The growing interest of governments and industry in developing healthy and natural alternative foods and beverages that will fulfil the consumer drive towards a healthy lifestyle and clean-label, natural diet has led to an increase in traditional lactic acid bacteria fermentation research. In particular, this research aims to address the organoleptic modulation of beverages using in situ-produced bacterial polysaccharides.
Methods and Results
Weissella cibaria MG1 is capable of producing exopolysaccharides (dextran) and oligosaccharides (glucooligosaccharides) during sucrose-supplemented barley-malt-derived wort fermentation. Up to 36·4 g l−1 of dextran was produced in an optimized system, which improved the rheological profile of the resulting fermentate. Additionally, small amounts of organic acids were formed, and ethanol remained below 0·5% (v/v), the threshold volume for a potential health claim designation.
The results suggest that the cereal fermentate produced by W. cibaria MG1 could be potentially used for the production of a range of novel, nutritious and functional beverages.
Significance and Impact of the Study
Using conventional raw materials and traditional processes, novel LAB-fermented beverages can be produced representing an innovative mechanism towards fulfilling the aim to decrease government and personal costs as well as potentially ameliorating consumer lifestyle regarding dietary-related disease.
In recent years, there has been an increase in the number of novel foods (EFSA 2010) designed to have specific health benefits, and the largest sector includes fermented dairy drinks. However, the number of lactose-intolerant individuals globally is around 20%, and in some Asiatic (Japan, China) or African countries, it can reach up to 100% (Swagerty et al. 2002), but varies considerably geographically (Vesa et al. 2000). As such, there is a need for variety in this sector with cereal-based fermented beverages offering a suitable choice or alternative for the obligate consumer, self-diagnosed patients (Nicklas et al. 2009) and lifestyle customers who prefer to avoid dairy products, including those which are lactose free, for other reasons. Other than beers, beverages, such as kvass, bouza, chichi and mahewu, represent traditional low alcoholic or nonalcoholic drinks produced in many parts of the world, which are based on microbial fermentation of cereal extracts (Waters et al. in press), with a generally acidic character due to the role of lactic acid bacteria (LAB) (Gadaga et al. 1999; Gotcheva et al. 2000; Muyanja et al. 2003; Dlusskaya et al. 2008; Prado et al. 2008; Aloys and Angeline 2009). The production of these cereal beverages is usually based on spontaneous fermentations by indigenous microbiota with yeasts and LAB being the dominant micro-organisms (Gotcheva et al. 2000; Muyanja et al. 2003; Zorba et al. 2003; Almeida et al. 2007).
Recently, in addition to these traditional beverages, there is an increasing interest in the production of cereal-based beverages with defined starter cultures for ease of control and reproducibility (Okafor et al. 1998; Zorba et al. 2003; Dlusskaya et al. 2008). Additionally, using a specific starter culture allows manipulation of the functionality, texture, flavour and other characteristics of the final product (Blandino et al. 2003; Prado et al. 2008; Ganzle et al. 2009; Waters et al. in press). This approach potentially allows the development of novel products capable of imparting the product with characteristic organoleptic features due to the production of microbial metabolites, which may include oligosaccharides (OS) and exopolysaccharides (EPS). Certain types of bioactive substances may also provide health benefits as prebiotics, which are defined as ‘a selectively fermented ingredient that allows specific changes, both in the composition and in the activity in the gastrointestinal microflora that confer benefits upon host well-being and health’ (Gibson et al. 2004; Roberfroid 2007) or similarly ‘a non-viable food component that confers a health benefit on the host associated with modulation of the microbiota’ (FAO 2008). Cereal-associated LAB produce a large structural variety of EPS and OS from sucrose through the activity of glycansucrases (Tieking and Ganzle 2005). Homopolysaccharides are polymers composed of either glucose or fructose units and are synthesized from sucrose through the action of extracellular glucansucrases or fructansucrases (glycansucrases), respectively (Korakli and Vogel 2006). In addition to EPS, glycansucrases synthesize OS by transferring glucose or fructose moieties to suitable acceptor carbohydrates, such as maltose (Ganzle et al. 2009).
Over the last two decades, comprehensive data have been generated regarding LAB-mediated polysaccharide production during sourdough fermentation (Brandt and Gaenzle 2006), with the primary focus on improving dough and bread quality, texture, flavour and shelf-life (Galle et al. 2011). However, transferring the existing knowledge from sourdough technology into beverage production exhibits an interesting and promising approach for novel products, albeit with certain technological challenges to overcome.
The aim of this research is to enrich and revalorize highly nutritious barley malt extract using traditional LAB fermentation, resulting in a novel alcohol-free beverage base with desirable textural properties.
