Eur Food Res Technol (2008) 227:261266
DOI 10.1007/s00217-007-0719-4
ORIGINAL PAPER
Microbial and chemical analysis of a kvass fermentation
Elena Dlusskaya Andr Jnsch Clarissa Schwab
Michael G. Gnzle
Received: 5 March 2007 / Revised: 10 May 2007 / Accepted: 28 June 2007 / Published online: 25 July 2007
Springer-Verlag 2007
Abstract Kvass is a fermented cereal beverage which pro-
duced from malt, rye Xour, stale rye bread, and sucrose and
is consumed mainly in Eastern Europe. Moreover, it is of
interest as one of the few examples of traditional, non-alco-
holic cereal-based beverages. In this study, a commercial
kvass sample was characterized with respect to the fermen-
tation microXora and the concentration of microbial metabo-
lites. Lactobacillus casei and Leuconostoc mesenteroides
were present in cell counts of 7.3 107 and 6.0 107 cfu
mL1, respectively. Saccharomyces cerevisiae was present
in cell counts of 3.0 107 cfu mL1. PCR-DGGE analysis
veriWed that all dominant fermentation organisms were cul-
tivated. Microbial metabolites in kvass were ethanol, lactate,
and acetate. One of the kvass isolates, Ln. mesenteroides
FUA 3086 harboured a putative dextransucrase gene and
formed dextran and isomalto-oligosaccharides from sucrose
and maltose. Fermentation of model kvass wort conWrmed
that all kvass isolates grew in the fermentation substrate,
moreover, formation of isomaltotriose by Ln. mesenteroides
FUA 3086 was observed in model kvass fermentation.
Introduction
Kvass is a non-alcoholic cereal beverage which is tradition-
ally produced from rye and barley malt, rye Xour, and stale
E. Dlusskaya A. Jnsch C. Schwab M. G. Gnzle (&)
Department of Agricultural, Food and Nutritional Science,
University of Alberta, 4-10 Ag/For Centre, Edmonton,
AB, Canada T6G 2P5
e-mail: mgaenzle@ualberta.ca
A. Jnsch
Technische Universitt Mnchen,
Lehrstuhl Technische Mikrobiologie, Freising, Germany
rye bread in eastern European countries. Two main kvass-
making techniques exist using either stale sourdough bread
or malt as main raw materials (Fig. 1). In kvass fermenta-
tions from stale sourdough bread, all sugars needed for
yeast fermentation are derived from the bread-making pro-
cess, in the second technique, gelatinized starch is cleaved
by malt enzymes. Prior to fermentation, the kvass batter is
diluted in boiling water and clariWed by sedimentation [1].
Sucrose is added to the kvass wort and fermentation is initi-
ated by addition of bakers yeast or a previous batch of
kvass. Fermentation is terminated by chilling of the product
to 4C before nutrients are exhausted. Kvass has a golden-
brown color, the pleasant Xavor of rye bread, low sweet-
ness, is not obviously alcoholic and has suYcient carbonate
to give kvass a sparkle on the palate. The Wnal product con-
tains carbohydrates, proteins and amino acids, lactic and
acetic acids and vitamins originating from the raw materials
or as a result of microbial activity. The predominant carbo-
hydrates are maltose, maltotriose, glucose, and fructose, the
ethanol content is 1% or less [2]. Kvass is considered to be
spoiled if ethanol accumulates to higher levels. As opposed
to sourdough bread and related fermented cereal products,
kvass undergoes no heat processing after fermentation and
thus contains high cell counts of viable yeast and lactic acid
bacteria.
The usage of artisanal starters which are maintained by
continuous back-slopping remains the most common way
to start kvass fermentations. The microXora of kvass fer-
mentation is consistently composed of lactic acid bacteria
and Saccharomyces cerevisiae [1, 3] but the composition
on species level is expected to be quite variable due to
diVerences in fermentation techniques and feedstock. To
date, microbial associations in kvass fermentations are
poorly investigated. In one study, Lactococcus lactis var.
diacetilactis was proposed as a starter culture in combination
123
262
Fig. 1 Flow charts of two main
kvass-making techniques Stale rye bread
Cutting to 2-3 cm cubes,
drying
Mixing with hot water (ca.
