Accepted Manuscript Microbial communities involved in Kashar cheese ripening Oğuz Aydemir, Henning Harth, Stefan Weckx, Muhammet Dervişoğlu, Luc De Vuyst PII:
S0740-0020(14)00252-4
DOI:
10.1016/j.fm.2014.10.002
Reference:
YFMIC 2281
To appear in:
Food Microbiology
Received Date: 25 August 2014 Revised Date:
1 October 2014
Accepted Date: 7 October 2014
Please cite this article as: Aydemir, O., Harth, H., Weckx, S., Dervişoğlu, M., De Vuyst, L., Microbial communities involved in Kashar cheese ripening, Food Microbiology (2014), doi: 10.1016/ j.fm.2014.10.002. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT 1
Microbial communities involved in Kashar cheese ripening
2 3
Oğuz Aydemir a,*, Henning Harth b, Stefan Weckx b, Muhammet Dervişoğlu c, Luc De
4
Vuyst b
5 6
a
7
18100 Çankırı, Turkey
RI PT
Department of Food Engineering, Faculty of Engineering, Çankırı Karatekin University,
8 9
b
Research Group of Industrial Microbiology and Food Biotechnology (IMDO), Department
of Bioengineering Sciences, Faculty of Sciences and Bioengineering Sciences, Vrije
11
Universiteit Brussel, Pleinlaan 2, B-1050 Brussels, Belgium
SC
10
13
c
14
Samsun, Turkey
M AN U
12
Department of Food Engineering, Faculty of Engineering, Ondokuz Mayıs University, 55139
15 16
*Corresponding author:
18
Asst. Prof. Dr. Oğuz Aydemir, Department of Food Engineering, Faculty of Engineering,
19
Çankırı Karatekin University, 18100 Çankırı, Turkey
20
Tel: +90 376 2189500–8353. E–mail:
[email protected] 23 24 25 26 27 28 29 30 31 32 33
EP
22
AC C
21
TE D
17
ACCEPTED MANUSCRIPT 34
ABSTRACT
35
The microbiota of non–starter lactic acid bacteria (NSLAB) and their concomitant community
37
dynamics during cheese ripening were investigated for traditional Turkish Kashar cheeses
38
made from raw cows’ milk. Five batches of 15 Kashar cheeses produced in different dairy
39
plants located in Kars were analyzed during their whole ripening phase up to 180 days.
40
Lactobacilli and lactococci were determined as the prevailing microbial groups. The
41
molecular classification and identification of 594 LAB isolates during Kashar cheese ripening
42
were performed through (GTG)5–PCR fingerprinting of their genomic DNA followed by
43
verification of the (GTG)5–PCR clusters obtained after numerical analysis through 16S rRNA
44
gene sequencing of representative isolates. Lactobacillus casei (247 isolates, 41.6 %),
45
Lactobacillus plantarum (77 isolates, 13.0 %), and Pediococcus acidilactici (58 isolates, 9.8
46
%) were the prevailing NSLAB species in all Kashar cheeses of the different dairy plants
47
investigated throughout cheese ripening. The data of the present study contribute to the
48
inventory of unique cheese varieties to enable the prevention of losses of microbial
49
biodiversity and the selection of starter cultures for controlled cheese manufacturing.
M AN U
SC
RI PT
36
50
Keywords Kashar cheese, Cheese ripening, Non–starter lactic acid bacteria
52 53
55
1. Introduction
EP
54
TE D
51
Raw milk cheeses produced from unpasteurized milk and following traditional
57
manufacturing procedures consist of a very diverse and rich microbial ecosystem, of which
58
the composition highly influences the quality of the cheeses (McSweeney et al., 1993;
59
Grappin and Beuvier, 1997; De Angelis et al., 2001; Marino et al., 2003). The coagulation
60
temperature of the milk will facilitate the growth of most microorganisms; however, the
61
temperature applied during curd processing has the potential to inhibit the growth of some of
62
them (Steffen et al., 1993). While variations in manufacturing parameters such as cooking
63
temperature and curd handling play an important role in determining the characteristics of
64
cheeses, the cheese ripening microbiota plays a critical and crucial role in the development of
65
the unique characteristics of each cheese variety (Beresford et al., 2001). Non–starter lactic
66
acid bacteria (NSLAB) from the environment are a significant proportion of the microbial
AC C
56
ACCEPTED MANUSCRIPT communities of most ripened cheeses. They do not contribute to acid production during
68
cheese manufacture as the LAB starter culture does, but do have an impact on flavour
69
development in the ripening cheese (Beresford and Williams, 2004). The primary
70
environmental factors controlling the growth of microorganisms in the cheese matrix are
71
water and salt content, pH, presence of organic acids and nitrate, redox potential, and ripening
72
temperature (Beresford et al., 2001). Increasing information on the natural microbiota present
73
in cheeses can help to prevent a loss of microbial biodiversity and consequently a loss of a
74
wide range of cheeses produced by different methods, whose typical features depend on local
75
and regional traditions and on the indigenous microbial communities present in raw milk and
76
selected for by the cheesemaking environment (Fortina et al., 2003).
