Flocculation of Saccharomyces cerevisiae

Flocculation of Saccharomyces cerevisiae

Food Research International, Vol. 31, No. 6±7, pp. 421±440, 1998 # 1999 Canadian Institute of Food Science and Technology Published by Elsevier Scienc...

319KB Sizes 0 Downloads 50 Views

Food Research International, Vol. 31, No. 6±7, pp. 421±440, 1998 # 1999 Canadian Institute of Food Science and Technology Published by Elsevier Science Ltd. All rights reserved Printed in Great Britain PII: S0963-9969(99)00021-6 0963-9969/99/$ - see front matter

Flocculation of Saccharomyces cerevisiae Yu-Lai Jina & R. Alex Speersb* a Department of Food Science, Yangzhou University, Yangzhou 225009, People's Republic of China Department of Food Science and Technology, Dalhousie University, PO Box 1000, Halifax, NS Canada B3J 2X4

b

This paper reviews our current understanding of cell ¯occulation with particular emphasis of the process in brewing fermentations. While cell ¯occulation has been examined for over a century and has been the subject to a number of reviews in the early part of this decade, our view of the process is cloudy. Flocculation is a€ected by cell genetic behavior, cell age as well as the chemical and physical nature of the surrounding medium. Recently, a number of advances in our understanding of the genes governing the process have occurred. In conjunction with these genetic advances, new assay methods have also been developed. This review will discuss and update our current knowledge of cell ¯occulation and its use in brewing fermentations. # 1999 Canadian Institute of Food Science and Technology. Published by Elsevier Science Ltd. All rights reserved Keywords: Saccharomyces cerevisiae, ¯occulation, aggregation, brewing yeast.

cells suspended in wort during both primary and secondary fermentation is a key factor in¯uencing fermentation speed, beer ¯avor, maturation and ®ltration. Although centrifugation can be applied to separate suspended cells, ¯occulation is still an important and necessary process for the removal of yeast. As mentioned, brewers and microbiologists have been exploring the mystery of ¯occulation of yeast cells since the establishment of pure yeast culture in the last century. Up to the 1960s, researchers examined the e€ect of environmental conditions such as salts, sugars, ethanol, pH, temperature, dissolved oxygen content and proteolysis on cell ¯occulation (Gilliland, 1951; Helm et al., 1953; Lindquist, 1953; Eddy, 1955a,c 1958; Eddy and Rudin, 1958; Harris, 1959; Mill, 1964). Since the mid1970s, genetic studies have helped our understanding of yeast ¯occulation at the molecular level (Johnson and Lewis, 1974; Lewis et al., 1976; Stewart et al., 1976; Stewart and Russell, 1977; Holmberg, 1978; Holmberg and Kielland-Brandt, 1978; Russell et al., 1980; Stewart and Russell, 1981; Teunissen et al., 1993, 1995; Teunissen and Steensma, 1995; Lo and Dranginis, 1996; Lo and Dranginis, 1998). The mechanism of lectin-like cell±cell interactions has been established to explain yeast ¯occulation in the past two decades (Taylor and Orton, 1978; Hough et al., 1982a; Miki et al., 1982a,b; Speers et al., 1992a, 1993b; Stratford, 1992c; Patelakis et al., 1998; Speers et al.,

INTRODUCTION In the food and biotechnological industries where fermentation occurs or cell reactors are employed, suspended cells must normally be separated from the media prior to further processing. Beverage alcohol production is a typical example of this process. Despite the study of cell ¯occulation since the time of Pasteur, many details of the process have been subject to debate and remain to be elucidated. Indeed to the occasional reader of cell ¯occulation literature, the process must seem obscure. While scientists from many disciplines have examined the ¯occulation of various cell species, the majority of research has focused on the ¯occulation of brewing yeast. Not surprisingly therefore, this paper will review our understanding of cell ¯occulation with particular reference to the brewing process. However, the speci®c ®ndings gained and techniques used from the study of Saccharomyces cerevisiae can often be applied to the ¯occulation of other cells in various biotechnological applications. In selecting a yeast strain for beer production, a brewer considers a number of factors. Yeast's ¯occulation ability is one of the major concerns. The number of *To whom correspondence should be addressed. Fax: +1902-420-0219; e-mail: [email protected] 421

422

Y.-L. Jin, R. A. Speers

1998). As the lectins found on S. cerevisiae cells have a speci®c role in yeast ¯occulation, they have been de®ned as zymolectins to distinguish them from lectins of other yeasts or other microorganisms (Speers et al., 1998). The detection and quanti®cation of zymolectins or yeast cell surface lectins have also been attempted but there has been no report on the correlation between zymolectin density and ¯occulation (Masy et al., 1992a; Ritcey, 1997; Patelakis et al., 1998). Since the 1980s, cell surface hydrophobicity (CSH) has been identi®ed as a second major factor responsible for ¯occulation onset and ¯occulence of brewing yeast (Straver et al., 1993b, 1994a; Straver and Kijne, 1996; Akiyama-Jibiki et al., 1997). Therefore, the relationship between cell surface hydrophobicity and ¯occulation also warrants further investigation. This paper will review various aspects of cell ¯occulation including its measurement, e€ect of cell wall structure, genetic and colloidal aspects, the e€ect of the environment and our current knowledge of the mechanism of the process. It is the authors' hope that this review will serve to summarize and clarify our current understanding of this fascinating phenomenon. YEAST AND YEAST FLOCCULATION Factors a€ecting attenuation A number of factors in¯uencing attenuation and the fermentation process have been recognized. Wort composition and pH are important for yeast growth and fermentation. For this reason, the levels of fermentable carbohydrates, a-amino acids and dissolved oxygen are usually monitored by the brewers. Lack of contact between yeast and oxygen over successive fermentations decreases attenuation from 67 to 44% where the wort of 0.5 mL/L dissolved oxygen was used (Hough et al., 1982b). Pitching rate greatly in¯uences the speed of fermentation and therefore in¯uences the time required for fermentation. The amount of yeast reproduced during fermentation depends on the pitching rate although the crop of yeast at the end of fermentation is almost independent of the pitching rate. On the other hand, the concentration of yeast cells suspended during the fermentation depends on mainly the ¯occulation behavior of the yeast, the pitching rate, agitation caused by convection currents in the fermenter due to rousing or stirring, and the size and geometry of the fermenter. It has been recently suggested that aging and senescence of yeast cells may play a role in the `hung fermentations' of the brewing process (Barker and Smart, 1996). Parameters such as fermentation temperature, pressure and time also a€ect fermentation pro®les. It is dicult to rank the order of importance of these parameters because of the variations among batches of fermentation, beer types and brewery plants.

As well, the importance of the yeast strain should be noted. Characteristics such as the ability of the yeast to ferment maltotriose greatly in¯uence the attenuation. Respiratory-de®cient mutants of brewing yeast may arise spontaneously at frequencies 0.5%. Such RD mutants are unable to respire glucose and usually produce high levels of vicinal diketones particularly diacetyl (Hough et al., 1982c). On contrary, transformed brewing yeast containing STA2 gene may super-attenuate the wort with pronounced utilization of dextrin in pilot scale of brewing (Hammond, 1995). Yeast taxonomy Yeasts are protists that possess many characteristics of higher cells, but show a simpler level of biological organization (Hough et al., 1982a). Classi®ed as fungi at the level of family, all yeasts are non-photosynthetic higher protists with rigid cell walls and exist as either unicellular organisms or mycelia. Under the genus of Saccharomyces, the old terms to which the fermentation industry is accustomed are: S. cerevisiae (ale yeast), S. carlsbergensis (later S. uvarum) (lager yeast), S. ellipsoideus later called S. cerevisiae var. ellipsoideus (wine yeast), S. oviformis later termed S. bayanus (employed in wine refermentations) and S. pastorianus (used in wine fermentation in cold climates) (Martini and Martini, 1989). It is noteworthy that there are over 1000 individual ale strains of S. cerevisiae (Stewart et al., 1975b; Russell, 1995). In the brewing industry, ale and lager beers are normally fermented with S. cerevisiae and S. carlsbergensis, respectively. Traditionally the ale yeasts or top yeasts are collected from the surface by the process of skimming while the lager yeasts or bottom yeasts are cropped from the bottom of the fermentation vessel. The di€erentiation of lagers and ales on the basis of bottom and top cropping has become less distinct as a result of the application of modern cylindroconical tanks and centrifuges (Russell, 1995). Thus, brewers select less ¯occulent strains for use in centrifugeequipped plants. However, due to cell ¯occulation variability, the understanding and monitoring of cell ¯occulation is a concern to all brewers. Originally identi®ed as S. carlsbergensis and S. cerevisiae, the lager and ale yeasts were distinguished on the basis of melibiose fermentation by the American Society of Brewing Chemists (ASBC, 1995). Lager yeast strains that possess the MEL gene(s) produce extracellular melibiase or a-galactosidase to utilize melibiose, whereas ale yeast strains that do not produce melibiase are unable to utilize melibiose. In 1970, S. carlsbergensis was renamed as S. uvarum as shown in Table 1 (Lodder, 1970). After exhaustive investigations based on such procedures as electrophoretic analysis of cellular enzymes, proton magnetic resonance spectrum, antigenic activity of cell wall mannans, and the percentages of guanine and cytosine (mol% G+C) of nuclear DNA, it was unequivocally shown that the yeasts used in the

Flocculation of Saccharomyces cerevisiae alcoholic fermentation industry consistently fall into the same species, S. cerevisiae in spite of their di€erences in some technological properties (Kreger-van Rij, 1984; Martini and Martini, 1989). However, such di€erences of the brewing yeast strains dismissed by the taxonomists are technically important to the brewers. Thus, brewers and brewing scientists use the labels S. cerevisiae and S. uvarum to denote ale and lager yeasts, respectively. Aggregation and ¯occulation of brewing yeast Microbial aggregation has been de®ned as a collection of microbial cells in intimate contact or as the gathering together of units to make a larger unit (Calleja et al., 1984; Calleja, 1987). Therefore, the aggregation may include agglutination (mating) and ¯occulation but not chain formation (Calleja, 1987). However, the European Brewery Convention (EBC) suggested that `the aggregation of yeast cells into ¯ocs which may be due to either non-separation of cells after budding or coalescence of single cells into clumps late in fermentation' (EBC Microbiologica, 1981). Chain formation has been generally held as a unique type of cell aggregation (Wilcocks and Smart, 1995; Stratford, 1996a).

423

Yeast mating involves haploid strains of a and a sexes of S. cerevisiae which exchange small peptide pheromones or a and a factors followed by aggregation before cell fusion to form a diploid cell. The cell-to-cell bonding in this case is due to protein-protein interactions between a and a agglutinins anchored in the complementary cell walls (Fig. 1a). Chain formation is caused by failure of daughter cells to separate from mother cells during cell division. The cells grow into chains of up to 100 cells by further budding of mother and daughter cells (Fig. 1b). Such chains can be dispersed irreversibly by mechanical shear rather than ethylenediaminetetraacetic acid (EDTA) since the chain structure is calcium-independent (Stratford, 1996a). Yeast ¯occulation has been de®ned as `the phenomenon wherein yeast cells adhere in clumps and sediment rapidly from the medium in which they are suspended' (Stewart et al., 1976). The responsible bonding has been proved to involve lectin-like protein-carbohydrate recognition and interaction in a manner of calciumdependent and sugar sensitive (Fig. 1c). A more restrictive de®nition has been suggested as `the non-sexual aggregation of yeast cells into clumps, dispersible by

Table 1. Changes in the nomenclature of S. cerevisiae (Martini and Martini, 1989)

S. cerevisiae S. willianus S. coreanus S. carlsbergensis S. uvarum S. logos S. bayanus S. pastorianus S. oviformis S. beticus S. heterogenicus S. chevalieri S. fructuum S. italicus S. steineri S. globosus

The Yeast, a Taxonomic Study, 3rd edn, 1984

The Yeast, a Taxonomic Study, 2nd edn, 1970

The Yeast, a Taxonomic Study, 1st edn, 1952 c

S. cerevisiae

c

S. coreanus

c

S. uvarum

c

S. bayanus

c

S. heterogenicus

c

S. chevalieri

c

S. italicus S. globosus

c

S. aceti S. prostoserdovii S. oleaginosus S. oleaceus S. capensis S. diastaticus S. hienipiensis S. inusitatus S. norbensis S. abuliensis S. cordubensis S. gaditensis S. hispalensis S. cerevisiae

c

S. cerevisiae

Fig. 1. Aggregation of brewing yeast (Calleja, 1987; Speers and Ritcey, 1995).

