Full text of DOI:10.1111/j.1365-2621.2002.tb08735.x

JFS: Food Microbiology and Safety
Effect of Yeast Extracts Containing
Propionic Acid on Bread Dough
Fermentation and Bread Properties
N. GARDNER, C.P. CHAMPAGNE, AND P. G ÉLINAS
ABSTRACT: Three commercial yeast extracts (Oxoid, Champlain, Lallemand) fermented with immobilized cells
of Propionibacterium freudenreichii were added in broths and bread formulations, and their effect on growth
and gas production by bakers yeast was determined. Appearance and preservation of the breads were examined.
There was significantly more gas produced by the yeast when fermented yeast extracts were added to the dough.
In broths, the initial growth rate was slower with media containing fermented yeast extract when compared to
their unfermented equivalents. Loaves formulated with propionic acid of the fermented yeast extract contained
less ethanol. Paired t-tests showed that breads formulated with fermented yeast extract were protected for a
longer period against mold than those that contained non-fermented yeast extract.
Keywords: yeast extracts, Propionibacterium, bread, leavening, preservation
Materials and Methods
Introduction
EAST EXTRACTS (YE) CONTAIN GROWTH
factors for the production of microor-
tained from Oxoid (Mississauga, Ont.,
Canada), Lallemand (Montreal, Quebec,
Canada and Champlain Industries (Corn-
wall, Ont., Canada). Fermented media
were then analyzed and dried using a Bü-
chi B-191 (Flawil,Switzerland) mini-spray-
drier, as described by Gardner and others
(1999).
Micro-organisms
Y
For dough leavening and bread assays,
compressed bakers’ yeast (Saccharomyces
cerevisiae; Fleischmann’s Yeast Ltd, Toron-
to, Ont., Canada ) was stored at 4 °C and
used within 1 wk of its receipt.
Strains of Penicillium sp. and Rhizopus
sp. were grown on PDA (potato dextrose
agar) (Difco, Detroit, Mich., U.S.A.) at 30 °C
for 14 d, allowing sporulation. A blend of
1% peptone (Difco) and 20% glycerol was
poured over the surface of the agar plate;
this was used to harvest spores in vials
containing 1.0 ml of suspension before
freezing at –70 °C. After 48 h, a vial was
thawed to enumerate the population on
PDA.
ganisms (Peppler 1982) and are found in
commercial media designed for the prepa-
ration of dairy cultures. They are also com-
monly used in soups, gravies, and sauces
to enhance or to simulate flavors such as
meat aroma.
In a previous study (Gardner and oth-
ers 1999), yeast extracts containing propi-
onic acid had also been produced by fer-
Bread making
Dough formulation contained per
400 g of bread-type flour (Robin Hood,
Multifoods, Montreal, Quebec, Canada),
sugar 12 g; salt 8 g; vegetable shortening
12 g; fresh compressed bakers’ yeast 12 g,
gluten 8 g; and according to the experi-
mental design, fermented or unferment-
ed yeast extract 4 g. For some assays, so-
dium propionate (Fisher) was added
(0.2 g of propionate per 100 g of bread).
Dry ingredients were mixed 15 s using a
Hobart mixer (model C-100). Water
(225 g) and 1160 ppm of ascorbic acid
were added to dry ingredients and mixed
for 5 to 7 min. After resting for 30 min,
dough was then divided in two-345 g por-
tions, sheeted-molded and fermented at
35 °C and 82% relative humidity until
dough reached upper surface of the one-l
bread pan. Fermentation times varied as
mentation to be used as
preservative.
a
food
Propionic acid is a preservative exten-
sively used against mold growth on
bread. However, the increasing trend to-
ward more natural foods favors the de-
velopment of products obtained from
fermentation. Javanainen and Linko
(1993) used Propionibacterium in the
lactic acid starter in sour dough bread.
The addition of propionic bacteria to a
bread culture is one approach to provide
propionic acid to the dough, but we opt-
ed for the addition of an ingredient
pre-fermented by Propionibacterium
freudenreichii used in the production of
Swiss-type cheese. These bacteria also
form acetic acid, which might increase
the antifungal properties of yeast ex-
tracts (Playnel 1985).
