Journal of Applied Microbiology 2001, 91, 1059±1066

Evaluation of microencapsulation of a Bi®dobacterium

strain with starch as an approach to prolonging
viability during storage

K. O'Riordan*, D. Andrews1, K. Buckle2 and P. Conway

School of Microbiology and Immunology and 2Department of Food Science and Technology, The University of
New South Wales, Sydney and 1Starch Australasia, Lane Cove, NSW, Australia

805/04/01: received 30 March 2001, revised 4 April 2001 and accepted 6 June 2001

K . O ' R I O R D A N , D . A N D R E W S , K . B U C K L E A N D P . C O N W A Y . 2001.
Aims: To optimize a spray coating process for the production of encapsulated microspheres
containing viable Bi®dobacterium cells and to determine whether the readily gelatinized
modi®ed starch coating used in this study improved bacterial survival in foods or under acid
conditions.
Methods and Results: An air inlet temperature of 100°C was demonstrated to be optimal for
the spray drying process, as it afforded good drying, low outlet temperatures (45°C) and
resulted in less than 1 log reduction in bi®dobacteria numbers during drying. Maximum
recovery yields of 30% were obtained after optimizing the air aspiration conditions. The
average size of the Bi®dobacterium PL1-containing starch microparticles was determined by
scanning electron microscopy to be of the order of 5 lm. The starch-coated cells did not
display any enhanced viability compared with free PL1 cells when exposed to acid conditions
for 6 h or in two dry food preparations over 20 d storage at ambient temperature (19±24°C).
Determination of 1491 nucleotides of the 16S rRNA gene from PL1 indicated that it shared
97% homology with a previously sequenced Bi®dobacterium ruminantium strain.
Conclusions: Our data demonstrated that, although spray drying is a valuable process for
encapsulating bi®dobacteria, further work is required to ascertain a more appropriate coating
material that will protect this strain against adverse environmental conditions.
Signi®cance and Impact of the Study: The production of small, uniformly coated
microspheres containing viable bi®dobacteria using an affordable and industrially convenient
process, such as spray drying, has commercial implications for the production of probiotic
products. Although popular for use as a coating polymer by the food industry, this study
indicated that modi®ed starches might not be suitable for use as an encapsulating material for
probiotic strains.

bene®t the health of the consumer' (Tannock et al. 2000)