424

Y.-L. Jin, R. A. Speers

EDTA or speci®c sugars and the subsequent removal of these clumps from medium' (Stratford, 1996a). In some cases, co-¯occulation may occur between ¯occulent and non-¯occulent strains where non-¯occulent cells adhere to ¯occulent cells (Curtis and Wenham, 1958; Miki et al., 1982a). So far co-¯occulation has not been found among lager strains (Enari, 1995). Eddy (1958) suggested the term of mutual ¯occulation for a pair of non-¯occulent yeast strains that ¯occulate in the presence of each other but do not ¯occulate separately. Measurement and classi®cation of yeast ¯occulation Microbial aggregates may be characterized by their strength, morphology, extent or rate of aggregation (Calleja, 1984). The measurements of the extent of cell aggregation are more appropriate for yeast ¯occulation quanti®cation. To measure the extent of ¯occulation, visual estimation is rather subjective due to judgment variations among observers. However, visual methods are simple and fast (Calleja and Johnson, 1977). Recently authors have suggested that the rate of agitation may substantially in¯uence the initial rate and the ®nal equilibrium of ¯occulation (Stratford, 1992c; Speers and Ritcey, 1995). After consideration of fermenter shear rates, it has been proposed to control agitation at an average rate of shear of 25 sÿ1 during ¯occulation determinations (Speers and Ritcey, 1995). For routine ¯occulation assays, Gilliland and Helm methods and Variants thereof have been chosen as standard methods (EBC Microbiologica, 1981; ASBC, 1986), respectively. In Gilliland's test, yeast strains are classi®ed into four classes. Class I strains are non-¯occulent or powdery yeasts. Class II strains are those that ¯occulate into small loose clumps towards the end of fermentation, therefore both Class I and Class II attenuate very well while Class II separate from the medium well (moderately ¯occulent). Class III strains can form lumpy masses towards the end of fermentation and sediment rapidly (strongly ¯occulent). Class IV strains are those that sediment very soon at the beginning of fermentation due to chain-formation. For Helm's test, suspension of Type 1 or ¯occulent yeast strains separates into two layers near the top. Such an interface falls rapidly and a falling boundary is measured after 10 min. For the Type 2 or non-¯occulent strains, an interface forms much more slowly near the bottom and a rising boundary is measured after 10 min. Both Gilliland's and Helm's methods are quick semi-qualitative approaches. Gilliland's method has been found to give variable results. The Helm's test was generally consistent in distinguishing ¯occulent strains from non-¯occulent ones but gave mixed results for identifying moderately ¯occulent strains (ASBC, 1993, 1994). To improve the Helm's method with regard to its reproducibility, the environmental factors which in¯uence ¯occulation must be included (D'Hautcourt and Smart, 1998).

During 1993±1996, the ASBC Microbiology Subcommittee selected a standard method suggested by Bendiak (ASBC, 1993±1996; Bendiak, 1994). The measurement is based on Helm's test and uses controlled inoculum, yeast growth, cell density of the suspension, and experiment temperature. The ¯occulation is expressed as the percentage of A600 reduction of the top 1.0 ml suspension in 6 min compared with EDTA-treated suspension. Strains can be classi®ed as non-¯occulent (<20%), very ¯occulent (>85%) and moderately ¯occulent (20±80%) (ASBC, 1996). As a modi®ed Helm's method, the test is not fundamental but relatively qualitative and fast. The on-line measurement of the rate of yeast settling has been attempted (Podgornik et al., 1997). As well other procedures based on sugar de¯occulation (Eddy, 1955b), thermal de¯occulation (Taylor and Orton, 1975), turbidity of suspension in a glass capillary (van Hamersveld et al., 1996), or hydrophobic interaction chromatography (Akiyama-Jibiki et al., 1997) have been employed. One might expect various procedures to be developed and these methods may coexist until a standard method is accepted by both brewers and researchers. THE ROLE OF THE CELL WALL IN YEAST FLOCCULATION It is generally agreed that the yeast cell wall is an important indicator of the rate and extent of cell wall ¯occulation (Calleja, 1987). Heat-killed cells will ¯occulate if they were originally ¯occulent (Mill, 1964). As well, isolated cell walls will ¯occulate if originally ¯occulent whereas walls from non-¯occulent cells will not ¯occulate (Eddy, 1955c). Composition and structure As an easy source of biomass and an important agent in alcoholic beverage industry, S. cerevisiae has been well studied, especially the composition and structure of the cell wall. The yeast cell wall surrounds the periplasmic space, spans 100±200 nm and represents some 15±25% of the total dry mass of the cell (Stratford, 1994). It is composed about 60±85% of carbohydrates, of which some are covalently linked to proteins. Beta-glucans and amannans equally comprise the majority of the carbohydrate of the wall. It has been reported that b-1,3-glucans have an estimated size of 1500 glucose units (or degree of polymerization or DP), whereas b-1,6-glucans have about 130±140 DP (Manners et al., 1973a,b). Alpha-mannan is a component of the secreted glycoproteins which are anchored in the cell wall. Four types of glycosylation in wall mannoproteins have been reported: (1) N-linkage to asparagine, (2) O-linkage to serine or threonine, (3) linkage of glucomanno-side chain to unknown amino acid (AA) residues and (4)

Flocculation of Saccharomyces cerevisiae attachment of a glycosyl phosphatidyl inositol (GPI) membrane anchor (Fig. 2). The N-linked glycosylation is important to yeast viability (Nagasu et al., 1992). The low molecular weight (MW) O-glycosylated proteins are extractable with sodium dodecyl sulphate (SDS), while the high MW N-glycosylated proteins are extractable by zymolyase or glucanase. A substantial amount of b-glucan is attached to the high MW mannoproteins, which suggests that such mannoproteins are secreted and then anchored to cell wall b-glucan. The attachment of aagglutinin, the sexual adhesion protein, is believed to be linked by O-glycans (Schreuder et al., 1993; Klis, 1994). The best characterized component of the cell wall is the a-mannan (Fig. 2). The molecule has an inner core of a-1,6-chain of mannose residues with a-1,2-, a-1,3linked short side chains. At the end of this inner core are two N-acetyl glucosamine residues (chitobiose) with the terminal residue attached to the side chain of an asparagine residue. At the other end of the inner core is an outer chain of 100±150 mannose residues linked by a1,6-backbone with a-1,2- and a-1,3-side chains, some of which contain phosphodiester linkages. This inner core and outer chain form N-glycosylated mannans. To the mannoprotein molecule, short a-1,2- and a-1,3-chains of mannose residues varying from 1 to 5 in length are attached via serine or threonine hydroxyl groups. The

425

glucomanno- side chains linked to unknown AA residues consist mainly of b-1,6-linked glucose residues and a-1,6-linked mannose residues. The C-terminal modi®cation of the mannoproteins is GPI-anchor-linked to the amino group of the AA residue at C-terminus. The GPI-anchor is the sole means of attachment of such proteins to the membrane. This GPI-anchor may exist in a modi®ed form lacking the inositol and phospholipid and function as a site linking to b-1,6-, b-1,3-glucans and chitin (Bacon, 1981; Hough et al., 1982a; Rose, 1993; Klis, 1994; Kollar et al., 1997). Proteins such as FLO (¯occulation) gene products are identi®ed as GPI-anchored, serine/threonine-rich wall proteins (Teunissen et al., 1993; Bidard et al., 1994). These proteins have hydrophobic C-terminals likely to be ®tted with GPI membrane anchors when secreted through the endoplasmic reticulum. It is also possible that O-glycosylation may serve to confer stability to proteins exposed to more hostile conditions outside the cell membrane (Stratford, 1994). Chitin is known to be an important constituent of bud scars with a major role of its synthesis in septum formation and cell division (Cabib et al., 1997). Cells lacking chitin were found to be resistant to Kluyveromyces lactis killer toxin, which suggested a role for chitin as the toxin receptor (Takita and Castilho-Valavicius, 1993).

Fig. 2. Glycosylation of cell wall proteins in Saccharomyces cerevisiae. The fragment in parentheses indicates the type of side chains rather than a repeating unit. n=10±15. A mannoprotein may not possess all four types of glycosylation (Bacon, 1981; Hough et al., 1982a; Rose, 1993; Klis, 1994; Kollar et al., 1997).

426

Y.-L. Jin, R. A. Speers

Extraction of yeast cells with hot alkali left an insoluble fraction consisting of b-1,3-, b-1,6-glucan and chitin. Since b-glucan is in itself soluble in alkali, this suggested that chitin may be responsible for the insolubility of this fraction (Klis, 1994). It has been shown that mutations a€ecting chitin cause osmotic sensitivity, abnormal morphology, aggregation and growth arrest with elongated buds (Takita and Castilho-Valavicius, 1993; Stratford, 1994). In spite of the very low (less than 1%) level of chitin in the S. cerevisiae cell wall, it seems to play an important role in wall structure. Cell wall proteins that are rich in glutamate and aspartate may act as either structural molecules or extracellular enzymes. Enzymes occupy the periplasmic space between the cell membrane and the cell wall. Both b-fructofuranosidase (invertase) and acid phosphatase have a large mannan component similar in structure to cell wall mannan (Bacon, 1981). Strains of lager yeast also possess extracellular melibiase (Hough et al., 1982a; ASBC, 1995). It has been reported that mannoprotein is attached to b-1,6-glucan through a remnant of a GPI anchor containing ®ve alpha-linked mannosyl residues. Beta-1,6glucan has some b-1,3-linked branches, and it is to these branches that the reducing termini of chitin chains appear to be attached via a b-1,4- or b-1,2-linkage. The reducing end of b-1,6-glucan is connected to the nonreducing terminal glucose of b-1,3-glucan through an as yet unknown linkage (Kollar et al., 1997). The cell wall structure is dynamic rather than static. It constantly changes its size and shape, to accompany growth of the cell. During cytokinesis, a specialized variant of the cell wall, the septum, is formed to separate the two dividing cells. Some researchers have suggested that lipids have certain structural roles to play in the cell wall. The lipid content of the wall varies both with yeast species and probably with the cultural conditions (Rogers, 1968). Presently little is known about the function of cell wall lipids. Functions The prime function of the cell wall is protection. Deprived of the wall, the cell would burst under the stress of osmotic pressure. Abnormal morphology is one of the reported mutational e€ects, which is often associated with yeast osmotic sensitivity. The wall acts as ®lter for large molecules, where the wall permeability determines passage of macromolecules both into and out from the cell. The wall also supports a number of external enzymes. These wall enzymes may be stabilized by phospholipids and wall exoglucanases may be glycosylated (Basco et al., 1993). However, cell wall is also responsible for sexual agglutination and ¯occulation of brewing yeast. Haploid a and strains form pointed morphologies and

adhere to each other via a- and a-agglutinins after exchange of pheromones. It is believed that cell surface zymolectins bind to carbohydrate receptors of the neighboring cell walls during ¯occulation of brewing yeast. Two types of zymolectins involved in the ¯occulation in S. cerevisiae have been described. Yeast cells of Flo1 phenotype possess a mannose-speci®c zymolectin (MS), while yeasts of NewFlo phenotype contain a zymolectin sensitive to mannose and glucose (GMS) (Stratford, 1992a,b; Stratford and Assinder, 1991; Bellal et al., 1995). Using protein±FITC conjugates, the zymolectin binding sites at the cell surface have been quanti®ed (Masy et al., 1992a; Ritcey, 1997; Patelakis et al., 1998). It is believed that zymolectin-mediated aggregation is the major factor in yeast ¯occulation (Speers et al., 1992a, 1993b). Those noncatalytic carbohydrate binding domains (CBDs) have been termed zymolectins and are de®ned as protein or glycoprotein structures associated with yeast cell wall speci®c CBDs which may cause or enhance ¯occulation of yeast cells (Patelakis et al., 1998; Speers, et al., 1999). GENETIC ASPECTS OF YEAST FLOCCULATION Since the 1950s, it has been recognized that ¯occulation is a hereditary characteristic (Gilliland, 1951; Thorne, 1951). Two dominant ¯occulation genes FLO1 and FLO2 and a recessive gene ¯o3 were ®rst identi®ed in the 1970s (Johnson and Lewis, 1974; Lewis et al., 1976). A dominant gene FLO4 was later mapped on chromosome I (Stewart et al., 1976; Stewart and Russell, 1977). Further tetrad analysis of genetic crosses demonstrated that FLO1, FLO2 and FLO4 were allelic, mapping onto chromosome I, and these genes were then consolidated into the FLO1 locus (Russell et al., 1980). The physical location of gene FLO1 was con®rmed to be 24 kb from the right end of chromosome I (Teunissen and Steensma, 1990; Teunissen et al., 1993). This gene confers ¯occulence when introduced into non-¯occulent yeast cells (Teunissen and Steensma, 1995). The FLO1 gene product (FLO1p) has been localized at the cell surface by immuno¯uorescent microscopy (Bidard et al., 1995). Further studies showed that the amount of FLO proteins in ¯occulent strains increased during yeast growth and the ¯occulation level was strongly correlated to the FLOp detected. The FLOp availability at the cell surface determined the ¯occulation degree of yeast. It is also believed that FLO proteins are polarly incorporated into the cell wall at the bud tip and the mother±daughter neck junction (Bony et al., 1998). It has been suggested that the FLO1p has putative GPIanchor at its C-terminal domain (Klis, 1994). Being rich in serine and/or threonine (Stratford, 1994) and with repeated domains of 70% of the sequence (Bidard et al., 1995), the GPI-anchored mannoproteins were suggested

