doi: 10.1111/j.1471-0307.2009.00464.

x
ORIGINAL
Blackwell
Oxford,
International
IDT
Society
1364-0307
1364-727X
XXX of
UK RESEARCH
Dairy
Publishing
Journal
Technology
of
LtdDairy2009
Technology

O RI G I NAL
RESEARCH Effects of storage temperature and water activity on the
survival of bifidobacteria in powder form
F UMIAKI AB E,* HIROF UM I M IYAUC HI, AYAKO UC HIJ IM A,
TOMOKO YAESHIMA and KE IJ I IWAT S UKI
Food Science and Technology Institute, Morinaga Milk Industry Co., Ltd., Kanagawa Prefecture 228-8583, Japan

Effects of water activity and storage temperature on survival of bifidobacteria in powder form were
investigated and kinetic analyses were performed to reveal characteristics of the stability. A significant
positive correlation was observed between water activity and natural logarithm of the inactivation rate
constant of bifidobacteria powder, indicating that higher water activity induced lower stability of
bifidobacteria in powder form. Also, higher temperature condition induced lower survival rate, which was
supported by that the stability was followed the Arrhenius theory. These findings constructed a prediction
model for bifidobacteria survival in powder form.
Keywords Bifidobacteria, Bifidobacterium longum, Water activity, Probiotics, Stability, Survival.

Zhang 2005). Furthermore, the effects of storage

*Author for correspondence. E-mail: f_abe@morinagamilk.co.jp

I N T RO D U C T I O N
Bifidobacteria, which are well known as probiotic
micro-organisms, have been investigated for their
beneficial effects on human health (Reddy and
Rivenson 1993; Yaeshima et al. 1997; Fujii et al.
2006). Now bifidobacteria are used globally as
probiotics in various food products including yogurt,
milk, infant formula, cereal, cheese and dietary
supplements (Vinderola et al. 2000; Mattila-
Sandholm et al. 2002; Charalampopoulos et al.
2002; Champagne et al. 2005; Philips et al. 2006),
and some bifidobacterial strains including Bifido-
bacterium longum BB536 are used in Japan as
functional ingredients in products designated as
Food for Specified Health Uses (Shortt 1999).
For industrial application, a powdered form of
bifidobacteria is necessary for addition to powder
products such as infant formulas. So far, spray
drying and freeze drying technologies used to
produce bifidobacteria powder have been
investigated (Lian et al. 2002; Saarela et al. 2005;
Simpson et al. 2005).
It is very important that the probiotic products
contain live probiotics at high enough viable counts
at the time of consumption (Saxelin et al. 1999). To
maintain living cells in powder foods for a long
time under ambient temperature conditions is not
easy, because drying damages the membranes,
proteins and nucleic acids (Castro et al. 1997;
Author for
correspondence. E-mail:
Potts 2001). Scientists have tried to develop
f_abe@morinagamilk.co.jp technologies to produce stable bacterial powder,
and many studies have investigated the effects of
© 2009 Society of drying on microbial cell survival (Kets et al. 1996;
Dairy Technology To and Etzel 1997; Carvalho et al. 2003; Zhao and
conditions on survival have been discussed recently
(Hernandez et al. 2007; Higl et al. 2007; Poltner
et al. 2007). However, few reports have investigated
the effects of storage conditions on the survival of
bifidobacteria in powder form (Damjanovic and
Rdulovic 1968; Nagawa et al. 1987; Simpson et al.
2005). For manufacturers of final food products
containing bifidobacteria, it is very important to
know what precautions should be taken during the
manufacturing process and storage. The present
article reports the effects of storage conditions,
both temperature and water activity, on the survival
rate of bifidobacteria, and proposes a prediction
model of product stability based on kinetic analyses
of data obtained from long-term stability tests.
M AT E R I A L S A N D M E T H O D S
Bifidobacteria powder
A commercial probiotic powder, lyophilized
Bifidobacterium longum ssp. longum BB536
(Morinaga Milk Industry Co., Ltd, Tokyo, Japan)
which is deposited into American Type Culture
Collection as strain BAA-999, was used in this
study. The bifidobacteria powder tested in this
study contained 1.2 × 1011 colony-forming units
(cfu)/g of B. longum BB536.
Preparation of test powders with various water
activities and storage tests
To prepare test powders with various water activi-
ties (aw), raw cornstarch (Nippon Syokuhin
Kako Co., Ltd, Tokyo, Japan) and dried cornstarch
(Matsutani Chemical Industry Co., Ltd, Hyogo,