Material and methods
Bacterial cultures and screening for exopolysaccharide and oligosaccharide production
Thirty-seven LAB isolates present in cereal environments were maintained as frozen stocks in 80% glycerol at −80°C. LAB were routinely grown on MRS agar plates (Merck, Darmstadt, Germany) under microaerophilic conditions for 48 h at 30°C.
To screen for EPS, the LAB were grown on modified MRS agar (mMRS) (Meroth et al. 2003) plates (Galle et al. 2010), supplemented with 10% sucrose (SucMRS). EPS-positive isolates were defined as those which displaying slimy colony morphology. To characterize EPS/OS production of the best producer strains in SucMRS broth, single colonies were picked from mMRS plates, subcultured once for 16 h in mMRS broth and subsequently inoculated at 1% in SucMRS broth and incubated anaerobically at 30°C for 48 h.
Barley malt and mashing regime
Commercial malt, made from the barley variety Sebastian, which was harvested in 2008, was purchased from the Cork Malting Company and used in these trials. The design of a mashing regime contributing maximum amounts of sucrose and maltose in wort was based on a RSM model as previously described (Mauch et al. 2011). The following mashing regime was applied; 28·1 min at 62°C, 25 min at 72°C and a mash off temperature of 78°C, with a heating rate of 1°C min−1 between the individual rests.
Wort production and fermentation (beverage base production)
A pilot-scale (1000 l) brewhouse was used for wort production and fermentation. First, the malt was milled with a two-roller mill (0·4 mm distance between rollers), and 120 kg of grist was mashed into 540 l of water at 62°C. Wort was boiled for 1 h and adjusted to a final extract content of 9% (w/w) by adding hot water; hops were not added to the wort. Precipitates in hot wort were removed using a whirlpool. For each of the three replicates, hot wort was transferred into a 20-l stainless steel container. Sucrose was dissolved in the hot wort to achieve final concentrations of 5 and 10% (w/v). The wort was then cooled to 30°C in a tempered water bath.
Strains of Weissella cibaria MG1 were subcultured twice in MRS broth and inoculated to the wort to obtain a concentration of 106 CFU ml−1. LAB were inoculated into the cooled wort with subsequent fermentation at 30°C for 72 h. Samples were aseptically withdrawn at 0 h (after inoculation), 24, 48 and 72 h. LAB cell counts and pH were determined after wort inoculation and throughout fermentation. The number of colony-forming units (CFU) was determined by standard microbiological plating assays. The characterization of EPS and OS produced by the strains was described above.
EPS characterization in sucrose-supplemented MRS and wort
EPS were isolated from SucMRS broth using chilled ethanol, with subsequent dialysis and freeze-drying steps. The EPS composition, structure and molecular weight were analysed as previously described (Galle et al. 2010). Briefly, EPS monosaccharide composition was determined after acid hydrolysis, and the linkage type was determined after digestion with dextranase and amyloglucosidase. The molecular weight was analysed with a Superdex 200 Column (GE Healthcare, Baie d'Urfe, QC, Canada) with water as the solvent and a flow rate of 0·4 ml min−1. Detection was by a refractive index detector (RID).
Determination of oligosaccharides produced in MRS and wort
OS patterns of fermented MRS and wort both initially containing 10% sucrose were analysed with a CarbopacPA20 column (Dionex, Oakville, ON, Canada) using water (A), 200 mmol l−1 NaOH (B) and 1 mol l−1 Na acetate (C) as solvents at a flow rate of 0·25 ml min−1. The following gradient was used: 0 min 30·4% B, 1·3% C, 22 min 30·4% B and 11·34% C, followed by washing and regeneration. OS were directly analysed from the cell-free supernatant using maltose and panose as external standards for peak identification. Detection was performed by a RID.
In advance of rheological analyses, fermented wort samples were centrifuged to remove LAB cells and any other particulate matter. The viscosities of these cell-free supernatants (cfs) were measured on a controlled stress rheometer (Physica MCR 301; Anton Paar, Graz, Austria) using a cone–plate system at 20°C. Two millilitres of cfs sample was placed on the plate, and the upper cone (diameter, 74·997 mm) was lowered to 0·04 mm with the removal of excess sample. Measurements were as a function of shear rate over a range of 10–1000 s−1.