SC
RI PT
67
Kashar cheese, a pasta–filata cheese made from raw ewes’ or cows’ milk or a mixture
78
of them, is one of the most important Turkish cheese varieties and is similar to Caciocavallo,
79
Provolone, Regusono, Kashkaval, and Mozzarella (Aran, 1998; Çetinkaya and Soyutemiz,
80
2006). However, nowadays, it is generally made from raw cow’s milk (Kamber, 2008).
81
Kashar cheese is typically produced in the city of Kars situated in the North–East of Turkey.
82
Kashar cheeses have typically a cylindrical shape of around 30 cm in diameter and a height of
83
20 cm, with an average weight of 12 kg.
M AN U
77
In general, Kashar cheeses are made in a traditional way, although their ripening phase
85
is a laborious and expensive process. At farmhouse level, Kashar cheeses are produced from
86
raw milk without the addition of any starter culture. According to traditional Kashar cheese
87
manufacturing, the raw milk is filtered through a cloth and transferred into vats. Then, rennet
88
is added to the milk at 30–33 °C. Coagulation of the milk is achieved in about 45 min,
89
followed by cutting of the curd in small pieces (size of rice grains), which are poured into a
90
cheese cloth. The remaining whey is drained off by placing heavy blocks on top of it. After
91
pressing, the curd is cut into large parts and then grated. The grated curd is put into a metal
92
basket with holes and cooked at 70–80 °C in water with 5–8 Baume degrees for about 2–3
93
min. The cooked curd is then kneaded like bread dough. In some dairies, dry salting is
94
performed during kneading. Then, the curd is placed into round metal moulds for 12–24 h.
95
The cheeses are sprinkled with salt and placed on wooden shelves in a ripening room. The
96
cheeses are periodically turned and stacked. After a first ripening period of 3-4 weeks at room
97
temperature, 5–6 wheels of cheeses are stacked between round wooden plates and put into
98
sacks for a second ripening in a cold room (3–4 °C) for at least one month to maximally two
99
years.
AC C
EP
TE D
84
ACCEPTED MANUSCRIPT Although several studies have been performed on the characterisation of Kashar
101
cheese, there is very little information available on the actual composition of its microbial
102
ecosystem (Aran, 1998; Soyutemiz et al., 2000; Gülmez et al., 2004; Kamber, 2005;
103
Çetinkaya and Soyutemiz, 2006; Özdemir and Demirci, 2006; Fırat, 2006; Var et al., 2006;
104
Sert et al., 2007; Cıbık et al., 2010). No information at all is available on the molecular
105
identification of the LAB communities involved in Kashar cheese production and their
106
evolution during cheese ripening. Therefore, the aim of the present study was to characterise
107
the LAB composition of Kashar cheeses during cheese ripening.
109
2. Materials and methods
110 111
2.1. Cheese manufacturing and sampling
M AN U
112
SC
108
RI PT
100
A total of 75 Kashar cheeses were manufactured from raw cows’ milk in five selected
114
dairy plants (further referred to as A, B, C, D, and E) located in Kars. Each dairy plant
115
produced 15 independent Kashar cheeses, each one weighing ca. 3 kg, according to their own
116
traditional procedures (Table 1). None of the dairy plants performed pasteurization of the raw
117
milk neither applied a starter culture or used CaCl2. All Kashar cheeses produced were kept in
118
the same ripening room of dairy plant A (at room temperature and 60–70 % relative humidity)
119
for 20 d. After this first ripening period, the Kashar cheeses were wrapped into baking paper
120
and placed into a sack. Then, the cheeses were ripened in the same cold room of dairy plant A
121
(at 3–4 °C and 70–80 % relative humidity) for 160 d. Hence, the Kashar cheeses were ripened
122
for a total period of 180 d. Three Kashar cheeses of each 15–cheese batch were sampled at 5
123
time points, namely after 3, 30, 60, 120, and 180 d following manufacture. These cheese
124
samples were transported to the laboratory for microbiological analysis.
