Article
Culture-dependent diversity profiling of spoilage yeasts
species by PCR-RFLP comparative analysis
Kereng M Corbett and Olga de Smidt
Abstract
Spoilage caused by yeasts is a constant, widespread problem in the beverage industry that can result in
major economic losses. Fruit juices provide an environment that allows the proliferation of yeast. Some
factories in South Africa are not equipped with laboratory facilities to identify spoilage yeasts and outsourcing
becomes a prolonged process which obstructs corrective action planning. This study aimed to establish
yeast diversity and apply a rapid method for preliminary identification of spoilage yeasts associated with a
small-scale fruit juice bottling factory. Yeast population in the factory was determined by isolation from the
production environment, process equipment and spoiled products. PCR-RFLP analysis targeting the 5.8S-ITS
region and D1/D2 sequencing was used for identification. A total of 207 yeasts belonging to 10 different
genera (Candida, Lodderomyces, Wickerhamomyces, Yarrowia, Zygosaccharomyces, Zygoascus,
Cryptococcus, Filobasidium, Rhodotorula/Cystobasidium and Trichosporon) were isolated and identified
from the production environment and processing equipment. Candida intermedia, C. parapsilosis and
Lodderomyces elongisporus were widely distributed in the factory. Zygosaccharomyces bailii, Z. bisporus,
Zygoascus hellenicus and Saccharomyces cerevisiae were isolated from the spoiled products. The data
provided a yeast control panel that was used successfully to identify unknown yeasts in spoiled products
from this factory using polymerase chain reaction-restriction length polymorphism (PCR-RFLP) comparative
analysis.
Keywords
Fruit juice, spoilage yeast, 5.8S-ITS region, yeast diversity, RFLP analysis
Date received: 12 March 2019; accepted: 22 May 2019
INTRODUCTION
The food and beverage industry is one of the most
important components of South Africa’s manufactur-
ing sector. Beverages account for just over 4% of all
manufacturing sales, while in the food and beverage
sector, beverages represent 24% of sales (Food and
Beverages Manufacturing Sector Education and
Training Authority. SSP 2011–2016). Increased con-
sumption of fruit juices has a direct influence on the
economy in a positive way, but becomes negatively
affected when foodborne disease outbreaks and spoil-
age problems occur (Tribst et al., 2009).
Yeast spoilage is a constant and widespread problem
in the beverage industry (Aneja et al., 2014). This type
of spoilage is predictable and mainly occurs in those
products where bacterial growth is either impeded or
prevented by predominating intrinsic, extrinsic and
processing factors. Typically, acidic low pH foods and
products with a high sugar content, such as fruit juice
concentrates, are affected (Tribst et al., 2009). The most
visible sign of yeast spoilage is recognised by swelling of
the product container due to gas production that
Centre for Applied Food Security and –Biotechnology (CAFSaB),
Central University of Technology, Bloemfontein, South Africa
Corresponding author:
Olga de Smidt, Centre for Applied Food Security and –
Biotechnology (CAFSaB), Central University of Technology, Free
State, Bloemfontein 9301, South Africa.
Email: odesmidt@cut.ac.za
Food Science and Technology International 25(8) 671–679
! The Author(s) 2019 Article reuse guidelines:
sagepub.com/journals-permissions
DOI: 10.1177/1082013219856779
journals.sagepub.com/home/fst
results from fermentation, referred to as ‘blowing’
(Kurtzman and James, 2006; Wang et al., 2016). This
form of spoilage causes major financial losses to the
bottling factories involved (Arias et al., 2002).
Industrial facilities harbour yeast strains that are
adapted to the processing environment and circumvent
the barriers set up to prevent them (Doyle, 2007;
Martorell et al., 2005). Precautionary measures to
avoid or reduce microbial spoilage require high expend-
iture. Despite quality control practices in bottling fac-
tories that entail rigorous microbial monitoring of
pulps, water, air and equipment, yeast spoilage con-
tinues to occur. The origin of spoilage yeast seldom
follows a set pattern and the modes of contamination
are often specific to each processing facility. It therefore
seems mandatory to study each particular manufactur-
ing plant to determine the origin of spoilers (Herna´ndez
et al., 2018).
To design adequate strategies to prevent spoilage, it
is advantageous to know the identity of the spoilage
organisms present in the product and gain insight into
the source of contamination (Loureiro, 2000).
Improved techniques with increased specificity, discrim-
inatory power and shorter detection times for the
identification of spoilage yeasts in foods and drinks
are becoming increasingly important in the food
sector (Casey and Dobson, 2004; Herna´ndez et al.,
2018). Many recent, complex and sensitive identifica-
tion tools for rapid identification of pathogens and
spoilers have become available, but for small-scale
fruit juice bottling factories, more basic methods are
still required, at least for preliminary identification
(Kesmen et al., 2018).
One such fruit juice bottling factory in
Bloemfontein, South Africa, experiences problems
with ‘blowing’ of certain juice concentrates on an
annual basis. The problem is aggravated by long wait-
ing periods for identification of the spoilage contamin-
ants, which have delayed corrective actions in the past.
Therefore, the aim of this research was to compile a
culture-dependent yeast diversity profile of the factory
environment/equipment and apply comparative PCR-
RFLP analysis to identify spoilage yeasts associated
with its products.
MATERIALS AND METHODS
Sampling protocol
The factory has been operating for 24 years producing
fruit juice concentrates of different flavours. It consists
of nine blending tanks and three filling lines.
Approximately 24,000 l of juice is produced per day.
Surface swabs were obtained from the production
environment and processing equipment. Areas included
the refrigerator, powder blenders, pipes, blending
tanks, holding tanks, nozzles, ramp, bottle and caps.