234 Vol 62, No 2 May 2009 International Journal of Dairy Technology

Vol 62, No 2 May 2009
Japan) were used. aw of the raw cornstarch and
dried cornstarch were 0.56 and 0.02, respectively,
when measured by the Rotronic Hygroskop DT
(Rotronic Instrument Corp., NY, USA). The two
types of cornstarch were mixed in various ratios to
prepare seven test powders with diverse aw ranging
from 0.04 to 0.56. Each mixed powder achieved
equilibrium 1 week after mixing (data not shown).
Commercial B. longum BB536 powder with a
viable count of 1.2 × 1011 cfu/g was then added to
each of the test powders at 0.1% (final count in
sample: approximately 1.2 × 108 cfu/g). After mixing
uniformly, 2 g of each bifidobacteria-containing
mixed powder was put into an aluminium bag
(PET/AL/PE, 9 cm wide × 7 cm tall × 70 μm
thick) and hermetically sealed for stability testing.
Aluminium bag samples of each aw were stored at
various temperature conditions: 5, 25, 37, 45 and
60°C. Viable counts of each sample were monitored
continuously during a storage period of 30 months
to observe survival under various conditions of
aw and storage temperature. The above stability
tests were performed twice using different lot of
B. longum powder, and all data were used in
regression analysis to estimate stability.
Enumeration of viable counts
Bifidobacterial counts were enumerated using
pour plate technique. One gram of storage powder
sample was suspended in 99 mL of buffer (4.5 g
potassium dihydrogenphosphate, 6.0 g disodium
hydrogenphosphate, 0.5 g l-cysteine, 0.5 g Tween-
80, 1.0 g agar and 1000 mL distilled water). All
ingredients of the buffer were purchased from
Kanto Chemical Co., Inc. (Tokyo, Japan). After
mixing thoroughly, the suspension was diluted
serially in sterilized saline. Blood Liver (BL) agar
medium (Nissui, Tokyo, Japan) was used as the
plating medium. The solidified agar plates were
incubated at 37°C for 72 h under anaerobic
condition (N2: 80%, H2: 10%, CO2: 10%) in an
anaerobic chamber (Anaerobox; Hirasawa Works,
Co. Ltd, Tokyo, Japan) to obtain the cfu.
Calculation of inactivation rate constant and
Statistical treatment
To obtain the rate of bifidobacteria inactivation
during storage, the regression line was determined
from the plot of the common logarithm of the
residual bifidobacterial count (log10cfu/g) versus
storage period (month) for each temperature
and aw condition. The absolute value of the slope
(regression coefficient) determined from the regres-
sion line was used as the inactivation rate constant
(log10cfu/g/month) at that specific condition only
when the correlation coefficient (R2) of the regres-
sion line was greater than 0.75 (R2 > 0.75).
Regression analysis was conducted to examine
the correlation between two factors, such as residual
© 2009 Society of Dairy Technology
bifidobacterial count vs storage period, inactivation
rate constant vs aw condition, and inactivation rate
constant vs inverse of absolute temperature for
storage. Where regression coefficient (R2) value
is discussed, this refers to the correlation of the
points on the line of best fit, 1 being a strong
positive correlation and 0 being no correlation
between the points.
R E S U LT S
Changes of bifidobacteria counts in the test
powders under various aw and storage temperature
conditions during the storage period are shown in
Figure 1a–e. In each figure, the regression lines were
determined and both the correlation coefficients
and the regression coefficients indicating slopes of
the line were calculated to estimate the bifidobacteria
loss rate under respective conditions. As shown in
Figure 1a, the test powder was very stable under
cool condition except for the highest aw condition
(0.56). As shown in Figure 1b, a significant negative
correlation was observed between the common
logarithm of surviving bifidobacteria count and the
storage period at aw conditions greater than 0.21.
However, no significant decrease was observed in
test powders with aw lower than 0.16 when stored
at 25°C. At both 37 and 45°C, a significant negative
correlation was observed between bifidobacterial
count and storage period in most test powders
except that with aw 0.04 stored at 37°C (Figure 1c
and b). Furthermore, at 60°C, rapid decreases of
bifidobacterial count were observed and the loss
was evident within very short-term even in test
powder with the lowest aw (Figure 1e).
In Figure 1, each regression line fitted the
following equation;
Nt = –kt + N0 (1)
where Nt is the common logarithmic value of
residual bifidobacterial count after a given storage
period (log10cfu/g), k is the absolute value of the
regression coefficient constant of each regression
line, t is the storage period (month), and N0 is the
common logarithm of initial bifidobacterial count
(log10cfu/g).
Because the k in Equation 1 can be read as a
level of the slope of each line in Figure 1, k was
defined as the inactivation rate constant (log10cfu/
g/month) in each storage condition. Table 1 shows
all the inactivation rate constants (k) obtained from
Figure 1, except for some conditions with low
correlation coefficients (R2 ≤ 0.75) or a small
number of samples (n ≤ 5).
When the relationship between aw and inactiva-
tion rate constant (k) obtained from Table 1 was
analysed at each storage temperature, a significant
correlation was observed between the natural
logarithm of k (lnk) and aw during storage for all
235
 Vol 62, No 2 May 2009

