From phenotyping to whole- genome sequencing

This section describes how techniques used for beer yeast typing and characterization evolved from low-resolution, error-prone, and labour-intensive phenotype-based genotyping to WGS. We focus on three species closely associated with beer brewing: the common brewer's yeast S. cerevisiae (used in ale production), the lager yeast S. pastorianus, and Brettanomyces (Dekkera) bruxellensis, an organism involved in spontaneous beer fermentation.

Before the emergence of molecular typing techniques, early yeast classification and speciation was based on the yeast's morphology and physiological properties (Boulton and Quain, 2009; Kurtzman and Fell, 1997). This is at the least remarkable, as the definition of species relies on the basic principle of genetic isolation, and not on the subjective appraisal and weighting of phenotypic properties (Tornai-Lehoczki and Dlauchy, 2000). Morphological differences between beer yeasts are rather limited; they are all unicellular fungi (Querol and Bond, 2009), the cellular morphology of which can vary from spheroidal to ovoidal, with multilateral budding observed when clonally reproducing. Cells occur isolated, in pairs, or sometimes as short chains or clusters, and in some conditions even as pseudohyphae (Boulton and Quain, 2009; Deak, 2008; Kurtzman and Fell, 1997; Voordeckers et al., 2012a). Besides highly similar cell morphology, brewing yeasts also generally form highly similar white, smooth colonies when plated on agar medium.

The shortcoming of brewing yeast classification based solely on morphological properties was already described by Pasteur in his ‘Etudes sur la biere' (Pasteur, 1876) and he clearly stressed the need for other classification criteria. Therefore, several new and more detailed typing methods were developed for the identification and classification of yeasts in the early twentieth century. These tests were based on phenotypic identification procedures in which several physiological and biochemical tests were used, typically targeting the yeast's ability to ferment different sugars, assimilate (= grow aerobically on) carbon and nitrogen compounds and grow in different stressful conditions (e.g. high or low temperatures or vitamin-free medium) (reviewed in Boulton and Quain, 2009; and Kurtzman and Fell, 1997).

Since there is often a need in the brewing industry to quickly and effectively distinguish ale (S. cerevisiae) and lager (S. pastorianus) yeasts, specific and rapid phenotypic tests to discriminate these two species were developed (Table 6.1) (see Chapter 4). First, these species were shown to differ in their flocculation behaviour. Flocculation is the ability of yeasts to forms flocs (clumps of cells) after the fermentation process. The flocculation process involves flocculins, which are lectin-like proteins that are associated with the cell wall of flocculating cells. These flocculins selectively bind mannose residues present in the cell wall of adjacent yeast cells and are activated by calcium ions present in the medium. Interestingly, S. cerevisiae ale yeast tends to adhere to the ascending CO2 bubbles towards the surface of the fermenting wort, whereas S. pastorianus lager yeast will sediment to the bottom of the fermentation vessel. This property has been ascribed to the different properties of the FLO1/Lg-FLO1 gene. The lager yeast-specific Lg-FLO1 gene can not only bind mannose residues, but also has a high affinity towards glucose residues, and is therefore responsible for the ‘NewFlo' phenotype of lager yeasts, and makes them to floc out to the bottom of the fermentor towards the end of the fermentation (Jin and Speers, 1998; Verstre- pen and Klis, 2006; Verstrepen et al., 2003) (see

Chapter 1). For this reason, S. cerevisiae yeast was dubbed top-fermenting yeast and S. pastorianus is generally known as the bottom-fermenting yeast. Second, they differ in their carbon metabolism. Most notably, S. pastorianus yeasts produce the extracellular enzyme a-galactosidase (melibiase, encoded by the MEL1 gene), which enables the hydrolysis of melibiose into the readily assimilable sugars galactose and glucose, while S. cerevisiae is unable to do so (Boulton and Quain, 2009; Deak, 2008; Gibson and Liti, 2014; Gibson et al., 2013a). Moreover, it was also shown that S. cerevisiae yeast only can partly ferment the trisaccharide raffinose (Deak, 2008). Third, ale and lager yeasts show differences in temperature tolerance. Whereas the optimal fermentation temperature of lager yeasts was shown to be below 15°C, it is generally higher for ale yeasts, typically between 20-30°C (Querol and Bond, 2009). Additionally, ale yeasts are more tolerant towards high temperatures, and can grow at up to 41 or even 42°C, whereas lager yeasts can grow only at temperatures up to 32-34°C (Deak, 2008; Hebly et al., 2015; Meersman, 2011; Mertens et al., 2015).