1:10 solid to liquid ratio)
Incubation for 12 h at
room temperature
with S. cerevisiae to improve aroma and organoleptic prop-
erties of kvass [4]. It was the aim of this study to character-
ize a kvass sample with respect to the microXora
composition and microbial metabolites.
Materials and methods
Microbial isolation and enumeration
Kvass (All Stars Bakery, Toronto, ON, Canada) was
obtained from a local supermarket. According to the sup-
pliers information, the kvass was produced from sour-
dough rye bread and malt using bakers yeast as starter
culture. Cell counts in kvass samples were determined by
surface plating of serial dilutions on modiWed MRS agar
[5] with addition of 100 mg L1 of cycloheximide (Sigma,
Oakville, Canada) or 300 mg L1 of chloramphenicol
(Sigma) for enumeration of bacteria and yeasts, respec-
tively. Plates were incubated for 48 h at 30C under
aerobic conditions for yeasts and both aerobic and micro-
aerophilic (10% CO2, 1% O2, balance N2) conditions for
bacteria. Colonies diVering in color and morphology were
counted separately. Cells from four colonies of each mor-
phological type were puriWed by serial dilution streaks on
modiWed MRS agar until uniform colony morphology was
obtained. Stock cultures of isolates were maintained at
80C in 30% glycerol. Subsequent cultivation of cultures
was done in mMRS at 30C for bacterial isolates and in
yeastpeptonedextrose medium (YPD, 10 g L1 yeast 1
L of template DNA. PCR fragments were separated on
extract, 20 g L1 peptone, 20 g L1 dextose, pH 6.5) at
30C for yeast isolates.
123
Eur Food Res Technol (2008) 227:261266
Rye malt Rye flour (boiled with
excess water)
Mixing and
saccharification
Clarification
Addition of starter (bakers yeast
or previous batch) and sugar;
fermentation for 12 h at ~20C
Decanting, bottling,
refrigerated storage.
DNA isolation
DNA from bacterial isolates and total DNA from kvass
were extracted using DNeasy Tissue Kit (QIAGEN, Mis-
sisauga, ON, Canada). DNA from yeasts was prepared
according the following protocol: Cell from an overnight
culture in YPD were resuspended in 200
L of lysis buVer
(2% Triton X-100, 1% SDS, 100 mM NaCl, 10 mM Tris
HCl, 1 mM EDTA, pH 8.0) and subjected to two cycles of
freezing tubes at 80C freezer followed by thawing in a
95C water bath. Tubes were vortexed for 30 s, 200
L of
chloroform was added and vortexed for two more minutes,
the aqueous layer was transferred to a tube containing
400
L of ice-cold 100% ethanol. After storage for 5 min to
allow for precipitation of DNA, tubes were centrifuged at
14,000 rpm for 5 min, the pellet was washed with 0.5 ml of
70% ethanol and air dried. DNA was resuspended in 20
L
of TE buVer (10 mmol L1 Tris, 1 mmol L1 EDTA, pH
8.0).
IdentiWcation of isolates
RAPD analysis was performed using DNA of all bacterial
isolates according to Mller et al. [6] with the oligonucleo-
tide primer M13V. The reaction volume was 50
L and
contained the following: 5
L 10 reaction buVer,
1.5 mM MgCl2, 200 nM each of deoxynucleotides, 1.5 U
Taq polymerase, 150 pmol of primer M13V (all reagents
from Invitrogen Corporation, Carlsbad, CA, USA) and
1% agarose (Invitrogen Corporation) gel in 1 TBE
buVer (80V). A 1 kb Plus DNA ladder was included as a
Eur Food Res Technol (2008) 227:261266
size marker. RAPD patterns were visualized by UV transil-
lumination after staining in ethidium bromide.
Three isolates of each bacterial strain detected by RAPD
analysis were identiWed to species level by sequencing of
the hypervariable region of bacterial 16S RNA. Bacterial
16S rRNA genes were ampliWed with universal primers
616V and 630R [7] and 28S rRNA genes of yeasts isolates
were ampliWed with primers P1 and P2 [8]. PCR products
were sequenced and analyzed by NCBI BLAST.
Denaturing gradient gel electrophoresis
PCR-DGGE analysis was carried out with DNA isolated
directly from kvass and DNA of bacterial strains isolated
from kvass samples. The hypervariable V3 regions (corre-
sponding to nucleotides 339539 in Escherichia coli) of the
16S RNA genes were ampliWed by using PCR with univer-
sal primers HDA1-GC and HDA-2 and DGGE was carried
out as described previously [9]. Experiments were per-
formed in duplicate.