Flocculation of Saccharomyces cerevisiae to extend 100±200 nm from plasma membrane in extended O-glycosylated form (Teunissen et al., 1993; Stratford, 1994, 1996a; Watari et al., 1994). One report has suggested that the FLO1p may function to activate a lectin rather than being a lectin itself (Stratford, 1994). The GPI-anchor may be modi®ed to serve as a site of linkage to b-1,6- and b-1,3-glucans and chitin (Caro et al., 1997; Kollar et al., 1997). This suggests that the FLO1p may be either modi®ed and GPI-anchored in the wall b-glucan network or GPI-anchored in the plasma membrane. The high glycosylation of FLO1p results in much higher apparent MW than predicted by the gene sequence (Bony et al., 1997). The hydrophobic C-terminus was found to be necessary for both the anchoring of FLO1p at the cell surface and the cell±cell interaction. The stabilization of FLO1p in the cell wall by its C-terminus suggested that the N-terminal region corresponds to the reacting domain at the cell surface. By expressing a truncated form of FLO1p with AAs 50 to 278 deleted, the N-terminal domain proved to be essential for the cell±cell interaction since the truncated form could not trigger a ¯occulent phenotype although the protein was detected in the cell wall (Bony et al., 1997). FLO5 was found to be a dominant gene not allelic to FLO1 and conferred strong ¯occulation (Johnson and Reader, 1983). It was mapped on chromosome VIII (Teunissen et al., 1995). FLO5p is also GPI-anchored into the cell wall at the bud tip and the mother-daughter neck junction (Bony et al., 1998). The FLO5 gene can trigger ¯occulation in a non-¯occulent strain (Bidard et al., 1994) and the FLO5p level of cell surface expression increases during growth and correlates with ¯occulation (Bony et al., 1998). Irreversible loss of ¯occulation conferred by FLO1 and FLO5 genes was achieved by treatment with pronase, proteinase K, trypsin or 2mercaptoethanol. However, the FLO1 strain was sensitive to chymotrypsin and stable to 70 C incubation whereas the FLO5 strain was thermolabile and chymotrypsin resistant (Hodgson et al., 1985). The FLO1 and FLO5 strains were therefore proposed to be di€erent phenotypes although only one strain of each was examined. Recessive genes ¯o6 and ¯o7 behaving in a semidominant manner have been considered as possible alleles of FLO1 (Johnson and Reader, 1983). The dominant gene FLO8 was mapped onto chromosome VIII (Yamashita and Fukui, 1983). It was later localized on chromosome I and was suggested to be allelic to FLO1 by genetic and physical mapping (Teunissen et al., 1995). However, this gene was further determined to encode a protein of 729 AA that has no pronounced hydrophobic regions. It was reported that the FLO8 gene has a signi®cant homogeneity with the S. cerevisiae chromosome V DNA sequence but no homology with the FLO1 gene. It was suggested that the FLO8 gene mediated ¯occulation by transcriptional activation of the FLO1 gene since the level of FLO1 gene transcription

427

was dependent on its transcription rate (Kobayashi et al., 1996). Another dominant gene FLO9 is not present in all strains (Teunissen and Steensma, 1995). The FLO9 product has a near similarity over 850 AAs to FLO1p and 74% identity over 970 AAs to FLO5p (Bossier et al., 1997). It also has strong similarity to FLO1p in the N-terminal region. Yet another FLO gene product, FLO10p has 58% similarity to FLO1p (Teunissen et al., 1995) and it is a member of the ¯occulin family including FLO1p, FLO5p and FLO9p (Bossier et al., 1997). FLO11 is another dominant gene related to the STA genes encoding secreted glucoamylase. The FLO11 gene was localized on chromosome IX. Yeast cells expressing FLO11 produce serine/threonine-rich and C-terminal GPI-anchored wall protein that has 26% identity to FLO1p and ¯occulate in a calcium-dependent manner while the null mutant cannot ¯occulate (Lo and Dranginis, 1996). The FLO11p is a `mucin-like protein' and is required for invasive growth of haploid cells and for diploid cells to form pseudohyphae in response to nitrogen starvation (Lo and Dranginis, 1998). The predicted properties of the FLO proteins are available on an internet database (Hodges et al., 1998; Proteome Inc., 1998) and summarized in Table 2. Semi-dominant genes fsu1 and fsu2 have been found to suppress the ¯occulence of FLO4 strains (Holmberg, 1978; Holmberg and Kielland-Brandt, 1978; Stewart and Russell, 1981). There are also mutations capable of causing yeast ¯occulation such as those of the TUP1 and CYC8 loci (Table 3). After distinguishing yeast strains by sugar and salt, acid inhibition, protease sensitivity and selective expression of ¯occulation, Stratford and Assinder (1991) suggested two phenotypes: Flo1 and NewFlo phenotype, mannospeci®c and gluco-/mannospeci®c respectively. The former phenotype includes all strains with known genes a€ecting ¯occulation whereas the latter phenotype comprises the majority of brewing ale strains. COLLOIDAL ASPECTS OF YEAST FLOCCULATION Yeast ¯occulation has been extensively studied in the past from a biochemical perspective, while its colloidal aspects have only recently interested researchers. Calleja (1984) outlined basic colloid theory and noted the importance of shear on the rate of ¯occulation. The in¯uence of agitation (shear) has been reported by other researchers (Kihn et al., 1988; Stratford et al., 1988; Stratford and Wilson, 1990; Stratford and Keenan, 1987, 1988; Stratford, 1989a; Speers, 1991; Speers et al., 1990; Speers et al., 1992b; Speers and Ritcey, 1995). Colloidal theory can help clarify our understanding of how yeast cells associate. While largely ignored in the

428

Y.-L. Jin, R. A. Speers Table 2. Comparison of predicted FLO gene products (Proteome Inc., 1998)

pI Mol. wt. Length N-term. Modif. C-term. Modif. Phosphorylation Glycosylation Identity to FLO1p

FLO1p

FLO5p

FLO9p

FLO10p

FLO11p

3.950 157 906 1513

4.030 109 355 1051

4.070 137 981 1322

4.050 119 556 1145

4.040 136 031 1367

GPI anchor unknown N- and O-linked

GPI anchor unknown O-linked 96%

unknown unknown O-linked 74% to FLO5p

GPI anchor unknown O-linked 58%

GPI anchor unknown unknown 26%

Table 3. Flocculence-related mutations of the S. cerevisiae genome (Stratford, 1992a) Mutation

Synonym

Selected characteristic

tup1 ¯k1 umr7 cyc9 pD7 s¯2 amm1 cyc8 ssn6

dTMP auxotrophy Maltase constitutive UV resistance Cytochrome c expression Abnormal mating Flocculation ARS1 stability Cytochrome c expression Invertase constitutive Envelope structure

tup1

cyc8 FH4C

between the particles. For yeast cells, the attractive potential energy, VA , due to van der Waals interactions can be estimated by the following equation: VA ˆ Af…h; T†r=12h

where Af…h; T† is the Hamaker function (10ÿ19ÿ10ÿ23 J), dependent on particle distance and temperature, but normally taken to be a constant, r is the particle radius, h is the distance between two particles. The contribution of electrostatic repulsion (VR ) to the interaction energy is: VR ˆ 2pee0 rc2 ln…1 ‡ eÿkh †

past, theories of the colloid science can help understand yeast ¯occulation. While DLVO (after Derjaguin, Landau, Verwey and Overbeek) theory has shown to be a poor predictor of the interaction energy of two approaching yeast cells, kinetic and rheological theories of ¯occulent yeast suspensions can provide an insight into the ¯occulation behavior of yeast cells (Speers, 1991; Speers et al., 1992b, 1993b). DLVO theory The classical theory of colloid stability, developed independently by Derjaguin and Landau, as well as Verwey and OverbeekÐnow called DLVO theoryÐdescribes the interaction between the charged particles immersed in a liquid medium, and assumes that the repulsive forces originate from the overlap of the electric double layers associated with the particles. Considerable experimental work on dispersions in aqueous media con®rm the essential validity of this approach, although there are observed experimental details which are not in full agreement with the theory. Few attempts to relate yeast suspension stability to particle (cell) charge have been made on cell dispersions. In general, it may be concluded that the DLVO theory provides an adequate explanation of coagulation data for non-biological systems containing low particle concentrations since the thickness of the double layer in media of low dielectric constant is larger compared with the average distance

…1†

…2†

where e is the dielectric constant of the medium, e0 is the permittivity of free space, is the surface potential (often taken to be the zeta potential x), k is the reciprocal of the double layer thickness. Changes to Hamaker and surface charge estimates can dramatically a€ect the attractive energies between the cells. The DLVO theory is useful for determining the qualitative importance of the van der Waals and electrostatic repulsion forces in yeast ¯occulation. While DLVO type forces may explain yeast ¯occulation at high surface charges (i.e. extreme pH ranges), it is believed that other processes of interaction such as zymolectin binding and hydrophobic interactions are more important in the cell±cell interactions (Speers et al., 1993a,b). Kinetic theory A second branch of colloid science which may be employed in the study of brewing yeast ¯occulation is that concerned with the rate at which particles collide and associate. There are essentially three mechanisms by which particles can associate: (1) by perikinetic aggregation due to Brownian motion, (2) by orthokinetic aggregation due to ¯uid ¯ow and (3) by ballistic aggregation arising from collisions during the settling of cells or ¯ocs. Due to the relatively large size of yeast cells, perikinetic aggregation is not important in brewing conditions. It was believed that orthokinetic and possibly ballistic aggregation are responsible for the majority of yeast cell±cell interactions (Speers, 1991).

Flocculation of Saccharomyces cerevisiae Laminar ¯ow Although orthokinetic aggregation can take place during either laminar or turbulent ¯ow, complete theories have only been developed for the aggregation of particles in laminar ¯ow ®elds. An expression describing ¯occulation rate of perfect spheres within a laminar shear ¯uid was ®rst developed by von Smoluchowski 80 years ago. It was later modi®ed by van de Van and Mason, (1977): _ 0 =†t Nt =N0 ˆ eÿ…4 0 '

…3†

where Nt is the number concentration of particles at time t; N0 is the initial number concentration of particles, a0 is the orthokinetic capture coecient, _ is the shear rate and j0 is the initial volume fraction of particles. This expression was said to hold for up to 80% reduction of particle numbers (Gregory, 1982). In eqn (3), the orthokinetic capture coecient is the most important term as it is directly proportional to the ¯occulation tendency of a yeast strain in a given environment. The value of a0 is determined by the forces acting on the cells as they approach one another in shearing ¯ow. Thus, a0 is a key parameter in orthokinetic ¯occulation. It has been demonstrated that if electrostatic forces dominate during the approach of cells in shearing ¯ow, the orthokinetic capture coecient would be a value near zero (<<0.0001). However, measured values of capture coecients were 0.0002, thus the binding mechanisms other than those described by the DLVO theory are important in brewing yeast ¯occulation (Speers et al., 1993a). Turbulent ¯ow Researchers did not control the rate of agitation prior to and during ¯occulation assays in the late 1980s when the e€ect of turbulent ¯ow on yeast ¯occulation was ®rst considered (Stratford and Keenan, 1987, 1988; Stratford et al., 1988; Stratford and Wilson, 1990). Flocculation rate of brewing yeasts subjected to laminar shear has been described (Speers, 1991; Speers et al., 1992b, 1993a). In the case of ¯occulation arising from turbulent ¯ow, a modi®cation may be made to the von Smoluchowski expression [eqn (3)] by including a term of an average shear rate (g_ ): 

Nt =N0 ˆ eÿ…4a0t g_ '0 =p†t

429

where d is the diameter of the particles and N is their number concentration. Equations (4)±(6) predict that the ¯occulation rate should vary according to the square root of the power input to the system and the square of the number of particles (yeast cell counts). It was found by experiment (Stratford and Keenan, 1987) that the initial ¯occulation rate was proportional to the square of cell concentration, but not equal to the square root of the power input to the system (the rotation rate of shake ¯asks) as predicted by eqn (6). This disagreement of theory with experiment may be caused by variation of the turbulent orthokinetic ¯occulation coecient with power input. Neither theory nor experiment has addressed the e€ect of mean shear rates on the turbulent orthokinetic ¯occulation coecient. Much research remains to be carried out on the e€ect of turbulent ¯ow on the ¯occulation of yeast cells. Rheology In the last few years, the rheological properties of yeast suspensions have been reported to be very sensitive to ¯oc structures which in turn are highly dependent on the attractive forces within the ¯oc (Speers, 1991; Speers et al., 1992b, 1993a; Speers and Ritcey, 1995). Yeast ¯oc structure is determined by a number of physicochemical and microbiological factors. Of all the factors (Fig. 3), it may be argued that the shear rate has the most important e€ect on suspension viscosity (Speers et al., 1992b). Recently an elastic ¯oc model has been described (Speers et al., 1993a). It predicts that ¯ow behavior of cell suspensions should follow the Bingham model above a critical shear rate (_gc ): s ˆ sy ‡ Z1 g_

…7†

where s is the shear stress, sy is the Bingham yield stress, and Z1 is the Bingham or in®nite shear viscosity.

