Fig. 4.35 C.ontd on next page

Fig. 4.35 Confocal micrographs of budding yeast cells yeast - bud scars stained with wheat germ agglutinin (kindly provided by Christopher Powell, Oxford Brookes University).

noprotein population as well as proteins that are believed to interact with cell wall glucans (Fleet & Manners, 1976). The enzymes divide into two groups; those which are involved in the metabolism of nutritional substrates and those involved in cell wall morphogenesis. The former class that includes the mannoproteins invertase and acid phosphatase are located in the 'never-never land' (Robinow & Johnson, 1991) of the periplasm - an ill defined region between the plasma membrane and the cell wall (see Fig. 4.1). The wall degrading enzymes, of which the |3-(l->3 )-glucanase is by far the best characterised, appears to have a roving brief around the cell wall. Finally, the lectin-like proteins represent the surface receptor proteins. These are discussed at length in Section

4.4.3 Cell wall and fermentation

The cell wall polysaccharides - glucan and mannan - each account for c. 5% of the yeast biomass during fermentation (Quain et al., 1981). Although somewhat low key compared to the arguably more exciting storage polysaccharide glycogen, the cell wall and its components play an important role in brewery fermentations.

Although its role in flocculation has received the plaudits (see below), the general contribution of the yeast cell wall to brewery fermentation has yet to be fully appreciated! One facet, which draws little attention, is the role of the yeast cell wall as a protective layer to the ever-changing environment in which the cell finds itself. In particular, brewery worts contain a melange of suspended particles that associate with the cell wall. Depending on production methods, wort can contain as much as 4% (v/ v) particulates or 'trub' (Lentini et al., 1994). As a rule of thumb, the more turbid the wort, the more trub it contains. Trub is a miscellany of substances that include insoluble proteins, complex carbohydrates, lipids, polyphenols and hop substances. There is no doubt that trub associates intimately with the cell wall. Trub lipids such as linoleic acid are adsorbed by yeast and can have a beneficial effect on fermentation, yeast growth and beer flavour (Schisler et al., 1982; Siebert et al., 1986; Lentini et al., 1994). However, O'Connor-Cox et al. (1996) have argued that the brewer should strive for bright and not cloudy worts. In their view, fermentation performance and product quality are compromised in cloudy worts. Intriguingly they propose that the 'yeast cells become coated with the "cloudy" material thereby preventing optimal activity' (O'Connor-Cox et al., 1996). Extending this concept, the association of trub with the cell wall is recognised as potentially interfering with the fining out of yeast (see Section There is no doubt that hop substances bind to the yeast cell wall (Hough & Hudson, 1961; Dixon & Leach, 1968; Laws et al., 1972) - so much so that significant 'bitterness' is lost because of fermentation, which can simply be recovered from the cell wall by washing with water (Kieninger & Durner, 1982).

The impact of serial repitching on the binding of trub and hop iso-a-acids is of interest. A model for what may happen has been reported by Kieninger and Durner (1982). Binding of iso-a-acids increased from 16.9 mg/100 g dry yeast after the first fermentation to 70 mg/100 g dry yeast after the fourth. Should similar accumulation occur with wort proteins, polyphenols and other substances, it is tempting to assume that such events may well influence flocculation, substrate transport and other wall functions. There is evidence that the charge of the cell wall becomes progressively more negative with serial repitching (Lawrence et al., 1989). Further, Bowen and Ventham (1994) have speculated that there is electrostatic attraction between ale yeasts and trub, which may enhance flocculation (see Section Acid washing. Of the many abuses yeast is exposed to during its life in the brewery, acid washing (see Section 7.3.3) to kill contaminating bacteria (but not contaminating yeasts!) is arguably the most extreme. 'Washing' yeast at pH 2.1 for two hours or so in the cold would be expected to damage the cell wall and result in generalised stress to the cell. Work by Simpson and Hammond (1989) has shown that although brewing yeasts are generally resistant to acid washing, the cell wall exhibits changes. In particular, surface blistering ('blebbing') was observed particularly at elevated temperature (25°C) (Fig. 4.36). Further, acid washing removes associated trub-like material (Fig. 4.37) and makes the yeast sticky (Fig. 4.38). Fining. The need to separate liquid and solids is a common theme in the brewing process. In an ideal world, the removal of yeast from beer is primarily achieved through flocculation (see Section 4.4.4). Green beer centrifuges are used on run down from fermenter to 'trim' yeast counts further or, where required, to remove powdery, poorly flocculent strains. However, inevitably, green beer will still contain yeast usually in the range of 1-3 x 106/ml. This loading is reduced significantly by the addition of isinglass finings either in conditioning tank (chilled and filtered products) or final package (cask beer). These materials interact with yeast cells to form increasingly large floes that sediment or 'fine' out of solution (for a review see Leather, 1994).

