| Collection Mémoires et thèses électroniques | |
| AccueilÀ proposNous joindre |
| Chapitre 3 Exopolysaccharide production during batch cultures with free and immobilized Lactobacillus rhamnosus RW-9595M | ||
|---|---|---|
 |  | |
Table des matières
Dirk Bergmaier 1 , Claude P. Champagne 2 and Christophe Lacroix 3
1 STELA Dairy Research Centre, Pavillon Paul Comtois, Université Laval, Québec, PQ, Canada, G1K 7P4
2 Food Research and Development Center, Agriculture and AgriFood Canada, St-Hyacinthe, PQ, Canada, J2S 8E3
3 Laboratory of Food Biotechnology, Institute of Food Science and Nutrition, Swiss Federal Institute of Technology, ETH Zentrum, LFO F18, CH-8092 Zurich, Switzerland
La production d'exopolysaccharide (EPS) par Lactobacillus rhamnosus RW-9595M a été étudiée lors des cultures en batch à pH contrôlé avec des cellules libres et lors des cultures en batch répété avec des cellules immobilisées par adsorption sur des supports poreux solides (ImmobaSil®). Les cultures ont été conduites à pH 6 dans un milieu de perméat de lactosérum supplémenté (SWP), contenant du perméat de lactosérum à une concentration de 5% ou de 8% (w/w). Pour des cultures en batch avec les cellules libres dans le milieu de 8% SWP, des comptes cellulaires (1.3 · 1010 CFU/ml) et une production d'EPS (2350 mg/l) très élevées ont été mesurées après respectivement 18 et 32 h de culture. Plusieurs méthodes pour la détermination de la biomasse immobilisée sur des supports solides basées sur l'analyse des composants de biomasse (protéines, ATP et ADN) ont été examinées. La méthode d'analyse d'ADN s'est avérée la plus appropriée dans ces circonstances. Cette méthode a dosé une haute biomasse immobilisée de 8.5 · 1011 CFU/ml après la colonisation des supports ImmobaSil®. Pendant les cultures en batch répété avec des cellules immobilisées, une concentration élevée d'EPS (1808 mg/l) a été mesurée après quatre cycles pour une courte période d'incubation de 7 h dans 5 % de SWP. La biomasse élevée dans le système de cellules immobilisées a augmenté la productivité volumétrique maximale d'EPS (258 mg/l·h après 7 h) comparée aux cultures en batch avec des cellules libres (110 mg/l·h après 18 h). Notre étude montre clairement le potentiel élevé de la souche et de la technologie des cellules immobilisées pour la production d'EPS comme ingrédient fonctionnel alimentaire.
The production of exocellular polysaccharides (EPS) by Lactobacillus rhamnosus RW-9595M was investigated during pH-controlled batch cultures with free cells and repeated-batch cultures with cells immobilized by adsorption on solid porous supports (ImmobaSil®). Cultures were conducted at pH 6 in supplemented whey permeate (SWP) medium containing 5% or 8% (w/w) whey permeate. For free-cell batch cultures in 8% SWP medium, very high maximum cell counts (1.3 · 1010 CFU/ml) and EPS production (2350 mg/l) were measured after 18 and 32 h, respectively. Several methods for the determination of immobilized biomass on solid supports based on analysis of biomass components (proteins, ATP and DNA) were tested. The DNA analysis method proved to be the most appropriate under these circumstances. This method revealed a high immobilized biomass of 8.5 · 1011 CFU/ml support after colonization of the ImmobaSil® supports. During repeated immobilized cell cultures, a high EPS concentration (1808 mg/l) was measured after four cycles for a short incubation period of 7 h in 5 % SWP. The high biomass in the immobilized cell system increased maximum EPS volumetric productivity (258 mg/l·h after 7 h culture) compared with free cell batch cultures (110 mg/l·h after 18 h culture). Our study clearly shows the high potential of the strain and immobilized cell technology for production of EPS as a functional food ingredient.
