Dairy Sci. & Technol. (2012) 92:75–90
DOI 10.1007/s13594-011-0045-2
O R I G I N A L PA P E R
Microbial diversity of the traditional Iranian cheeses
Lighvan and Koozeh, as revealed by polyphasic
culturing and culture-independent approaches
Mohammad Reza Edalatian & Mohammad Bagher Habibi Najafi &
Seyed Ali Mortazavi & Ángel Alegría & Mohammad Reza Nassiri &
Mohammad Reza Bassami & Baltasar Mayo
Received: 1 April 2011 / Revised: 18 July 2011 / Accepted: 19 July 2011 /
Published online: 11 August 2011
# INRA and Springer Science+Business Media B.V. 2011
Abstract The microbiota of two traditional Iranian cheeses (Lighvan and Koozeh)
made of raw ewe’s milk or mixtures of ewe’s and goat’s milk without starter addition
was explored by culture-independent and culture-dependent approaches. Three
batches of Lighvan and one of Koozeh were subjected to culture-independent
polymerase chain reaction (PCR)–denaturing gradient gel electrophoresis (DGGE)
analysis and sequencing of dominant bands to assess the microbial structure and
dynamics through manufacturing and ripening. In addition, culturing in elective
media for lactic acid bacteria (M17, MRS and KAA), isolation of single colonies
(n=130), molecular identification by PCR-amplified ribosomal DNA restriction
analysis and sequencing, and differentiation at the strain level by repetitive
extragenic palindromic PCR was also performed. DGGE analysis showed that the
dominant amplicons in all four cheese batches belonged to Lactococcus lactis and
Streptococcus parauberis. In addition, Escherichia coli and Lactococcus garvieae
were frequently identified in both Lighvan and Koozeh, while Streptococcus
Electronic supplementary material The online version of this article (doi:10.1007/s13594-011-0045-2)
contains supplementary material, which is available to authorized users.
M. R. Edalatian : M. B. H. Najafi : S. A. Mortazavi
Department of Food Science and Technology, Ferdowsi University of Mashhad, Mashhad, Iran
M. R. Nassiri
Department of Animal Science, Faculty of Agriculture, Ferdowsi University of Mashhad,
Mashhad, Iran
M. R. Bassami
Department of Clinical Science, Faculty of Veterinary Medicine, Ferdowsi University of Mashhad,
Mashhad, Iran
M. R. Edalatian : Á. Alegría : B. Mayo (*)
Departamento de Microbiología y Bioquímica, Instituto de Productos Lácteos de Asturias, (CSIC),
Carretera de Infiesto, s/n, 33300 Villaviciosa, Asturias, Spain
e-mail: baltasar.mayo@ipla.csic.es
76
M.R. Edalatian et al.
thermophilus was found occasionally. In contrast, Enterococcus faecium and
Enterococcus faecalis were found to be dominant among the isolates in all batches.
These species showed a high genetic diversity. The discrepancy between culturing
and DGGE results suggested that dominant populations were in a nonrecoverable
state in the used media. This reinforces the idea that culture-dependent and cultureindependent techniques provide complementary data, ultimately affording a better
description of cheese ecosystems. These data could be of help in the selection of
commercial starters for industrial-scale manufacture of Lighvan and Koozeh cheeses
using pasteurised milk. Alternatively, microbial analysis would allow the selection of
appropriate strains for designing of specific starters for traditional cheese
manufacture.
伊朗传统Lighvan和Koozeh干酪的微生物多样性
摘要采用纯培养和非培养方法研究了由生鲜羊奶或者混合羊奶(羊奶和山羊奶)自然发酵制
作的伊朗传统Lighvan和Koozeh干酪的微生物区系。对3批Lighvan干酪和1批Koozeh干酪进
行了非培养的PCR-DGGE分析和主要条带的测序,以此评价干酪制作和成熟过程中微生物
区系的结构和动力学。此外,采用选择性培养基(M17, MRS and KAA)对乳酸菌进行了培养,采
用PCR-ARDRA、基因测序以及rep-PCR方法对分离出的单个菌落(n=130)从分子水平上进
行鉴定。DGGE分析结果表明,在所有4批干酪样品中优势菌群为Lactococcus lactis a和Streptococcus parauberis。此外,在Lighvan和 Koozeh干酪中Escherichia coli和Lactococcus garviea的
检出频率较高,但只在几个干酪样品中检测到Streptococcus thermophilus。相反,在所有样品中
Enterococcus faecium 和Enterococcus faecalis也是主导菌群。微生物菌群之间表现出较高的生物多
样性。纯培养和非培养的DGGE结果之间的差异表明主导菌群在这些培养基中是不可回收的,
这种结果说明在进行干酪微生物生态系统的研究中,纯培养和非培养方法获得的数据可以互
补。可以从获得的菌株中筛选出具有潜在工业化生产Lighvan和Koozeh干酪的发酵剂,也可以
从中筛选出特定的菌株作为发酵剂用于传统干酪的生产。
Keywords Traditional cheeses . Biodiversity, Lactic acid bacteria . Denaturing
Gradient gel electrophoresis . DGGE
关键词 传统干酪 . 生物多样性 . 乳酸菌 . 变性梯度凝胶电泳 . DGGE
1 Introduction
Lighvan and Koozeh are among the best known and most appreciated of all
traditional Iranian cheeses. They are manufactured in the neighbouring northwestern
provinces of East and West Azerbaijan, respectively, from raw ewe’s milk or a
mixture of ewe’s and goat’s milk following ancient cheesemaking technologies
without addition of starter. The main steps of their respective technologies are
presented in Fig. 1. The cheeses demonstrate typicity in taste and flavour and their
popularity is increasing at Iranian market.