HPLC analysis of commercial and model kvass
Carbohydrates were separated using a Supelcosil LC-
NH2 (5
m, Supelco, Oakville, ON, Canada) column and
detected with an evaporative light scattering detector (All-
tech 500 ELSD, Alltech, Guelph, Canada). HPLC grade
water (A) and acetonitril (B) were used as mobile phase
(0 min 90% B, 1.5 mL min1; 25 min 75%, 1.5 mL min1;
26 min 50% B, 0.9 mL min1; 32 min 50% B,
0.9 mL min1, 33 min 30% B, 0.9 mL min1; 40 min 30%
B, 0.9 mL min1; 45 min 90% B, 1.5 mL min1). Peaks
were assigned using glucose, maltose, maltotriose, isomal-
totriose, fructose, sucrose (all Sigma), and an isomaltooli-
gosaccharide preparation (BioNeutra, Edmonton, Canada)
as external standards. Lactate, acetate and ethanol were
quantiWed using an Aminex HPX-87H column maintained
at 70C with 5 mM H2SO4 as solvent at 0.4 ml min1. All
concentrations of substrates and metabolites as determined
by HPLC were reported in mmol L1. HPLC analysis was
performed in duplicate and the experimental error was 10%
or less.
Oligo- and polysaccharide formation from sucrose
Ln. mesenteroides FUA3086 and L. casei FUA3087 were
grown in mMRS containing 100 g L1 sucrose, 100 g L1
sucrose and 20 g L1 maltose, or 20 g L1 maltose, and
20 g L1 sucrose. Oligosaccharides and polysaccharides
present within the supernatants were analysed using thin
layer chromatography. Carbohydrates were separated on
TLC plates (Whatman, London, UK). About 1 or 5
L of
supernatants and 2
L of standard solutions (maltose and
263
glucose 100 mM, isomaltooligosaccharides 100 mg mL1,
and dextran 150 g L1 (Sigma) were loaded on plates. TLC
plates were run thrice in a butanol : ethanol : water mixture
(5:5:3, v/v/v%) and stained with an ethanol solution con-
taining
-naphtol (w/v 0.3%) and H2SO4 (v/v 5%). The
staining was developed at 110C for 10 min [10]. Polysac-
charides were present when a dark purple ring was visible
after staining.
Dextransucrase gene in Lc. mesenteroides FUA3087
The genome of Ln. mesenteroides FUA3086 was screened
for the presence of a dextransucrase using PCR. Primers
LcGlu F (5GTTCTTCGTGTCTATTATGG-3) and
LcGlu R (5ATGGAACACTTTATCACTCG-3) targeting
a 563 bp region within dextransucrase genes of Leuconos-
toc spp. were constructed based on homologous regions of
dextransucrases of Ln. mesenteroides L0309
(AY743959.1), Ln. mesenteroides Lcc4 (AY017384), Ln.
mesenteroides B-512F (LMU81374).
Preparation of model kvass
Ln. mesenteroides FUA3086, L. casei FUA3087 and
S. cerevisiae FUA4009 were subcultured twice in overnight
mMRS, washed, and inoculated in a model kvass wort con-
taining 20 g L1 malt extract (Difco, Oakville, Canada) and
20 g L1 sucrose. Inoculum levels were approximately 106
and 104 CFU mL1 for bacteria and yeast, respectively.
Bacteria were inoculated alone, together and together with
yeast at room temperature and samples were taken after 8
and 24 h. Cell counts of bacteria were determined on
mMRS agar plates containing 4 mg L1 cycloheximide.
Yeasts were enumerated on mMRS plates containing
100 mg L1 chloramphenicol. Experiments were performed
in duplicate and results are reported as means SD.