Yeast isolates originating from weekly routine analysis
of surface swabs and air samples taken after Cleaning
in Place (CIP) protocols were carried out and were
isolated on chloramphenicol agar plates. All isolates
were collected over a period of one year and cryopre-
served in 15% glycerol at 80 C. Eight fruit juice sam-
ples, each representing a different batch affected by
spoilage (blowing) during the sampling period, were
collected, retained on ice during transportation and
analysed without delay. Spoiled sample flavours
included ice tea, fruit flavour, cordial and fruit/dairy
blend.
Enumeration and isolation
Surface swabs were suspended in 10ml peptone water
and serially diluted. Dilutions were plated onto Rose
Bengal Chloramphenicol (RBC) agar and incubated at
30 C for 48–120 h. Yeast colonies from chlorampheni-
col agar plates that were provided by the factory tech-
nician were also transferred to RBC agar and incubated
at the same conditions. The resulting colonies were
selected based on differences in colony morphology
and purified by repeated sub-culturing (Barata et al.,
2008).
Spoiled fruit juice samples were also serially diluted
in sterile peptone water. The series of dilutions were
plated onto RBC agar and Malt Extract agar, and incu-
bated for up to three days at 30 C. Of the resulting
colonies, approximately 10% were randomly selected
based on colony morphology and used as template in
whole cell PCR (Barata et al., 2008). Isolates were pur-
ified by repeated sub-culturing and cryopreserved in
15% glycerol at 20 C.
PCR amplification and RFLP analysis of
5.8S-ITS region
Whole cell PCR amplification and RFLP analysis of
the 5.8S-ITS region were performed on all isolates
and reference strains. For comparison and preliminary
identification, reference strains frequently isolated
from fruit juice were obtained from the UNESCO-
MIRCEN Biotechnological Yeast Culture Collection
of the University of the Free State. The 17 reference
strains, also cryopreserved as described before,
included Candida intermedia (UOFS Y-0649), Candida
parapsilosis (UOFS Y-0206), Candida tropicalis
(UOFS Y-0534), Dekkera anomala (UOFS Y-1062),
Hanseniaspora occidentalis (UOFS Y-0153),
Kluyveromyces marxianus (UOFS Y-0797), Lodderomyces
Food Science and Technology International 25(8)
672
elongisporus (UOFS Y-2394),Millerozyma farinosa (UOFS
Y-0203), Pichia kudriavzevii (UOFS Y-0814), Rhodotorula/
Cystobasidium slooffiae (UOFS Y-0972), Saccharomyces
bayanus (UOFS Y-0912), Saccharomyces cerevisiae
(UOFS Y-0792), Saccharomycodes ludwigii (UOFS
Y-0540), Torulaspora delbrueckii (UOFS Y-1016),
Wickerhamomyces anomalus (UOFS Y-0810),
Zygosaccharomyces bailii (UOFS Y-1535) and
Zygosaccharomyces rouxii (UOFS Y-0763).
Yeast cells from 48h single colonies were suspended in
50ml PCR reaction mix containing 0.52mM primer ITS 1
(50-TCCGTAGGTGAACCTGCGG-30) and ITS 4 (50-
TCCTCCGCTTATTGATATGC-30) (Rojo et al.,
2017), 0.2mM dNTPs, 1X reaction buffer ThermoPol,
and 1 U of NEB Taq ThermoPol (New England
Biolabs). Amplification conditions included one cycle at
95 C for 3min, followed by 30 cycles of 95 C for 30 s,
55 C for 30 s, 68 C for 1min. A final elongation step
was performed at 68 C for 7min. Successful amplifica-
tion was confirmed by agarose gel (1%) electrophoresis.
All amplicons (10ml) were digested with 1 unit of CfoI,
HaeIII andHinfI restriction enzymes using FastDigestTM
and TangoTM buffers (Thermo Scientific) in separate
reactions (Rojo et al., 2017). Fragments were separated
in a 3% agarose gel and digital images were captured
with the Molecular Imager Gel DocTM XR system
(BioRad Laboratories Inc.). Band sizes were calculated
with reference to a GeneRulerTM 1kb DNA ladder Plus
and GeneRulerTM 50bp DNA ladder (Thermo
Scientific), using Quantity One 1-D Analysis software
(BioRad Laboratories Inc.). Resulting PCR-RFLPs were
grouped according to profiles (Table 1), compared to the
reference strain profiles and yeast-ID database
(rank 20bp) for preliminary identification and Sanger
sequenced for confirmation.
Amplification and sequencing of D1/D2 domain
The D1/D2 domain of the 26S rRNA gene was ampli-
fied from at least three isolates representative of a spe-
cific PCR-RFLP profile and sequenced for identity
confirmation. Amplification of the D1/D2 domain
was performed using primers NL-1 (50-GCATA
TCAATAAGCGGAGGAAAAG-30) and NL-4 (50-
GGTCCGTGTTTCAAGACGG-30) (White et al.,
1990). Sequencing reactions were carried out using the
Big Dye Terminator v3.1 Cycle Sequencing Kit
(Applied Biosystems). Data were analysed using
Chromas LITE version 2.1.1 and compared with previ-
ously published sequences using the BLAST algorithm
(Altschul, 1997) for species identification (http://blast.
ncbi.nlm.nih.gov/Blast.cgi). Sequences were deposited
into the NCBI database with accession numbers
KU708234.1–KU708252.1.
RESULTS AND DISCUSSION
Identification of spoilage yeast using PCR-
RFLP comparative analysis
A total of 207 yeasts were isolated and identified
according to 5.8S-ITS polymorphisms (Ting et al.,
2018). The isolates showed different PCR product
sizes, ranging from 300 to 900 bp. The PCR products
digested with CfoI, HaeIII and HinfI enzymes were
analysed for all isolated strains and 18 distinct profiles
were obtained (Table 1), designated by a letter of the
alphabet. For comparison, PCR-RFLP of the 5.8S-ITS
region was applied simultaneously to reference strains
from the UNESCO-MIRCEN Biotechnological Yeast
Culture Collection as controls. Representatives of each
of the 18 profiles were confirmed by sequencing the D1/
D2 domain of the 26S rRNA gene.