Determination of sugars and metabolites in wort
Wort extract was determined using an automated beer analyser (Servo Chem, Tecator AB, Höganäs, Sweden). For the determination of sugars and organic acids, fermentation liquids were mixed in the ratio 1 : 1 with 7% perchloric acid, and proteins were precipitated at 4°C overnight. Maltose, fructose, lactate, acetate and ethanol were determined by high-performance liquid chromatography (HPLC) using an Agilent 1200 HPLC system with a RID. A REZEX 8μm 8% H organic acid column (300 × 7·8 mmol l−1; Phenomenex, Torrance, CA, USA) was used with 0·01N H2SO4 as the elution fluid, at a flow rate of 0·6 ml min−1, at 65°C. Substrate and end product peaks were identified by comparison of their retention times with pure compound standards, and their concentrations were also determined. Sucrose and glucose co-eluted and were not quantified separately.
Thirty-seven LAB strains that had been previously isolated from cereal environments and whose metabolic performance is suited to cereals were screened for their ability to produce EPS on a suitable agar media. Fourteen of the isolates showed slimy colony morphology (Fig. 1a,b) on SucMRS agar (Table 1) and were further investigated as EPS-positive strains (EPS+). These EPS+ isolates were belonging to the species L. arizonensis, L. plantarum and W. cibaria.
Table 1. Exopolysaccharide production by 37 LAB cereal isolates on SucMRS agar
EPS+ and EPS− mean EPS detected (+) through slimy colony morphology after growth on MRS agar plates supplied with 100 g l−1 sucrose or not detected (−), LAB strain codes beginning with ‘F’ or ‘MG’ are from wheat sourdough, ‘E’ is from oat sourdough, and ‘FST’ is from malt.
F17, F8, F7, F6, F13, F12
F10, F2, F21, F22, F25, F26, F3, F34, F36, F4, F9
FST 1.7, FST 1.9
E11, E2, E4
MG1, MG7, F33, F28, F27, F29, F27
EPS characterization in MRS and wort
The 14 EPS+ isolates were cultured in SucMRS broth, and the four strongest producer LAB strains were further analysed. These included the isolates, W. cibaria MG1, MG7, F28 and F33. The types of EPS (Table 2) and OS (Fig. 2) produced by these strains were analysed after 72 h; however, the levels produced remained unchanged after 48 h. The EPS produced by all strains was dextran, which has a glucan backbone consisting of α-1-6 linkages, and the OS were glucooligosaccharides (GOS), confirmed as previously reported (Galle et al. 2010). Dextran was produced in highest amounts by W. cibaria MG1, and as such, this strain was further applied in wort fermentations.
Table 2. Amounts of EPS (dextran, 5 × 106–4 × 107 Da) produced by the strongest LAB producer strains, all of which were from the Weissella genera
Amount of EPS produced by the LAB isolate in 10% sucrose-supplemented media broth after 72 h.
Weissella cibaria (MG1)
36·4 ± 0·6
W. cibaria (MG7)
13·7 ± 1·8
W. cibaria (F28)
13·4 ± 6·1
16·2 ± 1·8
Wort (0·3% sucrose)
W. cibaria (MG1)
1·4 ± 0·2
W. cibaria (MG7)
1·4 ± 0·1
W. cibaria (MG1)
8·6 ± 0·8
W. cibaria (MG7)
7·2 ± 0·4
W. cibaria (MG1)
14·4 ± 1·2
W. cibaria (MG7)
10·6 ± 0·7
Cell growth, pH, formation of metabolites and utilization of sugars during wort fermentations
During wort fermentations, bacterial growth was monitored through the determination of pH and cell counts. W. cibaria MG1 produced the highest amount of EPS amongst all strains in broth and wort fermentations and was therefore applied for further analysis. In the nonsucrose-supplemented wort (csucrose = 0·3%), W. cibaria MG1 grew from c. 106 CFU ml−1 to c. 2·0 × 108 CFU ml−1 after 24 h of growth. After 72 h, the amount of viable cells in the wort decreased during fermentation (6·4 × 107 ± 1·5 × 107 CFU ml−1). A decreased cell count was also observed for W. cibaria MG1 upon sucrose supplementation at a final concentration of 5 or 10%. The pH of all worts dropped to 4·0 after 24 h of fermentation, and this acidification profile was reflected in the lactic and acetic acids levels (Table 3). These organic acids were primarily produced during the first 24 h of fermentation, particularly for the former metabolite. Generally, sucrose-supplemented worts had slightly reduced lactic acid formation, whilst acetic acid formation was independent of sucrose concentration. The formation of ethanol by the W. cibaria MG1 was negatively correlated with the initial level of sucrose, and as such, the highest amounts were produced in nonsupplemented worts.
Table 3. Formation of acetic and lactic acids, fructose and ethanol as well as utilization of maltose during Weissella cibaria MG1 fermentation of worts containing different initial amounts of sucrose of c. 0·3, 5 and 10%
Product (g l−1)
Initial sucrose (% w/v)
ND, not detected.