Starter-free cheeses made from raw milk—such as Lighvan and Koozeh—rely for
their acidification and ripening on the action of their indigenous lactic acid bacteria
(LAB; Wouters et al. 2002). However, for such traditional cheeses to be competitive
in national and international markets, product standardisation is necessary and food
safety must be ensured. This requires the identification, characterisation and typing
of the key microorganisms which grow in them, and the selection of those
Microbial diversity of two Iranian cheeses
77
Lighvan cheese
Koozeh cheese
Mixtures of raw ewe’s or goat’s milk
Ewe’s or goat’s milk
Coagulation with homemade
lamb rennet at 28-32 C for 1 h
Coagulation with lamb or
commercial calf rennet at 33-34 C
for 45-60 min
Curd
Curd
Transferring to large, rectangular bags
and piling up for whey drainage at
room temperature
Mashing of the curd and putting on
a cloth; hanging it for whey
drainage (14-15 h)
Cut one-kilo cubes and putting
them in a 22%-salt brine for 6 h
Transferring of the curd to a large
cloth bag and adding dry salt onto
the surface
Turning the cubes upside down
for 9-15 times
Grinding the curd and
putting it into a pot
Keeping the cubes in a basin for
3-5 days for whey drainage
Ripening buried in the underground
at the shade for 2-3 months
Packing the cheese cubes in
metal tins with 10-12% salt brine
Ripening in deep-natural or manmade caves at 10-12 C for 3-4
months
Fig. 1 Diagram and flow chart of manufacturing and ripening stages of traditional, Iranian raw milk
cheeses Lighvan (a) and Koozeh (b)
appropriate for use as specific starters and adjunct cultures (Parente and Cogan
2004). The use of such starters would ensure that fermentations can be reliably
reproduced while preserving the typical, traditional bouquet of these cheeses
(Poznanski et al. 2004; Randazzo et al. 2008). The Lighvan and Koozeh ecosystems
78
M.R. Edalatian et al.
may also harbour LAB strains with unique flavour-forming capabilities that might be
advantageous in different areas of the dairy industry (Ayad et al. 2001) or in the
production of new, broad-range, natural antimicrobials (Ayad et al. 2002).
The microbiota of fermented foods can now be monitored by a variety of molecular
methods (Cocolin et al. 2007; Jany and Barbier 2008). Approaches combining classical
culturing followed by molecular identification of the isolates and application of
culture-independent molecular techniques are at present the most reliable choice to
obtain objective results about microbial structure and dynamics of any ecosystem
(Pogacic et al. 2010). The polymerase chain reaction (PCR) coupled with denaturing
gradient gel electrophoresis (DGGE)—PCR–DGGE—is one such culture-independent
molecular method that has been used to characterise the microbial populations of dairy
products (Cocolin et al. 2002; Ercolini et al. 2001; Flórez and Mayo 2006; Lafarge et
al. 2004). In fact, such characterisation has already been performed over the
manufacture and ripening of a number of traditional cheeses (Coppola et al. 2001;
Ercolini et al. 2004; Flórez and Mayo 2006; Pogacic et al. 2010; Randazzo et al. 2002).
Few microbial studies have been made on traditional Iranian cheeses. However,
pioneering work has identified the dominant LAB species in ripened Lighvan cheese
(Abdi et al. 2006; Barouei et al. 2008), as well as the cultivable lactobacilli that appear
during manufacture and ripening (Kafili et al. 2009). To our knowledge, studies on the
microbiota of Koozeh cheese have never been performed. In this study, PCR–DGGE
was used to type the major microbial populations during the manufacture and ripening
of three batches of Lighvan and one batch of Koozeh. In addition, a series of
cultivable LAB strains were isolated and subjected to molecular identification and
typing. The combined polyphasic approach allowed to assess the diversity and
dynamics of the dominant microbiota associated with these two cheeses.
2 Materials and methods
The general experimental outline applied for the microbiological characterisation of
Lighvan and Koozeh cheeses by culture-dependent and culture-independent
approaches is depicted in Fig. 2.
2.1 Sampling
Samples of raw milk, curd and cheeses (three batches of Lighvan made by different
producers and one batch of Koozeh) at different ripening periods (3, 7, 15, 30, 60
and 90 days) were sampled and transferred to the laboratory under refrigeration.
Culturing analyses were performed within 6 h after arrival. Samples were then
frozen at −20 °C until they were used to isolate total microbial DNA.
2.2 DGGE analysis of Lighvan and Koozeh cheeses
2.2.1 Extraction of total microbial DNA
Frozen samples were thawed at 4 °C and homogenised in 2% sodium citrate.
Homogenates were then used for the isolation of total microbial DNA employing the
Microbial diversity of two Iranian cheeses
Fig. 2 Diagram of the experimental scheme applied for the
microbiological characterisation
of Lighvan and Koozeh
cheeses
79
LIGHVAN, KOOZEH
Culturing
Dilutions
Plating:
M17, MRS, KAA,
MRS-vancomycin
PCR-DGGE
Isolation of
total DNA
PCR
of 16S rDNA
DGGE
PCR of
16S rDNA
from colonies
ARDRA
Sequencing
Isolation of
DNA from bands
Sequencing
Identification
by sequence
comparison
Identification
by sequence
comparison
rep-PCR
typing
QIAamp DNA Stool Mini Kit (Qiagen, Hilden, Germany) following the manufacturer’s instructions. DNA quantity and quality was measured by absorption at 260
and 280 nm using a Nanodrop spectrophotometer (Thermo Scientific, Wilmington,
DE, USA).