It is intriguing to speculate about the origins of 'white finings' or 'isinglass' as they are sourced from the swim bladders of tropical fish! Their success in removing yeast cells is down to the presence of collagen. This large linear protein carries a positive

Fig. 4.36 'Blebbing' of yeast cells after acid washing (from Hammond and Simpson (1989). with permission from the Institute of Brewing).

charge which, it is believed, interacts with the negatively charged yeast cell. The contribution of the yeast cell wall to fining is widely accepted, but it must be assumed that the phosphate moiety within the mannoprotein matrix (see Section is the key determinant of successful fining. Mindful of comments above about interactions with trub (4.4.3), recent work by Leather et al. (1997) has shown that increasing loadings of positively charged 'non-microbiological particles' (NMP) has progressively deleterious effects on isinglass fining performance (see Section Presumably, NPM associates with cell wall phosphate, which is then unable to interact with isinglass. Commercial applications. The yeast cell wall has been identified as a source of interesting and commercially useful polymers. In particular, glucan has 'functional characteristics' being able to hold water and to provide thickening properties (Johnson, 1977). More recently, mannoproteins from S. cerevisiae have been identified as being 'bioemulsifiers' (Cameron et al., 1988) and foam stabilisers (Kunst et al., 1996). A protein-rich fraction from yeast cell walls has been advocated for use in clarification processes such as fining of beer (Jackson, 1996).

4.4.4 Flocculation - an introduction

Yeast flocculation is a physical process and is of fundamental importance in brewery fermentations. Contrary to popular opinion, flocculation is not about sedimentation but the aggregation of single yeast cells into multicellular 'floes'. Perhaps 'aggregation' (Miki et al., 1982) or, less elegantly, 'clumping' are more useful descriptions than flocculation. Whatever, a good and popular definition of this process is 'the phenomenon wherein yeast cells adhere in clumps and sediment rapidly from the medium in which they are suspended or rise to the medium's

Fig. 4.37 Yeast cells (a) before and (b) after acid washing (from Hammond and Simpson (1989). with permission from the Institute of Brewing).

surface' (Stewart & Russell, 1981). In passing, flocculation is unrelated to another form of yeast cell aggregation - the small cell floes formed by certain chain forming brewing strains.

Flocculation is critical to yeast recovery be it through the skimming of open fermenters ('flotation') or, more typically, the cropping of closed cylindroconical vessels ('sedimentation' ). Because of this, it is tempting to conclude that such yeasts have been selected as strains which flocculate at the appropriate time so as to facilitate yeast recovery/pitching-on as well as controlling downstream yeast cell counts to racking or conditioning tank. Whether, in addition to making the brewer's life a little easier, flocculation offers some benefit to the yeast cell remains to be seen! Johnson (Stewart

Fig. 4.38 Sticky yeast post acid washing (from Hammond and Simpson (1989). with permission from the Institute of Brewing).

& Russell, 1981) and Iserentant (1996) have, quite reasonably, suggested that the formation of cell floes is one of the many responses of yeast to stress. 'Multicellular' floes, it is argued, offer greater protection to the external inclement environment.

Although not usually routinely monitored or measured, yeast flocculation is a high profile fermentation parameter second only to fermentation rate (PG drop). This is borne out by the readiness with which any in-process changes in yeast flocculence are seen through yeast solids or head formation. Such changes are thought to reflect genetic changes (see Section

Ironically, despite its importance, yeast flocculation in the brewing industry remains a 'given' with little real practical understanding of the factors that determine yeast flocculation. This is despite the efforts of legions of yeast physiologists, biochemists and geneticists who have exhaustively studied flocculation. Indeed, it has been estimated that in the last 20 years, 10-15 papers per year have been published on 'flocculation' (Speers & Ritcey, 1995). This 'hit rate' may well increase as, in recent years, flocculation has acquired a new interest and input from biotechnologists studying the removal of yeast cells from non-brewing fermentation broths.