INTRODUCTION
Probiotics have recently been de®ned as `live microbes
which transit the gastrointestinal tract and in doing so
Correspondence to: P. Conway, School of Microbiology and Immunology, The
University of New South Wales, Sydney 2052, Australia
(e-mail: P.Conway@unsw.edu.au).
*Present address: Channing Laboratory, Department of Medicine, Brigham and
Women's Hospital and Harvard Medical School, 181 Longwood Avenue, MA
02115, USA.
ã 2001 The Society for Applied Microbiology
differing somewhat from the earlier de®nitions which
focused on probiotic interactions with indigenous intestinal
microbes (Fuller 1989). Indeed, the functionality of these
bacteria in health promotion and disease prevention is
continuing to gain acceptance from health professionals,
consumers and the food industry. Probiotics have been
reported to play therapeutic roles by modulating immunity,
lowering cholesterol, improving lactose tolerance and pre-
venting cancer (reviewed by Kailasapathy and Chin 2000);
however, some controversy continues to exist regarding the
1060 K . O ' R I O R D A N ET AL.
probiotic rationale with concerns that there is insuf®cient
evidence to support usage of some of these strains in humans
(Reid 1999).
The viable microbial content and general quality of many
probiotic-containing products have often been questionable.
Analysis of products in several different countries has
con®rmed that probiotic strains exhibit poor survival in
traditional probiotic foods such as yoghurt and fermented
milks (Shah 2000). Microbial behaviour (growth, survival
and death) in foods is largely governed by properties of the
food (water availability, pH and buffering capacity) in
addition to the storage conditions (temperature, relative
humidity and atmosphere) (McMeekin et al. 1997). Several
approaches have been adopted in endeavouring to improve
probiotic survival during storage of dried starter cultures
and in foods. Workers have increased the buffering capacity
of products such as yoghurts with some positive bene®ts to
bacterial survival (Kailasapathy and Supriadi 1996). The
addition of protective agents such as high amylose maize
starch into the food has also been observed to enhance
probiotic survival at low pH (Wang et al. 1999) and in a
number of products (O'Riordan et al. 2000). For many
years researchers focused on selecting for acid- and bile-
resistant strains, although more recent reports suggest that
it may be possible to induce adaptive stress protective
mechanisms in bacteria that promote viability (Selmer-
Olsen et al. 1999; Shah 2000; K. O'Riordan and P. Conway,
unpublished). In addition, factors such as oxygen permea-
tion, reported to accelerate probiotic death, were success-
fully reduced by the use of glass rather than plastic storage
bottles (Shah 2000). Providing living cells with a physical
barrier against the external environment is an approach
currently receiving considerable interest. This study sought
to investigate whether encapsulating a potential probiotic
Bi®dobacterium strain within a starch coat would improve
viability of the cells during storage and in dry food
products.
M A T E R I A LS A N D M E T H O D S
Bacterial strains and culture conditions
Bi®dobacterium PL1, previously isolated from human faecal
material, was used in this study. This strain was routinely
transferred in Tryptone-Phytone-Yeast Extract (TPY) broth
(Scardovi 1986) under 37°C anaerobic conditions (MK3
anaerobic workstation; Don Whitley Scienti®c, Kensington
Gardens, SA, Australia; maintained at 10% CO2, 10% H2 and
80% N2). For spray drying purposes, Bi®dobacterium strain
PL1 was cultured in 10 l TPY broth for 16 h, after which the
cells were harvested by centrifugation (6000 g, 10 min),
washed in sterile deionized water and resuspended in one
