Although useful in broadly classifying brewing yeast strains, the real value of this approach is in selection of strains that are 'fit for the purpose'. Strains in group 1 (19% of 235 strains tested) are clearly best suited to fermentation in conical vessels (non-head forming and flocculent) whereas strains in group 5 (38%) are most suitable in traditional open vessels where cropping is through skimming the yeast head. Assimilation/fermentation. Taxonomists have long used fermentation and assimilation tests as a route to typing yeasts (Barnett et al., 1990). Typically, fermentation is judged by the formation of carbon dioxide (usually in a Durham tube) from a carbon source under anaerobic or 'semi-anaerobic' conditions. Conversely, assimilation of a carbon source is an aerobic process that is usually judged by measurement of yeast growth. Although normally performed in liquid culture (510 ml), assimilation can also be assessed on solid media. For example, auxanograms where pour plates of yeast nitrogen base (without a carbon source!) are seeded with crystals (approx. 5 mg) of each test compound on the agar surface. Where assimilation occurs growth is observed. Another approach - replica plating - involves the transfer of colonies from a master plate to a variety of agar plates containing different carbon sources. Transfer is achieved by placing a sterile velvet pad on the master culture, which is then used to 'inoculate' the test plates.

Although seemingly simple and effective, both fermentation and assimilation tests have been found to be technically wanting! The methods' very simplicity can be their undoing inasmuch that results from such tests are frequently unreliable and often ambiguous. For example, Barnett et al. (1990) note that the interpretation of fermentation tests in Durham tubes can be complicated by any aerobic metabolism of the test compound, background growth on contaminating substrates, flocculation and protracted incubation time. Similarly, Visser et al. (1990), in a wide ranging study of 75 yeast genera, noted the need for genuine anaerobic conditions and that visible gas production was unreliable as a criterion for fermentation - the measurement of ethanol being a more reliable approach.

Liquid culture assimilation tests require even more care! As noted by Barnett et al. (1990), the limited oxygen transfer compromises the use of static, tilted test tubes to test the aerobic assimilation of carbon sources. Although rocking/agitation will undoubtedly improve matters, the preferred route is to move from test tubes into triangular/conical flasks that are agitated in orbital incubators or shaking baths. Indeed, aerobic assimilation tests can only reliably be performed under truly aerobic incubation conditions.

These and other complications, were graphically demonstrated by Quain and Boulton (1987a) who studied the growth and metabolism of the sugar alcohol -mannitol - by over 40 strains of S. cerevisiae. The assimilation of mannitol was shown to be obligately aerobic, with growth simply stopping under anaerobic conditions and restarting when switched back to aerobiosis. Of the 20 or so strains capable of growth on mannitol (5%, w/v), some (but not all) were unable to grow at low substrate concentrations (1-2%, w/v). This is important as, typically, in taxonomic assimilation tests the concentration of the carbon source is less than 1% (Lodder, 1970; Barnett et al., 1990). Further, the composition of the media can have a dramatic impact on assimilation. Growth on mannitol could not be demonstrated in fully defined yeast nitrogen base but required the presence of yeast extract and 'salts'. As the taxonomic approach is to use yeast nitrogen base it is perhaps no surprise that mannitol utilising yeasts have been reported as being unable to use the sugar alcohol when subject to 'blind' typing by the National Collection of Yeast Cultures (David Quain, unpublished).

Despite these reservations, the fermentation and/or assimilation of various carbon sources by strains of S. cerevisiae has been described in the taxonomic texts as being 'variable' (Lodder, 1970; Barnett et al., 1990, Table 4.10). Although fermentation based tests have little application in the differentiation of brewing yeast strains, assimilation of various carbon sources may provide simple differentiation between strains. For example, Kirsop (1974a) was able to segregate 130 strains of S. cerevisiae into 25 groups dependent on their assimilation of melizitose, a-methyl D-glucoside, lactic acid, glycerol, ethanol and succinic acid.

One fermentation test that does find application is as noted in Section, the differentiation of lager and ale strains of S. cerevisiae by exploiting the presence of periplasmic a-galactosidase in lager yeasts. These strains are able to hydrolyse melibiose (galactosyl-a-l,6-sucrose) to the fermentable products, sucrose and galactose. Without the a-galactosidase, ale strains exhibit no growth on melibiose. Immunology. Over the years, there has been sporadic interest in using immunological methods to identify and differentiate Saccharomyces strains. More recently, such methods have perhaps found greater application in the identification and detection of beer spoilage bacteria (Eger et al., 1995; Russell & Dowhanick, 1996; Whiting et al., 1999).