It could be argued that an indigenous yeast control
panel might appear redundant, since yeast identifica-
tion databases constructed on RFLP data are available.
However, discrepancies can be observed in amplicon
sizes and restriction profiles among species in different
studies. It may be expected that differences in recorded
fragment sizes as great as 20 bp are possible simply due
to the manner in which the bands sizes were deter-
mined. This would account for many of the small size
variations reported by different researchers for the same
strain of a species (Coton et al., 2006; Jeyaram et al.,
2008; Pham et al., 2011; Satora et al., 2013; Sun and
Liu, 2014). Although innovative and user friendly, the
yeast-ID database (CECT-IATA, Spanish Type
Culture Collection, Universitat de Valencia, Valencia,
Spain; www.yeast-id.org) could identify only 22% of
the isolates (Table 1). It is also possible that the data-
base is limited to the specific culture collection and may
not contain the yeast species of interest, which was the
case with Candida quercitrusa, Candida sojae, C. sloof-
fiae, Filobasidium uniguttulatum and Trichosporon
ovoides isolated from the factory environment in this
study.
Yeasts that were isolated from eight spoiled fruit
juices each representing a different batch affected by
spoilage yielded populations ranging from 2.80 103
to 2.23 107 CFU/ml. PCR product lengths were
between 600 and 900 bp and was already an indication
that spoilage of the different juices was likely caused by
the presence of different yeast species. PCR products
were digested with CfoI, HaeIII and HinfI. Table 2 pre-
sents the digestion profiles of unknown yeasts.
Restriction profiles were compared to that of the con-
trol panel and subsequently identified. The restriction
profile of unknown yeasts 1 and 3 was not identical to
any of the isolates from the control panel and prelim-
inary identification was not possible. The isolates were
Corbett and de Smidt
673
T
a
b
le
1
.
C
h
a
ra
ct
e
ri
sa
tio
n
o
f
ye
a
st
is
o
la
te
s
b
a
se
d
o
n
5
.8
S
-I
T
S
re
g
io
n
P
C
R
a
n
d
R
F
L
P
d
a
ta
,
a
n
d
D
1
/D
2
d
o
m
a
in
se
q
u
e
n
ce
id
e
n
tif
ic
a
tio
n
s1
P
ro
fil
e
Is
o
la
te
s
R
F
L
P
b
a
se
d
id
e
n
tif
ic
a
tio
n
F
ra
g
m
e
n
t
le
n
g
th
s
(b
p
)
Id
e
n
tif
ic
a
tio
n
(D
1
/D
2
d
o
m
a
in
se
q
u
e
n
ci
n
g
)
U
N
E
S
C
O
-M
IR
C
E
N
B
io
te
ch
Y
e
a
st
C
u
ltu
re
C
o
lle
ct
io
n
Y
e
a
st
-I
D
.o
rg
d
a
ta
b
a
se
(%
id
e
n
tit
y)
P
C
R
C
fo
I
H
a
e
III
H
in
fI
G
6
N
o
m
a
tc
h
Y.
lip
o
ly
tic
a
(1
0
0
%
)
3
7
7
2
1
4
,
1
7
4
3
7
7
1
8
2
Y.
lip
o
ly
tic
a
C
1
2
7
C
.
in
te
rm
e
d
ia
C
.
in
te
rm
e
d
ia
/p
se
u
d
o
in
te
rm
e
d
ia
/c
a
te
n
u
-
la
te
/h
a
e
m
u
lo
n
is
(1
0
0
%
)
4
0
0
2
1
2
,
1
7
4
4
0
0
2
1
5
,
1
9
3
C
.
in
te
rm
e
d
ia
V
8
N
o
m
a
tc
h
C
ry
p
to
c
o
c
c
u
s
h
u
n
g
a
ri
c
u
s/
lu
te
o
lu
s
(8
8
%
)
4
8
0
2
7
2
4
7
6
2
3
4
T.
o
vo
id
e
s
A
1
3
8
C
.
tr
o
p
ic
a
lis
C
.
m
o
n
ta
n
a
(6
2
%
)
5
1
5
2
9
0
,
2
2
0
4
1
0
,
1
1
2
2
8
1
,
2
5
7
C
.
p
a
ra
p
si
lo
si
s
P
5
C
.
p
a
ra
p
si
lo
si
s
C
.
p
a
ra
p
si
lo
si
s
(1
0
0
%
)
5
1
5
2
5
2
,
2
0
2
,
6
4
3
9
9
2
3
6
C
.
la
u
re
n
tii
R
1
7
N
o
m
a
tc
h
C
ry
p
to
c
o
c
c
u
s
sk
in
n
e
ri
(6
2
%
)
5
2
7
2
9
3
,
2
3
8
4
6
8
2
7
4
C
.
so
ja
e
O
1
2
L
.
e
lo
n
g
is
p
o
ru
s
C
.
m
a
lto
se
/
L
.
e
lo
n
g
is
p
o
ru
s
(1
0
0
%
)
5
7
6
3
2
3
,
2
4
0
5
2
7
2
9
8
,
2
6
1
L
.
e
lo
n
g
is
p
o
ru
s
X
8
N
o
m
a
tc
h
C
.
m
u
lti
g
e
m
m
is
/C
.
va
ld
iv
ia
n
a
(7
5
%
)
5
9
6
3
1
7
4
2
3
,
1
4
1
3
1
8
C
.
o
le
o
p
h
ila
U
1
1
N
o
m
a
tc
h
C
.
la
u
re
n
tii
/W
ic
ke
rh
a
m
o
m
yc
e
s
b
o
vi
s/
P
ic
h
ia
d
ry
a
d
o
id
e
s
(7
5
%
)
6
1
1
6
1
2
5
8
6
3
3
9
,
2
6
5
C
.
