Part I: the influence of serial repitching ofSaccharomyces pastorianuson the uptake dynamics of metal ions and fermentable carbohydrates during the fermentation of barley and gluten-free buckwheat and quinoa wort

Abstract

Recently, research has been focusing on the use of alternative raw materials for brewing purposes and gluten-free beer-like beverages from malted buckwheat and quinoa are of commercial interest. A common commercial process involves the serial repitching of the yeast biomass, but this has not been described using buckwheat and quinoa wort fermentations. Our research studies (Parts I–III) explored the serial repitching of the yeast strain Saccharomyces pastorianus TUM 34/70 on the composition of a barley, buckwheat and quinoa fermentation medium. The present paper focuses on the fermentation performance and the uptake dynamics of metal ions and fermentable carbohydrates. Both pseudocereals showed high variations in all of the attributes examined during successive fermentations. In buckwheat the differences between successive fermentations were similar to those observed with barley, whereas differences in quinoa varied quite significantly from those observed with barley and showed a directional trend, suggesting a general weakening of the yeast from the sixth successive fermentation onward. In particular, the assimilation of the fermentable carbohydrates lessened and metal ion uptake appeared poorly controlled. It was concluded that buckwheat showed good potential for serial repitching of S. pastorianus TUM 34/70, whereas serial repitching of a quinoa wort appeared to be limited to five or six fermentations. Copyright © 2015 The Institute of Brewing & Distilling Barley has been the main brewing ingredient ever since the production of beer became widespread 1 and it has served its purpose well. However, with the discovery of the causes of coeliac disease, the use of barley in beer brewing can be a problem for some individuals. Barley and its dietary products contain gluten, a protein composite that triggers an autoimmune reaction in genetically predetermined individuals 2. Although beer is a highly diluted barley product, analyses of beer from barley even with hordein deletion have shown that all barley-based beers tested still contain hordeins 3. There are a few studies reporting that the gluten content in commercial beers is below the limits of Codex Alimentarius Standard (20 ppm) 4; however, precaution is still needed since the gluten content of the same beer type could vary significantly between batches and clinical sensitivity toward gluten differs substantially from patient to patient. Beer drinking is an important human habit that strengthens cultural ties in the majority of societies. Because in the past there have been relatively few commercial beers without gluten, gluten-sensitive individuals often tolerate side effects in order to take part in group activities. With gluten-free diets now established as a precautionary step for some autoimmune and metabolic disorders 5-8, there is now a growing demand for a wider choice of higher-quality and better-tasting gluten-free products, even among non-coeliac individuals. This has encouraged researchers and manufacturers in the last two decades to better explore the potential of gluten-free raw materials, such as the pseudocereals buckwheat and quinoa. In addition to the absence of gluten, buckwheat and quinoa grains also provide nutrient-dense and healthy nourishment 9, 10. A number of publications have reviewed the research findings of malting, mashing and brewing attempts from 100% buckwheat and quinoa malt 10-12 and have designated buckwheat and quinoa as potential brewing ingredients. However, knowledge on the optimal use of these alternative pseudocereals for beer production is still limited. The production of bottom-fermented buckwheat and quinoa beer-like beverages has recently been compared with the production of barley beer and it was concluded that both pseudocereals possess different, but promising, commercial potentials 13. Gluten-free beer-like beverages produced from malted buckwheat and quinoa for commercial production must deal with the higher expenses owing to the required technological adaptations of the process and the need for external enzyme supplementation during mashing 12, 13, as well as the relatively high price of the grains. Serial repitching of the yeast biomass represents one efficient and common mode for yeast expense reduction, especially on an industrial scale. Serial repitching has not yet been studied in detail for the fermentation of buckwheat and quinoa beer-like beverages and a series of papers (Parts I–III) are presented with detailed reports of daily changes in the fermentation medium during 11 successive fermentations of barley, buckwheat and quinoa wort. Changes in the yeast karyotype and protein profile owing to the successive fermentations of buckwheat and quinoa wort have recently been reported 14. Hopefully, therefore, it will be possible to elucidate a more direct link between the yeast's biochemistry and the chemistry of wort/fermentation medium. This paper deals with the fermentation properties, metal ion uptake and release dynamics, and the assimilation profile of