hundredth of the original volume with sterile deionized water.
ã 2001 The Society for Applied Microbiology, Journal of Applied Microbiology, 91, 1059±1066
Optimizing spray drying conditions (inlet
temperatures) for Bi®dobacterium PL1
Cultures prepared for spray drying were heated to 15°C and
maintained with continuous mixing using a magnet mixer
(motor speed 285 min±1) with a digital temperature probe
(IKAMAG; Crown Scienti®c, Sydney, Australia). Samples
were then dried with a laboratory spray dryer (model 190;
BuÈchi, Flawil, Switzerland) at a constant feed rate
(5 ml min±1), spray ¯ow of 700 normal litre (Nl) h±1 (Nl
is the volume of 1 l at 1 bar), under varying aspirator
settings (0±15) and air inlet conditions (60±140°C). The
resultant outlet air temperatures were recorded from the
digital outlet temperature display on the spray dryer. In
order to limit contamination of the spray-dried samples, the
compressed air supply (5±8 bar) to the spray dryer was ®tted
with a MICROalescerÒ ®lter (Wilkerson, Mettmann, Ger-
many) consisting of two sub®lters of 5á0 and 0á3 lm,
respectively and an active charcoal ®lter of 0á01 lm.
Determination of viability of Bi®dobacterium
PL1 in spray-dried powders
The viability of the Bi®dobacterium strain, before spraying
and in the spray-dried powders, was assessed by examining
TPY or Propionic Acid Medium [PAM; Reinforced
Clostridial Agar (Oxoid, Basingstoke, UK) containing
5 ml l±1 propionic acid (Sigma, Sydney, Australia) adjusted
to pH 5á0 (Beerens 1991)] agar after 3 d of anaerobic
incubation at 37°C. A volume (9á9 ml) of half-strength
Wilkinson-Chalgren (WC) anaerobe broth (Oxoid) was
added to 0á1 g spray-dried powder and the preparation
allowed to rehydrate before further dilutions were per-
formed and appropriate dilutions plated (0á1 ml) onto PAM.
The number of bacteria before drying was determined
(cfu g±1 dry weight was determined from cfu ml±1 after
completely drying a known volume of the original sample
used for plating) and compared with the levels of surviving
bacteria per gram spray-dried powder.
Spray coating of Bi®dobacterium
PL1 with starch
A modi®ed (octenyl succinated) waxy maize starch that
gelatinizes below 100°C was kindly provided by Starch
Australasia (Lane Cove, Sydney, Australia). A starch solu-
tion (10% w/v) was prepared in water and solubilized by
heating to 100°C. This starch solution was equilibrated to
20°C and then mixed with the required volume of bacterial
cells suspended in water. The amount of cells added was
varied to assess the in¯uence of different ratios of starch to
cells on the success of encapsulation. The solution was mixed
at 20°C for 30 min before commencement of spray drying
 ENCAPSULATION OF BIFIDOBACTERIA WITH STARCH
using a constant feed rate (5 ml min±1), various aspirator
settings from 0 to 15, spray ¯ow of 700 Nl h±1 and an inlet
temperature of 100°C. In order to determine the percentage
yield, the collecting ¯ask of the spray dryer was weighed
before and after spraying and the following equation used:
Y% ˆ [Ws ´ 100]/Wpc
where Ws is the weight of the sprayed sample in the ¯ask (g)
and Wpc corresponds to the dry weight of the polymer
dispersion (g) in addition to the dry weight of the cells (g). The
starch-coated cells were transferred to sterile bottles with
screw-on lids and stored as required for viability assessment.
Fluctuations in ambient temperature during storage were
recorded using a minimum±maximum mercury thermometer.
Electron microscopy of spray-dried powder
Spray-dried powders were attached to brass stubbs and
chromium coated (Xenosput 2000 chromium coater; Xeno-
sput, Boc Edwards, West Sussex, UK) using deposition
parameters of 0á08 sputter amps for 80 s. Coated preparations
were visualized with a scanning electron microscope (S900;
Hitachi, Tokyo, Japan) using an accelerating voltage of 2 kV.