2.2.2 PCR amplification of 16S rRNA sequences
One hundred nanogram of purified DNA was used as a template in PCR
amplifications of the V3 region of the bacterial 16S rRNA gene using the universal
primers F357-GC clamp (5′-TACGGGAGGCAGCAG-3′, to which a 39 bp GC
sequence was linked) and R518 (5′-ATTACCGCGGCTGCTGG-3′), as reported by
80
M.R. Edalatian et al.
Muyzer et al. (1993). Amplification of the D1 domain of the 26S rRNA gene of fungi
was accomplished using primers NL1-GC (5′-GCCATATCAATAAGCGGAGGAAAG3′, with a 39 bp GC clamp) and LS2 (5′-ATTCCCAAACAACTCGACTC-3′), as
reported by Cocolin et al. (2002). PCR was performed in 50 μL reaction volumes
containing 10 mmol.L−1 Tris-HCl, 50 mmol.L−1 KCl, 1.5 mmol.L−1 MgCl2,
0.2 mmol.L−1 of each dNTP, 0.2 mmol.L−1 of the primers, and 5 U of Taq polymerase.
2.2.3 Electrophoretic conditions
DGGE was performed in a DCode apparatus (Bio-Rad) in 8% polyacylamide gels at
60 °C. After initial trials, the best separation of PCR amplicons was observed at
denaturing gradients of 40–60% for bacteria and 30–50% for fungi. Electrophoresis
was performed at 75 V for 16 h and 130 V for 4.5 h for bacterial and fungal
amplifications, respectively. Bands were visualised under UV light after staining
with ethidium bromide (0.5 μg.mL−1) and photographed.
2.2.4 Identification of DGGE bands
Bands in the polyacrylamide gels were assigned to species by comparison with a control
ladder of known strains (Flórez and Mayo 2006), namely Lactococcus garvieae CECT
4531T, Lactobacillus plantarum CECT 748T, Leuconostoc mesenteroides CECT 219T,
Streptococcus parauberis DSMZ 6631T, Enterococcus faecium ATCC 19343T,
Enterococcus faecalis CECT 481T, Lactococcus lactis subsp. cremoris MG 1363,
Escherichia coli W3110, and Lactobacillus paracasei CECT 4022T. To construct the
ladder, purified DNA of the control strains was used in independent PCR–DGGE
reactions and equal amounts of amplicons (50 ng.mL−1 each) were mixed before
electrophoresis. DNA from bands that did not migrate to the positions of the controls
was isolated by elution, re-amplified with the same primer pair without the GC clamp,
sequenced by cycle extension in an ABI 373 DNA sequencer (Applied Biosystems,
Foster City, CA, USA), and the sequences compared with sequences in the GenBank
database using the Basic Local Alignment Search Tool programme (BLAST 2011) and
with those held by the Ribosomal Database Project (RDP 2011). Sequences showing
97% similarity or higher were deemed to belong to the same species (Palys et al. 1997;
Stackebrandt and Goebel 1994).
2.3 Culture-dependent approach
Milk samples were diluted in 0.1% sterile peptone water. Twenty five grams of curd
and cheese samples at day 30 (fresh cheese) and at day 90 (ripened cheese) were
homogenised in 225 ml of a sterile sodium citrate solution (2% w/v) using a
Stomacher 400 (Seward, Worthing, UK). Dilutions of milk, curd and cheese samples
were then plated in duplicate on agarified plates of M17 (Scharlab, Barcelona,
Spain), MRS (Merck, Darmstad, Germany), and KAA (Oxoid, Basingstoke–
Hampshire, UK) for enumeration and isolation of presumptive lactococci,
lactobacilli and enterococci, respectively. Plates were incubated aerobically in a
GasPack EZ system (BD, Franklin Lakes, NJ, USA) at either 30 °C (M17), 37 °C
Microbial diversity of two Iranian cheeses
81
(MRS) or 42 °C (KAA) for 24–72 h. Then, four to five representative colonies
(according to shape, size and colour) from the highest dilutions were picked up at
random, purified two or three times on the same media, and examined for Gram
staining, catalase production and morphology. All Gram-positive, catalase-negative
isolates were selected for further identification and stored frozen at −80 °C in MRS
broth containing 20% glycerol.
2.4 Molecular identification of LAB species
Cryopreserved cultures were recovered in the corresponding media of isolation and
identified by a polyphasic molecular procedure, including extraction of total DNA,
partial amplification of 16S rRNA genes, amplified ribosomal DNA restriction
analysis (ARDRA), sequencing and sequence comparison.
2.4.1 DNA extraction
For DNA isolation, single colonies were suspended in 50 μL of molecular grade
water (Sigma-Aldrich, St. Louis, MO, USA), heated at 98 °C for 10 min in a
thermocycler (Bio-Rad, Richmond, CA, USA), and centrifuged for 5 min at
16,000 rpm. Cell-free extracts were used as DNA templates for amplification of
a major part of the 16S rRNA gene by PCR. Isolates that did not produce
amplicon by this method were grown overnight in liquid media, centrifuged,
washed in sterile saline (0.9%) and suspended in the same volume of water. Cell
extracts were then obtained with glass beads (105 μm of diameter) in a Minibead
Beater apparatus (Biospec Products, Bartlesville, OK, USA) and centrifuged as
above.
2.4.2 Amplification of 16S rRNA genes
The primers used for the amplification of the 16S rRNA genes were 27FYM (5′AGAGTTTGATYMTGGCTCAG-3′) and 1492R (5′-GGTTACCTTGTTACGACTT3′), based on the conserved regions of the 16S rRNA gene. As reported above, PCR
was performed in 50 μL reaction containing 2 μL of each the cell-free extracts.
2.4.3 Amplified ribosomal DNA restriction analysis
Amplicons were subjected to ARDRA by digestion with the restriction enzymes
HaeIII and HhaI (Invitrogen Ltd., Paisley, UK). DNA fragments were separated by
electrophoresis in 1.5% agarose gels at 75 V for 90 min. Gels were then stained and
visualised as above.
