Identiﬁcation of Microﬂora in a Biological Brewer’s Wort Acidiﬁcation Process Run Continuously for 20 Years

: Biological acidiﬁcation is a common and beneﬁcial way for breweries to adjust the pH value of their mash or wort to improve enzymatic activity during mashing, raise yeast activity during fermentation, and increase the ﬂavor stability of the ﬁnished beer. The reactors are mostly run for many years without re-inoculating a fresh culture, creating the possibility of changes in the culture, genetic drifts, or the survival of different strains. In this study, a biological acidiﬁcation culture that had been continuously run for 20 years was analyzed by GTG5 PCR and IGS2-314 rDNA PCR ﬁngerprinting, as well as 16S and 26S rDNA sequencing, and real-time PCR was applied to differentiate the bacterial and yeast strains and species. The applied real-time PCR primers for Lactobacillus amylolyticus and Lactobacillus amylovorus have not been published yet. It was shown that only strains of the species L. amylolyticus were present, with low contamination of yeast strains from the species Saccharomyces cerevisiae. As the original starter culture was Lactobacillus amylolyticus , the acidiﬁcation plant ran for 20 years, and no Lactobacillus sp. cross-contamination could be analyzed using culture-dependent methods after 20 years. The microﬂora composition is a decisive factor for the ﬁnal beer ﬂavor.


Introduction
Deliberate biological mash acidification has a major positive impact on almost all steps in the brewery, from mashing through to the final beer [1]. It is caused by thermophilic lactic acid bacteria (LAB), which were first described approximately 130 years ago by P. Lindner [2]. Twelve years later, in 1896, Leichmann found bacterial cells producing lactic acid from malt mash [3]. Since then, it has become a widely used technique for improving the quality of beer and lowering the pH of the malt, sweet wort, and wort in the brewery. The first patent for an applied wort acidification process was claimed in 1906 by Otto Francke [4]. The process consists of a batch or continuous fermentation of diluted sweet wort (5-12 • P) by a culture of LAB, to a final concentration of 1% lactic acid. Acidification is usually complete within 24 h, and achieves a final pH value of 3.0-3.6 [1,5]. A ratio of 1-2% acidified sweet wort is added to acidify mash or wort. This addition changes the mash pH value by 0.1-0.3 pH units, and the wort pH value by 0.2-0.4 pH units [1] As the LAB applied are very sensitive to hops, and the mixture is heated to temperatures that are lethal to bacteria, an infection of the final product is unlikely [1]. Biological acidification is a natural way of adding lactic acid to the brew instead of using technical lactic or phosphoric acid to lower the pH value. It is therefore accepted under the German Purity Law (Reinheitsgebot), and is used by many German breweries as well as breweries around the world [6].
The benefits of biological acidification are, in particular, better fermentation, higher selective pressure on the yeast, better enzyme offerings, and activities in the mash (glucanolytic, proteolytic, and amylolytic), as well as good nutrient supply, filtration, sensory quality, and flavor stability of the beer in comparison with technical acidification (lactic and phosphoric acid) [7][8][9][10]. There are also minor disadvantages with biological acidification, such as a loss of bitter substances, lower hop isomerization, and a lower splitting range of DMS (dimethyl sulfide) precursors [9].
The most employed microorganism, Lactobacillus amylolyticus, originated from barley malt, but other LAB are also used, i.e., L. delbrueckii, L. fermentum, L. casei, L. amylovorus, L. helveticus, Pediococcus acidilacti, P. pentosaceus, and P. dextrinicus [1,7,11,12]. Important requirements in the selection of LAB strains are the production of the maximum amounts of L(+)-lactic acid as a metabolic end product, with relatively low or no production of diacetyl; it must be a natural isolate from barley or malt, and growth should be at 47-48 • C [10,12]. Strains of the species L. amylolyticus are Gram-positive rods, catalase-and oxidase-negative, homofermentative, thermophilic, and microaerophilic-they ferment maltose, dextrins, and starch, and are sensitive to hop compounds. In comparison with yeast propagation, which is re-inoculated with a fresh culture of pure yeast approximately every four months [13]; biological acidification is left to its own devices for many years by most breweries. The low pH value, as well as the high temperature, ensures a microbiologically stable process with a low contamination risk [1]. Moreover, most breweries using biological acidification report a typical "in-house taste" [6] that is related to their acidification process, and which could change after re-inoculating the LAB. This incidence could be due to strain-related flavor production or to contamination by thermophilic yeast (Kluyveromyces marxianus or Candida kefyr) other thermophilic bacteria (L. casei subspecies or L. coryniformis), or by a mixture of differing strains of the same species [1,6].
The aim of this study was therefore to identify the species of microorganisms in the biologically acidified sweet wort samples that had been continuously used for 20 years for biological acidification without re-inoculation in the Ottakringer brewery (Vienna, Austria).