4.4.5 Overview

Yeast flocculation has inspired numerous and lengthy reviews. In particular the reader is referred to the reviews of Stratford (1992a, 1996) and Speers el al. (1992) which ably detail the complexity and occasional ambiguity that has characterised work on flocculation. Although the majority of studies focus on brewing strains of S. cerevisiae, flocculation is found in other yeast genera (Stratford, 1996) such as Candida and Kluyeromyces.

Athough the debate continues, there appears to be general agreement that the physiology and biochemistry (but perhaps not genetics) of yeast flocculation are now broadly understood. From the viewpoint of brewery fermentations, the aggregation of cells into clumps ('floes') is an unusual, possibly unique process insomuch that individual yeast cells interact directly with each other. In summary, yeast flocculation occurs toward the end of fermentation (typically early stationary phase) and involves the interaction of cell wall proteins on one cell to carbohydrate 'branches' on the cell wall of another. Calcium is necessary for the 'activation' of the cell wall protein. As would be anticipated the process is reversible! On repitching flocculent yeast is 'deflocculated' by the presence of simple sugars or, in the laboratory, by the removal of calcium.

4.4.6 Mechanism

Given the history of 'flocculation' it seems inevitable that the present 'best bet' for the mechanism of this process will be subject to further development and embellishment. Certainly, as noted by Stratford (1996) 'several pieces of evidence have recently come to light that cast doubt on the (below) simple picture of yeast flocculation'. However, despite these slight reservations, the currently accepted theory of flocculation fits the majority, if not all, the facts and observations made on this process in the last 20 years or so.

The current concept of binding of surface proteins to carbohydrate receptors on neighbouring cell walls (Stratford, 1996) dates back to Eddy and Rudin (1958). However, what has now become, the 'lectin theory of flocculation' stems from a report by Miki et al. (1982). This seminal paper describes flocculation in terms of the recognition and interaction of cell wall factors. It is argued, through convincing experimental evidence, that the cell to cell interactions of flocculation are driven through 'lectin-like' proteins bonding with outer mannan chains. Lectin-like proteins. The involvement of lectin-like proteins or 'adhesins' in flocculation was first proposed by Miki et al. (1982). The interaction of the lectin-like proteins on flocculent cell surfaces (Fig. 4.39) with mannose receptors (Fig. 4.40) (see Section is now accepted as being at the heart of the flocculation (Fig. 4.41) theory. The involvement of either party can be simply demonstrated by the inhibition

• Determine flocculation

Fig. 4.39 Diagrammatic representation of cell wall lectins.



Cell wall

Receptors are present throughout fermentation. Receptors require at least 2-3 mannose residues.

Cell wall

Receptors are present throughout fermentation. Receptors require at least 2-3 mannose residues.

Fig. 4.40 Diagrammatic representation of cell wall receptors.

Fig. 4.41 Diagrammatic representation of yeast flocculation.

of flocculation by the presence of excess but related molecules. For example, a plant lectin (Concanavilin A) is able to competitively inhibit flocculation (Miki et al., 1982). Similarly, specific simple sugars (mannose, maltose etc.) are able to reversibly inhibit flocculation (Miki et al., 1982; Stratford & Assinder, 1991).

The inhibition of flocculation by sugars is worthy of fuller consideration. Although early work (Eddy, 1955) showed sucrose, mannose, maltose and glucose to be effective 'deflocculating agents', more recent work (Stratford & Assinder, 1991) with 42 strains of S. cerevisiae, has identified two distinct phenotypes. One group, Flol, were inhibited only by mannose which - as this sugar is not present in wort - implies such strains are typically heavily flocculent throughout fermentation. Conversely, mannose, maltose, sucrose and glucose inhibited the 'NewFlo' phenotype. Such strains are 'typical brewing strains'. To overcome any metabolic interference or confusion, sugars were added to heat killed but flocculent cells. A ready explanation of these observations is that there are two types of lectin-like proteins in yeast. In addition to differences in sugar inhibition (Stratford and Assinder, 1991), the two lectins were shown to differ in response to pH, protease digestion and inhibition by cations. These differences, it is argued, reflect the NewFlo proteins' greater suscept-ability or availability to these treatments. However, despite such differences, the preferred receptors of both Flol and NewFlo phenotypes are the outer branches of mannan (see Section below).