Results
Microbial analysis of kvass
The viable cell count of the kvass sample was 1.6 108 cfu
mL1, bacteria and yeasts accounted for 80 and 20% of the
total population, respectively. A total of total 32 bacterial
isolates and two yeast isolates were obtained from highest
dilution (106) plates. In agreement with previous reports
on RAPD typing of lactic acid bacteria [6], the patterns
generated for each strain contained 46 bands sized
between 300 and 3,000 bp (data not shown). RAPD pat-
terns from all bacterial isolates were virtually identical to
one of two diVerent RAPD patterns obtained from the
Kvass isolates (data not shown). RAPD typing is known to
123
264
discriminate on strain- or species level. As all 32 isolates
were obtained from the same sample, strains with indistin-
guishable RAPD patterns likely represent clonal isolates of
two diVerent strains or species. Three isolates of each of the
two RAPD patterns were subjected to taxonomic identiWca-
tion based on partial 16S RNA gene sequencing. Three
strains were identiWed as Lactobacillus casei (99% identity
in 573 bp to L. casei ATCC 334, Acc. No. AF404708), the
remaining three were identiWed as Leuconostoc mesentero-
ides (98.5 and 98.1% identity in 557 bp to Ln. mesentero-
ides spp. dextranicum CECT 912T, Acc. No. AB023246,
and Ln. mesenteroides spp. mesenteroides DSM 20343T,
Acc. No. M23034, respectively) [11] and assigned the
strain numbers L. casei FUA 3087 and Ln. mesenteroides
FUA 3086. Taking into consideration the diVerent colony
morphologies of the two bacterial species, the viable cell
counts were estimated as 7.3 107 and 6.0 107 cfu
mL1 for L. casei and Ln. mesenteroides, respectively.
Both yeast isolates were identiWed as Saccharomyces cere-
visiae after sequencing of their 28S RNA gene.
Culture-independent analysis of the bacterial microX-
ora was carried out to verify that all numerically domi-
nant microorganisms were cultivated. PCR-DGGE
analysis of DNA isolated from kvass indicated the pres-
ence of two dominant bacterial species in the sample
(Fig. 2). The bands could be assigned to the species Ln.
mesenteroides and L. casei by comparison with the band-
ing patterns obtained with strains L. casei FUA 3087 and
Ln. mesenteroides FUA 3086 isolated from the same
kvass sample.
Fig. 2 PCR-DGGE analysis of kvass. The following templates were
used, Lane 1, DNA isolated from kvass, lane 2, DNA from L. casei
FUA3087, lanes 3 and 4, DNA from Ln. mesenteroides FUA3086. The
experiment was performed in duplicate and representative results are
shown
123
Eur Food Res Technol (2008) 227:261266
Biochemical analysis of kvass
The pH of kvass was 3.45. Ethanol, lactate, and acetate
were present at concentrations 220, 6.5, and 4.1 mM,
respectively. Fructose, glucose, and maltose were present at
levels of 70, 38, and 22 mM, respectively. Maltotriose and
maltodextrins were present in much smaller concentrations.
Oligo- and polysaccharide production by Ln. mesenteroides
FUA 3086
The addition of sucrose to kvass wort enables the formation
of poly- and oligosaccharides during fermentation by bacte-
rial glycosyltransferases. Dextransucrase activity is fre-
quently found in strains of the genus Leuconostoc. To
detect exopolysaccharide synthesis by the two bacterial iso-
lates from kvass, Ln. mesenteroides FUA 3086 and L. casei
FUA3087 were grown in mMRS containing 100 g L1
sucrose and culture supernatants were analysed by TLC.
During growth in mMRS containing sucrose, Ln. mesen-
teroides FUA3086 formed a polysaccharide (data not
shown). An oligosaccharide was additionally detected
when Ln. mesenteroides FUA3086 was grown in the pres-
ence of maltose and sucrose (20 and 100 g L1 or 20 and
20 g L1, respectively). The oligosaccharide could be tenta-
tively identiWed as isomaltotriose using TLC (data not
shown). Neither oligo- nor polysaccharides were detected
in supernatants of L. casei FUA3087. Using primers LcGlu
F and LcGlu an amplicon of the expected size, 563 bp, was
obtained from genomic DNA of Lc. mesenteroides FUA
3086, indicating the presence of a dextransucrase gene.