fermentable carbohydrates during successive fermentations of barley, buckwheat and quinoa wort by the bottom-fermenting yeast strain S. pastorianus TUM 34/70. After water, metal ions are the single most important inorganic substance in wort and their presence in optimal amounts and bioavailable form is a general prerequisite for a satisfactory viability, vitality and fermentation performance of the yeast. The trace metal ions explored in this study, that is, iron, copper, zinc and manganese, play a crucial role, being a part of haem-proteins, cytochromes, redox pigments, enzyme cofactors, and others. Their relevance in the brewing process has been widely reviewed elsewhere 15, but for barley beer there are only a few reports regarding metal ion nutrition where the same yeast was used twice 16-18 or more 19. The physiological status of the yeast and the uptake rate and dynamics of a particular metal ion are primarily dependent both on its concentration in the wort and on its bioavailability. The latter is mainly governed by its solubility capacity and the presence of complexing chelators 20, as well as by the sugar and alcohol content in the fermentation medium 21. Differences in iron, copper, zinc and manganese concentrations between these three raw materials, as well as between successive fermentations, would be expected. Regardless of the raw material employed, a wort meant for beer preparation contains carbohydrates that the yeast is able to ferment, that is, fructose, glucose, sucrose, maltose and maltotriose, but their ratios and absolute concentrations will vary. The uptake of wort sugars by brewer's yeast has been a subject of close scrutiny, as befits their role in industrial fermentations 22. After pitching, the yeast (in an appropriate physiological condition) immediately takes up the monosaccharides, along with the simultaneous cleavage of sucrose by an invertase in the periplasm. The initial glucose concentration in the wort plays a key role in the order of the sugar consumption since it represses the utilization of other carbohydrates by an effect known as ‘carbon catabolite repression’. Thus high glucose levels in wort are not recommended as the yeast enzymatic system becomes adapted to the high glucose present and reduces or even halts maltose and maltotriose uptake 23. Fermentation performance is affected negatively if glucose is the predominant carbohydrate in wort 24. Yeast has been reported to exhibit higher viabilities in high maltose compared with glucose media 25. Since buckwheat and quinoa worts predominantly contain glucose 13, the question of whether there are there substantial differences in fermentable carbohydrate uptake as a function of raw material and successive fermentation was explored in this study. The same buckwheat and quinoa malt was used as reported in this journal in our 2014 study, with the same malting, mashing and hopping procedures followed as described therein 13. Successive fermentation design was identical to that described in our 2014 karyotype and protein profile study 14. Briefly, before each fermentation, yeast viability was ascertained by methylene blue staining and a cell count (microscope). The pitching volume was adjusted accordingly to achieve a final yeast concentration of 2 × 106 viable cells per millilitre of wort, with a wort real extract of ~10%. The wort samples, previously stored at 0°C, were frozen a day after production. Samples from the fermentation medium were taken every 24 h (±30 min) and, immediately after the sampling, 1 mL of 2 g/L sodium azide was added per 100 mL of sample to stop the fermentation. Sodium azide slows the growth of Saccharomyces cerevisiae and induces a lag, which increases steeply with the concentration 26. After 3 h, samples were centrifuged using a laboratory centrifuge Sigma 6K15 (Sigma, Germany) for 10 min at 4000g to remove the yeast. Supernatants were collected and frozen until the analysis. On the day of analysis, the samples were defrosted at room temperature, degassed in an ultrasonic water bath and filtered through a plain disc filter paper (diameter of 150 mm) and cellulose acetate membrane syringe filter unit (diameter of 25 mm, pore size of 0.45 µm; LLG Labware, Germany). Reagents and solutions were of appropriate quality and purity, and supplied by Sigma-Aldrich (Germany). The brewing attributes were determined in duplicate (n = 2) according to standard procedures from the current versions of either Analytica-EBC 27 or MEBAK 28, 29 protocols. The iron, copper, manganese and zinc standard solutions ‘Baker Instra-Analysed’, 1 g/L, were purchased from J. T. Baker (The Netherlands). Iron, copper and zinc were determined according to the Analytica-EBC methods 9.13.3, 9.14.3, and 9.20 27, respectively, and manganese according to Hoenig and Van Hoeyweghen 30 using Perkin Elmer AAnalyst 200 atomic absorption spectrometer (Perkin Elmer, USA) with Perkin Elmer Lumina™ hollow cathode lamp (Cu–Fe–Mn–Zn). The instrument was controlled with WinLab32 computer software (Perkin Elmer, USA). Fermentable sugars were determined according to Analytica-EBC method 9.27 27 using an HPLC apparatus (Agilent 1200 Series) equipped with a thermostatted autosampler, an ion-exclusion column (Bio-Rad Aminex HPX-87H, 