Fundamental to immunology is, of course, the antigen-antibody reaction (see Muller, 1991 for a review). An antibody (or immunoglobulin) is a protein synthesised by an animal in response to an antigen or 'foreign' macromolecule. Either antibodies can be highly specific for an antigen (monoclonal) or, alternatively, cheaper 'soups' of polyclonal antibodies can be raised against pure immunogens. In laboratory studies, antibodies are raised from rabbits by immunisation with immunogens (typically in this example, whole yeast cells or isolated cell walls will contain a number of antigens). After about five weeks, the antisera is recovered and treated with ammonium sulphate to isolate the immunoglogulin antibodies. In S. cerevisiae the immunodominant yeast antigens appear to be primarily the cell wall mannan side chains and, arguably, a group of glucan-free proteins (for a review see Fleet, 1991). Clearly, the complexity of the antigenic determinants potentially limits the application of immunology to differentiate brewing yeast. It would appear that there are common antigens as S. cerevisiae can be differentiated from other Saccharomyces species by two universal antigens (Umesh-Kumar & Nagarajan, 1991), whereas Cowan and Bryant (1981) reported 19 different antigens in a study of 43 brewing strains.

Methods differ in how antigen-antibody reactions are visualised and quantified. A range of approaches have been used in brewing microbiology including immunofluorescence (Cowland, 1968; Campbell, 1971; Gilliland, 1971a; Thompson & Cameron, 1971), immunofluorescence coupled with flow cytometry (Eger et al., 1995), immunoelectrophoresis (Cowan & Bryant, 1981). The advent of ELISA (enzyme-linked immunosorbent assay) (Russell & Dowhanick, 1996) has led to the rejuvenation of immunoassay-based methods. This is because ELISA-based protocols are comparatively simple, hugely flexible and, importantly, allow quantification.

It is telling that recent developments such as ELISA have found little application in the differentiation or identification of brewing yeasts. This perhaps reflects the success of other approaches together with fundamental problems inevitable to any phenotypic method of brewing yeast differentiation - the contribution of growth physiology and the close relatedness of brewing yeasts. For example, Cowland (1968) using immunofluorescence, clearly demonstrated that the antigenic determinants of brewing yeast are determined by cell physiology and age. Thompson and Cameron (1971) drew similar conclusions and also reported that the application of immunofluorescence was limited to the differentiation of top and sedimentary yeast strains. Attempts to differentiate within these classes were unsuccessful. Subsequent work using the more sensitive immunoelectrophoresis (Cowan & Bryant, 1981) extended earlier work (Bryant & Cowan, 1979) which had used fermentation characteristics to classify brewing yeasts. Building on this initial work, Cowan and Bryant (1981) raised antisera against representative strains of the five brewing groups (Bryant & Cowan, 1979) which revealed a total of 19 different antigens distributed amongst the 43 strains examined in this study. As with previous work (Thompson & Cameron, 1971) the 'best fit' was between specific antigens and the fermentation properties of top and sedimentary strains (Cowan & Bryant, 1981). Other approaches.

• Temperature - Lager and ale strains of S. cerevisiae can be simply differentiated by incubation at 37 + 0.5°C (Lawrence, 1983). Ale strains (and wild yeasts!) have a higher maximum growth temperature (37.5 to 39.8°C) than lager strains (31.6 to 34.0°C) (Walsh & Martin, 1977). The application of this seemingly simple method has been hampered by what is frequently poor temperature control of laboratory incubators. Ideally, to be meaningful, such studies should be performed in gradient heat blocks (Walsh & Martin, 1977) or gradient water baths (McCusker et al., 1994).

• ' Volatiles' - yeast metabolism during fermentation is responsible for the vast majority of esters and higher alcohols found in beer. Measurement of volatiles -either directly or organoleptically - has been described as part of broad approaches directed toward typing or differentiating brewing strains (Gilliland, 1971a; Thorne, 1975). Typically, analysis of volatiles would play a role in the sort of approaches previously described under 'Fermentation performance' (Section Although something of an aside, the ratio of the higher alcohols, 2- and 3-methyl-l-butanol can be used to differentiate between lager and ale strains of brewing yeast (lager = 0.23+0.026 [«=11 strains], ale = 0.14 + 0.012 [ra = 4 strains], David Quain, unpublished).

• Organic acids - brewing yeast produces a variety of organic acids as byproducts of metabolism. Coote and Kirsop (1974) tracked the formation of organic acids during fermentation and noted significant differences between brewing yeast strains. This potential approach to strain differentiation was extended in a series of publications (Bell et al., 1991a-c) which concluded that the organic acid profile could be used to discriminate after growth on the assimilatory carbon sources described by Kirsop (1974a, see Section but not after growth on glucose.