Acid tolerance assay
Samples of spray-dried Bi®dobacterium PL1 cells and
spray-dried, starch-coated Bi®dobacterium PL1 cells were
comparatively analysed for their tolerance to detrimental
environmental pH conditions. Measured amounts of the dried
powders (approximately 300 mg) were resuspended in
0á1 mol phosphate-buffered saline, pH 7á0, or in 0á1 mol
HCl-glycine buffers, pH 4á5 and 2á8, and incubated at 37°C.
Aliquots (1 ml) were removed after 0, 3 and 6 h, 10-fold serial
dilutions performed in half-strength WC broth and appro-
priate dilutions plated on PAM agar. Following 72 h of
anaerobic incubation, colonies on plates were enumerated.
Survival of starch-encapsulated cells in foods
Starch-coated Bi®dobacterium PL1 cells (approximately 2 g)
were mixed in 100 ml sterile DURAN (Schott Garsco,
Sydney, NSW, Australia) bottles with dry malted beverage
powder (20 g) or commercial muesli preparations (20 g)
which were obtained from local supermarkets. Following
thorough mixing, samples (2 g) were removed, 10-fold serial
dilutions performed in half-strength WC broth, plated on
PAM agar plates and incubated anaerobically as previously
described. Viability was assessed at regular intervals during
storage at ambient temperature for up to 20 d. Three
independent food trials were performed. Fluctuations in
ambient temperature during storage were recorded using a
minimum±maximum mercury thermometer.
ã 2001 The Society for Applied Microbiology, Journal of Applied Microbiology, 91, 1059±1066
1061
16S rRNA gene sequence determination
An aliquot (10 ll) of crude PL1 cell lysate was used as
template in polymerase chain reaction (PCR) ampli®cations
employing 25 pmol of each of the universal eubacterial
primers 27f and 1492r (Lane 1991). The PCR reactions were
prepared in a total volume of 50 ll using 5 ll 10´ buffer, 5 ll
2 mmol deoxyribonucleotide mix, 1á4 ll 25 mmol magnes-
ium chloride and 1 U Taq polymerase (all reagents supplied
by Biotech International, Brisbane, QLD, Australia). Ampli-
®cation of DNA was performed in an express thermal cycler
(Hybaid, London, UK) programmed for an initial denatur-
ation of 95°C for 5 min followed by 25 cycles of
(95°C ´ 30 s) + (50°C ´ 30 s) + (72°C ´ 30 s) and a ®nal
extension of 72°C for 7 min. The PCR products obtained
were puri®ed using the QIAquick kit (Qiagen, Valencia, CA,
USA). The ampli®ed product (1491 bp) was sequenced in its
entirety on both strands, using other universal primers when
necessary and checking ambiguous nucleotides until a de®n-
itive sequence was obtained.
DNA sequence analysis
DNA sequencing reactions were performed using an ABI
Prism Dye Terminator Cycle Sequencing Ready Reaction Kit
with Amplitaq DNA polymerase (Perkin Elmer, Boston, MA,
USA). Sequence determination was performed using an
automated DNA sequencer (377; Applied Biosystems, Foster
City, CA, USA) employing the synthetic oligonucleotides
described above (Life Technologies, Sydney, NSW,
Australia). Sequences were assembled using the AutoAssem-
bler package (Applied Biosystems). Database searches
(Altschul et al. 1990) were performed using the on-line server
maintained at the National Centre for Biotechnology Infor-
mation (Bethesda, MD, USA; (http://www.ncbi.nlm.nih.
gov/recipon/blast-search.html).
Nucleotide sequence accession number
The GenBank accession number for the 1491 bp of the 16S
rRNA gene from Bi®dobacterium sp. PL1 determined in this
study is AF306789.
RESULTS
Optimizing spray drying conditions
for Bi®dobacterium PL1
Bi®dobacterium cells were spray dried at various inlet
temperatures at a ®xed feed rate and the resultant outlet
temperatures recorded. At an inlet temperature of 60°C
the sample did not dry completely and moist sample was
observed to collect in both the cyclone and collecting
1062 K . O ' R I O R D A N ET AL.