126 127
EP
AC C
125
TE D
113
2.2. Community dynamics
128
Ten–g cheese samples were homogenised in 90 mL of a 2 % (w/v) solution of sodium
129
citrate (Carlo Erba, Milano, Italy) for 1 min in a stomacher (AES Chemunex, Bruz, France).
130
Decimal dilutions of the suspensions were prepared with Ringer’s solution (Merck,
131
Darmstadt, Germany). Microbiological analyses were performed using the following agar
132
media and incubation conditions: yeast–glucose–chloramphenicol agar (YGC agar; Merck)
133
for yeasts and moulds at 30 °C for 72 h (Manolopoulou et al., 2003), kanamycin–aesculin–
ACCEPTED MANUSCRIPT azide agar (KAA agar; Merck) for enterococci at 30 °C for 24 h (Manolopoulou et al., 2003),
135
Rogosa agar (RO agar; Merck) adjusted to pH 5.5 with acetic acid for lactobacilli at 30 °C for
136
5 d under anaerobiosis (Ferrazza et al., 2004), M17 agar (Merck) for lactococci at 30 °C for
137
72 h (Zarate et al., 1997), de Man–Rogosa–Sharpe (MRS) agar (Merck) supplemented with
138
vancomycin (30 µg/mL; Sigma–Aldrich, Steinheim, Germany) for leuconostocs at 30 °C for
139
72 h (Manolopoulou et al., 2003), violet–red–bile agar (Merck) for coliforms at 37 °C for 24 h
140
(Gonzalez et al., 2003), chromocult–trytpone–bile–X–glucuronide agar (Merck) for
141
Escherichia coli at 44 °C for 24 h (Gross et al., 2007), and Baird–Parker agar (Merck) for
142
Staphylococcus aureus at 37 °C for 24 h, the presence of which was confirmed by a positive
143
coagulase test (Turkish Standards, 2001). All microbial counts were expressed as colony
144
forming units (cfu) per g of cheese sample.
146
2.3. Isolation of lactic acid bacteria
147
M AN U
145
SC
RI PT
134
Three or four colonies were randomly isolated from each countable plate of the M17
149
(presumptive lactococci), RO (presumptive lactobacilli) and MRS–vancomycin (presumptive
150
leuconostocs) agar media for each cheese sample. In total, 594 colonies were isolated from 75
151
cheese samples. The isolates were subcultured in MRS medium (lactobacilli and
152
leuconostocs) or M17 medium (lactococci), supplemented with 15 % (v/v) glycerol as a
153
cryoprotectant, and stored at –20 °C.
154
TE D
148
2.4. (GTG)5-PCR fingerprinting of genomic DNA for classification and identification of the
156
lactic acid bacteria isolates
158
Isolates were grown overnight in MRS medium (lactobacilli and leuconostocs) or M17
AC C
157
EP
155
159
medium (lactococci) and 2–mL aliquots of the cultures were centrifuged (8000 × g, 15 min, 4
160
°C). The cell pellets were subjected to genomic DNA extraction and purification using the
161
NucleoSpin 96 Tissue kit (Macherey Nagel GmbH, Düren, Germany), following the
162
manufacturer’s instructions, after first applying an enzymatic lysis of the cells using 20 µg/µL
163
lysozyme (Sigma–Aldrich) and 0.5 U/µL mutanolysin (Sigma–Aldrich) in TET buffer (20
164
mM Tris–HCl; 2 mM EDTA; 1 % Triton X–100; pH 8.0).
165
Classification and identification of the isolates was performed by (GTG)5–PCR
166
fingerprinting of their genomic DNA, numerical cluster analysis of these DNA fingerprints
167
using the BioNumerics 5.1 software (Applied Maths, Sint–Martens–Latem, Belgium), and
ACCEPTED MANUSCRIPT 168
verification of the identity of the (GTG)5–PCR clusters through 16S rRNA gene sequencing
169
of representative isolates, as described previously (Ravyts et al., 2008).
170 171
2.5. Statistical analysis
172
Statistical analysis of the data as to assess the effect of the ripening time and other
174
processing factors on the size and nature of the microbial communities was performed by
175
one–way and two–factor randomized complete block design, using SPSS 12.0 statistical
176
software (SPSS Inc., Chicago, IL, USA). The mean differences were analyzed using Duncan’s
177
multiple–range test.
