Effects of pH and sugar concentration on Zygosaccharomyces rouxii growth and
time for spoilage of concentrated grape juice at isothermal and non-isothermal
conditions
M.C. Rojo1,3, F.N. Arroyo López2, M.C. Lerena1,3, L. Mercado3, A. Torres1,4
& M. Combina1,4*
1
Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET). Av.
Rivadavia 1917 (C1033AAJ). Ciudad Autónoma de Buenos Aires.
2
Food Biotechnology Department. Instituto de la Grasa (CSIC). Av. Padre García
Tejero 4. 41012, Seville, Spain.
3
Oenological Research Center. Estación Experimental Agropecuaria Mendoza. Instituto
Nacional de Tecnología Agropecuaria (EEA Mza INTA). San Martín 3853 (5507)
Luján de Cuyo, Mendoza, Argentine.
4
Microbiology and Inmunology Department. Facultad de Ciencias Exactas, Físico-
Químicas y Naturales. Universidad Nacional de Río Cuarto. Ruta Nacional 36, Km 601.
Running title: Effects of pH and sugar on Zygossacharomyces
*Corresponding author: Mariana Combina, Ph.D. Estación Experimental
Agropecuaria Mendoza. Instituto Nacional de Tecnología Agropecuaria (EEA Mza
INTA). San Martín 3853 (5507) Luján de Cuyo, Mendoza, Argentine. Tel: +54 261
4963020 ext. 295. e-mail: mcombina@mendoza.inta.gov.ar
1
1
Abstract
2
In the present survey, we assess the effects of pH (1.7 – 3.2) and sugar concentration
3
(64-68 ºBrix,) on the growth parameters of Zygosaccharomyces rouxii MC9 using
4
response surface methodology. Experiments were carried out in concentrated grape
5
juice inoculated with this spoilage yeast at isothermal conditions (23ºC) for 60 days. pH
6
was clearly the variable with the highest effect for both growth parameters (potential
7
maximum growth rate and lag phase duration). although the effects of sugar
8
concentration were also retained. In a second step, the time to produce spoilage by this
9
microorganism in concentrated grape juice was also evaluated at isothermal (23ºC) and
10
non-isothermal conditions, trying to reproduce storage and overseas shipping
11
temperature conditions respectively. Results show how pH was again the environmental
12
factor with the highest effect to delay the alteration of the product. Thereby, a pH value
13
below 2.0 was enough to increase the shelf life of the product for more than 60 days in
14
both constant and variable temperatures. The information obtained in the present work
15
could be very useful for producers and buyers to predict the growth and time for
16
spoilage of this yeast in concentrated grape juice.
17
18
Keywords: Response surface methodology; Zygosaccharomyces rouxii; concentrated
19
grape juice; pH; sugar concentration; spoilage.
2
20
1. Introduction
21
Grape juice and by-products represent an important part of the food industry in the
22
world. Argentina grape production is mainly devoted to the industry, where wine and
23
concentrated grape juices are the two mayor types of commercial products. Mendoza
24
and San Juan provinces are the main manufacturers of concentrated grape juices in the
25
country; 75% of their production is strictly wholesale to markets as United States,
26
Japan, Russia and México (Bruzone, 1998; INV, 2013). Concentrated grape juices have
27
a great importance as additive in several massive consume products. Due to their natural
28
qualities, it is employed to elaborate baby foods, pharmaceutical products, foods and
29
drinks (Bruzone, 1998). Concentrated juices are microbiologically more stable than
30
other fruit products and usually are stored at room temperature without any additional
31
treatment (ICMSF, 1980; Splittstoesser, 1987). However, these products are not free of
32
microbiological spoilage problems. The combination of high concentration of sugar and
33
low pH support the development of a reduced number of microorganism species.
34
Osmophilic yeasts represent the primary spoilage cause in high sugar food and drink
35
industries, where the genus Zygosaccharomyces is the most frequent described spoilage
36
microorganism (ICMSF, 1980; Deák and Beuchat, 1993; Worobo and Splittstoesser,
37
2005; Martorell et al., 2007).
38
The genus Zygosaccharomyces has a long history of spoilage in the food industry.
39
Three Zygosaccharomyces species, Z. bailii, Z. bisporous, and Z. rouxii, have been
40
associated with the spoilage of grape must, concentrated grape juice and wine
41
(Fugelsang and Edwards, 2007; Deák, 2008; Loureiro and Malfeito-Ferreira, 2003).
42
Spoilage by Zygosaccharomyces species can be categorized into two groups: i) visible
43
growth on the surface of the product, and ii) fermentative spoilage manifested by
3
44
alcoholic, esteric or other types of odours and/or visible evidence of gas production,
45
leading to bubbling of the product and/or expansion of flexible packaging (Legan et al.,
46
1991; Smith et al., 2004).
