Rating of the industrial application potential of yeast strains by molecular characterization

  • Alexander Lauterbach
  • Caroline Wilde
  • Dave Bertrand
  • Jürgen Behr
  • Rudi F. Vogel
Original Paper
  • 12 Downloads

Abstract

Each brewing yeast has its own unique impact on the formation of aroma compounds, and thus, on the properties of the final beer. The selection of the perfect strain for a specific brewing process results from physiological properties, which can be elucidated in brewing experiments. These properties result from genetic and proteomic features of each yeast strain. In the current study, 23 blind-coded yeasts were analyzed on a genomic level by microsatellite genotyping at 13 loci, on a sub-proteome level by MALDI-TOF MS, and on their phenotypic property by phenolic off flavor (POF) production assessment. These results were compared with the current application profile of each yeast strain. An expanded MALDI-TOF MS database was used to identify the blind-coded samples on species level, which was achieved to 100%. The samples belonged to top-fermenting Saccharomyces (S.) cerevisiae, bottom-fermenting S. pastorianus and S. cerevisiae var. diastaticus. Different groupings below species level were found with microsatellite analyses (classification on strain level) and MALDI sub-proteome (subdivision of yeasts into groups), which provided a prediction of application potential to beer styles for which they are currently used. The test for POF showed a wide variation and appears to be a strain-dependent property. However, this could serve as a starting point for the classification of yeast strains with respect to their usefulness for the production of specific beer styles or non-brewing applications.

Keywords

MALDI-TOF MS Microsatellite Phenolic off flavor Brewing yeast Saccharomyces 

Introduction

The main yeasts used for brewing are strains of Saccharomyces (S.) cerevisiae (top-fermenting/ale) or hybrids of S. pastorianus (bottom-fermenting/lager) [1]. For the fermentation of lager beer styles different strains of S. pastorianus are used, which have either flocculent or powdery properties [2]. Other properties of the lager yeast strains can be assigned to the genomic background: complete fermentation of raffinose (MEL1 & SUC2) [3], production of higher levels of sulfur dioxide and hydrogen sulfide than S. cerevisiae strains [4, 5], or enzymatic reduction of dimethyl sulphoxide to dimethyl sulfide [6].

Regarding S. cerevisiae strains, they possess a wider equipment of enzymes to create more fermentation by-products then lager strains like higher alcohols, ester, organic acids, vicinal diketone, sulfur compounds and aldehydes [7]. Fruitiness can be explained by the formation of aroma active esters, which are separated into two groups namely acetate esters and ethyl esters [8]. Higher alcohols define the final product as well and are formed via the Ehrlich pathway [8]. The clove-like aroma is associated with the production of phenolic off flavours (POF) [9]. Phenolic acids like ferulic acid are decarboxylated to volatile compounds such as 4-vinylguaiacol (clove) by the enzymes phenylacrylic acid decarboxylase (PAD1) and ferulic acid decarboxylase (FDC1) [10].

Considering the formation of metabolic compounds, a review on how brewing yeast strains, which result in unique aroma profiles, correlate to different beer styles, can be beneficial. For example, wheat beers are usually made with wheat beer strains, which introduce a clove to fruity flavor to the final product [11]. Likewise, the aroma profile of German Alt-Kölsch [12] beer styles is influenced by German Alt-Kölsch S. cerevisiae strains [13]. The flavor of these beers is associated with fruitiness, but tends to sulphurous notes, originating from the brewing yeast strain [14]. Finally, ale beer styles have a more or less fruity aromatic impression [15] which can be imparted to ale yeasts strains produce high concentrations of aroma-active compounds, and define a unique character to each ale beer style [16].

The characterization of Saccharomyces yeast strains for brewing or other food applications is important to be able to select the most appropriate strain(s). Over the years, we can observe an increasing number of publications on various Saccharomyces strains for industrial applications, aiming at understanding the importance of yeasts for mankind. For instance, studies looked for new hybridization events [17], a possible domestication [18] or the fermentation performance of selected strains [19]. This field is reflected below and gives an overview about genetic and non-genetic methods used to analyze yeasts of the Saccharomyces genus, and especially S. cerevisiae. Microsatellite loci analysis was used to match S. cerevisiae strains to their origins in bread, beer, wine, sake or flor aging [20, 21]. Likewise, genetic diversity and population structure among S. uvarum strains were analyzed by microsatellite genotyping, revealing fewer differences between strains of various origins, as compared to S. cerevisiae [22]. The use of genome sequences and phenomes of industrial S. cerevisiae strains showed a strong distinction between ecotypes [12, 18]. Thanks to those analyses, a novel crossing event was discovered among wheat beer strains of S. cerevisiae. Gallone et al. [18] showed that those wheat beer strains result from a cross of wine and ale strains of S. cerevisiae, which leads to a mosaic genome. Other methods described similar industrial Saccharomyces yeast strains, i.e., random amplified polymorphic [23, 24], pulsed-field gel electrophoresis [25, 26], pyrolysis mass spectrometry and Fourier transform infrared spectroscopy [27] and amplified fragment-length polymorphism [28, 29].

Regarding their phenotypic characteristics, the fermentation performance of some species of Saccharomyces was analyzed to characterize their impact on the final beer products. Gibson et al. [19] performed physiological and fermentation experiments to analyze S. pastorianus (Saaz/Frohberg) and S. eubayanus strains. It was found that beers produced with Saaz strains contain less aromatic compounds then those made with Frohberg or S. eubayanus strains. Likewise, large differences in fermentation performance were detected between those types [19]. Meier-Dörnberg et al. [16] applied genetic and phenotypic methods to characterize five ale yeast strains of S. cerevisiae. Genetic differences among all yeast strains were demonstrated at the IGS2-314 locus while fermentation dynamics, flocculation behavior and beer flavor varied considerably [16]. The flavor of the beers ranged from floral to fruity to clove (POF) [16]. Similar observations can be made for wine [30] or cider [31] applications.

In times of craft brewing, there is high interest in novel yeast strains able to produce different metabolic compounds. The detection of one metabolic activity can be shown by the production of POF. Since presence of POF is either desirable or unwanted in the final product, it can be used to differentiate within the species of S. cerevisiae. Different approaches exist for POF detection, namely chromatographic analysis [32, 33], plating tests combined with sniffing [12, 16] and a novel high-throughput absorbance-based screening method [34].

Besides genetic, phenotypic, and fermentation analyses, various experiments were realized to describe the yeast proteome. For example, proteome analyses were realized for a lager yeast to identify its parental strains [35], to characterize its response to stress [36, 37], or to compare transcriptomic and proteomic approaches between two commercial yeast strains [38]. Matrix-assisted laser-desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) and two-dimensional gel electrophoresis were used to examine the proteome of distiller’s yeast [39]. The Microflex LT (Bruker Daltonics) or VITEK MS (bioMérieux) MALDI TOF MS system were shown to be rapid and effective tools for biotyping of microorganisms at genus, species or strain levels [40, 41, 42]. With these systems, the mass spectra of unknown microorganisms are compared to reference spectra of known microorganisms from various ecotypes [43] implemented in proprietary databases. Furthermore, the Biotyper (Bruker Daltonics) and SARAMIS (bioMérieux) databases can be expanded with user entries. Biotyper compares peak patterns of sample and reference spectra using peak position, peak intensity, and peak frequency, yielding a log-score value [43]. SAMARI works similarly to Biotyper, but uses a confidence percentage for genus and species identification [43]. MALDI-TOF MS has traditionally been used in the clinical sector [44], but has also found applications in the analysis of food and beverages for the identification of contaminations [41, 45] or starting cultures [46, 47]. Furthermore, it has proven effective in several applications, including the differentiation of yeasts within the Saccharomyces genus [48, 49], the identification of S. cerevisiae and non-Saccharomyces wine yeasts [50], the sub-proteomic fingerprinting-based classification of wine strains to their application potential [51], and the matching of brewing Saccharomyces strains to major beer styles [52].

While genetic, proteomic or physiological properties have been used for the selection of optimal strains for specific brewing applications, few studies combined or compared those approaches and related genetic and proteomic profiles and phenotypical traits to characterize yeast strains. This wide characterization may be powerful enough to reflect the relation of strains to each other as well as to a specific application.

The aim of this study was to characterize a set of 23 blind-coded yeasts at the sub-proteomic (MALDI-TOF MS; Microflex LT), genetic (microsatellite analysis) and phenotypic levels (POF test), and to evaluate the potential of these methods to facilitate the optimal assignment of yeast strains to specific applications.