Figure 1 Effects of water activity and temperature on

survival of Bifidobacterium longum BB536 in powder form.
Storage temperatures were (a): 5°C, (b): 25°C, (c): 37°C,
(d): 45°C, and (e): 60°C. Water activities were 0.04 (),
0.10 (), 0.16 (), 0.21 (), 0.32 (), 0.40 () and
0.56 ().

storage temperatures, as shown in Figure 2. At all during storage, and C is the constant at each
temperatures, lnk correlated positively and signi- temperature. In Figure 2, the regression equations
ficantly with aw (R2 > 0.95). All regression lines were lnk = 12.21x–6.871 at 25°C, lnk = 13.24x–
fitted the following equation 4.342 at 37°C, lnk = 15.13x–3.051 at 45°C, and
lnk = 18.00x–0.459 at 60°C.
Lnk = k′x + C (2)
From Equations 1 and 2, an equation was derived
where k′ is the regression coefficient for each to estimate the survival rate of this bifidobacterial
temperature condition in Figure 2, x is the aw powder under each aw and temperature condition,
as shown below:
Table 1 Inactivation rate constants of the test bifidobacteria powder under various Log10(nt/n0) = –exp(k′x + C)/t (3)
storage conditions
where nt is the residual bifidobacterial count after
a given storage period (cfu/g), n0 is the initial
Water activity
bifidobacterial count (cfu/g), k′ is the regression
Temperature 0.56 0.40 0.32 0.21 0.16 0.10 0.04 coefficient for each temperature condition in Equa-
60°C ND a
ND ND 26.036 b
11.601 4.285 1.195 tion 2, x is the aw of each powder, C is the constant
45°C ND 21.509 5.361 0.913 0.681 0.176 0.100 for each temperature condition in Equation 2, and
37°C 11.011 4.198 1.294 0.404 0.070 0.046 0.015 t is the storage period (month). Using Equation 3,
25°C 0.856 0.155 0.056 0.012 ND ND ND the theoretical survival rate of a test bifidobacteria
powder in each condition can be calculated. To
a
ND, not determined because of low correlation coefficients (R2 ≤ 0.75) or small
numbers of samples (n ≤ 5).
validate Equation 3, the theoretical survival rates
b
Log10cfu/g/month. calculated from the equation and the actual survival
rates obtained from stability tests were compared