47
In a previous study, the osmotolerant and osmophilic yeast population in concentrated
48
grape juice from Argentina was characterized, being Z. rouxii the only yeast species
49
isolated from spoiled products. Moreover, in other samples without visible evidence of
50
spoilage, Z. rouxii was also isolated at high frequencies representing 76% of the total
51
yeast population (Combina et al., 2008). The unusual physiological characteristics of Z.
52
rouxii are largely responsible for their ability to cause spoilage, including resistance to
53
weak-acid preservatives, extreme osmotolerance, ability to adapt to high glucose
54
concentrations and high temperatures, ability to vigorously ferment glucose, and growth
55
at low pH values (Emmerich and Radler, 1983; James and Stratford, 2003; Martorell et
56
al., 2007).
57
Few studies have been carried out determining the effect of limiting factors on Z. rouxii
58
growth. All of them were done in culture media and some of them assessed each
59
variable in an independent way (Kalathenos et al., 1995; Praphailong and Fleet, 1997;
60
Membré et al., 1999). On the contrary, response surface (RS) methodology is a very
61
useful tool which has been previously applied to estimate the combined effects of
62
environmental variables on yeast growth (D’Amato et al., 2006; Arroyo-López et al.,
63
2006). This methodology has widely been used in predictive microbiology as a
64
secondary polynomial model to predict the microorganism response as a function of
65
environmental changes determining as the same time the interaction among them
66
(McMeekin et al., 1993).
4
67
In this work, we assess the combined effect of the two limiting factors (pH and sugar
68
concentration) on the growth parameters of a native strain of Z. rouxii (MC9) previously
69
isolated from spoiled concentrated grape juice. This task was accomplished using RS
70
methodology as secondary model. With the intention of results would be useful for
71
producers and buyers, the time for spoilage (TFS) was also determined in natural
72
substrate under storage (isothermal) and shipping (non-isothermal) temperatures into a
73
range of conditions usually found in Argentinean concentrated grape juices.
74
2. Material and methods
75
2.1. Yeast strain
76
The strain Z. rouxii MC9, previously isolated from spoiled concentrated grape juices,
77
was used in the present study. Strain was molecularly identified by sequencing of the
78
D1/D2 domain of 26S ribosomal gene and registered at the Oenological Research
79
Centre Microorganism Collection from INTA, Argentina (GenBank Accession Number
80
KF002711 ). This strain was selected from a previous study among several native Z.
81
rouxii strains because of its better adaptation to concentrated grape juice and fast growth
82
(data not shown).
83
2.2. Media and growth conditions
84
Z. rouxii MC9 was previously grown on YPD broth (40 g/l glucose, 5 g/l bacteriological
85
peptone, 5 g/l yeast extract, 20 g/l agar) during 1 day at 28ºC. Before inoculation in
86
natural substrate (concentrated grape juice), this strain was adapted to osmotic shock by
87
growing in a medium (MYGF) with an intermediate concentration of sugar (195 g/l
88
glucose, 195 g/l fructose, 20 g/l malt extract, 5 g/l yeast extract) with pH adjusted to 4.5
5
89
units by the addition of citric acid. This last medium was incubated during 3 days at
90
28ºC without shaking to reach the highest possible population (107 CFU/ml) just at the
91
end of exponential growth phase. Experiments were finally performed in concentrated
92
grape juice provided by a local company located in Mendoza region (Argentina).
93
2.3. Experimental design
94
The different runs (a total of 66) were carried out in 1 liter of concentrated grape juice
95
placed in sterile bag-in-box with Vitop® valve. Bags were placed in metal containers to
96
reproduce the conditions of storage (isothermal) and overseas shipping (non-isothermal)
97
and monitored for 60 days. The experimental design was obtained from the combination
98
of two variables (pH and sugar concentration) with 3 levels for each variable (Table 1).