¯ask. At an inlet temperature of 80°C, the sample did not when stored in sterile plastic containers with screw-on lids at
dry completely in the cyclone but usually only dry sample ambient temperatures (19±24°C). Results from triplicate
collected in the ¯ask. Neither of these two inlet temper- experiments indicated that numbers of free cells declined by
atures was considered in further experiments. An inlet 4 log, encapsulated cells (ratio of 5 : 1) by 3 log and
temperature of 100°C resulted in an outlet temperature of encapsulated cells (ratio of 10 : 1) by 2 log under these
45°C. Inlet temperatures of 120°C or greater were found storage conditions.
to yield high outlet temperatures (> 60°C). In order to
assess the effect of high outlet temperature on microbial
Scanning electron microscopy
survival, dried bi®dobacteria were exposed to a tempera-
of spray-dried powder
ture of 60°C (dry heat oven) for 20 min, over which time
a 3 log decrease in numbers was recorded. Bi®dobacterium Modi®ed starch microparticles demonstrated a `donut' effect
viability was evaluated prior to and following spraying when visualized under scanning electron microscopy (SEM)
using an inlet temperature of 100°C and it was observed (Fig. 1). The sizes of the starch-encapsulated microparticles
that the Bi®dobacterium population declined by less than 1 varied but were approximately 5 lm on average. As a visual
log (average log decrease 0á78 ‹ 0á1, n ˆ 6 assays). For control, free cells were observed after spray drying to be
this reason, an inlet temperature of 100°C was used in all clumped but bacterial cells could still be individualized
further experiments. The effect of air aspiration on the (Fig. 2). Cells encapsulated at ratios of starch to core of 5 : 1
production yield was also investigated. For this purpose, or 10 : 1 appeared similar to each other. Free cells could not be
microparticles were only collected from the collecting ¯ask
although it was noted that the cyclone was also usually
covered with a thick layer of microspheres. A range of
aspirator rates was used; spraying was otherwise per-
formed maintaining an inlet temperature of 100°C and a
feed rate of 5 ml min±1 and utilizing free cells or a
mixture of starch and cells as the drying material. A
sample of the yields obtained is given in Table 1.
Aspirator settings of higher than 12 did not result in
higher yields (data not shown) so, in further experiments,
an aspirator setting of 12 was used.