Materials and methods

Strains

The 23 yeast strains used in this study (blind-coded Lalld39–Lalld61) were provided by Lallemand Inc., Montreal, Canada. A POF-positive S. cerevisiae TMW 3.0250 and a POF-negative S. pastorianus TMW 3.0275 were used as controls for the phenolic off flavor (POF) test. Table S1 (see “Supplementary material”) represents 14 yeast strains of S. cerevisiae and S. pastorianus of different beer styles, which were added to the established MALDI-TOF MS database of brewing yeast stated in Lauterbach et al. [52].

Microsatellite genotyping

Genomic DNA was extracted by a classical phenol–chloroform method, and the microsatellite genotyping analysis was performed on 13 loci, divided into two multiplex PCR: Mastermix 1: C11, C8, YOR267C, C3, ScAAT3, C9, C5, and Mastermix 2: YPL009C, C6, ScAAT5, C4, ScAAT1, YKL172w. The genotyping protocol and the 13 loci were described in Legras et al. [20], Legras et al. [53] and Perez et al. [54]. PCRs were performed using the Qiagen Multiplex PCR kit, and the PCR products were sized on a capillary DNA sequencer (ABI 3500), and analyzed using the Genemapper 5.0 software. The chord distance Dc [55] was calculated between each strain with a laboratory-made program [20]. All trees were obtained from distance matrices derived with neighbor of the phylip 3.5 package, using mega 5.2 for tree-drawing. All trees were rooted by the midpoint method and arranged for balanced shape.

Saccharomyces cerevisiae var. diastaticus identification PCR

The S. cerevisiae var. diastaticus isolates were identified by the presence of an amplicon for the STA1 gene using primers SD-5A and SD-6B [56]. A PCR mastermix was prepared using 10 µM of each primer, 0.125 mM of dNTP, 1.5 mM of MgCl2, 0.5 U of Taq DNA polymerase (Invitrogen), and 200 ng of DNA. The amplification was carried out using the following conditions: denaturation at 94 °C for 40 s, followed by 45 cycles of 94 °C for 20 s, 64 °C for 30 s and 72 °C for 30 s. The final elongation was at 72 °C for 7 min.

Cultivation of the blind-coded yeast strains for MALDI-TOF MS analysis

Yeast strains were stored in glycerol stock media as described by Lauterbach et al. [52]. The yeast strains were grown on YPD (5 g/L yeast extract (Carl Roth GmbH & Co. KG, Karlsruhe, Germany), 10 g/L tryptone/peptone ex casein, granulated (Carl Roth GmbH & Co. KG, Karlsruhe, Germany), 20 g/L glucose (Merck, Darmstadt, Germany) and for agar plates 15 g/L agar; pH 6.5 (± 0.1)) at 30 °C for 2–3 days. The medium was sterilized at 121 °C for 15 min. Sugar was autoclaved separately and added to the media under a sterile bench after cooling to approximately 50 °C.

The cultivation of the yeast strains for identification was done as follows: after streaking and incubation on YPD agar plates, a single colony was picked to inoculate 15 ml YPD media in 50-ml flasks closed with cotton plugs and incubated aerobically at 30 °C for 18 h on a WisML02 rotary shaker with 180 rpm (Witeg Labortechnik GmbH, Wertheim, Germany). Subsequently, the samples were prepared for MALDI-TOF MS analysis and were identified.

The cultivation of the strains for the bioinformatics analysis by MALDI-TOF MS was done as described by Lauterbach et al. [52].

MALDI-TOF MS analysis of the blind-coded yeast strains

The preparation and measurement of the yeast strains by MALDI-TOF MS was done as previously described [52]. In summary, samples were treated with absolute ethanol (VWR, France) to precipitate cell material, after centrifugation, the supernatant was discarded and the pellet was air dried for 30 min. Subsequently, 70% formic acid and acetonitrile was added to the pellet for protein extraction and mixed thoroughly. One microliter of sample was spotted on a MALDI steel target, dried and overlaid with one microliter of alpha-cyano-4-hydroxy cinnamic acid (CHCA for MALDI-TOF MS ≥ 99% (HPLC) (Sigma Aldrich, Darmstadt, Germany)), prepared at a final concentration of 10 mg/ml (50.0% ACN, 2.5% trifluoroacetic acid (TFA) (Sigma Aldrich, Darmstadt, Germany)) and dried.

The analysis was carried out on a Microflex LT MALDI-TOF MS (Bruker Daltonics, Bremen, Germany) [52]. For discriminant analysis of principal components (DAPC), the same number of spectra for each strain was recorded as depicted in Usbeck et al. [51].

In total, nine recorded spectra (three biological replicates with technical triplicates) were applied to identify the blind-coded yeast samples by the Bruker Daltonics MALDI TOF MS database (6903; MALDI Biotyper 3 Version 6.0.0.0) and the expanded Saccharomyces genus database [52].

The calibration and validation of MALDI-TOF MS was done according to Lauterbach et al. [52].

Identification by MALDI-TOF MS

The first approach was to identify nine recorded spectra of each blind-coded sample from the Bruker Daltonics database (6903 entries; MALDI-Biotyper 3 Version 6.0.0.0) at the genus level, and if possible the species level. The software stipulates log-score values for identification to species and genus levels [48]. The range is covered from no reliable identification (0.000–1.699; red), probable species identification (1.700–1.999; yellow), secure genus and probable species identification (2.000–2.299; green) and highly probable species identification (2.300–3.000; green).

After genus classification, the recorded spectra were finally identified at the species level with the established Saccharomyces genus database [52], which was expanded with 14 strains (see “Supplementary material”: Table S1) from 52 to 66 entries. This database contains 38 top-fermenting S. cerevisiae brewing yeast strains (10 wheat beer, 11 ale, 13 German Alt-Kölsch, three Belgian beer styles, one opaque beer), seven S. cerevisiae var. diastaticus [57] and 21 bottom-fermenting S. pastorianus brewing yeast strains with different flocculation behavior (12 flocculent and 9 powdery).

Phenolic off flavor (POF) test

The POF test was realized to determine if the analyzed yeast strains (Table 1 A) have the ability to produce phenolic off flavors with various precursors. The experiment was done according to Michel et al. [58], with some modifications. YM medium contains 3 g/l malt extract (AppliChem GmbH, Darmstadt, Germany), 5 g/l tryptone/peptone ex casein (Carl Roth GmbH & Co KG, Karlsruhe, Germany), 3 g/l yeast extract (Carl Roth GmbH & Co KG, Karlsruhe, Germany), 10 g/l Glucose (Merck, Darmstadt, Germany), 15 g/l agar (Carl Roth GmbH & KG, Karlsruhe, Germany); pH 6.5 (± 0.1), with the sugar sterilized separately and added to the medium under sterile conditions. 5% ferulic acid or cinnamic acid and 1% coumaric acid stock solutions were prepared by dissolving the acids in absolute ethanol, and 2 ml/L of the of ferulic acid or cinnamic acid or 10 ml/L of the coumaric acid stocks were added to the YM agar at 45 to 50 °C under a sterile bench.

The yeast strains were spread on YPD plates from the cryogenic stocks and incubated at 30 °C for 2–3 days. After inoculation, colonies of the pure yeast strains were picked and streaked on YM agar plates containing one of the aforementioned acids. The plates were incubated at 30 °C for 2–3 days and scored by smelling as positive (+, POF present) or negative (−, no POF). Strains TMW 3.0250 and TMW 3.0275 were used as positive and negative controls, respectively. From each strain three colonies were picked and streaked on the three YM agar plates containing different acids to get three biological replicates for each acid.

Data analysis

Bioinformatic analysis of all 690 recorded sub-proteomic spectra, data exportation, pre-processing, and calculations were realized according to Usbeck et al. [59] based on an open shared-root computer cluster using Mass Spectrometry Comparative Analysis Package (MASCAP), which was implemented in Octave software (https://www.gnu.org/software/octave/) [52, 59]. Finally, we used the implementation of discriminant analysis of principal components (DAPC) in the adegenet R package [60] to infer similarity clusters from the recorded spectra [51]. Scatterplots were generated to visualize clustering of the investigated strains. This process discarded those strains which appeared as a discriminant group in one region in a step-by-step manner until the final group of strains remained [51].