2.4.4 Sequencing and sequence comparison
Representative amplicons of the different ARDRA profiles were sequenced using the
primer 27FYM. On average, 850 bp were obtained per sequence and the sequences
were compared as above.
82
M.R. Edalatian et al.
2.5 Typing of isolates
All isolates were grouped by repetitive extragenic palindromic PCR (rep-PCR)
typing using primer BoxA2R (5′-ACGTGGTTTGAAGAGATTTTCG-3′) and
employing the amplification conditions of Koeuth et al. (1995). PCR products were
then electrophoresed and visualised as above. Pattern similarity was expressed via
the simple matching (SM) coefficient, and patterns were clustered by the unweighted
pair group method using arithmetic averages (UPGMA).
3 Results
3.1 Microbial dynamics of Lighvan and Koozeh cheeses by DGGE
Three batches of Lighvan cheese manufactured by independent producers, plus
one batch of Koozeh cheese were analysed by PCR–DGGE using universal
primers to track both bacterial and eukaryotic populations (Figs. 3 and 4,
respectively; Online Resource Table 1). The number of bacterial DGGE bands
obtained for the different samples ranged between four (milk for Lighvan batch 2;
Fig. 3b, line 1) and 11 (60 day-old Lighvan batch 1; Fig. 3a, line 7); all were
identified at the species level. This large number of bacterial bands contrasts with
the small number of eukaryotic populations; only two bands were observed in a
60-day-old sample of Lighvan (data not shown), and between two and four bands
in samples of Koozeh over the entire experimental period (Fig. 4b). Eukaryotic
bands were mostly identified at the genus level. In total, 25 bands (20 bacterial and
five eukaryotic) were readily identified.
Similarities and differences in the DGGE profiles were noted between the two
cheeses and between the distinct batches of Lighvan. The predominant bands for all
cheeses over manufacture and ripening corresponded to the species S. parauberis
(band d) and L. lactis (band g). In Lighvan, a band of variable intensity
B
A
a
b
c
d
Ma 1
b
2
3
4
a
1
5
6
7
Mb
e
e
a
b
c
f
d
2
d
g
h
h
2
3
4
1
a
5
4
g
3
3
C
Ma 1
d
g
3
6
7
Mb
e
a
b
c
f
d
Ma 1
5
2
3
4
b
h
6
4
6
Mb
e
e
d
f
g
3
7
a
1
g
h
5
f
g
h
h
i
i
i
i
i
i
i
Fig. 3 Bacterial dynamics as shown by the DGGE profiles of the V3 variable region of the bacterial 16S rRNA
gene in three independent batches of Lighvan cheese (a, b, and c, respectively) throughout manufacturing and
ripening. Samples: milk, curd, and cheeses at days 3, 7, 15, 30, and 60 after manufacture. Ma and Mb,
DGGE markers used as a control and composed of amplicons of isolated strains, as follows: a L. garvieae, b
L. plantarum, c L. mesenteroides, d S. parauberis, e E. faecium, f E. faecalis, g L. lactis, h E. coli, and i L.
paracasei. Bands identified by sequencing are coded with a letter if corresponding to species on the markers
and with a number for other species, as follows: 1 L. raffinolactis, 2 Lactococcus plantarum, 3 S.
thermophilus (two bands), 4 S. haemolyticus, 5 L. salivarius, 6 M. caseolyticus
Microbial diversity of two Iranian cheeses
83
B
A
Ma
a
1
2
4
5
a
c
1
c
2
d
3
3
1
Mb
2
4
5
e
4
f
d
3
5
g
g
9
8
8
7
6
h
i
10
10
Fig. 4 Microbial dynamics in Koozeh cheese during manufacturing and ripening as judged from DGGE
profiles of the V3 variable region of the bacterial 16S rRNA gene (a) and the D1 domain of the 26S rRNA
gene of fungi (b) samples: cheese at 3, 7, 15, 30 and 60 days after manufacture. Ma and Mb as in Fig. 1
(except for absence of the L. plantarum amplicon; band b). Sequenced bands are denoted by a letter code
if corresponding to species on the markers or by a number if they were identified by reamplification,
sequencing and sequence comparison, as follows: 1 L. raffinolactis, 2 S. uberis, 3 C. maltomaricum, 4 L.
curvatus, 5 C. diazotrophica, 6 S. thermophilus; 7 V. tapetis, 8 Warcupia spp., 9 D. hansenii (two bands)
and 10 Penicillium spp. (two bands)
corresponding to L. garvieae (band a) was observed in most cheese samples, as
was a band for E. coli (band h). Lactococcus raffinolactis (band 1) and E. faecium
(band e) were also present in two batches of Lighvan at most times. A band with a
sequence matching that of Streptococcus thermophilus (band 3) was observed in
batch 1 of Lighvan cheese at all sampling times and in the milk and curd samples
respectively of the other two batches (Fig. 3). Bands identified occasionally
included L. plantarum and L. paracasei in batch 1 (Online Resource Table 1),
Staphylococcus haemolyticus in the milk samples of batches 2 and 3, Lactobacillus
salivarius in the milk samples of batch 3, and Macrococcus caseolyticus in all
samples of batch 3 cheese. DGGE profiles within a single Lighvan cheese batch
were almost identical over ripening. Greater changes were observed, however, for
the DGGE profiles of the Koozeh samples (Fig. 4). The DGGE profiles of the latter
cheese were dominated by S. parauberis (band d), followed by L. lactis which was
present in all samples (band g). However, L. raffinolactis was present only in 3and 7-day-old cheese samples. Streptococcus uberis (band 2) was identified in
days 3, 7, and 15 samples, Carnobacterium maltomaricum (band 3) was identified
in day 15, Lactobacillus curvatus (band 4) in days 30 and 60 samples, and S.
thermophilus (band 6) was seen in all samples except that for day 15. Surprisingly,
two bands present in most of the samples were identified as Celerinatantimonas
diazaotrophica (band 5) and Vibrio tapetis (band 7); microorganisms not usually
found in cheese. Bands corresponding to mould and yeast populations were
obtained in three of the five Koozeh samples. One of the eukaryotic bands was
84
M.R. Edalatian et al.
related to the ascomycete Warcupia spp. (band 8; Fig. 4) and two bands each were
identified as belonging to Debaryomyces hansenii (band 9) and Penicillium spp.
