OUP user menu

Dynamics of cell wall structure in Saccharomyces cerevisiae

Frans M. Klis, Pieternella Mol, Klaas Hellingwerf, Stanley Brul
DOI: http://dx.doi.org/10.1111/j.1574-6976.2002.tb00613.x 239-256 First published online: 1 August 2002


The cell wall of Saccharomyces cerevisiae is an elastic structure that provides osmotic and physical protection and determines the shape of the cell. The inner layer of the wall is largely responsible for the mechanical strength of the wall and also provides the attachment sites for the proteins that form the outer layer of the wall. Here we find among others the sexual agglutinins and the flocculins. The outer protein layer also limits the permeability of the cell wall, thus shielding the plasma membrane from attack by foreign enzymes and membrane-perturbing compounds. The main features of the molecular organization of the yeast cell wall are now known. Importantly, the molecular composition and organization of the cell wall may vary considerably. For example, the incorporation of many cell wall proteins is temporally and spatially controlled and depends strongly on environmental conditions. Similarly, the formation of specific cell wall protein–polysaccharide complexes is strongly affected by external conditions. This points to a tight regulation of cell wall construction. Indeed, all five mitogen-activated protein kinase pathways in bakers’ yeast affect the cell wall, and additional cell wall-related signaling routes have been identified. Finally, some potential targets for new antifungal compounds related to cell wall construction are discussed.

  • Yeast
  • Cell wall
  • Cell wall protein
  • Stress
  • Candida albicans

1 Introduction

This review focuses on recent advances in our understanding of the molecular organization of the cell wall of Saccharomyces cerevisiae and on the various protein–polysaccharide complexes found in its cell wall. We further discuss changes in composition and organization of the wall in response to cell wall stress and to various environmental conditions. Cell wall construction is tightly regulated, but with a few exceptions the underlying control mechanisms are still fragmentarily known. We will explore some important findings in this area. For earlier general reviews we refer to [13]. Other more specialized reviews will be mentioned in the relevant sections.

2 Cell wall architecture

2.1 Composition and properties of the cell wall

The cell wall is a sturdy structure providing physical protection and osmotic support [4]. Electron microscopic analysis of the wall using negative staining reveals a layered structure with an electron-transparent internal layer of about 70–100 nm thick depending on growth conditions and genetic background, and an electron-dense outer layer [5, 6]. In brewing yeast the electron-transparent inner layer may be as thick as 200 nm [7]. The mechanical strength of the wall is mainly due to the inner layer, which consists of β1,3-glucan and chitin, and represents about 50–60% of the wall dry weight (Table 1). The outer layer, which consists of heavily glycosylated mannoproteins emanating from the cell surface [5, 18], is involved among others in cell–cell recognition events [5, 1921]. It also limits the accessibility of the inner part of the wall and the plasma membrane to foreign enzymes such as cell wall-degrading enzymes in plant tissue [2224]. The carbohydrate side chains of the cell surface proteins contain multiple phosphodiester bridges, resulting in numerous negative charges at the cell surface at physiological pH values [25]. These side chains are responsible for the hydrophilic properties of the wall, and may be involved in water retention and drought protection. The outer protein layer accounts for about one third of the wall dry weight. Cell wall proteins are covalently linked to the β1,3-glucan–chitin network either indirectly through a β1,6-glucan moiety or directly (Sections 2.6 and 2.7). In addition, some proteins are disulfide-bonded to other cell wall proteins [26, 27].

View this table:
Table 1

Cell wall macromolecules in S. cerevisiae

MacromoleculeWall dry wt. (%)Site of synthesisMature formRefs.
Mannoproteins35–40Secretory pathway200aHighly branched[13]
β1,6-Glucan5–10(PM)140Highly branched[12, 14, 15]
β1,3-Glucan50–55PM1500Moderately branched[8, 16]
  • The data regarding the cell wall composition have been compiled from [810]. The mannoproteins are often heavily glycosylated; the protein content per se of isolated SDS-extracted walls accounts for only a small percentage of cell wall dry weight. The cell wall components are presented in the order in which they occur in the cell wall from the outside to the inside. The cell wall of C. albicans has a similar cell wall composition [11]. DP, degree of polymerization; PM, plasma membrane. The site of synthesis of β1,6-glucan is uncertain [12]. Note that the data presented here may vary depending on growth conditions.

  • aN-chains.

  • bChitin from bud scars.

The cell wall is highly elastic [28, 29]. When yeast cells are transferred to hypertonic solution, they rapidly shrink, and depending on the osmotic stress they may lose more than 60% of their initial volume. This process is reversible. When transferred back to the original medium, the cells immediately expand to their initial volume. This elasticity of the wall is probably due to the elastic properties of the β1,3-glucan chains (see Section 2.2). The elasticity of the wall explains why the wall of living cells is much more permeable than isolated cell walls. Whereas isolated walls are permeable only to molecules of molecular mass up to 760 Da [30], walls of living cells are permeable to much larger molecules especially under hypotonic conditions and also depending on growth conditions [22, 31, 32].

2.2 The β1,3-glucan network

The mechanical strength of the cell wall is mainly due to β1,3-glucan, which can be specifically stained with aniline blue [33, 34]. β1,3-Glucan chains belong to the so-called hollow helix family; in other words, they have a shape comparable to a flexible wire spring that can exist in various states of extension [35]. This property explains the above-mentioned elasticity of the cell wall. Using magic-angle spinning 13C-NMR on living cells, Krainer and co-workers have found that a portion of the β1,3-glucan indeed assumes a helical structure [36].

β1,3-Glucan is only slightly crystalline in lateral walls [37]. In stationary phase cells, β1,3-glucan molecules were found to consist of about 1500 glucose monomers [8]. This may, however, be an underestimation as a result of partial hydrolysis of glucan chains during the extraction procedure. Considerably higher degrees of polymerization have been reported, depending on the type of acid used for extraction [38]. In addition, the degree of polymerization may depend on environmental conditions, because yeast uses two β1,3-glucan synthase complexes, to an extent that depends on growth phase and carbon source [39]. In their mature form, β1,3-glucan chains of stationary phase cells are moderately branched and contain about 3–4%β1,6-linked glucose residues [16]. Conceivably, the degree of branching of β1,3-glucan may also depend on growth conditions.

The moderate degree of branching in mature β1,3-glucan molecules prevents the extensive crystallization seen on the surface of regenerating spheroplasts [6, 37, 40]. This is also consistent with the X-ray diffraction data of isolated walls, which point to a low level of crystalline β1,3-glucan [37]. On the other hand, the presence of uninterrupted chain segments of substantial length permits stable interchain association, thus allowing the formation of a three-dimensional network [16, 40]. Consistent with this, electron microscopic studies of isolated walls using negative staining reveal a fine network with meshes of about 20–60 nm wide [37]. As isolated walls are in a non-extended state, the pores of the glucan layer in living cells are expected to be wider, explaining why the permeability of the inner wall is only limiting for very large proteins [41].