Results

Microsatellite classification

The reference method based on microsatellite analysis of the 23 tested strains resulted in the dendrogram displayed in Fig. 1. Using 0.05% similarity as the arbitrary threshold to define biotypes, we divided the 23 strains into 19 groups. Lalld45 and Lalld58 (group E), Lalld43/Lalld53 (group F), and Lalld39/Lalld60/Lalld61 (group O) were above the threshold and were clustered in three groups.

Fig. 1

Strains clustering according to microsatellite analysis of 13 loci. The dashed line at 0.05% similarity displays the arbitrary threshold

Identification of blind-coded yeasts by MALDI-TOF MS

Nine recorded spectra per strain were taken to identify the 23 yeast samples using the Bruker Daltonics database (6903 entries; MALDI Biotyper 3 Version 6.0.0.0) and our in-house database (expanded to 66 entries [52]), as well as the ecotype prediction (Table 1). All strains were classified as Saccharomyces with a probable species identification to cerevisiae with the Bruker Daltonics database while the in-house database showed a clear separation to different species as well as variety. Four out of 23 strains belong to the species of Saccharomyces pastorianus (Lalld42, Lalld54, Lalld55 and Lalld56), two strains (Lalld44 and Lalld50) were matched to Saccharomyces cerevisiae var. diastaticus and 17 strains to Saccharomyces cerevisiae sensu stricto. The average score value increased from secure genus and probable species identification (2.000–2.299) to highly probable species identification (2.300–3.000). By considering the ecotype, we classified four strains to bottom-fermenting, 17 strains to top-fermenting and two strains as “high attenuator” (HT). In the case of the top-fermenting S. cerevisiae strains, it was possible to assign them to major beer styles like wheat beer (Lalld52 and Lalld57) or ale (Lalld41, Lalld43, Lalld51 and Lalld53), and the remaining 11 strains were matched to the beer style of German Alt-Kölsch.

Table 1

MALDI-TOF MS identification and ecotype prediction of the 23 blind-coded strains used in this study

Strain

Bruker Daltonics database (6903 entries)

In-house database of the genus Saccharomyces (66 entries)

Identification

ø Score value

Identification

ø Score value

Application type

Lalld39

S. cerevisiae

2.19

S. cerevisiae

2.32

AK

Lalld40

S. cerevisiae

2.19

S. cerevisiae

2.35

AK

Lalld41

S. cerevisiae

2.08

S. cerevisiae

2.45

Ale

Lalld42

S. cerevisiae

2.05

S. pastorianus

2.55

Lager

Lalld43

S. cerevisiae

2.02

S. cerevisiae

2.19

Ale

Lalld44

S. cerevisiae

2.08

S. cerevisiae var. diastaticus

2.47

High attenuator

Lalld45

S. cerevisiae

2.10

S. cerevisiae

2.31

AK

Lalld46

S. cerevisiae

2.22

S. cerevisiae

2.60

AK

Lalld47

S. cerevisiae

1.98

S. cerevisiae

2.27

AK

Lalld48

S. cerevisiae

2.21

S. cerevisiae

2.38

AK

Lalld49

S. cerevisiae

2.14

S. cerevisiae

2.35

AK

Lalld50

S. cerevisiae

2.02

S. cerevisiae var. diastaticus

2.43

High attenuator

Lalld51

S. cerevisiae

2.22

S. cerevisiae

2.36

Ale

Lalld52

S. cerevisiae

2.09

S. cerevisiae

2.50

Wheat beer

Lalld53

S. cerevisiae

2.03

S. cerevisiae

2.24

Ale

Lalld54

S. cerevisiae

1.96

S. pastorianus

2.55

Lager

Lalld55

S. cerevisiae

1.97

S. pastorianus

2.52

Lager

Lalld56

S. cerevisiae

2.03

S. pastorianus

2.44

Lager

Lalld57

S. cerevisiae

2.00

S. cerevisiae

2.43

Wheat beer

Lalld58

S. cerevisiae

2.03

S. cerevisiae

2.32

AK

Lalld59

S. cerevisiae

2.02

S. cerevisiae

2.40

AK

Lalld60

S. cerevisiae

1.91

S. cerevisiae

2.30

AK

Lalld61

S. cerevisiae

1.88

S. cerevisiae

2.21

AK

The ø score value is an average of nine recorded spectra per yeast strain. Ecotypes: AK = German Altbier-Kölsch

DAPC classification of the recorded MALDI-TOF MS spectra

The classification of the 23 yeast strains was achieved by a DAPC of the recorded MALDI-TOF MS spectra, displayed in Fig. 2. We used DAPC to visualize clustering of our strains and discarded step-by-step those strains which appeared as a discriminant group in one region until no more distinction was achieved. Subsequently, the provided scatterplot allowed for a graphical evaluation of the structures between the groups, enabling the stepwise deletion of separate groups (Fig. 2), which are highlighted with dashed black circles. Because of the high similarity of the recorded blind-coded samples, strains were counted to one group if a two-thirds majority was achieved. First, all S. pastorianus Lalld42, Lalld54, Lalld55 and Lalld56 clustered in a single group, (a), which differs from all other strains. The next step was to remove them from the data set, performing a new calculation and isolating strains Lalld43 and Lalld53 in cluster (b). Afterwards, strains Lalld46 and Lalld58 (c) were removed, followed by group (d) with Lalld40, Lalld45 and Lalld58, as well as Lalld39, Lalld60 and Lalld61 (e). The next eliminations contain strains starting with Lalld51 (f), Lalld44/Lalld50 (g), and Lalld48/Lalld49 (h). The final separation led to groups (i) (Lalld47), (j) (Lalld52 and Lalld57), and (k) (Lalld41). In total, 11 groups were found containing either single strains or groups from 2–4 yeast strains by the DAPC calculation.

Fig. 2

DAPC, displayed as a scatterplot where clearly distinguishable groups were eliminated stepwise. The groups were marked by the names of the included strains and highlighted with black dashed circles; in total 690 single spectra were analyzed in a. All 690 measurements were organized in three groups: b after the elimination of: group a, c after the elimination of group b, d after the elimination of group c, e after the elimination of group d, f after the elimination of group e, g after the elimination of group f, h after the elimination of group g. the last figure shows the separation in three different groups of i, j and k. The axes stand for x-component 1 and y-component 2 for all small figures

Results of POF plating test

Phenotypic property was determined with a POF plating test to reflect the sensory potential of the yeast strains. It was performed by sniffing to detect the production of POF from different precursors and Fig. 3 displays the results. We observed that when a sample exhibited a POF-positive phenotype, it was detectable regardless of the precursor used. Ten out of 23 samples are positive for the production of phenolic off flavors. More precisely, the “high attenuator” group showed a positive reaction (Lalld44 and Lalld50) similarly to the wheat beer strains of Lalld52 and Lalld57. Furthermore, strains, which are predicted to belong to the German Alt-Kölsch style, proved to have POF production like Lalld48, Lalld49, Lalld59 and Lalld47, and the ale strains Lalld41 and Lalld51. All other strains showed no formation of POF.

Fig. 3

Results of phenolic off flavor test on the production of 4-vinylphenol (4-VP), 4-vinylguaiacol (4-VG) and styrene. TMW Technische Mikrobiologie Weihenstephan, YM yeast mold agar, FA ferulic acid, CA coumaric acid, CIA cinnamic acid, positive control = TMW 3.0250; negative control = TMW 3.0275; check mark = smell of 4-VP, 4-VG or styrene; cross = no perception; exclamation mark = low perception; yellow star = produce POF; star half yellow and white = low concentration of POF; white star = no production of POF

Comparison of all typing approaches

Finally, the results of genetic, proteomic and physiological properties were compared to each other and summarized in Table 2. The actual brewer experience (Dr. Tobias Fischborn, Lallemand Inc., personal communication) is listed in Table 2, which is used to correlate our results with the industrial application.