Polysaccharides are widely used in the food industry as thickeners, viscosifiers, stabilizing or emulsifying agents, gelling agents or texturizers (Sutherland 1998). The majority of these additives are of plant origin, such as starch, locust bean gum and alginate. Until now, xanthan, curdlan and gellan have been the only FDA approved microbial polysaccharides although they are produced by non-GRAS (Generally Recognized As Safe) bacteria. Many food-grade microorganisms, such as lactic acid bacteria (LAB), also produce exocellular polysaccharides (EPS) (Cerning 1990; De Vuyst and Degeest 1999b). In situ-produced EPS play an important role in the manufacture of fermented dairy products such as yogurt, drinking yogurt, cheese, fermented cream and milk-based desserts (Duboc and Mollet 2001). They contribute to the texture, mouthfeel, taste perception and stability of the final products. EPS act as texturizers and stabilizers and so decrease syneresis and improve product stability. Furthermore, some EPS classified as prebiotics could contribute to human health and positively affect gut microflora (Ruas-Madiedo et al. 2002a).
In spite of these interesting properties, EPS from LAB are not yet exploited by food manufacturers as food additives (De Vuyst and Degeest 1999a). Their use could result in safe, natural end-products with improved rheological properties and enhanced stability, especially for the dairy industry. However, compared with dextran-producing Leuconostoc or Gram-negative EPS producers, the low production of EPS by LAB is a constraint for their commercial use as food additives. In this respect, fermentation technologies that can be used in the industry need to be developed in order to enhance EPS production and develop new EPS bioingredients.
It has been shown that immobilized cell technology (ICT) with LAB could increase the biomass concentration maintained in the fermentation system (Champagne et al. 1994; Norton et al. 1994a). Coupled with repeated-batch or continuous culture, ICT could largely increase process productivity and cell stability (Bertrand et al. 2001; Lamboley et al. 1997). Cell immobilization also protects cells against shear damage. Furthermore, reuse of cells and more constant product characteristics could facilitate downstream processing of EPS.
Immobilization by adsorption on porous preformed and inert supports could be an immobilization technique suited for EPS-producing cultures. The production of capsular and free EPS by the cells may promote their adsorption to the support and subsequent colonization (Dunne 2002). However, it is difficult to accurately determine the concentration of immobilized cells in solid supports, an important parameter of the fermentation process (Wang 1988). Disruption of the immobilization matrix and viable cell enumeration, commonly used with gel beads, will result in very high shear stress and mortality of cells, whereas biomass dry weight determinations will quantify not only biomass but also substances accumulated in the support matrix, such as EPS.
In this study, different methods for quantification of immobilized biomass, based on the extraction and analysis of specific components or molecules of the biomass such as proteins, adenosine triphosphate (ATP) and deoxyribonucleic acid (DNA), were compared. Bacterial growth and EPS production during batch and repeated-batch cultures with free and immobilized cells of Lb. rhamnosus RW-9595M, a strain that has shown very high EPS production in a whey permeate-based medium (Bergmaier et al. 2001; Macedo et al. 2002b) was studied. The mucoid properties of Lb. rhamnosus RW-9595M were used to promote adsorption on solid porous supports (ImmobaSil®) for cell immobilization.
Lb. rhamnosus RW-9595M, obtained from the Lactic Acid Bacteria Research Network (RBL Network) culture collection (Dairy Research Centre STELA, Université Laval, Québec, PQ, Canada), was isolated from reference culture Lb. rhamnosus ATCC 9595 and shown to produce very high EPS concentrations in the defined BMM medium (Dupont et al. 2000) and in supplemented whey permeate medium (Bergmaier et al. 2001; Macedo et al. 2002b). The strain was subcultured aerobically in supplemented whey permeate for 36 h at 37°C. The stock culture was kept frozen in 6% (v/v) rehydrated skim milk and 10% (v/v) glycerol at -80°C. For inoculum preparation, a fresh culture was propagated twice in 5% whey permeate with 1% yeast extract at 37°C for 10 h or until pH 4.8 was reached.