The two β1,3-glucan synthase complexes found in yeast contain either Fks1 or Gsc2/Fks2 depending on environmental conditions. Both are multiple-spanning transmembrane proteins that are essential for the synthesis of β1,3-glucan [39] and the corresponding genes are well conserved among fungi [42, 43]. It is generally assumed that Fks1 and Gsc2 represent the catalytic subunits, but this has still to be validated experimentally, because, for example, they lack the known UDP-Glc binding site (K/RXGG) found in glycogen synthase [44] and in an α1,3-glucan synthase in Schizosaccharomyces pombe [45]. Alternatively, they may represent a pore-forming protein that guides the newly formed chain through the plasma membrane to the outside. An elegant electron microscopic study, using snap-frozen freeze-etched cells, has identified particles in the plasma membrane, from which, on close examination, fibrils with a diameter of 5 nm can be seen emerging [46]. These fibrils disappear into the inner regions of the cell wall, suggesting that they may represent β1,3-glucan fibrils. A follow-up of this older work in combination with the use of molecular genetic tools promises a more complete understanding of in vivo β1,3-glucan synthesis. Interestingly, a plant homologue of FKS1 has been identified in cotton fibers [47].

It is unknown whether growing β1,3-glucan chains in yeast are extended at the reducing or non-reducing end. However, β1,3-glucan from the fungus Sclerotium rolfsii is extended at the non-reducing end [48]. Also, bacterial cellulose, diatom chitin, and plant homogalacturonan seem to be extended at their non-reducing ends [4951], suggesting that this is generally the case for processive glycosyltransferases such as β1,3-glucan synthase. This would make the reducing end of the growing chain immediately available for coupling to Pir cell wall proteins (see Section 2.5) and for processing enzymes such as Bgl2 [52, 53].

As β1,3-glucan chains are synthesized as linear chains [54], the presence of branches in the mature molecules implies the existence of β1,3-glucan processing enzymes. Additional enzymes may be needed for integrating newly synthesized β1,3-glucan molecules into the growing wall during isotropic growth. Enzymes potentially involved in these steps are Gas1, an endotransglycosylase that may be involved in extending and rearranging β1,3-glucan chains [55, 56], Bgl2, an endotransglycosylase that introduces intrachain β1,6-linkages [52, 57], and, possibly, also the Crh1 [58] and the SUN family [57, 59]. Their exact function in vivo is, however, still incompletely understood.

2.3 Chitin as an intrinsic part of the lateral walls

Chitin is believed to occur as linear chains in the chitin ring, in and around the bud scars and also to a minor extent uniformly dispersed in the lateral walls of the mother cell [6062]. Chitin isolated from bud scars consists of about 190 N-acetylglucosamine monomers [17], but it is unknown whether this is also a valid estimate for chitin in the lateral walls. Calcofluor white is a fluorescent, anionic dye that preferentially binds to β1,4-glucans such as chitin, chitosan, and cellulose. It reacts to a lesser extent with β1,3-glucan [63, 64]. In S. cerevisiae, Calcofluor white is frequently used to visualize chitin.

Chitin synthesis in S. cerevisiae involves three chitin synthases and is tightly regulated [1, 62]. For example, under normal conditions the deposition of chitin in the lateral walls takes place after cytokinesis [61]. This results in a relatively low chitin level in the lateral walls: 0.1–0.2% excluding the bud scar(s) and 1–2% including the bud scars (see also Table 1). In cells with a (genetically) weakened wall, however, chitin synthesis is activated as part of a salvage mechanism and the levels in the lateral walls of those cells may become as high as 20% of the wall dry weight (see also Section 3.4) [10, 66]. As to be expected for a salvage mechanism, chitin is then also deposited in the wall of the growing bud [66]. The responsible synthase is encoded by CHS3 [67]. Chs3 is a multi-spanning membrane protein with its active site at the cytosolic side of the plasma membrane [68]. Like other fungal chitin synthases, it is a processive enzyme with the diagnostic motif D,D,D,QRRRW [69]. Diatom chitin synthase extends the growing chitin chain at the non-reducing end using UDP-GlcNAc as substrate [50]. Assuming that fungal chitin synthases behave similarly, the reducing end of a chitin chain is directly available for coupling to the acceptor sites of β1,3- and β1,6-glucan molecules [70, 71], when it emerges from the pore through the plasma membrane.

2.4 β1,6-Glucan

β1,6-Glucan is in its mature form a highly branched, water-soluble polymer consisting on average of about 130 glucose monomers [14]. It is unknown whether β1,6-glucan is synthesized as a water-insoluble linear molecule similar to β1,3-glucan [72], or as a water-soluble branched molecule similar to the mixed β1,3-/β1,6-glucan produced by S. rolfsii [48]. This is particularly relevant regarding the development of an in vitro assay for β1,6-glucan synthase.

β1,6-Glucan is used in the cell wall to connect GPI-dependent cell wall proteins to the β1,3-glucan network (see Fig. 4, Section 2.7). It may also function as acceptor site for chitin, particularly in case of cell wall stress (see Section 2.6 and Fig. 2) [71, 73]. The pioneering studies by Bussey and co-workers have identified several genes that affect β1,6-glucan levels in the cell wall [74]. From these studies it has emerged that various ER-resident proteins, Golgi-resident proteins, and cell surface proteins strongly affect β1,6-glucan levels in the cell wall. This might be interpreted in various ways: (i) the biogenesis of β1,6-glucan is a stepwise process that begins in the ER; (ii) the synthesis of β1,6-glucan takes place at the plasma membrane but requires the stepwise synthesis of a primer; or (iii) the synthesis of β1,6-glucan takes place at the plasma membrane but the activity of the synthesizing complex is highly sensitive to various defects in the secretory pathway. Using immunogold labeling, Montijn and co-workers could not detect intracellular β1,6-glucan, not even in a temperature-sensitive mutant that had accumulated secretory vesicles at the restrictive temperature for 2 h [12]. In contrast, plasma membrane-derived vesicles and the cell walls reacted strongly. This seems to exclude the first hypothesis. Because an in vitro assay for β1,6-glucan is not yet available, differentiation between the two other hypotheses is not simple, the more so since they are not mutually exclusive. Summarizing, a clear picture of how β1,6-glucan is synthesized is still lacking.

Figure 4

Putative molecular organization of the cell wall of S. cerevisiae. The mechanical strength of the wall is due to an elastic three-dimensional network of β1,3-glucan molecules kept together by hydrogen bonding between locally aligned chains [40]. As the arrows in the model indicate, the terminal non-reducing ends of the β1,3-glucan side chains are believed to function as acceptor sites for β1,6-glucan and chitin, whereas the reducing end of the β1,3-glucan molecules may be involved in the linkage to Pir-CWPs. The β1,3-glucan layer forms the inner layer of the cell wall, whereas cell wall proteins form the outer layer. The two most abundant CWP–polysaccharide complexes are shown here: GPI-CWP–β1,6-glucan–β1,3-glucan and Pir-CWP–β1,3-glucan. The β1,6-glucan molecules are highly branched and thus water-soluble, tethering GPI-CWPs to the β1,3-glucan network. GPI-CWPs represent the major component of the outer protein layer. The Pir-CWPs are directly linked to the β1,3-glucan network through an alkali-sensitive linkage and are distributed throughout the β1,3-glucan network. This model is based on [2, 65, 70, 71, 73, 80, 115, 137]. The model is believed to be valid for C. albicans as well [11]. GPI-CWP, cell wall protein linked through a GPI remnant to β1,6-glucan (β1,6-Glc); Pir-CWP, a cell wall protein from the PIR family linked through an alkali-sensitive linkage to β1,3-glucan.