sp
a
n
d
o
ve
n
si
s
A
1
4
N
o
m
a
tc
h
C
yb
e
rl
in
d
n
e
ra
b
im
u
n
d
a
lis
/
W
ic
ke
rh
a
m
o
m
yc
e
s
st
ra
sb
u
rg
e
n
si
s
(8
8
%
)
6
1
2
2
8
1
,
3
1
0
5
0
8
3
5
7
,
2
8
0
F.
c
a
p
su
lig
e
n
u
m
H
8
N
o
m
a
tc
h
K
u
ra
is
h
ia
c
a
p
su
la
te
/
C
.
e
rg
a
te
n
si
s
(7
9
%
)
6
2
4
3
9
2
,
3
0
8
4
8
7
2
6
1
,
2
4
2
,
1
4
6
F.
u
n
ig
u
ttu
la
tu
m
I1
5
N
o
m
a
tc
h
C
.
sa
n
ta
m
a
ri
a
e
/a
tla
n
tic
a
S
c
h
w
a
n
n
io
m
yc
e
s
p
se
u
d
o
p
o
ly
m
o
rp
h
u
s
Ya
m
a
d
a
zy
m
a
M
ex
ic
a
n
a
(9
2
%
)
6
3
4
3
2
0
,
2
2
6
,
6
6
4
2
0
,
1
3
7
3
1
2
C
.
q
u
e
rc
itr
u
sa
E
4
N
o
m
a
tc
h
C
yb
e
rl
in
d
n
e
ra
m
ey
e
ra
e
/
F.
c
a
p
su
lig
e
n
u
m
6
2
%
)
6
3
5
3
8
5
5
3
6
3
5
0
,
2
9
2
C
.
sa
ito
i
F
2
5
W
.
a
n
o
m
a
lu
s
W
.
a
n
o
m
a
lu
s
(8
8
%
)
6
4
0
5
6
9
6
4
0
3
0
8
W
.
a
n
o
m
a
lu
s
K
1
7
N
o
m
a
tc
h
Z
.
h
e
lle
n
ic
u
s/
W
ic
ke
rh
a
m
o
m
yc
e
s
b
is
-
p
o
ru
s
(9
2
%
)
6
4
3
3
3
1
6
4
3
3
4
3
,
1
7
1
,
1
2
2
Z
.
h
e
lle
n
ic
u
s
W
1
0
R
.
sl
o
o
ff
ia
e
C
.
g
la
e
b
o
sa
(7
5
%
)
6
4
7
6
4
7
6
4
7
3
2
5
,
2
5
3
R
.
sl
o
o
ff
ia
e
R
6
N
o
m
a
tc
h
S
a
c
c
h
a
ro
m
yc
e
s
kl
u
yv
e
ri
(7
3
%
)
6
8
7
3
4
1
,
3
2
1
3
8
7
,
2
0
9
2
2
1
,
2
1
5
,
1
0
3
R
.
d
a
ir
e
n
e
n
si
s
D
6
Z
.
b
a
ili
i
Z
.
b
a
ili
i
(6
9
%
)
7
8
2
3
3
6
,
2
8
4
,
8
9
7
1
2
3
3
1
,
2
3
0
,
1
6
0
Z
.
b
a
ili
i
1
D
a
ta
a
rr
a
n
g
e
d
a
cc
o
rd
in
g
to
P
C
R
p
ro
d
u
ct
si
ze
;
fr
a
g
m
e
n
ts
sm
a
lle
r
th
a
n
5
0
b
p
w
e
re
n
o
t
in
cl
u
d
e
d
.
Food Science and Technology International 25(8)
674
sequenced (D1/D2 domain) and identified as
Zygosaccharomyces bisporus and S. cerevisiae, respect-
ively. The restriction profiles of unknown yeasts 2 and 4
were identical to that of Z. bailii and Zygoascus helle-
nicus and were verified by D1/D2 domain sequencing.
Spoiled ice tea and fruit flavours contained only
Z. bailii. Cordial contained Z. bailii, Z. bisporus and
Z. hellenicus. Both Z. bailii and S. cerevisiae were iso-
lated from the fruit/dairy blend.
Yeast diversity in the production environment
and processing equipment
Many types of yeasts are potential spoilage agents of
fresh and concentrated fruit juices due to favourable
pH conditions and high sugar levels of these beverages
(Tribst et al., 2009). Contamination may originate from
any step along the manufacturing process. Raw mater-
ials, factory environment, packaging and processing
equipment are all potential sources of contamination
(Herna´ndez et al., 2018). In a ‘forensic approach’ to
spoilage of soft drinks, Davenport (1996) noted that
most yeast contaminants encountered could be divided
into four categories, which he referred to as Groups
1–4. Group 1 constitutes spoilage yeasts that are
fermentative and preservative resistant. Group 2 com-
prises spoilage or hygiene types and Group 3 are indi-
cators of poor factory hygiene. Group 4 yeasts are
‘aliens’ which are out of their normal environment
(Davenport, 1996).
The yeast population from the production environ-
ment and processing equipment comprised 10 different
genera. Ascomycetous yeast species included C. inter-
media (13%), C. parapsilosis (18%), C. sojae (3%), C.
quercitrusa (2%), C. spandovensis (5%), C. oleophila
(4%), L. elongisporus (6%), W. anomalus (12%),
Yarrowia lipolytica (3%), Z. bailii (3%) and
Z. hellenicus (3%). Yeast diversity in this factory was
dominated by Ascomycetes, which was not surprising
since most ascomycetous yeasts have been found in
environments with high concentrations of sugar
(Molna´rova´ et al., 2014). Basidiomycetous yeasts were
represented by C. slooffiae (5%), Rhodotorula dairenen-
sis (3%), Cryptococcus laurentii (2%), Cryptococcus
saitoi (2%), F. uniguttulatum (4%), Filobasidium capsu-
ligenum (7%) and T. ovoides (4%). Basidiomycetous
yeasts are generally found in soil, plant materials and
bird droppings, and are not usually associated with
spoilage and industrial processes, but are regarded as
hygiene-indicator species (Tekolo et al., 2010). As such,
these yeasts were also less abundant in the factory
equipment compared to ascomycetous yeasts
(Figure 1).