The supplemented whey permeate (SWP) medium was prepared as follows. Whey permeate powder (Foremost, Baraboo, WI, USA) was rehydrated to give a final concentration of 5% or 8% (w/w) and the pH was adjusted to pH 5.0 with 1 N HCl. The solution was autoclaved (121°C, 15 min), cooled to room temperature and filtered on a AP25 prefilter (Millipore Corporation, Bedford, MA, USA). The other components MgSO4 · 7H2O, MnSO4 · H2O and Tween-80, were added to give final concentrations of 0.5 g/l, 0.05 g/l and 1 ml/l, respectively, and the solution was autoclaved again. A yeast extract (YE) solution was prepared by adding 40 g of YE (Rosell Institute Inc., Montreal, PQ, Canada) to 100 ml deionized water. This YE concentrated solution was autoclaved (121°C, 15 min) and added to the whey permeate medium to give a final YE concentration of 1%.
Fermentations were carried out in a 1.5 l bioreactor (Bioflo III, New Brunswick Scientific, Edison, NY, USA) with a total working volume of 1.2 l. Temperature was controlled at 37°C and pH at 6 by addition of 6 N NH4OH. Free-cell batch cultures were performed in 5 or 8% SWP medium with 1.5% (v/v) inoculum size. Samples of culture broth were taken for analysis at different time intervals, depending on the base addition rate. Cells for DNA analysis were sampled during exponential growth phase as estimated by the observed base addition rate.
For repeated batch cultures with immobilized cells in 5% SWP medium, the bioreactor was filled with 150 ml ImmobaSil® supports (Ashby Scientific Ltd., Coalville, Leicestershire, UK). After support conditioning according to the manufacturer's instructions, the bioreactor was filled aseptically with 500 ml SWP and inoculated at 4% (v/v). A stirring regime of 5 min at 50 rpm followed by 30 min with no stirring was applied for 4 h. After this time, the first fermentation was started with stirring set at 100 rpm and pH controlled at 6. After 8 h, fresh medium was added to the bioreactor for a total volume of 1.2 l. The fermentation medium was totally replaced when a volume of 80 ml 6 N NH4OH was used for pH control, which corresponded to late exponential growth phase of the culture. A total of four repeated batch cultures were performed.
Viable cell counts (CFU per milliliter) were estimated by plating diluted samples on solid MRS agar (Difco Laboratories, Richmond, CA, USA). Plates were incubated aerobically at 37°C for 36 h. Reported data are means for triplicate analyses. For determination of the cell dry weight in the broth, 10 ml of broth were centrifuged at 10,000 g for 15 min at 4°C (Sorval Instrument, Dupont, Newtown, CO, USA). The cell pellet was washed twice with distilled water and then transferred to a preweighed aluminium dish. The cells were dried to a constant weight in a vacuum oven at 70°C. Reported data are means for duplicate analyses.
The immobilized cell concentration was determined by measuring the DNA content using the method described by Wang (1988) with some modifications. For calibration, the DNA content of cells was correlated with viable cell counts and biomass dry weight using effluent samples from free-cell pH-controlled batch cultures with cells in the exponential growth phase. As a consequence of the high EPS production by Lb. rhamnosus RW-9595M, it was impossible to accurately determine the cell dry weight (fluffy pellet and obstruction of filters). Therefore, the strain Lb. rhamnosus ATCC 9595 was used for this part of the study. This strain is the parent strain from which the variant strain Lb. rhamnosus RW-9595M was isolated and it produces only a very small amount of EPS (Dupont et al. 2000).