300 × 7.8 mm), and a refractive index detector (Agilent 1100 Series). The chromatographic conditions were as follows: injection volume, 5 μL; mobile phase, 5 mm H2SO4; isocratic elution, flow rate, 0.5 mL/min; optical unit temperature, 40 °C; and positive polarity. Because of the high reproducibility of the HPLC analysis, samples were analysed only once. The relative standard deviation of each analyte was determined in a preliminary intra-laboratory method validation 13. For all the statistical calculations, the Prism 5 computer software (GraphPad, USA) was used. Statistically significant differences between the data groups of interest were evaluated by one-way analysis of variance (ANOVA) followed by Tukey's Multiple Comparison Test (differences at p > 0.05 were not considered statistically significant). Non-linear regression was performed using the ‘log(inhibitor) vs response − variable slope’ model, inherent to Prism 5 computer software. Tables and graphs were constructed using Microsoft Office Excel 2007. For clarity purposes only mean values are presented in the tables and figures. The brewing attributes of the malt and wort did not differ significantly from previously reported data 13. However, fermentation performance differed substantially, most likely as 19.5 L NC Cornelius steel tanks were used instead of sphero-conical fermenters. Fig. 1 shows the rate of extract consumption and ethanol production during 11 successive fermentations (F1–F11) of barley, buckwheat and quinoa wort. Because the attenuation time (defined as the last day when a daily change in the extract content was ≥0.1%) was raw material- and successive fermentation-dependent, the time axis has been relativized throughout this paper. Similarly, because worts from two (buckwheat and quinoa) or four (barley) different brews were used, the dependent attribute is expressed as a percentage of its content in wort. The profile of extract consumption and ethanol production differed among the successive fermentations. In addition, significant differences were observed between the same successive fermentations of a particular raw material. In order to quantitatively describe these differences, a non-linear regression was performed, fitting the four-parameter symmetrical sigmoidal curve to the experimental data of extract and ethanol content. Three statistics, namely AT50 (the percentage of attenuation time needed to assimilate a half of total consumed extract or to produce a half of final ethanol), curve slope (describes the steepness of the curve; higher absolute values belong to steeper curves) and R2 (percentage of the goodness of fit) were considered for further discussion. Since the yeast was not genetically and phenotypically adapted to buckwheat and quinoa wort, the first three fermentations appeared to represent a phase during which a major acclimatization to a new environment (i.e. new substrate type) occurred (Table 1). As would be expected, this phase was missing in the case of barley. A second phase, which was characterized by a one-day increase in attenuation time, possibly represented a minor acclimatization event. In a consequent phase the attenuation time gradually lessened to a 6 day-long fermentation in all three raw materials. Regardless of the attenuation time differences, the extract consumption and ethanol production within a raw material type remained very consistent over the entire successive fermentation, with a relative standard deviation (RSD) of <3.90%. In general, the goodness of curve fit was very high, most being >99.7% (Table 1). The most obvious deviations from this observation were with the quinoa, where both the extract and ethanol R2 values were <99.7% from the sixth and ninth fermentation on, respectively. The successive fermentations of each raw material differed greatly from each other in their AT50 values (Table 1, Fig. 1). These differences were raw material-dependent and smallest in the case of barley, followed by buckwheat and quinoa. The steepness of the fitted curves for barley extract consumption and ethanol production did not show any overall tendency for change. In the case of buckwheat and quinoa, the initial absolute values of curve steepness were almost twice as high and steadily decreased over successive fermentations. Compared with earlier fermentations, the later ones began more intensively with their extract consumption and ethanol consumption profile closer to a straight line. Regarding final ethanol content, it was previously hypothesized that the lower alcohol production in buckwheat and quinoa was causally connected to lower relative concentrations of the 37.1 kDa protein band in these two pseudocereals, since this could be due to lower expressions of the ADH1 and ADH2 genes 14. General aspects of iron, copper, zinc and manganese in brewery fermentations are discussed in detail elsewhere 15, as well as their content in beer-like beverages from buckwheat and quinoa 13. The uptake and/or release dynamics of these four crucial trace metal ions for brewery fermentation are shown in Figs. 2-4. No levels for iron and copper for barley, and iron for buckwheat, are shown since they were below the limit of quantification for the method (0.05 mg/L). Table 2 shows the