Figure 2

Overview of the CWP–polysaccharide complexes in yeast. GPI-CWP, cell wall protein linked through a GPI remnant to β1,6-glucan (β1,6-Glc); Pir-CWP, a cell wall protein from the PIR family linked through an alkali-sensitive linkage to β1,3-glucan (β1,3-Glc). Kollar and co-workers have identified several of the interconnecting linkages between cell wall macromolecules [70, 71], but the interpolymer linkages between GPI-CWP and Pir-CWP on the one hand and β1,3-glucan on the other hand are unknown.

2.5 Cell wall proteins (CWPs)

The mannoproteins that form the outer cell wall layer are highly glycosylated with a carbohydrate fraction that often amounts to over 90% (w/w) [1]. Biotinylation of intact cells using a sulfonated derivative of biotin that does not cross the plasma membrane is a convenient tool to distinguish between authentic cell wall proteins and adventitiously bound proteins [75]. The outer layer of mannoproteins is much less permeable to macromolecules than the internal fibrillar layer [24]. This is largely due to the presence of the long and highly branched carbohydrate side chains linked to asparagine residues [1, 22] and to the presence of disulfide bridges [32]. Serine and threonine residues, which may carry short oligomannosyl chains, are often clustered, resulting in relatively rigid rod-like regions of the polypeptide backbone [76, 77]. Finally, due to phosphodiester bridges in both N- and O-linked mannosyl side chains the cell surface of yeast contains numerous negative charges [1, 25]. Interestingly, the sensitivity of yeast cells to the antifungal plant protein osmotin, which has a high isoelectric point and thus is positively charged at physiological pH values, depends on the presence of mannosyl phosphate groups at the cell surface [78]. In addition, the cell wall may bind positively charged, cytosolic proteins originating from lysed cells. Phosphodiester groups can be visualized with Alcian blue [25].

There are two main classes of proteins covalently coupled to cell wall polysaccharides (Fig. 4, see Section 2.7).

(a) GPI-dependent cell wall proteins (GPI-CWPs). They are generally indirectly linked to β1,3-glucan through a connecting β1,6-glucan moiety. In the genome of S. cerevisiae about 60–70 GPI proteins have been identified ([79]; J.G. De Nobel, unpublished data). About 40 of them are destined for the plasma membrane whereas the others become covalently linked to β1,6-glucan [71, 8084]. They often contain repeats and serine- and threonine-rich regions. The most extensively studied GPI-CWP is Sag1, which is involved in sexual agglutination [5, 77, 84, 85]. Mature proteins only have a remnant of the original GPI anchor, that links them to β1,6-glucan [71, 81]. The core structure of this remnant [71] is formed by –ethanolamine–Pi–(Man)4–, and is probably substituted with additional ethanolamine phosphate groups [8688]. Interestingly, β1,6-glucan extracted from cell walls by hot acetic acid may contain a minor amount of galactose [14]. Conceivably, in some genetic backgrounds this is part of the GPI anchor remnant of GPI-CWPs [89]. The presence of galactose in the cell wall of S. cerevisiae is consistent with the evidence for a UDP-galactose transporter in bakers’ yeast [90].

(b) Pir proteins (Pir-CWPs). They are presumably directly linked to β1,3-glucan through an alkali-sensitive linkage. In S. cerevisiae a family of four such proteins has been found [65, 91, 92]. They are all similarly organized (SP–Kex2–repeat(s)–CX(66)CX(16)CX(12)C), consisting of an N-terminal signal peptide, a Kex2 site, followed by a repeat-containing region with up to 11 repeats, and a highly conserved carboxy-terminal region with four cysteine residues in a conserved spacing pattern. Pir1, Pir2/Hsp150, Pir3, and Pir4/Cis3 all have been localized to the cell wall immunologically [27, 93, 94]. Several additional proteins, such as Pau1 and its homologues, and Sps100, which is believed to contribute to maturation of the spore wall [95], and also Ygp1, which is induced by nutrient limitation [96], are predicted to have an N-terminal signal peptide, but not an addition signal for a GPI anchor. Possibly, their final destination is not the medium, but the cell wall, to which they may become linked in a Pir-CWP-like fashion.

In addition, several cell wall proteins such as Bar1, a protease [27], Aga2 [5, 97], the active subunit of the sexual agglutinin complex in MATa cells, Pir4/Cis3 [27], and some known or potential cell wall glycanases such as Sun4/Scw3 [57], can be released from intact cells using a reducing agent. This suggests that they might be disulfide-linked to other cell wall proteins. Reducing agents are also expected to release soluble, intermediate forms of GPI-CWPs [98]. Finally, SDS extraction of isolated walls releases many proteins. With a few exceptions, like the transglucosylase Bgl2 [52] and the chitinase Cts1 [99, 100], they are not authentic cell wall proteins and their presence is due to contamination with membrane fragments [11, 57, 101].

Members of the Hsp (heat-shock protein) family [102] and abundant glycolytic enzymes such as Tdh1, Tdh2, and Tdh3 [103] are often found at the cell surface. They can be extracted from intact cells with a reducing agent such as mercaptoethanol under slightly alkaline conditions, suggesting that they are either trapped inside the wall or are ionically bound to cell surface proteins. It is not clear whether these proteins originate from lysed cells or as frequently claimed are exported by a non-conventional secretory mechanism [102104]. Heat-shock proteins and glycolytic enzymes have also been found in the medium of regenerating spheroplasts, which raises the same issue [104106].

Cell wall proteins may have various functions, which are summarized in Table 2. In many cases their precise function is unknown. Some proteins, like the very small but presumably abundant GPI-CWPs Ccw12 and Ydr134c with a predicted unprocessed size of 133 and 66 amino acids, respectively, and a codon adaptation index of 0.870 and 0.646, respectively, may be used as a means to present mannan to the cell surface (Fig. 1). Various GPI-CWPs are involved in adhesion events like sexual agglutination and flocculation of yeast cells. Others appear to have an enzymatic function like Crh1, Crh2, and Crr1 [58]. Fig. 2 is required for the normal width of the conjugation tube [108], suggesting that it may be involved in remodeling of β-glucan in the conjugation tube.

View this table:
Table 2

Known or putative functions of CWPs

Presentation of mannan chainsCCW12, YDR134CFig. 1
Cell wall permeability[22, 24, 41]
Retention of waterSection 2.1
MatingSAG1, AGA1, AGA2, FIG2[5, 77, 107109]
FlocculationFLO1, FLO5, FLO9, FLO10[20, 109111]
Invasive/pseudohyphal growthFLO11[109]
Biofilm formationFLO11[21]
Cell wall processing
Glucan remodelingCRH1, CRH2, CRR1, FIG2 (speculative), BGL2[52, 58, 106, 108]
Cell separationEGT2, PRY3, CTS1, DSE2, DSE4, SCW11[57, 100, 112]
Cell wall strengtheningCWP1, PIR1, PIR2/HSP150, PIR3, PIR4/CIS3[65, 113116]
Isotropic growthPIR1, PIR2/HSP150, PIR3, PIR4/CIS3Section 3.1; 117
Growth conditions
Stationary phaseSED1, SPI1[119, 120]
Anaerobic growthDAN1, DAN4, TIP1, TIR1, TIR2, TIR3, TIR4[121123]
Low temperatureTIP1, TIR1, TIR2, TIR4[121, 122, 125]
Iron uptakeFIT1, FIT2, FIT3[127]
Figure 1

Schematic representation of the GPI-CWP Ccw12. The mature form of Ccw12 is estimated to consist of less than 100 amino acids. Ccw12 has three predicted attachment sites for N-chains. As N-chains may contain up to 200 mannose residues each [13], they may together have a molecular mass of about 90 kDa. Interestingly, CCW12 has a codon adaptation index (CAI) of 0.87, indicating that it encodes a very abundant protein. Deletion of CCW12 results in a reduced growth rate of about 40% and increased sensitivity to Calcofluor white and Congo red [128]. SP, signal peptide; grey box, GPI anchor addition signal.