Production of the propionic-
fermented YE
Three yeast extracts were fermented by
Propionibacterium freudenreichii ssp. sher-
manii as described by Gardner and others
(1999). The yeast extract medium was
composed of (g.L–1): yeast extracts 70; lac-
tate (Fisher, Nepean, Ont., Canada) 20;
glucose 10; CaCl2(BDH, Toronto, Ont.,
Canada) 5. The media were autoclaved at
121 °C, 15 min and fermented at 37 °C for
48 h with an immobilized culture of algi-
nate beads of Propionibacterium freuden-
reichii ssp. shermanii, representing 20% of
the fermentation volume. Two native yeast
extracts from bakers’ yeast and 1 native
yeast extract from brewers’ yeast were ob-
a
function of the ingredients used.
Dough was baked at 210 °C for 12 min. Af-
ter cooling for 60 min, loaf volume was de-
termined by rapeseed displacement (Na-
tional Mfg. Co, Lincoln, Neb., U.S.A.).
Breads were then wrapped and frozen
at –20 °C.
The aim of this study was to evaluate
the effect of the addition of a fermented
yeast extract containing propionic acid on
bakers yeast leavening power as well as
bread characteristic and preservation.
© 2002 Institute of Food Technologists
Vol. 67, Nr. 5, 2001—JOURNAL OF FOOD SCIENCE 1855

Yeast extracts with propionic acid . . .
Table 2—Some properties of the breads formulated with various YE. The control is constituted by a regular formula-
tion without addition of YE. Propionate control is constituted by a formulation without the addition of YE but in which
propionic acid was added.
Loaf
volume
(cc)
1575a
1560a
Surface
openings
(mm2)
1715ab
1501ab
Fraction of
holes within
Gas
Ethanol
(g/100 g
Propionate
(g/100 g
Days
production in
without
Treatment
Control
total surface (%)
Risograph (cc) of bread)
of bread)
apparent mold
30
29
219a
182b
0.49
0.44
0
8.5a
Ն16b
Propionate
Oxoid
0.23a
UYE1
FYE2
1508a
1428ab
28
31
206c
0.38
0.49
0
8.5a
1607a
1724ab
229ad
0.18b
10ac
Lallemand UYE
FYE
1500a
1538a
1344a
28
31
211ac
0.51
0.44
0
6.8a
1703ab
223ad
0.16b
9.0ac
Champlain UYE
FYE
1531a
1885b
33
34
212ac
226ad
0.53
0.47
0
5.5a
1460a
1745b
0.17b
9.0ac
1
2
a, b
UYE = native unfermented yeast extracts
FYE = yeast extracts previously fermented by Propionibacterium.
Numbers followed by the same letter are not significantly different (P Յ 0.05)
=
Chemical composition
Table 1—Concentration (g/100g) of
organic acids in fermented yeast ex-
tracts (FYE)
37 °C. Yeast gas production was followed
during 90 min using a Rhisograph (Rde-
sign, Pullman, Wash., U.S.A.) coupled to a
computer for data acquisition. Three jars
per treatment were prepared and 2 inde-
pendent trials were done.
To analyze the organic acids, ethanol,
and the sugar content of the loaves, 10 g
of bread was homogenized in 100 ml of
H2SO4 0.008N (Baker, Toronto, Ont., Can-
ada), which constituted the mobile
phase of the HPLC analysis. Samples
were then centrifuged at 3000 ´ g for
20 min, then the supernatant was fil-
tered using a 0.45 mm membrane (HVLP
filters, Millipore, Milford, Mass., U.S.A.).
The chromatography procedure de-
scribed by Doyon and others(1991) was
FYE
Acetic
Propionic
Oxoid
Champlain
Lallemand
7.4
8.2
8.9
17.7
19.2
18.0
Effect on growth of the yeast in
broth
Saccharomyces cerevisiae was isolated
from the commercial compressed yeast
(used for bread making). YM broth (Difco)
was inoculated and incubated at 32 °C for
24 h. Every 4 d, 0.1 ml of the culture was
transferred into 10 ml of YM broth and in-
cubated 24 h at 32 °C.
