J. Dairy Sci. 2007. 90:5361-5373. doi:10.3168/jds.2007-0273
© 2007 American Dairy Science Association ®
Enhancement of Functional Characteristics of Mixed Lactic Culture Producing Nisin Z and Exopolysaccharides During Continuous Prefermentation of Milk with Immobilized Cells
F. Grattepanche*,
P. Audet
and
C. Lacroix
* STELA Dairy Research Center, Pavillon Paul-Comtois, Université Laval, Québec, Québec, G1K 7P4 Canada
École des sciences des aliments, de nutrition et d’études familiales, Pavillon Jacqueline-Bouchard, Université de Moncton, Moncton, New Brunswick, E1A 3E9 Canada
Laboratory of Food Biotechnology, Institute of Food Science and Nutrition, Swiss Federal Institute of Technology, ETH-Zürich, Schmelzbergstrasse 7, 8092 Zürich, Switzerland
1 Corresponding author: christophe.lacroix{at}ilw.agrl.ethz.ch
 |
ABSTRACT
|
|---|
Antagonistic phenomena between strains often occur in mixed cultures containing a bacteriocinogenic strain. A nisin Z producer (Lactococcus lactis ssp. lactis biovar. diacetylactis UL719) and 2 nisin-sensitive strains for acidification (Lactococcus lactis ssp. cremoris ATCC19257) and exopolysaccharide (EPS) production (Lactobacillus rhamnosus RW-9595M) were immobilized separately in gel beads and used to continuously preferment milk at different temperatures, with pH controlled at 6.0 by fresh milk addition. The process showed high volumetric productivity, with an increase from 8.0 to 12.5 L of prefermented milk per liter of reactor volume and hour as the temperature was increased from 27 to 35°C. Lactococcus lactis ssp. lactis biovar. diacetylactis UL719 counts in prefermented and fermented (22-h batch fermentation) milks were stable during 3 wk of continuous fermentation (8.1 ± 0.1 and 8.9 ± 0.2 log cfu/mL, respectively). The L. lactis ssp. cremoris population (estimated with real-time quantitative PCR) decreased rapidly during the first week of continuous culture to approximately 4.5 log cfu/mL and remained constant afterward. Lactobacillus rhamnosus counts in prefermented and fermented milks significantly increased with prefermentation time, with no temperature effect. Nisin Z reached high titers in fermented milks (from 177 to 363 IU/mL), with EPS concentration in the range from 43 to 178 mg/L. Immobilization and continuous culture led to important physiological changes, with Lb. rhamnosus becoming much more tolerant to nisin Z, and Lb. rhamnosus and L. lactis ssp. lactis biovar. diacetylactis UL719 exhibiting large increases in milk acidification capacity. Our data showed that continuous milk prefermentation with immobilized cells can stimulate the acidification activity of low-acidifying strains and produce fermented milks with improved and controlled functional properties.
Key Words: immobilization • exopolysaccharide mixed culture • continuous fermentation
|
INTRODUCTION
|
|---|
Lactic acid bacteria play a major role in aroma development, microbiological safety, and texture of fermented dairy products. In recent years, strains with new functional properties such as bacteriocin and exopolysaccharide (EPS) production have been studied, and some have been used for fermented milk and cheese production. Bacteriocins are defined as proteinaceous antimicrobial compounds with bacteriostatic or bactericidal activities, or both, against closely related species and a broader host spectrum for some bacteriocins, such as lantibiotics (Tagg et al., 1976; Klaenhammer, 1993). Lactic acid bacteria bacteriocins have been extensively studied to control the growth of pathogenic and spoilage microorganisms in fermented and nonfermented foods (Cleveland et al., 2001). Lactic acid bacteria can also produce and secrete EPS, which contributes to the texture, mouthfeel, taste perception, and stability of the final products (De Vuyst et al., 2001; Jolly et al., 2002). Moreover, several authors have suggested that EPS have physiological functions, such as antitumor and immunomodulating activities and improvement of probiotic bacterial growth (Ruas-Madiedo et al., 2002). These functional strains may exhibit a limited capacity to grow and produce acid in milk, as shown for EPS (Dupont et al., 2000; Grattepanche et al., 2007) and nisinogenic strains (Roberts et al., 1992; De Vuyst and Vandamme, 1994), and must usually be combined with acidifying cultures. However control of mixed cultures with bacteriocin producers is difficult, because bacteriocin production may limit the growth of bacteriocin-sensitive strains (Grattepanche et al., 2007).
Cell immobilization has been studied for continuous starter culture and metabolite productions, and continuous milk inoculation and prefermentation, with many advantages compared with free-cell cultures: high cell density leading to high process productivity, control of strain ratios in mixed cultures, reuse of biocatalysts, enhanced tolerance to environmental stresses, retention of cell plasmids, improved resistance to contamination, stimulation of production and secretion of secondary metabolites, and physical and chemical protection of the cells (Doleyres and Lacroix, 2005; Lacroix et al., 2005).