The expression of PIR genes is also up-regulated in case of cell wall stress (see Section 3.3) [113, 114], consistent with the idea that their gene products might be involved in cell wall strengthening. Disruption of all four genes results in swollen cells that grow slightly slower and are more sensitive to Calcofluor white and Congo red, indicating that their cell wall is indeed weakened [92]. Interestingly, cells that overexpress PIR2 are more resistant to the plant antifungal protein osmotin whereas cells that lack PIR1, PIR2, and PIR3 are more sensitive to it [93]. Differences in tolerance to osmotin depend on the presence of an intact cell wall, because after spheroplasting the cells are equally sensitive [88]. Possibly, Pir-CWPs make the cell wall less permeable to osmotin and other proteins.

2.6 CWP–polysaccharide complexes

Analysis of CWP–polysaccharide complexes requires a relatively small set of reagents. Recombinant β1,3-glucanase and β1,6-glucanase are now commercially available. For detection, antibodies specific for β1,3-glucan are commercially available, whereas antibodies specific for β1,6-glucan are easy to raise [129]. The lectins concanavalin A for mannoproteins and wheat germ agglutinin for chitin oligomers further complement this list.

Cell wall proteins may be linked to the β1,3-glucan network in various ways (Fig. 2). In cells grown in rich medium the GPI-CWP–β1,6-glucan–β1,3-glucan complex is the most abundant one (Fig. 2A). It consists of cell wall proteins such as Ccw12, Tip1, Ssr1, Cwp1, or Sag1 covalently linked through a GPI remnant to a non-reducing end of β1,6-glucan, which in turn is linked to a non-reducing end of β1,3-glucan [2, 71, 80, 81]. Because the GPI remnant (core structure: –ethanolamine–Pi–(Man)4–) contains a phosphodiester bridge, GPI-CWPs in this complex can be specifically released by using aqueous hydrofluoric acid [80]. Pir-CWP–β1,3-glucan represents another, abundant CWP–polysaccharide complex in which a protein from the PIR family is linked through an alkali-sensitive linkage to β1,3-glucan (Fig. 2D) [65, 75, 92]. The nature of this linkage is not known, but in view of its sensitivity to mild alkali it is tempting to postulate that it involves an O-linked side chain. Interestingly, immunogold labeling of Pir-CWPs shows that the signal is uniformly dispersed over the chitin–β1,3-glucan layer of the cell wall [94]. In contrast, immunogold labeling of the GPI-CWPs Sag1 and Flo1 results in strong labeling of the outer fibrillar wall layer [5, 110]. This indicates that Pir-CWPs are part of the inner layer. A possible interpretation of these observations is presented in Fig. 3, in which it is proposed that a β1,3-glucan molecule is covalently linked through its reducing end to an O-linked side chain of a Pir-CWP, possibly present in the latter's internal repeats.

Figure 3

Hypothetical linkage between a Pir-CWP and a β1,3-glucan polymer. According to this scheme a β1,3-glucan molecule is linked with its reducing end to an oligomannoside, O-linked to a serine or threonine residue of a Pir-CWP. The glycopeptide linkage is alkali-sensitive.

Some GPI-CWPs such as Cwp1 may also be bound directly to β1,3-glucan through an alkali-sensitive bond presumably in a Pir-CWP-like fashion [115]. As a result, two additional GPI-CWP–polysaccharide complexes can be distinguished, which are either single- or double-linked (Fig. 2B,C) [115]. Importantly, Cwp1 and possibly also other potentially double-linked GPI-CWPs play an important role in the response of cells towards cell wall stress [113, 116]. Finally, a fifth complex has been identified in which a GPI-CWP is linked to a β1,6-glucan molecule, which is directly interconnected to chitin (Fig. 2E). Although this type of complex is relatively rare in cells growing on rich medium, under special conditions it can become much more prominent (see Section 3.4 and Table 3) [73]. Interestingly, chitin in the lateral walls, which represents ∼10% of the total chitin, seems to be specifically associated with β1,6-glucan [60]. This may mean that chitin in the lateral walls is predominantly present in the form of the complex GPI-CWP–β1,6-glucan–chitin. Finally, chitin is also directly linked to β1,3-glucan [70]. Chitin is, however, omitted from the first four complexes shown in Fig. 2, because (i) such complexes may exist without a chitin molecule attached to it, for example, in the growing bud, and (ii) a chitin molecule may become linked to another β1,3-glucan molecule than the one shown in the various cell wall protein–polysaccharide complexes, i.e. the linkage between β1,6-glucan and chitin may be quite indirect.

View this table:
Table 3

Changes in cell wall composition and organization in response to genetically caused cell wall defects

Less β1,3-glucanMore chitin in the lateral walls[10, 73, 116, 130, 131]
More GPI-CWP-β1,6-glucan-chitin
Less β1,6-glucanMore GPI-CWPs in the medium[65, 84]
More Pir-CWPs in the wall
More β1,3-glucan and chitin in the wall
Possibly, increased usage of the alkali-sensitive linkage between GPI-CWPs and β1,3-glucan
Defective N- and O-glycosylationMore chitin in the lateral walls[2, 132135]

As the synthesis of chitin and β1,3-glucan takes place at the plasma membrane [54, 68] and the coupling of GPI-CWPs to β1,6-glucan also occurs outside the plasma membrane [12, 84], the interconnections between the cell wall macromolecules also have to be made outside the plasma membrane. This strongly points to the existence of various classes of cell wall assembly enzymes – presumably transglycosylases – responsible for interconnecting (a) GPI-CWPs to β1,6-glucan, (b) β1,6-glucan to β1,3-glucan, (c) chitin to β1,3-glucan, (d) chitin to β1,6-glucan, and (e) cell wall proteins directly to β1,3-glucan through an alkali-sensitive linkage. Interestingly, cells growing in the presence of gentiobiose, which consists of two β1,6-linked glucose monomers, incorporate a GPI-CWP reporter protein less efficiently into the cell wall than wild-type cells, and this is accompanied by increased sensitivity to Zymolyase [136]. This suggests that gentiobiose inhibits a cell wall assembly step involving β1,6-glucan. At present, none of the postulated assembly enzymes has been identified. Also, the sequence of coupling events is largely unknown (see Section 2.7).

