Yeast flocculation

Yeast flocculation typically refers to the clumping together (flocculation) of brewing yeast once the sugar in a wort has been fermented into beer.In the case of "top-fermenting" ale yeast (Saccharomyces cerevisiae), the yeast creates a "head" on the top of the liquid, unlike with "bottom-fermenting" lager yeast (Saccharomyces pastorianus) where the yeast falls to the bottom of the brewing vessel.

Cell aggregation occurs throughout microbiology, in bacteria, filamentous algae, fungi and yeast (Lewin, 1984; Stratford, 1992). Yeast are capable of forming three aggregates; mating aggregates, for DNA exchange; chain formation; and flocs as a survival strategy in adverse conditions (Calleja, 1987). Industrial brewing strains rarely mate. Therefore, only chain formation and flocculation are of relevance to the brewing industry.

Yeast flocculation is distinct from agglomeration (‘grit’ formation), which is irreversible and occurs most commonly in baker's yeast when strains fail to separate when resuspended (Guinard and Lewis, 1993). Agglomeration only occurs following the pressing and rehydration of yeast cakes and both flocculent and non-flocculent yeast strains have been shown to demonstrate agglomeration (Guinard and Lewis, 1993). It is also distinct from the formation of biofilms, which occur on a solid substrate.

Louis Pasteur is erroneously credited with first describing flocculation of brewer’s yeast. Brewer's yeast flocculation has been the subject of many reviews (Stewart et al., 1975; Stewart and Russell, 1986; Calleja, 1987; Speers et al., 1992; Jin and Speers, 1999). Flocculation has been defined as the reversible, non-sexual aggregation of yeast cells that may be dispersed by specific sugars (Burns, 1937; Lindquist, 1953, Eddy, 1955; Masy et al., 1992) or EDTA (Burns, 1937; Lindquist, 1953). The addition of nutrients other than sugars has been demonstrated not to reverse flocculation (Soares et al., 2004). This is as opposed to mating aggregates formed as a prelude to sexual fusion between complementary yeast cells (Calleja, 1987).

For flocculation to occur the yeast must be flocculent and certain environmental conditions (such as agitation, absence of sugars, a microamount of Ca²⁺, ethanol, etc.; Jin and Speers 1999) must be present. Several factors are important in cell-to-cell binding such as surface charge, hydrophobic effects and zymolectin interactions (see following). The importance of these forces in brewing yeast flocculation was unrecognized in the past but work by Speers et al. (2006)^[1] have indicated the importance of zymolectin and hydrophobic interactions. As the cells are too large to be moved by Brownian motion, for binding of two or more cells to occur the cells must subjected to low level of agitation.

Zymolectin Interaction Theory

The accepted mechanism of flocculation involves a protein-carbohydrate model (Miki et al., 1982) (figure 1.3). Fully flocculent yeast cells exhibit carbohydrate α-mannan receptors and protein zymolectins (section 1.5.4). Zymolectins are so termed as they may not be true bivalent lectins (Speers, Smart, Stewart and Jin,1998)^[2] It has been suggested that zymolectin interactions between the protein and mannan moities results in the flocculation phenotype (section 4.1) with Ca²⁺ ions required for the correct conformation of the zymolectins. Coflocculation between Kluyveromyces and Schizosaccharomyces has been shown to be by a “lectinic” mechanism (El-Behhari et al., 2000). This theory explains the essential role of calcium and how deproteinisation affects flocculation.

Flocculation Zymolectins and Phenotypes

Three flocculation phenotypes have been elucidated based on the zymolectins they produce: Flo1 (Stratford and Assinder, 1991) NewFlo (Stratford and Assinder, 1991) and Mannose Insensitive (MI) (Masy et al., 1992; Dengis and Rouxhet, 1997). These flocculation phenotypes differ in the time of the onset of flocculation and the sugar inhibition of flocculation. Flocculation has also been classified according to time of onset and floc morphology.