Typically, further complexity has been introduced by the identification of a third phenotype (Dengis et al., 1995) which is neither Flol or NewFlo. 'Mannose insensitive' (MI) top-fermenting strains are not inhibited by mannose or sucrose and require ethanol (5-10%, v/v) but not calcium for flocculation (Dengis & Rouxhet, 1997). These MI strains may not obey the current rules of flocculation, being 'governed by non-specific interactions solely, or by non-lectin specific interactions (e.g. protein-protein)' (Dengis & Rouxhet, 1997).

The involvement of inorganic salts in yeast flocculation has been recognised for many years (see Stratford, 1992a). It is now clear that calcium ions are unequivocally required for flocculation. Indeed, the presence of chelating agents (that remove calcium) causes deflocculation which, in turn, can be reversed by the addition of further calcium. However, what is not clear is the precise role of calcium in flocculation. An attractive proposal is that calcium is somehow involved in maintaining the correct structural conformation of the lectin-like protein for bonding with flocculation receptors (Stratford, 1989). This remains to be seen but there is supporting evidence from work with calcium-dependent animal lectins (Stratford, 1992b).

The lectin-like proteins - not the receptors (see below) - determine both the capability and onset of flocculation. Working with a flocculent brewing strain, Stratford and Carter (1993) showed that although lectins were synthesized throughout growth they were only available for flocculation during early stationary phase. Although poorly understood, the activation process involves no new protein synthesis and can be mimicked in the laboratory by the rapid boiling and cooling of 'non-flocculent' cells.

Isolation and characterisation of the lectin-like protein would clearly add credibility to its proposed role in flocculation. However, as if in support of Stratford's (1996) clouding of the 'simple picture', the reports detailing various putative lectin-like proteins have not fully supported the expectations of this protein. For example, two reports have described surface proteins from yeast with lectin-like properties (Straver et al., 1994; Shankar & Umesh-Kumar, 1994).

One report (Straver et al., 1994a) describes the recovery of an 'agglutinin' from cell walls of both flocculent and non-flocculent cells. Intriguingly, the protein mimics flocculation in being sensitive to mannose, pH and is calcium dependent. Further addition of the partially purified protein to the same strain of yeast stimulated flocculation in flocculating (stationary phase) and in non-flocculating (exponential) cells. However, in the latter case, stimulation although measurable was 'weak' and required significantly more agglutinin to trigger a response. Although clearly part of the flocculation process, this lectin-like protein differs from that postulated in the model in that it is easily released from yeast cells by agitation during a laboratory flocculation protocol. This would suggest that there are other anchored lectin-like proteins, that release through agitation is artefactual or, fundamentally, that the model is flawed. Although a conclusion has not been reached, the authors present an alternative model where the released agglutinin cross links a non-lectin glycoprotein ('flocculin') found only in flocculent cells (Straver et al., 1994b). As the authors conclude, there is a need for 'more insight into the interplay between flocculin, fimbriae-like structures and agglutinin' (Straver et al., 1994a).

The putative lectin-like protein isolated by Shankar and Umesh-Kumar (1994) is perhaps a better contender for the hypothetical protein. Unlike the Straver et al. (1994b) agglutinin, this protein is not found in non-flocculating yeast. Further, in a particularly elegant experiment, the protein was shown to bind to inert surfaces coated with yeast mannan but only in the presence of calcium. Similarly the protein was shown to bind to a non-flocculent 'mutant'. As noted by the authors, 'this is probably the first report on the isolation of active lectins associated with flocculation' (Shankar & Umesh-Kumar, 1994). Receptors. The identity of the flocculation receptors has proved less controversial. The inhibition of flocculation by mannose is consistent with the lectin-like protein interacting with this sugar. As mannan - a polysaccharide consisting of mannose - is a major component of the yeast cell wall it is hardly surprising that mannan became the major candidate for the flocculation 'receptor'. Further evidence came from Miki et al., (1982) who showed inhibition of the process by a plant lectin (Concanavilin A) which binds selectively to a-mannan.