99
Variables levels were established into a range of conditions usually found in
100
Argentinean concentrated grape juices. The pH values were 1.7, 2.5 and 3.2 and the
101
concentration of sugar (expressed as ºBrix, which is a unit more frequently used by
102
producers and buyers) were 64 (779 g/l reducing sugar; aw: 0.778±0.003), 66 (810 g/l;
103
aw: 0.767±0.003) and 68 º Brix (842 g/l; aw: 0.744±0.003). Two intermediate pH values
104
(1.9 and 2.1) were also evaluated in the sugar concentration condition more frequently
105
required by the market (68º Brix), making a total of 11 different treatments run by
106
triplicate. To achieve the different pH values, grape juices were passed through ion
107
exchange column to obtain the desired pH prior to concentration in order to reproduce
108
industrial conditions. The ºBrix of the concentrated grape juices were confirmed using a
109
digital hand-held refractometer (Atago PAL-2, Japan) and the pH was determined using
110
a digital pH meter (ALTRONIX, United States). Each treatment was inoculated with 1.2
111
x 102 CFU/ml of the strain Z. rouxii MC9, which represent the maximum limit
112
recommended for fungi and yeasts count by the buyers. Un-inoculated bags for each
6
113
experimental series were also introduced as negative control. The assays were
114
conducted at two different temperature systems, a constant temperature (23±0.5°C), to
115
simulate the most frequent condition of product storage (isothermal condition), and non-
116
isothermal (variable) conditions, trying to reproduce the overseas shipping temperature.
117
This last temperature profile was designed with the data recorder during wine shipping
118
to destinations in the Northern Hemisphere (Leinberger, 2006; Hartley, 2008). Internal
119
and external temperatures were monitored by iButton® temperature data logger placed
120
inside the bag and outside of the metal containers in the control treatments. The
121
recorded non-isothermal profile is shown in Figure 1, which has three different regions.
122
The first section of the graph shows the container on the dockside in Santiago (Chile)
123
harbour in summer. The middle section shows a hypothetical sea journey from Santiago
124
(Chile) to the Asian continent (eg. Russia or Japan). The final section of the graph again
125
shows the containers on the dockside in destination harbour in winter. Simulated
126
shipping time was 45 days, after that, the treatments were maintained at room
127
temperature (23ºC) until completion of the test (60 days).
128
2.4. Modeling yeast growth and time for spoilage
129
Concentrated grape juice samples at isothermal (23ºC) and non-isothermal conditions
130
were aseptically taken every 2 days to follow Z. rouxii MC9 growth. Samples were
131
decimal diluted in 30% (w/v) glucose to prevent osmotic shock and allow the recovery
132
of sublethally injured cells. Dilutions were then spread in two culture media. Selective
133
high sugar media: MY50G (aw 0.89) (Beuchat, 1993) and TGY media (Beuchat et al.,
134
2001) were chosen to detect Z. rouxii present in the concentrated samples. Plates were
135
incubated during 3 to 5 days at 28 °C before counting. Growth parameters (µmax,
136
potential maximum growth rate; λ, lag phase duration) were calculated from each
7
137
treatment contemplated in the experimental design by directly fitting plate count (log10
138
CFU/mL) versus time (days) using the Baranyi and Roberts model (1994). For this
139
purpose, DMfit Software 2.1 was used.
140
In addition, bags at constant and variable temperatures were also daily examined for
141
signs of fermentative activity such as gas production, bubbling and expansion of
142
flexible packaging, and the time required for spoilage (TFS, days) recorded. This
143
alteration was correlated with a Zygosaccharomyces population level in the
144
concentrated grape juice always higher than 106 CFU/ml (data not shown).
145
In the secondary modeling step, growth parameters (µmax and λ) and TFS of Z. rouxii
146
MC9 in both temperature systems (IC, isothermal conditions; NIC, non-isothermal
147
conditions) were adjusted to a RS equation of the general form:
148
Y=ßo+ß1X1+ß11X12+ß2X2+ß22X22+ß12X1X2+.
(1)
149
where Y is the parameter modelled (µmax, λ, TFS-IC or TFS-NIC), ßo is the
150
mean/intercept term, ßi are the coefficients to be estimated during the RS fitting (ß1 is
151
the coefficient for the linear effect of X1, ß11 for the quadratic effect of X1, ß12 for the
152
interaction between variables X1 and X2, and so on),  is the term for error and X1 and
153
X2 are the environmental variables under study (pH and sugar concentration (ºBrix),
154
respectively). For the main effects, regression coefficients can be interpreted as the
155
increase or decrease (depending of the positive or negative coefficient sign) in the
156
response when the factor changes one unit. Analysis of the RS was made using the
157
Experimental Design module of the Statistica 7.0 software package, using the pure
158
error, derived from repetitions of experiments, as option in the corresponding
8
159
ANOVAs. Model performance was also checked by the lack of fit test and the
160
determination coefficient R2 (percentage of variability in the response that can be
161
explained by the model).
162
3. Results
163
The combined effect of pH and sugar concentration on the growth parameters of the
164
spoilage yeast Z. rouxii MC9 was evaluated in concentrated grape juice at isothermal
165
temperature. Moreover, the TFS of concentrated grape juice inoculated with this
166
microorganism was modeled at isothermal and non-isothermal temperatures, obtaining
167
representative and valuable data for industry of the shelf life of the product.
