and a fatty acyl-CoA ester (Nordstrom, 1962, 1963, 1964). Second, by esterases working in a reverse direction (Soumalainen, 1981). Schermers et al. (1976) contended that a positive correlation existed between reverse esterase activity and the concentrations of ethyl acetate and isoamyl acetate found in beers, thereby implicating these enzymes in ester synthesis. Nevertheless, it is generally accepted that the weight of evidence favours the first of these routes for ester synthesis in brewing yeasts. Biosynthesis involves two enzymes, an acyl-CoA synthetase and an alcohol acetyl transferase.

RiCOOH + ATP + CoASH R^OSCoA + AMP + PPi Acyl Coenzyme A synthetase RiCOSCoA + R2OH R!COOR2 + CoASH Alcohol acyl transferase

It will be appreciated that the process requires energy and this suggests that esters serve a metabolic role. It is frequently stated in the literature that the acetyl Co-A precursor of esters arises from the action of pyruvate dehydrogenase. In fact, as discussed previously this enzyme is almost certainly inactive under the conditions of fermentation, and therefore acyl-CoA synthetase must be responsible. Evidence for the involvement of the alcohol acetyltransferase has been provided by genetic studies. The gene coding for the enzyme in S. cerevisiae, ATFI, has been cloned (Tamai, 1996) and it has been demonstrated that ester formation is dependent on its expression (Lyness et al., 1997).

Much work has been performed in order to establish a unifying mechanism explaining the formation of esters during fermentation and to provide an insight into how their concentrations in beer may be regulated. In fact, the complete mechanism is still far from clear. It has long been brewing dogma that availability of acetyl-CoA exerts a controlling influence on ester synthesis. This premise precludes any role for reverse esterase activity. Changes in fermentation conditions, which influence the size of the pool of acetyl-CoA, should also affect the extent of ester synthesis. In particular, an inverse correlation between extent of yeast growth and ester synthesis has been demonstrated.

Thurston et al. (1981) have argued that ester synthesis may have metabolic sig nificance in regulating the relative cellular concentration of acetyl-CoA and its precursor, CoASH. Certainly, at the mid-point of fermentation, the specific rate of ester synthesis increases more than two-fold at the same time that acetyl-CoA consumption through lipid synthesis ceases. Further evidence for the importance of maintaining an appropriate ratio of CoASH and acetyl-CoA was provided by the demonstration that, at the same time as the induction of ester synthesis, an acetyl-CoA hydrolase was induced (Thurston et al., 1981). Although it is tempting to propose that ester synthesis enables 'fine tuning' of the cellular concentration of acetyl-CoA and CoASH, definitive evidence required the measurement of these metabolites during fermentation. Unpublished work (Pat Thurston and David Quain) reviewed by Quain (1988) confirmed that the intracellular concentration of acetyl-CoA increased in response to the shutting down of lipid synthesis. Indeed, the levels of this metabolite changed in tandem with the specific rate of ester synthesis. During the decline in both the rate of ester synthesis and in the level of acetyl-CoA, CoASH increased significantly.

The hypothesis that acetyl-CoA availability controls ester formation may be explored by examining the effects of fermentation conditions. With respect to wort composition, an increase in amino nitrogen concentration leads to increased ester synthesis. This may be interpreted as increased availability of acyl-CoA esters due to high rates of dissimilation of wort amino nitrogen. Conversely, where wort with low levels of assimilable nitrogen are used, such as might be the case with a high-gravity wort containing a large proportion of adjunct, yeast growth is restricted and in consequence the acetyl-CoA pool is also depleted.

Wort components, which promote yeast growth, tend to decrease ester levels. Thus, high trub levels are often accompanied by low ester levels. This can be explained in that trub may exert general stimulatory effects on yeast growth by favouring removal of carbon dioxide due to its ability to provide nucleation sites. In addition, trub may contain lipids and zinc, both of which promote yeast growth.