Stratford (1992c) using mannoprotein mutants elegantly demonstrated the involvement of mannoproteins in flocculation. Although structurally complex (see Fig. 4.33), only the 'outer chain' of the mannoprotein molecule is involved in flocculation. Using a selection of mnn mutants, Stratford (1992c) was able to probe the requirements for receptor structure by coflocculation of the non-flocculent mutants with flocculent strains and by aggregation using concanavalin A (a plant lectin that binds to the flocculation receptors). Two mutants - mnn2 and mnn5 - were unable to coflocculate. As mnn2 mutants lack the side chains on the outer chain and mnn5 contain truncated side chains with only one mannose unit, it is clear that the receptors are the di/trisaccharide side chains of the outer mannan chain.

Given the apparent structural importance of mannoproteins in the yeast cell wall, it is perhaps not surprisingly that functional receptors are present throughout growth and, consequently, do not determine flocculence. Stratford (1993) demonstrated this for 11 strains of S. cerevisiae using the simple but powerful coflocculation and concanavalin A methods used in the above work on receptor structure. Interaction between receptors and lectin-like proteins. The timing of the flocculation event is of clear importance. Early flocculation will result in 'stuck', poorly attenuated fermentations whereas late flocculation will lead to over-attenuated beers and difficulties in yeast cropping. Typically however, flocculation occurs in laboratory aerobic fermentations when cell division has stopped (Smit et al., 1992) and the culture has entered stationary phase (Amri et al., 1982). It seems likely that this is governed by the inhibitory presence of sugars as apparently non-flocculent exponentially growing yeast can be flocculated by washing and resuspension in calcium containing buffer (Stratford, 1992a). This is in keeping with the above observations that receptors (Stratford, 1993) and the lectin-like proteins (Stratford & Carter, 1993) are present throughout fermentation.

The vast majority of flocculation studies are performed aerobically in shake flask cultures. Although technically more convenient, the physiology of aerobically grown yeast is often different and distinct from yeast grown anaerobically. In this case it is heartening to report that in anaerobic fermentations of wort, flocculation of (presumably) NewFlo brewing strains begins when cell division stops (Kempers et al., 1991; Straver et al., 1993) at the so-called 'mid-point' of fermentation (Thurston et al., 1981). Although probably true of many production brewing strains the work of Gilliland (1951) clearly demonstrated at least four different classes of flocculence. It is tempting to conclude that the differentiation of brewing strains into four groups (see Section reflects a non-flocculent strain (class 1), a top-fermenting NewFlo strain (class 2), a bottom-fermenting NewFlo strain (class 3) and an early, heavily flocculent Flol strain.

Although seemingly contradictory, flocculation requires agitation! The early 'fortuitous' observations of Stratford and Keenan (1987) have been further developed into a persuasive argument (for a review see Stratford, 1992a). Initial laboratory studies (Stratford & Keenan 1987) showed that the harder flocculent yeasts were shaken the better they flocculated. Indeed, without agitation, a flocculating culture was unable to flocculate. Further, the rate of flocculation increased in parallel with increasing mechanical agitation.

On the face of it, such observations are in conflict with the real world of brewery fermentations where mechanical mixing is either unavailable or ineffectual. However, to put 'agitation' into perspective, Stratford and Keenan's (1987) experiments involved relatively gentle mixing (70-120 rpm) to trigger flocculation. Consequently, as brewery fermentations are indirectly 'mixed' though the release of carbon dioxide, it is easier to integrate the involvement of agitation in flocculation. Indeed, at the 'mid-point' of fermentation, flocculation starts against a background of highly active 'mixed' fermentation.

The bridge-like bond between the lectin-like protein and receptors confers the strength and stability of the floes. There is evidence that hydrogen bonding is important in floe stability. Agents which disrupt hydrogen bonding - such as urea or elevated temperature (50°C) - disrupt floes (Speers et al., 1992) or the interaction in vitro between isolated (putative) lectin and mannose (Shankar & Umesh-Kumar, 1994).