Formation of isomaltotriose during model kvass
fermentation
Cell counts, pH, organic acid formation and carbohydrate
utilization of the kvass isolates was evaluated during a
model kvass fermentation. Model kvass wort contained
58.0 mM sucrose, 24.4 mM glucose, 9.7 mM maltose,
3.1 mM maltotriose and higher maltooligosaccharides, and
the initial pH was 4.5. In single culture, Ln. mesenteroides
FUA 3086 and L. casei FUA3087 grew to cell counts of
3.3 1 107 and 4.1 4.5 107 cfu mL1 within 24 h,
respectively, and acidiWed the model kvass wort to pH 3.4
and 3.5, respectively. In keeping with their identiWcation as
obligate heterofermentative and homofermentative organ-
isms, Ln. mesenteroides formed lactate and acetate in a
molar ratio of 3:1 whereas lactate was the sole product of
hexose metabolism by L. casei. In mixed culture experi-
ments, Ln. mesenteroides and L. casei grew to cell counts
of 1.9 1.6 107 and 2.3 1.3 106 cfu mL1, respec-
tively. Cell counts of S. cerevisiae FUA4009 increased to
5.1 2.1 105 cfu mL1 in 24 h. Oligosaccharide analysis
Eur Food Res Technol (2008) 227:261266
revealed that isomaltotriose was formed by Ln. mesentero-
ides during growth in model kvass (Fig. 3). In single cul-
ture of Ln. mesenteroides , 0.8 0.1 g L1 isomaltotriose
were formed; 1.0 0.1 and 0.8 0.1 g L1 were found in
model kvass inoculated with both bacterial kvass isolates
and after fermentation with both bacteria and S. cerevisiae,
respectively.
Discussion
This study provides a microbial and chemical analysis of a
commercial kvass fermentation. The kvass microXora was
characterized by a combination of lactic fermentation with
alcoholic fermentation. Two species of lactic acid bacteria
were identiWed by sequencing of their 16S rRNA
sequences, Lc. mesenteroides and L. casei. Culture-inde-
pendent PCR-DGGE analysis conWrmed that cultivation
retrieved all bacterial species that were present in relevant
cell counts. Because rDNA rather than rRNA was used as
template for PCR-DGGE analysis, this method also detects
cells that are dead or dormant after processing or transport
[12]. The cell counts of bacterial isolates in model kvass
fermentations were in agreement with the cell counts in the
kvass sample.
The nomenclature of the L. casei/L. paracasei group is a
matter of ongoing discussion. Three phylogenetic clusters
are readily diVerentiated on the basis of multi-locus
sequence analysis [1315]. One of these units comprises
strains of L. rhamnosus, a second unit comprises L. zeae
LMG 17315T and L. casei ATCC 393T, and a third unit
comprises L. casei ATCC 334 and L. paracasei ATCC
Fig. 3 HPLC analysis of model kvass fermentations using Ln. mesen-
teroides FUA3086 (lower trace), L. casei FUA3087 (middle trace), or
Ln. mesenteroides FUA3086, L. casei FUA3087 and S. cerevisiae
FUA4009 in mixed culture (upper trace). The model kvass wort con-
tained malt extract (20 g L1) 20 g L1 sucrose and was fermented for
24 h at room temperature ( 20C)
265
25598T. Our nomenclature thus follows the suggestion to
use L. casei ATCC 334 as neotype for L. casei [13, 16].
Kvass is similar to boza, a Turkish, non-alcoholic fer-
mented beverage, with respect to the composition of the
Wnal product as well as the microXora [17] but the charac-
terisation of kvass microbiota on species level has hitherto
not been reported. The high cell counts of S. cerevisiae are
obviously derived from the use of bakers yeast to initiate
the fermentation and the yeast produced the lions share of
ethanol and carbon dioxide present in the sample. The lac-
tic acid bacteria present in the sample likely originate from
malt which is used as raw material, from bakers yeast, or
originate from the bakery environment as sourdough bread
is produced in the same facility. Bakers yeast is frequently
contaminated with high cell counts of lactic acid bacteria
that grow to high cell counts upon prolonged fermentation
in cereal substrates, e.g. sponge doughs [18]. The coexis-
tence of homo- and heterofermentative lactic acid bacteria
and yeast is typical for some specialty beers such as Ber-
liner Weie and has been frequently reported for traditional
(type I) sourdough fermentations [19, 20] Ln. mesentero-
ides is frequently found in fermented food of plant origin
and has been repeatedly isolated from sourdough [21, 22].
Strains of the L. casei group are found in a variety of man-
made and natural habitats, including cheese, sourdoughs,
and the intestine of mammals [21, 23, 24]. Current com-
mercial applications focus on its use as probiotic culture
[25]. The isolation of L. casei from kvass conWrms previous
observations that the spectrum of Lactobacillus species
found in the intestine and certain cereal fermentations
strongly overlap [9, 12, 26, 27]. Cereal fermented foods are
thus a suitable alternative to milk-based carriers for probi-
otic bacteria and kvass containing high cell counts of viable
lactic acid bacteria may serve as a blueprint for the devel-
opment of cereal-based probiotic functional beverages.