Effect of the addition of YE on
bread appearance
After slicing, bread portions were
scanned and images digitalized using HP
deskscan software. Photoshop 3.0 (Adobe,
Ottawa, Ont., Canada) and Sigmascan
Pro 2.0 (SPPS, Chicago, Ill, U.S.A.) were
used to create the image and to measure
the total surface as well as the mean sur-
face of breadcrumb cells. Values from 2
bread slices per trial were used to calculate
the mean of total surface and mean sur-
face of bread cells.
used. Samples were injected using
a
717 sampler (Waters, Mississauga, Ont.,
Canada) on an Aminex HPX 87H (300
mm ´ 7.8 mm) column from BioRad
(Mississauga, Ont., Canada) maintained
at 40 °C. Flow rate was maintained at
0.4 ml/min. Metabolites were detected
For automated spectrophotometry as-
says, base medium was composed (g.L–1):
10 g glucose, 2 g malt extract, 5 g ferment-
ed or non-fermented yeast extracts. The
pH was adjusted to 5.4 with NaOH and
then the broth was sterilized by filtration
using a micro-porous membrane of 0.22
mm (HVLP filters, Millipore). The medium
was then inoculated with 1% (v/v) of the
yeast pre culture, and 250 mL of inoculated
broth were distributed in a 100 well plate
of an automated spectrophotometer plate
(Bioscreen, Labsystems Corp., Helsinki,
Finland). Plates were incubated at 37 °C
for 24 h and OD was read at 600 nm every
10 min following 20 s of plate agitation
(medium setting). Blanks were constitut-
ed of the un-inoculated medium. Maximal
optical density and maximal growth rate
(mmax) were used as an indication of biom-
ass production. The biomass values in the
spectrophotometric assay system were
transformed into Ln values of biomass
and plotted against time for the exponen-
tial growth part of the curve (Sigma Plot
using
a refractive index monitor 410
(Waters) and data were collected using
the Millennium 2.10 software (Waters).
Determination of the mold-free
storage period
Statistical analysis
Analyses of variance and t-tests for
means comparisons were done using the
GraphPad Instat 3.0 software (San Diego,
Calif., U.S.A.).
Spore suspensions were diluted to ob-
tain 100000 colony forming units for each
mold strain. Two 10 mL droplets were dis-
pensed on the bread surface. Breads were
wrapped in plastic bags so that the bag
did not touch the surface of the loaf. Inoc-
ulated breads were then transferred to a
chamber at 20 °C. Mold growth was moni-
tored daily to determine the time required
Results and Discussion
Mold inhibition of the yeast
extract
Breads containing fermented yeast ex-
tract showed similar mold development to
the control (Table 2). However, paired t-
tests between fermented and non-fer-
mented yeast extracts suggest that breads
formulated with fermented yeast extract
were protected for a longer period against
mold than those that contained unfer-
mented yeast extract. In addition to delay-
ing the appearance of mold colonies, fer-
for the appearance of
a mold colony
(mold-free storage period) and by mea-
suring the diameter of the spreading
mycelium colony.
Leavening power of the yeast
Dough (50 g) (production described software, Jandel). The slope of the curve,
previously) was inserted in glass jars that giving the ␮max, was obtained by linear
were placed in a water bath regulated at regression.
1856 JOURNAL OF FOOD SCIENCE—Vol. 67, Nr. 5, 2002

Yeast extracts with propionic acid . . .
Table 3—Effect of yeast extract on the
growth properties in broths (ODmax
bread pan. After baking, propionic acid comparison between the 6 yeast extracts.
concentration remained above 0.15 g/100 Little growth was observed in the base
,
m
) and gas production in dough
Rmhaixsograph assays of Saccharomyces
g of bread (Table 2). The initial concentra- medium without yeast extract (Figure 1).
tions of acetic and propionic acid in the The initial growth rate was lower with me-
yeast extract preparation were not signifi- dia containing fermented yeast extract
cantly different from each other (Table 1). when compared to their unfermented
Control breads in which propionic acid equivalents (Figure 1; Table 3). However,
was added directly without yeast extract there was a tendency to achieve higher
contained 0.23 g per 100g of bread of pro- ODmax values with the fermented yeast ex-
pionic acid. This concentration slightly ex- tract (Table 3). In the propionic acid-con-
ceeds the maximal allowed concentration taining broths without yeast extracts, no
cerevisiae
Gas
mmax (cc)
Treatment
ODmax
Control
Oxoid
0.200
0.683
0.871
0.18 219
0.32 206
0.28 229
UYE1
FYE2
Lallemand UYE
FYE
0.849
0.934
0.37 211
0.28 223
of 0.20 g per 100 g of bread.