Increased provision of wort oxygenation encourages yeast growth and this is also accompanied by a reduction in ester levels. Again a plausible explanation is that because of the increase in yeast growth a large proportion of the acetyl-CoA pool is utilised for biosynthetic reactions and thereby the quantity available for ester formation is reduced. Changes in yeast pitching rate may also influence ester formation. In this instance the effects are less predictable. Maule (1967) reported that ester levels were reduced by large increases in pitching rate. A large increase in pitching rate would be expected to result in an overall reduction in yeast growth extent due to a restricted supply of oxygen available to each yeast cell and this should lead to elevated ester levels. However, the presence of high cell concentrations early in fermentation would be predicted to lead to rapid utilisation of wort amino nitrogen and oxygen for biosynthetic reactions. In which case the acetyl-CoA pool may be depleted and ester synthesis reduced.

Fermentation temperature and ester formation are directly related. The explanation for this effect is unclear. Peddie (1990) considered that higher temperatures could make cell membranes more fluid and this could possibly modulate the activity of the membrane-bound alcohol acetyl transferase or simply increase diffusion rates of esters from cells into the beer. However, elevated temperature also promotes increased formation of higher alcohols and this would provide greater concentrations of pre cursor for ester synthesis. In fact, the explanation for the effects of fermentation variables on ester formation is more likely to reside in control of the activity of alcohol acetyl transferase, as opposed to regulation via availability of substrate. In particular, the effects of lipids on ester formation appear to be crucial.

Alcohol acetyl transferase has proven a difficult enzyme to study since it is unstable in cell-free extracts. The instability is a consequence of its association with membranes. Thus Minetoki et al. (1993) reported that the enzyme is associated with the microsomal fraction of the yeast cell. It is well recognised that such proteins tend to be unstable when solubilised. Minetoki et al. (1993) purified the enzyme from a strain of S. cerevisiae used for saké production. In this beverage isoamyl acetate is the main flavour-active ester and it was reported that isoamyl alcohol was the best substrate, with acetyl-CoA for the alcohol acetyl transferase. The Km values for acetyl-CoA and isoamyl alcohol differed by nearly 200-fold, indicating, in the opinion of the authors, that supply of isoamyl alcohol and not acetyl-Co A would be rate-determining. It was suggested that varying substrate affinities could account for altered patterns of ester formation associated with different yeast strains. It was also demonstrated that the enzyme was strongly inhibited by unsaturated fatty acids.

Several reports (Thurston et al., 1982; Nakatani et al., 1991) have observed that ester synthesis occurs at low rates during the period of active yeast growth and then proceeds at high rates when yeast growth rate declines in mid to late fermentation. These authors claimed a correlation between the time of onset of rapid ester formation and cessation of lipid synthesis. It was speculated that the ratio of acetyl-CoA to free Co A could exert a regulatory influence on control of carbon flux between ester formation and lipogenic pathways. Crucially it was demonstrated that supplementation of worts with linoleic acid (50mgl *) resulted in an 80% decrease in ester formation. This was attributed to a possible inhibition of alcohol acetyl transferase.

Yoshioka and Hashimoto (1982a, b; 1984) partially purified alcohol acetyl transferase from brewers' yeast and observed inhibition by unsaturated fatty acids. Furthermore, activity could be correlated with the unsaturated fatty acid concentration of the cell membrane. It is suggested, therefore, that ester formation and lipid synthesis are related in an inverse manner. This allows a re-interpretation of the mechanism by which changes in fermentation conditions influence ester formation. Notably, the decrease in ester formation due to aeration can be explained in terms of inhibition due to increased synthesis of unsaturated fatty acids.