2.7 Molecular organization of the cell wall

The cell wall forms a bilayered, supramolecular structure that surrounds the entire cell (Fig. 4). Its mechanical strength is largely based on an internal layer of moderately branched β1,3-glucan molecules that form a three-dimensional network that is kept together by hydrogen bonding between laterally associated chains [40]. Methylation analysis has revealed that each β1,3-glucan polymer has multiple side chains [16]. Their terminal non-reducing ends are believed to function as acceptor sites for β1,6-glucan and chitin [70, 71], whereas the reducing end of the β1,3-glucan molecules may also be involved in the linkage to Pir-CWPs (see Section 2.6) [115]. Methylation analysis of mature β1,6-glucan has shown that it is a highly branched molecule [14], which explains why it is water-soluble [15, 40]. Chitin is deposited in the lateral walls after cytokinesis has taken place, probably stiffening the cell wall [61]. It represents less than 10% of the total chitin content of the walls and may be coupled to either β1,3-glucan or β1,6-glucan [70, 71].

The biosynthetic pathway of the sexual agglutinin Sag1 has been studied in detail and several intermediate glycoforms have been identified including two extracellular glycoforms, namely, a plasma membrane-bound form and a later, soluble form [84, 98]. In contrast to the mature wall-bound form, both extracellular intermediates are not yet linked to β1,6-glucan. This suggests that the formation of the GPI-CWP–β1,6-glucan–β1,3-glucan complex proceeds as follows: (i) β1,6-glucan is connected to β1,3-glucan, and (ii) a GPI-CWP is coupled to β1,6-glucan. Nothing is known about how the formation of other CWP–polysaccharide complexes proceeds.

GPI-CWPs, Pir-CWPs, as well as disulfide-linked cell wall proteins all have been successfully used to target fusion proteins to the cell wall for covalent attachment [27, 83, 85, 138145].

Finally, there is clear evidence for limited release of cell wall proteins into the medium [146], possibly as a result of remodeling of the cell wall during isotropic growth or when a new bud is formed.

2.8 Predictive value of the cell wall model for ascomycotinous yeasts

There is strong evidence that the cell wall of Candida albicans is similarly organized as the wall of S. cerevisiae [11]. The model presented here seems also valid for other Candida species. For example, in Candida glabrata an authentic GPI-CWP involved in adhesion to human epithelial cells has been identified that, after expression in S. cerevisiae cells, permits the latter to efficiently adhere to epithelial cells [147]. Pir-CWPs have been identified in C. albicans [94, 148], Kluyveromyces lactis [91], and in Zygosaccharomyces rouxii [91]. The cell wall model presented here is only partially valid for the fission yeast Schizosaccharomyces pombe. First, unlike S. cerevisiae, the wall of S. pombe contains a considerable amount of α1,3-glucan. Second, Southern analysis did not reveal PIR-like genes in S. pombe [91]. For a more detailed discussion of the cell wall of mycelial fungi from the Ascomycotina, see [42, 149].

2.9 Identification of cell wall mutants

De Groot and co-workers have shown that cell wall construction in S. cerevisiae is a tightly regulated process and that about 1200 genes directly or indirectly affect formation of the cell wall [150]. Using a hierarchical screening approach, the gene products can be divided in various classes such as proteins involved in the synthesis of cell wall macromolecules, proteins involved in remodeling or interconnecting cell wall polymers and proteins involved in the regulation of cell wall construction. Primary screens for the detection of mutants with an impaired cell wall are often based on increased sensitivity towards compounds that are known to interfere – either directly or indirectly – with normal cell wall construction, such as β1,3-glucanase [151, 152], Calcofluor white [150152] and the related compound Congo red [150, 153], SDS, and caffeine [150]. These compounds have in common that their deleterious effect on growth can be alleviated by osmotically stabilizing the medium [23, 150, 154, 155]. Increased sensitivity to sonication [150, 156] and hypotonic shock [157, 158] has also been used to detect mutants with a cell wall defect. 2-Deoxyglucose, which is incorporated in the cell wall and causes cell lysis in growing areas, may presumably also be effectively used to identify cell wall mutants [159, 160]. Alternatively, the culture filtrate may be screened for increased amounts of cell wall proteins or for incomplete CWP–polysaccharide complexes indicative of defective cell wall assembly [84, 150]. In general, cell wall perturbants appear to be more effective at 37°C than at 30°C, possibly because of the combined effect of increased turgor pressure on the wall and of defective cell wall construction due to thermal inactivation of cell wall assembly enzymes at 37°C (see Section 3.3).

Hypoglycosylation mutants represent a special category of cell wall-related mutants. They are often more resistant to orthovanadate and hypersensitive to hygromycin, which facilitates their identification [13]. Defects in O-glycosylation can be easily detected using the heavily O-glycosylated endochitinase Cts1 as a marker protein [99], whereas invertase is a useful marker for defects in N-glycosylation [161]. For a more detailed analysis of O- and N-glycosylation, FACE (fluorophore-assisted carbohydrate electrophoresis) is useful [162].

Mutants with an endocytotic block show a rare cell wall-related phenotype, namely, the mother cells of such mutants form multiple cell wall layers [163].

3 Cell wall dynamics

An estimated 1200 genes of S. cerevisiae affect the composition and organization of the cell wall [150]. Changes in the newly formed cell wall occur depending on the phase of the cell cycle, nutrient availability, and environmental conditions such as pH, temperature, and the availability of oxygen. Evidence is accumulating that all major signaling pathways such as the mitogen-activated protein (MAP) kinase pathways and the cAMP-dependent protein kinase A (PKA) pathway directly or indirectly may affect cell wall construction. The Pkc1-controlled MAP kinase pathway (the Slt2/Mpk1 MAP kinase pathway) is directly involved in signaling cell wall damage. The expression of cell wall proteins is particularly tightly controlled. For example, transcript levels of Cwp1, an abundant GPI-CWP, are cell cycle-controlled, but they are also up-regulated during late sporulation, under hypertonic conditions, and in case of cell wall stress [132, 133]. They are down-regulated during growth at low oxygen concentrations, low temperatures and during nitrogen depletion. This points to an intricate network of signaling mechanisms controlling the level of Cwp1. Regulation of the GPI-CWP Muc1/Flo11, which is involved in pseudohyphal and invasive growth and in biofilm formation (Table 2), is equally complex [164166]. These observations show that cell wall construction is a tightly regulated process.

3.1 Cell cycle-related changes in the cell wall

Pulse labeling of cells with FITC-concanavalin A, which specifically binds to mannoproteins, is an elegant means to monitor cell surface growth during the cell cycle [167, 168]. This approach has revealed that cells growing in rich medium switch from polarized growth to isotropic growth and vice versa during the cell cycle. Cells with a small bud show mainly tip growth, but during the G2–M phase they switch to isotropic growth [163, 169]. As a result, they acquire their characteristic oval shape. The formation of a septum between mother and daughter cell directs cell wall formation temporarily to the neck region. Released daughters tend to be smaller than mother cells [170]. They grow mainly isotropically until they have reached a critical size before a new budding cycle is initiated.