In the present work, we studied cell immobilization and continuous culture for milk inoculation and prefermentation with a functional mixed culture containing antagonistic strains. The tested mixed culture contained 2 nisin-sensitive strains, an acidifier strain (Lactococcus lactis ssp. cremoris ATCC19257), an EPS-producing strain (Lactobacillus rhamnosus RW-9595M), and a high nisin Z producer (Lactococcus lactis ssp. lactis biovar. diacetylactis UL719). The effects of temperature and prefermentation age were examined on microbial populations in prefermented milks directly from the continuous culture and in prefermented milks that were incubated for an additional 22 h in batch, to model fresh cheese fermentation and examine metabolite production.
|
MATERIALS AND METHODS
|
|---|
Bacterial Strains
Lactobacillus rhamnosus RW-9595M was selected for its capacity to produce high amounts of EPS in dairy media (Dupont et al., 2000; Bergmaier et al., 2003); the structure of this EPS has been studied by Van Calsteren et al. (2002). Lactococcus lactis ssp. lactis biovar. diacetylactis UL719 and Pediococcus acidilactici UL5 were used as nisin Z producer and indicator strains for bacteriocin activity assay, respectively (Meghrous et al., 1997). All strains were obtained from the Lactic Acid Bacterial Research culture collection (Dairy Research Center STELA, University Laval, Québec, Québec, Canada). Lactococcus lactis ssp. cremoris ATCC19257 (American Type Culture Collection, Manassas, VA) was selected for its ability to acidify milk. Before use, cultures from frozen stocks (–80°C) were grown twice for 24 h in M17 broth (Difco, Detroit, MI) at 30°C for lactococci and in de Man, Rogosa, Sharpe (MRS) broth (Difco) at 30 and 37°C for P. acidilactici UL5 and Lb. rhamnosus RW-9595M, respectively.
Cell Immobilization Procedure
Each strain was immobilized separately in mixed polysaccharide gel beads by using a 2-phase dispersion procedure, as described by Lamboley et al. (1997). The 
-carrageenan and locust bean gum polymer solution [2.75 and 0.25% (wt/vol), respectively] was inoculated at 2% (vol/vol) with an overnight pure culture of each strain. Gel beads with diameters ranging from 1 to 2 mm were selected by wet sifting with a 0.2 M KCl solution. To increase the immobilized biomass, beads with pure culture of each strain were colonized separately during 2 successive batch cultures, carried out in a 1-L bioreactor (Biogénie, Ste-Foy, Québec, Canada) with a bead inoculum of 10% (vol/vol), in M17 at 30°C for 6 h and in MRS at 37°C for 20 h, for lactococci and Lb. rhamnosus RW-9595M, respectively, with pH controlled at 6.0 using 6 N NaOH. The M17 and MRS media were supplemented with 50 g/L of lactose and 0.2 M KCl to avoid sugar limitation and to maintain bead integrity, respectively. After this colonization step, the immobilized biomass reached 3.16 ± 0.24 x 1010, 1.52 ± 0.15 x 1011, and 7.23 ± 0.58 x 1010 cfu/g of gel beads for L. lactis ssp. cremoris ATCC19257, L. lactis ssp. lactis biovar. diacetylactis UL719, and Lb. rhamnosus RW-9595M, respectively.
Continuous Prefermentation of Milk
The continuous inoculation and prefermentation of milk was performed as described by Sodini-Gallot et al. (1995), in a custom-made glass bioreactor with a useful volume of 100 mL. A commercial microfiltrated skim milk (PurFiltre, Québec, Québec, Canada), supplemented with 0.5 g/L of MgSO4·7H2O and 0.05 g/L of MnSO4·H2O to stimulate growth of Lb. rhamnosus RW- 9595M (Bergmaier et al., 2003) and acid production in milk (data not shown), was used for the feed medium. The system was inoculated with 25 mL of colonized gel beads from the 3 strains, with a strain ratio of 1:1:1 (by vol). The pH was continuously maintained at 6.0 by inflow of milk, and mixing was provided by a magnetic bar at 120 rpm. Before entering the bioreactor, milk at 2°C was preheated to the set temperature by immersing the tubing in a water bath. The milk feed flow rate was used to control the pH in the bioreactor (PID control software, Biogénie). The useful volume of the bioreactor was maintained by positioning the out-flow tube and setting the outlet peristaltic pump at a higher flow rate value than the feeding pump. To retain the beads, a stainless-steel screen was installed at the outlet tubing and periodically purged with a gas pulse of nitrogen to prevent clogging of the mesh by the clotted milk and beads. The milk flow rate was monitored continuously by recording the power supply of the feed pump and was estimated daily by volumetric measurements by using a graduated cylinder.
At the end of each week, the continuous prefermentation was stopped. The beads were stored in a solution of 0.1% peptone, 0.2 M KCl, and 0.03 M citrate buffer at pH 5.6, and the bioreactor was washed (Sodini et al., 1997b). At the beginning of each week, the beads were transferred to the bioreactor, which was rinsed with 1 L of milk fed at high flow rate (3 L/h) to eliminate the bead storage solution. The continuous prefermentation was then stabilized for 2 d before sampling, except for the second week, for which steady state, indicated by a stable milk flow rate, was obtained after 1 d. Three temperatures were tested. To evaluate the effect of storage conditions, the same prefermentation temperature of 31°C was used before stopping and to restart the system. The 2 other temperatures (27 and 35°C) were randomly applied and repeated 4 times over the 3-wk experiment (Figure 1
). The temperature was changed daily and a pseudo steady state, indicated by variations in the dilution rate of less than 0.4 h–1 within 4 h of continuous culture, was reached in less than 7 h. Three 200-mL samples of prefermented milk were taken during steady-state conditions, at 30-min intervals, for chemical and microbial analyses and for batch fermentations, after a 22-h fermentation with a new condition.
Final Fermentation of the Prefermented Milk
Three 100-mL prefermented milk samples were transferred into 150-mL Erlenmeyer flasks, which were placed into a water bath set at the prefermentation temperature. The pH was recorded every minute during incubation for 22 h (Biogénie software). Acidification parameters were calculated as described previously (Grattepanche et al., 2007): maximal acidification rate (|Vmax|, milli-pH unit/min; mupH/min), time and pH at which |Vmax| occurred (tmax and pHmax, respectively), time during which the acidification rate was equal to or higher than |Vmax |/2 (dt50), time to reach pH 5.25 (tpH 5.25), and final pH (pHF; Lamboley et al., 1997). At the end of acidification, samples were collected for chemical analyses and cell enumeration.