…4†

where a0t is the orthokinetic ¯occulation constant in turbulent ¯ow and the average shear rate is a function of the power added to the system (P), the viscosity of the medium (Z) and the volume of the system (V):

_ ˆ …P=ZV†0:5

…5†

Equation (4) may also be presented as shown by O'Melia (1972): dNt =dt ˆ ÿ2a0t d3 …P=ZV†0:5 N2 =3

…6†

Fig. 3. Factors a€ecting apparent viscosity of yeast suspensions (Speers, 1991).

430

Y.-L. Jin, R. A. Speers

As well, the Bingham yield stress was predicted to be a function of the energy associated with the collision of ¯ocs (E) in a laminar shear ®eld when g_ > g_ c (Hunter, 1984; Speers et al., 1990; Speers, 1991): sy ˆ 3a0 jf E=p2 r3

…8†

where jf is the cell volume fraction and r is the cell radius. According to the elastic ¯oc model, the critical shear rate represents the minimum rate of shear at which doublet ¯ocs are separated in the shear ®eld. By observing the critical shear rate, the force required to separate the cells can be estimated on the basis of analysis of rotation of two connected spheres in a laminar ®eld: Fr ˆ Zr2 gc C

…9†

where Fr is the minimum value of the separation force, Z is the viscosity of the suspending medium, and C is a constant. This constant depends on the distance between the two spheres and has been calculated to be (A) 19.33 in the case of sphered red blood cells connected by a polymer bridge of 20 nm (Tha and Goldsmith, 1986) and (B) values ranging from 8.4 (Curtis and Hocking, 1970) to 38.45 (Goren, 1971) in the case of two spheres directly connected to one another. Calculation of the minimum separation force in an industrial strain led to a value of 2.210ÿ11 N which is roughly equal to the force of a lectin bond (Speers et al., 1993b). PHYSIOLOGICAL ASPECTS OF YEAST FLOCCULATION Inorganic ions Reports on the e€ect of ions particularly calcium and magnesium ions were confusing until it was demonstrated that calcium ions are directly required for yeast ¯occulation (Mill, 1964; Taylor and Orton, 1975; Nishihara et al., 1982; Stratford, 1989c). The ®rst report by Sey€ert (1896) stated a loss of yeast ¯occulation can be caused by using soft water while ¯occulation may be recovered by adding lime. There is an overwhelming consensus that calcium ions promote ¯occulation of dispersed ¯occulent yeast cells (Lindquist, 1953; Eddy, 1955a; Harris, 1959; Mill, 1964; Lyons and Hough, 1970; Lyons and Hough, 1971; Taylor and Orton, 1973; Taylor and Orton, 1975; Stewart and Goring, 1976; Miki et al., 1982a; Nishihara et al., 1982). Removal of calcium by such chelators as EDTA results in de¯occulation (Taylor and Orton, 1973; Beavan et al., 1979; Miki et al., 1982a). The inhibition of ¯occulation by EDTA can be overcome by adding more calcium ions at a trace level of 10ÿ8 M (Taylor and Orton, 1975).

Strontium and barium as calcium analogues have been reported to inhibit ¯occulation by competition (Taylor and Orton, 1973; Nishihara et al., 1982; Kuriyama et al., 1991). However, it has been reported that magnesium can induce ¯occulation indirectly by stimulating release of intracellular calcium ions (Stratford, 1989c). There are also reports on ¯occulation caused by addition of sodium or potassium salts (Stewart and Goring, 1976) and antagonization of calcium-induced ¯occulation by the addition of sodium ions (Mill, 1964; Nishihara et al., 1982), and the inhibition of magnesium-induced ¯occulation by sodium or potassium ions (Miki et al., 1982a). As noted by Stratford (1989c), sodium or potassium ions may cause calcium e‚ux from yeast cells and result in ¯occulation. However, at high sodium or potassium concentrations, ¯occulation is competitively inhibited whereas the magnesiuminduced ¯occulation is not competitively inhibited (Stratford, 1989c). EDTA or ethylene glycol-bis((-aminoethyl ether) N,N,N0 ,N0 -tetraacetic acid (EGTA) inhibition of ¯occulation can be overcome by excess calcium ions rather than excess magnesium or transition element ions (Stratford, 1989c). Calcium speci®city helps explain cell±cell interactions during yeast ¯occulation. The calcium-bridging hypothesis proposed that calcium ions link ¯occulent yeast cells by binding surface-carboxyl groups (Harris, 1959; Mill, 1964). However, this theory does not explain the action of mannose and mannose-like compounds in blocking ¯occulation. The lectin hypothesis developed by Miki et al. (1982a,b) suggests that zymolectins on the ¯occulent cell surfaces bind speci®cally to mannose residues of wall mannan on adjacent cells. The role of calcium ions has been interpreted as necessary to maintain the correct conformation of the zymolectins (Taylor and Orton, 1978; Miki et al., 1982a,b). Ethanol The e€ect of ethanol and other alcohols in promoting ¯occulation has been reported by some authors (Amory et al., 1988; Eddy, 1955c, 1958; Mill, 1964; Patel and Ingledew, 1975). Inhibition of ¯occulation by ethanol has also been observed (Kamada and Murata, 1984). Addition of ethanol did not in¯uence ¯occulation of stationary cells for bottom yeast but induced ¯occulation of stationary ale yeast (Dengis et al., 1995). A relationship was found between ¯occulation behavior and the dielectric constant of suspensions containing organic solvents at high concentrations and was explained by a decrease in ionization of salt bonds and an increase in the strength of hydrogen bonds (Mill, 1964) which implies an increase of hydrophobicity. Considering the yeast cell as a hydrophobic colloid, the e€ect of ethanol may be a result of its adsorption at the cell surface. It may cause a reduced local dielectric constant and a decreased cell±cell electrostatic repulsion, or a decreased

Flocculation of Saccharomyces cerevisiae steric stabilization by allowing the protrusion of mannoproteins for speci®c or non-speci®c binding. Organic solvents would then promote the cell±cell interactions (Dengis et al., 1995). pH Flocculation occurs towards the end of primary fermentation where the pH value of wort has fallen from 5.2 to around 4.5 to 4.0. The hydrogen-ion concentration was considered as an important factor promoting ¯occulation by early researchers as noted by Calleja (1987) and Stratford (1992c). Lower surface charge caused by increased hydrogen-ion concentration may have a role to play for the colloidal theory of ¯occulation. One would expect that lowered surface charge may make cell± cell contact easier and increase the rate of ¯occulation (Stratford, 1992c). According to Stratford and Assinder (1991), some ale strains ¯occulate in fermenting wort but show little or no ¯occulation in laboratory culture media like YEPD (yeast extract peptone dextrose) or YB (yeast protein base). The pH of YB medium falls from initial value of 4.5 rapidly to 2.3 during growth because of its very small bu€ering capacity. The YEPD medium with an initial pH 6.2 and a stronger bu€ering capacity generates a ®nal pH of about 5.3 which is not appropriate for some strains to ¯occulate (Stratford, 1996b). However, the pH value is generally considered as a minor factor under brewing conditions and yeast cells may ¯occulate anywhere between pH 2 and 8, varying with strains, with optimum values of pH 3±6 (Calleja, 1987), usually around 4.5 (Smit et al., 1992; van der Aar, 1996). Temperature In early studies it was found that ¯ocs were dispersed by heating (50±60 C) and such a `melting' of ¯ocs was readily reversible on cooling (Mill, 1964). The thermal de¯occulation occurred at approximately 54 C and the temperature at which ¯ocs dissociated (TF ) was de®ned as a measure of yeast ¯occulation (Taylor and Orton, 1975). For yeast Kluyveromyces lactis, it was found that ¯ocs remained stable at temperatures below 50 C and the thermolabile structures involved in the ¯occulation were attached to the cell surface rather than dissolved in the medium (Bellal et al., 1995). Such temperature range infers that hydrogen bonding is involved (Mill, 1964; Shankar and Umesh-Kumar, 1994). The fact that urea causes cell de¯occulation adds credence to the hydrogen bonding theory. It is worth noting that when yeast were sampled before the onset of ¯occulation and brought to a boil and cooled rapidly, ¯occulation was induced or activated (Stratford and Carter, 1993) although the mechanism is yet unknown. Studies on the ¯ow behavior of an ale yeast suspension have shown increased Bingham yield stress with

431

increased temperature. This indicates the hydrophobic nature of yeast ¯ocs, and therefore, an increase in cell± cell attraction (Speers et al., 1990). With a Photometric Dispersion Analyzer, it was found that ¯oc size and settling rate increased when medium temperature was increased from 3 to 15 C while complete dispersion of ¯ocs was achieved at 30 C. The lower dissociation temperature of ¯ocs was suggested as a result of the in¯uence of medium composition and strain variability (van Hamersveld et al., 1996). In fact, there are reports suggesting temperature sensitive strains with ¯occulation suppressed at temperatures as low as 25 C which may be caused by mutations (Stratford, 1992c). There is a report that the volume of settled yeast increased at 5 C with yeast cells reported as non-¯occulent at 20 C (GonzaÂlez et al., 1996). Generally however, there is little in¯uence of temperature on ¯occulation of brewing yeast at physiological temperatures of 15±32 C (Stewart et al., 1975a). Dissolved oxygen Oxygen consumption occurs during the very early stages of beer fermentation. The consumed molecular oxygen is used for the production of unsaturated fatty acids and sterols, which are essential constituents of yeast cell membrane. When oxygen is insucient for membrane synthesis, yeast cells would fail to grow. It was suggested that shortage of sterols and unsaturated fatty acids precedes ¯occulence under brewing conditions (Straver et al., 1993b). There are reports on both promoting and inhibitory e€ect of aeration on ¯occulation as cited by Calleja (1987). Insucient aeration may hasten the onset of ¯occulation (Gilliland, 1951). In an aerobic culture, ¯occulence was found to be constant (Miki et al., 1982b). The ¯occulating ability of yeast cells during fermentation in wort was triggered after growth limitation by oxygen shortage and coincided with a sharp increase in CSH of the cells. Presence of oxygen in the pitched wort in¯uenced ®nal cell number, ¯occulence and CSH (Straver et al., 1993b). It was reported, for ethanol fermentation with a ¯occulation airlift bioreactor, that ¯oc size changed with di€erent aeration rates (Sousa et al., 1994). Protein and protein denaturation Protein synthesis is required for development of yeast ¯occulation (Baker and Kirsop, 1972; Lands and Gra€, 1981; Stratford and Carter, 1993). Irreversible loss of ¯occulation caused by papain treatment of yeast ¯ocs suggested involvement of cell-surface proteins in ¯occulation (Eddy and Rudin, 1958). Besides papain, pronase E, proteinase K, trypsin, chymotrypsin and pepsin and protein modi®ers like mercaptoethanol, urea and guanidine can cause irreversible de¯occulation (Miki et al.,

432

Y.-L. Jin, R. A. Speers

1982a; Nishihara et al., 1982; Stratford and Assinder, 1991; Stratford, 1992c). Other degradative enzymes like lipase, RNAse (ribonuclease), DNAse (deoxyribonuclease) and lysozyme showed no e€ect on ¯occulation (Nishihara et al., 1982). Di€erences in chymotrypsin sensitivity and heat stability between two strains of FLO1 and FLO5 genes suggested di€erent cell wall proteins may be responsible for ¯occulation (Hodgson et al., 1985). Scanning electron micrographs have indicated an existence of mucus-like attachment among adhering cells (Simpson and Hammond, 1989; Speers et al., 1993a). Pronase E treated cells exhibited reduced cell±cell contact and less of this mucus-like covering (Speers et al., 1993a). Sugar inhibition The inhibition of yeast ¯occulation by sugars was ®rst observed by Lindner (1901), and later con®rmed by Burns (1937), Lindquist (1953), Eddy (1955b) and Mill (1964). It was found that maltose and mannose were most e€ective inhibitors whereas sucrose and glucose were less e€ective. Sugars like galactose and fructose were ine€ective. Flocculation can be inhibited speci®cally by mannose and derivatives (Taylor and Orton, 1978; Miki et al., 1982a) such as a-methyl-d-mannopyranoside (a-MM) at 25 mM (Smit et al., 1992). Using sugars and derivatives with a hydrophobic substituent, the ¯occulation of stationary bottom yeast cells was inhibited by 500 mM glucose and mannose, 10 mM phenyl-a-d-mannoside and p-nitrophenyl-a-d-mannoside, although no inhibition was found for a top yeast strain studied (Dengis et al., 1995). The higher speci®city of the (-linked aromatic glycosides of phenyl-a-d-mannoside and p-nitrophenyla-d-mannoside suggests hydrophobic sugar binding sites on the cell surface although it was not reported by the authors (Dengis et al., 1995). Two distinct groups of strains have been classi®ed according to sugar speci®city: the Flo1 phenotype which is mannose sensitive only and the NewFlo phenotype which is sensitive to glucose, maltose in addition to mannose (Stratford, 1989b; Stratford and Assinder, 1991). However, some researchers also classi®ed ¯occulent yeast strains into three groups as MS (mannose sensitive), GMS (glucose±mannose sensitive) and MI (mannose insensitive) (Masy et al., 1992a,b). Although metabolizable by yeast, these sugars discussed above have been proven to a€ect ¯occulation directly. Yeast suspensions metabolically controlled by low temperature (4 C) or heat (60 C, 3 min) still showed sugar inhibition of ¯occulation (Stratford, 1989b; Stratford and Assinder, 1991). The inhibition of ¯occulation by sugars, ignored by the calcium-bridging hypothesis, can be easily explained by the lectin hypothesis. Indeed, sugar inhibition is key evidence for lectin-like theory of ¯occulation.