Media and Growth Conditions
Two biologically acidified sweet wort samples were supplied by the Ottakringer brewery in Vienna, Austria. Each of the samples was serially diluted using 10-fold volumes of physiological salt solution and cultured on wort agar with tetracycline (WAT) (20 µL/mL) (Weihenstephan Brewery, Freising, Germany), on sweet wort agar (SW) at 11 • P (Weihenstephan Brewery), and MRS agar (Sifin, Berlin, Germany) using the spread plate technique. Cultivation was performed at 28 and 48 • C under aerobic (WAT) and anaerobic (WAT, SW, and MRS) conditions for 2-7 days. Anaerobic cultivation was performed with Anaerocult A (Merck, Darmstadt, Germany) in a 2.5 L anaerobiosis jar (Merck). Microorganisms with distinct colony morphologies were selected, and the cell morphology was microscopically determined. The isolates were sub-cultured in the same medium and used as stock cultures in subsequent analyses. Yeasts were numbered in the following from sample SG1 to SG100. All bacteria are numbered starting from SG 101 and continuing in ascending order.

Preparing Cryogenic Cultures
Single colonies were inoculated into liquid media, and the growth cultures were used to prepare cryogenic cultures-Roti-Store yeast cryo-vials (Roth, Karlsruhe, Germany) for yeasts, and a Roti-Store cryo-vial (Roth) for bacteria. The cultures were stored at −80 • C.

PCR and Fingerprinting
DNA isolation was executed as follows: microbial cultures (100 µL) were transferred into sterile reaction vessels and centrifuged at 12,000× g for 2 min; cell pellets were re-suspended in the InstaGene Matrix (Biorad, München, Germany). DNA was then isolated according to the manufacturer's instructions and was used for fingerprint typing, real-time PCR, and for 16S rDNA and 26S rDNA sequence analyses.
Amplicons were initially analyzed by the gel electrophoresis system Agilent 2100 (Agilent Technologies, Santa Clara, CA, United States), using the Agilent DNA 7500 kit ((GTG) 5 analysis) and the Agilent DNA 1000 kit (IGS2 314 analysis). Phylogenetic analysis was conducted using Bionumerics 7.1 software (Applied Maths, Sint-Martens-Latem, Belgium). The sequence similarity was estimated using the Pearson correlation with an optimization of 0.5% and curve smoothing of 0.5%, and the cluster analysis was constructed using the UPGMA (Unweighted Pair Group Method with Arithmetic Mean) method with highest overall similarity.

Sequencing of 16S and 26S rDNA
The primers 16S-27f and 1492r (Table 1) were used for the analysis of 16S rDNA and bacteria identification, and the PCR parameters were chosen as follows: (i) 95 • C for 5 min; (ii) 35 cycles of 95 • C for 25 s, 55 • C for 40 s, and 72 • C for 2 min; and (iii) 72 • C for 5 min.
The primers NL1 and NL4 (Table 1) were used for the analysis of 26S rDNA of the yeast, and the following protocol was used for the PCR: (i) 95 • C for 5 min; (ii) 35 cycles of 95 • C for 25 s, 52 • C for 1 min, and 72 • C for 1 min; and (iii) 72 • C for 10 min.