Cell Enumeration
Cell populations in prefermented and fermented milks and in gel beads were enumerated by using the appropriate dilutions in 0.1% peptone water (wt/vol) plated in triplicate on agar media. Enumeration of L. lactis ssp. lactis biovar. diacetylactis UL719 and Lb. rhamnosus RW-9595M was performed on differential KMK (Kempler and McKay, 1980) medium (Geneq Inc., Montreal, Québec, Canada) as described previously (Grattepanche et al., 2007). Lactococcus lactis ssp. cremoris ATCC19257 had low concentrations in prefermented and fermented milks and was estimated by using a real-time PCR method (Grattepanche et al., 2005).
Dispersion of immobilized populations in gel beads was performed by using 2 methods, depending on the enumeration method used. For L. lactis ssp. lactis biovar. diacetylactis UL719 and Lb. rhamnosus RW-9595M, a bead sample of approximately 1.0 g, accurately weighed, was placed in 9 mL of 0.1% peptone water (wt/vol) and treated with an UltraTurrax (Janke and Kunkel, IKA Laborte, Staufen, Germany) for 1 min at 20,000 rpm in ice. The mixture was then diluted in peptone water and plated. For L. lactis ssp. cremoris ATCC19257, approximately 0.5 g of gel beads, accurately weighed, was placed in a tube containing 9.5 mL of water and held at 90°C, with periodic vortexing until complete dissolution of the gel. After centrifugation at 10,000 x g at room temperature for 10 min, the supernatant was discarded and the cell pellet was washed twice with sterile water. The pellet was then resuspended in 400 µL of TES buffer (50 mM Tris-HCl, 1 mM EDTA, 6.7% sucrose, pH 8.0) before DNA extraction and quantification by real-time PCR. Microbiological analyses were performed in duplicate.
Quantification of Nisin Z Activity
Nisin Z in prefermented and fermented milk samples was extracted and quantified as described previously (Grattepanche et al., 2007). Analyses were performed in triplicate.
Determination of Nisin Sensitivity
Sensitivity of Lb. rhamnosus RW-9595M to nisin Z was measured by using the agar diffusion bioassay. Samples of prefermented milk from d 1 and 22 were diluted in peptone water to obtain a final cell concentration for Lb. rhamnosus RW-9595M of 106 cfu/mL. A 30-mL volume of MRS containing 7.5 g/L of agar at 45°C was inoculated at 1% (vol/vol) with the bacterial suspension and poured into a Petri dish. After agar solidification, wells (6 mm in diameter) were made and filled with 100-µL pure nisin Z solutions at concentrations varying from 47 to 1,288 IU/mL. The plates were incubated at 45°C for 48 h, and inhibition zones were measured and plotted as a function of the logarithm of nisin concentration. An overnight culture of Lb. rhamnosus RW-9595M in MRS broth at 45°C from the frozen stock was used as a control. To test the reversibility of nisin tolerance for Lb. rhamnosus RW-9595M cells, prefermented milk samples at d 22 were subcultured 8 times in MRS medium with a 1% inoculum and incubated at 45°C for 18 h. Reported data are means of triplicate analyses.
Determination of EPS Concentration
The EPS concentration in fermented milk samples was determined as described previously (Grattepanche et al., 2007). Analyses were performed in triplicate.
Acidification Capacity of L. lactis ssp. lactis biovar. diacetylactis UL719 and Lb. rhamnosus RW-9595M
Additional acidification tests were carried out with prefermented milk harvested at d 22 and milk inoculated with pure cultures (controls) of L. lactis ssp. lactis biovar. diacetylactis UL719 and Lb. rhamnosus RW-9595M, under conditions selecting for the growth of each strain, to compare their acidifying capacity in the 2 types of samples. Cell concentrations of pure culture inocula were adjusted to those in prefermented milk samples (9.2 x 107 and 2.9 x 106 cfu/mL for L. lactis ssp. lactis biovar. diacetylactis UL719 and Lb. rhamnosus RW-9595M, respectively). Incubation was performed at 30°C for 48 h to preferentially grow L. lactis ssp. lactis biovar. diacetylactis UL719 and at 45°C for Lb. rhamnosus RW-9595M. The pH was monitored and acidification parameters were calculated. Experiments were performed in duplicate.
Statistical Analyses
Multiple regression (JMP In Software, version 4, SAS Institute Inc., Cary, NC) was used to fit bacterial populations in prefermented and fermented milk, acidification parameters, and metabolite responses as first- and second-order functions of temperature (T) and age of prefermentation (A). The regression model equation for temperature, between 27 and 35°C, and age of prefermentation, between 2 and 22 d, was in the form:
 |
where xT and xA were means of temperature (31°C) and age of prefermentation (11.6 d), respectively. Regression coefficients 
ij were interpreted as the intercept (00), the linear and quadratic effect of temperature (10 and 20, respectively), and the linear and quadratic effect of age of the prefermentation (01 and 02, respectively), and 
was the residual. Significance of the overall model was tested with the Fisher test. Student’s t-test was used to test the significance of model coefficients. P-values of less than 0.05 for a 95% confidence interval indicate a statistically significant nonzero coefficient.
|
RESULTS
|
|---|
Milk Dilution Rate
The dilution rate of prefermented milk increased from 8.0 ± 1.1 to 12.5 ± 1.5 h–1 as the temperature increased from 27 to 35°C. The prediction model for dilution rate explained 80% of the total variation and was highly significant (P < 0.01). Analyses of variance of the effects showed that only the temperature factor was significant (P = 0.0011) on the milk dilution rate.
Microbial Populations in Prefermented and Fermented Milks
The immobilized population of L. lactis ssp. lactis biovar. diacetylactis UL719 remained stable at 10.72 ± 0.04 log cfu/g of gel beads over the 3 wk of continuous fermentation (Figure 1
). Populations of L. lactis ssp. cremoris ATCC19257 decreased by 1 log after 2 wk and remained constant afterward. A decrease in immobilized cell population of Lb. rhamnosus RW-9595M from 10.44 ± 0.04 to 9.28 ± 0.11 log cfu/g was observed for the first week (Figure 1).