Flocculation phenotypes

Gilliland Class	Flocculation Characteristics
I	Completely Dispersed
II	Flocculating into small, loose lumps late in fermentation
III	Flocculating into dense masses late in fermentation
IV	Flocculating very early in fermentation owing to non-separation of daughter cells

The genetic control of yeast flocculation has not been extensively studied. Recent reports suggest genes encoding lectin-like proteins exhibit close sequence homology (Jin and Speers, 1991, 1999; Smart, 2001). Furthermore, it seems that FLO genes have interchangeable functions that can compensate for one another (Guo et al., 2000).

Flocculation Phenotypes

The Flo1 phenotype is inhibited by mannose (Burns, 1937; Miki et al., 1982; Nishihara and Toraya, 1987; Kihn et al., 1988; Stratford, 1989; Stratford and Assinder, 1991) occurs in both ale and lager strains (Miki, 1982; Stratford and Assinder, 1991; Masy et al., 1992; Smit et al., 1982; Stratford, 1993; Stratford and Carter, 1993; Teunissen et al., 1993; Teunissen et al., 1995a, b; Bony et al., 1997; Braley and Chaffin, 1999; Fleming and Pennings, 2001; He et al., 2002; Verstrepen et al., 2003) and is associated with the FLO1 gene (Watari, 1991 Masy et al., 1992; Stratford, 1993; Stratford and Carter, 1993; Teunissen et al., 1993; Teunissen et al., 1995a, b; Bony et al., 1997; Braley and Chaffin, 1999).

The NewFlo phenotype differs from that of Flo1 in several ways. Firstly NewFlo flocculation is inhibited by mannose, glucose and maltose (Stratford and Assinder, 1991; Masy, 1992; Rhymes, 1999). Secondly the NewFlo lectin is putatively encoded by the FLO10 gene (Guo et al., 2000; Smart, 2001) and may not expressed until stationary phase onset (Stratford, 1989; Stratford and Assinder, 1991; D’Hautcourt and Smart, 1999). This is not clear since studies by the Speers group have indicated little change in zymolectin levels through a fermentation. It is argued that lectin maturation occurs some fourteen hours after the cessation of cell division (Stratford, 1989; Stratford and Assinder, 1991; Masy 1992; D’Hautcourt and Smart, 1999) and is therefore not concomitant with entry into stationary phase, although this is strain dependent (D’Hautcourt and Smart, 1999; Verstrepen et al., 2003). However the molecular proof of this maturation is near non-existent. The picture is complicated by changes in cell surface hydrophobicity and CO₂ driven shear which confounds flocculation measurements often erroneously solely attributed to zymolectin interactions.While flocculation (clumping and settling) occurs at this time, this flocculation occurs as a result of changes in hydrophobicity and a decline in shear due to which is in turn to low CO₂ evolution.

The MI phenotype appears to occur in ale but not lager strains (Masy et al., 1992; Dengis and Rouxhet, 1997; Jin and Speers, 1999) and is considered to be a rare phenotype. The FLO11 gene has however been identified as being essential for flocculation in S. bayanus (Ishigami et al., 2004) and characteristics such as invasive growth and pseudohyphal formation in Saccharomyces cerevisiae (Lo and Dranginis, 1998; Gagiano et al., 1999; Gancedo, 2001; Gagiano et al., 2002; Gagiano et al., 2003; Verduzco-Luque et al., 2003; Vivier et al., 2003; Guldener et al., 2004). Although this flocculation phenotype has not been fully characterised, it is differentiated from other flocculation phenotypes by a lack of inhibition of the zymolectin-like reaction in the presence of mannose (Dengis and Rouxhet, 1997; Guo et al., 2000; Smart, 2001).

References

1. Speers, R.A., Wan, Y-Q., Jin, Y-L., and R. J. Stewart, R.J. 2006. Effects of fermentation parameters and cell wall properties on yeast flocculation. J. Inst. Brew. 112:246-254.
2. Speers, R.A., Smart, K., Stewart, R., Jin, Y-L., 1998. Zymolectins in Saccharomyces cerevisiae. Letter J. Inst. Brew., 104:298.