absolute concentrations of these metal ions in wort and in the final beverage for barley, buckwheat and quinoa. What all three raw materials, regardless of the particular successive fermentation, had in common was a marked uptake of iron, copper and zinc during the first 24 h, although both absolute and relative amounts varied considerably. This was probably a consequence of biosorption, which is an immediate and fairly non-specific biophysical attachment of metal ions to negatively charged cell wall moieties 15. In contrast, manganese was taken up more slowly and to a lesser degree than the other metal ions; however, the absolute overall uptake was similar between the raw materials despite the large differences in initial concentrations. After the first 24 h, the initial drop in the iron concentration during 11 successive fermentations (Fig. 4) differed significantly being the lowest in F1 and F2, the highest in F3-F6, F8 and F10, and intermediate in F7, F9 and F11. Iron concentrations between 0.055 and 0.165 mg/L are said to be sufficient for normal fermentation 15 but during F1, yeast assimilated almost a 30× higher amount of this metal ion, most likely not by controlled active transport. It appears that no earlier than after the third day of F2, yeast was able to resist the surplus and also probably toxic concentrations of iron. In the following four fermentations (F3–F6), the iron concentrations in the fermentation media were maintained at relatively high levels and from F7 on, it appeared that the yeast began to augment its iron accumulation to some intermediate level. This increase in the iron uptake by yeast could be due to an increased physiological need for iron, a weakening of yeast vitality, changes in yeast cell surface or other factors. In general, the assimilation profiles of copper during the successive fermentations of buckwheat and quinoa wort were very similar (Figs. 3 and 4), as were their initial and final absolute concentrations (Table 2). Nevertheless, one clear difference exists between buckwheat and quinoa – the initial rate of copper uptake. The majority of the copper was assimilated on the first day and later only minor fluctuations in concentration occurred. Day-by-day change did not exceed 20% of initial concentration. Both the first day uptake and the degree of fluctuations declined with increasing number of successive fermentations. The final concentrations of copper were very low after F1–F4 and much higher in later fermentations. The first day uptake of copper was also prominent but lower compared with quinoa. Further uptake gradually declined during the first half of the attenuation time. Inter-day fluctuations were similar to quinoa, but the first day uptake of copper only declined from F1 to F6 and then stabilized around 50% of the initial level. Final concentrations were similar to each other and a general trend in their increase was noticeable. The uptake dynamics of zinc showed that its assimilation by yeast was highly raw material-dependent (Figs. 2-4), although the influence of successive fermentations should not be neglected. In all cases, zinc was rapidly taken up by yeast at the beginning of fermentation, which is in agreement with previous reports that zinc is absorbed by the yeast biomass even before fermentation commences, that is, just after the cells are dispensed into wort 16-18. In our case, the overall absolute uptake appeared to be primarily a function of the initial zinc concentration, except when this concentration was very high, then the influence of successive fermentations started to prevail (Table 3). The first day uptake was in the range of the final day; however, only in the first fermentation did the zinc content remain at its minimal level throughout the process. In all subsequent fermentations, it increased once as the fermentation progressed and dropped again before its end. In all cases, this increase reached a peak value in the first half of the fermentation, whereas its relative value was higher in later fermentations. This marked release of zinc by yeast has been observed previously 19 and it was causally related to a drop in pH during the fermentation process, as that drop decreases the apparent stability constant of the zinc metal ion binders 31. This occurrence was presumably more considerable the more times the yeast slurry was reused. On the basis of our results it can be concluded that successive fermentations of barley wort influence only the temporary release of zinc during fermentation (i.e. the time and intensity of a peak value) and not the initial and overall uptake, the latter being in discordance with other reports 18, 19, most likely because of the low zinc concentration in our wort. The relative uptake of zinc during the first day varied between 15 and 70% of the initial level, but at the end of fermentation, the concentration was <10% regardless of successive fermentation numbers (Fig. 3). In terms of absolute concentrations, overall uptake represented 5- to 10-fold higher values compared with barley (i.e. ~2 mg/L), an amount that would be tolerated by yeast with no excessive harm. Final concentrations were achieved in a more or less linear manner between 20 and 40% of fermentation progression, which