In prefermented milk, the population of L. lactis ssp. lactis biovar. diacetylactis UL719 of 8.11 ± 0.12 log cfu/ mL was not significantly changed by age of prefermentation and temperature (Figure 1 and Table 1). The populations of L. lactis ssp. cremoris ATCC19257 and Lb. rhamnosus RW-9595M in milk effluent decreased and increased, respectively, with age of prefermentation time, whereas temperature showed no significant effect (P > 0.005; Figure 1 and Table 1). The model for these 2 strains was significant and explained 65 and 76% of the total variation for L. lactis ssp. cremoris ATCC19257 and Lb. rhamnosus RW-95959M cell counts, respectively (Table 1). Figure 1 shows the effect of covariate age at 31°C on L. lactis ssp. cremoris ATCC19257 and Lb. rhamnosus RW-95959M populations in prefermented milk. The covariate models for these 2 populations indicate a maximum increase of approximately 1.8 log and a decrease of approximately 1.9 log for Lb. rhamnosus RW-95959M and L. lactis ssp. cremoris ATCC19257, respectively, with age of prefermentation. The highest cell count measured for Lb. rhamnosus RW-9595M (7.85 ± 0.03 log cfu/mL) was reached at d 20 at 35°C. The population of L. lactis ssp. cremoris ATCC19257 decreased by more than 3 log cfu/ mL over a 3-wk continuous fermentation, as shown by counts measured on d 2 and 22, both at 31°C (Figure 1).
View this table:
[in this window]
[in a new window]
|
Table 1. Regression coefficients for models for bacterial populations in prefermented and fermented milks as a function of temperature (T) and age of the prefermentation
|
|
For fermented milks, age of prefermentation and temperature had no significant effects on L. lactis ssp. lactis biovar. diacetylactis UL719 counts, which were higher (plus 0.67 ± 0.16 log cfu/mL) than in prefermented milks (Table 1 and Figure 2). Analyses of variance showed that age of prefermentation had highly significant linear effects on L. lactis ssp. cremoris ATCC19257 and Lb. rhamnosus RW-9595M populations, and the models explained 85 and 83% of total variations (Table 1). The effect of age of prefermentation on cell counts was also greater for these 2 strains in fermented than in prefermented milk, with L. lactis ssp. cremoris ATCC19257 counts decreasing by approximately 4.2 log during a 22-d culture and Lb. rhamnosus RW-9595M increasing by approximately 2.84 log between d 2 and 14 and decreasing slightly thereafter (Figure 1 and 2). Cell counts for L. lactis ssp. cremoris ATCC19257 and Lb. rhamnosus RW-9595M were also significantly affected by temperature (Table 1). In fermented milk, the highest population of Lb. rhamnosus RW-9595M (9.13 ± 0.03 log cfu/mL) was reached at d 20 for 35°C. After 20 d, the population of L. lactis ssp. cremoris ATCC19257 in fermented milk was less than 2 log cfu/mL for all 3 temperatures. A sharp decrease in cell counts for this strain was observed on d 11, 13, and 20 when fermentation was conducted at 35°C.
Batch Acidification of Prefermented Milk
Figure 3 shows the changes in pH during a 22-h batch incubation of prefermented milks produced at different temperatures. The maximal acidification rate (|Vmax|), time to reach pH 5.25 (tpH 5.25), and final pH (pHF) values were all significantly affected by incubation temperature and age of prefermentation (Table 2). The lowest pH (4.02 ± 0.01) and tpH 5.25 (243 ± 28 min) were measured on d 20 for the highest temperature of 35°C (Figure 3). The |Vmax| decreased significantly with both temperature and age of the prefermentation (Table 2). The highest value of |Vmax| (2.95 ± 0.10 mupH/min) was measured on d 20 (35°C), whereas the lowest |Vmax| (0.81 ± 0.03 mupH/min) was tested on d 3 (27°C). However, tmax (145 ± 19 min) and pHmax (5.77 ± 0.03), corresponding to the time and pH at which |Vmax| occurred, were not significantly changed by temperature and age of prefermentation (data not shown). For d 21 (27°C) and 22 (31°C), 2 |Vmax| were recorded (Figure 3). The second |Vmax| values of 2.08 ± 0.11 and 2.77 ± 0.14 mupH/min for d 21 and 22 were recorded for pHmax of 4.99 ± 0.07 and tmax of 957 ± 10 and 585 ± 12 min, respectively.
View larger version (18K):
[in this window]
[in a new window]
|
Figure 3. Change of pH during a 22-h batch incubation of prefermented milks produced at different temperatures and different ages of fermentation. Bars represent the standard deviations calculated from 3 repetitions.
|
|
View this table:
[in this window]
[in a new window]
|
Table 2. Regression coefficients for models for final pH (pHF), time to reach pH 5.25 (tpH 5.25) and maximal acidification rate (|Vmax|) measured during a 22-h batch incubation of prefermented milks
|
|
Nisin and EPS Productions
Nisin Z was not detected in prefermented milk produced during continuous culture with a detection limit of the method of 40 IU/mL. Nisin Z concentration in fermented milks after a 22-h incubation varied from 177 ± 38 to 363 ± 30 IU/mL (Figure 4). However, no significant effects were observed for age of prefermentation and temperature of fermentation, and the regression model was not significant (Table 3).