SC
RI PT
173
178 179
3. Results
181
M AN U
180
3.1. Effect of the ripening time on the microbial community dynamics
182
Presumptive lactobacilli (MRS agar counts), lactococci (M17 agar counts),
184
leuconostocs (RO agar counts), enterococci (KAA agar counts), and yeasts and moulds (YGC
185
agar counts) were the different microbial groups present during the ripening of the Kashar
186
cheeses investigated (Table 2). Coliforms, E. coli, and Staph. aureus were not found in any of
187
the cheese samples analysed.
TE D
183
In dairy plant A cheeses, the lactobacilli counts of the samples ranged from 4.35 to
189
7.84 log cfu/g upon cheese ripening. At 30 d of ripening, there was a significant (p < 0.05)
190
increase of the counts of the lactobacilli. Afterwards, there was no significant difference in
191
their counts. The lactococci counts showed a slight increase on the 30th day of cheese
192
ripening (p > 0.05). Thereafter, their counts showed a decreasing trend. The lactobacilli
193
counts outnumbered the lactococci counts after 120 d. The counts of the leuconostocs
194
increased from 4.08 log cfu/g after 3 d to 8.09 log cfu/g after 60 d of cheese ripening (p
0.05) and remained constant
200
throughout the ripening phase.
AC C
EP
188
ACCEPTED MANUSCRIPT Concerning dairy plant B cheeses, the lactobacilli counts of the samples did not
202
change significantly during cheese ripening (p > 0.05). The lactococci counts ranged from
203
3.62 to 6.74 log cfu/g during cheese ripening. Only after 60 d of ripening, there was a
204
significant (p < 0.05) decrease in the counts of the lactococci. Afterwards, there was no
205
significant difference in their counts. The lactobacilli counts outnumbered the lactococci
206
counts at all time points sampled during cheese ripening. The counts of the leuconostocs
207
showed a very significant increase on the 30th day (p < 0.01). Thereafter, the leuconostoc
208
counts remained more or less constant. The yeast and mould counts showed an irregular
209
progress during cheese ripening and reached their highest counts after 30 d (p < 0.05).
210
However, a very significant decrease of the yeast and mould counts was found after 60 d of
211
ripening (p < 0.01). The enterococci counts decreased significantly after 30 d (p < 0.05) to
212
remain constant thereafter. Enterococci were not found in the cheese samples of the last two
213
sampled time points during the ripening phase.
M AN U
SC
RI PT
201
In dairy plant C cheeses, the lactobacilli counts ranged from 2.01 to 2.72 log cfu/g and
215
there was no significant change during cheese ripening (p > 0.05). The counts of the
216
lactococci exhibited a significant increase after 30 d of ripening (p < 0.05); afterwards, their
217
counts did not change significantly (p > 0.05). The counts of the lactococci were higher than
218
those of the lactobacilli at all time points sampled during the ripening phase. The leuconostoc
219
counts did not show a significant change during cheese ripening (p > 0.05). The yeast and
220
mould counts showed a slightly increasing trend during cheese ripening. The counts of the
221
yeast and moulds were significantly higher after 180 d of cheese ripening than those at the 3rd
222
and 60th day (p < 0.05). Enterococci, at a level of 1.03 log cfu/g, were only found after 3 d of
223
cheese ripening, to disappear upon.
EP
TE D
214
Concerning dairy plant D cheeses, the counts of the lactobacilli showed a significant
225
increase after 30 d of cheese ripening (p < 0.05), to remain more or less constant afterwards.
226
In contrast with the lactobacilli counts, the lactococci counts remained constant till the 120th
227
day of cheese ripening and decreased significantly (p < 0.05) at the end of the ripening phase.
228
The counts of the lactococci were higher than those of the lactobacilli till the 120th day, but
229
the lactobacilli counts outnumbered the lactococci counts after 180 d of ripening. The counts
230
of the leuconostocs increased from 7.05 to 8.55 log cfu/g after 30 d of cheese ripening (p
0.05). The counts of the
240
lactobacilli showed an increase till the 30th day of cheese ripening, to exhibit a decreasing
241
trend afterwards. The counts of the lactococci showed a decreasing trend till the 60th day of
242
cheese ripening and then remained constant. The counts of the lactococci were higher than
243
those of the lactobacilli at all time points sampled during the cheese ripening phase. The
244
counts of the leuconostocs ranged between 2.19 and 3.09 log cfu/g and there was no
245
significant change during cheese ripening (p > 0.05). The yeast and mould counts tended to
246
remain constant during cheese ripening. The enterococci counts decreased significantly after
247
30 d of cheese ripening (p < 0.05) and disappeared after 60 d.