Standard deviation was below 10% for all data reported.
Fructose formation by W. cibaria MG1 in wort was positively correlated with the initial concentration of sucrose present (Table 3), with the latter being depleted as fermentation progressed. An approximate conversion rate of 98% was found in 10% sucrose-supplemented wort, and the use of maltose was greater in the 5 and 10% sucrose-supplemented worts by up to 34 and 47%, respectively (Table 3). The combined HPLC peak representing sucrose and glucose showed that the W. cibaria MG1 isolate used consumed more than 90% of these sugars after 72 h of fermentation in all worts (data not shown).
Formation of exopolysaccharides and oligosaccharides
The maximal EPS production during W. cibaria MG1 wort fermentations was exhibited in the 10% sucrose-supplemented sample (Table 2). Fermentation of sucrose nonsupplemented wort resulted in the lowest EPS production levels (1·4 g l−1).
The OS patterns of original and fermented worts are shown in Fig. 2, but were not quantified due to the lack of commercially available standards. The fermented wort OS patterns were comparable to previous reports (Galle et al. 2010) in W. cibaria MG1-fermented 10% sucrose-supplemented MRS broth. Panose was identified by external standards with subsequent peaks likely representing the OS: 6′glucosylpanose, 6′6′diglucosylpanose and higher OS of the same series (Fig. 2). Sucrose nonsupplemented W. cibaria MG1-fermented wort was used as a control and showed baseline levels of OS production.
Rheology of fermented worts
The apparent viscosities (η) as a function of the shear rate of the fermentation liquids of W. cibaria MG1-fermented (24 and 72 h) and unfermented wort are shown in Fig. 3a–c. All fermented worts displayed a shear-thinning behaviour with W. cibaria MG1 after 72 h of fermentation exhibiting the highest viscosities, when the wort was supplemented with 5 or 10% sucrose (Fig. 3b,c). Sucrose unsupplemented unfermented wort exhibited a slight shear-thinning behaviour (Fig. 3a); however, the addition of sucrose increased the viscosities, and the worts developed Newtonian behaviour (Fig. 3b,c). Increasing initial concentrations of sucrose in worts fermented with W. cibaria led to increased differences in viscosities to the respective unfermented wort, at both low and high shear rates, indicating the formation of polymers.
LAB screened in this research originated from cereal or processed cereal environments. In total, 38% of the cereal isolates proved to be EPS-producing strains, with all of these strains also producing OS, as seen for W. cibaria MG1 in Fig. 1. The importance of in situ production of EPS and OS through natural fermentation technologies lies in their functional applications regarding their textural amelioration abilities as well as potential health effects. Certain OS such as galacto-/fructo-/gluco-/arabinoxylan OS are recognized as prebiotics (Rycroft et al. 2001; Grootaert et al. 2007), in the prevention of allergic diseases (Kukkonen et al. 2007), to support growth of gut-friendly microbial probiotics (Kondepudi et al. 2012) and to aid immunity through suppression of inflammatory markers (cytokines) (Macfarlane et al. 2009; Bengmark 2012). Additionally, the benefits of these EPS and OS producer strains combined with probiotic LAB could potentially lead to the production of a synbiotic, clean-label, cereal-based beverage (Grimoud et al. 2010).
Weissella cibaria MG1 dominated the (sucrose-supplemented) barley wort fermentation and produced 14 g l−1 EPS. The high maltose concentration, compared with sucrose, in wort diverted dextransucrase activity from EPS production to the formation of panose-oligosaccharides. Because the LAB analysed in this research originated from cereal or processed cereal environments, their suitability to dominate wort during fermentation was presumed. This theory was tested for the highest EPS producer LAB, W. cibaria MG1. Even though the strain dominated the (sucrose-supplemented) wort fermentation, it was found to produce EPS with lower efficiency than in SucMRS. Higher maltose concentration in wort led to the increased formation of OS at the expense of EPS production levels.
LAB naturally produce EPS and OS, thus distinguishing them as microbes with potential for the development of technologically functional ingredients, which can be produced in situ to formulate alternative, nonalcoholic, cereal-based beverages. Additionally, diversifying the brewhouse product portfolio through incorporation of traditional fermentation technologies represents an economical approach to raw material revalorization. This evolution of product conceptualization allows brewers to implement these technologies in their existing plants, whilst concurrently continuing their regular brewing objectives, thus incurring no additional investment.