View larger version (10K):
[in this window]
[in a new window]
|
Figure 4. Nisin Z concentration in fermented milks after 22-h incubation of prefermented milk produced at different ages and temperatures. Bars represent standard deviations calculated from analyses of 3 samples.
|
|
View this table:
[in this window]
[in a new window]
|
Table 3. Regression coefficients of the models for nisin Z and EPS productions in fermented milks after a 22-h incubation
|
|
Concentrations of EPS in fermented milks ranged from 43 ± 7 to 178 ± 29 mg/L (Figure 5). The regression model for EPS production in fermented milk was significant and explained 74% of the total variation (Table 3). Both linear and quadratic effects of age of prefermentation and a linear effect of temperature were significant (Table 3). Concentrations of EPS predicted by the covariate model increased with age of prefermentation from 67 to 145 mg/L for 27 or 31°C, and from 104 to 182 mg/L at d 14 for 35°C (Figure 5). The highest amount of EPS (178 ± 29 mg/L) was measured on d 11 and at 35°C (Figure 5). A significant correlation was also calculated between cell counts of Lb. rhamnosus RW-9595M and EPS concentrations (R2 = 0.73 and P < 0.0001).
View larger version (11K):
[in this window]
[in a new window]
|
Figure 5. Exopolysaccharide (EPS) concentration in fermented milks after a 22-h incubation of prefermented milk produced at different ages and temperatures. The dashed lines show the EPS concentration in fermented milks predicted by the regression model for different ages of prefermentation and temperature. Bars represent the standard deviations calculated from analysis of 3 samples.
|
|
Nisin Sensitivity
Figure 6 shows nisin Z sensitivities for Lb. rhamnosus RW-9595M produced in different culture conditions. Cells in prefermented milks (d 1 and 22) were less sensitive to nisin Z, tested at different concentrations, than cells from control-batch cultures. Moreover, cells in prefermented milk on d 22 were much more tolerant toward nisin Z than on d 1, especially for high nisin concentrations. Nisin tolerance of cells produced during subcultures of cells in prefermented milk on d 22 decreased progressively with the number of subcultures and reached a sensitivity level similar to that of cells in prefermented milk on d 1 after 8 subcultures (Figure 6).
View larger version (12K):
[in this window]
[in a new window]
|
Figure 6. Sensitivity to nisin Z of Lactobacillus rhamnosus RW-9595M produced during control free-cell batch culture in de Man, Rogosa, Sharpe broth ( ), in prefermented milks harvested at d 1 ( ) and d 22 ( ; S0) and after 3, 6, 7, and 8 subcultures of cells from prefermented milk harvested at d 22 (; dashed lines, S3, S6, S7, and S8). Bars represent standard deviations calculated from 3 repetitions.
|
|
Acidification Capacity of L. lactis ssp. lactis biovar. diacetylactis UL719 and Lb. rhamnosus RW-9595M
During incubation at 30°C, milk acidification in prefermented milk started later than in the control (Figure 7). Nevertheless, after a 48-h incubation, the final pH was significantly different (Table 4). For both cultures, |Vmax| and tpH 5.25 were not significantly different (Table 4). However, dt50, the time during which the acidification rate was equal to or higher than |Vmax|/2, was approximately 1.6 higher for prefermented milk (1,137 ± 58 min) than the control (721 ± 33 min). Cell counts of L. lactis ssp. lactis biovar. diacetylactis UL719 in experimental and control fermented milks were not significantly different (Table 4), whereas counts of Lb. rhamnosus RW-9595M in prefermented milk increased from 6.46 to 7.27 ± 0.01 log cfu/mL.
View this table:
[in this window]
[in a new window]
|
Table 4. Acidification parameters and cell counts, after a 48-h incubation at 30 and 45°C, of prefermented milk samples harvested at d 22 and control pure cultures of Lactococcus lactis ssp. lactis biovar. diacetylactis UL719 and Lactobacillus rhamnosus RW-9595M1
|
|
At 45°C, only Lb. rhamnosus RW-9595M grew during incubation of prefermented milk, and L. lactis ssp. lactis biovar. diacetylactis UL719 was not detected after a 48-h incubation (Table 4). Milk acidification was delayed compared with Lb. rhamnosus RW-9595M control cultures (Figure 7), but |Vmax| was approximately 2-fold higher and tpH 5.25 was 45% less (Table 4). After a 48-h incubation, pHF reached 3.86 ± 0.06 and 5.06 ± 0.03 for prefermented milk and the control, respectively, whereas final cell counts were identical (Table 4).
|
DISCUSSION
|
|---|
The volumetric productivity of the continuous milk prefermentation process with immobilized cells is directly related to the acidifying capacity of bacteria, because milk inflow is used to control milk pH, which was set at 6.0 in this study. The maximum volumetric productivity (12.5 ± 1.5 L/L·h was measured for 35°C, and corresponds to mean residence times of 4.8 or 3.6 min calculated for the entire or extraparticular fermentation volumes, respectively. This temperature corresponds to the optimal growth and milk acidification temperature for L. lactis ssp. lactis biovar. diacetylactis UL719 (Grattepanche et al., 2007). The acidifying capacity of L. lactis ssp. cremoris ATCC19257 was optimum at 30°C (unpublished data), and Lb. rhamnosus RW-9595M exhibited limited growth and lactic acid production in milk over the temperature range from 30 to 38°C (Grattepanche et al., 2007). Sodini-Gallot et al. (1995) reported dilution rates approximately 50% higher for a similar system with an immobilized mixed culture of 4 mesophilic lactic acid bacteria for fresh cheese production operated with similar conditions. Several factors could explain this difference in dilution rate, including different growth and acidifying capacities of lactic cultures in milk, different milk compositions, and lower cell loads in beads. However, the inoculation of milk with 1 to 2 x 108 cfu/mL was very similar in both studies.