M AN U
SC
RI PT
238
248 249
3.2. Molecular identification of the lactic acid bacteria isolates and community dynamics at
250
species level
251
A total of 594 strains isolated from MRS, M17 and RO agar media were subjected to a
253
molecular identification. The species and number of isolates identified are summarized in
254
Table 3 and the community dynamics of these species present during cheese ripening for each
255
dairy plant is represented in Fig. 1. Generally, the species variety was large on the 3rd day of
256
cheese ripening and decreased as long as ripening progressed.
EP
TE D
252
In dairy plant A cheeses, Lactobacillus fermentum was the prevailing species (20 %)
258
during the first cheese ripening period. In this period, other prevailing species were
259
Lactobacillus delbrueckii subsp. bulgaricus and Pediococcus acidilactici. On the 30th day,
260
Lactobacillus casei (58 %) and Lactobacillus plantarum (12 %) prevailed. Lactobacillus casei
261
and Staphylococcus saprophyticus subsp. saprophyticus were the species most frequently
262
isolated from cheese samples of 60 and 120 d of ripening. At the end of the cheese ripening
263
phase, Lb. casei (73 %) was the prevailing species. Hence, in cheeses from dairy plant A, Lb.
264
casei (48 %), Staph. saprophyticus subsp. saprophyticus (9 %), and Lb. plantarum (7 %) were
265
the predominant species throughout cheese ripening. Enterococcus faecium was found with
266
frequencies ranging from 5 to 7 % during the whole ripening phase. Only few isolates were
267
identified as Streptococcus thermophilus, Enterococcus durans, Lactobacillus paracasei,
268
Lactobacillus reuteri, Leuconostoc lactis, and Weissella halotolerans.
AC C
257
ACCEPTED MANUSCRIPT In dairy plant B cheeses, S. thermophilus (25 %) and P. acidilactici (21 %) were the
270
prevailing species on the 3rd day of cheese ripening. Staphylococcus saprophyticus subsp.
271
saprophyticus (38 %) and Lb. plantarum (25 %) were the prevailing species on the 30th day.
272
On the 60th day, Lb. casei (25 %), Lb. plantarum (25 %), and P. acidilactici (21%) prevailed.
273
Lactobacillus casei (33 %), Lb. plantarum (20 %), and Lactobacillus coryniformis subsp.
274
torquens (20 %) prevailed at 120 d of ripening. At the end of the ripening period, Lb. casei
275
(77 %) and Lb. plantarum (12 %) were the prevailing species. Hence, in dairy plant B
276
cheeses, the predominant species throughout cheese ripening were Lb. casei (32 %), Lb.
277
plantarum (19 %), and P. acidilactici (13 %). Only few isolates were identified as E. faecium,
278
E. durans, and W. halotolerans.
SC
RI PT
269
In dairy plant C cheeses, P. acidilactici (42 %), Lb. casei (19 %), and Lb. reuteri (15
280
%) were the prevailing species during the first cheese ripening period. On the 30th day of
281
cheese ripening, Lb. plantarum (36 %) and P. acidilactici (29 %) prevailed. After 30 d of
282
cheese ripening, Lb. casei prevailed. Hence, in cheeses from dairy plant C, Lb. casei (44 %),
283
P. acidilactici (17 %), Lb. plantarum (9 %), and Lb. reuteri (9 %) were the predominant
284
species throughout cheese ripening. Only few isolates were identified as Staph. saprophyticus
285
subsp. saprophyticus, S. thermophilus, Lb. coryniformis subsp. torquens, Leuc. lactis, and E.
286
durans.
TE D
M AN U
279
In dairy plant D cheeses, Lb. casei was prevailing during 120 d of cheese ripening.
288
Afterwards, Lb. plantarum (42 %) became prevailing in the cheese samples. Hence, in dairy
289
plant D cheeses, Lb. casei (53 %) and Lb. plantarum (18 %) were the predominant species
290
during cheese ripening. The presence of E. durans (18 %) during the first cheese ripening
291
period was remarkable. Moreover, among all dairy plants, E. durans and E. faecium were
292
present at the highest rate (10 %) in cheeses of dairy plant D during cheese ripening. Only few
293
isolates were identified as S. thermophilus, P. acidilactici, Lb. fermentum, Lb. reuteri, and
294
Leuc. lactis.