Diversity distribution varied among the different
equipment. C. slooffiae, Z. hellenicus, W. anomalus
and Z. bailii were isolated from the air samples taken
from the refrigerator where the concentrated pulps were
stored. C. slooffiae, previously known as Rhodotorula
slooffiae (Yurkov et al., 2015), was the dominant yeast
isolated from this area.
In the powder blenders that are used for mixing
powder ingredients, eight different yeast species were
detected, of which C. intermedia and C. parapsilosis
dominated. Both yeasts have been isolated previously
from reconstituted fruit juice (Maciel et al., 2013) and
C. parapsilosis has been reported as an opportunistic
pathogen responsible for various mycoses (Jacques
and Casaregola, 2008). It is reasonable to assume that
contamination of equipment by C. parapsilosis is a
result of food handlers, since this organism is frequently
found in blood, skin and nails (including hands
of healthcare workers) (Nosek et al., 2009).
Furthermore, Welthagen and Viljoen (1998) reported
that workers’ hands and aprons were also responsible
Table 2. Characterisation of yeast isolated from spoiled fruit juices based on 5.8S-ITS region PCR and RFLP data, and
D1/D2 domain sequence identifications
RFLP based identification Fragment lengths (bp)
Identification
(D1/D2 domain
sequencing)Isolate Juices
Factory control
panel yeast PCR CfoI HaeIII HinfI
Unknown 1 Cordial No match 765 286, 252 692 393, 228, 148 Z. bisporus
Unknown 2 Ice tea
Fruit
Cordial
Fruit/dairy
blend
Z. bailii 781 322, 270, 85 697 349, 220, 160 Z. bailii
Unknown 3 Fruit/dairy
blend
No match 844 357, 335, 127 306, 224, 169 356, 114 S. cerevisiae
Unknown 4 Cordial Z. hellenicus 630 317 623 327, 161, 110 Z. hellenicus
Corbett and de Smidt
675
for a high rate of yeast contamination in other food
processing environments. The isolates from the
powder blenders comprised Group 2 yeasts that are
spoilage and hygiene types (Davenport, 1996). This
group is able to cause spoilage of fruit juices, but
only if complications arise during manufacturing,
such as low level or absence of preservative, ingress of
oxygen, pasteurisation failure or poor standards of
hygiene (Davenport, 1997).
The pipes that connect the powder blenders to the
blending tanks were largely colonised by C. parapsilo-
sis. T. ovoides is classified by Davenport (1996, 1997,
1998) as belonging to Group 3 organisms that are
hygiene indicators, not causing spoilage. These species
are able to utilise different carbohydrates and carbon
sources and degrade urea, but members of this genus
are non-fermentative. W. anomalus, C. sojae and
R. dairenensis were also isolated from the pipes. Not
unexpectedly, the blending tanks shared similar diver-
sity with the pipes and also contained F. capsuligenum,
C. intermedia, L. elongisporus, C. slooffiae, Z. hellenicus,
C. quercitrusa and C. spandovensis. The high sugar con-
tent and low water activity of the ingredients in the
blending tanks favour the growth of yeasts, which con-
tributed to the considerable diversity isolated from this
equipment.
C. parapsilosis, L. elongisporus and C. saitoi were
present in the nozzles that release the fruit juice into
the bottles during filling. These yeast are common con-
taminants in bottling factories, but can be effectively
controlled if Good Manufacturing Practices (GMPs)
are strictly adhered to (Davenport, 1996).C. parapsilosis,
L. elongisporus, C. sojae, C. oleophila and C. laurentii
were isolated from the bottles and caps used for packa-
ging during filling. Caps and bottles are not washed
prior to filling and are stored in the roof area, which is
not properly insulated. The latter is likely to introduce
soil and dust-related yeasts, such as L. elongisporus and
C. laurentii, into the packaging material (Cloete et al.,
2010; Sla´vikova´ et al., 2007; Stratford and James, 2003).
Lodderomyces 
elongisporus
Candida sojae
Saccharomyces cereisiae
Wiskerham
omyces anom
alus
Yarrowia lipolytica
Zygoascus hellenicus
Can
dida
 spa
ndo
ven
sis
Ca
nd
ida
 qu
erc
itru
sa
Ca
nd
ida
 pa
ra
ps
ilo
sis
Ca
nd
id
a 
ol
eo
ph
ila
Ca
nd
id
a 
in
te
rm
ed
iaTrichosporon ovoides
Rhodotorula slooffiae
Rhodotorula dairenensis
Filobasidium uniguttulatum
Filobasidium capsuligen
um
Crypto
coccu
s sato
i
Cry
pto
coc
cus
 lau
ren
tii
Zy
go
sa
cc
ha
ro
my
ce
s b
isp
or
us
Zy
go
sa
cc
ha
ro
m
yc
es
 b
ai
lii
Yeast present in
area
Refrigerator
Powder blenders
Pipes
Blending tanks
Holding tanks
Nozzles
Bottles & caps
Ramp
Fruit juice
Figure 1. Data wheel showing the yeast distribution in the fruit juice bottling factory and spoiled products. View clock-
wise from C. intermedia to Z. bisporus shows the distribution of Ascomycetes, and from C. laurentii to T. ovoides that of
Basidiomycetes. The coloured legend indicates the area/origin sampled and a red dot when the specific yeast species
was detected.