This study shows that the LAB W. cibaria MG1 can dominate a barley-malt-derived wort environment to produce up to 14·4 g l−1 of the homoexopolysaccharide dextran, resulting in a novel and nutritious beverage base with prospects for further advancements using tailored microbial cultures. The yield of EPS is dependent on sucrose concentration, the LAB strain used and growth conditions, amongst others. Barley malt wort produced for beer production naturally contains between 0·3–6·0 g l−1 of sucrose and 39–60 g l−1 of maltose (Boulton and Quain 2001), characteristics which can be manipulated through optimization of the mashing programme; however, sucrose supplementation is still necessary to obtain sufficient yields of EPS. The barley malt invertase/glucansucrase enzymes function optimally at a lower temperatures than β-amylase maltogenesis (Narziß 2004). As such, using an elevated temperature during the initial step of wort production will maximize saccharification. As such, to reiterate, regardless of mash parameters, sucrose supplementation is necessary to allow efficient EPS and OS production by W. cibaria MG1 in wort.
As the ratio of sucrose to maltose increased, analyses of the W. cibaria MG1 wort fermentation revealed that EPS production was positively correlated. This effect was previously reported for a purified dextransucrase of Leuconostoc mesenteroides (Paul et al. 1986; Heincke et al. 1999). The depletion of maltose, a strong acceptor sugar for dextransucrase (Killey et al. 1955), in sucrose-supplemented worts indicates OS formation, as confirmed by the W. cibaria MG1 SEC-HPLC analyses. In the presence of maltose, W. cibaria MG1 is known to produce panose and higher OS of the panose series (Galle et al. 2010), with the concomitant accumulation of fructose during fermentation indicating an almost complete conversion of sucrose. In keeping with previous observations (Galle et al. 2010), W. cibaria MG1 did not reduce fructose to mannitol (data not shown).
During W. cibaria wort fermentation, due to a slightly negative correlation between lactic acid production and sucrose levels combined with the positive correlation with acetic acid production, the fermentation quotients decreased. These low acetate formation levels are beneficial in food processing due to the negative unpleasant flavour attributes associated with higher amounts (Galle et al. 2010). Additionally, ethanol formation by the heterofermentative W. cibaria MG1 during wort fermentation was negatively correlated with sucrose content. Additionally, as sucrose levels increase, one of its hydrolysis products (glucose not fructose) acts as building blocks for EPS production, whilst maltose concurrently acts as a an acceptor sugar for the glucansucrases catalysing the transferase reaction (Galle et al. 2010).
The formation of EPS by W. cibaria MG1 significantly influenced the rheological behaviour of wort. Because dextran formed by W. cibaria MG1 solely consists of unbranched α-1,6 glycosidic linkages (Galle et al. 2011), it is linear in character. The levels of EPS formed were positively correlated with the initial sucrose concentration, and EPS production continued throughout the complete fermentation time. Additionally, because the EPS remained stable after the completion of fermentation, this dextran potentially represents a suitable viscosity enhancer for beverages over their storage time.
Given the current government, industry and consumer drive towards a healthy lifestyle, dietary amendments are a positive way to implement a widespread, effective and simple societal reform. This can have a significant and positive repercussive effect on personal and government medical costs related to noncommunicable diseases, such as cardiovascular, kidney and liver diseases, certain cancers and obesity-related chronic illnesses, particularly in Western developed countries. However, these trends place substantial pressure on the food and beverage industries to be innovative by reformulating current, or reinventing novel products, in line with consumer expectations and regulatory authority stipulations. Thus, using conventional locally produced raw materials (such as cereal grain extracts) and traditional processes (such as fermentation and brewing), to produce a novel product (such as LAB-fermented beverages), represents an innovative yet technologically challenging mechanism towards fulfilling this aim. Therefore, producing a naturally fermented cereal-based beverage from wort easily accomplishes these requirements, with the capacity to produce tailor-made products addressing specific consumer requirements, such as low-calorie, high-vitamin and high-mineral, cholesterol-lowering, prebiotic and flavoursome beverage with desirable organoleptic attributes. Furthermore, the fermented beverage produced in the current work had ethanol levels <0·5% (v/v), the threshold volume for beverages with a health claim designation (Kreisz et al. 2008). Additionally, the viscosity obtained using the in situ EPS production by LAB also contributes to a shelf-stable desirable body in the final formulation. Finally, all these attributes can potentially be realized in a functional beverage using standard brewing equipment and widely available raw materials.
Funding for this research was provided under the Irish National Development Plan, through the Food Institutional Research Measure, administered by the Department of Agriculture, Fisheries & Food, Ireland. The authors would like to thank Clarissa Schwab, Dan Walsh and Roland Kerpes for their help and assistance.