It would be anticipated that the number of interactions between flocculating cells would be substantial. As perhaps a reflection of the technical challenge, relatively few attempts have been made to quantify these interactions. A preliminary report noted that up to five contacts per individual cell have been estimated in a single plane (Miki et al., 1982). What this means in terms of cell-cell interactions is not clear but may be gauged from observations quantifying the number of lectin sites per cell. Elegant work with fluorescent mannose or galactose probes (Masy et al., 1992) suggests the number of lectins per cell ranges from 4 x 106 (Flol strain) to 2 x 107 (NewFlo strain). Other work (Speers, unpublished) using fluorescent avidin (that contains dimannose) suggests comparable lectin numbers (4 x 106 per cell). It is not clear how many lectin molecules on a cell actively bind to receptors on other cells. The qualitative and quantitative contribution of mannan receptors (as 'mannoproteins' - see Section might suggest that the stability of flocculation is due to the number of cell to cell interactions. Genetics. The genetics of flocculation (for a review see Jin & Speers, 1998) in many ways exemplifies our understanding of the flocculation process in general. Like the measurement of flocculation, the genetics of the process is subject to great but confusing activity. The MIPS search reported in Table 4.17 is revealing in both the number of FLO genes and the debate as to their role. Although by no means definitive, there is growing evidence that the product of FLOl is a cell wall protein involved in flocculation. The work of Teunissen and Steensma (1995) is persuasive in that this gene confers flocculence when introduced into a non-flocculent strain. Further, flocculation was strongly correlated with the amount of FLOl protein that was expressed. More recently, Javadekar et al. (2000) have provided further fuel for the role of this gene in flocculation. These workers isolated a cell surface lectin that had at least 70% homology with the predicted N-terminal sequence of the putative FLOl as well as FL05 gene products. This protein, which was isolated from a highly flocculent strain of S. cerevisiae, was shown to play a role in flocculation and to bind to the branched trimannoside cell wall receptor. These collective observations suggest that the riddle of flocculation genes is getting ever closer to resolution! Premature flocculation. Although rarely monitored, gross changes in yeast flocculence are observed through increased cell counts at beer run-down (less flocculent) or, conversely, elevated racking gravity (more flocculent). Early or 'premature' flocculation has received greater attention, perhaps because it impinges more on product quality or, simply, is more common!

Although genetic instability may play a role in premature flocculation (see Section 4.4.4), there is growing evidence that wort polysaccharides adhere to yeast cell walls and trigger early flocculation. To explain this, Stratford (1992a) has proposed that large 'multivalent' wort polysaccharides overcome sugar inhibition by binding to cell wall lectins and, consequently, flocculation occurs early.

Work by Herrera and Axcell (1991a and b) showed premature flocculation to be triggered by a lectin-like, gum type polysaccharide present in the malt husk. Addition of the semi-purified polysaccharide to normal fermentations resulted in early flocculation (Herrera and Axcell, 1991a). In an accompanying publication (Herrera & Axcell, 1991b), use of enzyme-linked immunosorbent assays (ELISA) showed the polysaccharide to be present in normal and problem worts. However, more (65%) was found in the wort causing premature flocculation. More recent work (Nakamura et al., 1997) has shown the 'premature yeast flocculation' factor to be present in barley husks. A screening method is described where the propensity of barley to trigger premature yeast flocculation can be determined in four days.

These observations are in keeping with the growing realisation that the subtleties of wort composition have a much greater impact on fermentation performance than previously believed. It is tempting to speculate that such 'premature yeast flocculation' factors already contribute a further layer of complexity to the flocculation story in commercial fermentations. Further work in this area should be encouraged! Hydrophobicity. One of the few certainties in the great flocculation debate is that the cell wall plays an essential role! However, in addition to providing 'lectin-like proteins' and 'receptors', the hydrophobicity of the cell surface has been impli cated in flocculation. The key determinant of cell hydrophobicity appears to be the phosphate content of the outer layer of the cell wall (Mestdagh et al., 1990), hydrophobicity being associated with low phosphate and vice versa. Another view is that hydrophobicity is related to the number of bud scars and, by implication, cell age (Akiyama-Jibiki et al., 1997).