Cell surface properties There is evidence of an increase of the water contact angle at the beginning of ¯occulation. It was demonstrated that a relation existed between cell division arrest, the increase of CSH and initiation of ¯occulence during fermentation (Straver et al., 1993b). A high level of CSH may facilitate cell±cell contact in an aqueous medium resulting in more speci®c lectin±carbohydrate interactions (Straver and Kijne, 1996). However, it was recently reported that no signi®cant di€erences in hydrophobicity or surface concentration of proteins, polysaccharides or hydrocarbons were found between ¯occulating and non-¯occulating cells at stationary and exponential phases of growth, respectively (Dengis et al., 1995; Dengis and Rouxhet, 1997). It has been demonstrated with hydrophobic interaction chromatography that CSH is strongly correlated with ¯occulence of brewing yeast (Akiyama-Jibiki et al., 1997). Previous work revealed that loss of CSH coincided with a signi®cant increase in ¯occulation ability of the cells during incubation before ¯occulation assay (Straver et al., 1994a). The appearance and loss of CSH was also correlated with appearance or loss of ®mbriaelike structures on the cell surface (Straver et al., 1994a). Mannose present during incubation inhibited the loss of these factors. Our experiments have con®rmed that hydrophobic probe anilinonaphthalene sulphonate binding was inhibited by a-methyl-d-mannopyranoside (unpublished data) which suggested that the mannose binding site itself is hydrophobic. To date, there is no signi®cant correlation between cell surface charge and ¯occulation of yeast (Stewart et al., 1976; Speers et al., 1993a; Dengis and Rouxhet, 1997; Ritcey, 1997). However, it has been suggested that surface charge and the non-separation of progeny from mother cells rather than surface hydrophobicity in¯uences the chain-formation of brewing yeast (Wilcocks and Smart, 1995). Cell age It was noticed by early workers that ¯occulation developed towards the late stationary phase of growth near the end of the primary fermentation (Helm et al., 1953). Stationary cells were found to ¯occulate in media separated from exponential and stationary stages but not in fresh medium. The stationary top yeast cells ¯occulated in the fresh medium with added ethanol (more than 3% v/v) which suggested no inhibition by sugar in the fresh medium for the strain studied. The exponential cells did not ¯occulate in fresh medium nor media separated from exponential and stationary phases supplemented with ethanol up to 10% (v/v) (Dengis et al., 1995). Comparing the cell surface properties at di€erent physiological stages, the mobility curves obtained in 1 mM KNO3 were identical for the exponential and stationary cells (Dengis et al., 1995).

Flocculation of Saccharomyces cerevisiae In a recent report, the ¯occulation of a Flo1 phenotype strain was not a€ected by the physiological stage of cells throughout the growth (Patelakis et al., 1998) whereas a NewFlo phenotype strain possessed cyclic ¯occulation ability. `NewFlo' cells progressively lost their ¯occulation ability in the early period of growth and recovered it towards the end of exponential phase of growth (Soares and Mota, 1996). A morphological study found that the progressive crenellation and wrinkling of the cell wall during aging may increase the potential surface area of contact compared with that of smooth younger cells and therefore promote cell±cell adhesion (Barker and Smart, 1996). It is also reported that new daughter cells have lower ¯occulation ability than mother cells (Bielecki and Brzeski, 1989). However, the increase of ¯occulation towards the end of fermentation can hardly be explained only by the increase of mother cells since virgin daughter cells in a normal population is close to 50% of the total cells (Deans et al., 1997). In NewFlo strains, zymolectins were proved to be synthesized continuously from an early stage of growth and rapidly inserted into the cell wall remaining inactive before being activated at ¯occulation onset by an unknown mechanism (Stratford and Carter, 1993). Nutrient limitation or starvation resulted in a reduction in subsequent ¯occulation of lager and ale yeasts (Rhymes and Smart, 1996). Physical properties of yeast cell wall were in¯uenced by starvation. Wall thickness and density of surface material were decreased. For lager yeast strain, cell wall surface phosphate content reduced after starvation and the cell surface charge was decreased. However, the CSH as well as the content of mannoprotein and b-glucan were not a€ected. For an ale strain, increased cell surface charge and decreased CSH were observed in starved cells. The starvation-induced reduction in ¯occulation was retained by subsequent generations (Rhymes and Smart, 1996). When repitched in subsequent fermentation, both ale and lager strains showed increased ¯occulation (Teixeira et al., 1991; Smart and Whisker, 1996). The ale strain exhibited a reduction in ¯occulation after 25 serial repitchings and the extended serial repitching signi®cantly a€ected cropped yeast ¯occulation due to the physiological stress (Smart and Whisker, 1996). Flocculation onset To further understand the process better, some researchers have been searching for the change (or changes) in yeast cell characteristics at the moment of ¯occulation onset. It has been demonstrated that the receptors for zymolectin binding are available throughout growth of all yeast strains studied (Stratford, 1993). The lectin itself is synthesized and inserted into cell wall at an early stage of growth and is believed to be activated later during fermentation (Stratford and Carter, 1993). Also, ¯occulation onset coincides with budding arrest (Straver

433

et al., 1993b) and glucose limitation (Straver et al., 1993b; Soares and Mota, 1996). There is an increase of CSH at the time of ¯occulation onset (Straver et al., 1993b) and it can be concluded from reports of Straver et al. (1993b) and Dengis et al. (1995) that there exists a hydrophobic region in the sugar binding site. The increase in hydrophobicity may be a sign of zymolectin activation and/or exposure of its hydrophobic binding site. MECHANISM OF YEAST FLOCCULATION A summary of main features of yeast ¯occulation Flocculation is a genetically controlled inducible characteristic of S. cerevisiae. FLO gene-disrupted yeast cells fail to ¯occulate (Kobayashi et al., 1995; Lo and Dranginis, 1996) and insertion of a FLO gene can trigger ¯occulation in a non-¯occulent strain (Bidard et al., 1994; Kobayashi et al., 1995). Binding of zymolectins of ¯occulent cell walls to sugar residues of adjacent cell walls results in ¯occulation (Stratford, 1992a) and the availability of cell surface FLOp in¯uences the ¯occulation level (Bony et al., 1998). Flocs can consist of up to thousands of cells (Calleja, 1984) and cell ¯occulation is based on wall-wall interactions (Eddy, 1955c; Masschelein and Devreux, 1957). As the cells are not a€ected by Brownian motion, agitation is essential for ¯occulation (Calleja, 1984; Stratford and Keenan, 1987; Kihn et al., 1988; Stratford et al., 1988; Stratford and Wilson, 1990; Speers, 1991; Speers et al., 1992b; Stratford, 1992c; Speers and Ritcey, 1995). The ¯occulating ability of brewing yeast cells is regained at the late exponential or early stationary phase of growth (Helm et al., 1953; Dengis et al., 1995). In other words, cells of non-¯occulent strains are always non-¯occulating while those ¯occulent ones may not always be ¯occulent. However, it is noteworthy that constant ¯occulation of a constitutive haploid Flo1 phenotype strain has been observed (Soares and Mota, 1996; Patalakis et al., 1998). Yeast ¯occulation is calcium-dependent (Lindquist, 1953; Eddy, 1955a; Harris, 1959; Mill, 1964; Lyons and Hough, 1970; Lyons and Hough, 1971; Taylor and Orton, 1973, 1975; Stewart and Goring, 1976; Miki et al., 1982a; Nishihara et al., 1982; Stratford, 1989c) and sugar speci®c (Taylor and Orton, 1978; Miki et al., 1982a; Stratford, 1989b; Stratford and Assinder, 1991; Masy et al., 1992a,b; Smit et al., 1992; Dengis et al., 1995). It is also known that cell surface proteins are involved in the ¯occulation process (Eddy and Rudin, 1958; Baker and Kirsop, 1972; Lands and Gra€, 1981; Miki et al., 1982a; Nishihara et al., 1982; Stratford and Assinder, 1991; Stratford, 1992c; Speers et al., 1993a; Stratford and Carter, 1993). Hydrogen bonding and hydrophobic interactions are also involved in cell±cell ¯occulation (Mill, 1964; Taylor and Orton, 1978; Speers et

434

Y.-L. Jin, R. A. Speers

al., 1990; Shankar and Umesh-Kumar, 1994; Stratford, 1996a). As well, cells of some non-¯occulent strains may mutually ¯occulate or co-¯occulate with ¯occulent cells (Eddy, 1958; Curtis and Wenham, 1958; Miki et al., 1982a). Forces involved in cell±cell interactions Repulsive forces The presence of negative charges on the yeast cell surfaces at physiological pH values has been established for almost ®ve decades (Jansen and Mendlik, 1951). Both carboxyl and phosphodiester groups may be involved in cell-surface anionic generation (Rose, 1984) although only carboxyl groups were thought to mediate calcium bridging in ¯occulation (Beavan et al., 1979). Electrostatic repulsion may keep the cell surfaces at a distance of about 10 nm from one another (Speers, 1991; Dengis et al., 1995). Such repulsion keeps cells dispersed before they become ¯occulent and acts as one of the barriers to ¯occulation. Repulsion may also arise from steric hindrance and the need to displace ionic atmospheres or electric double layers (Stratford, 1992c; Dengis et al., 1995). By observing the falling of yeast cells which are too large to be in colloidal suspension, it was identi®ed that the movement of the cells did not show Brownian motion in spite that molecules on the cell surfaces may possess Brownian-like vibration (Stratford, 1992c). However, the di€usion of counter-ions in the electric double layers may be caused by Brownian motion or thermal randomization (Calleja, 1994). Attractive forces It has been suggested that yeast aggregates are bonded together by van der Waals forces (Stratford, 1996a). In the absence of any other type of attractive forces, van der Waals forces are important for yeast cells to overcome electrostatic repulsive forces to allow cell ¯occulation. Hydrogen bonding is also involved in cell±cell interactions as evidenced by the reversible de¯occulation by heat, urea, guanidinium chloride, and SDS (Mill, 1964; Taylor and Orton, 1978; Shankar and Umesh-Kumar, 1994). Some researchers have emphasized the importance of CSH in yeast ¯occulation (Iimura et al., 1980; Amory et al., 1988; Speers et al., 1990; Smit et al., 1992; Straver et al., 1993a,b; Azeredo et al., 1997; AkiyamaJibiki et al., 1997). A large hydrophobic e€ect indicates that the attraction of cells to water molecules is lower than that among cells to cells. It has been proposed that hydrophobic interactions consist of the long-range van der Waals attractions and the short-range interactions, particularly hydrogen bonds (van Oss and Giese, 1995). Hypotheses of yeast ¯occulation The calcium-bridging hypothesis The calcium-bridging hypothesis was a favored theory until the early 1980s to explain the cell±cell interactions

during ¯oc formation. It was outlined by Harris (1959) and proposed by Mill (1964). It was supposed that calcium ions form bridges between ¯occulating cells by binding to negative charges on the cell surface. Such bridges may be stabilized by hydrogen bonding between carbohydrate hydrogen atoms and hydroxyl groups as the thermal dissociation of ¯ocs (50±60 C) suggested the involvement of hydrogen bonds in structure maintenance. There were several facts supporting this hypothesis. Flocculation can be inhibited irreversibly by 1,2-epoxypropane as a carboxyl esteri®er (Mill, 1964; Jayatissa and Rose, 1976). The density of carboxyl groups on cell surfaces was correlated to ¯occulation (Beavan et al., 1979). De¯occulation by proteolysis or protein denaturation also suggested that carboxyl groups associated with cell wall mannoproteins may be involved in this mechanism (Eddy and Rudin, 1958; Stewart et al., 1973; Nishihara et al., 1977, 1982). Phosphoester groups in wall phosphomannan were suggested to be alternative binding sites for calcium ions (Lyons and Hough, 1970, 1971) and an increase of ¯occulation was found after excision of wall phosphodiester bonds (Jayatissa and Rose, 1976). No correlation was found between wall phosphate content and ¯occulation onset (Jayatissa and Rose, 1976; Beavan et al., 1979). However, a decline in phosphate content was later reported (Amory et al., 1988). The ®nding of sugar inhibition was not easily accommodated in this theory (Rose, 1993). The theory also ignored the phenomena of mutual ¯occulation and co¯occulation and the presence of mannoprotein carboxyl groups on the non-¯occulent cell surfaces. This theory has received extensive criticism since it fails to explain the speci®city of cell±cell interactions (Calleja, 1987; Speers et al., 1992a). It is also unlikely that the microamount of calcium ions needed for ¯occulation would allow divalent bridging in solutions (Stratford and Assinder, 1991; Stratford, 1992c). The lectin hypothesis Since the early 1980s, the lectin hypothesis has become a more convincing and more putative mechanism of yeast ¯occulation. It proposes that speci®c surface proteins known as zymolectins present on ¯occulent yeast cells bind to mannose residues of mannan molecules on neighboring cell surfaces. Calcium ions are believed to maintain a correct conformation of the zymolectin binding site. The involvement of such a protein-carbohydrate interaction was suggested as ¯occulation can be inhibited speci®cally by mannose (Taylor and Orton, 1978). It was later mentioned in Malting and Brewing Science (Hough et al., 1982a) and recognized as `lectinlike' hypothesis as shown in Fig. 4 (Miki et al., 1982a,b). The lectin hypothesis was supported by involvement of cell surface proteins in yeast ¯occulation. Further evidence supporting this theory was the mutual ¯occulation and co-¯occulation of non-¯occulent cells that