(band 10; Fig. 4).
3.2 Identification and typing of LAB species from Lighvan and Koozeh by culturing
Counts (in colony forming units; cfu) on the different media and sampling points
ranged widely. In general, counts in M17 and MRS were shown to be very similar
and usually one logarithmic unit higher than those in KAA. The highest counts were
recorded in curd samples (7.58±0.15 cfu.g−1 in M17, 7.91±0.01 cfu.g−1 in MRS and
6.44±0.04 cfu.g−1 in KAA), decreasing one or two logarithmic units from that point
onwards. As none of these three media is selective, isolated colonies were purified
and subjected to a polyphasic molecular identification scheme. In total, 130 isolates
(82 from Lighvan and 48 from Koozeh) were identified from the counting plates of
M17 (47 isolates), MRS (59 isolates) and KKA (24 isolates) (Table 1). Regardless of
Table 1 Species and numbers of majority cultured microorganisms identified through manufacturing and
ripening of the traditional Iranian cheeses Lighvan and Koozeh
Species
Stage of manufacture
Milk
Curd
Medium of isolation
(number of isolates)
Total
30 days 90 days
Lighvan cheese
L. lactis subsp. lactis
3
1
–
–
M17 (4)
4
L. plantarum
3
8
9
–
MRS (20)
20
L. brevis
2
–
–
3
MRS (5)
5
E. faecium
3
11
11
13
M17 (18), MRS (15) KAA (5)
38
11
E. faecalis
–
2
–
9
KAA (7), M17 (4)
Enterococcus durans
1
–
–
–
MRS (1)
1
Enterococcus casseliflavus –
–
1
–
M17 (1)
1
Entererococcus italicus
1
–
–
–
M17 (1)
1
Micrococcus luteus
–
–
–
1
MRS (1)
1
Total Lighvan
13
22
21
26
M17 (28), MRS (42) KAA (12)
82
E. faecium
7
29
M17 (13), MRS (13), KAA (10) 36
E. faecalis
–
5
M17 (4), MRS (1)
Koozeh cheese
5
Enterococcus durans
–
1
M17 (1)
1
Enterococcus casseliflavus
–
2
KAA (2)
2
L. plantarum
1
–
MRS (1)
1
L. brevis
1
–
MRS (1)
1
S. haemolyticus
–
1
MRS (1)
1
Aerococcus viridans
–
1
M17 (1)
1
9
39
M17 (19), MRS (17) KAA (12)
48
30
65
M17 (47), MRS (59) KAA (24)
130
Total Koozeh
Total
13
22
Microbial diversity of two Iranian cheeses
85
the enzyme used (HaeIII or HhaI), the isolates were grouped into 11 different
ARDRA patterns (Online Resource Fig. 1). ARDRA profiles were compared with in
silico analysis of several LAB species and with the profiles of the type strains used
to construct the DGGE ladder. In addition, to unequivocally identify patterns at the
species level, representative amplicons were sequenced and the sequences compared
with those in the GenBank and RDP databases. Table 1 shows the molecular
identification results. Nine different bacterial species were identified in Lighvan and
eight in Koozeh. Enterococcus spp., including E. faecium, E. faecalis, E. durans and
other species, made up the major part of the cultures. Of these, E. faecium (74
isolates) was dominant in both Lighvan and Koozeh at all sampling points, followed
by E. faecalis (16 isolates). Enterococci species were isolated from all counting
plates, indicating they constitute the dominant cultivable population (Table 1).
Although in small numbers, L. plantarum (21 isolates) and Lactobacillus brevis (six
isolates) were also recovered from both cheeses. In contrast to enterococci,
lactobacilli were only identified from MRS agar plates (Table 1), suggesting they
are subdominant populations in both Lighvan and Koozeh. In fact, counts recorded
in MRS were usually slightly lower than those in M17. All isolates from the KKA
medium were identified as Enterococcus spp. These results differed with those
obtained in DGGE analysis since the species producing the most intense bands were
almost absent in the cultures (only four L. lactis isolates from Lighvan cheese milk
and curd were obtained, while S. parauberis isolates were never recovered).
Enterococcus spp. and L. plantarum isolates were all subjected to rep-PCR typing
to evaluate intra-species diversity. Figure 5 shows the profiles obtained with the 38
E. faecium isolates from Lighvan cheese, plus the similarity dendrogram for the
different typing patterns clustered by the UPGMA method and using the SM
coefficient. Given the reproducibility of the assay (around 90%; Online Resource
Fig. 2), isolates sharing a percentage of similarity of >88% (an arbitrary figure) were
considered to be the same strain (Fig. 5). Twenty-four different profiles were
considered to represent different strains. Twenty-four different stains were also found
among the 36 E. faecium isolates from Koozeh cheese (Online Resource Fig. 3). A
wide genetic diversity was further detected among the L. plantarum, E. faecalis and
E. casseliflavus isolates (data not shown).
4 Discussion
Few microbial studies of Lighvan cheese have been undertaken (Abdi et al. 2006;
Kafili et al. 2009), and to our knowledge this is the first microbiological description
of Koozeh cheese. A similar number of species was detected by molecular
identification of the isolates and by the PCR–DGGE approach. Eight and nine
different species were identified among the cultured isolates from Koozeh and
Lighvan respectively, and four through 11 bands corresponding to an equal number
of species were obtained in DGGE analyses. However, the different methods showed
discrepancies in terms of the microbial populations identified; they therefore
provided complementary results (El-Baradei et al. 2007; Flórez and Mayo 2006;
Poznanski et al. 2004; Randazzo et al. 2002) allowing a better description of these
cheese ecosystems.
