Brettanomyces is a well-adapted fermentation scavenger
The occurrence of Brettanomyces in natural habitats is only sporadically described (Renouf and Lon- vaud-Funel, 2007), but it has been shown that they frequently colonize man-made ecological niches, such as alcoholic fermentation processes (wine, beer, bioethanol, cider, etc.), soft drinks, dairy products, kombucha, and sourdough (Crauwels et al., 2015a; Steensels et al., 2015). A common thread in these niches is the presence of harsh environmental conditions that are lethal for many microbes: high ethanol concentrations, low pH, the absence of readily fermentable nitrogen and carbon sources, low oxygen, etc. While resistance to these stressors is not uncommon in microbes, there are few species that combine all of these traits, thus withstanding such challenging environments. In this section, we will further elaborate on how Brettanomyces evolved to become such a highly specialized fermentation organism and how it has acquired different mechanisms to outcompete, or rather outlive, its main fungal competitor in alcoholic fermentations, S. cerevisiae.
Comparative analysis of the B. bruxellensis genomes revealed some interesting properties that could be linked to their behaviour and ecological niches. First, similar to S. cerevisiae, B. bruxellensis seems to have evolved a mechanism that allows them to accumulate and be highly tolerant to high concentrations of a toxic metabolite (ethanol), even in the presence of oxygen, a strategy that heavily contributes to their dominance over ethanol-sensitive microbes in sugar-rich environments, and is called the ‘make-accumulate-consume' strategy (Piskur et al., 2006). Recently, a study performed by Rozpe- dowska and coworkers (2011) deciphered how B. bruxellensis has evolved this phenotype similarly to, but independently of, Saccharomyces yeasts: both lineages used the same strategy relying on global promoter rewiring to change the expression pattern of respiration-associated genes. Interestingly, B. bruxellensis seems to have evolved an additional strategy to outcompete other microbes. Besides ethanol, they are also capable of producing, accumulating, and later consuming acetic acid in aerobic conditions, and withstand the resulting low-pH environment. The (lineage-specific) duplication of oxidoreductase genes might explain the capacity to produce acetate under aerobic conditions (Curtin et al., 2012; Piskur et al., 2012). It is important to note, however, that not all Brettanomyces species share this trait. For example, B. naardenensis, which separated approximately 100 million years ago from B. bruxellensis, is unable to grow in the absence of oxygen and their metabolism is completely respiratory (and thus no ethanol or acetic acid is formed), indicating that they are Crabtree-negative (Rozp^dowska et al., 2011).
Second, it was shown that B. bruxellensis is well-equipped to withstand nutrient-poor environments. For example, many B. bruxellensis strains are equipped with a gene cluster containing a nitrate transporter, nitrate reductase, nitrite reductase, and two Zn(ll)2 Cys6-type transcription factors, which enables the utilization of nitrate as a sole nitrogen source (Borneman et al., 2014; Crauwels et al., 2014, 2015b; Woolfit et al., 2007). This trait might provide an important fitness advantage over other species, such as S. cerevisiae (typically unable to utilize nitrate), in low-nitrogen environments like molasses, soft drinks, or late stages of beer and wine fermentations. Additionally, B. bruxellensis lacks (or only shows very limited) glycerol 3-phosphate phosphatase activity, resulting in no (or very low) glycerol production (Tiukova et al., 2013; Wijsman et al., 1984). This trait can be disadvantageous in certain conditions (e.g. it reduces growth speed in environments deprived from oxygen, since glycerol is an important factor in maintaining the redox balance in anaerobiosis), but it provides a competitive advantage over S. cerevisiae in nutrient-limiting environments (where energy efficiency is pivotal), since glycerol production is an energy-consuming process. All these adaptations to harsh, nutrient- limiting environments support the scavenging lifestyle of B. bruxellensis.