Flocculation of Saccharomyces cerevisiae

Fig. 4. Lectin hypothesis of yeast ¯occulation. FLO1 cells possess both binding sites and their receptors while ¯o1 cells only have receptors on the surface (Miki et al., 1982a).

suggested two distinct parts be involved in cell±cell interactions. As discussed earlier, ¯occulent yeast strains have been classi®ed into Flo1 and NewFlo phenotypes or alternatively MS, GMS and MI strains. The existence of Flo1 and NewFlo phenotypes were explained by di€erent sugar speci®cities indicating two distinct zymolectin mechanisms (Stratford, 1992c). The mannose-insensitivity (MI) is probably resulted from very low speci®city to monosaccharide since lectins may have much greater anity to tri- or polysaccharides than for simple sugars (Stratford, 1992c). The zymolectin anity for mannose would be MS>GMS>MI while that for other polysaccharides may be MS
435

identi®ed in early work showing its accumulation in the wall of ¯occulent cells and absence in non-¯occulent cells (Stewart and Goring, 1976). A 13 kDa peptide was also isolated from the alkaline extract of FLO4 (now FLO1) yeast cell wall (Holmberg, 1978). As reviewed by Moradas-Ferrira et al. (1994), a number of protein bands on SDS electrophoresis were reported to be 13±67 kDa in size. A 300 kDa agglutinin termed `¯occulin' was isolated from cell walls of both ¯occulent and non-¯occulent brewing yeasts (Straver et al., 1994b,c). This ¯occulin stimulated ¯occulation of ¯occulent cells without showing lectin-like activity. A 13 kDa zymolectin was shown to bind to mannose as well as yeast mannan immobilized in the wells of microtitre plates in a calciumdependent manner (Shankar and Umesh-Kumar, 1994). A single 65 kDa protein from gel ®ltration generated a single peak at 13 kDa following 4 M urea treatment. This 13 kDa fraction showed spontaneous aggregation during SDS-PAGE. When the protein was eluted from Con A-Sepharose with mannose, the sugar bound tightly to the eluted protein (Shankar and UmeshKumar, 1994). Implied by the self-aggregation and binding to both mannan and Con A-Sepharose, the mannoprotein may possess both zymolectin and its receptor. As the aggregate (65 kDa) contains ®ve 13 kDa units, we believe the possibility that one binding site and one receptor domain exists on each 13 kDa zymolectin molecule. Using immobilized yeast cells in polyacrylamide gel, three proteins of 20, 36 and 66 kDa in size were isolated by the anity chromatography (Stewart et al., 1995). Cells lost their ability to ¯occulate following extraction of these wall proteins while the extract inhibited ¯occulation of ¯occulent yeast cells (Stewart et al., 1995). This indicates neither coexistence of receptor on zymolectin molecule nor di-/multi-valence of the zymolectin. All the previously reported proteins could not be complete FLO products which have a predicted MW higher than 100 kDa. A recent report revealed that hot SDS-extracted FLO1p is larger than 200 kDa due to glycosylation (Bony et al., 1997). The N-terminal region of FLO1p was indirectly demonstrated to contain the sugar recognition domain (Kobayashi et al., 1997) although there is no in vitro evidence for its lectin-like activity (Bony et al., 1997). The contradictories arise from mainly di€erent strains and extraction conditions used by di€erent researchers. To date, there is still scarce information on the characteristics of zymolectins and their genetic basis.

Detection of surface proteins mediated in ¯occulation

CONCLUDING REMARKS

Many researchers have searched for surface proteins or zymolectins from ¯occulent cells that are absent on non¯occulent cell surfaces. A speci®c peptide of 37 kDa was

As so many papers have been published on the subject of yeast ¯occulation, it is not impossible that some of them escaped our attention. Generally speaking, much

436

Y.-L. Jin, R. A. Speers

progress has been made in understanding the nature of yeast ¯occulation in the last 20 years. Scientists now accept the lectin hypothesis since it helps interpret the known features of ¯occulation. While seeking for better understanding of the zymolectin molecules, some researchers have explored quanti®cation of the zymolectin density on the ¯occulent cell surface (Masy et al., 1992a; Ritcey, 1997; Patelakis et al., 1998). However, little attention has been paid to the e€ect of environmental factors on the e€ective density of zymolectin nor the possible correlation between zymolectin density and the ¯occulating ability of the yeast cells. As a second causative agent of ¯occulation, hydrophobicity has attracted research due to its contribution to ¯occulation onset and its in¯uence on yeast ¯occulence (Iimura et al., 1980; Amory et al., 1988; Speers et al., 1990, 1992a; Smit et al., 1992; Straver et al., 1993a,b; Akiyama-Jibiki et al., 1997; Azeredo et al., 1997). Again, there is little known about the in¯uence of environment conditions on the CSH and the relationship between CSH and ¯occulation of yeast cells. REFERENCES Akiyama-Jibiki, M., Ishibiki, T., Yamashita, H. and Eto, M. (1997) A rapid and simple assay to measure ¯occulation in brewer's yeast. MBAA Tech. Quart. 34, 278±281. Amory, D. E., Rouxhet, P. G. and Dufour, J. P. (1988) Flocculation of brewery yeast and their surface properties: chemical composition, electrostatic charge andhydrophobicity. J. Inst. Brew. 84, 79±84. ASBC (1986) Report of the subcommittee on measurement of yeast ¯occulation. J. Am. Soc. Brew. Chem. 44, 133±134. ASBC (1993) Yeast ¯occulation determination. J. Am. Soc. Brew. Chem. 51, 188±190. ASBC (1994) Yeast ¯occulation determination by the Helm assay. J. Am. Soc. Brew. Chem. 52, 188±191. ASBC (1995) Report of the subcommittee on di€erentiation of ale and lager yeast by melibiose. J. Am. Soc. Brew. Chem. 53, 219±222. ASBC (1996) Yeast ¯occulation by absorbance method. J. Am. Soc. Brew. Chem. 54, 245±248. Azeredo, J., Ramos, I., Oliveira, R. and Teixeira, J. (1997) Yeast ¯occulation: a new method for characterizing cell surface interactions. J. Inst. Brew. 103, 359±361. Bacon, J. S. D. (1981) Nature and disposition of polysaccharides within the cell envelope. In Yeast Cell Envelopes: Biochemistry, Biophysics, and Ultrastructure., ed. W. N. Arnold, Vol. I, pp. 68±83. CRC Press, Boca Raton, FL. Baker, D. A. and Kirsop, B. H. (1972) Flocculation in Saccharomyces cerevisiae as in¯uenced by wort composition and by actidione. J. Inst. Brew. 78, 454±458. Barker, M. G. and Smart, K. A. (1996) Morphological changes associated with the cellular aging of a brewing yeast strain. J. Am. Soc. Brew. Chem. 54, 121±126. Basco, R. D., Munoz, D., Hernandez, L. M., Vazquez, de Aldana, C. and Larriba, G. (1993) Reduced eciency in the glycosylation of the ®rst sequon of Saccharomyces cerevisiae exoglucanase leads to the synthesis of a new glycoform of the molecule. Yeast 9, 221±234. Beavan, M. J., Belk, D. M., Stewart, G. G. and Rose, A. H. (1979) Changes in electrophoretic mobility and lytic enzyme

activity associated with development of ¯occulating ability in Saccharomyces cerevisiae. Can. J. Microbiol. 25, 888±895. Bellal, M., Boudrant, J., Elfoul, L. and Bonaly, R. (1995) Flocculation dispersion in Kluyveromyces lactis. Proc. Biochem 30, 641±648. Bendiak, D. S. (1994) Quanti®cation of the Helm's ¯occulation test. J. Am. Soc. Brew. Chem 52, 120±122. Bidard, F., Blondin, B., Dequin, S., Vezinhet, F. and Barre, P. (1994) Cloning and analysis of a FLO5 ¯occulation gene from Saccharomyces cerevisiae yeast. Curr. Gent. 25, 196± 201. Bidard, F., Bony, M., Blondin, B., Dequin, S. and Barre, P. (1995) The Saccharomyces cerevisiae FLO1 ¯occulation gene encodes for a cell surface protein. Yeast 11, 809±822. Bielecki, S. and Brzeski, H. (1989) Characterization of non¯occulent cells isolated from a culture of ¯occulent Saccharomyces cerevisiae NCYC1001. FEMS Microbiol. Lett. 61, 189±194. Bony, M., Thines-Sempoux, D., Barre, P. and Blondin, B. (1997) Localization and cell surface anchoring of the Saccharomyces cerevisiae ¯occulation protein Flo1p. J. Bacteriol 179, 4929±4936. Bony, M., Barre, P. and Blondin, B. (1988) Distribution of the ¯occulation protein, Flop, at the cell surface during yeast growth: the availability of Flop determines the ¯occulation level. Yeast 14, 25±35. Bossier, P., Goethals, P. and Rodrigues-Pousada, C. (1997) Constitutive ¯occulation in Saccharomyces cerevisiae through overexpression of the GTS1 gene, coding for a `Glo'-type Zn-®nger-containing protein. Yeast 13, 717±725. Burns, J. A. (1937) J. Inst. Brew., 43, 31 [cited by Jansen, H. E. (1958) Flocculation of Yeast. In The Chemistry and Biology of Yeasts, ed. A. H. Cook, pp. 635±667. Academic Press, New York]. Cabib, E., Drgon, T., Drgonova, J., Ford, R. A. and Kollar, R. (1997) The yeast cell wall, a dynamic structure engaged in growth and morphogenesis. Biochemical Society Transactions 25, 200±203. Calleja, G. B. (1984) Microbial Aggregation. CRC Press, Boca Raton, FL. Calleja, G. B. (1987) Cell Aggregation. In The Yeast, ed. A. H. Rose and J. S. Harrison, Vol. 2, 2nd edn. pp. 165±238. Academic Press, London. Calleja, G. B. (1994) Hooks, loops, and the shallow minimum: mechanistic aspects of yeast ¯occulation. Colloid. Surf. B: Biointerfaces 2, 133±149. Calleja, G. B., Atkinson, B., Reichenbach, H., Garrod, D. R., Rose, A. H., Hirsch, P., Schink, B., Jenkins, D., Vincent, B., Johnson, B. F. and Wilderer, P. A. (1984) Aggregation. In Microbial Adhesion and Aggregation, ed. K. C. Marshall pp. 303-21. Springer±Verlag, New York. Calleja, G. B. and Johnson, B. F. (1977) A comparison of quantitative methods for measuring yeast ¯occulation. Can. J. Microbiol 23, 68±74. Caro, L. H. P., Tettelin, H., Vossen, J. H., Ram, A. F. J., VandenEnde, H. and and Klis, F. M. (1997) In silicio identi®cation of glucosul-phosphatidylinositol-anchored plasmamembrane and cell wall proteins of Saccharomyces cerevisiae. Yeast 13, 1477±1489. Curtis, A. S. G. and Hocking, L. M. (1970) Collision eciency of equal spherical particles in a shear ¯ow. The in¯uence of London-van der Waals forces. Trans. Faraday Soc. 66, 1381±1390. Curtis, N. S. and Wenham, S. (1958) Some observations on the declining attenuative power of brewery yeast. J. Inst. Brew. 64, 421±428. Deans, K., Pinder, A., Catley, B. J. and Hodgson, J. A. (1997) E€ects of cone cropping and serial re-pitch on the distribution