236 © 2009 Society of Dairy Technology

Vol 62, No 2 May 2009
Figure 2 Effects of water activity on the natural logarithmic
value of inactivation rate constant (ln k) at various storage
temperatures. Temperature condition: () 25°C, ( ) 37°C,
() 45°C and () 60°C.
Figure 3 Comparison between theoretical and actual
measurement values of survival rate of the tested powder at
25°C (a) and 37°C (b).
(a) (❍) theoretical value at aw 0.21, () at aw 0.32, ( ) at aw
0.40 () at aw 0.56.
() actual measurement value at aw 0.21, () at aw 0.32,
() at aw 0.40 and () at aw 0.56.
(b) () theoretical value at aw 0.10, () at aw 0.21, () at aw
0.32. () actual measurement value at aw 0.10, () at aw 0.21
and () at aw 0.32.
under various aw conditions as shown in Figure 3a
and b. There were no great differences between the
theoretical and the actual survival rates for various
conditions, confirming that the theoretical survival
rates obtained from Equation 3 can be used as
prediction models of stability for the bifidobacteria
powder. Also, using the Equation 3, the effects of
aw on the survival rate of bifidobacteria at 25°C in
© 2009 Society of Dairy Technology
Figure 4 Prediction model of the survival rate of test
bifidobacterial powder at 25°C.
() Survival rate after 12 months, () after 18 months,
() after 24 months and () after 36 months.
Figure 5 Arrhenius plot for the inactivation of the test
bifidobacteria powder at various water activities (aw).
aw conditions: () 0.40, ( ) 0.32, () 0.21, (❍) 0.16,
() 0.10 and () 0.04.
the test powder were calculated at different storage
periods and are illustrated in Figure 4.
On the other hand, when the relationship between
the inactivation rate constant (k) obtained from
Table 1 and the storage temperature was analysed,
the natural logarithm of the inactivation rate
constant (lnk) showed an inverse dependence on
absolute temperature at all aw conditions, as
shown in Figure 5. These results indicate that the
inactivation rate of the test bifidobacteria powder
follows the Arrhenius theory.
DISCUSSION
Recently, probiotic powders are widely used in
recent food formulations, notably in infant formulas
and milk powders. To maintain a long shelf-life
for powder foods containing probiotics, attention
should be paid to stability of the probiotics under
ambient temperature.
In the present study, the effects of aw and storage
temperature on the stability of bifidobacteria in
powder form during long-term storage were inves-
tigated using a commercial B. longum BB536
powder. From the data of stability tests under various
storage conditions, we observe that stability can be
well described as a first-order kinetic reaction, as
indicated by Equation 1, which supports previous
237
 Vol 62, No 2 May 2009
reports (Damjanovic and Radulovic 1968; To and
Etzel et al. 1997; Andersen et al. 1999; Ziadi et al.
2005).
From Figure 1, the inactivation rate constants (k)
were determined for various storage conditions, as
shown in Table 1. These constants represent the
level of stability under various storage conditions.
Figure 2 shows that the natural logarithm of the
inactivation rate increases in proportion to aw at all
storage temperatures, indicating that aw has very
strong effect on the survival of bifidobacteria in the
test powder. These results suggest that when food
products containing the bifidobacterial powder are
produced, the aw of the food should be considered.
It is easier to maintain the stability of bifidobacteria
in powder products with lower aw. This finding may
be the same as a previous study on B. longum con-
ducted by Nagawa et al. (1987), although they
measured water content, not aw.
Equation 3 induced from both Equations 1 and
2 constructed a prediction model for survival of
bifidobacteria in powder product considering aw
as shown in Figure 4. This prediction model or
equation may be very useful for manufacturers in
producing probiotics products to decide the shelf
life, to expect the survival after storage or to design
the products. Equally, to construct a prediction
model from various stability tests is very helpful to
understand the characteristics of the stability of
probiotics. Also, the survival kinetics of bifidobac-
teria in this powder product followed the Arrhenius
rate law (Figure 5), as was reported in previous
studies (Damjanovic and Radulovic 1968; To and
Etzel 1997; Achour et al. 2001), indicating that the
storage temperature has great effect on the stability.
On the other hand, it was reported that aw was
temperature dependent and aw changed in accord-
ance with temperature changes, although water
content was not varied (Barbosa-Canovas et al.