In fact the real effects of lipids on ester synthesis appear to be more complex than simple inhibition at the enzyme level. Thus the apparent regulatory effects of fatty acids could be interpreted as non-specific inhibition because of the ability of these compounds to inhibit enzyme activity due to non-specific detergent effects (Pande & Mead, 1968). Several authors have reported that oxygen and unsaturated fatty acids exert their effects by repressing the expression of the alcohol acetyl transferase gene. Thus, Malcorps et al. (1991) concluded that the step increase in ester synthesis, which occurred at the end of the yeast growth phase, was due to induction of alcohol acetyltransferase activity. The process required de novo protein synthesis and the induction was prevented by linoleic acid and oxygen. The same regulatory mechanism was considered to be involved for the synthesis of both ethyl and acetyl esters. Fuji et al. (1997) used Northern analysis to show that the alcohol acetyl transferase gene

(ATF1) was repressed by oxygen and unsaturated fatty acids. They concluded that the degree of repression exerted was probably the mechanism by which activity of alcohol acetyl transferase was controlled. Lyness et al. (1997) concluded that ATF1 gene expression and ethyl acetate concentration showed a close correlation; however no similar relationship was demonstrated with isoamyl acetate. These authors observed no effect on ATF1 expression and the addition of linoleic acid.

These apparently conflicting results may be explained in that there is good evidence that yeast cells possess more than one alcohol acetyl transferase and that these have differing substrate specificities. Dufour and Malcorps (1994) concluded that there may be a single enzyme which shows activity exclusively for ethyl acetate formation and another with broader specificity which can produce both acetyl and isoamyl acetate. Working with ATF1 mutants, Fuji et al. (1996) also concluded that the differing patterns of ethyl and isoamyl acetate formation could be explained by the presence of more than one enzyme.

It is apparent that regulation of ester synthesis is complex and intimately related to other pathways that utilise acetyl-CoA, in particular lipid formation. The evidence suggests that alcohol acetyl transferase is a key activity and several distinct enzymes may be present in the cell, each responsible for the formation of one or a few different esters. The activity of some, but not all, of these enzymes is regulated at the gene level, in a negative fashion, by oxygen and/or unsaturated fatty acids. Alcohol acetyl transferases may also be regulated by the availability of the alcohol substrate. In addition, it is not possible to exclude the role of esterases either acting in a reverse synthetic direction or selectively hydrolysing specific pre-formed esters. The characteristic patterns of esters formed by different yeast strains largely reflects the spectrum of alcohol acetyl transferases present, their substrate specificity and the precise manner of their regulation in vivo.

The metabolic role for ester formation is unknown. A plausible explanation is that it provides a route for nullifying the toxic effects of fatty acids. Thus, when lipid synthesis ceases, for example, due to oxygen depletion, the cell continues to produce medium-chain length fatty acids from acetyl-CoA. These are potentially toxic and the cell counters this threat by esterification and removal by diffusion into the medium.

3.7.4 Carbonyls

Some 200 carbonyl compounds have been detected in beers (Berry and Watson, 1987). Of importance to beer flavour and aroma and influenced by yeast metabolism are acetaldehyde and several other aldehydes and VDK.

Aldehydes have flavour threshold concentrations significantly lower than the corresponding alcohols. Almost without exception they have unpleasant flavours and aromas, variously described as 'grassy', 'fruity', 'green leaves' and 'cardboard', depending on the actual compound (Meilgaard, 1975). An aldehydic note is characteristic of the aroma of wort. This character is lost during a normal fermentation. In the case of low or zero alcohol beers that are made by limited fermentation, the 'worty' notes may be retained and this is considered an undesirable feature of such beers. Apart from aldehydes which are formed during wort mashing and boiling, others may arise as part of the catabolic and anabolic routes for higher alcohol formation, as described in Section 3.7.2. In beer, carbonyls form addition compounds with sulphur dioxide. In this form they may not be available for enzymatic reduction.

Reduction of aldehydes by yeast during fermentation is complex and apparently involves several enzyme systems with differing substrate specificities. Alcohol dehydrogenases are involved. Debourg et al. (1994) demonstrated that yeast with fermentative metabolism was more efficient at reducing aldehydes than respiratory sufficient cells. This was explained in that in fermentative cells the ADHI gene is fully expressed and the alcohol dehydrogenase coded for by this gene is the one responsible for aldehyde reduction. By implication the products of ADHII and ADHIII, the respiratory alcohol dehydrogenases, are not involved in aldehyde reduction.