For DNA extraction, the fermented medium was centrifuged (10,000 g at 4°C for 15 min). The pellet was resuspended in 10 ml of a saline-citric solution (SCS) containing 8.77 g/l NaCl and 1.47 g/l Na3C6H5O7 · 2H2O at pH 7 and centrifuged again. The pellet was resuspended in 2 ml of SCS solution with 0.25 N perchloric acid (pH 1.45) and incubated for 30 min at 0°C. After centrifugation (10,000 g at 0°C for 10 min), 4 ml 0.5 N perchloric acid solution was added to the pellet, the tube was sealed and incubated in a water bath at 70°C for 30 min. The tubes were cooled to room temperature for exactly 25 min and centrifuged (5,000 g, 10 min, 20°C). The supernatant was kept at 4°C for a maximum of 1 day until analysis. The DNA concentration was determined by the diphenylamine method (Burton 1956). The supernatant was diluted with 0.5 N perchloric acid to be in the standard curve concentration range. Two ml of diphenylamine reagent, prepared the same day with 1.5 g diphenylamine in 100 ml demineralized water, 1.5 ml concentrated H2SO4 and 0.5 ml 1.6% acetaldehyde, were added to a 1-ml diluted sample. A standard curve of deoxyribose in a concentration range from 0 to 50 μg/ml was prepared. Standards and samples were incubated at 30°C for 17 h. Absorbance was measured spectroscopically at 595 nm with background correction at 650 nm. Linear correlations between DNA content, cell dry weight and cell counts were determined.
To determine the DNA concentration in the solid supports, a sample of about 3 to 4 supports was taken and the wet weight was determined. The supports were washed twice with SCS solution and dried on a filter paper. The supports were transferred into glass tubes with 3 ml SCS and 2.5 N perchloric acid was added to give a final concentration of 0.25 N (pH 1.45). After incubation at 0°C for 30 min, the excess liquid was removed on a filter paper. For DNA extraction the supports were transferred to a 10 ml centrifugation tube and 3 ml 0.5 N perchloric acid were added. The samples were incubated in a water bath at 70°C for 25 min and cooled to room temperature for 25 min. The DNA content of the clear supernatant was analyzed by the diphenylamine method, as described above.
As only wet weight of colonized supports was determined during the DNA extraction method, a correlation between support wet and dry weights and support volume was established. The wet weight of SCS washed colonized supports was determined in the same way as during DNA extraction. Support volume was measured by water displacement in a graduated cylinder and the supports were dried in a vacuum oven at 70°C until constant weight was obtained.
The immobilized cell concentration was calculated with the correlation between DNA content, viable cell counts and biomass, assuming that the DNA content per cell of immobilized Lb. rhamnosus RW-9595M and free cells of Lb. rhamnosus ATCC-9595 is the same per cell. Immobilized cell concentration was expressed as colony-forming units or cell dry weight per support dry weight (CFU/g or g/g, respectively). Each analysis was repeated three times.
The quantity of ATP in a sample containing planktonic cells was determined using the firefly bioluminescence method as described by Gikas and Livingston (1993). For ATP extraction from cells, one ml of sample was mixed with 9 ml of a Tris-EDTA buffer and incubated at 100°C for 3 min. Then the extracted ATP reacted with luciferin (LH2) in a reaction catalyzed by the luciferase enzyme and produced luminescence that was measured using a Wallac Victor-1420 microplate reader. The ATP Bioluminescence Assay Kit CLS II (Boehringer Mannheim GmbH, Germany) was used according to the manufacturers standard protocol.
Broth samples (25 ml) were centrifuged (7500 g at 4°C for 15 min) and the pellet was washed three times with 0.95% NaCl solution. For protein extraction, 10 ml 0.1 M NaOH were added to the pellet and the sample was heated at 100°C for 30 min. After centrifugation (7500 g at 4°C for 15 min) the protein concentration in the supernatant was determined using the DC Protein Assay (Bio-Rad Laboratories, Hercules, CA, USA) with bovine serum albumin as standard. For determination of protein content of immobilized cells, 1 g of support was treated as described for cell pellets except that cells were ultrasonicated (Aquasonic Model 250T, VWR Scientific, West Chester, PA) for 2 h in 0.1 M NaOH before the heating step.
Homogeneous culture broth samples were heated to 100°C for 15 min, cooled and centrifuged (13,200 g at 4°C for 25 min). Supernatants were stored at -20°C. Polysaccharides in supernatant were partially purified by a new ultrafiltration method (Bergmaier et al. 2001) (Chapitre 2), and total carbohydrate was quantified by the phenol-sulfuric acid method (Dubois et al. 1956) with glucose as standard. The concentration of EPS produced during fermentation was calculated by subtracting the polysaccharide concentration in the fresh medium from that measured in the fermented broth.