Table 2

Comparison of brewer experience to the results of the molecular characterization

Strain

Species level

Brewer experience

MALDI prediction

Microsatellite grouping

DAPC

POF

Lalld39

S. cerevisiae

Ale

AK

O

e

Lalld40

S. cerevisiae

Ale

AK

E

d

Lalld41

S. cerevisiae

Wheat beer

Ale

H

k

+

Lalld42

S. pastorianus

Lager

Lager

A

a

Lalld43

S. cerevisiae

Ale

Ale

G

b

Lalld44

S. cerevisiae var. diastaticus

HT

HT

R

g

+

Lalld45

S. cerevisiae

Ale

AK

F

d

Lalld46

S. cerevisiae

CF

AK

L

c

Lalld47

S. cerevisiae

BBS

AK

P

i

+

Lalld48

S. cerevisiae

BBS

AK

M

h

+

Lalld49

S. cerevisiae

BBS

AK

N

h

+

Lalld50

S. cerevisiae var. diastaticus

HT

HT

S

g

+

Lalld51

S. cerevisiae var. diastaticus

BBS

Ale

Q

f

+

Lalld52

S. cerevisiae

Wheat beer

Wheat beer

I

j

+

Lalld53

S. cerevisiae

Ale

Ale

G

b

Lalld54

S. pastorianus

Lager

Lager

C

a

Lalld55

S. pastorianus

Lager

Lager

B

a

Lalld56

S. pastorianus

Lager

Lager

D

a

Lalld57

S. cerevisiae

Wheat beer

Wheat beer

J

j

+

Lalld58

S. cerevisiae

Unknown

AK

F

d

Lalld59

S. cerevisiae

Kölsch

AK

K

c

+/−

Lalld60

S. cerevisiae

Ale

AK

O

e

Lalld61

S. cerevisiae

Ale

AK

O

e

All strains are listed with their coding, species or variety level, classification to brewer experience and MALDI-TOF MS. Furthermore, groupings according to microsatellite (capital letters) as well as discriminant analysis of principal components (DAPC; small letters) are included and phenolic off flavor (POF) characteristics positive (+) and negative (−)

Brewer’s experience = Dr. Tobias Fischborn, Lallemand Inc., personal communication

AK German Alt-Kölsch, HT high attenuator, BBS Belgian beer style, CF cold fermentation

Considering the total count of groups, we observed 19 groups based on microsatellite analysis (A–S) while 11 groups (a–k) are detected based on DAPC classification (Table 2). Most of the strains are classified differently by both methods, yet some strains share the same sub-proteomic and genetic groups, which can be retraced in Table 2. S. pastorianus strains were detected with both techniques, which are used for the production of lager beer styles. All strains are POF negative. Industrial ale strains like Lalld43 and Lalld53 formed one cluster in the microsatellite analysis (G) as well as by DAPC (b) and were predicted to the ale style. In fact, both strains are highly related to each other, because it was isolated from the same brewery several years apart. The genetic approach clustered strains of Lalld39, together with Lalld60 and its second-generation derivative Lalld61 (O), as observed in the sub-proteomic analysis (e). DAPC generated three groups of one strain each, namely Lalld41 (k), Lalld47 (i), and Lalld51 (f), which concurred with the microsatellite analysis results. Furthermore, all three strains possess the property to produce POF and are used for the production of wheat beer (Lalld41) or Belgian beer styles (Lalld47/Lalld51). Regarding Lalld51, a difference was observed within the identification and classification. Using the STA1 gene-specific PCR method as described in Yamauchi et al. [56] (data not shown), Lalld51 was identified as S. cerevisiae var. diastaticus, whereas the MALDI-TOF MS comparison did not allow identification at the variety level. The different classifications of Lalld51 are visualized in Table 2, which is listed with the MALDI prediction for an ale style and the brewer experience to a Belgian beer style. On the contrary, ale strains Lalld40, Lalld45 and the unknown Lalld58 were classified to one group by DAPC (d) while microsatellite analysis assigned them to two groups (Lalld40 (E) and Lalld45/Lalld58 (F)). Two out of three wheat beer strains were classified to this beer style by the MALDI-TOF MS approach, which is observed on the forming of two groups by the DAPC and the identification by the database. Lalld59 is applied for the production of Kölsch, which is confirmed to the database comparison. Furthermore, this strain formed one group with the cold fermenter of Lalld46 within the DAPC analysis and one clade to the microsatellite tree. A clear classification of Lalld48 and Lalld49 to one beer style could not be achieved by the database approach. However the DAPC grouped them together, which corresponds to the brewer’s experience, because both strains are used for the production of Belgian beer styles. In total, 11 strains were predicted correct to their industrial application, but using microsatellite as well as DAPC it was possible to characterize each strain to their industrial application as well as their relation to each other.

Discussion

We used different approaches based on sub-proteomic, genetic and phenotypic properties to characterize a set of 23 yeast strains. Different groupings and relations between strains were observed, which partially overlap with their application in beer styles.

Grouping by microsatellite analysis

Multilocus microsatellite typing was used as a genetic approach to classify strains according to their ecotype, and it also showed the unique profile of each strain. Considering the grouping by this genomic approach, three groups were detected that reached the minimum similarity threshold of 0.05%. All other strains do not show significant similarity to each other. Furthermore, the formation of a bottom-fermenting cluster could be observed, which reflects the unique nature of the S. pastorianus strains. Microsatellite analysis is suitable to distinguish S. cerevisiae strains from different origins, which was illustrated by Legras et al. [20] In this study, we could separate yeast strains at the genus, species, and intra-species levels. S. cerevisiae var. diastaticus strains Lalld44, Lalld50 and Lalld51 could be distinguished from the other S. cerevisiae strains, which was similarly observed in the microsatellite analysis of Legras et al. [20]. Within the major beer styles the wheat beer strains formed their own cluster, which is noteworthy because of the mosaic genome of these strains [18]. A wide genetic variation is observed in the remaining strains, which sometimes reflects the unique character. However, a clustering of some groups of strains is still observed. Similar to Legras et al. [53], yeast strains grouped below the similarity threshold of 0.05% show a direct relation to each other, which can occur, i.e., through the same strain or isolation source. In contrast, a grouping above the threshold possibly displays an application type or variety level. This is for example observed for strains used for wheat beer or Belgian beer styles.

MALDI-TOF MS-based clustering

Two databases for identification at the genus and species levels were used to classify yeast samples with nine recorded spectra per entry. The identification of all yeast strains with the current MALDI Biotyper 3 database (6903 entries) showed a correct identification on genus level of Saccharomyces even though only 13 entries of one species are available within the current database version, namely S. cerevisiae. Those strains were from culture collections, isolated from the beverage industry or clinical samples, and were grown on different media. One general reason for the limited recognition of these different yeast isolates at species level may result from the significant influence of growth conditions on the yeast proteome [51, 59]. Nevertheless, the addition of our in-house database of Saccharomyces yeasts from major beer styles [52] to the commercial MALDI-Biotyper database improved the species identification, distinguishing S. cerevisiae, S. cerevisiae var. diastaticus and S. pastorianus. Assignment of the samples to ecotypes was also achieved by taking the established database into account using the in-house database. Thereby, we were able to classify the analyzed yeast strains to beer styles and to visualize those groups by DAPC. All S. pastorianus strains were predicted to the bottom-fermenting lager style. This prediction was confirmed by DAPC, which clearly distinguished S. pastorianus from S. cerevisiae strains. By restricting DAPC to the S. cerevisiae samples [51], we could divide them according to the predicted beer style groups. We were able to identify a recently described wheat beer cross event of S. cerevisiae [18] and show the formation of a separated cluster by DAPC. Two out of the three highly attenuating S. cerevisiae var. diastaticus were identified and formed their own group. Using MALDI-TOF-MS, it was possible to predict that four strains belong to the ale style, and grouped into three clusters. The remaining strains were classified as German Alt-Kölsch strains. However, according to DAPC we could divide those strains into five clusters, allowing visualization of their relatedness.

DAPC proved its usefulness for the analysis of sub-proteome spectra and to distinguish yeast strains from different ecotypes, as successfully demonstrated for Saccharomyces wine strains [51]. Moreover, the identification of unknown samples combined with a prediction of application style enabled a simple overview of the strain grouping. This bioinformatics analysis can be used to confirm known grouping of strains or to further distinguish strains within a larger group, as was applied for the German Alt-Kölsch strains.

Grouping according to phenotypic property

The formation of phenolic aroma compounds was used to distinguish on phenotypic property within the yeast strains. Two major groups were observed, which are either positive or negative for phenolic off flavors. Actually, POF is described as an off flavor in beverages caused by non- or wild-Saccharomyces yeasts [61, 62] and is undesirable in beer styles like lager, German Alt-Kölsch and ale. However, wheat beers are associated to aromatic notes like clove [11, 12]. We found at least ten S. cerevisiae strains, which are able to form POF compounds. Only two out of ten strains were wheat beer strains and all the other strains were predicted for other beer styles or are S. cerevisiae var. diastaticus. One strain, Lalld59, is able to produce only a low amount of POF compounds. It is known that the responsible genes, PAD1 and FDC1 [63], have a wide distribution within different S. cerevisiae ecotypes. Goncalves et al. [12], Meier-Dörnberg et al. [16] and Mertens et al. [34] detected the POF property in various application styles of S. cerevisiae like wheat beer strains and many ale strains. It is described that the sequence of PAD1 and FDC1 are available in different S. cerevisiae strains, but are inactive due to the presence of premature stop codons or frameshift mutations [10, 12, 64].