168
3.1. Modeling yeast growth at isothermal conditions
169
Experiments were carried out under storage temperature at 23ºC (isothermal), which
170
allowed to built the growth curves and calculate the growth parameters for the different
171
conditions assayed. Table 1 shows the µmax and λ values obtained for the different
172
conditions included in the experimental design. As can be easily deduce from Table 1, λ
173
and µmax at 23 ºC changed as a function of the pH and sugar concentration modified.
174
Thereby, µmax ranged from 0.000 (experiments with pH 1.7) to 0.530 (log10 CFU/mL
175
days-1) (pH 2.5 and 64º Brix), while λ ranged from 0.000 (experiments with the highest
176
pH at 3.2) to >60 days (experiments with the lowest pH value 1.7).
177
Table 2 shows the results obtained from the ANOVA analysis of regression for the
178
growth parameters µmax and λ of Z. rouxii MC9 as a function of environmental factors.
179
The variance in the response that was explained by both models was high: 90.2% for
180
µmax and 92.8% for λ. In this way, both models can be considered appropriated to
9
181
describe the effects of environmental variables (pH and sugar concentration) on yeast
182
growth.
183
The RS equations that predict the response of both growth parameters as a function of
184
environmental variables can be easily deduced from Table 2, by replacing the
185
appropriate coefficients in the general equation 1. These polynomial equations in
186
physical terms can be used by the industry to estimate the behaviour of Z. rouxii as a
187
function of diverse combinations of pH and sugar concentration within the studied
188
experimental range (interpolation region), which was established into a range of values
189
with application for producers and buyers. Figures 2a and 3a show the graphical
190
representation of these equations for µmax and λ, respectively. Both RSs show a clear
191
curvature effect along the pH axis. When the pH decreased, the µmax value decreased
192
(positive correlation) while λ increased (negative correlation), which is indicative of a
193
clear inhibitory effect of this factor producing a delay of the yeast growth. Figures 2b
194
and 3b show the Pareto chart for the standardized effect of environmental factors for
195
µmax and λ, respectively. Except the quadratic effect of ºBrix for µmax, all regression
196
coefficients were retained at a p-value <0.05. As can be clearly deduced, the highest
197
effects for both growth parameters were observed for pH (linear and quadratic effect in
198
this order), followed by the interaction ºBrix*pH and the linear effect of ºBrix (in the
199
case of µmax) and the linear effect of ºBrix, interaction ºBrix*pH and the quadratic effect
200
of ºBrix (in the case of λ). Probably, the lower effect of sugar concentration compared to
201
pH was due to the good adaptation of this yeast to high osmotic pressures.
202
3.2. Modeling time for spoilage at isothermal and non-isothermal conditions
10
203
TFS of Z. rouxii MC9 in concentrated grape juice was modeled at both isothermal (IC,
204
23ºC) and non-isothermal (NIC, variables) conditions. Table 3 shows the TFS values
205
obtained for the different conditions assayed in the experimental design. As can be
206
easily deduce from this table, TFS changed as a function of the pH and sugar
207
concentration evaluated. Thereby, TFS-IC (23ºC) ranged from 13.66 (experiments with
208
pH 3.2 and 64ºBrix) to >60 days (experiments with pH 1.7), while TFS-NIC (variables
209
temperatures) ranged from 12.66 (pH 3.2 and 64ºBrix) to >60 days (again runs with the
210
lowest pH value). In general, the treatments performed at non-isothermal conditions
211
showed a slight lower microbial stability than those made at isothermal conditions (see
212
Table 3). In the most growth favourable pH value (3.2), the increase of sugar
213
concentration from 64 to 68 ºBrix doubled microbial stability of the product for both
214
temperature systems.
215
Table 4 shows the results obtained from the ANOVA analysis of regression for the
216
TFS-IC and TFS-NIC of Z. rouxii MC9 as a function of environmental factors. The
217
variance in the response that was explained by the models was the highest, with 96.7
218
and 97.6% for TFS-IC and TFS-NIC, respectively. All the terms of regression were
219
retained in the mathematical equations, with very similar values for both temperature
220
systems. In this way, from Table 4 is possible to deduce the RS equations that predict
221
the TFS of the concentrated grape juice as a function of pH and sugar concentration at
222
constant and overseas shipping temperatures. By replacing the appropriate terms in the
223
general equation 1, the respective RS equations for TFS are as follow:
224
TFS-IC (days) = 2849.592 – 79.148 (ºBrix) + 0.596 (ºBrix)2 – 232.069 pH + 20.771
225
(pH)2 + 1.537 (ºBrix) * (pH)
(2)
11
226
TFS-NIC (days) = 2771.561 – 75.864 (ºBrix) + 0.573 (ºBrix)2 – 228.12 pH + 27.544
227
(pH)2 + 0.956 (ºBrix) * (pH)
228
Both equations (very similar between them) allow estimation of the time (days) that Z.