Flocculation of Saccharomyces cerevisiae of cell ages in brewery yeast. Proc. Eur. Brew. Conv. Cong., Maastricht, The Netherlands, pp. 469±476. Dengis, P. B., Nelissen, L. R. and Rouxhet, P. G. (1995) Mechanisms of yeast ¯occulation: Comparison of top- and bottom-fermenting strains. Appl. Environ. Microbiol 61, 718±728. Dengis, P. B. and Rouxhet, P. G. (1997) Flocculation mechanisms of top and bottom fermenting brewing yeast. J. Inst. Brew. 103, 257±261. D'Hautcourt, O. and Smart, K. A. (1998) The measurement of brewing yeast ¯occulation. Paper P-14, Presented at the 64th ASBC Annual Meeting, Boston, MA. EBC Microbiologica (1981) Method 2.5.2. Flocculation. J. Inst. Brew. 87, 319±320. Eddy, A. A. (1955a) Flocculation characteristics of yeasts. I. Comparative survey of various strains of Saccharomyces cerevisiae. J. Inst. Brew. 61, 307±312. Eddy, A. A. (1955b) Flocculation characteristics of yeasts. II. Sugars as dispersing agents. J. Inst. Brew 61, 313±317. Eddy, A. A. (1955c) Flocculation characteristics of yeasts. III. General role of ¯occulating agents and special characteristics of a yeast ¯occulated by alcohol. J. Inst. Brew. 61, 318±320. Eddy, A. A. (1958) Composite nature of the ¯occulation process of top and bottom strains of Saccharomyces. J. Inst. Brew. 64, 143±151. Eddy, A. A. and Rudin, A. D. (1958) Part of the yeast surface involved in ¯occulation. J. Inst. Brew 64, 19±21. Enari, T.-M. (1995) One Hundred Years of Brewing Research. J. Inst. Berw., Centennial edn. The Dorset Press, Dorchester, UK. Garsoux, G., Haubursin, H., Bilbault, S. and Dufour, J. P. (1993) Yeast ¯occulation: biochemical characterization of yeast cell wall components. Proc. Eur. Brew. Conv. Prog., Oslo, pp. 275±282. Gilliland, R. B. (1951) The ¯occulation characteristics of brewing yeast during fermentation. Proc. Eur. Brew. Conv. Cong., Brighton, UK, pp. 35±58. GonzaÂlez, M. G., Fernandez, S. and Sierra, J. A. (1996) E€ect of temperature in the evaluation of yeast ¯occulation ability by the Helm's method. J. Am. Soc. Brew. Chem 54, 29±31. Goren, S. L. (1971) The hydrodynamic forces on touching spheres along the line of centers exerted by a shear ®eld. J. Colloid Interface Sci 36, 94±96. Gregory, J. (1982) Particle interactions in ¯owing suspensions. Adv. Colloid Interface Sci. 17, 149±160. Hammond, J. R. M. (1995) Genetically-modi®ed brewing yeasts for the 21st century. Progress to date. Yeast 11, 1613± 1627. Harris, J. O. (1959) Possible mechanism of yeast ¯occulation. J. Inst. Brew. 65, 5±11. Helm, N., Nohr, B. and Thorne, R. S. W. (1953) The measurement of yeast ¯occulence and its signi®cance in brewing. Wallerstein Lab. Commun 16, 315±355. Hodges, P. E., Payne, W. E. and Garrels, J. I. (1998) The Yeast Protein Database (YPD): a curated proteome database for Saccharomyces cerevisiae. Nucleic Acids Res 26, 68±72. Hodgson, J. A., Berry, D. R. and Johnston, J. R. (1985) Discrimination by heat and proteinase treatments between ¯occulent phenotypes conferred on Saccharomyces cerevisiae by the genes FLO1 and FLO5. J. Gen. Microbiol 131, 3219±3227. Holmberg, S. (1978) Isolation and characterization of a polypeptide absent from non-¯occulent mutants of Saccharomyces cerevisiae. Carlsberg Res. Commun 43, 401±413. Holmberg, S. and Kielland-Brandt, M. C. (1978) A mutant of Saccharomyces cerevisiae temperature sensitive for ¯occulation.

437

In¯uence of oxygen and respiratory de®ciency on ¯occulence. Carlsberg Res. Commun 43, 37±47. Hough, J. S., Briggs, D. E., Stevens, R. and Young, T. W. (1982a) Biology of yeast. In Malting and Brewing Science, Vol. 2, 2nd edn., pp. 527±565. Chapman and Hall, New York. Hough, J. S., Briggs, D. E., Stevens, R. and Young, T. W. (1982b) Brewery Fermentations. In Malting and Brewing Science, Vol. 2, 2nd edn, pp.644±686. Chapman and Hall, New York. Hough, J. S., Briggs, D. E., Stevens, R. and Young, T. W. (1982c) Microbial Contamination in Breweries. In Malting and Brewing Science. Vol. 2, 2nd edn, pp. 741±775. Chapman and Hall, New York. Hunter, R. J. (1984) Rheological and sedimentation behaviour of strongly interacting colloidal systems. In Modern Trends of Colloid Science in Chemistry and Biology, ed. H.-F. Eicke, pp. 184±202. Birkhauser Verlag, Boston, MA. Iimura, Y., Hara, S. and Otsuka, K. (1980) Cell surface hydrophobicity as a pellicle formation factor in ®lm strain of Saccharomyces. Agric. Biol. Chem. 44, 1215±1222. Jansen, H. E. and Mendlik, F. (1951) A study on yeast ¯occulation. Proc. Eur. Brew. Conv. Cong., Brighton, UK, pp. 59±81. Jayatissa, P. M. and Rose, A. H. (1976) Role of wall phosphomannan in ¯occulation of Saccharomyces cerevisiae. J. Gen. Microbiol 96, 165±174. Johnson, J. R and Lewis, C. W. (1974) Genetic analysis of ¯occulation in Saccharomyces cerevisiae and tetrad analysis of commercial brewing and baking yeasts. In Second International Symp. on the Genetics of Industrial Microorganisms. ed. K. D. Macdonald. pp. 339±355. Academic Press, London, UK. Johnson, J. R. and Reader, H. P. (1983) Genetic control of ¯occulation. In Yeast Genetics, Fundamental and Applied Aspects. eds. J. F. T. Spencer, D. M. Spencer and A. R. W. Smith. pp. 205±224. Springer±Verlag, New York, NY. Kamada, K. and Murata, M. (1984) On the mechanism of brewer's yeast ¯occulation. Agric. Biol. Chem. 48, 2423± 2433. Kihn, J. C., Masy, C. L. and Mestdagh, M. M. (1988) Yeast ¯occulation: Competition between nonspeci®c repulsion and speci®c bonding in cell adhesion. Can. J. Microbiol. 34, 773±778. Klis, F. M. (1994) Review: cell wall assembly in yeast. Yeast 10, 851±869. Kobayashi, O., Hayashi, N. and Sone, H. (1995) The FLO1 genes determine two ¯occulation phenotypes distinguished by sugar inhibition. Proc. Eur. Brew. Conv. Cong., Brussels, pp. 361±368. Kobayashi, O., Suda, H., Ohtani, T. and Sone, H. (1996) Molecular cloning and analysis of the dominant ¯occulation gene FLO8 from Saccharomyces cerevisiae. Mol. Gen. Genet 251, 707±715. Kobayashi, O., Hayashi, N., Unemoto, N. and Sone, H. (1997) The region of Flo1p responsible for sugar recognition. Yeast 13, s233. Kollar, R., Reinhold, B. B., Petrakova, E., Yeh, H. J., Ashwell, G., Drgonova, J., Kapteyn, J. C., Klis, F. M. and Cabib, E. (1997) Architecture of the yeast cell wall. b-(16)glucan interconnects mannoprotein, (-(1(3)-glucan, and chitin. J. Biol. Chem. 272, 17762±17775. Kreger-van Rij, N. J. W. (1984) The Yeasts, a Taxonomic Study, 3rd edn. North-Holland, Amsterdam. Kuriyama, H., Umeda, I. and and Kobayashi, H. (1991) Role of cations in the ¯occulation of Saccharomyces cerevisiae and discrimination of the corresponding proteins. Can J. Microbiol 37, 397±403.

438

Y.-L. Jin, R. A. Speers

Lands, W. E. M. and Gra€, G. (1981) E€ects of fatty acid structure on yeast cell physilogy. In Current Developments in Yeast Research. Advances in Biotechnology, eds G. G. Stewart and I. Russell. pp. 331±343. Pergamon Press, Toronto. Lewis, C. W., Johnson, J. R. and Martin, P. A. (1976) The genetics of yeast ¯occulation. J. Inst. Brew. 82, 158±160. Lindner, P. (1901) Jahrb. Versuchs-u, Vol. 4, p. 309. Lehranstalt Brauerei, Berlin. [cited by Jansen, H. E. (1958). Flocculation of Yeast. In The Chemistry and Biology of Yeasts, ed. A. H. Cook, pp. 635±667]. Academic Press, New York. Lindquist, W. (1953) On measuring the mechanism of yeast ¯occulation. J. Inst. Brew. 59, 59±61. Lo, W. S. and Dranginis, A. M. (1996) FLO11, a yeast gene related to the STA genes, encodes a novel cell surface ¯occulin. J. Bacteriol. 178, 7144±7151. Lo, W. S. and Dranginis, A. M. (1998) The cell surface ¯occulin Flo11 is required for pseudohyphae formation and invasion by Saccharomyces cerevisiae. Mol. Biol. Cell. 9, 161±171. Lodder, J. (1970) The yeast, a Taxonomic Study, 2nd edn. North-Holland, Amsterdam. Lyons, T. P. and Hough, J. S. (1970) Flocculation of brewer's yeast. J. Inst. Brew. 76, 564±571. Lyons, T. P. and Hough, J. S. (1971) Further evidence for the cross-bridging hypothesis for ¯occulation of brewer's yeast. J. Inst. Brew. 77, 300±305. Manners, D. J., Masson, A. J. and Patterson, J. C. (1973a) The structure of a b-(13)-d- glucan from yeast cell walls. Biochem. J. 135, 19±30. Manners, D. J., Masson, A. J., Patterson, J. C., Bjorndal, H. and Lindberg, B. (1973b) The structure of a b-(16)-d-glucan from yeast cell walls. Biochem. J. 135, 31±36. Martini, A. V. and Martini, A. (1989) A proposal for correct nomenclature of the domesticated species of the genus Saccharomyces. In Biotechnology Applications in Beverage Production, ed. C. Cantarelli and G. Lanzarini, pp.1±30. Elsevier Applied Science, New York. Masschelein, C. A. and Devreux, A. (1957) Flocculence et structure de la paroi cellulaire de la levure. Proc. Eur. Brew. Conv. Cong., Copenhagen, pp. 194±211. Masy, C. L., Henquinet, A. and Mestdagh, M. M. (1992a) Fluorescence study of lectin-like receptors involved in the ¯occulation of the yeast Saccharomyces cerevisiae. Can. J. Microbiol 38, 405±409. Masy, C. L., Henquinet, A. and Mestdagh, M. M. (1992b) Flocculation of Saccharomyces cerevisiae: Inhibition by sugars. Can. J. Microbiol. 38, 1298±12306. Miki, B. L. A., Poon, N. H., James, A. P. and Seligy, V. L. (1982a) Possible mechanism for ¯occulation interactions governed by gene FLO1 in Saccharomyces cerevisiae. J. Bacteriol 150, 878±889. Miki, B. L. A., Poon, N. H., James, A. P. and Seligy, V. L. (1982b) Repression and induction of ¯occulation interactions in Saccharomyces cerevisiae. J. Bacteriol. 150, 890±809. Mill, P. J. (1964) The nature of interactions between ¯occulent cells in the ¯occulation of Saccharomyces cerevisiae. J. Gen. Microbiol. 35, 61±68. Morades-Ferrira, P., Fernandes, P. A. and Costa, M. J. (1994) Yeast ¯occulationÐthe role of cell wall proteins. Colloid. Surf. B: Biointerfaces 2, 159±164. Nagasu, T., Shima, Y. I., Nakanishi, Y., Kuromitsu, J., Iwama, K., Nakayama, K. I., Suzuki, K. and Jigami, Y. (1992) Isolation of new temperature-sensitive mutants of Saccharomyces cerevisiae de®cient in mannose outer chain elongation. Yeast 8, 535±547. Nishihara, H. T., Toraya, T and Fukui, S. (1977) E€ect of a chemical modi®cation of cell surface components of a brewer's yeast on the ¯oc-forming ability. Arch. Microbiol. 115, 19±23.