POF shall be used to characterize S. cerevisiae strains and reflect their impact on the final product. For instance, this property shows that not only wheat beer strains form spicy flavors. It is possible to produce a beer to the wheat beer style without using a traditional wheat beer strain.

Overlap or differences between genetics and proteomics

Both techniques led to different grouping patterns. The approach on microsatellite allowed a classification of the different strains because of the simple sequence repeats (SSR). Those microsatellites are often used as genetic markers for genetic mapping, population genetics or reflection of the biodiversity [53, 65]. Considering the sub-proteome, MALDI-TOF MS can be used in different ways. One approach is to identify unknown samples of microorganisms on genus, species, or strain level, which are compared to an available library of reference spectra from different sectors [41, 47, 66]. This typing allows to visualize the response to stressful environments on sub-proteomic levels [67, 68]. In other cases, MALDI-TOF MS can be used not only to identify microorganisms, but also to predict their specific ecotypes or application potentials [51, 52, 69]. Both, MALDI-TOF MS and microsatellite typing measure elements to classify or identify microorganisms, which are mostly stable and enable a reliable measurement. Because of that, differentiation at the species and even at the strain level can be achieved. In fact, the classification of the blind-coded samples showed a separation of top- and bottom-fermenting yeast strains using both methods. In total, 100% were correctly identified at the species level by MALDI-TOF MS and were confirmed with the microsatellite analysis. However, a 100% differentiation on variety level could not be achieved, because Lalld51 was not recognized as S. cerevisiae var. diastaticus by MALDI-TOF MS.

Considering the grouping of the blind-coded samples, both techniques presented five similar groups. A separation of species was obtained on genetic and sub-proteomic level. The 19 S. cerevisiae strains were differentiated primarily on strain level using genetic markers. The DAPC similarity calculation for the recorded sub-proteomic spectra instead achieved a grouping and matched them to different brewing ecotypes, including the database classification.

Reflection of application range

After we looked for overlaps and differences between the groupings, we integrated the actual application of those strains in order to reflect the prediction of the different typing. This is shown in Table 2, which includes the actual brewer’s experience (Dr. Tobias Fischborn, Lallemand Inc., personal communication) of each strain. Half of our analyzed strains corresponded to their true application, as discussed below.

The characterization of all S. pastorianus as lager beers was achieved with all techniques and correspond to the brewer’s experience. However, further characterization, such as flocculation, could not be achieved by the established database of Saccharomyces. One solution to distinguish them could be classifying S. pastorianus strains to Saaz and Frohberg-types by MALDI-TOF MS. Gibson et al. [19] showed, e.g. the different fermentation performance of Saaz and Frohberg, which resulted in unique lager beer styles. Furthermore, these types of S. pastorianus can also be genetically differentiated, e.g., by their ITS sequences/RFLP profiles [70] and ploidy [71]. Considering this, not only a separation between top- and bottom-fermenting yeasts can be achieved, but S. pastorianus can also be characterized to a focused brewing application using different analyses.

In case of S. cerevisiae, all approaches reflect in different ways the application range to top-fermenting beer styles. Yeasts used for wheat beer production were recognized with all techniques as well as the typical POF characteristic was assigned to those strains. The microsatellite analysis illustrated that all strains used for wheat beer formed one cluster; however, Lalld52 and Lalld57 are more related to each other on a genetic basis than to Lalld41. On sub-proteomic level these strains are assigned to two groups, which was displayed by a wheat beer classification of Lalld52 and Lalld57 as well as an ale matching of Lalld41. Because of the wide distribution of the POF property, it is possible to produce beers of the wheat beer style with different S. cerevisiae strains as illustrated by these three strains.

The group of S. cerevisiae var. diastaticus displays a good coverage within the analysis compared to the industrial experience. Lalld44 and Lalld50 are described as high attenuating S. cerevisiae var. diastaticus, which was confirmed by MALDI-TOF MS as well as genetically by STA1 gene amplification. These strains possess the property to produce phenolic off flavor, which was confirmed by the POF agar test. This phenotypic characteristic is typical for S. cerevisiae var. diastaticus strains [72]. Regarding the sub-proteomic results, it is useful to define a new application style for this variety, namely “high attenuator”, because of the shared property of super-attenuation and because the name diastaticus does not define an application. Considering the results of the gene-specific PCR, Lalld51 was classified to the group of “HT” as well, because of the presence of STA1. Within the microsatellite tree, Lalld51 clustered next to the S. cerevisiae var. diastaticus clade. Furthermore, Lalld51 is applied in the brewing application for a Belgian beer style (Table 2). This shows the contrast within the variety and importance of a good characterization. On one hand, many breweries consider S. cerevisiae var. diastaticus as a contaminant resulting in product damages or loss of image [73]. On the other hand, this variance is used in old beer styles like Saison beers and also gives brewers a tool to create innovative beer styles with unique aromatic profiles. A good characterization of Lalld51 with different molecular techniques illustrates these results: (I) unique microsatellite profile; (II) presence of STA1; (III) a low score value using the database of Saccharomyces, which means that this strain distinguishes on sub-proteomic level to the database entries; (IV) positive for POF.

The four S. cerevisiae strains of Lalld59, Lalld46, Lalld48 and Lalld49 are used in different applications. Lalld59 is actually applied for the production of a Kölsch beer style and Lalld46 is an ale yeast used in cold fermentation. In the case of Lalld48 and Lalld49, both strains are isolates from Belgium, which are POF positive. In case of those two groups detected with DAPC, a rename of the application style might be useful. Lalld48 and Lalld49 represents the “Belgian beer style”, which shall be implemented within the MALDI-TOF MS library [52] for a better detection of this application type. This “Belgian beer style” for the MALDI-TOF MS database needs to include a variation of strains from Lambic, Trappist, Wit beer, Saison-style and more. The expansion will help to classify yeast strains like Lalld47, Lalld48, Lalld49 as well as Lalld51 to specific Belgian styles. Lalld59 and Lalld46 shall be renamed likewise with respect to White, Zainasheff [14], who described the classification of brewing yeast strains based on their fermentation character. They describe a specific group of yeast strains as “Hybrid Ale strains”, because those are strains that brewers commonly ferment at temperatures cooler than the average top-fermenting temperature [14].

Furthermore, microsatellite and DAPC illustrated high relations of four strains. Both techniques enabled the classification of Lalld43 to Lalld53 and Lalld61 to Lalld60, which were re-isolations from two breweries after years of application. Subsequently, strains of Lalld39, Lalld60 and Lalld61 possess the same phenotypic properties. Besides not producing phenolic off flavors, these strains do not use maltotriose and have a low flocculation (Dr. Tobias Fischborn, Lallemand Inc., personal communication).

In conclusion, both techniques reflected very well to the actual application style. The combination with the POF agar test showed the wide distribution of the POF characteristic within the yeast strains. Major differences within the classification on sub-proteomic and genetic level were observed, but both techniques have noticeable advantages. The microsatellite analysis enables an insight into the genetic material and is more useful to distinguish on strain level. Afterwards, it is possible to determine the origin or application style for each strain using MALDI-TOF-MS. MALDI-TOF MS enables an identification of yeasts using a fast and easy sample preparation and a rapid sample measurement. The identification is realized by means of an established database, which differentiates on species or strain level within brewing yeasts. A classification on ecotype or application style is possible simultaneously to the identification. To achieve a better differentiation on variety level, we suggest an ongoing expansion of the MALDI-TOF MS database.

A final test for the fermentation behavior is always needed, but preliminary molecular methods can serve as a starting point for the classification of yeast strains with respect to their usefulness for the production of specific beer styles or non-brewing applications.

Notes

Acknowledgements

Part of this work was supported by the German Ministry of Economics and Technology and the Wifö (Wissenschaftsförderung der Deutschen Brauwirtschaft e.V., Berlin, Germany) in project AiF 17698 N. We would like to thank our technical assistant Sabine Forster, and Dr Tobias Fischborn (Lallemand Inc.) for their support with the experiment, as well as Dr Damien Biot-Pelletier (Lallemand Inc.) and Viktor Eckel (Technische Mikrobiologie Weihenstephan) for critical reading of the manuscript. We are also grateful for the information about the origin of the new yeast strains for the MALDI-TOF MS library expansion from Dr. Mathias Hutzler and Tim Meier-Dörnberg.