Culture broth samples were centrifuged (13,200 g at 4°C for 25 min), diluted and filtered on an Acrodisc LC13 filter (0.2 µm; Gelman Sciences, Ann Arbor, MI, USA). Residual sugars and lactic acid concentration in the supernatant were determined by HPLC (Waters, Millipore Co. Montreal, QC, Canada) using a Phenomenex ion column (Phenomenex, Torrance, CA) and H2SO4 0.0064N as eluent at a flow rate of 0.4 ml/min. Analysis was performed in duplicate.
In this study, pH-controlled batch cultures in a whey permeate based medium with Lb. rhamnosus RW-9595M were carried out. Bacterial growth and EPS production during free-cell batch cultures were compared with repeated-batch cultures with cells immobilized on solid porous supports.
Data obtained for two free-cell batch fermentations in 8% SWP medium are presented in Fig. 3.1. The free-cell counts increased rapidly during the pH-controlled culture and reached a very high maximum value of 1.31 ± 0.05·1010 CFU/ml after 18 h. This data shows the suitability of the whey-permeate-based medium for biomass growth. Lactose as the only carbon source in this medium with an initial concentration of 64.9 ± 0.3 g/l was completely consumed after 32 h culture. Lactic acid was the only organic acid produced, for a final concentration in the fermented broth of 58.7 ± 0.8 g/l. When the maximum population was reached (18 h), only 43 g/l of lactic acid had been produced (Fig. 3.1). Uncoupling between growth and lactic acid production was thus observed after 18 h of fermentation, and approximately 25% of acid production occurred during this phase. Such uncoupling between growth and acidification has been reported with lactobacilli (Oyaas et al. 1996). Although EPS production kinetics were similar to that of lactic acid production, it is noteworthy that little EPS production was observed after 18 h of fermentation. Therefore, in free-cell fermentations, EPS production appeared more closely linked to growth than was acidification. After 32 h culture, a high total polysaccharide concentration in the fermented broth of 2728 ± 38 mg/l was measured. The final EPS production calculated by subtracting the initial polysaccharide content in the medium (378 ± 13 mg/l), which is mainly due to YE supplementation (Bergmaier et al. 2001; Torino et al. 2000), was 2350 ± 51 mg/l. The EPS volumetric and specific productivities were maximum after 18 h, equal to 110.3 mg/l·h and 8.45·10-12 mg/CFU·h, respectively.
Fig. 3.1 Viable cell counts (●), lactose (■) and lactic acid (◆) concentrations, and EPS production (▲) during pH-controlled batch cultures with Lb. rhamnosus RW-9595M. Means and standard deviations for two fermentations.
The EPS production obtained in this study, more than 2.3 g/l, is one of the highest reported for lactobacilli in the literature. Only Macedo et al. (2002b) reported higher values (2775 mg/l) for the same strain in a complex and rich medium composed of whey permeate supplemented with yeast extract and vitamins, salts and amino acids of the defined BMM medium (Morishita et al. 1981). However, the EPS production by Lb. rhamnosus RW-9595M is still much lower than the high yields obtained with dextran-producing LAB and Gram-negative EPS producers such as X. campestris with very high productions of 10-25 g/l (Becker et al. 1998). De Vuyst and Degeest (1999b) stated that, from an economic point of view, a production of 10-15 g/l would be required in order to use EPS from LAB as functional (thickener, gelling agent) food additives.
In order to increase biomass in the fermentation system, and thus achieve higher productivity, Lb. rhamnosus RW-9595M cells were immobilized on ImmobaSil® solid porous supports which were used during repeated pH-controlled cultures. Data obtained for four consecutive batch cultures with immobilized cells in 5% SWP medium are presented in Fig. 3.2. Each culture was carried out until about 80 ml of 6 N base was used to control pH at 6. This base volume corresponds to a theoretical lactic acid production of about 41 g/l, and to nearly complete conversion of lactose in the medium (43 g/l) into lactic acid (Bergmaier et al. 2001) (Chapitre 2). This was verified by HPLC analysis of samples taken at the end of the four consecutive batch cultures. In these samples, residual lactose concentration ranged between 0 and 5 g/l and a high concentration of lactic acid between 33 and 38 g/l was tested.