Food Science and Technology International 25(8)
676
Relevance of detected spoilage yeasts
Z. bailii, Z. bisporus, Z. hellenicus and S. cerevisiae
isolated from the spoiled fruit juices are categorised
as Group 1 yeasts (Davenport, 1996) that have been
described as spoilage organisms adapted to growth in
fruit juices and being able to cause spoilage from very
low cell numbers. The characteristics of Group 1 yeasts
are osmotolerance, aggressive fermentation, resistance
to preservatives (particularly weak organic acids) and a
requirement for vitamins. The high sugar concentra-
tions in fruit juice concentrates favour the growth of
yeasts with a higher fermentative activity (Brugnoni
et al., 2012).
Osmotolerant Z. bailii is commonly encountered in
high sugar (40–70%) environments responsible for con-
siderable economic losses in the beverage industry
(Loureiro, 1994; Rojo et al., 2014; Stratford et al.,
2013; Wang et al., 2016). Z. bailii was detected in all
the spoiled fruit juice flavours, which was not unex-
pected given its extreme resistance to preservatives
and ability to grow in concentrations of preservatives
in excess of the legally permitted levels (Harrison et al.,
2011). The low permeability of Z. bailii to weak acid
preservatives at low pH values, and its ability to metab-
olise acid compounds, even in the presence of glucose,
are some of the physiological traits associated with its
high tolerance to acidic environments (Sousa et al.,
1996). The factory from which the spoiled fruit juices
were obtained uses sodium benzoate and sodium meta-
bisulphite preservatives, which are classified as weak
acid preservatives, and are thus easily resisted by mem-
bers of the genus Zygosaccharomyces.
Z. bailii was not widely distributed in the processing
equipment analysed. This observation could be linked
to the fact that it exhibits the lowest capacity of adhe-
sion to stainless steel and also the lowest percentage of
hydrophobicity (Brugnoni et al., 2007). Stainless steel is
the most frequently used food contact material in the
fruit juice processing industry and the factory investi-
gated was no exception. The combination of the two
parameters pertaining to adhesion and hydrophobicity
leads to a significant decrease of the adhesion capacity
of this species (Brugnoni et al., 2007). Although
Z. bailii may be present in low numbers on stainless
steel, it can show gradual proliferation in concentrates
and only one cell per container of diluted stock can
cause spoilage (Wareing and Davenport, 2005).
Although Z. bisporus is isolated from foods at a sub-
stantially lower frequency than Z. bailii, it has a similar
ability to cause food spoilage and is also preservative
resistant (Barata et al., 2012).
Zygoascus hellenicus is highly fermentative and has
been described extensively as being associated with
grape berries or must in the winery environment
(Barata et al., 2008; Simo˜es and Gomes, 2015). The
cordial, which was the only fruit juice type contami-
nated by Z. hellenicus, consisted of strawberry, cran-
berry and raspberry pulp. It could be possible that
this species originated from the pulp since it has been
associated with contamination of berries. It has also
been described as a contaminant often associated with
damaged grapes (Barata et al., 2008) and some studies
indicated that it has been isolated from fruit juices
(Maciel et al., 2013; Nyanga et al., 2013). It was not
unusual to isolate S. cerevisiae, also a fermentative
yeast associated with microbial decomposition of fruit
juices (Turtoi, 2014). This yeast can grow in a range of
conditions and is characterised for its optimal growth
on high sugar media. S. cerevisiae has also been isolated
from a variety of dairy products, especially those con-
taining sugar and fruit (Mayoral et al., 2005). Not sur-
prisingly in this study, the fruit/dairy blend containing
milk powder was populated by S. cerevisiae.
CONCLUSION
Culture-dependent PCR-based techniques are primarily
chosen to identify yeasts due to their technical simplicity
and relatively low cost. Restriction enzyme digestion of
the ITS regions is still commonly used for yeast species
identification. A culture-dependent approach was also
chosen for this study to capitalise on the sampling
protocol and yeast enumeration methods that already
formed part of the factory’s microbial quality control.
Consequently, this approach not only ensured consist-
ent standard operating procedures and sufficient cover-
age of the area for diversity analysis, but also a control
panel of yeasts that could be used as comparative ref-
erences to identify isolates from spoiled products. Yeast
diversity data revealed that the processing environment
and equipment of the fruit juice bottling factory inves-
tigated harboured a variety of yeast species, despite
rigorous cleaning and disinfection. The dominant pres-
ence of hygiene indicator species C. intermedia, C. para-
psilosis and W. anomalus in the factory environment
after CIP suggests that the current GMP requires atten-
tion. Z. bailii and Z. hellenicus were both mainly iso-
lated from the air in the refrigerator where the fruit
pulp was stored, which might imply the refrigerator as
the likely source of contamination. S. cerevisiae and Z.
bisporus were not detected in the factory environment,
which showed that spoilage yeasts did not necessarily
originate from the factory environment. Although it is
unlikely that these micro-organisms were missed during
sampling, it cannot be excluded as a possibility. It is,
however, more reasonable to assume that they origi-
nated from an external source, presumably the concen-
trated pulps. Pre-screening fruit pulp for spoilage yeast
could be a useful approach towards spoilage prevention
for this particular factory.
Corbett and de Smidt
677
ACKNOWLEDGEMENTS
Thanks to Dr Daleen Struwig, medical writer/editor, Faculty
of Health Sciences, University of the Free State,
Bloemfontein, South Africa, for technical and editorial prep-
aration of the manuscript.
DECLARATION OF CONFLICTING INTERESTS
The author(s) declared no potential conflicts of interest with
respect to the research, authorship, and/or publication of this
article.
FUNDING
The author(s) disclosed receipt of the following financial sup-
port for the research, authorship, and/or publication of this
article: This material is based upon work financially sup-
ported by the National Research Foundation (NRF). Any
opinion, findings and conclusions or recommendations
expressed in this material are those of the author(s) and there-
fore the NRF does not accept any liability in regard thereto.
ORCID ID
Olga de Smidt https://orcid.org/0000-0002-0247-3624
REFERENCES
Altschul SF. (1997). Evaluating the statistical significance of
multiple distinct local alignments. In: Suhai S (ed.)