Compliance with ethical standards

Conflict of interest

The authors declare that they have no conflict of interest.

Compliance with Ethics requirements

This communication does not contain any studies with human participants or animals performed by any of the authors.

Informed consent

Informed consent was obtained from all individual participants included in the study.

Supplementary material

217_2018_3088_MOESM1_ESM.docx (17 kb)
Supplementary material 1 (DOCX 16 KB)

References

  1. 1.
    Barnett JA (1992) The taxonomy of the genus Saccharomyces meyen ex reess: A short review for non-taxonomists. Yeast 8(1):1–23.  https://doi.org/10.1002/yea.320080102 CrossRefGoogle Scholar
  2. 2.
    Verstrepen KJ, Derdelinckx G, Verachtert H, Delvaux FR (2003) Yeast flocculation: what brewers should know. Appl Microbiol Biotechnol 61(3):197–205.  https://doi.org/10.1007/s00253-002-1200-8 CrossRefGoogle Scholar
  3. 3.
    Donalies UEB, Nguyen HTT, Stahl U, Nevoigt E (2008) Improvement of Saccharomyces yeast strains used in brewing, wine making and baking. In: Stahl U, Donalies UEB, Nevoigt E (eds) Food Biotechnology. Springer, Berlin, pp 67–98  https://doi.org/10.1007/10_2008_099 CrossRefGoogle Scholar
  4. 4.
    Yoshida S, Imoto J, Minato T, Oouchi R, Sugihara M, Imai T, Ishiguro T, Mizutani S, Tomita M, Soga T (2008) Development of bottom-fermenting Saccharomyces strains that produce high SO2 levels, using integrated metabolome and transcriptome analysis. Appl Environ Microbiol 74(9):2787–2796CrossRefGoogle Scholar
  5. 5.
    Nakao Y, Kanamori T, Itoh T, Kodama Y, Rainieri S, Nakamura N, Shimonaga T, Hattori M, Ashikari T (2009) Genome sequence of the lager brewing yeast, an interspecies hybrid. DNA Res 16(2):115–129.  https://doi.org/10.1093/dnares/dsp003 CrossRefGoogle Scholar
  6. 6.
    Hansen J, Bruun SV, Bech LM, Gjermansen C (2002) The level ofMXR1gene expression in brewing yeast during beer fermentation is a major determinant for the concentration of dimethyl sulfide in beer. FEMS yeast research 2(2):137–149.  https://doi.org/10.1111/j.1567-1364.2002.tb00078.x Google Scholar
  7. 7.
    Bokulich NA, Bamforth CW (2013) The microbiology of malting and brewing. Microbiol Mol Biol Rev 77(2):157–172.  https://doi.org/10.1128/MMBR.00060-12 CrossRefGoogle Scholar
  8. 8.
    Pires EJ, Teixeira JA, Branyik T, Vicente AA (2014) Yeast: the soul of beer’s aroma–a review of flavour-active esters and higher alcohols produced by the brewing yeast. Appl Microbiol Biotechnol 98(5):1937–1949.  https://doi.org/10.1007/s00253-013-5470-0 CrossRefGoogle Scholar
  9. 9.
    Coghe S, Benoot K, Delvaux F, Vanderhaegen B, Delvaux FR (2004) Ferulic acid release and 4-vinylguaiacol formation during brewing and fermentation: indications for feruloyl esterase activity in Saccharomyces cerevisiae. J Agric Food Chem 52(3):602–608.  https://doi.org/10.1021/jf0346556 CrossRefGoogle Scholar
  10. 10.
    Mukai N, Masaki K, Fujii T, Iefuji H (2014) Single nucleotide polymorphisms of PAD1 and FDC1 show a positive relationship with ferulic acid decarboxylation ability among industrial yeasts used in alcoholic beverage production. J Biosci Bioeng 118(1):50–55.  https://doi.org/10.1016/j.jbiosc.2013.12.017 CrossRefGoogle Scholar
  11. 11.
    Schneiderbanger H, Koob J, Poltinger S, Jacob F, Hutzler M (2016) Gene expression in wheat beer yeast strains and the synthesis of acetate esters. J Inst Brew 122(3):403–411CrossRefGoogle Scholar
  12. 12.
    Goncalves M, Pontes A, Almeida P, Barbosa R, Serra M, Libkind D, Hutzler M, Goncalves P, Sampaio JP (2016) Distinct domestication trajectories in top-fermenting beer yeasts and wine yeasts. Curr Biol 26(20):2750–2761.  https://doi.org/10.1016/j.cub.2016.08.040 CrossRefGoogle Scholar
  13. 13.
    Focke K, Jentsch M (2013) Kleine Zelle—große Wirkung—Hefestämme, Hefelagerung und der Einfluss auf den Biercharakter. Brauindustrie 11:16–18Google Scholar
  14. 14.
    White C, Zainasheff J (2010) Yeast: the practical guide to beer fermentation. Brewers Publications, BoulderGoogle Scholar
  15. 15.
    Dornbusch HD (2010) The Ultimate almanac of world beer recipes: a practical guide for the professional brewer to the world’s classic beer styles from A to Z. Cerevisia Communications, Bamberg, Germany.Google Scholar
  16. 16.
    Meier-Dörnberg T, Michel M, Wager RS, Jacob F, Hutzler M (2017) Genetic and phenotypic characterization of different top-fermenting Saccharomyces cerevisiae Ale Yeast Isolates. Brew Sci 70:9–25Google Scholar
  17. 17.
    Gonzalez SS, Barrio E, Gafner J, Querol A (2006) Natural hybrids from Saccharomyces cerevisiae, Saccharomyces bayanus and Saccharomyces kudriavzevii in wine fermentations. FEMS Yeast Res 6(8):1221–1234.  https://doi.org/10.1111/j.1567-1364.2006.00126.x CrossRefGoogle Scholar
  18. 18.
    Gallone B, Steensels J, Prahl T, Soriaga L, Saels V, Herrera-Malaver B, Merlevede A, Roncoroni M, Voordeckers K, Miraglia L, Teiling C, Steffy B, Taylor M, Schwartz A, Richardson T, White C, Baele G, Maere S, Verstrepen KJ (2016) Domestication and divergence of Saccharomyces cerevisiae beer yeasts. Cell 166(6):1397–1410 e1316.  https://doi.org/10.1016/j.cell.2016.08.020 CrossRefGoogle Scholar
  19. 19.
    Gibson BR, Storgards E, Krogerus K, Vidgren V (2013) Comparative physiology and fermentation performance of Saaz and Frohberg lager yeast strains and the parental species Saccharomyces eubayanus. Yeast 30(7):255–266.  https://doi.org/10.1002/yea.2960 CrossRefGoogle Scholar
  20. 20.
    Legras JL, Merdinoglu D, Cornuet JM, Karst F (2007) Bread, beer and wine: Saccharomyces cerevisiae diversity reflects human history. Mol Ecol 16(10):2091–2102.  https://doi.org/10.1111/j.1365-294X.2007.03266.x CrossRefGoogle Scholar
  21. 21.
    Legras JL, Erny C, Charpentier C (2014) Population structure and comparative genome hybridization of European flor yeast reveal a unique group of Saccharomyces cerevisiae strains with few gene duplications in their genome. PloS one 9(10):e108089.  https://doi.org/10.1371/journal.pone.0108089 CrossRefGoogle Scholar
  22. 22.
    Masneuf-Pomarede I, Salin F, Borlin M, Coton E, Coton M, Jeune CL, Legras JL (2016) Microsatellite analysis of Saccharomyces uvarum diversity. FEMS Yeast Res 16(2):fow002.  https://doi.org/10.1093/femsyr/fow002 CrossRefGoogle Scholar
  23. 23.
    Couto MB, Eijsma B, Hofstra H, van der Vossen J (1996) Evaluation of molecular typing techniques to assign genetic diversity among Saccharomyces cerevisiae strains. Appl Environ Microbio 62(1):41–46Google Scholar
  24. 24.
    Laidlaw L, Tompkins T, Savard L, Dowhanick T (1996) Identification and differentiation of brewing yeasts using specific and RAPD polymerase chain reaction. J Am Soc Brew Chem 54(2):97–102Google Scholar