229
rouxii needs to produce visible spoilage in concentrated grape juice at isothermal and
230
non-isothermal conditions.
231
The graphical representation of combined effect of the limiting growth factors
232
extending the microbial stability of the product at isothermal and non-isothermal
233
conditions is shown in Figures 4a and 5a, respectively. Both RSs had a very similar
234
morphology. The highest effects were noticed for the linear and quadratic effects of pH
235
(see Figures 4b and 5b). Thereby, a gradual decrease of pH was directly correlated with
236
the increase in the microbial stability of the product, showing a greater difference in the
237
highest concentrations of sugar evaluated. At pH values below 2.5, small variations in
238
pH units represented significant increase in product shelf life (see Figures 4a and 5a).
239
Concentrated grape juice with pH adjusted below 2.0 allowed to increase the shelf life
240
of the product for more than 60 days.
241
4. Discussion
242
It has been previously reported that the spoilage of various high-sugar products, such as
243
honey, maple syrup, dried fruits, concentrated fruit juices, raw sugar cane, jams and
244
jellies, was caused mainly by the activity of osmophilic yeasts (Deák and Beuchat,
245
1993; Tokouka, 1993). Yeasts have been reported to be significant spoilage organisms,
246
especially in foods with low pH and high sugar and salt concentrations and in products
247
containing sorbate and benzoate as preservatives, as well as in the presence of alcohol
248
where most bacterial species are inhibited (Evans et al., 2004; Praphailong and Fleet,
(3)
12
249
1997). Many environmental factors affect the yeast growth, but the response to any
250
particular condition varies within the species (Praphailong and Fleet, 1997)
251
In order to design adequate strategies to prevent spoilage, it is advantageous to know the
252
identity of the spoilage organisms present in the products and to get an insight into the
253
source of contamination (Loureiro, 2000). In a previous study, Z. rouxii has been
254
described as the main microbiological agent that causes spoilage in concentrated grape
255
juice from Argentina (Combina et al., 2008).
256
Many yeast species can tolerate a wide range of pH, from pH 1.5 to 10.0. In fact, most
257
yeasts prefer a slightly acidified medium, between 3.5 and 6.0, which is the pH found in
258
most fruit juices, beverages and soft drinks. aw of foods is also a very important factor
259
limiting yeast growth. While most yeasts will easily grow in 20% (w/v) glucose, only a
260
limited number of yeast species are able to grow at low aw caused by the presence of
261
high concentrations of sugar (60% w/v) (Tilbury, 1980ab). Moreover, Z. rouxii is able
262
to grow at a wide range of pH values, such as pH 1.8 to 8.0 in the presence of high
263
concentrations of glucose (Tokuoka, 1993) or pH 1.5 to 10.5 in 12% glucose medium
264
(Restaino et al., 1983). In a well-documented review of spoilage yeast, Fleet (1992)
265
mentioned that a high sugar concentration may either increase or decrease the low-pH
266
tolerance of yeast and emphasized that further study of these influences would be
267
needed to clarify the discrepant to observations that have been described. In this paper,
268
we study the combined effects of pH and high sugar concentrations on the Z. rouxii
269
growth parameters and TFS evaluated in natural substrate under two temperature
270
conditions to mimic product storage and shipping overseas.
13
271
Membré et al. (1999) described the combined effects of pH and high sugar
272
concentration on Z. rouxii growth rate in laboratory culture media. Our results shown a
273
partial agreement with these authors, who found that increasing the sugar concentration
274
from 300 (aw: 0.957) to 800 g/L (aw: 0.843) resulted in a reduction of the specific
275
growth rate, and the growth rate at high concentrations, such as 875 (aw: 0.810) and 950
276
(aw: 0.788) g/L, was very low. A pH of 2.5 resulted in a 30% reduction in the growth
277
rate, and no growth occurred at pH 2.0 at any sugar concentration assayed. In this work,
278
a pH of 1.9 allowed a slow growth of Z. rouxii in concentrated grape juice at 23ºC,
279
while no growth was observed after 60 days at pH 1.7. The minimum pH value which
280
allows the growth of Z. rouxii will be dependent on the strain, the culture medium
281
employed and the compound used to acidified the medium. For instance, Martorell et al.