Real-Time PCR
Species classification was verified using real-time PCR (Light Cycler 480 II, Roche, Penzberg, Germany).

Phenolic Off-Flavor Test
For the stock solution of ferulic/cinnamic acid, 1 g of the instant phenolic acid was diluted in 20 mL of 96% (v/v) ethanol. Then 2 mL of the stock solution was added to 1000 mL yeast and mold (YM) agar media (add distilled water to 3.0 g malt extract, 3.0 g yeast extract, 5.0 g peptone, 11.0 g glucose monohydrate, and 20.0 g agar, to 1000 mL) at 45-50 • C under sterile conditions. The investigated pure yeast strains were taken from the agar plates and spread on the YM agar plate containing the described acid. Samples were incubated for three days at 28 • C. LeoBavaricus-TUM 68 ® (a common phenolic off-flavor (POF)-positive wheat beer strain [22]) and Frisinga-TUM 34/70 ® were also taken into account as a positive (TUM 68) and a negative (TUM 34/70) control. Evaluation was performed by seven trained panelists, who judged the smell of the plates with 4-vinylguaiacol (4-VG; clove-like) or 4-vinylstyrene (4-VS; styrofoam-like) positive or negative.

Results
The aim of this study was to identify the species and strains of microorganisms in the biological acidification reactor that were continuously used for 20 years without re-inoculation in the Ottakringer brewery (Vienna, Austria). A total of 22 of the 50 microorganism isolates recovered from biologically acidified sweet wort samples were microscopically identified as yeasts, and 28 isolates were identified as LAB.

Identification of Bacteria
On the basis of the phylogenetic tree of the fingerprint type (GTG) 5 (Figure 1), four different LAB were chosen for further analyses (SG109, SG115, SG126, and SG139). As can be seen, the strains are divided into two main clusters that had a similarity of 78%. A similarity value of >70% implies that all the strains belong to the same species [23]. Therefore, two strains of each cluster were chosen for further analyses. The primer 1492r (see Table 1) was used for the 16S rDNA gene sequencing. The results of BLAST are shown in Table 2. The sequences obtained with the 1492r primer gave ambiguous results: bacteria with the highest consensus score corresponded to the targeted species, L. amylolyticus, but were not significantly different from the bacteria with the second highest score, the closely related species L. amylovorus. The primers 16S-27r, 612r, and 926r were then additionally used for the 16S rDNA gene sequencing. According to the obtained sequences, the bacteria were identified as L. amylolyticus. The 16S rDNA sequence obtained from strain L. amylolyticus SG117 was deposited in the Genbank database under the accession number KT 891013. All selected strains can be identified at species level.
The identification of bacteria at the species level, L. amylolyticus, was confirmed by real-time PCR, according to Brandl [20]. All of the 28 samples were determined to be L. amylolyticus. The primer 1492r (see Table 1) was used for the 16S rDNA gene sequencing. The results of BLAST are shown in Table 2. The sequences obtained with the 1492r primer gave ambiguous results: bacteria with the highest consensus score corresponded to the targeted species, L. amylolyticus, but were not significantly different from the bacteria with the second highest score, the closely related species L. amylovorus. The primers 16S-27r, 612r, and 926r were then additionally used for the 16S rDNA gene sequencing. According to the obtained sequences, the bacteria were identified as L. amylolyticus. The 16S rDNA sequence obtained from strain L. amylolyticus SG117 was deposited in the Genbank database under the accession number KT 891013. All selected strains can be identified at species level. The identification of bacteria at the species level, L. amylolyticus, was confirmed by real-time PCR, according to Brandl [20]. All of the 28 samples were determined to be L. amylolyticus.

