Yeast KRE Genes Provide Evidence for a Pathway of Cell Wall/ -Glucan Assembly

The Saccharomyces cerevisiae KRE1 gene encodes a Ser/Thr-rich protein, that is directed into the yeast secretory pathway, where it is highly modified, probably through addition of O-linked mannose residues. Gene disruption of the KRE1 locus leads to a 40% reduced level of cell wall (l~6)-/~glucan. Structural analysis of the (1---6)-/~-glucan fraction, isolated from a strain with a krel disruption mutation, showed that it had an altered structure with a smaller average polymer size. Mutations in two other loci, KRE5 and KRE6 also lead to a defect in cell wall (1-*6)-~-glucan production and appear to be epistatic to KRE/. These findings outline a possible pathway of assembly of yeast cell wall (l~6)-fl-glucan. ~ -G LUCA~S, homopolymers of glucose, are an abundant class of polysaccharides that includes cellulose, and appears to serve structural, functional, and morphological roles at the cell surface of fungi, bacteria, and plants (Fleet and Phaff, 1981; Sharp et al., 1984; Inon de Iannino and Ugalde, 1989; Kato, 1981). Despite their widespread occurrence, there has been surprisingly little work to address the basis of cell wall glucan biosynthesis at the genetic and molecular level in eukaryotes. In vitro enzymatic reactions resulting in glucan synthesis have been defined and partially characterized for several systems (Kang and Cabib, 1986; Aloni et al., 1982), although components of the synthetic machinery have eluded purification. The isolation of mutants defective in the production of cell wall glucan should define genes that encode biosynthetic enzymes as well as other products, for example those that regulate glucan synthesis or generate glucan precursors. A mutant approach has been valuable in understanding the synthesis of such other cell wall polysaccharides, as mannan (Ballou, 1982) and chitin (Silverman et al., 1988; Bulawa et al., 1986). Mixed linked /3-D-glucans consisting of glucopyranosyl residues joined through (1-.3) and (1-*6)-linkages are common to fungi belonging to the Ascomycetes, Basidomycetes, and Oomycetes (Wessels and Sietsma, 1981). Fractionation studies of the Saccharomyces cerevisiae cell wall demonstrated the presence of several glucan subclasses, which could be structurally distinguished by polymer length and the ratio of (1-.3) to (1-*6)-/3-D-linkages (Fleet and Manners, 1976). Much of the yeast cell wall glucan is isolated from whole cells as an alkali insoluble fraction that was found to contain two distinct types of polymers. The most abundant alkali insoluble glucan consists predominantly of repeating units of linear (l~3)-/~-linked residues, 3% of which are branched through a (l~6)-/~-linkage (Manners et al., 1973a). This gluean has a degree of polymerization estimated to be 1,500 and has been proposed to determine the shape and stability of the yeast cell wall (Zlotnik et al., 1984). The other alkali-insoluble glucan has a degree of polymerization estimated to be 140 and contains residues that are predominantly connected through linear (1-*6)-/~-linkages (Manners et al., 1973b). This glucan will be referred to as (l~6)-/3-glucan, although in addition to linear (1-*6)linked units it is composed of some linear (l-*3)-linked residues and a relatively high proportion of (1-'3, 1-~6) linked branched residues (14%). Yeast (1-*6)-/~-glucan accounts for ,x,20 % of the alkali insoluble glucan or 3 % of the total cellular dry weight. The K1 killer toxin of S. cerevisiae provides a selection scheme for the isolation of mutants defective in (I~6)-/~-D glucan production. This toxin is a protein secreted by killer yeast strains which kills sensitive (nonkiller) strains. K1 toxin displays a lectin-like affinity for linear (I~6)-/~-D glucan and must bind to the wall of sensitive yeast in order to initiate the killing process (Bussey et al., 1979). Mutations in the KRE/gene result in killer toxin resistance and are associated with an abnormal production of the cell wall (1-*6)-/3-glucan (Hutchins and Bussey, 1983). We describe here that the KRE/gene encodes a protein directed into the yeast secretory pathway. The (1~6)-/~ glucan fraction which remained in a krel mutant yeast strain had an altered structure with a smaller average polymer size and suggests that (l~6)-/3-glucan is synthesized in a stepwise manner. We address this possibility through the isolation of additional killer resistant mutants, some of which are required for (1-*6)-/~-glucan biosynthesis and appear to be epistatic to KRE/. Gene products required for fungal cell wall biosynthesis have been recognized as potential targets for specific antifungal antibiotics and the KRE genes are discussed in this context. © The Rockefeller University Press, 0021-9525190/05/1833/I 1 $2.00 The Journal of Cell Biology, Volume 110, May 199