Immunofluorescence labeling and GFP tagging studies have shown that the expression of several cell wall proteins is cell cycle regulated [27, 58, 100, 171173]. This has been confirmed by transcript analysis, indicating that the expression of over 50% of all cell wall proteins is cell cycle regulated [117, 174]. For example, several cell wall proteins belong to the SIC1 expression cluster of cell cycle-regulated genes. This cluster is named for its prototype gene SIC1, which encodes an inhibitor of the Cdc28–Clb protein kinase complex. It comprises at least 27 genes that are maximally expressed in late M and/or early G1 [117]. Using YFP (yellow fluorescent protein) reporter constructs controlled by the promoter from the genes in the SIC1 cluster, Colman and co-workers have identified 10 genes in this group that show daughter-specific expression [100]. These include the endochitinase Cts1 [99], the putative glucanases Scw11, Dse2 and Dse4 [57, 100], and the GPI-CWP Pry3. The GPI-CWP Egt2, the loss of which causes a delay in cell separation [112], also shows daughter-specific expression, but only in cultures reaching saturation [100]. In view of their enzymatic activities, these proteins seem to be involved in cell separation by degradation of the septum from the daughter-proximal side. This also explains why the birth scar is much less obvious than the bud scar. Consistent with this, chitinase has been specifically localized in the cell wall at the daughter side of the bud neck [100]. Daughter-specific expression of this subset of SIC1 genes involves preferential localization of the transcription factor Ace2 in the nucleus of the daughter cell, possibly as a complex with Cbk1 and Mob2 [100].

The Pir-CWP-encoding genes PIR1, PIR, and PIR3 also belong to the SIC1 expression cluster and show a similar expression pattern [117]. Because the encoded proteins follow the secretory pathway, they will be incorporated in the cell wall during early G1, which is a period of isotropic growth. This raises the question whether they might have a function related to this mode of growth. PIR4/CIS3, on the other hand, is maximally expressed at the G2 phase [117]. In agreement with this, Pir4/Cis3 is predominantly found in the wall of growing buds [27]. The GPI-CWP Crh1 shows a particularly interesting expression pattern. It is deposited at the site of the future bud and, later, in the neck region of large-budded cells, consistent with peaking of transcript levels of CRH1 both in late G1 and in M/G1 [58].

Normally, the time of expression during the cell cycle seems to correlate with where at the cell surface a cell wall protein is incorporated. The GPI-CWP Cwp1, however, behaves differently [173]. Similar to the GPI-CWP Cwp2, its transcript levels are maximal at G2, but whereas Cwp2 is incorporated in the walls of the growing bud, Cwp1 is incorporated later and targeted specifically to the birth scar of the daughter cell. For further discussion of these and related topics, the reader is referred to [137, 166, 175].

3.2 Stationary phase

When cells enter the stationary phase, they form different walls. For example, they become thicker, more resistant to β1,3-glucanase digestion [22, 23, 119], and less permeable to macromolecules [22]. The GPI-CWP Sed1 becomes the most abundant cell wall protein [119], and the transcript level of SPI1, which encodes a related GPI-CWP [83], also strongly increases [120]. The level of mannosyl phosphorylation of cell wall proteins increases at the late-exponential and stationary phase of cell growth [176]. Also, the number of disulfide bridges in the wall increases 6–7-fold [22]. Force-deformation studies with single yeast cells entering stationary phase indicate that they strengthen their walls by simply increasing wall thickness, but not by increased cross-linking [4]. Interestingly, stationary phase cells have a considerably higher turgor pressure, compared to exponentially growing cells [29], which is probably caused by the strong increase in the level of trehalose [177179]. This is consistent with the presence of a thicker cell wall.

Evidence is emerging that the Ras/cAMP-dependent signal transduction pathway modulates cell wall integrity. Deletion of PDE2, which encodes the high affinity cAMP phosphodiesterase, results in up-regulation of this pathway and in increased sensitivity to hypotonic shock, indicating a loss of cell wall strength [158]. Conversely, overexpression of PDE2, which results in down-regulation of the pathway, suppresses the sorbitol dependence of a mutant strain with fragile walls and thus seems to strengthen the wall [158]. Similarly, overexpression of PDE2 leads to earlier mannosyl phosphorylation of cell wall proteins [176]. Entrance of the stationary phase is accompanied by low PKA activity [180]. As PKA negatively regulates expression of genes with stress response elements (STREs) in their promoter region [162], STRE-controlled genes, which include several cell wall-related genes, will become more active, thus contributing to the changes in cell wall properties observed in stationary phase.

3.3 Activation of the Slt2/Mpk1 MAP kinase pathway by cell wall stress

A GFP-tagged form of Pkc1, which drives the Slt2 MAP kinase pathway, is rapidly relocalized in cells challenged with the cell wall-degrading enzyme Zymolyase. Whereas in unchallenged cells, Pkc1 localizes in areas of polarized growth, it moves out within minutes to multiple spots outside this area [181]. Possibly, redistribution of Pkc1 under cell wall stress conditions is involved in targeting the secretory vesicles to sites of cell wall damage. Similarly, the β1,3-glucan synthase complex is redistributed in case of cell wall stress [182]. Both observations suggest that the cell is capable of locally repairing the cell wall in case of cell wall damage. The Slt2 MAP kinase pathway itself is also activated by cell wall stress. For example Zymolyase, which degrades the β1,3-glucan network, and cell wall perturbing compounds such as Calcofluor white or Congo red activate the Slt2 MAP kinase pathway [23, 183]. Caffeine, vanadate and SDS, which probably indirectly weaken the cell wall, also activate the Slt2 MAP kinase pathway [155, 184]. In addition, mutant cells with a weakened cell wall such as fks1Δ and gas1Δ show constitutive activation of the Slt2 MAP kinase pathway [23].

When cells are transferred to a medium of lower osmotic strength, resulting in increased water uptake and increased turgor pressure [185], an almost immediate activation of Slt2 is observed. Interestingly, the Slt2 MAP kinase pathway is also activated by high growth temperatures (37–39°C) [155, 186]. Compared to the rapid activation of the Slt2 MAP kinase pathway in cells challenged with hypotonic stress, which is maximal within 1 min and then slowly tapers off to the original state within 30 min, this temperature stress response is delayed, becoming maximal after 15–30 min. This raises the question whether activation of the Slt2 MAP kinase pathway may be an indirect effect of heat stress. The following observation is pertinent to this question. When cells are transferred to high growth temperatures (37–40°C), the cytosolic concentration of trehalose increases rapidly within minutes and continues to rise for up to 90 min to an estimated level of over 0.5 M [187191]. In addition, the intracellular glycerol level increases [192]. This raises the possibility that these responses, and the concomitant rise in turgor pressure, are largely or entirely responsible for activating the Slt2 MAP kinase pathway during heat stress. These observations can also explain why mutations in a member of the Slt2 MAP kinase pathway result in cell lysis only at high growth temperatures, but not at lower temperatures. Conceivably, thermal inactivation of proteins involved in cell wall construction may also result in cell wall weakening, thus aggravating the effect of increased turgor pressure. Finally, in cells treated with α-pheromone not only the Fus3 MAP kinase pathway is activated but also the Slt2 MAP kinase pathway. This is again a late response coinciding with the appearance of the mating structure [183, 193195]. It seems likely that the extensive remodeling of the cell wall required to form a mating structure leads to local softening of the wall, thus explaining the activation of the cell wall salvage pathway in the presence of pheromone. At the base of the mating structure chitin is deposited [183, 196]. Conceivably, the cell forms a zone of CWP-GPI–β1,6-glucan–chitin complexes in this area.

