation, end4Δ and ypt7Δ deletion strains were derived from a homothallic strain, which was then streaked onto MEA medium and observed microscopically. After incubation for 3 days, the mating efficiency (ME) of wild-type cells was 71.9%, but that of endocytic mutants was 25.4% (end4Δ) and 48.8% (ypt7Δ). Figure 8 shows the morphology of spores under these conditions. The asci of end4Δ cells resemble those of wild-type cells, but the asci of ypt7Δ cells appear aberrant. The ypt7Δ cells could not differentiate into mature spores, since free spores were not observed after an additional day of incubation. When these strains were incubated in MEL medium for 24 h, wild-type cells formed large aggregates, including diploids, spore-containing zygotes and haploids. Although ypt7Δ cells aggregated, end4Δ cells did not. This observation indicates that End4p is in part responsible for mating, and that the membrane fusion process requiring the Ypt7p function is crucial for spore formation.

Figure 8.

Mating and sporulation of the homothallic strains. Wild-type, end4Δ and ypt7Δ cells were cultured on ME plates at 28 °C for 4 days, and then observed using Nomarski optics


In this study we have demonstrated that End4p is indispensable for endocytosis in fission yeast though the construction and functional analysis of an end4Δ mutant. End4p is 1092 amino acids in length and shows structural similarity with Sla2p/End4p of budding yeast. A high degree of sequence similarity between the proteins occurs in the amino-terminal region (47%), which is known as the ENTH domain (Kay et al., 1999) and the carboxyl-terminal region (32%), which contains the I/LWEQ module (McCann and Craig, 1997, 1999), showing considerable homology to talin, an actin filament-binding protein that is found at focal cell adhesions (Moulder et al., 1996).

The budding yeast SLA2/END4 was originally characterized as end4 (Raths et al., 1993) and two other alleles of END4, SLA2 and MOP2, identified on the basis of the mutant phenotype (Holtzmann et al., 1993; Na et al., 1995). The fission yeast end4Δ cells showed similar phenotypes to the budding yeast sla2/end4 mutants, i.e. temperature-sensitive growth (Figure 1), abnormal morphology (Figures 2–6), actin depolarization (Figure 2) and a defect in endocytosis (Figures 3, 5A, 6), indicating functional similarity for actin organization and endocytosis of both proteins.

The fission yeast ypt7+ has been known to be involved in endocytosis and vacuole fusion (Murray and Johnson, 2001; Bone et al., 1998), suggesting that vacuole fusion in Sz. pombe is likely to depend on cellular machinery similar to that in the budding yeast and Ypt7ps has the same functions in Sz. pombe and S. cerevisiae. Consistent with this explanation, fission yeast ypt7Δ cells showed similar phenotypes to the S. cerevisiae ypt7 mutant, including highly fragmented vacuoles (Figure 4), defects of vacuolar protein transport and maturation (Figure 7) (Wada et al., 1992), and a defect of endocytosis (Figure 5A) (Wichmann et al., 1992; Schimmöller and Riezman, 1993).

We have shown that both the Sz. pombe end4Δ and ypt7Δ mutants are defective in fluid-phase endocytosis. However, observations of the clearance of marker molecules on the cell surface, FM4-64 and Map3–GFP, and sorting defects of SpCPY revealed that End4p is responsible for the internalization process and Ypt7p is involved at a later stage, through endosomes and at the vacuoles (Figures 3, 6, 7).

A phenotype common to fission yeast end4Δ and ypt7Δ cells is AZC sensitivity (Figure 5C), suggesting that the stability and activity of amino acid transporters could not be regulated by endocytosis. The cellular response to AZC would be of potential use for screening of an endocytosis-defective mutant. Unlike S. cerevisiae, amino acid permeases responsible for proline uptake are unidentified in fission yeast, although four genes encoding a protein similar to budding yeast Gap1p are present in the fission yeast genome (unpublished results). If the proline permease is identified in fission yeast, its contribution to AZC sensitivity will be revealed.

These endocytosis-defective mutants of fission yeast have defects in sexual cellular differentiation, end4Δ in mating and ypt7Δ in sporulation. During sexual differentiation, the polarization of Sz. pombe cells was accompanied by rearrangements of the F-actin cytoskeleton, and cortical actin patches concentrated at the projection tip (Petersen et al., 1998). Some fission yeast mutants carrying cytoskeletal defects show a defect in mating (Lee et al., 2000; Wu et al., 2001). In S. cerevisiae, some of the actin mutants including sla2/end4 have defects in sporulation (Whitacre et al., 2001). Candida albicans could not produce hyphae, when the SLA2 homologue Ca SLA2 was deleted (Asleson et al., 2001). Actin depolarization of Sz. pombe end4Δ cells would be responsible in part for the mating defects. The phenotype of Sz. pombe endocytic mutant ypt7Δ can be accounted for by vacuolar protein sorting defects. The missorting of vacuolar proteases would reduce the amount of proteases in the vacuoles, which are required for the nitrogen starvation-induced protein degradation that accompanies sporulation. An alternative possibility is that the defects in endocytosis affect the ability of cells to mate and sporulate. If the defect in endocytosis leads to a defect in sporulation, then endocytosis has a role in sporulation; however, what this role would be remains unknown.

In budding yeast, SLA2/END4 shows a genetic and physiological interaction with RVS167 and RVS161 (Wesp et al., 1997; Drees et al., 2001). The S. cerevisiae genes RVS167 and RVS161 were identified in a screen for mutants that had reduced viability when starved of glucose, nitrogen or sulphur (Crouzet et al., 1991). rvs161 and rvs167 mutants show a reduced rate of fluorescent dye uptake, a disorganized actin cytoskeleton and improper budding patterns (Munn et al., 1995; Brizzio et al., 1998; Colwill et al., 1999; Bauer et al., 1993; Sivadon et al., 1995). In fission yeast, it was reported that a mutant of the RVS161 homologue hob3+ and the RVS167 homologue hob1+ were dispensable for endocytosis, despite the mislocalization of actin in the hob3 mutant (Routhier et al., 2001, 2003). We found that vesicular trafficking including fluid-phase endocytosis, uptake of FM4-64 and vacuolar protein transport is normal in the double disruption mutant hob3Δ hob1Δ (data not shown). Our observations indicate that Hob1p and Hob3p may not be components of the endocytic machinery. In S. cerevisiae, Rvs167p physically interacts with Abp1p, actin-binding protein, and consists of a complex including Abp1p, Sla2p, Srv2p and Rvs161p (Wendland et al., 1998). Yet, the roles of the Abp1p and Srv2p homologues for endocytosis are still unknown in fission yeast. Although other complexes participating in the internalization step, including the endocytic regulatory proteins, are found in budding yeast, the components may be different in fission yeast. Now that almost all fission yeast genes have been sequenced, the function of other factors that are potentially involved in the endocytic pathway will be investigated, and the similarities and differences between budding yeast and fission yeast endocytosis will be further explored.


We are grateful to Dr Chikashi Shimoda for providing the plasmids and Taro Nakamura for stimulating discussions. We thank Scott Emr for providing the S. cerevisiae strains. This work was partly supported by the Project for Development of a Technological Infrastructure for Industrial Bioprocesses on R&D of New Industrial Science and Technology Frontiers, by the Ministry of Economy, Trade and Industry (METI), and entrusted by a New Energy and Industrial Technology Development Organization (NEDO) fellowship.