In other reports the same authors have confirmed that alcohol dehydrogenase is active in reducing wort aldehydes, in particular pentanal and pentenal. In addition, several other enzyme activities have been detected. There were two NADPH reductases specific for 3-methylbutanal and one for pentanal. Aldo-ketoreductases with broad specificity were also detected (Colin et al., 1991; Debourg et al., 1993; Laurent et al., 1995).

Acetaldehyde must be considered separately to other longer chain aldehydes because of its importance as an intermediate in the formation of ethanol or acetate. In some circumstances acetaldehyde may persist in beers at concentrations above its flavour threshold value of 10-20 ppm. In this case it produces an unpleasant 'grassy' flavour and aroma (Meilgaard, 1975).

Acetaldehyde formation in beer occurs in early to mid-fermentation during the period of active yeast growth. Later in the stationary phase acetaldehyde levels usually decline (Pessa, 1971; Geiger & Piendl, 1976). Accumulation of acetaldehyde is dependent on the kinetic properties of the enzymes responsible for its formation and dissimilation, pyruvate decarboxylase and acetaldehyde, and alcohol dehydrogenase, respectively. Alcohol dehydrogenase (ADHI), the fermentative enzyme, exhibits highest specific activity during the period of active primary fermentation. In late fermentation specific activity usually declines (Fig. 3.15).

Two acetaldehyde dehydrogenases are found in S. cerevisiae. One is mitochondrial,

Fig. 3.15 Specific activity of alcohol dehydrogenase (ADHI) and ethanol concentration during a high-gravity lager fermentation (W. Tessier, unpublished data).

activated by K+ and NAD+ or NADP + -linked, and is reportedly implicated solely with oxidative growth (Jacobsen & Bernofsky, 1974). The other is cytosolic, activated by magnesium and NADP +-linked (Dickinson, 1996). As discussed previously, acetaldehyde dehydrogenase probably forms the major route for generation of cytosolic acetyl-CoA in brewing yeast under the conditions of fermentation, suggesting that the cytosolic acetaldehyde dehydrogenase would be involved in this role. However, since this enzyme is NADP +-linked it would not participate in the cycle of NAD + /NADH redox-balancing reactions in which pyruvate dehydrogenase is implicated.

There is evidence that the mitochondrial aldehyde dehydrogenase is active under brewery fermentation conditions (Fig. 3.16). The data presented shows the activity of the cytosolic and mitochondrial acetaldehyde dehydrogenases during the course of a high-gravity lager fermentation. Surprisingly, the cytosolic enzyme is apparently rapidly induced during the aerobic phase of fermentation, reaches a peak of activity in early fermentation and then declines to undetectable levels after some 130 hours. In contrast, the mitochondrial activity could be detected throughout the fermentation and would appear to be implicated with the fall in beer acetaldehyde concentration which is characteristic of the later stages of fermentation. This again serves to emphasise the common theme that certain mitochondrial functions are essential under conditions of repression and anaerobiosis.

Fig. 3.16 Specific activities of the cytosolic (Mg2 + -ALDH) and mitochondrial (K + -ALDH) acetaldehyde dehydrogenases, together with the concentration of acetaldehyde in beer during a high-gravity lager fermentation (W. Tessier, unpublished data).

Fig. 3.16 Specific activities of the cytosolic (Mg2 + -ALDH) and mitochondrial (K + -ALDH) acetaldehyde dehydrogenases, together with the concentration of acetaldehyde in beer during a high-gravity lager fermentation (W. Tessier, unpublished data).