282
(2007) found that pH 2.2 was the minimal pH for growth of two Z. rouxii strains in a
283
culture medium modified by adding HCl; and Restaino et al. (1983) showed that the
284
growth of Z. rouxii was inhibited at a pH as low as 1.5 evaluated under similar
285
conditions. However, when the pH value was adjusted by citric acid or other
286
inorganic/organic buffers, the minimal pH value to support Z. rouxii growth was 2.0
287
(Praphailong and Fleet ,1997; Membré et al., 1999). In this survey, the decrease in pH
288
was obtained by passing the grape juice through ion exchange columns. This fact
289
highlights the importance of evaluating the growth of spoilage yeast in natural
290
substrates using acidification methods normally employed by the industry.
291
Praphailong and Fleet (1997) concluded that 700 g/L of glucose was the maxima sugar
292
concentration in which Z. rouxii proliferated. This value is not consistent with results
293
obtained in the present study, since increasing of sugar concentration in concentrated
294
grape juice to levels as high as 842 g/L (68 ºBrix) did not greatly affected Z. rouxii
14
295
growth. Our results are in agreement with others authors, who have reported the growth
296
of Z. rouxii at low aw levels, such as 0.650 (Legan and Voyset, 1991; Tokuoka, 1993).
297
Moreover, Martorell et al. (2007) have reported that two Z. rouxii strains isolated from
298
spoiled syrup were able to grow in medium containing 900 g/L of glucose.
299
We have shown using RS methodology that the main limiting factor that affected Z.
300
rouxii growth was the pH, mainly when its value was below 2.1. In a recent work,
301
Vermeulen et al. (2012) studied the influence of the environmental stress factors on the
302
growth/no growth boundary of Z. rouxii. The authors found that a pH decrease from 7.0
303
to 3.5 had almost no effect on the time to detection of the different Z. rouxii strains.
304
Only pH values below 2.5 had a significant effect on the time to detection with an
305
increase from approximately 4-40 days. However, this pH was outside the relevant
306
range for the target product (chocolate filling) and it was not included in the model
307
design. Furthermore, even the most stringent conditions of pH (5.0) and aw (0.76)
308
evaluated, was not sufficient to prevent Z. rouxii growth (Vermeulen et al., 2012).
309
These results are in line with the obtained in our work, where increasing sugar
310
concentration was not enough to inhibit the growth of this spoilage specie in
311
concentrated grape juice.
312
The decrease in the pH values in concentrated grape juice leads to an increase in the
313
time required for spoilage. An extension of the self life for over 30 days represents a
314
huge marketing advantage, since at that time the product has arrived to destination. TFS
315
was above 60 days when pH was below 2.1 units, independently of the sugar
316
concentration and temperature condition. Time to show microbial spoilage of
317
concentrated grape juice was slight lower when a non-isothermal profile was applied.
318
This may be due to the extreme and oscillating temperatures during the first week of
15
319
this assay which produced water evaporation and subsequent condensation on the
320
substrate surface, increasing aw and promoting the onset of spoilage. It is important to
321
note, that the concentrated grape juice were inoculated with 102 CFU/mL, which
322
represent the maximum limit recommended for fungi and yeasts count by the buyers.
323
This fact places us in the worst scenario, because the concentrated grape juice contain
324
the maximum yeast count tolerated and they are all osmophilic yeasts. Therefore, the
325
data showed represents the minimum microbial stability period for this substrate.
326
5. Conclusions
327
According to results obtained from the mathematical models, it is advisable to obtain
328
pH values below 1.7 to achieve total inhibition of Z. rouxii MC9 in concentrated grape
329
juices. However, this pH value could be difficult to achieve under industrial conditions.
330
Thereby, reducing the pH up to values of 2.0 may be enough to produce a significant
331
extension of the shelf life of the product at 68ºBrix for both storage and shipping
332
overseas temperatures.
333
Acknowledgements
334
This research was supported by the Technological Project MZASJ51007: Support for
335
Regional Viticulture Development - INTA. The authors wish to thank to concentrating
336
grape juice industry for providing all the samples used in the present study. M.C. Rojo
337
is a Doctoral Fellowship of CONICET, whereas F.N. Arroyo-López wants to thank
338
CSIC and Spanish government for his Ramón y Cajal postdoctoral research contract.
339
References
16
340
Arroyo-López, F.N., Durán Quintana, M.C., & Garrido Fernández, A. (2006). Use of
341
the generalized z-value concept to study the effects of temperature, NaCl
342
concentration and pH on Pichia anomala, a yeast related to table olive fermentation.
343
International Journal of Food Microbiology, 106, 45-51.
344
345
346
Baranyi J., Roberts T.A. (1994). A dynamic approach to predicting bacterial growth in
food. International Journal of Food Microbiology, 23, 277–294.