With this repeated-cycle batch fermentation strategy, a decrease of fermentation time was observed with an increase in the cycle number, with 16, 9, 7, and 7 h from the first to the fourth batch culture. The enumeration of free-cells in the fermentation broth showed very high cell concentrations at the end of incubation before medium change, with a mean of 1.02 ± 0.14·1010 CFU/ml (Fig. 3.2). This cell concentration is equal (p<0.05) to that obtained during batch cultures in 8% SWP medium in this study (1.31 ± 0.05 · 1010 CFU/ml) and to that measured during batch fermentations in 5% SWP (1.1 ± 0.1 · 1010 CFU/ml) (Bergmaier et al. 2001) (Chapitre 2). Between two consecutive cultures the medium was completely changed and the new medium was mixed with the supports. The mean free-cell count in the broth after this change was high (3.9 ± 0.5 · 108 CFU/ml) probably due to incomplete broth removal in the void spaces of the support and cell release from the support. The rapid build up of free cells during incubation resulted from both free cell growth and immobilized cell growth associated with cell release from the solid supports (Lamboley et al. 1997).
The knowledge of the content of immobilized biomass in solid supports is important for process analysis and optimization. However, conventional methods such as dry weight or viable cell enumeration are not suitable for cells immobilized in solid supports, particularly for EPS-producing strains. Biomass dry weight determination is greatly influenced by products accumulated in the porous support and viable cell count determination requires an initial step to release immobilized cells with no loss of viability before cell enumeration, which is not possible with solid supports. Therefore, several methods for immobilized biomass estimation were tested in this study. Specific cell components, proteins, ATP and DNA, were quantified during free-cell cultures and calibration curves were developed to relate the component concentration with viable cell counts or biomass concentration. These calibrations could then be used to estimate the immobilized biomass or cell concentration from the specific component concentration tested in the solid support.
One major constituent of microbial cells are proteins. About 57% of the dry weight of a typical bacterium is protein (Atkinson and Mavituna 1991). In the present study, the amount of proteins in planktonic cells was determined and correlated with biomass. An amount of 252.4 mg proteins per g dry weight of bacteria was determined (results not shown). However, protein extraction experiments with colonized carriers resulted in an important variance for the value of immobilized biomass (results not shown), probably due to incomplete removal of contaminating proteins contained in the medium or to interference of peptidoglycans. Similar results were also observed for immobilized Acinetobacter calcoaceticus cells (Wang 1988). In contrast, Chen et al. (1998) evolved and successfully used a protein detection method for sludge biomass estimation of cells immobilized in phosphorylated polyvinyl alcohol gel beads. Also Freeman et al. (1982) developed and used an assay based on protein content determination in order to quantify immobilized biomass in synthetic and native polymer-gel beads.
In ecological research the extraction and analysis of ATP is often used for biomass estimation (Holm-Hansen and Karl 1987; Karl 1980). There are two important assumptions for the use of this method: first, that ATP exists only in viable cells and is rapidly degraded after the cell dies (Holm-Hansen and Karl 1987); and second, that the quantity of ATP in the bacterial cell does not vary significantly for a particular strain (Stanley 1986). Thus, the viable cell concentration can be estimated by the amount of cellular ATP. Gikas and Livingston (1993) used ATP content and the dry weight of planktonic and immobilized cells growing on 3,4-dichloraniline to evaluate cell viability. The biomass during continuous culture on whey permeate of Lb. casei cells immobilized on sintered glass particles was estimated by the estimation of cellular ATP (Krischke et al. 1991).
In our study, culture broth samples from free-cell cultures with Lb. rhamnosus RW-9595M were analyzed for ATP content. Preliminary data showed that the amount of ATP per cell decreased dramatically (from 1.2 · 10-19 to 2 · 10-20 mol ATP/CFU) during the exponential growth phase (see Annexe 1 ). For this reason no further experiments were made with this method. Chappelle et al. (1987) also pointed out that the ATP concentration in bacterial cells varies over the growth cycle and decreases as the population ages.