probably correlates well with yeast proliferation. As opposed to barley, only during F2 and F3 did the zinc content temporarily increase. The profile of zinc uptake during successive fermentations was reminiscent of the uptake profile of iron, especially that of F1 and F2 (Fig. 4). In particular, it appears that during the second fermentation, yeast was able to withstand the toxic zinc concentrations and this ability improved continuously in subsequent fermentations. Indeed, the zinc uptake values of later successive fermentations reached the so-called ‘tolerable concentration’ of ~2 mg/L, as ascertained for buckwheat above. Furthermore, no temporary single increase in zinc concentration was observed. Fluctuations with no obvious pattern were noted. In general, the manganese uptake profile was characterized by relatively high fluctuations during the fermentation process, with no distinctive patterns (Figs. 2-4). Moreover, the overall uptake was always <50% of the initial manganese concentration. The highest first-day uptake concentration corresponded to F1–F3 and from F4 to F6 initial uptake declined. From F7 onward, steady or increased values were observed. Final concentrations of manganese were 50–80% of the initial level. Manganese concentrations after the first day were 50–120%, and at the end 50–100% of the initial level, with no explicit trend of change over successive fermentations. Relative profiles of manganese uptake in F1 and F2 appeared similar to those of iron and zinc (Fig. 4). In brewing the recommended values of manganese are in the range of 0.11–0.22 mg/L 15, thus as for iron and zinc, yeast may have undergone a phenotypical adaptation to excessive manganese concentrations. In none of the fermentations (with the above-mentioned exception of F1 and F2) were the manganese concentrations significantly changed after the first day. The same was observed for the final concentrations with the exception of F1. The initial content of fermentable carbohydrates (FC) and their ratios in barley, buckwheat and quinoa wort differed considerably (Table 3), similarly to the reports in our previous study 13, where obtained values were discussed in detail. In this study we have focused primarily on the uptake dynamics of FC as a function of successive fermentation and raw material type (Figs. 5-7). For the purpose of quantitative data interpretation, the same principle as in the case of fermentation performance was followed, that is, non-linear regression was performed and three statistics, namely AT50, curve slope and R2 (Table 3), were used to support the discussion. Fructose was exempted from the non-linear regression analysis because of its very low R2 values (data not shown). Furthermore, an in-depth analysis of fructose uptake was dispensable as its low absolute values in wort would not influence the fermentation process substantially. All other R2 values were >95% (99.45% on average). All three raw materials shared a common characteristic, that is, after the sixth successive fermentation there was a small (<5%) increase in relative glucose utilization, mainly at the expense of lower DP2 uptake (Supporting Information 1, 2 and 3). There were also some minor distinctions observed between different raw materials regarding the total FC assimilation (Figs. 5-7). In the case of barley the most FC was assimilated overall (80-90%), followed by buckwheat (~80%) and quinoa (65–75%). This observation correlates with the low overall extract consumption and low final ethanol production in quinoa (Fig. 1). In addition to the relatively low content of FC in quinoa wort, their overall assimilation was also relatively low. The curve slope of the barley did not change much over the course of successive fermentations, whereas with the other two, there was an overall increase. Contrary to barley, in buckwheat and quinoa there was a general tendency over successive fermentations for a faster beginning of fermentation, a situation which is mostly quantitatively explained by a decreasing steepness of curves. As a preferred fermentation carbohydrate, glucose uptake normally begins immediately after yeast is pitched. In some barley fermentations (Fig. 5) the values were even higher after the first day in wort. An enzymatic process took place that yielded glucose, that is, the hydrolysis of sucrose by a periplasm-located invertase 22. Most likely, this occurred in buckwheat (Fig. 6) and quinoa (Fig. 7) but because of their high initial glucose levels this was not noticeable. In all cases, glucose was fermented completely, but sooner in barley than in buckwheat and quinoa. This observation was quantitatively supported by the average values of AT50. From the curve slopes of buckwheat and quinoa, it can be concluded that, after F2, yeast was able to ferment glucose more rapidly from the beginning resulting in a steeper fitting curve. Overall, differences between the barley and both pseudocereals were most likely due to the large differences in initial glucose concentrations. It has already been hypothesized that the differences in absolute and relative concentrations of glucose between worts are connected to different relative concentrations of the 26.3 kDa protein band 14. In particular, this can be due to