  25. 25.
    Krogerus K, Magalhaes F, Vidgren V, Gibson B (2015) New lager yeast strains generated by interspecific hybridization. J Ind Microbiol Biotechnol 42(5):769–778.  https://doi.org/10.1007/s10295-015-1597-6 CrossRefGoogle Scholar
  26. 26.
    Sheehan CA, Weiss AS, Newsom IA, Flint V, O’Donnell DC (1991) Brewing yeast identification and chromosome analysis using high resolution CHEF gel electrophoresis. J Inst Brew 97(3):163–167CrossRefGoogle Scholar
  27. 27.
    Timmins EM, Quain DE, Goodacre R (1998) Differentiation of brewing yeast strains by pyrolysis mass spectrometry and Fourier transform infrared spectroscopy. Yeast 14 (10):885–893. (10.1002/(SICI)1097-0061(199807)14:10<885::AID-YEA286>3.0.CO;2-G)CrossRefGoogle Scholar
  28. 28.
    Azumi M, Goto-Yamamoto N (2001) AFLP analysis of type strains and laboratory and industrial strains of Saccharomyces sensu stricto and its application to phenetic clustering. Yeast 18(12):1145–1154.  https://doi.org/10.1002/yea.767 CrossRefGoogle Scholar
  29. 29.
    de Barros Lopes M, Rainieri S, Henschke PA, Langridge P (1999) AFLP fingerprinting for analysis of yeast genetic variation. Int J Syst Bacteriol 2(2):915–924.  https://doi.org/10.1099/00207713-49-2-915 CrossRefGoogle Scholar
  30. 30.
    Cappello MS, Bleve G, Grieco F, Dellaglio F, Zacheo G (2004) Characterization of Saccharomyces cerevisiae strains isolated from must of grape grown in experimental vineyard. J Appl Microbiol 97(6):1274–1280.  https://doi.org/10.1111/j.1365-2672.2004.02412.x CrossRefGoogle Scholar
  31. 31.
    Gonzalez Flores M, Rodriguez ME, Oteiza JM, Barbagelata RJ, Lopes CA (2017) Physiological characterization of Saccharomyces uvarum and Saccharomyces eubayanus from Patagonia and their potential for cidermaking. Int J Food Microbiol 249:9–17.  https://doi.org/10.1016/j.ijfoodmicro.2017.02.018 CrossRefGoogle Scholar
  32. 32.
    McMurrough I, Madigan D, Donnelly D, Hurley J, Doyle AM, Hennigan G, McNulty N, Smyth MR (1996) Control of ferulic acid and 4-vinyl guaiacol in brewing. J Inst Brew 102(5):327–332CrossRefGoogle Scholar
  33. 33.
    Vanbeneden N, Gils F, Delvaux F, Delvaux FR (2008) Formation of 4-vinyl and 4-ethyl derivatives from hydroxycinnamic acids: Occurrence of volatile phenolic flavour compounds in beer and distribution of Pad1-activity among brewing yeasts. Food Chem 107(1):221–230.  https://doi.org/10.1016/j.foodchem.2007.08.008 CrossRefGoogle Scholar
  34. 34.
    Mertens S, Steensels J, Gallone B, Souffriau B, Malcorps P, Verstrepen K (2017) Rapid screening method for phenolic off-flavor (POF) production in yeast.  https://doi.org/10.1094/ASBCJ-2017-4142-01
  35. 35.
    Joubert R, Brignon P, Lehmann C, Monribot C, Gendre F, Boucherie H (2000) Two-dimensional gel analysis of the proteome of lager brewing yeasts. Yeast 16 (6):511–522. (10.1002/(SICI)1097-0061(200004)16:6<511::AID-YEA544>3.0.CO;2-I)CrossRefGoogle Scholar
  36. 36.
    Trabalzini L, Paffetti A, Scaloni A, Talamo F, Ferro E, Coratza G, Bovalini L, Lusini P, Martelli P, Santucci A (2003) Proteomic response to physiological fermentation stresses in a wild-type wine strain of Saccharomyces cerevisiae. Biochem J 370(Pt 1):35–46.  https://doi.org/10.1042/BJ20020140 CrossRefGoogle Scholar
  37. 37.
    Kobi D, Zugmeyer S, Potier S, Jaquet-Gutfreund L (2004) Two-dimensional protein map of an “ale"-brewing yeast strain: proteome dynamics during fermentation. FEMS Yeast Res 5(3):213–230.  https://doi.org/10.1016/j.femsyr.2004.07.004 CrossRefGoogle Scholar
  38. 38.
    Zuzuarregui A, Monteoliva L, Gil C, del Olmo M (2006) Transcriptomic and proteomic approach for understanding the molecular basis of adaptation of Saccharomyces cerevisiae to wine fermentation. Appl Environ Microbiol 72(1):836–847.  https://doi.org/10.1128/AEM.72.1.836-847.2006 CrossRefGoogle Scholar
  39. 39.
    Hansen R, Pearson SY, Brosnan JM, Meaden PG, Jamieson DJ (2006) Proteomic analysis of a distilling strain of Saccharomyces cerevisiae during industrial grain fermentation. Appl Microbiol Biotechnol 72(1):116–125.  https://doi.org/10.1007/s00253-006-0508-1 CrossRefGoogle Scholar
  40. 40.
    Kern CC, Vogel RF, Behr J (2014) Differentiation of lactobacillus brevis strains using matrix-assisted-laser-desorption-ionization-time-of-flight mass spectrometry with respect to their beer spoilage potential. Food Microbiol 40:18–24CrossRefGoogle Scholar
  41. 41.
    Wieme AD, Spitaels F, Aerts M, De Bruyne K, Van Landschoot A, Vandamme P (2014) Identification of beer-spoilage bacteria using matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. Int J Food Microbiol 185:41–50.  https://doi.org/10.1016/j.ijfoodmicro.2014.05.003 CrossRefGoogle Scholar
  42. 42.
    Guo L, Ye L, Zhao Q, Ma Y, Yang J, Luo Y (2014) Comparative study of MALDI-TOF MS and VITEK 2 in bacteria identification. J Thorac Dis 6(5):534–538.  https://doi.org/10.3978/j.issn.2072-1439.2014.02.18 Google Scholar
  43. 43.
    Demirev P, Sandrin TR (2016) Applications of Mass Spectrometry in Microbiology. Springer, HeidelbergCrossRefGoogle Scholar
  44. 44.
    Croxatto A, Prod’hom G, Greub G (2012) Applications of MALDI-TOF mass spectrometry in clinical diagnostic microbiology. FEMS Microbiol Rev 36(2):380–407.  https://doi.org/10.1111/j.1574-6976.2011.00298.x CrossRefGoogle Scholar
  45. 45.
    Kern CC, Vogel RF, Behr J (2014) Identification and differentiation of brewery isolates of Pectinatus sp. by matrix-assisted-laser desorption–ionization time-of-flight mass spectrometry (MALDI-TOF MS). Eur Food Res Technol 238(5):875–880.  https://doi.org/10.1007/s00217-014-2173-4 CrossRefGoogle Scholar
  46. 46.
    Nacef M, Chevalier M, Chollet S, Drider D, Flahaut C (2017) MALDI-TOF mass spectrometry for the identification of lactic acid bacteria isolated from a French cheese: The Maroilles. Int J Food Microbiol 247:2–8.  https://doi.org/10.1016/j.ijfoodmicro.2016.07.005 CrossRefGoogle Scholar
  47. 47.
    Pavlovic M, Mewes A, Maggipinto M, Schmidt W, Messelhausser U, Balsliemke J, Hormansdorfer S, Busch U, Huber I (2014) MALDI-TOF MS based identification of food-borne yeast isolates. J Microbiol Methods 106:123–128.  https://doi.org/10.1016/j.mimet.2014.08.021 CrossRefGoogle Scholar
  48. 48.
    Moothoo-Padayachie A, Kandappa HR, Krishna SBN, Maier T, Govender P (2013) Biotyping Saccharomyces cerevisiae strains using matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS). Eur Food Res Technol 236(2):351–364.  https://doi.org/10.1007/s00217-012-1898-1 CrossRefGoogle Scholar
  49. 49.
    Blattel V, Petri A, Rabenstein A, Kuever J, Konig H (2013) Differentiation of species of the genus Saccharomyces using biomolecular fingerprinting methods. Appl Microbiol Biotechnol 97(10):4597–4606.  https://doi.org/10.1007/s00253-013-4823-z CrossRefGoogle Scholar