The high productivity of the continuous milk prefermentation is explained by the high viable population in beads, which did not vary significantly during 3 wk, but was approximately 40% lower than those reported by Sodini-Gallot et al. (1995) and Sodini et al. (1997a, 1998) during continuous milk prefermentation with 3 lactococci and one leuconostoc. After 2 wk of continuous prefermentation, L. lactis ssp. lactis biovar. diacetylactis UL719 accounted for almost 90% of the total immobilized cell population, with the other 2 strains comprising 5% each. This dominance of L. lactis ssp. lactis biovar. diacetylactis strains in mixed fermentations was observed previously by Sodini et al. (1997b) and Lamboley et al. (2001) during continuous fermentation of milk and whey medium with immobilized cells,
and was tentatively explained by their ability to use citrate. In addition, the high sensitivity of L. lactis ssp. cremoris ATCC19257 and Lb. rhamnosus RW-9595M to nisin Z (MIC of 5 and 11 IU/mL) shown by Grattepanche et al. (2005) may explain their low populations in beads, even though nisin activity was not detected in gel beads or prefermented milks at the respective detection limits of 70 and 40 IU/mL. Lactobacillus rhamnosus RW-9595M had a limited capacity to grow in milk in pure culture with free cells, presumably explained by a deficient proteolytic system (Grattepanche et al., 2007), which could contribute to low populations in beads.
In fermented milks, L. lactis ssp. cremoris ATCC19257 counts decreased with age of prefermentation and were lowest at 35°C. These data could be explained by both the high nisin Z titer in fermented milks and a suboptimal temperature for growth of L. lactis ssp. cremoris (Mundt, 1986). It should be emphasized that real-time PCR methods that use DNA as a target do not distinguish between live and dead cells. However, in a continuous fermentation operated with conditions favoring growth, as in this study, we expect low counts of dead cells. In contrast to immobilized cells, Lb. rhamnosus RW-9595M counts in both prefermented and fermented milks increased with age of prefermentation and accounted for 27% of the total population in prefermented milk on d 20 for a temperature of 35°C. These data could be due to physiological changes of the strain developing with time, as discussed below.
The increase in acidifying capacity of prefermented milks with age of prefermentation and temperature can be partly explained by the increase of Lb. rhamnosus RW-9595M cell counts in prefermented milks with time. In addition, the acidification activities of both L. lactis ssp. lactis biovar. diacetylactis UL719 and Lb. rhamnosus RW-9595M in prefermented milks were shown to be greatly enhanced after 22 d of continuous culture. For samples taken during the last 2 d (d 21 at 27°C, and d 22 at 31°C), 2 |Vmax| values were measured for pHmax, 5.80 ± 0.04 and 4.95 ± 0.08, corresponding to values recorded for pure cultures of L. lactis ssp. lactis biovar. diacetylactis UL719 and Lb. rhamnosus RW-9595M at the same temperatures (Grattepanche et al., 2007).
A shift from homo- to heterofermentative metabolism resulting in a decrease of lactic acid yields has been reported during repeated-batch fermentation of a chemically semidefined medium for lactic acid production by an immobilized culture of Lb. plantarum (Krishnan et al., 2001). Lamboley et al. (1999) also showed a slight but significant increase in acidifying activity of an immobilized mixed mesophilic lactic starter as a function of fermentation time, which was tentatively related to an increase in the immobilized population of L. lactis ssp. lactis biovar. diacetylactis and a change in culture composition with time. In the present study, cell immobilization and continuous fermentation were clearly shown to enhance the acidification properties of L. lactis ssp. lactis biovar. diacetylactis UL719 and Lb. rhamnosus RW-9595M.
Nisin Z activities tested in fermented milk (177 to 363 IU/mL) did not change with prefermentation time or temperature and were in the active range for controlling spoilage microorganisms (Roberts and Zottola, 1993). The lack of detection of nisin Z in prefermented milks can be explained by the high detection limit of the activity test and the very short mean residence time of milk in the continuous reactor. Similarly, Desjardins et al. (2001) showed that nisin production in steady-state conditions of continuous cultures was limited by some biosynthesis steps of the mature peptide, such as posttranslational reactions, transport, or maturation.
We previously showed that the EPS of Lb. rhamnosus RW-9595M improved the structural stability of milk gel at a low concentration of 30 mg/L (Grattepanche et al., 2007). Furthermore, an increase in EPS content from 30.0 to 47.8 mg/L increased the initial apparent viscosity of milk gel. In addition, Doleyres et al. (2005) reported that addition of EPS from Lb. rhamnosus RW-9595M at 125 to 500 mg/L resulted in yogurts with lower yield stress and viscoelastic moduli compared with control yogurts without EPS. Therefore, we can expect that EPS concentration in the fermented milk at 35°C (165 ± 26 mg/L) significantly improved the rheology of the milk gel. The EPS production of Lb. rhamnosus RW-9595M was previously shown to correlate with cell counts for batch and continuous cultures in supplemented whey permeate (Bergmaier et al., 2005). The amount of EPS tested in milks fermented was maximal at 35°C and similar to that reported by Dupont et al. (2000) during milk batch acidification with Lb. rhamnosus RW-9595M at 37°C for 21 h, although Lb. rhamnosus RW-9595M cell counts were approximately 10-fold lower in fermented milks from continuous prefermentation. However, they were 3.5-fold higher than for milk inoculated with mixed culture (108 cfu/mL of each strain) and batch fermented at 34°C because of nisin Z production by L. lactis ssp. lactis biovar. diacetylactis UL719 and high nisin sensitivity (Grattepanche et al., 2007). In this work, Lb. rhamnosus RW-9595M cell counts in fermented milk largely increased with age of prefermentation, even though high nisin Z titers (>177 IU/mL) were produced. These data may be explained by the large increase in nisin Z tolerance shown for this strain (particularly for high bacteriocin concentrations), which was reversible during 8 successive subcultures.