AC C
295
EP
287
In dairy plant E cheeses, P. acidilactici (39 %), Lb. casei (22 %), and E. durans (17
296
%) were the prevailing species during the first cheese ripening period. On the 30th and 60th
297
day, Lb. casei and Lb. plantarum prevailed. Pediococcus acidilactici (75 %) and Lb. casei (25
298
%) prevailed on the 120th day. At the end of the cheese ripening phase, Lb. coryniformis
299
subsp. torquens (45 %), Lb. casei (20 %), and S. thermophilus (20 %) became prevailing.
300
Hence, in the dairy plant E batch of cheeses, Lb. casei (25 %), P. acidilactici (15 %), and Lb.
301
coryniformis subsp. torquens (10 %) were the predominant species during cheese ripening.
ACCEPTED MANUSCRIPT 302
Only few isolates were identified as E. durans, E. faecium, Lb. sakei, Lb. paracasei, W.
303
halotolerans, Weissella paramesenteroides, and Leuconostoc mesenteroides.
304 305
3.3. Effect of the dairy plant and ripening time on Kashar cheese ripening
306
The dairy plant and cheese ripening time had a significant effect on the presumptive
308
lactobacilli, leuconostoc, lactococci, and enterococci counts (p < 0.01 in each case). The
309
lowest and highest lactobacilli counts were on the 3rd and 30th day of cheese ripening,
310
respectively. On the 30th day, an increase of the lactobacilli, leuconostoc, and lactococci
311
counts was found in all cheese samples, except for the lactococci counts of dairy plant E.
312
Afterwards, the lactobacilli counts tended to remain constant, while the lactococci counts of
313
the cheese samples decreased. On the 60th day of cheese ripening, small decreases in the
314
leuconostoc counts were found in all cheese samples, except in those from dairy plant A.
315
Thereafter, the leuconostoc counts of the cheese samples showed some decreases, except for
316
dairy plant C. The highest average enterococci counts of the cheese samples during Kashar
317
cheese ripening were on the 3rd day (p < 0.01). In Kashar cheeses of dairy plants B, C, and E,
318
the enterococci counts showed a decreasing trend throughout cheese ripening and disappeared
319
on the 120th, 30th, and 60th day of ripening, respectively.
TE D
M AN U
SC
RI PT
307
Kashar cheeses of dairy plants A and D displayed higher lactobacilli, leuconostoc,
321
lactococci, and enterococci counts than the other batches of cheeses during cheese ripening (p
322
< 0.05 in each case). There was no difference in lactococci counts of the Kashar cheeses
323
between dairy plants A and D in terms of mean values (p > 0.05), while the other cheese
324
batches were significantly different from these two batches and from each other (p < 0.01).
325
The lactococci counts of the Kashar cheese samples of the different dairy plants were in
326
increasing order C = E < B < A < D. Cheeses of dairy plant D showed the highest enterococci
327
counts throughout cheese ripening (P < 0.05). All these data indicate an influence of the lower
328
cooking temperature and time applied in the Kashar cheeses of dairy plants A and D.
AC C
329
EP
320
The Kashar cheeses of all dairy plants investigated displayed high yeast and mould
330
counts. During cheese ripening, there was no difference between the cheeses of dairy plants
331
A, B and C in terms of their mean values (p > 0.05). While the highest average values of yeast
332
and mould counts were found in cheese samples from dairy plant D (p < 0.05), cheese
333
samples from dairy plant E showed the lowest values during cheese ripening (p < 0.01). All
334
cheese samples showed increases of the yeast and mould counts on the 30th day, but they all
ACCEPTED MANUSCRIPT 335
showed decreases on the 60th day in the cold room. Afterwards, yeast and mould counts in all
336
cheese samples increased slightly upon further ripening in the cold room.