Third, gene content analysis revealed a relative enrichment in cell membrane-related genes compared with two genetically closely related species (Pichia angusta and Komagetaella pastoris) and S. cerevisiae (Curtin et al., 2012). Interestingly, many of the enriched membrane-related genes (e.g. FIG2, FLO1, FLOS, FLO9, HKR1, and MUC1) might be advantageous for survival in wine or beer matured in oak barrels, where they could mediate the adhesion of the cells to the internal wall of the barrel and protect them from washing out during high-pressure cleaning (Christiaens et al., 2012; Verstrepen and Klis, 2006).
Finally, B. bruxellensis shows some interesting trends in carbon-source utilization, which can sometimes even be variable amongst different isolates (see further). Compared to S. cerevisiae, the genome of B. bruxellensis shows a significant enrichment for membrane-associated nutrient transporters and genes involved in the metabolism of alternative carbon sources (such as chitin, N-acetylglucosamine, galactose, mannose, and lactose) (Curtin et al., 2012; Woolfit et al., 2007). Moreover, it shows very efficient sucrose utilization, accounted for by the expression of a high-efficiency sucrose transporter for which no homologues exist in S. cerevisiae. This trait might be key for the high competitiveness of B. bruxellensis in sucrose-based fermentations such as certain types of molasses (De Barros Pita et al., 2011; Tiukova et al., 2013). Additionally, B. bruxellensis shows a higher affinity for glucose in carbon-limiting conditions, possibly mediated by an orthologue of the Candida albicans HGT1 gene, encoding a high-affinity H+-symport glucose transporter (Leite et al., 2013).
In short, the behaviour of B. bruxellensis shows strong similarities with Saccharomyces species, which is reflected in the numerous niches that they share. However, B. bruxellensis seems to have evolved an extensive specialization in more challenging conditions, which are for example created during the course of beverage fermentation processes.
Intraspecific Brettanomyces bruxellensis variability: adaptations to wine and beer environment?
Genetic diversity studies using classic DNA fingerprinting techniques have revealed significant genotypic intraspecific variability of B. bruxellensis (Fig. 6.3) (Conterno et al., 2006; Crauwels et al., 2014; Martorell et al., 2006; Miot-Sertier and Lon- vaud-Funel, 2007; Mitrakul et al., 1999; Vigentini et al., 2012). Moreover, much like in S. cerevisiae, some of these studies reported a correlation between genotype groups of B. bruxellensis and their source of isolation (e.g. beer or wine) (Conterno et al., 2006; Crauwels et al., 2014; Vigentini et al., 2012), suggesting niche adaptation. Furthermore, recently, this correlation was also established for B. bruxellensis phenotypes (Crauwels et al., 2015b). In this latter
Figure 6.3 Brettanomyces bruxellensis phylogenetics. Neighbour-joining tree of B. bruxellensis strains from diverse ecological niches, including beer (indicated in green), soft drinks (orange), and wine (red). Studied strains were genotyped using seven established DNA fingerprinting techniques. Data analysis was performed on the combined dataset as described in Crauwels et al. (2014).
study, the authors performed high-throughput phenotyping experiments (using Biolog Phenotype Microarrays), and identified several consistent differences between strains from different origins, such as wine and beer fermentations. For example, the ability to assimilate particular a- and в-glycosides as well as a- and ^-substituted monosaccharides was shown to be highly variable amongst isolates, but consistent within strains from the same origin. While strains isolated from wine were able to utilize D-galactose, this is not the case for beer isolates. Coinciding with these observations, strains unable to grow on galactose were found to lack at least one of the genes involved in the Leloir pathway of galactose metabolism. Further, brewing strains are, in contrast to wine strains, not capable of using the в-glycoside disaccharides cellobiose and gentio- biose as well as the в-substituted monosaccharides в-methyl-glucoside and arbutine, suggesting that these strains lack the enzyme(s) responsible for the breakage of specific в-bonded sugars. Indeed, WGS revealed that while wine strains contain two (distinct) в-glucosidase genes, beer strains lack one of these genes (Crauwels et al., 2015b). Interestingly, these в-glucosidases are industrially very relevant, since they also play an important role in flavour development of beer and wine (Daenen et al., 2004). They enable the hydrolysis of locked, gly- cosidically bound flavour compounds and can thus enrich the flavour profile of these fermented beverages (Pogorzelski and Wilkowska, 2007). However, further research is needed to investigate the exact role of these в-glucosidases in flavouring capability of B. bruxellensis strains.