Straver and Kijne (1996) have proposed that 'a high level of cell surface hydrophobicity may facilitate cell-to-cell contacts in an aqueous medium, leading to a more specific interaction between the cells, i.e. calcium-dependent lectin-sugar binding'. Certainly, in some strains of S. cerevisiae, there is a step increase in hydrophobicity at the end of logarithmic growth just prior to the onset of flocculation in aerobic (Smit et al., 1992) or anaerobic cultures (Straver et al., 1993). However, no changes in hydrophobicity were seen during the anaerobic growth of a non-flocculent mutant (Straver et al., 1993). In a later paper, the same group, working with five strains of varying flocculence, demonstrated a positive correlation between hydrophobicity and flocculation (Straver & Kijne, 1996). Similar conclusions were drawn by van der Aar (1996) working with two strains and Azeredo et al. (1997) working with four strains. Particularly convincing evidence of a relationship between cell surface hydrophobicity and flocculation was reported by Akiyama-Jibiki et al. (1997) working with 75 lager strains.

However, others (Garsoux et al., 1993; Dengis et al., 1995; Dengis & Rouxhet, 1997) have been unable to demonstrate any significant differences in hydrophobicity between flocculent and non-flocculent cells. Nor have they been able to associate increased hydrophobicity with the onset of flocculation (Wilcocks & Smart, 1995).

It is not clear why there should be such confusion as to whether (or not) cell surface hydrophobicity is involved in flocculation. Perhaps a simple explanation is the diversity of methods used to quantify hydrophobicity. This ranges from 'water contact angle determinations' (Smit et al., 1992; Garsoux et al., 1993; Dengis et al., 1995; Azeredo et al., 1997) through cell adhesion to polystyrene petri dishes (Amory et al., 1988; Smit et al., 1992) or magnetic beads (Straver & Kijne, 1996) to hydrophobic interaction chromatography (Akiyama-Jibiki et al., 1997). Perhaps more telling is the 'old chestnut' of yeast strain differences. This, of course, can be exacerbated where the conclusions are drawn from work with only one or two strains - a point that applies to key papers from both camps (e.g. Smit et al., 1992; Dengis et al., 1995)!

Cell surface hydrophobicity however may explain why some yeast strains are 'top-fermenting'. These strains are more hydrophobic than bottom-fermenting strains (Hinchliffe et al., 1985; Mestdagh et al., 1990; Dengis et al., 1995; Dengis & Rouxhet, 1997) and, as a consequence, are more able to adhere to carbon dioxide bubbles and to form yeast heads at the top of the fermenter. Inevitably these observations are linked to flocculation as the strains used by Dengis and co-workers are those unusual yeasts described as being 'mannose insensitive' (see Section Zeta potential. Measurement of zeta potential of brewing yeast and its involvement (or not) in flocculation has, in recent years, attracted much attention. Zeta potential is a measure of surface charge (for a full explanation see Lawrence et al., 1989) which is primarily determined by phosphate found in cell wall manno-

proteins (Section (Speers et al., 1992) and the pH of the surrounding medium (Lawrence et al., 1989). Indeed, a curvilinear relationship has been found between increasing phosphate concentration and progressively more negative zeta potential (Amory et al., 1988; Mestdagh et al., 1990).

During fermentation, there some confusion as to whether the zeta potential declines (i.e. becomes more positive) (Lawrence et al., 1989; Bowen & Ventham, 1994) or increases during active fermentation only to decline again at the end (Brown, 1997b). There is general agreement however that 'such a lessening in zeta potential would reduce electrostatic repulsion between individual cells and so favour floccu-lation, even if it is not the cause of flocculation' (Bowen & Ventham, 1994). The same workers also note that at the end of fermentation, trub has a positive zeta potential that they suggest would be attracted to the negatively charged yeast cells. Bowen and Ventham (1994) also speculate that trub could enhance flocculation and cite as evidence work on wort proteins triggering premature flocculation (see Section

4.4.7 Measurement

Unfortunately, there is no agreement on a 'standard' flocculation test. The vast majority of workers in this field appear driven to develop their own method, which frequently involves a minor modification of an existing method. This frustrating state of affairs was succinctly captured by Speers and Ritcey (1995) who noted that between 1990 and 1995, '17 different flocculation methods (or variations thereof) have been proposed in 30 articles written by 21 authors'. Indeed, the methods issue is now so big that it has spawned a number of reviews that seek to make sense of the diversity of approaches (Stratford, 1992; Speers et al., 1992; Speers & Ritcey, 1995; Soares & Mota, 1997). Arguably, real progress in understanding yeast flocculation awaits the advent of a standard and universally accepted method of measuring yeast flocculation.