transcriptional repressor Nrg1p, which mediates glucose catabolite repression and negatively regulates a variety of other processes. In addition, high glucose levels may cause significantly lower relative concentrations of the 38.3 kDa protein band in buckwheat and quinoa 14, which could be a consequence of a lower Reg2p expression, involved in glucose-induced proteolysis of maltose permease. The initial fructose concentrations were very low and thus comparable amongst the raw materials (Figs. 5-7). In almost every fermentation, concentrations rose dramatically after the first day and they stayed above the initial level for a substantial amount of time. In buckwheat, fructose was assimilated to a similar degree as in barley whereas in quinoa only 20% of the initial level at most was assimilated. In quinoa the fructose uptake and release appeared different from the continuous profiles in barley and buckwheat. The topologies of di- and trisaccharide uptake profile did not differ (Figs. 5-7). The uptake of the sugars of two and three degrees of polymerization (DP2 and DP3) should begin just after the majority of glucose has been fermented, as glucose exerts the well-known effect of ‘carbon catabolite repression’. The assimilation in barley started immediately, whereas in buckwheat and quinoa, the consumption of DP3 during the first 40% of attenuation time was rather low. Only in quinoa were there significant differences between successive fermentations, both in slower rate and lower overall consumption of DP2 and DP3, and these differences also showed a directional trend. In barley and quinoa wort fermentations differences between average AT50 values of glucose and DP2/3 were relatively high, whereas in the case of buckwheat, the average AT50 value of glucose differed only slightly from those of DP2 and DP3. This suggests that the fermentation of glucose, DP2 and DP3 proceeded more or less simultaneously, probably because of a higher expression of the 73.8 kDa protein band in the case of buckwheat 14, which raises the possibility that the Mal11p, a high-affinity maltose transporter, was expressed at a higher level. The investigation demonstrated that both the type of raw materials used and the number of times a yeast was repitched influenced fermentation performance, metal ion assimilation dynamics and wort sugar uptake profile. Based on the results, evaluations of buckwheat and quinoa were made from two distinct points of view: (a) whether the pseudocereal is suitable to be commercially used for serial repitching of S. pastorianus TUM 34/70 per se; and (b) whether this suitability includes a substitutional potential for barley as a brewing raw material. In general, both pseudocereals showed higher variations than barley of all examined attributes in the course of successive fermentations. However, in buckwheat the differences were closer to those of barley and fluctuated with no obvious tendency, whereas differences between successive fermentations in quinoa were very prominent and often showed a directional trend. This suggested a general and integral weakening of the yeast, especially from the sixth successive fermentation on. The assimilation of fermentable carbohydrates, especially DP2 and DP3, lessened and the metal ion uptake appeared to be poorly controlled. Comparing buckwheat with barley, fermentability and ethanol production were similar and it appears that, once yeast is fully adapted to a new substrate, its fermentation performance becomes stable. This adaptation phase could perhaps be shortened if yeast starter cultures were pre-conditioned in a similar substrate. Buckwheat wort did not contained toxic concentrations of any of the four most important trace metal ions that would impair the fermentation process, even when the yeast was serially repitched 11 times. Regarding fermentable carbohydrates, a high glucose to DP2 ratio can be avoided with the proper use of supplemental enzymes. With buckwheat, di- and trisaccharides were taken up simultaneously with glucose and fructose, overall assimilation of fermentable carbohydrates was very consistent among successive fermentations and the zinc uptake appeared to be better regulated. Quinoa however showed little substitutional potential for barley. As for buckwheat, the glucose to DP2 ratio could be optimized, but the major issue would be low levels of total fermentable carbohydrates and very high levels of iron, zinc and manganese. It can be concluded from the experimental work, that buckwheat shows good potential for serial repitching of S. pastorianus TUM 34/70, both per se and for the preparation of a gluten-free substitute of barley beer. On the other hand, the serial repitching of quinoa wort with this yeast appears to be limited to five fermentations and only for a preparation of a unique beverage with little resemblance to a barley beer. This work was financially supported by the Slovenian Research Agency (grant no. 01-1222/1-2010) and Slovene Human Resources Development and Scholarship Fund (grant no. 11012-1/2012-7). The authors would like to thank Centre Weihenstephan for Brewing and Food Quality (Freising, Germany) for kindly providing the yeast strain TUM 34/70. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.