High levels of acetaldehyde in beer at the end of fermentation are associated with non-standard performance. This could take the form of premature separation of yeast from wort due to induced changes in yeast flocculence, such that metabolism of the pre-formed acetaldehyde pool is restricted. It may be indicative of a loss of fermentation control. Use of too high a temperature, over-oxygenation of worts and high pitching rates are all reported to result in high beer acetaldehyde levels (Geiger & Piendl, 1976). More commonly it is indicative of poor yeast quality. In the latter case it is not easy to ascribe simple cause and effect. Thus, a failure to provide adequate control of fermentation conditions may have stressed the yeast such that acetaldehyde metabolism is impaired and in consequence it may accumulate to abnormally high concentrations in beer.

Conversely, it may be the formation of acetaldehyde that provides the initial stress to the yeast. Jones (1989) has argued that acetaldehyde is highly toxic to cells. Thus its ability to form Schiff bases with amino residues can lead to inactivation of enzymes and disruption of synthesis of proteins and nucleic acids. These effects may be manifested as disruption of cell growth, cell death or mutagenesis. This has led Jones (1989) to suggest that acetaldehyde may account for some of the toxic effects ascribed to ethanol. Furthermore, toxic effects of acetaldehyde may explain why ethanol generated intracellularly is apparently more harmful than ethanol added extra-cellularly at the same concentration (Section 3.6.2). Stanley and Pamment (1993) concluded that the toxic effects of acetaldehyde might be magnified since cells accumulated this metabolite. Thus ethanol was shown to exit from yeast more rapidly than acetaldehyde. Vicinal diketones. Vicinal diketones (VDK) have progressively higher flavour threshold concentrations with increase in molecular weight (Meilgaard, 1975). Most are considered to have unpleasant flavours. With respect to beer the two most important members of this group are diacetyl (2,3-butanedione) and 2,3-pentanedione. Both have the flavour and aroma of butterscotch but the threshold concentration of diacetyl is almost ten times lower than 2,3-pentanedione, 0.15 ppm and 0.9 ppm respectively (Mielgaard, 1975). In some beers, for example UK topfermented ales, VDK contributes to the overall palate and aroma. Wainwright (1973) reported that up to 0.6 ppm diacetyl was acceptable in some beers. In some wines up to 8 ppm diacetyl may be found. In lagers they are considered undesirable and an essential part of fermentation management is to ensure that the finished beer contains VDK at concentrations below their flavour thresholds.

The occurrence of very high concentrations of diacetyl during fermentation is indicative of microbial contamination, either with Lactobacillus or Pediococcus spp (see Section 8.1.2). In the past this condition was referred to as 'Sarcina sickness' in which beers developed a characteristic sickly buttery aroma (Shimwell & Kirkpatrick, 1939). Indeed, it was popular belief that all diacetyl arose from the effects of contamination. However, it is now well established that a proportion devolves from the activity of brewing yeast and is part of the normal fermentation process. Peak concentrations of diacetyl and diacetyl precursor (a-acetolactate) produced by yeast during the course of fermentation are within the range 1-5 ppm, depending on the type of beer being produced.

Diacetyl and 2,3-pentanedione arise as an indirect result of yeast metabolism. It is generally accepted that the pathway for the formation of VDK and their subsequent dissimilation is that shown in Fig. 3.17. The precursors are a-acetohydroxy acids, which are intermediates in the pathways in the biosynthesis of valine and isoleucine. During the early to middle parts of primary fermentation some of the intracellular pools of these a-acetohydroxy acids are excreted into the fermenting wort where they undergo spontaneous oxidative decarboxylation to form diacetyl and 2,3-pentanedione. From middle to late fermentation extracellular VDK is metabolised by yeast cell reductases to form acetoin and 2,3-butanediol from diacetyl and 2,3-pentanediol from 2,3-pentandione. These metabolites persist in beers but since they are much less flavour-active than VDK, their presence can be tolerated.


vate a-Ketobutyrate

4 Thiaminpyrophosphate

-acetaldehyde CO, + 2H



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