86
M.R. Edalatian et al.
M 1 2
3
4
5 6 7
8
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 M
5.0
3.0
2.0
1.5
0.7
0.6
0.5
0.3
0.1
UPGMA
0.52
0.6
0.68
0.76
37
16
13
23
22
33
15
11
12
31
18
3
17
14
7
6
21
20
24
36
35
34
32
30
19
8
26
29
9
10
4
25
28
5
27
2
38
1
0.84
0.92
L77-R
L54-F
L39-F
L75-R
L73-R
L59-R
L50-F
L36-F
L37-F
L44-F
L41-F
L13-M
L58-R
L40-F
L20-C
L19-C
L68-R
L66-R
L76-R
L74-R
L67-R
L61-R
L46-F
L43-F
L62-R
L23-C
L15-C
L32-C
L31-C
L33-C
L17-C
L78-R
L21-C
L18-C
L16-C
L9-M
L42-F
L1-M
1.0
Simple Matching Coefficient
Fig. 5 Typing REP-PCR profiles obtained with primer BoxA2R among the 38 E. faecium isolates from
Lighvan cheese. Below, dendogram of similarity of the different typing patterns clustered by the UPGMA
method using the Simple Matching coefficient. M molecular weight marker GeneRulerTM (Fermentas, St.
Leon-Rot, Germany). The broken line denotes the arbitrary percentage of similarity (88%) used to
consider isolates as different strain; this percentage of similarity was lower than the assay reproducibility
(90%; Supplementary Figure 2)
The bacterial and fungal population dynamics recorded by DGGE for the two
cheeses were similar to those reported for other traditional cheeses (Flórez and Mayo
2006; Pogacic et al. 2010; Randazzo et al. 2006). The bacterial diversity of dominant
populations was usually greater in milk and curd samples, as some species were
never identified in cheese (Table 1). Two to four high intensity bands were observed
for the different batches of the two cheeses at all sampling times, which were
accompanied by up to nine bands of lower intensity. The intensity of an individual
DGGE band is assumed to be a semiquantitative measure of the corresponding
microbe’s abundance in the sample (Muyzer et al. 1993). In Lighvan and Koozeh,
DGGE identified the dominant populations as belonging to S. parauberis and L.
lactis species, with contributions at certain sampling times from S. thermophilus in
both cheeses and L. curvatus in Koozeh. Celerinatantimonas diazotrophica, L.
garvieae, L. raffinolactis and E. coli were found in most batches of the cheeses in
subdominant numbers. Although Enterococcus spp. bands were found in several
batches, high intensity bands of enterococcal species were only encountered in
Lighvan cheese batch 1 (Fig. 3a). Thus, it was surprising to discover that species of
this genus accounted for a large proportion of the cultured microorganisms (73.8%
of the isolates; Table 1). These results, however, are not unusual, as DGGE bands
corresponding to enterococcal species have been reported in some cheese types
Microbial diversity of two Iranian cheeses
87
(Coppola et al. 2001; Ercolini et al. 2003; Ogier et al. 2004) and not in others
(Ercolini et al. 2004; Ogier et al. 2004; Randazzo et al. 2002). As enterococci were
identified among isolates from the highest cultivable dilutions in all media, we must
assume that the dominant species detected by PCR–DGGE are in a nonrecoverable
state in the counting media utilised (although preferential lysis and/or DNA
amplification cannot be excluded (Cocolin et al. 2007)). Thought nonselective,
M17, MRS and KAA may be a highly stressful media, impeding growth of a part of
the population, as has been reported elsewhere for other cheese systems when
comparing count in M17 and MRS as compared to count on PCA (Alegría et al.
2009; Flórez and Mayo 2006). The paradigm in this study is KAA, which
consistently showed counts one or two logarithmic units lower than those in the
other two media, while enterococci strains were equally isolated from all three.
Of note in the sampled cheeses is the presence of bands related to S. thermophilus in
most batches; these Iranian cheeses might therefore be considered a good source of
new strains of this important cheese starter. DGGE bands of S. thermophilus have also
recently been reported in a traditional Spanish starter-free cheese made from raw cow’s
milk (Alegría et al. 2009). The identified strains of S. thermophilus in the latter cheese
have been isolated and are currently being characterised and compared with industrial
starter strains (unpublished). Similarly, L. garvieae, a lactic acid bacterium similar to L.
lactis (Fernández et al. 2010), has been identified by culturing and molecular methods
in other cheese types (Alegría et al. 2009; Flórez and Mayo 2006).
The fact that cultivable strains belonging to the populations of the most prominent
bands (S. parauberis and L. lactis) were not readily recovered on the enumeration
plates strongly suggests that these microbial species are in a noncultivable state in
Lighvan and Koozeh. Similar results have been reported elsewhere for other
traditional cheeses (Ercolini et al. 2004; Randazzo et al. 2002). However, the DGGE
results indicate that these two populations reach high densities at the beginning of
the manufacturing process; certainly, they are present in the milk. S. parauberis has
been associated with subclinical and clinical mastitis (Pitkälä et al. 2008); therefore,
its presence in cheese is not desirable. In contrast, L. lactis is the typical LAB
species of cheese. It enjoys ‘generally regarded as safe’ status and its enzymes
contribute towards the typical taste and aroma profiles of cheese (Mayo 2010;
Parente and Cogan 2004). Specific starters for these Iranian cheeses should therefore
include strains of this species. Isolation of such L. lactis strains would have to be
undertaken at the beginning of cheese manufacture (curdling of the milk,
acidification, whey drainage, etc.).