It is not precisely known how the cell detects cell wall stress. The observations discussed in the previous section can be most easily explained by assuming that the cell actually responds to acute plasma membrane stretch. This is also consistent with the observation by Kamada and co-workers [186] who found that chlorpromazine, which is supposed to cause stretching of the plasma membrane, also activates the Slt2 MAP kinase pathway. Membrane stretch is accompanied by opening of stretch-activated membrane channels [197] and a rapid increase in the cytosolic levels of free Ca2+ [198]. However, it is not known whether these responses are related to the cell wall salvage pathway. It is conceivable that an additional pathway may be activated using the Ca2+/calmodulin-regulated protein phosphatase calcineurin, which co-regulates the transcript level of FKS2/GSC2 and thus influences the synthesis of β1,3-glucan (see Section 3.5) [199]. Several putative sensor proteins for cell wall stress have been identified that function upstream from the Slt2 MAP kinase pathway, such as the type I transmembrane proteins Slg1, Wsc2/3/4, Mid2, and Mtl1 [183, 195, 200]. They are located in the plasma membrane and uniformly distributed over the cell surface [195, 200] and act through Rom2, a guanine nucleotide exchange factor for Rho1 [201]. Of these, Mid2 seems to play a major role in sensing cell wall stress and in the coupled activation of the MAP kinase Slt2 [23, 155]. Mid2 is a single span transmembrane protein with its C-terminus in the cytosol and its N-terminus outside the plasma membrane. An important feature of the part of the protein outside the plasma membrane is the presence of an extended serine- and threonine-rich region, which is essential for its function [183, 201]. This is consistent with the observation that O-glycosylation of Mid2 by the protein mannosyl transferase Pmt2 is essential for its signaling function [201]. Conceivably, Mid2, and also the other cell wall integrity sensors, which have a similar organization, are linked to the β1,3-glucan network through one or more of its O-linked side chains, in a Pir-CWP-like fashion.

Recently, it has been reported that mutations in SLG1/WSC1 and the other WSC genes result in increased sensitivity towards hydrogen peroxide and ethanol [202]. Intriguingly, these treatments are known to induce the intracellular accumulation of trehalose [187], and, thus, cause the turgor pressure to rise. This suggests that wsc mutants are hypersensitive to these stress treatments, because they cannot adapt to the accompanying increase in turgor pressure.

Generally, cells with a defective cell wall as a result of either a genetic defect or treatment with a cell wall perturbing compound can be rescued by increasing the osmotic value of the medium, for example by adding 10% (w/v) sorbitol [154, 155]. This sounds paradoxical, because increasing the osmotic strength of the medium is expected to activate the Hog1 MAP kinase pathway. As a result, the intracellular glycerol level is expected to increase, thus counteracting the effect of increasing the osmotic strength of the medium. This raises the question whether the Hog1 MAP kinase pathway is suppressed in case of cell wall stress. For further discussion of the cell wall salvage pathway, the reader is referred to [175, 203207].

3.4 Ultrastructural changes in the wall in response to cell wall defects

Recent studies using genomic transcript analysis have shown that constitutive activation of the Slt2 MAP kinase pathway results in the up-regulation of mainly cell wall-related genes [113, 114]. These include genes that code for biosynthetic enzymes such as Gfa1 and Chs3, which are involved in chitin synthesis, and Gsc2/Fks2, which is involved in the synthesis of β1,3-glucan, cell wall proteins such as Cwp1, Crh1, and the Pir-CWPs, and the MAP kinase Slt2 itself [113, 114]. A simple way to constitutively activate the Slt2 MAP kinase pathway is based on using cell wall mutants with a reduced amount of β1,3-glucan or β1,6-glucan [23]. In these mutants various changes in cell wall composition and organization have been detected (Table 3) [65, 73, 116, 130, 131]. Although their cell wall organization differs considerably, it seems unlikely that the cell is able to sense specific cell wall defects and respond accordingly. Presumably, by activating the Slt2 MAP kinase pathway, cell wall stress leads in all cases to the activation of multiple cell wall reinforcing reactions, which – depending on the cell wall defect involved – may result in various changes in cell wall organization. Generally, the chitin level in the wall of cells confronted with cell wall stress is strongly elevated [10, 65]. This is consistent with the observation that many mutants with a cell wall defect are synthetic lethal with chs3Δ and hypersensitive to nikkomycin, an inhibitor of chitin synthesis [55, 135]. Chitin deposition in response to cell wall stress seems to be independent of Chs6 [208], which under non-stress conditions is required for proper Chs3 targeting [209].

3.5 Control of GSC2/FKS2

FKS1 and GSC2/FKS2 encode alternative subunits of the β1,3-glucan synthase complex. FKS1 is preferentially expressed under optimal growth conditions and is cell cycle-regulated [210, 211]. Regulation of GSC2/FKS2 expression is independently controlled by various inputs [199]. For example, growth at 39°C leads to increased expression of FKS2 through parallel inputs from the Slt2 MAP kinase pathway and from calcineurin, acting on different parts of the 5′ upstream region of GSC2/FKS2 [199]. GSC2/FKS2 is also independently up-regulated upon entry into the stationary phase and at low glucose levels [199]. As discussed above, transcript levels of GSC2/FKS2 also rise in case of cell wall stress in an Slt2-dependent fashion [23, 113].

3.6 The Hog1 MAP kinase pathway

The Hog1 MAP kinase pathway is activated in response to hyperosmotic stress. This seems to be accompanied by transient usage of an altered set of cell wall proteins. The transcript levels of the GPI-CWP-encoding genes CWP1, SED1, and SPI1, and of SPS100 and YGP1, which encode putative cell wall proteins (see Section 2.5), increase, whereas the transcript levels of the GPI-CWP-encoding genes CWP2 and EGT2 decrease [212214]. Cells quickly adapt to hypertonic stress, resulting in down-regulation of many of the up-regulated genes.

There is clear evidence that the Hog1 MAP kinase signaling pathway is also required for cell wall construction under non-stress conditions. First, both deletion of PBS2, which codes for the MAP kinase kinase of this pathway, and of HOG1, results in increased sensitivity of intact cells to β1,3-glucanase, whereas multicopy expression of PBS2 results in lower sensitivity to β1,3-glucanase [192, 215]. In addition, β1,6-glucan levels are slightly increased in pbs2Δ cells and slightly lowered in cells that overexpress PBS2 [216]. Third, in hog1Δ cells the Golgi mannosyl transferase Mnn1 is mislocalized [217]. Fourth, as discussed in Section 3.8, cells kept at low pH become highly resistant to β1,3-glucanase in a Hog1-dependent fashion [115]. Finally, pbs2Δ and hog1Δ are more resistant to the cell wall perturbant Calcofluor white than wild-type cells, again indicating that the mutant cell wall differs from the wild-type wall [66]. How the Hog1 MAP kinase pathway affects cell wall construction still remains unclear.

3.7 Responses to anaerobic conditions

When oxygen levels decrease, the cell seems to switch from one set of GPI-CWPs to another [121]. Northern analysis reveals that the transcript level of CWP1 and CWP2, which encode two abundant GPI-CWPs, decrease, whereas the transcript levels of DAN1/CCW13, DAN2, DAN3, and TIR1, TIR2, TIR3, and TIR4, and TIP1 are strongly up-regulated. Similarly, cells growing in stagnant cultures, which results in hypoxic growth conditions, show a strong increase in the transcript level of TIR1 and the level of Tir1 in the wall [123]. Interestingly, tir1Δ, tir2Δ, and tir3Δ cells arrest in the unbudded phase when transferred to anaerobic conditions, suggesting that Tir1, Tir2, and Tir3 are required for growth during G1 [121]. Genome-wide transcript analysis of chemostat-cultured cells grown under anaerobic conditions are largely consistent with these data and also reveal up-regulation of MUC1/FLO11, a GPI-CWP involved in pseudohyphal growth, and strong down-regulation of SPS100 [218]. It has further been shown that nearly the entire family of PAU genes (see Section 2.5) is up-regulated under anaerobic conditions [124, 218]. Biochemical data of the molecular organization of cell walls from cells grown under anaerobic conditions are scarce. For information about the regulation of CWP-encoding genes, induced under anaerobic conditions, the reader is referred to [124, 219221].