337 338
4. Discussion
339
Lactobacillus casei (41.6 %), Lb. plantarum (13.0 %), and P. acidilactici (9.8 %) were
341
the most isolated LAB species from all Kashar cheeses of the different dairy plants
342
investigated throughout cheese ripening. Whereas P. acidilactici was present at the beginning
343
of the cheese ripening phase, Lb. casei and Lb. plantarum were present after a first period of
344
cheese ripening. In each Kashar cheese, Lb. casei was most prevalent, with an increasing
345
trend in the cheeses of dairy plants A, B, and C upon cheese ripening. Also, Aran (1998)
346
found that Lb. casei subsp. casei (34.3 %) and Lb. plantarum (7.5 %) are the most frequently
347
isolated LAB species during Kashar cheese ripening, with an increasing proportion of the
348
former upon four months of cheese ripening. Lactobacillus casei and Lb. plantarum are
349
commonly present in cheeses, being predominant in for instance Cheddar cheese (Peterson
350
and Marshall, 1990; Jordan and Cogan, 1993). These mesophilic lactobacilli are often part of
351
the NSLAB secondary microbiota as are Lb. paracasei, Lb. curvatus, and pediococci
352
(Fitzsimons et al., 1999). Whereas Lb. coryniformis subsp. torquens was found during the late
353
stages of cheese ripening in the Kashar cheese samples of dairy plants B and E, Lb. reuteri
354
was found at the early stages of cheese ripening, especially in Kashar cheeses of dairy plant
355
C. Lactobacillus coryniformis subsp. torquens is sometimes found in dairy environments and
356
has been isolated from cheeses occasionally (Coppola et al., 2003; Martin et al., 2005; Dolci
357
et al., 2008). Lactobacillus reuteri has been isolated from cheeses seldomly (Dellaglio et al.,
358
1981). Pediococcus acidilactici was present in high numbers at early stages of the Kashar
359
cheese ripening, in particular in cheeses from dairy plants B, C, and E. This LAB species is
360
most frequently isolated from cheeses; also, pediococci are used as adjunct starter cultures in
361
Feta and Cheddar cheese productions (Beresford and Williams, 2004). Streptococcus
362
thermophilus was found in all cheese samples with similar frequencies (4.7–7.9 %)
363
throughout cheese ripening as a minor component of the Kashar cheese LAB microbiota. This
364
LAB species often accompanies mesophilic LAB species during cheese manufacturing
365
(Champagne et al., 2009). In the present study, Leuconostoc and Lactococcus species were
366
hardly found among the LAB species during Kashar cheese ripening. Leuconostoc lactis (1.2
367
%) was the predominant leuconostoc. Only one isolate of Lc. lactis subsp. cremoris in Kashar
368
cheeses of dairy plant A and two isolates of Lc. lactis subsp. lactis in cheeses of dairy plant E
AC C
EP
TE D
M AN U
SC
RI PT
340
ACCEPTED MANUSCRIPT were identified during cheese ripening. Aran (1998) found 4.8 % and 8.0 % of leuconostocs
370
among fresh and ripened Kashar cheese isolates, respectively. Further, they reported that Lc.
371
lactis subsp. lactis was important only in the curd to acidify the milk and is inactivated during
372
cheese texturing. Cıbık et al. (2010) identified 37 strains of Lc. lactis subsp. lactis and 5
373
strains of Lc. lactis subsp. cremoris isolated from white pickled and Kashar cheeses,
374
respectively. Remarkably, Staph. saprophyticus subsp. saprophyticus was found in cheeses
375
from dairy plants A, B, and C. This species has already been isolated from different cheeses
376
(Rea et al., 2007); it may originate from the udder skin (Braem et al., 2013). Staphylococcus
377
saprophyticus is known to cause urinary tract infections in women (Kuroda et al., 2005).
RI PT
369
The LAB microbiota of Kashar cheese slightly differed from other pasta‒filata
379
cheeses. For instance, in Caciocavallo cheese, Lb. brevis, Lb. coryneformis subsp.
380
coryneformis, Lb. fermentum, Lb. parabuchneri, Lb. paracasei subsp. paracasei, Lb.
381
pentosus, and/or Lb. plantarum have been determined as the predominant species (Corsetti et
382
al, 2001; Gobbetti et al. 2002; Coppola et al., 2003; Piraino et al., 2005). In Mozzarella
383
cheese, Lc. Lactis, Lactobacillus spp. such as Lb. fermentum and Lb. plantarum and S.
384
thermophilus were found as the prevailing species (Coppola et al., 2001; De Angelis et al.,
385
2006). In Provolone cheese, Lb. rhamnosus, S. macedonicus, and S. thermophilus were
386
predominant (Aponte et al., 2008). Cultivation-independent methods might be valuable to
387
complement the results of the present study.
TE D
M AN U
SC
378
The presumptive lactobacilli counts of the Kashar cheese samples of the present study
389
ranged from 1.69 to 8.74 log cfu/g and their average counts were 4.51 log cfu/g during cheese
390
ripening. Whereas Sert et al. (2007) determined 5.70 log cfu/g of lactobacilli in raw–milk
391
Kashar cheese and 5.59 log cfu/g of lactobacilli in Kashar cheese made with a starter culture,
392
Aran (1998) reported MRS agar counts from 7.24 to 8.11 log cfu/g during Kashar cheese
393
ripening. The present study unravelled a significant effect of the dairy plant (e.g., cooking
394
temperature and time, as shown for dairy plants A and D) and cheese ripening time on the
395
lactobacilli counts, which increased mainly after 30 d of ripening to remain constant upon.