Identification of Yeasts
On the basis of the phylogenetic trees of IGS2 314 (Figure 2), the 10 most differing yeasts were chosen for further analyses (SG1, SG14, SG16, SG22, SG82, SG85, SG93, and SG100). The primer NL4 was used for the 26S rDNA gene sequencing of the isolated yeast species. The results of BLAST are shown in Table 2, and all of the 10 yeast samples were identified as S. cerevisiae. For the BLAST analysis of the sequence 26S rDNA, the yeast with the highest consensus score corresponded to the yeast of detection and was significantly different from the yeast species with the second-highest score. All of the selected yeast strains could be identified at the species level. Moreover, differentiation was established between the top-fermenting (S. cerevisiae) and bottom-fermenting (S. pastorianus) yeast (Table 3), according to Hutzler [15] and Salinas et al. [21]. All the selected samples were confirmed to be S. cerevisiae. As wild S. cerevisiae strains are mostly able to produce phenolic off-flavors (POF), their ability to decarboxylase cinnamic acid and ferulic acid to the corresponding flavor components was analyzed [24,25]. The results can be reviewed in Table 4. They show that the strains vary in their ability to produce phenolic off-flavors. Strains SG16 and SG29 do not produce any phenolic off-flavors, in contrast to the other analyzed strains.

Discussion
The identification of yeast and lactic acid bacteria found in the sample of a biological acidification plant continuously run for 20 years uncovered two main species. The yeast species S. cerevisiae was detected with a variety of differing strains, as shown in Figure 2 and Table 4. Some of them were POF-positive as well as POF-negative, suggesting that a certain number of the strains might have been present for some time. All yeast strains are considered as spoilers in a biological acidification plant [5,6]; however, the presence of POF-negative S. cerevisiae strains suggests these were present for a long time, as a missing ability to produce phenolic off-flavors is a link to the domestication process [26]. A comparison of the yeast fingerprints ( Figure 2, IGS2 314) with the IGS2 314 fingerprint of the culture yeast of the brewery showed low equality (data not shown), excluding the eventual case of a contamination by the applied culture brewing yeast. The acidified wort, however, did not have a phenolic off-flavor, which indicates a low contamination of the yeast strains. A contamination with cell counts higher than 1 × 10 4 cell/mL would show a phenolic off-flavor in the produced wort [27]. The low pH value of the wort (pH value of approximately 3.2) and the high temperatures (>45 • C) might have lowered the growth potential of the S. cerevisiae strains [9].
The presence of differing strains of the species L. amylolyticus, but no other detected bacteria species, shows a potential mutation of the original pure culture, or that the original culture was already a strain mixture (strain mixtures were often popular in biological acidification plants in the 1980s and 1990s). The phylogenetic tree of the ((GTG) 5 ) fingerprints suggests three main clusters with very high similarity. The change from the mixed culture to a pure culture changed the taste of the product significantly. This showed that the different strains had an impact on the flavor. A re-inoculation with the old mixed culture produced the same taste as before, suggesting that the differing strains influence the flavor (as reported by the brewery).

Conclusions
In a biological acidification process that has run continuously for 20 years, only strains of the inoculated species Lactobacillus amylolyticus could be cultivated, and no cross-contamination by other LAB species could be analyzed. Lactobacillus amylolyticus was also the initial starter culture species. Thus, it could be confirmed that only strains of species Lactobacillus amylolyticus could be identified, and that they belonged to the same species that was initially inoculated 20 years ago. Only Lactobacillus amylolyticus strains could be identified, and no other LAB species could be cultivated. Minor contamination with yeast might not be averted, but as shown, their growth rate can be described as low, because a pH value of around 3 and temperatures at 48 • C will keep the spoilage under control. However, the small change in the culture and resulting variation in the strains, as shown in the phylogenetic tree in Figure 1, has an impact on the flavor and the "in-house taste" of the final beer. A new set up with a pure culture of one of the present L. amylolyticus strains produced a considerably different taste to the mixture of all the strains (as reported by the brewery).