Interestingly, the genes involved in the Leloir pathway as well as the above-mentioned в-glucosidase genes are clustered in a =36 kb region encompassing 13 genes, the majority of which are involved in carbon metabolism. This region was found to be completely absent in the beer strain ST05/12.22 (Crauwels et al., 2014). Moreover, a more thorough investigation revealed that this cluster has been gradually lost over time in beer strains: some lack only a few genes, other lack all 13 genes, but all beer strains lack the в-glucosidase gene. In contrast, this cluster of genes was entirely present in wine strains (at least in eight of the nine studied strains) (Table 6.3) (Crauwels et al., 2015b). Interestingly, this genomic region is also prone to CNVs and loss-of-heterozygosity (Crauwels et al., 2015b). Based on these findings, it may be speculated that this gene cluster carries a fitness cost (e.g. a metabolic burden) for B. bruxel- lensis in certain fermentation systems such as beer brewing, thereby providing a selective pressure for its loss. These observations are reminiscent of the concerted loss of the galactose catabolism cluster in Japanese Saccharomyces kudriavzevii isolates compared to European isolates (Hittinger et al., 2010). However, further research is required to draw firm conclusions.
Despite the increasing knowledge about these genetic and phenotypic differences between B. bruxellensis strains, only little is known about the behavior of different B. bruxellensis strains in different ecological niches. Recent research by Crauwels et al. (2016) has shown that sugar consumption and aroma production are determined by both the yeast strain and the composition of the medium. Furthermore, this study reinforces the hypothesis of niche adaptation of B. bruxellensis, most clearly for wine strains. For example, only strains originally isolated from wine were able to thrive well and produce the typical Brettanomyces-related phenolic off-flavors 4-ethylguaiacol and 4-ethylphenol when inoculated in red wine. Sulfite tolerance was found as a key factor explaining the observed differences in fermentation performance and off-flavor production.
It is also interesting to note the possibility that the ploidy level may be linked to the yeast's ecological niche. More particularly, triploidy seems to be predominant in the Australian B. bruxellensis population, since it is observed in 92% of all isolates from Australian wines (Borneman et al., 2014; Curtin et al., 2007). Moreover, microsatellite typing hints towards the existence of similar populations in French and South African wine isolates (Albertin et al., 2014). In contrast, the majority of B. bruxellensis beer strains are found to be diploid (Crauwels et al., 2015b). This interesting allotriploid genome structure is not rare in fungi and is evocative of the interspecific hybrids identified in the Saccharomyces sensu stricto clade, such as the lager yeast S. pasto- rianus and the S. cerevisiae/S. kudriavzevii hybrids isolated from wine and ale beer fermentations (Gonzalez et al., 2006, 2008). In the case of Saccha- romyces hybrids, it was suggested that the additional set of chromosomes confers a selective advantage in an industrial environment (see above), but it
Table 6.3 Distribution of 13 clustered genesa across different Brettanomyces bruxellensis strains originating from beer and winebc
aClustered genes in a region of =36 kb, corresponding to a part of B. bruxellensis AWRI 1499 (ST05.12/62) contig AHIQ01000280.
bDetermined by PCR amplification using primers targeting the almost complete ORF, followed by agarose gel electrophoresis. Grey, target band detected; black, target band not detected.
cAdapted from Crauwels et al. (2015b).
remains to be determined whether similar evolutionary driving forces are at play in the allotriploid B. bruxellensis strains. It was suggested that the ability of wine strains to withstand high levels of sulfite, the main antimicrobial agent in wine fermentations, might be at least partially explained by the triploidy state (Borneman et al., 2014; Curtin et al., 2012), but further research is required.