growth occurred inhibiting completely
Champlain UYE
FYE
0.473
0.463
0.29 212
0.25 226
In this series of experiments, the effect yeast metabolism (data not shown).
of extract-based supplements on yeast
1
2
UYE = native unfermented yeast extracts
FYE = yeast extracts previously fermented by
Effect of yeast extract on bread
properties
metabolism was studied. It is recognized
that propionic acid may decrease yeast ac-
tivity (Reed and Nagodawithana 1991)
but it was hoped that the yeast extract
could counterbalance this decrease in
bakery products.
Propionibacterium.
The addition of yeast extracts, native or
fermented, did not alter significantly the
appearance of the bread as the loaf vol-
ume with yeast extract were comparable to
the control (Table 2). The difference be-
tween the surface openings was not signif-
icant, even though there is a tendency to
larger surface openings with fermented
yeast extracts compared to unfermented
yeast extracts from Oxoid and Lallemand.
In the Rhisograph, all the dough formulat-
ed with fermented yeast extracts showed
higher volumes than with the non-fer-
mented yeast extracts indicating a more
active yeast. Also, the gas production was
higher with fermented yeast extracts than
the control that did not contain yeast ex-
tract.
In broths, the use of fermented yeast
extract has a negative effect on mmax, com-
pared to the corresponding unfermented
yeast extract, but the opposite was mostly
the case with ODmax data. It could be hy-
pothesized that the growth-promoting
property of the fermented yeast extract
was initially masked by the presence of
mented yeast extract reduced the spread-
ing rate of the colony. Thus, when the un-
fermented yeast extract was added in the
formulation, bread surfaces were com-
pletely covered with mycelium after 14 d
while those formulated with fermented
yeast extract harbored mold colonies of 2
cm in diameter.
Yeast extracts are often used in micro-
biological media, since they contain
growth-promoting compounds, particular-
ly amino acids and vitamins (Smith and
others 1975). Thus enrichment of bread
with unfermented yeast extract for nutri-
tional or technological purposes could
have the disadvantage of reducing the
preservation of products.
In all Rhisograph trials, there was sig-
nificantly more gas produced by the yeast
when fermented yeast extract was added
to the dough compared to the unferment-
ed yeast extract (Table 3). The source of
the yeast extract did not influence the
leavening capacity of the yeast. Yeast in
dough formulated with unfermented
yeast extract from Lallemand, Oxoid, and
Champlain produced a mean of 208 ml of
gas compared to the fermented yeast ex-
tract that produced 225 ml, whatever the
origin of the extract. Paired t-tests showed
that this small difference (7%) was statisti-
cally significant.
The addition of propionic acid in
dough formulation inhibited the fermen-
tation since 20% less gas was produced
compared to the control.
Effect of yeast extract on yeast
activity
It took between 30 and 45 min for
dough to reach the upper surface of the
Effect of yeast extract on growth of
bakers’ yeast
To evaluate the effect of yeast extract the organic acids. It appears, however,
on the growth of the yeast, various con- that fermentation by Propionibacterium
centrations of non-fermented yeast ex- may improve the biomass yield as well as
tracts were added to a broth medium and the yeast activity shown by the increase in
the optical density (OD) was followed with gas production obtained from yeast ex-
an automated spectrophotometer (Bio- tract.
ScreenJ). In order to favor growth without
Physiology of yeast cells seems affect-
outgrowing the linear portion of the curve ed by prior propionic fermentation of
in optical density, preliminary trials were yeast extract. There was more gas pro-
done with 3 yeast extract concentrations, duced in the baking test with dough when
namely 0.5 g/L, 1.0 g/L and 1.5 g/L, so as to fermented yeast extracts were used for 2
obtain at least 1 treatment with a maxi- yeast extracts among 3. Final biomasses
mum optical density (ODmax) value under were more important in broths containing
1.0. In all cases, the addition of non-fer- fermented yeast extract as shown by high-
mented yeast extract favored the growth er final turbidity for
2 yeast extracts
of Saccharomyces compared to the controls among 3 (Figure 1). However, when the
(data not shown). Since growth curves in fermented yeast extract was present in the
the media containing 1.0 and 1.5 g/L YE broths, a lag in growth as shown by lower
Figure 1—Comparison of various yeast
extracts (fermented and non-ferment-
ed) on the growth of Saccharomyces
cerevisia. Controls are media without
additives
were not significantly different, and since
mmax was observed with Saccharomyces cer-
their ODmax values were often over 1.0, a evisiae. Therefore, results suggest that the
concentration of 0.5 g/L was used for the presence of organic acids may be inhibito-
Vol. 67, Nr. 5, 2002—JOURNAL OF FOOD SCIENCE 1857

Yeast extracts with propionic acid . . .