396
Özdemir and Demirci (2006) noticed a similar trend for the lactobacilli counts of Kashar
397
cheeses. Sert et al. (2007) and Moatsou et al. (2001) determined an increase in lactobacilli
398
counts at the beginning of the ripening phase followed by a decrease afterwards in the Kashar
399
and Kasseri cheeses investigated, respectively. In general, mesophilic lactobacilli form a
400
significant portion of the microbiota of most cheese varieties during ripening (Beresford et al.,
401
2001). Despite the fact that mesophilic lactobacilli are inhabitants of raw milk and the dairy
402
environment, they are frequently overgrown by strong acidifiers of the genus Lactococcus.
AC C
EP
388
ACCEPTED MANUSCRIPT Alternatively, they do gain access during the cheesemaking process, so that they are often
404
found as a secondary microbiota during the ripening phase of different cheeses. This is
405
especially true for raw-milk cheeses, but mesophilic lactobacilli are also common in cheeses
406
manufactured with modern technologies (Wouters et al., 2002). The presumptive lactococci
407
counts of the Kashar cheese samples investigated varied between 2.27 and 9.24 log cfu/g and
408
their average counts were 5.71 log cfu/g during cheese ripening. Fırat (2006) reported an
409
average of 6.36 log cfu/g of lactococci in Kashar cheese made without a starter culture during
410
90 d of cheese ripening. Mesophilic lactococci are generally associated with the milk
411
environment (Wouters et al., 2002). The evolution of the lactococci counts in the cheese
412
samples of the present study is in agreement with the results of previous studies on Kashar
413
cheeses (Çetinkaya and Soyutemiz, 2006; Fırat, 2006; Sert et al., 2007). In several artisan
414
cheeses produced from raw milk, indigenous lactococci are found as the predominant LAB,
415
while in other cheeses, they are just dominant during the first period of the ripening phase
416
(Pogačić et al., 2011). In the present study, a significant effect of the dairy plant (e.g., cooking
417
temperature and time, as shown for dairy plants A and D) and cheese ripening time on the
418
lactococci counts was found. As leuconostocs were identified only sporadically in the Kashar
419
cheese samples investigated, the RO agar counts most probably represented lactobacilli as
420
well. The presumptive leuconostoc counts of the cheese samples ranged from 2.19 to 8.55 log
421
cfu/g and their average was 4.71 log cfu/g. As for lactococci, the natural habitat of
422
leuconostocs is plant material containing fermentable carbohydrates. They are introduced in
423
the dairy environment through the green pasture and via fodder fed to the cows (Vedamuthu,
424
2006).
EP
TE D
M AN U
SC
RI PT
403
Enterococcus faecium (3.9 %) and E. durans (3.4 %) were isolated from all cheese
426
samples during Kashar cheese ripening, except from cheeses of dairy plant C. Aran (1998)
427
found that E. faecium (28.4 %) is the second predominating LAB species in Kashar cheese
428
during cheese ripening. In general, enterococci are an important component of the microbiota
429
of cheeses produced in Italy, Spain, Portugal, Greece, Turkey, and Egypt. Enterococcus
430
faecalis, E. faecium and E. durans are the most frequently isolated enterococcal species, as
431
they may be present in raw milk (Beresford and Williams, 2004; Zamfir et al., 2006). High
432
enterococci counts in artisan cheeses are usually associated with poor hygienic practices
433
(Franz et al., 1999). Indeed, enterococci may get into milk either directly via human or animal
434
faeces or indirectly through contaminated water sources, milking equipment, or bulk storage
435
tanks (Giraffa, 2003). However, they may contribute to functionalities such as antibiosis and
436
flavour formation (Foulquié Moreno et al., 2006; De Vuyst et al., 2011). As they occur
AC C
425
ACCEPTED MANUSCRIPT 437
frequently in a wide range of dairy products, several strains show potential as functional
438
starter cultures for food fermentations based on their biochemical properties (Wouters et al.,
439
2002; Wessels et al., 2004; Foulquié Moreno et al., 2006; De Vuyst et al., 2011). The presumptive enterococci counts of the cheese samples ranged from