Since Brettanomyces and Saccharomyces yeasts are genetically so distinct and their characteristics often differ widely, development of quick screening tests to distinguish between the two was relatively straightforward. Especially in the wine industry, where Brettanomyces is a vicious spoilage organism, much research has been dedicated to develop reliable methods for detection and identification of Brettanomyces in the fermentation environment. One of the most widely used methods involves a plating assay on semi-selective media with ethanol as carbon source, combined with bromocresol

Table 6.1 Overview of phenotypic differences between the lager yeast S. pastorianus and the ale yeast S. cerevisiae (adapted from Deak, 2008)

Characteristics

Lager strains

Ale strains

Mode of flocculation

Bottom

Top

Fermentation temperature

4-15°C

15-24°C

Maximum growth temperature

32-34°C

38-42°C

Utilization of maltotriose

More complete

Less efficient

Utilization of melibiose

Yes

No

SO2 production

> 4 mg/l

< 2 mg/l

Fructose transport

Active proton symport

Facilitated diffusion

Sporulation

No

Yes

green and/or phenolic precursors such as hydrocinnamic acids to distinguish the genus Brettanomyces from other yeasts after a long period of cultivation (Rodriguez et al., 2014).

However, it soon became apparent that yeast classification solely based on morphologic and phenotypic characteristics was insufficient to deal with the wide variety of yeast used and found in the fermentation industry. Moreover, diverse process developments and changes in the beer-brewing industry have undermined some of the earlier mentioned physiologic classification markers. For example, the use of large cylindroconical fermentation vessels induces ale yeast to sediment after the main fermentation to the cone, a property characteristic of lager yeast (Boulton and Quain, 2009). Therefore, new techniques, based on the genomic rather than phenotypic features of the yeast, were developed for the detection, identification, and classification of yeasts (Campbell, 1972; Tornai- Lehoczki and Dlauchy, 2000). Moreover, these techniques led to a new area for ecological surveys, and enabled researchers to have a closer look into the population dynamics of fermentative yeasts (Legras and Karst, 2003).

The first developed molecular method for the differentiation between lager and ale beer yeast was based on mitochondrial DNA (mtDNA) restriction profiling (Aigle et al., 1984). Gel electrophoresis of the resulting genomic fragments revealed that there were some clear and consistent differences between ale and lager yeasts. Moreover, the obtained patterns ofbeer yeasts (both ale and lager) were in turn very different to the patterns of non-beer yeasts, suggesting that it also is a good technique to detect possible contaminations.

In the following decades, the portfolio of molecular typing techniques for beer yeast differentiation was further expanded: DNA-DNA homology (Tornai-Lehoczki et al., 1996; Vaughan Martini and Kurtzman, 1985; Martini and Martini, 1987), electrophoretic karyotyping (Tornai-Lehoczki et al., 1996; Vezinhet et al., 1990), random amplified polymorphic DNA analysis (RAPD) (Baleiras Couto et al., 1994), amplification of interdelta regions (Ness et al., 1993), and ribosomal RNA coding DNA restriction fragment length polymorphism (RFLP) (Baleiras Couto et al., 1996; Messner and Prillinger, 1995; Smole Mozina et al., 1997), or a combination of different techniques (Tornai-Lehoczki and Dlauchy, 2000). Later, more advanced DNA-based techniques with higher resolution, such as microsatellite comparison (Goddard et al., 2010; Katz Ezov et al., 2006; Legras et al., 2005, 2007b); restriction site-associated sequencing (RAD-seq) (Cromie et al., 2013); multilocus sequence typing (Bing et al., 2014; Fay and Benavides, 2005; Ramazzotti et al., 2012; Wang et al., 2012); tiling array hybridization (Schacherer et al., 2009) and ultimately WGS (Liti et al., 2009) were developed and established for yeast characterization. The advent of WGS revolutionized the way to investigate and characterize genetic and phenotypic diversity in yeast. The analysis of whole genomes rapidly progressed from the study of one or a handful of yeast isolates to simultaneous investigation of tens or even hundreds of individuals, enabling the development of a population genetic perspective. Examining genome-wide patterns of sequence variation within and between closely related species is providing the first comprehensive view of the evolutionary history of S. cerevisiae, S. pastorianus and

B. bruxellensis.

 
Source
< Prev   CONTENTS   Source   Next >