Doleyres et al. (2004) previously reported that cell immobilization and continuous fermentation increased the tolerance of Bifidobacterium longum and L. lactis ssp. lactis biovar. diacetylactis strains to different environmental stresses, including nisin, antibiotics, H2O2, and simulated gastric and intestinal juices. These authors suggested that conditions occurring in gel beads, such as limited diffusion of substrate and inhibitory products, leading to a steep pH gradient and quorum sensing, can induce a nonspecific stress adaptation of immobilized cells. Proteomic and transcriptomic studies have shown that cell immobilization modified expression of genes implicated at all cellular levels: membrane, metabolism, DNA replication, transcription, translation and elongation factors, motility, adaptation, protection, protein folding, and nucleotide biosynthesis (Donlan, 2002; Junter and Jouenne, 2004). Additional work is needed to elucidate the stress adaptation mechanisms of immobilization.
To conclude, continuous prefermentation with immobilized cells allowed high milk inoculation and volumetric productivity that could likely be further improved by selecting a nisintolerant acidifying strain. Lactococcus lactis ssp. lactis biovar. diacetylactis UL719 cell counts and nisin production were high in fermented milk throughout the 22-d continuous fermentation. Moreover, it was possible to control populations of Lb. rhamnosus RW-9595M in fermented milk by changing the incubation temperature. A high concentration of nisin-sensitive strain Lb. rhamnosus RW-9595M was measured in fermented milks, leading to an enhanced amount of EPS in fermented milk compared with traditional inoculation of milk. We showed that cell immobilization and continuous fermentation resulted in large improvements in nisin tolerance for this strain and acidification parameters for both Lb. rhamnosus RW-9595M and L. lactis ssp. lactis biovar. diacetylactis UL719. This technology permitted the production of fermented milks with EPS and nisin Z concentrations, which are relevant for their effects on texture and the control of bacterial contaminants.
|
ACKNOWLEDGEMENTS
|
|---|
The authors wish to thank the Natural Sciences and Engineering Research Council of Canada (Research Partnerships Program–Research Network on Lactic Acid Bacteria), Agriculture and Agri-Food Canada, Novalait Inc., Dairy Farmers of Canada, and Institut Rosell Inc. for financial support.
Received for publication April 11, 2007.
Accepted for publication August 21, 2007.
|
REFERENCES
|
|---|
Bergmaier, D., C. P. Champagne, and C. Lacroix. 2003. Exopolysaccharide production during batch cultures with free and immobilized Lactobacillus rhamnosus RW-9595M. J. Appl. Microbiol. 95:1049–1057.[CrossRef][Medline]
Bergmaier, D., C. P. Champagne, and C. Lacroix. 2005. Growth and exopolysaccharide production during free and immobilized cell chemostat culture of Lactobacillus rhamnosus RW-9595M. J. Appl. Microbiol. 98:272–284.[CrossRef][Medline]
Cleveland, J., T. J. Montville, I. F. Nes, and M. L. Chikindas. 2001. Bacteriocins: Safe, natural antimicrobials for food preservation. Int. J. Food Microbiol. 71:1–20.[CrossRef][Medline]
Desjardins, P., J. Meghrous, and C. Lacroix. 2001. Effect of aeration and dilution rate on nisin Z production during continuous fermentation with free and immobilized Lactococcus lactis UL719 in supplemented whey permeate. Int. Dairy J. 11:943–951.[CrossRef]
De Vuyst, L., F. Devin, F. Vaningelgem, and B. Degeest. 2001. Recent developments in the biosynthesis and applications of heteropoly-saccharides from lactic acid bacteria. Int. Dairy J. 11:687–707.[CrossRef]
De Vuyst, L., and E. J. Vandamme. 1994. Nisin, a lantibiotic produced by Lactococcus lactis subsp. lactis: Properties, biosynthesis, fermentation and applications. Pages 151–221 in Bacteriocins of Lactic Acid Bacteria. L. De Vuyst and E. J. Vandamme, ed. Blackie Academic & Professional, London, UK.
Doleyres, Y., I. Fliss, and C. Lacroix. 2004. Increased stress tolerance of Bifidobacterium longum and Lactococcus lactis produced during continuous mixed-strain immobilized-cell fermentation. J. Appl. Microbiol. 97:527–539.[CrossRef][Medline]
Doleyres, Y., and C. Lacroix. 2005. Technologies with free and immobilized cells for probiotic bifidobacteria production and protection. Int. Dairy J. 15:973–988.[CrossRef]
Doleyres, Y., L. Schaub, and C. Lacroix. 2005. Comparison of the functionality of exopolysaccharides produced in situ or added as bioingredients on yogurt properties. J. Dairy Sci. 88:4146–4156.[Abstract/Free Full Text]
Donlan, R. M. 2002. Biofilms: Microbial life on surfaces. Emerg. Infect. Dis. 8:881–890.[Medline]
Dupont, I., D. Roy, and G. Lapointe. 2000. Comparison of exopolysaccharide production by strains of Lactobacillus rhamnosus and Lactobacillus paracasei grown in chemically defined medium and milk. J. Ind. Microbiol. Biotechnol. 24:251–255.[CrossRef]
Grattepanche, F., P. Audet, and C. Lacroix. 2007. Milk fermentation by functional mixed culture producing nisin Z and exopolysaccharides in a fresh cheese model. Int. Dairy J. 17:123–132.[CrossRef]
Grattepanche, F., C. Lacroix, P. Audet, G. La, and G. Pointe. 2005. Quantification by real-time PCR of Lactococcus lactis subsp. cremoris in milk fermented by a mixed culture. Appl. Microbiol. Bio-technol. 66:414–421.[CrossRef][Medline]
Jolly, L., S. J. F. Vincent, P. Duboc, and J. R. Neeser. 2002. Exploiting exopolysaccharides from lactic acid bacteria. Antonie Van Leeuwenhoek 82:367–374.[CrossRef][Medline]
Junter, G. A., and T. Jouenne. 2004. Immobilized viable microbial cells: From the process to the proteome . . . or the cart before the horse. Biotechnol. Adv. 22:633–658.[CrossRef][Medline]
Kempler, G. M., and L. L. McKay. 1980. Improved medium for detection of citrate-fermenting Streptococcus lactis subsp. diacetylactis. Appl. Environ. Microbiol. 39:926–927.[Abstract/Free Full Text]
Klaenhammer, T. R. 1993. Genetics of bacteriocins produced by lactic acid bacteria. FEMS Microbiol. Rev. 12:39–85.[Medline]
Krishnan, S., L. R. Gowda, M. C. Misra, and N. G. Karanth. 2001. Physiological and morphological changes in immobilized L. plantarum NCIM 2084 cells during repeated batch fermentation for production of lactic acid. Food Biotechnol. 15:193–202.[CrossRef]
Lacroix, C., F. Grattepanche, Y. Doleyres, and D. Bergmaier. 2005. Immobilised cell technologies for the dairy industry. Pages 295–319 in Applications of Cell Immobilisation Biotechnology. Focus on Biotechnology Series. Vol. 8b. V. Nedovic and R. Willaert, ed. Springer Verlag, New York, NY.