Nishihara, H. T., Toraya, T and Fukui, S. (1982) Flocculation of cell walls of brewer's yeast and e€ects of metal ions, protein-denaturants and enzyme treatments. Arch. Microbiol. 131, 112±115. O'Melia, C. R. (1972) Coagulation and ¯occulation. In Physicochemical Processes for Water Quality Control, ed. W. J. Weber, pp.61±109. John Wiley and Sons Ltd., New York. Patel, G. B. and Ingledew, W. M. (1975) A ¯occulation test for moderately ¯occulent brewing yeast. J. Inst. Brew. 81, 123± 126. Patelakis, S. J. J., Ritcey, L. L. and Speers, R. A. (1998) Density of lectin-like receptors in the FLO1 phenotype of Saccharomyces cerevisiae. Letters in Appl. Microbiol 26, 279± 282. Podgornik, A., Koloini, T. and Raspor, P. (1997) On-line measurement and analysis of yeast ¯occulation. Biotechnol. and Bioeng 53, 179±184. Proteome, Inc. (15 May, 1998). Yeast Protein Database 24 July, 1998. Rhymes, M. R., Smart, K. A. E€ect of starvation on the ¯occulation of alelager brewing yeasts. J. Am. Soc. Brew. Chem 54, 50±56. Ritcey, L. L. (1997) Cell wall properties a€ecting brewing yeast ¯occulation, M.Sc. thesis, Technical University of Nova Scotia, Halifax, NS. Rogers, H. J. (1968) The yeast cell wall. In Cell Walls and Membranes, ed. H. J. Rogers. pp. 135±152. Academic Press, London. Rose, A. H. (1984) Physiology of cell aggregation: ¯occulation by Saccharomyces cerevisiae as a model system. In Physiology of Cell Aggregation, ed. K.C. Marshall, pp.323±335. Springer±Verlag, New York. Rose, A. H. (1993) Composition of the envelope layers of Saccharomyces cerevisiae in relation to ¯occulation and ethanol tolerance. J. Appl. Bacteriol. 74, 110s±118s. Russell, I. (1995) Yeast. In Handbook of Brewing, ed. W.A. pp. 168±202. Hardwick, Marcel Dekker, New York. Russell, I., Stewart, G. G., Reader, H. P., Johnson, J. R. and Martin, P. A. (1980) Revised nomenclature of genes that control yeast ¯occulation. J. Inst. Brew. 86, 120±121. Schreuder, M. P., Brekelmans, S., van der Ende, H. and Klis, F. M. (1993) Targeting of a heterologous protein to the cell wall of Saccharomyces cerevisiae. Yeast 9, 399±409. Sey€ert, H. (1896) Remarks on pure culture yeasts and their nourishment. Z. ges. Brauw., 19, 318±322 [as cited by Stratford, M. (1996a) Yeast ¯occulation: Restructuring the theories in line with recent research. Cerevisia 21, 38±45].. Shankar, C. S. and Umesh-Kumar, S. (1994) A surface lectin associated with ¯occulation on brewing strains of Saccharomyces cerevisiae. Microbiology 140, 1097±1101. Simpson, W. J. and Hammond, J. R. M. (1989) The response of brewing yeasts to acid washing. J. Inst. Brew. 95, 347± 354. Smart, K. A. and Whisker, S. (1996) E€ect of serial repitching on the fermentation properties and condition of brewing yeast. J. Am. Soc. Brew. Chem. 54, 41±44. Smit , G., Straver, M. H., Lugtenberg, B. J. J. and Kijne, J. W. (1992) Flocculence of Saccharomyces cerevisiae cells is induced by nutrient limitation, with cell surface hydrophobicity as a major determinant. Appl. Environ. Microbiol. 58, 3709±3714. Soares, E. V. and Mota, M. (1996) Flocculation onset, growth phase, and genealogical age in Saccharomyces cerevisiae. Can. J. Microbiol. 42, 539±547. Sousa, M. L., Mota, M. and Teixeira, J. A. (1994) In¯uence of operational parameters on the start-up of a ¯occulation airlift bioreactor. Colloids and Surfaces B: Biointerfaces 2, 181±188.

Flocculation of Saccharomyces cerevisiae Speers, R. A. (1991) Rheological and colloidal properties of commercial brewing yeast suspensions. Ph.D. thesis, University of British Columbia, Vancouver, BC. Speers, R. A., Durance, T. D. and Tung, M. A. (1990) Flow behavior of commercial brewing yeast suspensions. In Rheology of Food, Pharmaceutical and Biological Materials, ed. R.E. Carter, pp. 74±100. Elsevier, London. Speers, R. A., Tung, M. A., Durance, T. D. and Stewart, G. G. (1992a) Biochemical aspects of yeast ¯occulation and its measurement: A review. J. Inst. Brew 98, 293±300. Speers, R. A., Tung, M. A., Durance, T. D. and Stewart, G. G. (1992b) Colloidal aspects of yeast ¯occulation: a review. J. Inst. Brew 98, 525±531. Speers, R. A., Durance, T. D., Odense, P., Owen, S. and Tung, M. A. (1993a) Physical properties of commercial brewing yeast suspensions. J. Inst. Brew. 99, 159±164. Speers, R. A., Durance, T. D., Tung, M. A. and Tou, J. (1993b) Colloidal properties of ¯occulent and non-¯occulent brewing yeast suspensions. Biotechnol. Prog. 9, 267±272. Speers, R. A. and Ritcey, L. L. (1995) Towards an ideal ¯occulation assay. J. Am. Soc. Brew. Chem. 53, 174±177. Speers, R. A., Smart, K. A., Stewart, R. J. and Jin, Y. (1998) Zymolectins in Saccharomyces cerevisiae. J. Inst. Brew. 104, 298. Stewart, G. G., Garrison, I. F., Goring, T. E., Meleg, M., Pipasts, P. and Russell, I. (1976) Biochemical and genetic studies on yeast ¯occulation. Kemia-Kemi 3, 465±479. Stewart, G. G. and Goring, T. E. (1976) E€ect of some monovalent and divalent metal ions on the ¯occulation of brewer's yeast strains. J. Inst. Brew. 82, 341±342. Stewart, G. G. and Russell, I. (1977) The identi®cation, characterization and mapping of a gene for ¯occulation in Saccharomyces sp. Can. J. Microbiol. 23, 441±447. Stewart, G. G. and Russell, I. (1981) Yeast ¯occulation. In Brewing Science, Vol. 2, ed. J. R. A. Pollock, pp. 61±91. Academic Press, New York. Stewart, G. G., Russell, I. and Garrison, I. F. (1973) Further studies on ¯occulation and co-¯occulation in brewer's yeast strains. Proc. Am. Soc. Brew. Chem. 31, 100±106. Stewart, G. G., Russell, I. and Goring, T. (1975) NatureNuture anomalies-further studies in yeast ¯occulation. Proc. Am. Soc. of Brew. Chem. 33, 137±147. Stewart, G. G., Russell, I. and Garrison, I. F. (1975) Some considerations of the ¯occulation characteristics of ale and lager yeast strains. J. Inst. Brew. 81, 248±257. Stewart, R. J., Russell, I. and Stewart, C. G. (1995) Characterization of anity-puri®ed cell wall-binding proteins of Saccharomyces cerevisiae: Possible role in ¯occulation. J. Am. Soc. Brew. Chem. 53, 111±116. Stratford., M. (1989a) Yeast ¯occulation: the in¯uence of agitation. Yeast 5, 97s±102s. Stratford, M. (1989b) Evidence for two mechanisms of ¯occulation in Saccharomyces cerevisiae. Yeast 5, 441s±445s. Stratford, M. (1989c) Yeast ¯occulation: calcium speci®city. Yeast 5, 487±496. Stratford, M. (1992a) Yeast ¯occulation: reconciliation of physiological and genetic viewpoints. Yeast 8, 25±38. Stratford, M. (1992b) Yeast ¯occulation: receptor de®nition by mnn mutants and concanavalin A. Yeast 8, 635±645. Stratford, M. (1992c) Yeast ¯occulation: a new perspective. In Advances in Microbiol. Physiol, Vol. 33, ed. A. H. Rose. pp. 1±71. Academic Press, New York. Stratford, M. (1993) Yeast ¯occulation: ¯occulation onset and receptor availability. Yeast 9, 85±94. Stratford, M. (1994) Another brick in the wall? Recent developments concerning the yeast cell envelope. Yeast 10, 1741±1752. Stratford, M. (1996a) Yeast ¯occulation: restructuring the theories in line with recent research. Cerevisia 21, 38±45.

439

Stratford, M. (1996b) Induction of ¯occulation in brewing yeasts by change in pH value. FEMS Microbiol. Lett. 136, 13±18. Stratford, M. and Assinder, S. (1991) Yeast ¯occulation: Flo1 and NewFlo phenotypes and receptor structure. Yeast 7, 559±574. Stratford, M. and Carter, A. T. (1993) Yeast ¯occulation: lectin synthesis and activation. Yeast 9, 371±378. Stratford, M., Coleman, H. P. and Keenan, M. H. J. (1988) Yeast ¯occulation: a dynamic equilibrium. Yeast 4, 199± 208. Stratford, M. and Keenan, M. H. J. (1987) Yeast ¯occulation: kinetics and collision theory. Yeast 3, 201±206. Stratford, M. and Keenan, M. H. J. (1988) Yeast ¯occulation: quanti®cation. Yeast 4, 107±115. Stratford, M. and Wilson, P. D. G. (1990) Agitation e€ects on microbial cell±cell interactions. Lett. Appl. Microbiol 11, 1± 6. Straver, M. H. and Kijne, J. W. (1996) A rapid and selective assay for measuring cell surface hydrophobicity of brewer's yeast cells. Yeast 12, 207±213. Straver, M. H., Kijne, J. W. and Smit, G. (1993a) Cause and control of ¯occulation in yeast. Trends in Biotechnol 11, 228±232. Straver, M. H., van der Aar, P. C., Smit, G. and Kijne, J. W. (1993b) Determinants of ¯occulence of brewer's yeast during fermentation in wort. Yeast 9, 527±532. Straver, M. H., Smit, G. and Kijne, J. W. (1994a) Induced cell surface hydrophobicity in¯uences ¯occulation of brewer's yeast in a ¯occulation assay. Colloid. Surf. B: Biointerfaces 2, 173±180. Straver, M. H., Smit, G. and Kijne, J. W. (1994b) Puri®cation and partial characterization of a ¯occulin from brewer's yeast. Appl. Environ. Microbiol 60, 2754±2758. Straver, M. H., Traas, V. M., Smit, G. and Kijne, J. W. (1994c) Isolation and partial puri®cation of mannose-speci®c agglutinin from brewer's yeast involved in ¯occulation. Yeast 10, 1183±1193. Takita, M. A. and Castilho-Valavicius, B. (1993) Absence of cell wall chitin in Saccharomyces cerevisiae leads to resistance to Kluyveromyces lactis killer toxin. Yeast 9, 589±598. Taylor, N. W. and Orton, W. L. (1973) E€ect of alkaline-earth metal salts on ¯occulence in Saccharomyces cerevisiae. J. Inst. Brew 79, 294±297. Taylor, N. W. and Orton, W. L. (1975) Calcium in ¯occulence of Saccharomyces cerevisiae. J. Inst. Brew 81, 53±57. Taylor, N. W. and Orton, W. L. (1978) Aromatic compounds and sugars in ¯occulation of Saccharomyces cerevisiae. J. Inst. Brew 84, 113±114. Teixeira, J. M., Teixeira, J. A., Mota, M., Manuela, M., Guerra, B., Machado Cruz, J. M. and S'Almeida, A. M. (1991) The in¯uence of cell wall composition of a brewer's ¯occulant lager yeast on sedimentation during successive industrial fermentations. Proc. Eur. Brew. Conv. Cong., Lisbon, pp. 241±248. Teunissen, A. W. R. H. and Steensma, H. Y. (1990) Physical location of the FLO1 gene of Saccharomyces cerevisiae. Yeast 6, s35.. Teunissen, A. W. R. H. and Steensma, H. Y. (1995) Review: The dominant ¯occulation genes of Saccharomyces cerevisiae constitute a new subtelomeric gene family. Yeast 11, 1001±1013. Teunissen, A. W. R. H., van den Berg, J. A. and Steensma, H. Y. (1993) Physical localization of the ¯occulation gene FLO1 on chromosome I of Saccharomyces cerevisiae. Yeast 9, 1±10. Teunissen, A. W. R. H., van den Berg, J. A. and Steensma, H. Y. (1995) Localization of the dominant ¯occulation genes

440

Y.-L. Jin, R. A. Speers

FLO5 and FLO8 of Saccharomyces cerevisiae. Yeast 11, 735±745. Tha, S. P. and Goldsmish, H. L. (1986) Interaction forces between red cells agglutinated by antibody. I. Theoretical Biophys. J. 50, 1109±1116. Thorne, R. S. W. (1951) Some aspects of yeast ¯occulence. Proc. Eur. Brew. Conv. Cong., Brighton, UK, pp. 21± 34. Van de Van, T. G. M. and Mason, S. G. (1977) The microrheology of colloidal dispersions. VII. Orthokinetic doublet formation of spheres. Colloid and Polymer Sci 255, 468± 479. Van der Aar, P. C. (1996) Consequences of yeast population dynamics with regard to ¯occulation. Ferment 9, 39±42. Van Hamersveld, E. H., van Loosdrecht, M. C. M., van der Lans, R. G. J. M. and Luyben, K. Ch. A. M. (1996) On the measurement of the ¯occulation characteristics of brewer's yeast. J. Inst. Brew 102, 333±342.

van Oss, C. J. and Giese, R. F. (1995) The hydrophilicity and hydrophobicity of clay minerals. Clays and Clay Minerals 43, 474±477. Watari, J., Takata, Y., Ogawa, M., Sahara, H., Koshino, S., Onnela, M. L., Airaksinen, U., Jaatinen, R., Pentilla, M. and Keranen, S. (1994) Molecular cloning and analysis of the yeast ¯occulation gene FLO1. Yeast 10, 211±225. Wilcocks, K and Smart, K. A. (1994) The importance of surface charge and hydrophobicity for the ¯occulation of chain-forming brewing yeast strains and resistance of these parameters to acid washing. FEMS Microbiol. Lett. 134, 293±297. Yamashita, I. and Fukui, S. (1983) Mating signals control expression of both starch fermentation genes and a novel ¯occulation gene FLO8 in the yeast Saccharomyces. Agric. Biol. Chem. 47, 2889±2896.

(Received 3 September 1998; accepted 17 January 1999)