Theoretical and Computational Methods in Genome
Research. New York: Plenum, pp. 1–14.
Aneja KR, Dhiman R, Aggarwal NK and Aneja A. (2014).
Emerging preservation techniques for controlling spoilage
and pathogenic microorganisms in fruit juices.
International Journal of Microbiology 2014: 1–14.
Arias CR, Burns JK, Friedrich LM, Goodrich RM and
Parish M. (2002). Yeast species associated with orange
juice: Evaluation of different identification methods.
Applied and Environmental Microbiology 68(4): 1955–1961.
Barata A, Malfeito-Ferreira M and Loureiro V. (2012). The
microbial ecology of wine grape berries. International
Journal of Food Microbiology 153(3): 243–259.
Barata A, Seborro F, Belloch C, Malfeito-Ferreira M and
Loureiro V. (2008). Ascomycetous yeast species recovered
from grapes damaged by honeydew and sour rot. Journal
of Applied Microbiology 104(4): 1182–1191.
Brugnoni LI, Cubitto MA and Lozano JE. (2012). Candida
krusei development on turbulent flow regimes: Biofilm for-
mation and efficiency of cleaning and disinfection pro-
gram. Journal of Food Engineering 111(4): 546–552.
Brugnoni LI, Lozano JE and Cubitto MA. (2007). Potential
of yeast isolated from apple juice to adhere to stainless
steel surfaces in the apple juice processing industry. Food
Research International 40(3): 332–340.
Casey GD and Dobson ADW. (2004). Potential of using real-
time PCR-based detection of spoilage yeast in fruit juice –
A preliminary study. International Journal of Food
Microbiology 91(3): 327–335.
Cloete KJ, Przybylowicz WJ, Mesjasz-Przybylowicz J,
Barnabas AD, Valentina AJ and Botha A. (2010).
Micro-particle-induced X-ray emission mapping of elem-
ental distribution in roots of a Mediterranean-type sclero-
phyll, Agathosma betulina (Berg.) Pillans, colonized by
Cryptococcus laurentii. Plant Cell Environment 33(6):
1005–1015.
Coton E, Coton M, Levert D, Casaregola S and Sohier D.
(2006). Yeast ecology in French cider and black olive nat-
ural fermentations. International Journal of Food
Microbiology 108(1): 130–135.
Davenport RR. (1996). Forensic microbiology for soft drinks
business. Soft Drinks Management International 1996:
34–35.
Davenport RR. (1997). Forensic microbiology II. Case book
investigations. Soft Drinks International 1997: 26–30.
Davenport RR. (1998). Microbiology of soft drinks.
In: Ashurst PR (ed.) Chemistry and Technology of Soft
Drinks and Fruit Juices. Sheffield: Sheffield Academic
Press, pp. 197–216.
Doyle ME. (2007). Microbial Food Spoilage: Losses and
Control Strategies. Madison: Food Research Institute,
University of Wisconsin-Madison. Available at: https://
pdfs.semanticscholar.org/ec5b/16cb60801c2f626d64df499
2aeee0e184342.pdf (accessed 3 November 2018).
Food and Beverages Manufacturing Sector Education and
Training Authority. Sector skills plan (SSP 2011–2016).
Available at: http://www.foodbev.co.za/wp-content/
uploads/2015/03/FoodBev-SSP-15-16-update-2011.pdf
(accessed 4 November 2018).
Harrison E, Muir A, Stratford M and Wheals A. (2011).
Species-specific PCR primers for the rapid identification
of yeasts of the genus. Zygosaccharomyces. FEMS Yeast
Research 11(4): 356–365.
Herna´ndez A, Pe´rez-Nevado F, Ruiz-Moyano S, Serradilla
MJ, Villalobos MC, Martı´n A, et al. (2018). Spoilage
yeasts: What are the sources of contamination of foods
and beverages? International Journal of Food
Microbiology 286: 98–110.
Jacques N and Casaregola S. (2008). Safety assessment of
dairy microorganisms: The hemiascomycetous yeasts.
International Journal of Food Microbiology 126(3):
321–326.
Jeyaram K, Singh WM, Capece A and Romano P. (2008).
Molecular identification of yeast species associated with
‘Hamei’ – A traditional starter used for rice wine produc-
tion in Manipur, India. International Journal of Food
Microbiology 124(2): 115–125.
Kesmen Z, O¨zbekar E and Bu¨yu¨kkiraz ME. (2018).
Multifragment melting analysis of yeast species isolated
from spoiled fruits. Journal of Applied Microbiology
124(2): 522–534.
Kurtzman CP and James SA. (2006). Zygosaccharomyces and
related genera. In: Blackburn CDW (ed.) Food Spoilage
Microorganisms. Boca Raton, FL: CRC Press,
pp. 289–305.
Loureiro V. (1994). Wine Spoilage Yeasts: A Problem to be
Solved. Paris: Assemble´ General de l’Office International
de la Vigne et du Vin.
Loureiro V. (2000). Spoilage yeasts in food and beverages:
Characterization and ecology for improved diagnosis
and control. Food Research International 33(3–4): 247–256.
Food Science and Technology International 25(8)
678
Maciel NOP, Pilo´ FB, Freitas LFD, Gomes FC, Johann S,
Nardi RM, et al. (2013). The diversity and antifungal sus-
ceptibility of the yeasts isolated from coconut water and
reconstituted fruit juices in Brazil. International Journal of
Food Microbiology 160(3): 201–205.
Martorell P, Ferna´ndez-Espinar MT and Querol A. (2005).
Molecular monitoring of spoilage yeasts during the pro-
duction of candied fruit nougats to determine food con-
tamination sources. International Journal of Food
Microbiology 101(3): 293–302.