2007). Because in present study we adjusted aw of
the test powder at 25°C, the real aw in aluminium
bag at each temperature may be different from our
observation. Although we should consider the
differences in cases where the temperature varied
widely from 25°C, because most manufacturers of
infant formula or powder products generally
produce their products and check the aw at room
temperature, our observation may be of benefit to
them as index when these products are kept at dif-
ferent temperature widely. Also, Barbosa-Canovas
et al. (2007) indicated that nonenzymatic or chemical
reactions were accelerated after formation of water
monolayer on surface of substance as shown in a
food stability map developed by them (Barbosa-
Canovas et al. 2007). In the present study, from
over 0.21 aw, cell numbers decreased significantly
at 25°C (Figure 1b). There may be a relationship
between the bacterial inactivation and nonenzy-
matic or chemical reactions indicated on the food
238 © 2009 Society of Dairy Technology
stability map. However, as some scientists indicated,
the bacterial stability in powder form was very
variable and depended on various conditions
(Andersen et al. 1999; Simpson et al. 2005;
Schoug et al. 2006). Further investigations are
required to clear up the reason why bacteria
powders are reduced during storage or the relation-
ship between bacterial inactivation in powder form
and nonenzymatic or chemical reactions.
The fact that both aw and temperature have
considerable impact on the stability is in principle
consistent with other bacterial powders (Higl et al.
2007). We also observed the same tendency in our
preliminary studies using other bifidobacterial
species, although the inactivation rate constants
differed from strain to strain (not data shown).
From these results, the stability of all bacterial
powders may depend on both aw and temperature
conditions. Also, although this study used cornstarch
as a carrier, knowledge of the effects of different
carriers like milk powder will promote further
understanding of the bifidobacterial stability. Further
kinetic analyses of other factors may produce
useful hints for improving long-term stability.
These additional findings will be more useful for
the manufacturing of probiotic products.
CONCLUSIONS
Effects of aw and storage temperature on stability
of B. longum in powder were investigated through
long-term stability test. It was observed that both
higher aw and storage temperature induced higher
inactivation rate constant of bifidobacteria powder.
Also, a significant correlation was observed between
the natural logarithm of inactivation rate constants
of bifidobacteria and aw condition. The kinetic
analyses of the stability data induced an equation
indicating a prediction model of survival of bifido-
bacteria. Furthermore, the stability followed the
Arrhenius theory. These observations and kinetic
analyses will be useful for manufacturers to produce
probiotic products because they can predict survival
of adding bifidobacteria in their products.
A C K N OW L E D G E M E N T S
We thank Mr Hamano of the Nutritional Science
Laboratory in Morinaga Milk Industry Co., Ltd.
for his kind assistance with the statistical analysis.
REFERENCES
Achour M M, Cornelius N, Zgouli C, Mahjoub S, Thonart A P H
and Hamdi M (2001) Application of the accelerated
shelf life testing method (ASLT) to study the survival
rates of freeze-dried Lactococcus starter cultures. Journal
of Chemical Technology and Biotechnology 76 624–
628.
Vol 62, No 2 May 2009
Andersen A B, Fog-Petersen M S, Larsen H and Skibsted L H
(1999) Storage stability of freeze-dried starter cultures
(Streptococcus thermophilus) as related to physical
state of freezing matrix. Lebensmittel-Wissenschaft und
Technologie 35 540–547.
Barbosa-Canovas V G, Fontana J A, Jr., Shmidt S J and
Labuza P T (2007) Water Activity in Foods. Ames: IFT
Press, Blackwell Publishing.
Carvalho A S, Silva J, Ho P, Teixeira P, Malcata F X and
Gibbs P (2003) Effect of addition of sucrose and salt, and
of starvation upon thermotolerance and survival during
storage of freeze-dried Lactobacillus delbrueckii spp.
bulgaricus. Journal of Food Science 68 2538–2541.
Castro H P, Teixeira P M and Kirby R (1997) Evidence of
membrane damage in Lactobacillus bulgaricus following
freeze drying. Journal of Applied Microbiology 82 87–94.
Champagne C P, Gardner N J and Roy D (2005) Challenges
in the addition of probiotic cultures to foods. Critical
Review Food Science and Nutrition 45 61–84.
Charalampopoulos D, Wang R, Pandiella S S and Webb C
(2002) Application of cereals and cereal components in