Spray coating of Bi®dobacterium

PL1 with starch
Cells were encapsulated with starch with different ratios
(w/w) of starch coat to core material (cells). Ratios of 5 : 1
(starch : cells) or 10 : 1 were evaluated for survival over 5 d

Table 1 Effect of aspirator setting and polymer : core ratio on the

production yield (%) of microparticles in the collecting ¯ask (inlet
temperature 100 °C)

Aspirator Yield (%)* No. of assays

Sample setting (average ‹ S.D.) (n = 2 per assay)

Free cells 0 11á26 ‹ 9á9 4

Free cells 12 29á62 ‹ 12á2 3
Starch-encapsulated 12 26á84 ‹ 4á6 6
cells (5 : 1)
Starch-encapsulated 12 30á18 ‹ 3á1 6
cells (10 : 1)

*Y% ˆ [sprayed sample in ¯ask (g) ´ 100]/[initial, dry weight of

starch (g) + dry weight of cells (g)]. Results are expressed as the Fig. 1 Scanning electron micrograph of Bi®dobacterium PL1-con-
average percentage with S.D. values. taining spray-dried starch microparticle. Bar ˆ 3 lm

ã 2001 The Society for Applied Microbiology, Journal of Applied Microbiology, 91, 1059±1066
 ENCAPSULATION OF BIFIDOBACTERIA WITH STARCH
Fig. 2 Scanning electron micrograph of Bi®dobacterium PL1 cells after
the spray-drying process. Bar ˆ 3 lm
visualized in either sample (n ³ 10 ®elds). The `donut' effect
in Fig. 1 appeared to be an inherent property of this starch
when spray dried. As may be observed in Fig. 3, although the
starch appeared as granules (approximately 10 lm) before
gelatinization (Fig. 3a), after spraying the starch appeared
similar to the spray-coated cell samples (Fig. 3b).
Acid tolerance assay
Free cells and cells encapsulated with starch at ratios of
coating polymer : core of 5 : 1 or 10 : 1 were evaluated for
their ability to withstand buffered environments of pH 7á0,
4á5 and 2á8 for as long as 6 h at 37°C. The results, as
presented in Table 2, indicated that encapsulation offered
no enhanced protection under the acid conditions tested.
Survival of starch-encapsulated cells in foods
Starch-encapsulated cells were incorporated into muesli mix
and their viability over 14 d when stored at ambient
temperature in sealed glass bottles was compared with free
cells that had also been spray dried. Levels of free and
encapsulated cells declined by 2 log counts over 5 d and no
counts were detectable when assayed after 14 d. Free and
encapsulated cells were also included into a dry malted
beverage powder and their survival over 20 d monitored. No
improved survival was observed for encapsulated cells
(Fig. 4).
Determination of the 16S rRNA gene sequence
from Bi®dobacterium PL1
Primers 27f and 1492r were used to amplify a 1491
nucleotide region of the 16S gene. This fragment, when
sequenced on both strands, displayed 97% similarity, over
the available sequence, to Bi®dobacterium ruminantium
(strain JCM8222 (Japanese Collection of Micro-organisms);
ã 2001 The Society for Applied Microbiology, Journal of Applied Microbiology, 91, 1059±1066
1063
accession no. D86197) and 96% identity, over the 1491
nucleotides, to a B. dentium strain (accession no. D86183).
DISCUSSION
The aim of this study was to devise a cost-effective method
of producing microspheres containing viable bi®dobacterial
cells and to investigate whether the starch material used as
the coating polymer afforded the probiotic strain enhanced
survival in adverse environmental conditions. Encapsula-
tion, entrapment or immobilization are terms used almost
interchangeably in the food industry to refer to the provision
of an outer protective coat or layer to protect the core
material from damage. The application of these methodol-
ogies to improving probiotic survival in foods and during
gastrointestinal transit is relatively new. To date, most
published reports have focused on the utilization of
coacervation methods to coat probiotic strains with calcium
alginate and have documented various degrees of success.
Encapsulation in calcium alginate improved the survival
of Lactobacillus in ice milk (Sheu et al. 1993) and of
Bi®dobacterium in mayonnaise (Khalil and Mansour 1998)
but other studies reported that immobilization of bi®do-
bacteria in alginate offered no protection in Cresenza cheese
compared with free cells (Gobbetti et al. 1998) and a recent
report has demonstrated that survival of calcium alginate-
entrapped bi®dobacteria was dependent on several factors
including alginate concentration and bacterial species (Lee
and Heo 2000). Moreover, the scale-up of coacervation-type
processes for commercialization purposes can be problem-
atic. Conversely, the process of spray drying is economical,
easily scaled-up and uses equipment readily available in the
food industry where it is routinely used to produce good
quality microcapsules containing ¯avours, arti®cial sweet-
eners or a wide variety of other food ingredients (Gibbs
et al. 1999). Coating in a ¯uidized bed was initially
considered in this study due to the low temperatures that
may be used in this process. Preliminary work, however,
demonstrated that the small particles produced did not
behave in a ¯uid-like manner, a problem often associated
with ¯uidized systems, which resulted in an excessive
amount of carry-over.
The spray coating procedure was successfully optimized
for the production of starch-coated microspheres containing
viable Bi®dobacterium PL1 cells. Initial work focused on
optimizing spraying conditions for the strain used in this
study. Previous reports indicated that survival of probiotics
or lactic acid bacteria during spray drying decreased with
increasing inlet temperatures (Mauriello et al. 1999). It was
observed that inlet temperatures of 60 or 80°C resulted in
poor drying and moist sample often collected in the cyclone
and sometimes in the collecting ¯ask. Inlet temperatures of
greater than 120°C resulted in high outlet temperatures
1064 K . O ' R I O R D A N ET AL.