Lamboley, L., C. Lacroix, J. M. Artignan, C. P. Champagne, and J. C. Vuillemard. 1999. Long-term mechanical and biological stability of an immobilized cell reactor for mixed-strain mesophilic lactic starter production in whey permeate. Biotechnol. Prog. 15:646–654.[CrossRef][Medline]
Lamboley, L., C. Lacroix, C. P. Champagne, and J. C. Vuillemard. 1997. Continuous mixed strain mesophilic lactic starter production in supplemented whey permeate medium using immobilized cell technology. Biotechnol. Bioeng. 56:502–516.[Medline]
Lamboley, L., C. Lacroix, I. Sodini, M. J. Lemay, and C. P. Champagne. 2001. Effect of inoculum composition and low KCl supplementation on the biological and rheological stability of an immobilized-cell system for mesophilic lactic starter production. Biotechnol. Prog. 17:1071–1078.[CrossRef][Medline]
Meghrous, J., C. Lacroix, M. Bouksaim, G. Lapointe, and R. E. Simard. 1997. Genetic and biochemical characterization of nisin Z produced by Lactococcus lactis ssp. lactis biovar. diacetylactis UL719. J. Appl. Microbiol. 83:133–138.[CrossRef][Medline]
Mundt, J. O. 1986. Lactic acid streptococci. Pages 1063–1071 in Bergey’s Manual of Systematic Bacteriology. Vol. 2. P. H. A. Sneath, N. S. Mair, M. E. Sharpe and G. J. Holt, ed., Williams and Wilkins, Baltimore, MD.
Roberts, R. F., E. A. Zottola, and L. L. McKay. 1992. Use of a nisin-producing starter culture suitable for Cheddar cheese manufacture. J. Dairy Sci. 75:2353–2363.[Abstract]
Roberts, R. F., and E. A. Zottola. 1993. Shelf-life of pasteurized process cheese spreads made from Cheddar cheese manufactured with a nisin-producing starter culture. J. Dairy Sci. 76:1829–1836.[Abstract]
Ruas-Madiedo, P., J. Hugenholtz, and P. Zoon. 2002. An overview of the functionality of exopolysaccharides produced by lactic acid bacteria. Int. Dairy J. 12:63–171.
Sodini, I., C. Y. Boquien, G. Corrieu, and C. Lacroix. 1997a. Microbial dynamics of co- and separately entrapped mixed cultures of mesophilic lactic acid bacteria during the continuous prefermentation of milk. Enz. Microb. Technol. 20:381–388.[CrossRef]
Sodini, I., C. Y. Boquien, G. Corrieu, and C. Lacroix. 1997b. Use of immobilized cell bioreactor for the continuous inoculation of milk in fresh cheese manufacturing. J. Ind. Microbiol. Biotechnol. 18:56–61.[CrossRef][Medline]
Sodini, I., L. Lagace, C. Lacroix, and G. Corrieu. 1998. Effect of continuous prefermentation of milk with an immobilized cell bio-reactor on fermentation kinetics and curd properties. J. Dairy Sci. 81:631–638.[Abstract]
Sodini-Gallot, I., G. Corrieu, C. Y. Boquien, E. Latrille, and C. Lacroix. 1995. Process performance of continuous inoculation and acidification of milk with immobilized lactic acid bacteria. J. Dairy Sci. 78:1407–1420.[Abstract]
Tagg, J. R., A. S. Dajani, and L. W. Wannamaker. 1976. Bacteriocins of gram-positive bacteria. Bacteriol. Rev. 40:722–756.[Free Full Text]
Van Calsteren, M. R., C. Pau-Roblot, A. Bégin, and D. Roy. 2002. Structure determination of the exopolysaccharide produced by Lactobacillus rhamnosus RW-9595M and R. Biochem. J. 363:7–17.[CrossRef][Medline]
TOP
ABSTRACT
INTRODUCTION
), estimated by quantitative PCR, Lactococcus lactis ssp. lactis biovar. diacetylactis UL719 (
, •) and Lactobacillus rhamnosus RW-9595M (
), determined by plate counts, during continuous immobilized cell culture at different temperatures. For prefermented milk, data with more than a 1-d interval corresponded to weekend interruptions. The dashed lines correspond to the effects of age of prefermentation shown on L. lactis ssp. cremoris ATCC19257 (• • • •) and Lb. rhamnosus RW-9595M (—) cell counts predicted by the regression model for a temperature of 31°C. Bars represent standard deviations calculated from triplicate analyses.
) and 45°C (