The ecology of wine fermentation: a model for the study of complex microbial ecosystems.

The general interest in microbial ecology has skyrocketed over the past decade, driven by technical advances and by the rapidly increasing appreciation of the fundamental services that these ecosystems provide. In biotechnology, ecosystems have many more functionalities than single species, and, if properly understood and harnessed, will be able to deliver better outcomes for almost all imaginable applications. However, the complexity of microbial ecosystems and of the interactions between species has limited their applicability. In research, next generation sequencing allows accurate mapping of the microbiomes that characterise ecosystems of biotechnological and/or medical relevance. But the gap between mapping and understanding, to be filled by "functional microbiomics", requires the collection and integration of many different layers of complex data sets, from molecular multi-omics to spatial imaging technologies to online ecosystem monitoring tools. Holistically, studying the complexity of most microbial ecosystems, consisting of hundreds of species in specific spatial arrangements, is beyond our current technical capabilities, and simpler model systems with fewer species and reduced spatial complexity are required to establish the fundamental rules of ecosystem functioning. One such ecosystem, the ecosystem responsible for natural alcoholic fermentation, can provide an excellent tool to study evolutionarily relevant interactions between multiple species within a relatively easily controlled environment. This review will critically evaluate the approaches that are currently implemented to dissect the cellular and molecular networks that govern this ecosystem. KEY POINTS: • Evolutionarily isolated fermentation ecosystem can be used as an ecological model. • Experimental toolbox is gearing towards mechanistic understanding of this ecosystem. • Integration of multidisciplinary datasets is key to predictive understanding.

Combinatorial analysis of population dynamics, metabolite levels and malolactic fermentation in Saccharomyces cerevisiae/ Lachancea thermotolerans mixed fermentations.

The outcome of co- or sequential inoculation of Lachancea thermotolerans in winemaking remains unpredictable due to a lack of integrated data regarding the impact of grape juice composition on L. thermotolerans fermentation behaviour. Here, we investigate the impact of nitrogen composition on fermentation characteristics and aroma compound production in grape juice sequentially inoculated with commercial L. thermotolerans and S. cerevisiae strains. Subsequently, all treatments were subjected to malolactic fermentation (MLF) using two commercial strains of Oenococcus oeni. Addition of amino acids led to faster growth for S. cerevisiae fermentations, compared to the nitrogen-equivalent addition of diammonium phosphate (DAP). L. thermotolerans persistence in the mixed fermentations was significantly higher following DAP addition, with higher glycerol and lactic acid production. Interestingly, the lower total Nitrogen content in DAP-treated musts compared to other treatments did not alter the subsequent growth of S. cerevisiae. MLF was more similar between musts fermented with L. thermotolerans, regardless of nutrient regime, whereas significant differences in MLF completion times were observed for different nitrogen treatments in S. cerevisiae fermentations. Collectively, the data present an integrated view of the impact of nitrogen treatment on multispecies co-inoculation (growth kinetics and aromatic outcomes) and the downstream impact on MLF.

Candida pyralidae killer toxin disrupts the cell wall of Brettanomyces bruxellensis in red grape juice.

AIMS: The control of the wine spoilage yeast Brettanomyces bruxellensis using biological methods such as killer toxins (instead of the traditional chemical methods, e.g. SO2 ) has been the focus of several studies within the last decade. Our previous research demonstrated that the killer toxins CpKT1 and CpKT2 isolated from the wine yeast Candida pyralidae were active and stable under winemaking conditions. In this study, we report the possible mode of action of CpKT1 on B. bruxellensis cells in red grape juice. METHODS AND RESULTS: Brettanomyces bruxellensis cells were exposed to CpKT1 either directly or through co-inoculation with C. pyralidae. This exposure yielded a temporary or permanent decline of the spoilage yeast population depending on the initial cell concentration. Scanning electron microscopy revealed cell surface abrasion while propidium iodide viability staining showed that CpKT1 caused plasma membrane damage on B. bruxellensis cells. Our data show that the exposure to CpKT1 resulted in increased levels of ß-glucan, suggesting a compensatory response of the sensitive cells. CONCLUSIONS: The toxin CpKT1 causes cell membrane and cell wall damage in B. bruxellensis. SIGNIFICANCE AND IMPACT OF THE STUDY: Candida pyralidae shows potential to be used as a biocontrol agent against B. bruxellensis in grape juice/wine.