The description of dextran formation by Leuconostoc
species (betacoccus) dates back to Orla-Jensen [28], who
described their ability to form polysaccharides from
sucrose and the formation of dextran during the extraction
of sugar beets. In the past decade, more than a dozen dex-
transucrases from Leuconostoc spp. and related genera
were characterized on the genetical and biochemical level
[29]. The enzymes generally catalyse the synthesis of di-
and oligosaccharides in addition to polymer formation; the
oligosaccharides formed include leucrose, panose, isomal-
tose, and the corresponding higher oligosaccharides [30
32]. Ln. mesenteroides FUA 3086 formed dextran or isom-
alto-oligosaccharides from sucrose or sucrose and maltose,
respectively, in mMRS. Isomaltotriose was detected in
model kvass fermentation and isomaltotriose formation in
kvass may be attributable to a dextransucrase detected in
strain FUA 3086 by PCR-screening. The presence of isom-
altotriose as the major oligosaccharide formed in model
123
266
kvass reXects the availability of glucose as dominant accep-
tor carbohydrate. Isomaltotriose was not detected in the
kvass sample, possibly because the overall carbohydrate
turnover in kvass was mainly attributable to yeast metabo-
lism. Sucrose hydrolysis by yeast invertase competes with
transglucosylation by dextransucrase.
Isomaltotriose and higher isomalto-oligosaccharides
only partially digested in the small intestine [33], therefore,
they are currently applied in food as alternative, low caloric
sweetener and as prebiotic. Prebiotic compounds are not
adsorbed in the small intestine and stimulate growth and
metabolism of biWdobacteria, resulting in an increase in the
number of biWdobacteria in the intestine [33, 34]. Several
studies reported a prebiotic activity of isomalto-oligosac-
charide [35, 36]. The amount of isomaltotriose in model
kvass was well below the dose of 10 g isomalto-oligosac-
charides that were reported to exert a biWdogenic eVect in
humans [37], however, the current knowledge on oligosac-
charide synthesis by glucosyltransferases enables a directed
optimisation of oligosaccharide formation in kvass. Other
non-digestible poly- and oligosaccharides present in kvass
may further contribute to a prebiotic eVect.
In conclusion, we have identiWed L. casei, Ln. mesen-
teroides and S. cerevisiae as dominating members of the
microXora of a kvass fermentation. The yeast originates
from the use of bakers yeast as inoculum. Although the
composition on species level of kvass samples of diVerent
origin can be expected to be quite variable due to diVer-
ences in fermentation techniques and feedstock, the study
provides a Wrst step towards the characterization of this
fermented, cereal-based soft drink. Kvass is currently
produced predominantly at the artisanal and household lev-
els, limiting the economic prospects for starter culture
development. However, kvass may serve as an example for
non-alcoholic cereal-based beverages. The presence of high
cell counts of viable lactic acid bacteria and the possibility
to produce prebiotic carbohydrates during fermentation
provides new opportunities for the development of func-
tional, cereal -based fermented beverages.