functional foods: a review. International Journal of Food
Microbiology 15 131–141.
Damjanovic V and Rdulovic D (1968) Predicting the stability
of freeze-dried Lactobacillus bifidus by the accelerated
storage test. Cryobiology 5 101–104.
Fujii T, Ohtsuka Y, Lee T, Kudo T, Shoji H, Sato H, Nagara
S, Shimizu T and Yamashiro Y (2006) Bifidobacterium
breve enhances transforming growth factor beta1 signal-
ing by regulating Smad7 expression in preterm infants.
Journal of Pediatric Gastroenterology and Nutrition 43
83–88.
Hernandez A, Weekers F, Mena J, Pimentel E, Zamora J,
Borroto C and Thonart P (2007) Culture and spray-drying
of Tsukamurella paurometabola, C-924: stability of
formulated powders. Biotechnology Letters 29 1723–
1728.
Higl B, Kurtamann L, Carlsen C U, Ratjen J, Först P, Skibsted
L H, Kulozik U and Rsibo J (2007) Impact of water activ-
ity, temperature, and physical state on the storage stability
of Lactobacillus paracasei ssp. paracasei freeze-dried in
a lactose matrix. Biotechnology Progress 23 784–800.
Kets E P W, Teunissen P J M and Bont J A M (1996) Effect
of compatible solutes on survival of lactic acid bacteria
subjected to drying. Applied and Environmental Micro-
biology 62 259–261.
Lian W C, Hisao H C and Chou C C (2002) Survival of
bifidobacteria after spray-drying. International Journal
of Food Microbiology 4 79–86.
Mattila-Sandholm T, Myllärinen P, Crittenden R, Mogensen G,
Fondén R and Saarela M (2002) Technological challenges
for future probiotic foods. International Dairy Journal 12
173–182.
Nagawa M, Nakabayashi A and Fujino S (1987) Preparation
of the bifidus milk powder. Journal of Dairy Science 71
1777–1782.
Philips M, Kaliasapathy K and Tran L (2006) Viability of
commercial probiotic cultures (L. acidophilus, Bifidobac-
© 2009 Society of Dairy Technology
terium sp., L. casei, L. paracasei and L. rhamnosus) in
cheddar cheese. International Journal of Food Micro-
biology 108 276–280.
Poltner D C, Leuschner R G K and Murry B S (2007)
Optimising the viability during storage of freeze-dried
cell preparations of Campylobacter jejuni. Cryobiology
54 265–270.
Potts M (2001) Desiccation tolerance: a simple process.
Trends in Microbiology 9 553–559.
Reddy B S and Rivenson A (1993) Inhibitory effect of
Bifidobacterium longum on colon, mammary, and liver
carcinogenesis induced by 2-amino-3-methylimidazo
[4,5-f]quinoline, a food mutagen. Cancer Research 53
3914–3918.
Saarela M, Virkajärvi I, Alakomi H L, Mattila-Sandholm T,
Vaari A, Suomalainen T and Mättö J (2005) Influence of
fermentation time, cryoprotectant and neutralization of
cell concentrate on freeze-drying survival, storage stability,
and acid and bile exposure of Bifidobacterium animalis
ssp. lactis cells produced without milk-based ingredients.
Journal of Applied Microbiology 99 1330–1339.
Saxelin M, Grenov B, Svensson U, Fonden R, Reniero R and
Mattila-sandholm T (1999) The technology of probiotics.
Trends in Food Science and Technology 10 387–392.
Schoug Å, Olsson J, Carlfors J, Schnürer J and Håkansson S
(2006) Freeze-drying of Lactobacillus coryniformis Si3 —
effect of sucrose concentration, cell density, and freezing
rate on cell survival and thermophysical properties.
Cyobiology 53 119–127.
Shortt C (1999) The probiotic century: historical and current
perspectives. Trends in Food Science and Technology 10
411–417.
Simpson P J, Stanton C, Fitzgerald G F and Ross R P (2005)
Intrinsic tolerance of Bifidobacterium species to heat and
oxygen and survival following spray drying and storage.
Journal of Applied Microbiology 9 493–501.
To B C S and Etzel M R (1997) Survival of Breveibacterium
linens (ATCC9174) after spray drying, freeze drying, or
freezing. Journal of Food Science 62 167–189.
Vinderola C G, Prosello W, Ghiberto T D and Reinheimer J A
(2000) Viability of probiotic (Bifidobacterium, Lactobacillus
acidophilus and Lactobacillus casei) and nonprobiotic
microflora in Argentinian Fresco cheese. Journal of Dairy
Science 83 1905–1911.
Yaeshima T, Takahashi S, Matsumoto N, Ishibashi N,
Hayasawa H and Iino H (1997) Effect of yogurt containing
Bifidobacterium longum BB536 on the intestinal environ-
ment, fecal characteristics and defecation frequency: a
comparison with standard yogurt. Bioscience and Micro-
flora 16 73–77.
Zhao G and Zhang G (2005) Effect of protective agents,
freezing temperature, rehydration media on viability of
malolactic bacteria subjected to freeze-drying. Journal of
Applied Microbiology 99 333–338.
Ziadi M, Touhami Y, Achour M, Thonart P Hand Hamdi M
(2005) The effect of heat stress on freeze-drying and
conservation of Lactococcus. Biochemical Engineering
Journal 24 141–145.
239