Beuchat, L., Frandberg, E., Deák, T., Alzamora, S., Chen, J.,Guerrero, S., López-Malo,
347
A., & Ohlsson, I. (2001). Performance of mycological media in enumerating
348
desiccated food spoilage yeasts: an interlaboratory study. International Journal of
349
Food Microbiology, 70, 89–96.
350
351
352
353
354
Beuchat, L.R. (1993). Selective media for detecting and enumerating foodborne yeasts.
International Journal of Food Microbiology, 19, 1–14.
Bruzone, A. (1998). Cadenas alimentarias: Jugo de uva concentrado. Alimentos
Argentinos, 8, 42-45.
Combina, M., Daguerre, C., Massera, A., Mercado, L., Sturm, M.E, Ganga, A., &
355
Martinez, C. (2008). Yeasts identification in with grape juice concentrates from
356
Argentina. Letters in Applied Microbiology, 46(2), 192-197.
357
D’Amato, D., Corbo, M.R., Del Nobile, M.A., & Sinigaglia, M. (2006). Effects of
358
temperature, ammonium and glucose concentrations on yeast growth in a model
359
wine system. International Journal of Food Science and Technology, 41, 1152-1157.
360
Deák, T. (Ed.) (2008). Yeasts in specific types of foods. In: Handbook of Food Spoilage
361
Yeasts (2nd ed.) (pp. 117–201). Boca Raton: CRC Press Taylor and Francis Group.
362
363
Deák, T., & Beuchat, L. (1993). Yeast associated with fruit juice concentrates. Journal
of Food Mycology, 56, 777-782.
17
364
365
366
Emmerich, W. & Radler, F. (1983). The anaerobic metabolism of glucose and fructose
by Saccharomyces bailii. Journal of General Microbiology, 129, 3311–3318.
Evans, D. G., Everis, L. K., & Betts, G. D. (2004). Use of survival analysis and
367
classification and regression trees to model the growth/no growth boundary of
368
spoilage yeasts as affected by alcohol, ph, sucrose, sorbate and temperature.
369
International Journal of Food Microbiology, 92, 55–67.
370
Fleet, G. (1992). Spoilage yeasts. Critical Review in Biotechnology, 12, 1–44.
371
Fugelsang, K.C & Edwards, C.G. (Eds.), (2007). Yeasts. In: Wine Microbiology.
372
Practical Applications and Procedures (2nd ed.) (pp. 3-14). New York: Springer
373
Science
374
Hartley, A. (2008). Bulk shipping of wine and its implications for product quality.
375
Project code MSG009.
376
http://www.wrap.org.uk/sites/files/wrap/Bulk%20shipping%20&%20wine%20qualit
377
y%20May'08.pdf. Last Access date: January 5th, 2012.
378
ICMSF – International Commission on Microbiological Specifications of Foods (1980)
379
Bebidas no alcoholicas, zumos de frutas naturales, concentrados y mermeladas. In:
380
Ecología Microbiana de los Alimentos (Volumen II) (pp. 652–677). Zaragoza:
381
Editorial Acribia.
382
James, S., & Stratford, M. (2003). Spoilage yeasts with emphasis on the genus
383
Zygosaccharomyces. In: T. Boekhout, V. Robert (Eds), Yeasts in Food (pp. 171–86)
384
Cambridge: Woodhead Publ. Ltd.
385
Kalathenos, P., Baranyi, J., Sutherland, J.P., & Roberts, T.A. (1995). A response surface
386
study on the role of some environmental factors affecting the growth of
387
Saccharomyces cerevisiae. International Journal of Food Microbiology, 25, 63-74.
18
388
389
390
Legan, J.D., & Voyset, P.A. (1991). Yeast spoilage of bakery products and ingredients.
Journal of Applied Bacteriology, 70, 361-371.
Leinberger, D. (2006). Temperature and humidity in ocean containers.
391
http://www.ista.org/forms/LEINBERGER_Dimensions06_paper.pdf. Last accession
392
date: January 5th, 2012.
393
394
395
Loureiro, V., & Malfeito-Ferreira, M. (2003). Spoilage yeasts in the wine industry.
International Journal of Food Microbiology, 86, 23–50.
Loureiro, V. (2000). Spoilage yeasts in foods and beverages: characterization and
396
ecology for improved diagnosis and control. Food Research International, 33, 247–
397
256.
398
Martorell, P., Stratford, M., Steels, H., Fernández-Espinar, M. T., & Querol, A. (2007).
399
Physiological characterization of spoilage strains of Zygosaccharomyces bailii and
400
Zygosaccharomyces rouxii isolated from high sugar environments. International
401
Journal of Food Microbiology, 114, 234–242.
402
McMeekin, T.A., Olley, J.N., Ross, T., & Ratkowsky, D.A. (Eds) (1993). Predictive
403
Microbiology: Theory and Application. New York: John Wiley & Sons, Inc.