Enterococci have been repeatedly reported to constitute major populations in
artisanal, traditional cheeses made from raw milk (for a review, see Giraffa 2003). The
high enterococci counts in Lighvan and Koozeh agree well with the low pH of these
cheeses (average, 4.64) and the high concentration of salt present during ripening (up
to 5.27% at day 90). Under such harsh conditions, Enterococcus spp. may thrive better
than other LAB species. Though the presence of high numbers of enterococci in foods
is controversial (Ogier and Serror 2008), strains of some species have been proposed
as starters or adjunct cultures for several cheese types (Giraffa 2003).
A high intraspecies diversity was found among the enterococci isolates using repPCR, which suggests a high subsequent phenotypic diversity (Fig. 5). Moreover,
several strains were shown to produce bacteriocins, such as enterocin A, B, P and X,
88
M.R. Edalatian et al.
or a combination of these (unpublished). Certainly, enterococci strains play pivotal
roles in the production of aroma compounds (El-Baradei et al. 2007; Foulquié
Moreno et al. 2006; Giraffa 2003). However, for their safe use in foods, candidate
strains would have to be subjected to a complete characterisation, guaranteeing the
absence of recognised virulence factors and atypical, potentially transferable
antibiotic resistances (Foulquié Moreno et al. 2006; Ogier and Serror 2008).
C. diazotrophica and V. tapetis are both recently reported marine bacteria, which
may well come from the salt added to the cheeses. The occasional development of
these pathogenic microorganisms and the recurrent presence of opportunistic
populations (i.e., Enterococcus spp. and S. parauberis) argue for a need of
improvement of the safety conditions of these two traditional cheeses.
5 Conclusions
This work contributes to the microbiological characterisation of the most important
traditional Iranian cheeses, Lighvan and Koozeh. The fact that some microbial
populations were detected by one identification method only stresses the importance
of combined approaches for fully describing the microbiota of naturally fermented
cheeses. Microbial data could either help in the selection of appropriate commercial
starters for industrial scale manufacture or in designing specific starters for
traditional cheese manufacture. The results may further provide the basis for the
future award of protected designation of origin status or an equivalent quality label
for these cheeses.
Acknowledgements This research was partially supported by a project from the Spanish Ministry of
Science and Innovation (MICINN) to BM (Ref. AGL2007-61869-ALI). AA was awarded a scholarship of
the Severo Ochoa programme from FICYT (Ref. BP08-053). The authors wish to thank the Iranian
Ministry of Industries and Mines, as well as Razavi Dairy Industry (Mashhad, Iran) and the Office of
Industrial Relationships (OIR) of Ferdowsi University of Mashhad (FUM).
References
Abdi R, Shikh-Zeinoddin M, Soleimanian-Zad S (2006) Identification of lactic acid bacteria isolated from
traditional Iranian Lighvan cheese. Pak J Biol Sci 9:99–103
Alegría A, Álvarez-Martín P, Sacristán N, Fernández E, Delgado S, Mayo B (2009) Diversity and
evolution of the microbial populations during manufacture and ripening of Casín, a traditional
Spanish, starter-free cheese made from cow’s milk. Int J Food Microbiol 136:44–51
Ayad EHE, Verheul A, Engels WJ, Wouters JT, Smit G (2001) Enhanced flavour formation by combination
of selected lactococci from industrial and artisanal origin with focus on completion of a metabolic
pathway. J Appl Microbiol 90:59–67
Ayad EHE, Verheul A, Wouters JTM, Smit G (2002) Antimicrobial-producing wild lactococci isolated
from artisanal and non-dairy origins. Int Dairy J 12:145–150
Barouei J, Karbassi A, Ghoddusi HB, Mortazavi A (2008) Lactic microflora present in Liqvan ewe’s milk
cheese. Int J Food Properties 11:407–414
BLAST (2011) Basic Local Alignment Search Tool, National Center for Biotechnology Information,
National Library of Medicine (NLM), National Institutes of Health (NIH), USA. http://www.ncbi.nlm.