Fig. 3 Scanning electron micrographs of starch granules (a) prior to gelatinization and spray drying and (b) after the spray drying process.
Bar ˆ 6 lm

Table 2 The in¯uence of starch coating on the survival of Bi®dobacterium PL1 in buffers at pH 7á0, 4á5 and 2á8

pH 7á0 pH 4á5 pH 2á8

Starch- Starch- Starch- Starch- Starch- Starch-

Incubation encapsulated encapsulated encapsulated encapsulated encapsulated encapsulated
time (h) Cells cells* (5 : 1) cells (10 : 1) Cells cells (5 : 1) cells (10 : 1) Cells cells (5 : 1) cells (10 : 1)

0 9á26 8á54 8á07 9á45 8á35 7á87 9á21 6á63 6á27

3 8á39 7á59 6á97 3á80 3á32 ND ND ND ND
6 7á79 7á39 5á83 ND ND ND ND ND ND

Results are expressed as the mean log cfu g)1 from duplicate experiments (n = 2 per assay).
ND, No counts detectable (detection limit 103 cfu g)1).
*Ratios of starch coat to cell core of 5 : 1 and 10 : 1 assessed in the study.

(> 60°C) and an unacceptable decrease in viable numbers, as
assessment of viability of PL1 at 60°C indicated a 3-log
decrease over a 20-min period. Other probiotic bacteria have
similarly been reported to exhibit considerable strain-
speci®c lack of viability at such temperatures (Gardiner
et al. 2000). Thus, 100°C was chosen to be the optimal inlet
temperature as it afforded good drying and the resultant
outlet temperatures (mean temperature of 45°C) permitted
good bacterial survival. The product yield recovered from
the collecting ¯ask of the spray dryer was observed to be
ã 2001 The Society for Applied Microbiology, Journal of Applied Microbiology, 91, 1059±1066
approximately 30% of the original sample (total dry weight
of starch coat and cells) after optimizing aspirator param-
eters. Although lower than the average yield for the mini
spray dryer used in this study (approximately 45% yield
when spray drying skim milk powder), it is considerably
higher than the 4% yield recovery of starch microspheres
containing rhizobacteria reported by Amiet-Charpentier
et al. (1998).
Scanning electron microscopy indicated consistent and
uniform coverage of cells and a good average microsphere
 ENCAPSULATION OF BIFIDOBACTERIA WITH STARCH
0 2 4 6 8 10 12 14 16 18 20 22
Fig. 4 Viable numbers (cfu g±1) of free (s) and encapsulated
Bi®dobacterium PL1 cells with varying coat : core ratios of 5 : 1 (d)
and 10 : 1 (j) in a dry malted beverage powder over 20 d at ambient
storage temperatures (19±24 °C)
size (< 10 lm), as particles of this size should not affect the
mouthfeel properties of most foods. The ¯attened appear-
ance of starch microspheres (Fig. 1) has been previously
observed (Amiet-Charpentier et al. 1998). From SEM
analysis (Fig. 3) of spray-dried starch samples this effect
appeared to be an inherent characteristic of the starch itself.
The study also sought to investigate whether modi®ed
starch used as the coating polymer would adequately protect
the cells in acid conditions or in food. The strain used,
Bi®dobacterium sp. PL1, has been previously demonstrated
to exhibit bene®cial traits but did not survive well in several
food products examined (P. Conway, unpublished). Starch
was chosen as the coating material in this study as starches
modi®ed by the addition of octenyl succinate have been
reported to exhibit many good spray drying properties and,
in addition, may be used at high infeed solid levels (Shahidi
and Han 1993). Moreover, Kailasapathy (2000) reported that
bioencapsulation of Bi®dobacterium and Lactobacillus strains
using a coat of calcium alginate and starch sustained greater
survival of these probiotic strains than calcium alginate
coating alone. Different ratios (5 : 1 and 10 : 1) of the starch
coating polymer to the bacterial core material were assessed
by comparing survival of spray-dried free cells with these
spray-encapsulated cells directly after spraying and follow-
ing storage at ambient temperature for a number of days.
Viability was improved by the presence of the starch coat
under these conditions with numbers of encapsulated cells as
much as 2 log higher than free cells after 5 d. However,
when added into acid buffers, both free and encapsulated
cells demonstrated similar rates of decline at pH 7á0, 4á5 and
ã 2001 The Society for Applied Microbiology, Journal of Applied Microbiology, 91, 1059±1066
1065
2á8 over a 6-h period, indicating that the starch coat offered
no protection against environmental acid stress. The viab-
ility pro®les of encapsulated microspheres and free cells
were also assessed in foods. Microparticles were ®rst
incorporated into muesli but no viable cells could be
cultured from encapsulated or free cell-containing powders
after 2 weeks. This poor survival may have been exacerbated
by the presence of fruit pieces in this food, the presence of
which was previously determined to have a negative effect
on probiotic survival (M. Playne, personal communication),
probably attributable to the water activity of the fruit. For
this reason, a dry malted beverage powder was also
investigated. Again, starch coating did not promote survival;
in fact, the mean from three independent trials indicated
that levels of free cells were 1 log higher after 20 d. The
reasons for the poor viability observed can only be
speculated upon. The hydrophilic nature of starch may
have caused absorption of intracellular ¯uid of the bacteria
as was suggested for the complete loss of viability of starch-
coated rhizobacteria within 1 week of storage (Amiet-
Charpentier et al. 1998). The enhanced survival observed
during storage of starch-coated compared with free cells may
have been due to the shorter monitoring period of 5 d. The
food trials, which were performed for a longer period of
20 d, may represent more accurately the in¯uence of the
starch coat on bacterial viability or possibly the moisture
content of the dry foods also contributed to microbial death.
The 16S rDNA sequence of Bi®dobacterium PL1 was
determined as it is a promising probiotic candidate but its
speciation is unknown. The 1491 nucleotide sequence
displayed highest similarities to a B. ruminantium strain
(97%) and a B. dentium strain (96%). These two strains were
themselves reported to be 97á1% similar and clustered with
B. adolescentis (Miyake et al. 1998). A value of 97% is
the threshold for being considered the same species
(Stackebrandt and Goebel 1994) thus, the PL1 strain could
tentatively be considered as B. ruminantium until further
16S rRNA sequence information for Bifdobacterium spp.
becomes available.
From the literature and the results obtained in this study,
further work is required to establish successful encapsula-
tion processes for probiotic strains. The lack of suitable
food-grade polymers is a problem. The so-called `enteric'
polymers would be an attractive choice for use and have
been demonstrated to provide excellent protection for
bi®dobacteria and lactic acid bacteria against simulated
in vivo conditions (Kim et al. 1988; Rao et al. 1989) and
should also sustain viability in high acid foods. Such
polymers, however, are not currently permitted for use in
foods although extensive pharmacopoeia information is
available. The results of this study indicated that, although
the starch coating used did not offer any protection to the
Bi®dobacterium strain in two dry food products investigated,
1066 K . O ' R I O R D A N ET AL.