  50. 50.
    Gutierrez C, Gomez-Flechoso MA, Belda I, Ruiz J, Kayali N, Polo L, Santos A (2017) Wine yeasts identification by MALDI-TOF MS: Optimization of the preanalytical steps and development of an extensible open-source platform for processing and analysis of an in-house MS database. Int J Food Microbiol 254:1–10.  https://doi.org/10.1016/j.ijfoodmicro.2017.05.003 CrossRefGoogle Scholar
  51. 51.
    Usbeck JC, Wilde C, Bertrand D, Behr J, Vogel RF (2014) Wine yeast typing by MALDI-TOF MS. Appl Microbiol Biotechnol 98(8):3737–3752.  https://doi.org/10.1007/s00253-014-5586-x CrossRefGoogle Scholar
  52. 52.
    Lauterbach A, Usbeck JC, Behr J, Vogel RF (2017) MALDI-TOF MS typing enables the classification of brewing yeasts of the genus Saccharomyces to major beer styles. PloS one 12(8):e0181694.  https://doi.org/10.1371/journal.pone.0181694 CrossRefGoogle Scholar
  53. 53.
    Legras JL, Ruh O, Merdinoglu D, Karst F (2005) Selection of hypervariable microsatellite loci for the characterization of Saccharomyces cerevisiae strains. Int J Food Microbiol 102(1):73–83.  https://doi.org/10.1016/j.ijfoodmicro.2004.12.007 CrossRefGoogle Scholar
  54. 54.
    Perez M, Gallego F, Martinez I, Hidalgo P (2001) Detection, distribution and selection of microsatellites (SSRs) in the genome of the yeast Saccharomyces cerevisiae as molecular markers. Lett Appl Microbiol 33(6):461–466CrossRefGoogle Scholar
  55. 55.
    Cavalli-Sforza LL, Edwards AW (1967) Phylogenetic analysis: models and estimation procedures. Evolution 21(3):550–570CrossRefGoogle Scholar
  56. 56.
    Yamauchi H, Yamamoto H, Shibano Y, Amaya N, Saeki T (1998) Rapid methods for detecting Saccharomyces diastaticus, a beer spoilage yeast, using the polymerase chain reaction. J Am Soc Brew Chem 56(2):58–63Google Scholar
  57. 57.
    Bayly JC, Douglas LM, Pretorius IS, Bauer FF, Dranginis AM (2005) Characteristics of Flo11-dependent flocculation in Saccharomyces cerevisiae. FEMS Yeast Res 5(12):1151–1156.  https://doi.org/10.1016/j.femsyr.2005.05.004 CrossRefGoogle Scholar
  58. 58.
    Michel M, Kopecká J, Meier-Dörnberg T, Zarnkow M, Jacob F, Hutzler M (2016) Screening for new brewing yeasts in the non-Saccharomyces sector with Torulaspora delbrueckii as model. Yeast 33:129–144CrossRefGoogle Scholar
  59. 59.
    Usbeck JC, Kern CC, Vogel RF, Behr J (2013) Optimization of experimental and modelling parameters for the differentiation of beverage spoiling yeasts by matrix-assisted-laser-desorption/ionization-time-of-flight mass spectrometry (MALDI-TOF MS) in response to varying growth conditions. Food Microbiol 36(2):379–387.  https://doi.org/10.1016/j.fm.2013.07.004 CrossRefGoogle Scholar
  60. 60.
    Jombart T, Collins C (2015) A tutorial for discriminant analysis of principal components (DAPC) using adegenet 2.0.0. Imperial College London, MRC Centre for Outbreak Analysis and Modelling, LondonGoogle Scholar
  61. 61.
    Jespersen L, Jakobsen M (1996) Specific spoilage organisms in breweries and laboratory media for their detection. Int J Food Microbiol 33(1):139–155.  https://doi.org/10.1016/0168-1605(96)01154-3 CrossRefGoogle Scholar
  62. 62.
    Heresztyn T (1986) Metabolism of volatile phenolic compounds from hydroxycinnamic acids by Brettanomyces yeast. Arch Microbiol 146(1):96–98CrossRefGoogle Scholar
  63. 63.
    Mukai N, Masaki K, Fujii T, Kawamukai M, Iefuji H (2010) PAD1 and FDC1 are essential for the decarboxylation of phenylacrylic acids in Saccharomyces cerevisiae. J Biosci Bioeng 109(6):564–569.  https://doi.org/10.1016/j.jbiosc.2009.11.011 CrossRefGoogle Scholar
  64. 64.
    Chen P, Dong J, Yin H, Bao X, Chen L, He Y, Wan X, Chen R, Zhao Y, Hou X (2015) Single nucleotide polymorphisms and transcription analysis of genes involved in ferulic acid decarboxylation among different beer yeasts. J Inst Brew 121(4):481–489CrossRefGoogle Scholar
  65. 65.
    Li YC, Korol AB, Fahima T, Beiles A, Nevo E (2002) Microsatellites: genomic distribution, putative functions and mutational mechanisms: a review. Mol Ecol 11(12):2453–2465.  https://doi.org/10.1046/j.1365-294X.2002.01643.x CrossRefGoogle Scholar
  66. 66.
    Marklein G, Josten M, Klanke U, Muller E, Horre R, Maier T, Wenzel T, Kostrzewa M, Bierbaum G, Hoerauf A, Sahl HG (2009) Matrix-assisted laser desorption ionization-time of flight mass spectrometry for fast and reliable identification of clinical yeast isolates. J Clin Microbiol 47(9):2912–2917.  https://doi.org/10.1128/JCM.00389-09 CrossRefGoogle Scholar
  67. 67.
    Usbeck JC, Behr J (2012) Vogel RF Differentiation of top-and bottom-fermenting brewing yeasts and insight into their metabolic status by MALDI-TOF MS, In: Poster World Brewing Congress, PortlandGoogle Scholar
  68. 68.
    Schurr BC, Behr J, Vogel RF (2015) Detection of acid and hop shock induced responses in beer spoiling Lactobacillus brevis by MALDI-TOF MS. Food Microbiol 46:501–506.  https://doi.org/10.1016/j.fm.2014.09.018 CrossRefGoogle Scholar
  69. 69.
    Christner M, Trusch M, Rohde H, Kwiatkowski M, Schluter H, Wolters M, Aepfelbacher M, Hentschke M (2014) Rapid MALDI-TOF mass spectrometry strain typing during a large outbreak of Shiga-Toxigenic Escherichia coli. PloS one 9(7):e101924.  https://doi.org/10.1371/journal.pone.0101924 CrossRefGoogle Scholar
  70. 70.
    Pham T, Wimalasena T, Box W, Koivuranta K, Storgårds E, Smart K, Gibson B (2011) Evaluation of ITS PCR and RFLP for differentiation and identification of brewing yeast and brewery ‘wild’ yeast contaminants. J Inst Brew 117(4):556–568CrossRefGoogle Scholar
  71. 71.
    Walther A, Hesselbart A, Wendland J (2014) Genome sequence of Saccharomyces carlsbergensis, the world’s first pure culture lager yeast. Genes Genomes Genet 4(5):783–793  https://doi.org/10.1534/g3.113.010090
  72. 72.
    Spencer JF, Spencer DM (1983) Genetic improvement of industrial yeasts. Annu Rev Microbiol 37(1):121–142.  https://doi.org/10.1146/annurev.mi.37.100183.001005 CrossRefGoogle Scholar
  73. 73.
    Meier-Dörnberg T, Jacob F, Michel M, Hutzler M (2017) Incidence of Saccharomyces cerevisiae var. diastaticus in the beverage industry: cases of contamination, 2008–2017. Master Brew Assoc Am 54(4):140–148.  https://doi.org/10.1094/TQ-54-4-1130-01 Google Scholar

Copyright information

© Springer-Verlag GmbH Germany, part of Springer Nature 2018

Authors and Affiliations

  • Alexander Lauterbach
    • 1
  • Caroline Wilde
    • 3
  • Dave Bertrand
    • 3
  • Jürgen Behr
    • 1
    • 2
  • Rudi F. Vogel
    • 1
  1. 1.Lehrstuhl für Technische MikrobiologieTechnische Universität MünchenFreisingGermany
  2. 2.Bavarian Center for Biomolecular Mass SpectrometryFreisingGermany
  3. 3.Lallemand Inc.QuébecCanada

Personalised recommendations