References
1. Korolev DA (1963) Russkii kvas (Russian kvass) Moscow, Pishc-
hepromizdat
2. Ivanova EG, Kiseleva LV, Lenets NG (2006) Pivo I Napitki 2:50
51
3. Ermakov AA, Sitalo NA, Ivanov VS, Yakubovich FF, Emelyano-
va NA (1982) Pishevaya Promyshlennost 4:3031
4. Kosminskii GI, Tsed EA, Yakirevich LM (1996) Izvestiya Vyss-
hikh Uchebnykh Zavedenii Pischevaya Tekhnologiya 5/6:4748
5. Stolz P, Bcker G, Hammes WP, Vogel RF (1995) Z Lebensm Un-
ters Forsch 201:9196
6. Mller MRA, Wolfrum G, Stolz P, Ehrmann MA, Vogel RF
(2001) Food Microbiol 18:217227
123
Eur Food Res Technol (2008) 227:261266
7. Ehrmann MA, Mller MRA, Vogel RF (2003) Int J Syst Evol
Microbiol 53:713
8. Sandhu GS, Kline BC, Stockman L, Roberts GD (1995) J Clin
Microbiol 33:29132919
9. Walter J, Tannock GW, Tilsala-Timisjarvi A, Rodtong S, Loach
DM, Munro K, Alatossava T (2000) Appl Environ Microbiol
66:297303
10. Chung CH, Day DF (2002) J Ind Microbiol Biotechnol 29:196
199
11. Chelo IM, Z-Z L, Tenreiro R (2007) Int J Syst Evol Microbiol
57:276286
12. Van Beek S, Priest FG (2002) Appl Environ Microbiol 68:297
305
13. Dobson CM, Chaban B, Deneer H, Ziola B (2004) Can J Microbiol
50:482488
14. Felis GE, Dellaglio F, Mizzi L, Torriani S (2001) Int J Syst Evol
Microbiol 51:21132117
15. Vsquey A, Molin G, Petterrsson B, Antonsson M, Ahrn S (2005)
Syst Appl Microbiol 25:430441
16. Dellaglio F, Felis GE, Torriani S (2002) Int J Syst Evol Microbiol
52:285287
17. Hancioglu , Karapinar M (1997) Int J Food Microbiol 35:271
274
18. Gnzle MG (2005) In: Brandt MJ, Gnzle MG (eds) Mikrobiolo-
gie des Sauerteiges. Behrs Verlag, Hamburg, pp 77107
19. Hammes WP, Brandt MJ, Francis KL, Rosenheim J, Seitter MFH,
Vogelmann SA (2005) Trends Food Sci Technol 16:411
20. Vogel RF, Knorr R, Mller MRA, Steudel U, Gnzle MG, Ehr-
mann MA (1999) Antonie Van Leeuwenhoek 76:403411
21. Ehrmann MA, Vogel RF (2005) Trends Food Sci Technol 16:31
42
22. Stiles ME, Holzapfel WH (1997) Int J Food Microbiol 36:129
23. Fitzsimons NA, Cogan TM, Condon S, Beresford T (1999) Appl
Environ Microbiol 65:34183426
24. Walter J (2005) In: Tannock WG (ed) Probiotics, Prebiotics Cais-
ter. Academic Press, Norfolk, UK, pp 5182
25. Ouwehand AC, Salminen S, Isolauri E (2002) Antonie Van Leeu-
wenhoek 82:279289
26. Johansson M-L, Molin G, Jeppsson B, Nobaek S, Ahrn S, Beng-
mark S (1993) Appl Environ Microbiol 59:1520
27. Tieking M, Korakli M, Ehrmann MA, Gnzle MG, Vogel RF
(2003) Appl Environ Microbiol 69:945952
28. Orla-Jensen S (1943) The lactic acid bacteria (1943) Det Kongel-
ige Danske Videnskabernes Selskab Biologiske Skrifter Bind II Nr
3 Kopenhagen
29. van Hijum SAFT, Kralj S, Ozimek LK, Dijkhuizen L, van Geel-
Schutten IGH (2006) Microbiol Mol Biol Rev 70:157176
30. Dols M, Simeon MR, Willemot R-M, Vignon MR, Monsan PF
(1998) Carbohydr Res 305:549559
31. Rodrigues S ,Lona LMF, Franco TT (2005) Bioprocess Biolsyst
Eng 28:914
32. Seo E-S, Nam S-H, Kang H-K, Cho J-Y, Lee H-S, Ryu H-W, Kim
D (2006) Enzyme Microb Technol e-pub ahead of print
33. Van Loo J, Cummings J, Delzenne N, Englyst H, Franck A, Hop-
kins M, Kok N, Macfarlane G, Newton D, Quigley M, Roberfroid
M, van Vliet T, van der Heuvel E (2000) Br J Nutr 81:121132
34. Bhm G, Jelinek J, Stahl B, van Laere K, Knol J, Fanaro S, Moro
G, Vigi V (2004) J Clin Gastroenterol 38(6S):S76S79
35. Rycroft CE, Jones MR, Gibson GR, Rastall RA (2001) J Appl
Microbiol 91:878887
36. Vernazza CL, Gibson GR, Rastall RA (2006) J Appl Microbiol
100:846853
37. Kohmoto T, Kukui F, Takaku H, Mitsuoka T (1991) Agric Biol
Chem 55:21572159