nih.gov/BLAST/. Accessed Feb 2011
Cocolin L, Aggio D, Manzano M, Cantoni C, Comi G (2002) An application of PCR-DGGE analysis to
profile the yeast populations in raw milk. Int Dairy J 12:407–411
Microbial diversity of two Iranian cheeses
89
Cocolin L, Diez A, Urso R, Ratnsiou K, Comi G, Bergmaier I, Beimfohr C (2007) Optimization of
conditions for profiling bacterial populations in food by culture-independent methods. Int J Food
Microbiol 120:100–109
Coppola S, Blaiotta G, Ercolini E, Moschetti G (2001) Molecular evaluation of microbial diversity in
different types of Mozzarella cheese. J Appl Microbiol 90:414–420
El-Baradei G, Delacroix-Buchet A, Ogier JC (2007) Biodiversity of bacterial ecosystems in traditional
Egyptian Domiati cheese. Appl Environ Microbiol 73:1248–1255
Ercolini D, Moschetti G, Blaiotta G, Coppola S (2001) The potential of a polyphasic PCR-DGGE
approach in evaluating microbial diversity of natural whey cultures for water-buffalo Mozzarella
cheese production: bias of culture-dependent and culture-independent analyses. Syst Appl Microbiol
24:610–617
Ercolini D, Hill PJ, Dood CER (2003) Bacterial community structure and location in Stilton cheese. Appl
Environ Microbiol 69:3540–3548
Ercolini D, Mauriello G, Blaiotta G, Moschetti G, Coppola S (2004) PCR-DGGE fingerprints of microbial
succession during a manufacture of traditional water buffalo mozzarella cheese. J Appl Microbiol
96:263–270
Fernández E, Alegría A, Delgado S, Mayo B (2010) Phenotypic, genetic and technological
characterization of Lactococcus garvieae strains isolated from a raw milk cheese. Int Dairy J
20:142–148
Flórez AB, Mayo B (2006) Microbial diversity and succession during the manufacture and ripening of
traditional, Spanish, blue-veined Cabrales cheese, as determined by PCR-DGGE. Int J Food
Microbiol 110:165–171
Foulquié Moreno MR, Sarantinopoulos P, Tsakalidou E, de Vuyst L (2006) The role and application of
enterococci in food and health. Int J Food Microbiol 106:1–24
Giraffa G (2003) Functionality of enterococci in dairy products. Int J Food Microbiol 88:215–222
Jany J-L, Barbier G (2008) Culture-independent methods for identifying microbial communities in cheese.
Food Microbiol 25:839–848
Kafili T, Razavi SH, Emam Djomeh Z, Naghavi MR, Álvarez-Martín P, Mayo B (2009) Microbial
characterization of Iranian traditional Lighvan cheese over manufacturing and ripening via culturing
and PCR-DGGE analysis: identification and typing of dominant lactobacilli. Eur Food Res Technol
229:83–92
Koeuth T, Versalovic J, Lupski JR (1995) Differential subsequence conservation of interspersed repetitive
Streptococcus pneumoniae BOX elements in diverse bacteria. Genome Res 5:408–418
Lafarge V, Ogier J-C, Girard V, Maladen V, Leveau J-Y, Gruss A, Delacroix-Buchet A (2004) Raw
cow milk bacterial population shifts attributable to refrigeration. Appl Environ Microbiol
70:5644–5650
Mayo B (2010) Dairy lactococci. In: Heldman RD, Bridges A, Hoover D, Wheeler M (eds) Encyclopedia
of biotechnology in agriculture and food. Taylor & Francis Group, London. DOI: 10.1081/E-EBAF120044696
Muyzer G, de Waal EC, Uitterlinden AG (1993) Profiling of complex microbial populations by denaturing
gradient gel electrophoresis analysis of polymerase chain reaction-amplified genes encoding for 16S
rRNA. Appl Environ Microbiol 59:695–700
Ogier JC, Serror P (2008) Safety assessment of dairy microorganisms: the Enterococcus genus. Int J Food
Microbiol 126:291–301
Ogier J-C, Lafarge V, Girard V, Rault A, Maladen V, Gruss A, Leveau J-Y, Delacroix-Buchet A (2004)
Molecular fingerprint of dairy microbial ecosystems by use of temporal temperature denaturing
gradient gel electrophoresis. Appl Environ Microbiol 70:5628–5643
Palys T, Nakamura LK, Cohan FM (1997) Discovery and classification of ecological diversity in the
bacterial world: the role of DNA sequence data. Int J Syst Bacteriol 47:1145–1156
Parente E, Cogan TM (2004) Starter cultures: general aspects. In: Fox PO (ed) Cheese: chemistry, physics
and microbiology, 3rd edn. Elsevier, Oxford, pp 123–147
Pitkälä A, Koort J, Björkroth J (2008) Identification and antimicrobial resistance of Streptococcus uberis
and Streptococcus parauberis isolated from bovine milk samples. J Dairy Sci 91:4075–4081
Pogacic T, Samarjiza D, Coich V, D’Andrea M, Kagli DM, Giacomini A, Majhenic AC, Rogelj I (2010)
Microbiota of Karakacanski skakutanak, an artisanal fresh sheep cheese studied by cultureindependent PCR-ARDRA and PCR-DGGE. Dairy Sci Technol 90:461–468
Poznanski E, Cavazza A, Cappa F, Cocconcelli PS (2004) Indigenous raw milk microbiota influences
the bacterial development in traditional cheese from an alpine natural park. Int J Food Microbiol
92:141–151
90
M.R. Edalatian et al.
Randazzo CL, Torriani S, Akkermans ALD, de Vos WM, Vaughan EE (2002) Diversity, dynamics, and
activity of bacterial communities during production of an artisanal Sicilian cheese as evaluated by 16S
rRNA analysis. Appl Environ Microbiol 68:1882–1892
Randazzo CL, Vaughan EE, Caggia C (2006) Artisanal and experimental Pecorino Siciliano cheese:
microbial dynamics during manufacture by culturing and PCR-DGGE analyses. Int J Food Microbiol
109:1–8
Randazzo CL, Pitino I, De Luca S, Scifò GO, Caggia C (2008) Effect of wild strains used as starter
cultures and adjunct cultures on the volatile compounds of the Pecorino Siciliano cheese. Int J Food
Microbiol 122:269–278
RDP (2011) Ribosomal Database Project, Center for Microbial Ecology, Michigan State University, USA
http://rdp.cme.msu.edu/index.jsp. Accessed Feb 2011
Stackebrandt E, Goebel BM (1994) Taxonomic note: a place for DNA-DNA reassociation and 16S rRNA
sequence analysis in the present species definition in bacteriology. Int J Syst Bacteriol 44:846–849
Wouters JTM, Ayad EHE, Hugenholtz J, Smit G (2002) Microbes from raw milk for fermented dairy
products. Int Dairy J 12:91–109