• Oxygen requirement - the amount of oxygen necessary to support a successful standard fermentation varies between brewing strains (Kirsop, 1974b; Jacobsen & Thorne, 1980). Although of significance in terms of process performance and of interest in terms of yeast physiology, differing oxygen requirements are too cumbersome for the differentiation of brewing yeast strains.

4.2.6 'Modern' methods

In recent years great strides have been made in the differentiation of brewing strains. Unlike the 'traditional' methods, the 'modern' approaches have succeeded in the unequivocal identification and differentiation of individual brewing strains. This success has primarily been a consequence of the explosion of activity in the molecular biology of the Saccharomyces. The development of techniques to probe the yeast genome has spawned spin-off techniques that have been enthusiastically applied by geneticists to the differentiation of brewing yeasts. Although a simple premise that different strains are genetically different, the challenge has been to employ methods that reveal these differences. The three approaches to DNA fingerprinting described below have all been shown to achieve the differentiation of closely related brewing strains. Genetic fingerprinting - RFLP. Arguably one of the most successful approaches to yeast differentiation has been the RFLP (restriction fragment length polymorphism) DNA fingerprinting approach (for applied reviews see Meaden, 1990, 1996; Walmsley, 1994). The application of RFLP DNA fingerprinting to the differentiation of commercial brewing strains has been described (Schofield et al., 1995; Wightman et al., 1996), as has the its role in the routine QA of yeast supply to breweries (Quain, 1995). Not surprisingly, this approach is one of a number of fingerprinting methods that has found favour in characterising and tracking clinical isolates (see Section of S. cerevisiae (demons et al., 1997; McCullough et al., 1998).

The RFLP approach to DNA fingerprinting is neither simple nor quick! Indeed, the introduction of RFLP into a brewing laboratory can be costly (hardware) and, from a zero base, can involve a steep learning curve. Although RFLP uses what are now standard molecular protocols, it is prudent to seek hands-on training with subsequent access to guidance and support. This can be important, as experience has shown that the interpretation of routine genetic methods by microbiologists can differ significantly from that of molecular biologists. This can lead to a variety of unfathomable problems and complications that can often be labyrinthine to unravel.

RFLP DNA fingerprinting can be viewed as a series of modular steps (Fig. 4.7) which take in all about five days from start to finish. In outline (see Meaden, 1996 for more detail) DNA is extracted and digested with a restriction enzyme. The enzymes -of which over 100 are now commercially available - are bacterial endonucleases, which cut DNA into fragments. Different restriction enzymes cut DNA at different

Fig. 4.7 The steps in RFLP DNA fingerprinting.

recognition sequences producing a different set of fragments. This soup of restriction fragments is separated by size (range 200-20 kilobases) using gel electrophoresis and is then subject to 'Southern blotting'. This step - named after its inventor, Ed Southern - transfers the separated DNA fragments from the gel to a nylon membrane.

This matrix enables the DNA fragments to be 'probed' with another piece of yeast DNA. This is necessary as restriction patterns alone support little or no differentiation between strains (Casey et al., 1990). The choice of DNA probe (and restriction enzyme) is of critical importance to the differentiation of closely related brewing yeasts. Two multi-locus probes - poly GT (Walmsley et al., 1989; Walmsley, 1994) and Ty 1 (Schofield et al., 1995, Wightman et al., 1996) - have been used which bind, respectively, to the many poly GT tracts or complimentary 'Ty elements' to be found in the yeast DNA fragments. This 'hybridisation' is visualised by radiolabelling or, preferably, by labelling the probe with a plant steroid, digoxigenin that is itself detected by a subsequent colour reaction.

The elegance of the colorimetric reactions that finally result in a 'fingerprint' is worthy of explanation (see Fig. 4.8). The digoxygenin (DIG) labelled probe is bound to an antibody ('ant-DIG') which in turn is conjugated with the enzyme alkaline phosphatase. This enzyme forms an insoluble blue dye when the chromogenic X-phosphate' is added.

The complexity of RFLP DNA fingerprints is shown in Fig. 4.9. Here three restriction enzymes - Hindlll, EcoRl and Pstl - have been used to digest three different ale strains. It is apparent that each enzyme creates a different fingerprint for each yeast strain which, in turn, can differentiate between strains. This approach -Hindlll and Tyl-15 probe - is routinely used to differentiate 24 commercial production yeasts (Quain, 1995) and periodically to successfully identify strains 'blind' (Philip Meaden, unpublished observations). Genetic fingerprinting - PCR. The polymerase chain reaction is a powerful molecular tool that exploits the way that DNA is naturally replicated. It is increasingly widely used in food microbiology (Candrian, 1995), particularly in the detection

DNA on membrane

-DIG label

. probe DNA

chromogenic substrates blue dye blue dye

Fig. 4.8 The colourimetric 'cold' method for visualising an RFLP fingerprint.

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