Another important component of viable cells is DNA. The DNA extraction method was used for biomass determination in the domain of waste water treatment (Raebel and Schlierf 1980) and marine sediment characterization (Dell'anno et al. 1998). The quantification of Acinetobacter calcoaceticus immobilized on Celite (Wang 1988) and Comamonas acidovorans immobilized on different supports (Buchtmann et al. 1997) was also tested by a DNA method.
In this study, the microbial DNA extraction and quantification method was used to estimate biomass immobilized on ImmobaSil® carriers. Correlations between DNA content, dry weight and free-cell counts of Lb. rhamnosus ATCC-9595 were first established. This low EPS producer was used because Lb. rhamnosus RW-9595M, a natural mutant isolated from Lb. rhamnosus ATCC-9595, produces very high amounts of EPS that interfere with cell separation. Cell samples from the exponential growth phase of free-cell pH-controlled batch cultures with concentrations from 1.7·108 to 1.3·1010 CFU/ml were used to develop calibration curves between DNA content and dry weight or viable cell counts. Highly significant (p<0.001) correlations were obtained:
(2)
(3)
where Xw and Xc are the viable cell count (CFU/ml) and dry biomass concentration (mg/ml), respectively, and CDNA is the DNA concentration (µg/ml).
After extraction, the amount of DNA in colonized supports during the four consecutive cultures was measured and used to estimate biomass growth in the supports. After only a 4-h stirring regime, an immobilized biomass of about 1.1 · 109 CFU/ml support was measured, indicating the good adhesion properties of Lb. rhamnosus RW-9595M cells on the silicon rubber supports (Fig. 3.2). The immobilized biomass grew continuously during the four successive batch cultures and reached a very high value of 8.5 · 1011 CFU/ml support at the end of the fourth culture.
This result was obtained with the assumptions that DNA content of parental strain, Lb. rhamnosus ATCC-9595, used for calibration of this method was the same as for the EPS producing variant Lb. rhamnosus RW-9595M and that DNA content of immobilized cells was the same as for cells produced during free-cell cultures. Lb. rhamnosus ATCC 9595 was used for method calibration, because it produced only very low amounts of EPS compared with the variant strain RW-9595M. The high production of EPS by this strain prevented the separation of cells from the broth medium by physical techniques (filtration, centrifugation), which is required for the development of a calibration curve for the method. A very rapid degradation of DNA was observed for this culture with cell death as a result of a very active DNA degrading enzyme (Peant and LaPointe 2001). Therefore, the immobilized cell content estimated with this method mainly corresponded to living cell DNA that was not degraded.
The viable immobilized biomass tested in our study (8.5 · 1011 CFU/ml support) is high, compared with data previously reported for other LAB, although it is difficult to compare data for immobilized biomass, because different microorganisms, immobilization procedures, reactor designs and measurement methods were used. For example, immobilized viable cell counts of 2.7 · 1011 CFU/ml gel beads for Lactobacillus casei subsp. casei (Arnaud et al. 1992) and of 1.3 · 1011 CFU/g beads for three strains of lactococci (Lamboley et al. 1997) immobilized in κ-carrageenan/locust bean gum gel beads were measured during continuous fermentation. Modelling of the density of Lactococcus lactis cells in gel matrix indicated that cell concentration at the bead periphery could reach maximum values as high as 1 · 1012 cells/ml support (Cachon et al. 1995). Doleyres et al. (2002b) also reported high bifidobacteria viable counts (6.8 · 1010 CFU/g) in gellan gum gel beads during continuous cultures.
The established correlation between DNA content and dry biomass concentration was used to estimate the bacterial dry weights in the carriers. The highest immobilized biomass value of 384.9 ± 1.7 mg/ml was obtained after the fourth batch culture. This data is in agreement with data on immobilized biomass dry weights in porous solid supports reported in the literature. Lb. casei immobilized by adsorption on porous glass/ceramic beads for lactic acid fermentation reached a biomass concentration of 0.105 g/g support during continuous culture in whey permeate (Krischke et al. 1991). For Lb. casei cells immobilized on polyethyleneimine-coated foam glass particles in a recycle batch reactor with a whey-based medium, biomass concentration reached 0.490 g/g support after 12 cycles (Senthuran et al. 1997). In a study with Lb. rhamnosus immobilized on different inert adsorbent supports, mainly made of porous glass, in a continuously recycled packed reactor, a maximum immobilized biomass load of the support of 34.4 g dry weight/l was tested (Gonçalves et al. 1992).