Cynically, there are two drivers for the development of a 'new' flocculation test. First, the current approaches simply do not work or are strain specific and, second, the methods are so simple that 'improvements' by fine-tuning are almost irresistible! It may well be that the existing raft of methods, which fundamentally are very similar, will never (irrespective of fine tuning) provide a definitive, standard method.

The vast majority of methods are derived from the test of Burns (1941) (see Section who simply washed yeast and resuspended (at 5% w/v) in an acetate buffer at pH 4.6. After 10 minutes standing, flocculation (or as Burns noted 'clumping power') was determined by measurement of the sediment. As with all subsequent flocculation methods, this test tracks the transition from a population of exclusively single cells to a mixed population of single cells and flocculated clumps (see Fig. 4.42). However, as noted by Stratford (1992), quantification of flocculation requires two of the three possible measurements from the relationship:

Free cell fraction + Flocculated fraction = Total cell count

Many of the methods used today operate with a standard total cell count and quantify the 'free cell fraction' after a period of flocculation. Typically, for convenience, cell counts are derived from measurements of absorbance in a spectro-

Single cell fraction 9 •

Single cell fraction 9 •

Flocculated fraction

Equilibrium position im t

Flocculated fraction


Fig. 4.42 Phase diagram during flocculation.

photometer. Flocculation is then expressed as a percentage of settled or flocculated cells. Alternatively, some publications describe methods where flocculation is assessed in situ in a spectrophotometer. In such circumstances, flocculation is quantified by measurement of the rate of change in absorbance. Current methods. Although there are a host of variants, there are essentially three methods used to measure flocculence (see Table 4.22). The 'Helm' sedimentation test (Helm et al., 1953) - which is an extension of the Burns method - is perhaps the most popular method, having been subject to numerous modifications and 'improvements' (see Speers et al., 1992; D'Hautcourt & Smart, 1999). As perhaps a measure of the method's universal acceptability, the Helm test is 'recommended' in its original subjective form by the Institute of Brewing (1997) and, as a more objective absorbance method, by the American Society of Brewing Chemists (Bendiak, 1994, 1996). The 'Stratford' approach although by no means as popular has been used in support of an impressive body of work on the physiology of yeast flocculation (Stratford, 1989; Stratford & Assinder, 1991; Stratford, 1993; Stratford & Carter, 1993).

Both of the above methods operate at the end of an in vitro flocculation process, when equilibrium has been achieved between the single-cell fractions and floes (see Fig. 4.42). The third approach to the measurement of flocculence (Miki et al., 1982) determines the (initial) rate of flocculence directly by change in absorbance in a spectrophotometer. This approach has also been subject to numerous (minor) modifications and has found wider usage in the more fundamental, academic studies on yeast flocculation.

Ironically, given the seemingly never-ending debate about the measurement of flocculation, the above three approaches have surprisingly much in common! As can be seen in Table 4.22, the three approaches agree on the need to wash and resuspend (at a pH of 4.6) cells, to add calcium and to mix to promote the flocculation event. Essentially the differences in protocol are restricted to volume, incubation time and, as noted above, philosophy of measurement at the beginning or at the end of the flocculation process.

Table 4.22 Current approaches to the quantification of flocculation.

Was this article helpful?

0 0
Brew Your Own Beer

Brew Your Own Beer

Discover How To Become Your Own Brew Master, With Brew Your Own Beer. It takes more than a recipe to make a great beer. Just using the right ingredients doesn't mean your beer will taste like it was meant to. Most of the time it’s the way a beer is made and served that makes it either an exceptional beer or one that gets dumped into the nearest flower pot.

Get My Free Ebook

Post a comment