the spray-coating process optimized in this study resulted in Lane, D.J. (1991) 15S/23S rRNA sequencing. In Nucleic Acid
small-sized microparticles, excellent coverage of cells and Techniques in Bacterial Systematics ed. Stackebrandt, E. and
adequate viability of Bi®dobacterium PL1 was maintained Goodfellow, M. pp. 115±147. Chichester: John Wiley.
during the drying process. Lee, K.-Y. and Heo, T.-R. (2000) Survival of Bi®dobacterium longum
immobilized in calcium alginate beads in simulated gastric juices and
bile salt solution. Applied and Environmental Microbiology 66,
ACKNOWLEDGEMENTS 869±873.
Mauriello, G., Aponte, M., Andol®, R., Moschetti, G. and Villani, F.
The authors wish to thank Jed Siva for technical assistance, (1999) Spray-drying of bacteriocin-producing lactic acid bacteria.
Henk Roubos for invaluable discussion and assistance with Journal of Food Protection 62, 773±777.
spray drying and Bettina Wolpensinger for guidance with McMeekin, T.A., Brown, J., Krist, K., Miles, D., Neumeyer, K.,
SEM. An Australian Research Council±Strategic Partner- Nichols, D.S., Olley, N.J., Presser, K., Ratkowsky, D.A., Ross, T.,
ship with Industry, Research and Training (ARC±SPIRT) Salter, M. and Soontranon, S. (1997) Quantitative microbiology: a
grant funded the work. basis for food safety. Emerging Infectious Diseases 3, 541±549.
Miyake, T., Watanabe, K. and Oyaizu, H. (1998) Phylogenetic analysis
of the genus Bi®dobacterium and related genera based on 16S rDNA
REFERENCES sequences. Microbiology and Immunology 42, 661±667.
O'Riordan, K., Wang, X., Siva, J. and Conway, P.L. (2000) Survival of
Altschul, S.F., Gish, W., Miller, W., Myers, E.W. and Lipman, D.J.
probiotic strains in foods. Microbiology Australia 21, A45.
(1990) Basic local alignment search tool. Journal of Molecular Biology
Rao, A.V., Shiwnarain, N. and Maharaj, I. (1989) Survival of
215, 403±410.
microencapsulated Bi®dobacterium pseudolongum in simulated gastric
Amiet-Charpentier, C., Gadille, P., Digat, B. and Benoit, J.P. (1998)
and intestinal juices. Canadian Institute of Food Science and
Microencapsulation of rhizobacteria by spray-drying: formulation
Technology Journal 22, 345±349.
and survival studies. Journal of Microencapsulation 15, 639±659.
Reid, G. (1999) The scienti®c basis for probiotic strains of Lactobacillus.
Beerens, H. (1991) Detection of bi®dobacteria by using propionic acid
Applied and Environmental Microbiology 65, 3763±3766.
as a selective agent. Applied and Environmental Microbiology 57,
Scardovi, V. (1986) The genus Bi®dobacterium. In Bergey's Manual
2418±2419.
of Systematic Bacteriology, Vol. 2. ed. Holt, J. London: Williams &
Fuller, R. (1989) Probiotics in man and animals. Journal of Applied
Wilkins.
Bacteriology 66, 365±378.
Selmer-Olsen, E., Birkeland, S.-E. and Sorhaug, T. (1999) Effect of
Gardiner, G.E., O'Sullivan, E., Kelly, J., Auty, M.A.E., Fitzgerald, G.F.,
protective solutes on leakage from and survival of immobilized
Collins, J.K., Ross, R.P. and Stanton, C. (2000) Comparative
Lactobacillus subjected to drying, storage and rehydration. Journal
survival rates of human-derived probiotic Lactobacillus paracasei and
of Applied Microbiology 87, 429±437.
L. salivarius strains during heat treatment and spray drying. Applied
Shah, N.P. (2000) Probiotic bacteria: selective enumeration and
and Environmental Microbiology 66, 2605±2612.
survival in dairy foods. Journal of Dairy Science 83, 894±907.
Gibbs, B.F., Kermasha, S., Alli, I. and Mulligan, C.N. (1999)
Shahidi, F. and Han, X.-Q. (1993) Encapsulation of food ingredients.
Encapsulation in the food industry: a review. International Journal
Critical Reviews in Food Science and Nutrition 33, 501±547.
of Food Science and Nutrition 50, 213±224.
Sheu, T.-Y., Marshall, R.T. and Heymann, H. (1993) Improving
Gobbetti, M., Corsetti, A., Smacchi, E., Zocchetti, A. and de Angelis, M.
survival of culture bacteria in frozen desserts by microentrapment.
(1998) Production of Cresenza cheese by incorporation of bi®do-
Journal of Dairy Science 76, 1902±1907.
bacteria. Journal of Dairy Science 81, 37±47.
Stackebrandt, E. and Goebel, B.M. (1994) Taxonomic note: a place for
Kailasapathy, K. (2000) Probiotic bacteria: improved delivery by bio-
DNA±DNA reassociation and 16S rRNA sequence analysis in the
encapsulation. Probiotica 9, 1±5.
present species de®nition in bacteriology. International Journal
Kailasapathy, K. and Chin, J. (2000) Survival and therapeutic potential
of Systematic Bacteriology 44, 846±849.
of probiotic organisms with reference to Lactobacillus acidophilus and
Tannock, G.W., Munro, K., Harmsen, H.J.M., Welling, G.W., Smart, J.
Bi®dobacterium spp. Immunology and Cell Biology 78, 80±88.
and Gopal, P.K. (2000) Analysis of the fecal micro¯ora of human
Kailasapathy, K. and Supriadi, D. (1996) Effect of whey protein
subjects consuming a probiotic product containing Lactobacillus
concentrate on the survival of Lactobacillus acidophilus in lactose
rhamnosus DR20. Applied and Environmental Microbiology 66,
hydrolysed yoghurt during refrigerated storage. Milchwissenschaft 51,
2578±2588.
565±569.
Wang, X., Brown, I.L., Evans, A.J. and Conway, P.L. (1999) The
Khalil, A.H. and Mansour, E.H. (1998) Alginate encapsulated bi®do-
protective effects of high amylose maize (amylomaize) starch
bacteria survival in mayonnaise. Journal of Food Science 63, 702±705.
granules on the survival of Bi®dobacterium spp. in the mouse
Kim, H.S., Kamara, B.J., Good, I.C. and Enders, G.L. (1988) Method
gastrointestinal tract. Journal of Applied Microbiology 87, 631±639.
for the preparation of stabile microencapsulated lactic acid bacteria.
Journal of Industrial Microbiology 3, 253±257.

ã 2001 The Society for Applied Microbiology, Journal of Applied Microbiology, 91, 1059±1066