Mayoral MB, Martı´n R, Sanz A, Herna´ndez PE, Gonza´lez I
and Garcı´a T. (2005). Detection of Kluyveromyces marx-
ianus and other spoilage yeasts in yoghurt using a PCR-
culture technique. International Journal of Food
Microbiology 105(1): 27–34.
Molna´rova´ J, Vadkertiova´ R and Stratilova´ E. (2014).
Extracellular enzymatic activities and physiological pro-
files of yeasts colonizing fruit trees. Journal of Basic
Microbiology 54(1): S74–S84.
Nosek J, Holesova Z, Kosa P, Gacser A and Tomaska L.
(2009). Biology and genetics of the pathogenic yeast
Candida parapsilosis. Current Genetics 55(5): 497–509.
Nyanga LK, Nout MJR, Smid EJ, Boekhout T and
Zwietering MH. (2013). Fermentation characteristics of
yeasts isolated from traditionally fermented masau
(Ziziphus mauritiana) fruits. International Journal of Food
Microbiology 166(3): 426–432.
Pham TA, Kawai S, Kono E and Murata K. (2011). The role
of cell wall revealed by the visualization of Saccharomyces
cerevisiae transformation. Current Microbiology 62(3):
956–961.
Rojo MC, Arroyo Lo´pez FN, Lerena MC, Mercado L,
Torres A and Combina M. (2014). Effects of pH and
sugar concentration in Zygosaccharomyces rouxii growth
and time for spoilage in concentrated grape juice at iso-
thermal and non-isothermal conditions. Food
Microbiology 38: 143–150.
Rojo MC, Torres Palazzolo C, Cuello R, Gonza´les M,
Guevara F, Ponsone ML, et al. (2017). Incidence of osmo-
philic yeasts and Zygosaccharomyces rouxii during the
production of concentrate grape juices. Food
Microbiology 64: 7–14.
Satora P, Droz_dz_ I, Duda-Chodak A and Skrzypiec K. (2013).
The use of PCR-RFLP technique for the analysis of yeast
contaminants in winemaking. Potravina´rstvo 7(Special
issue): 16–19.
Simo˜es J and Gomes AC. (2015). Isolation and selection of
conventional and non-conventional fermentative yeasts.
In: Ravishankar RV (ed.) Advances in Food
Biotechnology. West Sussex: John Wiley & Sons Ltd,
pp. 243–262.
Sla´vikova´ E, Vadkertiova´ R and Vra´nova´ D. (2007). Yeasts
colonizing the leaf surfaces. Journal of Basic Microbiology
47(4): 344–350.
Sousa MJ, Miranda L, Coˆrte-Real M and Lea˜o C. (1996).
Transport of acetic acid in Zygosaccharomyces bailii:
Effects of ethanol and their implications on the resistance
of the yeast to acidic environments. Applied and
Environmental Microbiology 62(9): 3152–3157.
Stratford M and James SA. (2003). Non-alcoholic beverages
and yeasts. In: Boekhout and Robert V (eds) Yeasts in
Food. Beneficial and Detrimental Aspects. Cambridge:
Woodhead Publishing Limited, pp. 309–345.
Stratford M, Steels H, Nebe-von-Caron G, Novodvorska M,
Hayer K and Archer DB. (2013). Extreme resistance to
weak-acid preservatives in the spoilage yeast
Zygosaccharomyces bailii. International Journal of Food
Microbiology 166(1): 126–134.
Sun Y and Liu Y. (2014). Investigating of yeasts species in
wine fermentation using terminal restriction fragment
length polymorphism method. Food Microbiology 38:
201–207.
Tekolo OM, McKenzie J, Botha A and Prior BA. (2010). The
osmotic stress tolerance of basidiomycetous yeasts. FEMS
Yeast Research 10(4): 482–491.
Ting J, Xu R and Xu J. (2018). Molecular identification and
distribution of yeasts in fruits. In: Sheika EL, Levin RE
and Xu J (eds) Molecular Techniques in Food Biology:
Safety, Biotechnology, Authenticity and Traceability.
Oxford: Wiley, pp. 117–144.
Tribst AAL, Sant’Ana AS and de Massaguer PR. (2009).
Review: Microbiological quality and safety of fruit juices
– Past, present and future perspectives. Critical Review in
Microbiology 35(4): 310–339.
Turtoi M. (2014). Inactivation of Saccharomyces cerevisiae
using new non-thermal technologies. Romanian
Biotechnological Letters 19(1): 8901–8909.
Wang H, Hu Z, Long F, Guo C, Niu C, Yuan Y, et al. (2016).
Combined effect of sugar content and pH on the growth of
a wild strain of Zygosaccharomyces rouxii and time for
spoilage in concentrated apple juice. Food Control 59:
298–305.
Wareing P and Davenport RR. (2005). Microbiology of soft
drinks and fruit juices. In: Ashurst PR (ed.) Chemistry and
Technology of Soft Drinks and Fruit Juices, 2nd ed.
Oxford: Blackwell Publishing Ltd., pp. 279–299.
Welthagen JJ and Viljoen BC. (1998). Yeast profile in Gouda
cheese during processing and ripening. International
Journal of Food Microbiology 41(3): 185–194.
White TJ, Bruns T, Lee S and Taylor J. (1990). Amplification
and direct sequencing of fungal ribosomal RNA genes for
phylogenetics. In: Innis MA, Gelfand DH, Sninsky JJ and
White TJ (eds) PCR Protocols: A Guide to Methods and
Applications. San Diego, CA: Academic Press,
pp. 315–322.
Yurkov AM, Kachalkin AV, Daniel HM, Groenewald M,
Libkind D, de Garcia V, et al. (2015). Two yeast species
Cystobasidium psychroaquaticum f.a. sp. nov. and
Cystobasidium rietchieii f.a. sp. nov. isolated from natural
environments, and the transfer of Rhodotorula minuta
clade members to the genus Cystobasidium. Antonie Van
Leeuwenhoek 107(1): 173–185.
Corbett and de Smidt
679