ry at the beginning of yeast growth; other
Marked differences were obtained on tures or fermented ingredients have been
factors contained in the yeast extracts the growth of the yeast in broths on the proposed to enhance aroma and flavor in
stimulated growth. This was observed for Bioscreen J. Twice as much growth was ob- bread (Gélinas and Lachance 1995). Bread
all yeast extracts used. The 3 yeast extracts tained with media supplemented with fer- with cheese flavor could be produced, as
used, fermented or not, favored yeast mented yeast extracts from Lallemand propionic acid is one of the main flavor
growth in the broths as shown by the high- than with fermented yeast extract from constituent of Swiss type cheese. More-
er turbidity and growth rate compared to Champlain. Fermentation of this extract over, propionibacteria are able to produce
control. However, growth response to did not increase final biomass and even vitamin B12 and could therefore increase
these factors does not seem to relate to the slowed down the growth of Saccharomyces. the nutritive value of the bread (Cavin
gas production in solid fermentation. The Yeast extract from Champlain did not be- and others 1985).
reasons for these observations between have the same way as the others. The fact
References
leavening and broth-related data were not that it originates from a brewers’ yeast
determined. In one case, the effect of could be responsible for this observation.
yeast extract was evaluated on fermenta- The hop resins are present as beer solids
tion activity while it was evaluated on and this yeast cream undergoes debitter-
yeast growth in the other case. During ing steps that may not have eliminated all
leavening, there is little growth, as the the compounds. Moreover, previous stud-
time was insufficient to allow for more ies showed lower nitrogen content in some
than 1 or 2 generations to occur. During native brewers’ yeast extract than in bak-
this time, in the dough, nutrients from the ers’ yeast extract (Gaudreau and others
flour were probably not limiting as in the 1999). Autolysis conditions generate dif-
broths experiments. In the broth experi- ferent native yeast extracts with specific
ments, growth was monitored over a 24 h properties that affect microbial growth
period as it only was for 90 min in the (Champagne and others 1999). Bakers’
dough experiments. Yeast extracts certain- yeast extracts have the disadvantage of
ly influence yeast activity, but they could being more expensive. It would be of in-
also affect dough texture as well. There- terest to examine brewers’ yeast extracts
fore, the method of automated spectro- from other sources.
Bibek R, Daeschel M. 1985. Food biopreservatives of
microbial origin. Vol. 3. In: Comprehensive Bio-
technology, Blanch HW, Dew S, Wang DIC, editors.
New York: Permagon Press.
Cavin JF, Saint C, Divies C. 1985. Continuous produc-
tion of Emmental cheese flavours and propionic acid
starters by immobilized cells of a propionic acid
bacterium. Biotechnology Letters 7(11):821-826.
Champagne CP, Barrette J, Goulet J. 1999. Interac-
tion between autolysis and bacterial contamina-
tion on the production of yeast extracts. Food Re-
search Intl. 32:575-583.
Doyon G, Gaudreau G, St-Gelais D, Beaulieu Y, Ran-
dail CJ. 1991. Simultaneous HPLC determination
of organic acids, sugars and alcohols. Can. Inst.
Food Sci. Technol. J. 24(2):87-94.
Gardner NJ, Champagne CP, Déry C. 1999. Produc-
tion d’un extrait de levure riche en acide propion-
ique. Sci. des Aliments. 19(5):579-592.
Gaudreau H, Champagne CP, Conway J, Degré R. 1999.
Effect of ultrafiltration of yeast extracts on their
ability to promote lactic acid bacteria growth. Can.