Adjustment of trehalose metabolism in wine Saccharomyces cerevisiae strains to modify ethanol yields.

The ability of Saccharomyces cerevisiae to efficiently produce high levels of ethanol through glycolysis has been the focus of much scientific and industrial activity. Despite the accumulated knowledge regarding glycolysis, the modification of flux through this pathway to modify ethanol yields has proved difficult. Here, we report on the systematic screening of 66 strains with deletion mutations of genes encoding enzymes involved in central carbohydrate metabolism for altered ethanol yields. Five of these strains showing the most prominent changes in carbon flux were selected for further investigation. The genes were representative of trehalose biosynthesis (TPS1, encoding trehalose-6-phosphate synthase), central glycolysis (TDH3, encoding glyceraldehyde-3-phosphate dehydrogenase), the oxidative pentose phosphate pathway (ZWF1, encoding glucose-6-phosphate dehydrogenase), and the tricarboxylic acid (TCA) cycle (ACO1 and ACO2, encoding aconitase isoforms 1 and 2). Two strains exhibited lower ethanol yields than the wild type (tps1Δ and tdh3Δ), while the remaining three showed higher ethanol yields. To validate these findings in an industrial yeast strain, the TPS1 gene was selected as a good candidate for genetic modification to alter flux to ethanol during alcoholic fermentation in wine. Using low-strength promoters active at different stages of fermentation, the expression of the TPS1 gene was slightly upregulated, resulting in a decrease in ethanol production and an increase in trehalose biosynthesis during fermentation. Thus, the mutant screening approach was successful in terms of identifying target genes for genetic modification in commercial yeast strains with the aim of producing lower-ethanol wines.

Expression of the Aspergillus aculeatus endo-beta-1,4-mannanase encoding gene (man1) in Saccharomyces cerevisiae and characterization of the recombinant enzyme.

The endo-beta-1,4-mannanase encoding gene man1 of Aspergillus aculeatus MRC11624 was amplified from mRNA by polymerase chain reaction using sequence-specific primers designed from the published sequence of man1 from A. aculeatus KSM510. The amplified fragment was cloned and expressed in Saccharomyces cerevisiae under the gene regulation of the alcohol dehydrogenase (ADH2(PT)) and phosphoglycerate kinase (PGK1(PT)) promoters and terminators, respectively. The man1 gene product was designated Man5A. Subsequently, the FUR1 gene of the recombinant yeast strains was disrupted to create autoselective strains: S. cerevisiae Man5ADH2 and S. cerevisiae Man5PGK1. The strains secreted 521 nkat/ml and 379 nkat/ml of active Man5A after 96 h of growth in a complex medium. These levels were equivalent to 118 and 86 mg/l of Man5A protein produced, respectively. The properties of the native and recombinant Man5A were investigated and found to be similar. The apparent molecular mass of the recombinant enzyme was 50 kDa compared to 45 kDa of the native enzyme due to glycosylation. The determined K(m) and V(max) values were 0.3 mg/ml and 82 micromol/min/mg for the recombinant and 0.15 mg/ml and 180 micromol/min/mg for the native Man5A, respectively. The maximum pH and thermal stability were observed within the range of pH 4-6 and 50 degrees C and below. The pH and temperature optima and stability were relatively similar for recombinant and native Man5A. Hydrolysis of an unbranched beta-1,4-linked mannan polymer released mannose, mannobiose, and mannotriose as the main products.