The calculated total population in the ICT system after the first batch fermentation (1.09 · 1014 CFU) was approximately eight fold higher than that for the free-cell system (1.57 · 1013 CFU). This is unusual, since total populations in ICT systems are generally lower than those in free-cell systems after a single batch fermentation (Champagne et al. 1993). This suggests that DNA data used to estimate the CFU population slightly overestimates Lb. rhamnosus RW-9595M viable cell counts. It must be remembered that the DNA-CFU relationship was established with the parent ATCC 9595 strain. After the first batch, immobilized cells represented approximately 60% of the total population in the ICT system. This is typical of data obtained in ICT systems following a first batch fermentation (Champagne et al. 1993).
At the end of the fourth consecutive culture, which lasted only 7 h, a high total polysaccharide content in the broth of 2176 ± 95 mg/l was obtained (Fig. 3.2), which corresponded to an EPS production of 1808 ± 107 mg/l. At this time, the carbon source in the medium, lactose, was completely consumed and the culture entered stationary growth phase. It can be presumed that cell growth and/or EPS production could be further enhanced by using a higher carbon concentration in SWP medium, such as 8% instead of 5% whey permeate in SWP. Even though batch cultures with immobilized cells were limited by the carbon source, the high biomass maintained in the system more than doubled the maximum EPS volumetric productivity (258 mg/l·h after 7 h during the fourth batch culture) compared with free-cell batch cultures in 8% SWP (110 mg/l·h, calculated for an EPS production of 1985.2 mg/l after 18 h culture). Considering interruption periods for equipment cleaning and inoculum preparation, the EPS volumetric productivity of repeated immobilized cell batch cultures is approximately 4–5 fold higher than for free-cell batch cultures. Taking into account free and immobilized biomass, a specific productivity of 6.24·10-12 mg/CFU h was calculated for the fourth immobilized cell fermentation, which is approximately 20% lower than that for free-cell batch fermentations (8.45·10-12 mg/CFU h). This data can be explained by mass transfer limitations due to immobilization which create a different microenvironment for immobilized cells compared with that for planktonic cells, affecting not only nutrient uptake but also removal of products like lactic acid (Champagne et al. 1994; Lamboley et al. 1997). Product accumulation may lower growth rate and change the physiology of immobilized LAB cells (Cachon et al. 1998; Champagne et al. 1994; Krishnan et al. 2001).
In this study, a whey permeate-based medium was fermented by Lb. rhamnosus RW-9595M for high EPS production. This is a promising approach for using whey permeate, a by-product of the dairy industry available in large amounts and with limited value. A method for biomass determination, based on extraction and quantification of microbial DNA, was developed. This method was useful for the estimation of biomass in solid supports, which cannot be quantified by traditional methods, such as biomass dry weight or cell enumeration. Repeated batch cultures clearly showed the high potential of immobilized cells on solid supports for EPS production, increasing the maximum volumetric productivity compared with free-cell cultures. In addition, this technology results in shorter shut-down periods of fermentation systems, with more than three production batches per day, and reuse of microbial cells, which further enhances industrial productivities. Our data shows the great potential of an immobilized cell system with this strain for the production of EPS-type functional and eventually nutraceutical food ingredients in a whey permeate-based medium. This system is currently being studied for continuous EPS production in supplemented whey permeate medium.
Copyright Dirk Bergmaier
| | | |
| Chapitre 2 New method for exopolysaccharide determination in culture broth using stirred ultrafiltration cells |  | Chapitre 4 Exopolysaccharide production during chemostat cultures with free and immobilized Lactobacillus rhamnosus RW-9595M |