3.8 Responses to extreme environmental pH values

In its natural environment, for example in rotting fruits, S. cerevisiae is often confronted with low pH values. Growth at low pH leads to various adaptations in cell wall composition and organization. First, the cells become highly resistant to the recombinant β1,3-glucanase Quantazyme [115]. As this is not due to an increase in chitin content of the wall, this suggests that the mannoprotein layer has become less permeable to Quantazyme and, presumably, to other fungal cell wall-degrading enzymes normally found in plant tissues as well. Second, the cells use the alkali-sensitive linkage between cell wall proteins and β1,3-glucan more extensively, resulting in more efficient incorporation of Pir-CWPs and an increase in double-linked GPI-CWPs [115]. Genomic transcript analysis shows that the transcript levels of the GPI-CWP-encoding genes CWP1, HOR7, and SPI1 are increased [115]. Also the transcript level of the putative cell wall protein-encoding gene YGP1 (Section 2.5) is increased. A similar genomic transcript analysis, comparing pH 4 and 8, identified only two alkali-inducible GPI-CWP-encoding genes, namely, FIT2, and FIT3 [222], believed to be involved in iron uptake [127]. Interestingly, the increase in Quantazyme resistance at low pH is dependent on Hog1, but not on Slt2 [115], supporting the notion that also the Hog1 MAP kinase pathway influences cell wall organization (see also Section 3.6).

4 Perspectives

The molecular organization of the cell wall of S. cerevisiae, growing in rich medium, has now largely been elucidated. This, however, is only part of the answer. In view of the dynamic nature of the cell wall, a study of cell wall organization using chemostat-cultured cells limited for carbon, nitrogen and phosphate may reveal the limits of the variability in cell wall composition of yeast. In addition, a study of cell wall construction in pseudohyphal cells, in which polarized growth predominates and isotropic growth is limited, seems worthwhile.

The construction of the fungal cell wall is an attractive target for the development of new antifungal compounds. The fact that the molecular organization of the cell wall is now largely elucidated makes it possible to have a closer look at more specific targets such as cell wall polymer synthases, cell wall assembly enzymes and enzymes involved in remodeling of cell wall polymers. With some exceptions, however, detailed research of these enzymes in yeasts and other fungi is still in its infancy. Above all, biochemical and also crystallographic studies of these enzymes are urgently needed to complement the vast amount of molecular genetic information currently available.

The cell wall of yeast has long been viewed as a relatively static structure with limited changes in composition and structure depending on environmental changes. It is now clear that cell wall construction just like in bacteria is a very dynamic process and that the cell tends to continually adapt its newly formed wall to changing conditions both in terms of cell wall organization and with respect to the cell wall proteins presented at the cell surface. How all this is regulated is still largely unclear. Additional questions are how the cell behaves when it is confronted with multiple stress conditions simultaneously. For example, yeast cells growing on rotting grapes are confronted with a rather acidic, extremely glucose-rich environment in which also fungal cell wall-degrading enzymes abound. Are perhaps some signaling pathways epistatic with respect to each other? For example, one may imagine that the cAMP PKA pathway, which when activated leads to a decrease in cell wall strength, is down-regulated in case of cell wall stress. Better understanding of these regulatory processes may help to further and more efficiently exploit this organism, and other fungi, for the benefit of mankind.


We would like to acknowledge the financial support from the Netherlands Technology Foundation (STW) and from EUROFAN II. We also thank Dr. Han de Winde for critical reading of the manuscript.


  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].
  26. [26].
  27. [27].
  28. [28].
  29. [29].
  30. [30].
  31. [31].
  32. [32].
  33. [33].
  34. [34].
  35. [35].
  36. [36].
  37. [37].
  38. [38].
  39. [39].
  40. [40].
  41. [41].
  42. [42].
  43. [43].
  44. [44].
  45. [45].
  46. [46].
  47. [47].
  48. [48].
  49. [49].
  50. [50].
  51. [51].
  52. [52].
  53. [53].
  54. [54].
  55. [55].
  56. [56].
  57. [57].
  58. [58].
  59. [59].
  60. [60].
  61. [61].
  62. [62].
  63. [63].
  64. [64].
  65. [65].
  66. [66].
  67. [67].
  68. [68].
  69. [69].
  70. [70].
  71. [71].
  72. [72].
  73. [73].
  74. [74].
  75. [75].
  76. [76].
  77. [77].
  78. [78].
  79. [79].
  80. [80].
  81. [81].
  82. [82].
  83. [83].
  84. [84].
  85. [85].
  86. [86].
  87. [87].
  88. [88].
  89. [89].
  90. [90].
  91. [91].
  92. [92].
  93. [93].
  94. [94].
  95. [95].
  96. [96].
  97. [97].
  98. [98].
  99. [99].
  100. [100].
  101. [101].
  102. [102].
  103. [103].
  104. [104].
  105. [105].
  106. [106].
  107. [107].
  108. [108].
  109. [109].
  110. [110].
  111. [111].
  112. [112].
  113. [113].
  114. [114].
  115. [115].
  116. [116].
  117. [117].
  118. [118].
  119. [119].
  120. [120].
  121. [121].
  122. [122].
  123. [123].
  124. [124].
  125. [125].
  126. [126].
  127. [127].
  128. [128].
  129. [129].
  130. [130].
  131. [131].
  132. [132].
  133. [133].
  134. [134].
  135. [135].
  136. [136].
  137. [137].
  138. [138].
  139. [139].
  140. [140].
  141. [141].
  142. [142].
  143. [143].
  144. [144].
  145. [145].
  146. [146].
  147. [147].
  148. [148].
  149. [149].
  150. [150].
  151. [151].
  152. [152].
  153. [153].
  154. [154].
  155. [155].
  156. [156].
  157. [157].
  158. [158].
  159. [159].
  160. [160].
  161. [161].
  162. [162].
  163. [163].
  164. [164].
  165. [165].
  166. [166].
  167. [167].
  168. [168].
  169. [169].
  170. [170].
  171. [171].
  172. [172].
  173. [173].
  174. [174].
  175. [175].
  176. [176].
  177. [177].
  178. [178].
  179. [179].
  180. [180].
  181. [181].
  182. [182].
  183. [183].
  184. [184].
  185. [185].
  186. [186].
  187. [187].
  188. [188].
  189. [189].
  190. [190].
  191. [191].
  192. [192].
  193. [193].
  194. [194].
  195. [195].
  196. [196].
  197. [197].
  198. [198].
  199. [199].
  200. [200].
  201. [201].
  202. [202].
  203. [203].
  204. [204].
  205. [205].
  206. [206].
  207. [207].
  208. [208].
  209. [209].
  210. [210].
  211. [211].
  212. [212].
  213. [213].
  214. [214].
  215. [215].
  216. [216].
  217. [217].
  218. [218].
  219. [219].
  220. [220].
  221. [221].
  222. [222].
View Abstract