J. Microbiol. 45 (11):891-897.
Gélinas P, Lachance O. 1995. Development of fer-
mented dairy ingredients as flavor enhancers for
bread. Cereal Chem. 70:17-21.
Javanainen P, Linko YY. 1993. Factors affecting rye
sour dough fermentation with mixed-culture pre-
ferment of lactic and propionic acid bacteria. J.
Cereal Sci. 18:171-185.
Lai CS, Davis AB , Hoseney RC. 1984. Isolation of a
fermentation stimulant from yeast protein con-
centrate. Cereal Chem. 61(5):428-431.
McKinnon CM, Gélinas P, Simard RE. 1996. Wine yeast
preferment for enhancing bread aroma and fla-
vor. Cereal Chem. 73 (1):45-50.
Peppler HJ. 1982. Yeasts extracts. In: AH Rose, editor. Fer-
mented Foods. London: Academic Press. p. 511-516
Ranum P. 1999. Encapsulated mold inhibitors - The
greatest thing since slice bread? Cereal Foods
World 44(5):370-371.
Reed G. 1972. Yeast food. Baker’s Dig. 46(6):16-18.
Reed G, Nagodawithana TW.1991. Bakers’ yeast pro-
duction. In Reinhold: Yeast Technology, New York:
Van Nostrand Reinhold. p 261-369.
Suomalainen TH, Mayra-Makinen AM. 1999. Propi-
onic acid bacteria as protective cultures in fer-
mented milks and breads. Lait 79:165-174.
Smith JS, Hillier AJ, Lees GJ. 1975. The nature of the
stimulation of the growth of Streptococcus lactis by
yeast extract. J. Dairy Research 42:123-138.
MS20010375 Submitted 7/21/01, Accepted 11/27/01,
Received 11/27/01
photometry used to evaluate the effect of
Growth curves in broths also point to
an ingredient on yeast metabolism should differences in ODmax and maximum
be used only for screening purposes. The growth rate (mmax). It is logical to assume
final outcome of the product depends that acetate and propionate in fermented
more on yeast activity and the way the in- yeast extracts slowed the growth of
gredient will affect the dough during leav- S. cerevisiae. However, fermented yeast ex-
ening and baking. Thus, the effect of tracts seem to contain higher levels of es-
yeast extract on dough volume could be as sential growth factors.
much an effect of yeast metabolism as
This study suggests that yeast extracts
that of dough chemistry. When the loaves previously fermented by Propionibacteri-
were baked, there was no significant dif- um would be better supplements than
ference in final volume as there was in the non-fermented yeast extracts in bread
Rhisograph experiments. Since there was making, with respect to leavening proper-
no difference in the organic acid composi- ties as well as in preservation. Although
tion of the yeast extract added, it could be propionic acid bacteria are now recognized
assumed that yeast extract itself may have for their preservation role in foods, pro-
influenced the gas retention of the dough cesses optimization are needed. The com-
during baking. The fact that the addition bination of Lactobacillus rhamnosus and
of yeast extract does not significantly af- Propionibacterium shermanii in bakery
fect bread properties constitutes an ad- products was proposed as it is more pro-
vantage to the baker.
tective against yeasts and molds than the
We acknowledge the technical and scientific support of
Carole McKinnon, Diane Montpetit, Julie Fontaine as well
as the scientific support of Claude Déry, Carole Beaulieu and
Riszard Brezensky. We also thank Byong Lee for providing
some yeast extracts.
Lai and others (1984) added a protein culture alone (Suomalainen and Mayra-
concentrate from bakers’ yeast to a bread Makinen 1999).
dough and obtained an increase of 20 to
Conclusions
HE TREND TOWARDS SHORTER BREAD
25% in gas production which was not the
case in the present study. Time required
to leaven dough to a specific volume
would be a better indicator of yeast activi-
ty than total volume reached.
Authors are with the Food Research and Develop-
ment Center, Agriculture and Agri-Food Canada,
3600 Casavant Blvd.West, Saint.Hyacinthe (Que-
bec) J2S 8E3, Canada. Direct inquiries to author
Gardner (E-Mail:gardnern@em.agr.ca).
T
making processes can limit the devel-
opment of bread flavor and aroma
(McKinnon and others 1996). Bacterial cul-
1858 JOURNAL OF FOOD SCIENCE—Vol. 67, Nr. 5, 2002