404
Membré, J.M., Kubaczka, M., 6 Chene, C. (1999). Combined effects of pH and sugar
405
on growth rate of Zygosaccharomyces rouxii, a bakery product spoilage yeast.
406
Applied and Environmental Microbiology, 65,4921-4925.
407
INV - Instituto Nacional Vitivinicola (2013). Estadistica de exportacion de mostos.
408
http://www.inv.gov.ar/PDF/Estadisticas/Exportaciones/2013/ComparativoMosto.pdf
409
. Last accession date: February 2nd, 2013.
410
Praphailong, W., & Fleet, G. H. (1997). The effect of pH, sodium chloride, sucrose,
411
sorbate and benzoate on the growth of food spoilage yeasts. Food Microbiology, 14,
412
459–468.
19
413
Restaino, L., Bills, S., Tscherneff, K., & Lenovich, L.M. (1983). Growth characteristics
414
of Saccharomyces rouxii from chocolate syrup. Applied and Environmental
415
Microbiology, 45, 1614-1621.
416
Smith, J.P., Daifas, D.P., El-Khoury, W., & Koukoutsis, J., (2004). Shelf life and safety
417
concerns of bakery products. Critical Reviews in Food Sciences and Nutrition,
418
44,19-55.
419
420
421
Splittstoesser, D.F. (1987). Fruits and fruit products. In: L.R. Beuchat (Ed). Food and
Beverage Mycology (2nd ed.) (pp. 101-122). New York: AVI Book.
Tilbury, R.H. (1980a). Xerotolerant yeasts at high sugar concentrations. In: G.W. Gould
422
& J.E.L. Corry (eds.). Microbial Growth and Survival in Extremes of Environment.
423
Society for Applied Bacteriology Technical Series (Volumen 15) (pp. 103–128).
424
London: Academic Press.
425
Tilbury, R.H. (1980b). Xerotolerant (osmophilic) yeasts. In: F.A. Skinner, S.M.
426
Passmore & R.R. Davenport (eds.). Biology and Activity of Yeasts. Society for
427
Applied Bacteriology, Symposium Series (pp. 153–179) London: Academic Press.
428
429
430
Tokuoka, K. (1993) A review: sugar and salt tolerant yeasts. Journal of Applied
Bacteriology, 74, 101-110.
Vermeulen, A., Daelman, J., Van Steenkiste, J., & Devlieghere, F. (2012). Screening of
431
different stress factors and development of growth/no growth models for
432
Zygosaccharomyces rouxii in modified Sabouraud medium, mimicking intermediate
433
moisture foods (IMF). International Journal of Food Microbiology, 32, 332-341.
434
Worobo, R.W, & Splittstoesser, D.F. (2005). Microbiology of fruit products. In: D.M.
435
Barret, L. Somogyi & H. Ramaswamy (eds). Processing fruit. (2nd ed.) (pp: 161-
436
284) Boca Raton: CRC Press, Taylor and Francis Group.
437
20
438
Figure Legends
439
Figure 1. Inside and outside temperature profiles (non-isothermal conditions) recorded
440
during the experiment to reproduce overseas shipping of concentrated grape juice from
441
south Hemisphere (summer time) to north Hemisphere (winter time).
442
Figure 2. Response surface graph (a) and Pareto chart standardized estimate effect of
443
the regression coefficients (b) for potential maximum growth rate (µmax, log10 cfu/mL
444
days-1) of Zygosaccharomyces rouxii MC9 as a function of pH and sugar concentration
445
(ºBrix) at isothermal conditions (23ºC) in concentrated grape juice.
446
Figure 3. Response surface graph (a) and Pareto chart standardized estimate effect of
447
the regression coefficients (b) for lag phase duration (λ, days) of Zygosaccharomyces
448
rouxii MC9 as a function of pH and sugar concentration (ºBrix) at isothermal conditions
449
(23ºC) in Argentinean concentrated grape juice.
450
Figure 4. Response surface graph (a) and Pareto chart standardized estimate effect of
451
the regression coefficients (b) for time for spoilage (TFS, days) produced by
452
Zygosaccharomyces rouxii MC9 as a function of pH and sugar concentration (ºBrix) at
453
isothermal conditions (IC, 23ºC) in concentrated grape juice.
454
Figure 5. Response surface graph (a) and Pareto chart standardized estimate effect of
455
the regression coefficients (b) for time for spoilage (TFS, days) produced by
456
Zygosaccharomyces rouxii MC9 as a function of pH and sugar concentration (ºBrix) at
457
non-isothermal conditions (NIC, variable temperatures) in concentrated grape juice.
21