CHAPITRE 2 : Bifidobacterium longum ATCC 15707 cell production during free- and immobilized-cell cultures in MRS-whey permeate medium

Table des matières

Y. Doleyres, C. Paquin, M. Leroy et C. Lacroix

Centre de recherche en sciences et technologie du lait STELA

Université Laval

Québec, Qc, G1K 7P4, Canada

Publié dans "Applied Microbiology and Biotechnology", 60  : 168-173, 2002.

La production de Bifidobacterium longum ATCC 15707 dans le milieu de culture MRS supplémenté en perméat de lactosérum (MRS-WP) a été étudiée pendant des fermentations batch avec cellules libres et en continu avec cellules immobilisées. Une concentration cellulaire très élevée fut dénombrée après 12 h de fermentation en batch à pH= 5.5 dans le milieu MRS-WP (1.7±0.5x1010 CFU/ml), ce qui est environ deux fois supérieur aux comptes cellulaires dans le MRS non supplémenté. Notre étude a montré que le perméat de lactosérum est une source bon marché de lactose et d’autres composés qui pourrait être utilisé pour augmenter la production de bifidobactéries dans le milieu MRS. La fermentation continue dans le MRS-WP avec des cellules de B. longum immobilisées dans des billes de gel a produit les plus fortes concentrations cellulaires (4.9±0.9x109 CFU/ml) dans l’effluent pour un taux de dilution de 0.5 h-1. Cependant la productivité volumétrique maximale pendant la fermentation (6.9±0.4x109 CFU•ml-1•h-1) a été obtenue pour un taux de dilution de D=2.0 h-1, et était approximativement 9.5 fois plus élevée que celle obtenue lors des fermentations batch avec cellules libres (7.2x108 CFU•ml-1•h-1) au pH optimal de croissance des bifidobactéries de 5.5.

Bifidobacterium longum ATCC 15707 cell production was studied in MRS medium supplemented with whey permeate (MRS-WP) during free-cell batch fermentations and continuous immobilized-cell cultures. Very high populations were measured after 12 h batch cultures in MRS-WP medium controlled at pH=5.5 (1.7±0.5x1010 CFU/ml), which was approximately 2-fold higher than in non-supplemented MRS. Our study showed that whey permeate is a low-cost source of lactose and other components that can be used to increase bifidobacteria cell production in MRS medium. Continuous fermentation of MRS-WP with B. longum immobilized in gellan gum gel beads produced the highest cell concentrations (4.9±0.9x109 CFU/ml) in the effluent at a dilution rate of 0.5 h-1. However maximal volumetric productivity (6.9±0.4x109 CFU•ml-1•h-1) during continuous cultures was obtained for a dilution rate of D=2.0 h-1, and was approximately 9.5-fold higher than during free-cell batch cultures at optimal pH of 5.5 (7.2x108 CFU•ml-1•h-1).

Bifidobacteria are natural hosts of the human and animal large intestines which have beneficial effects on the hosts (Chandan, 1999). They are now used in the preparation of a large variety of pharmaceuticals and fermented milk foods, such as fresh cheeses, fermented milks and health supplements (Tamime et al., 1995). However, bifidobacteria are fastidious bacteria that require complex and expensive media for propagation, such as the MRS medium (de Man et al., 1960) with the addition of growth-promoting factors, due to their stringent growth requirements (Ibrahim and Bezkorovainy, 1994). Low-cost whey-based media were developed to continuously produce cultures of bifidobacteria (Corre et al., 1992). Concentrated suspensions of Bifidobacterium bifidum were obtained during continuous cultures with these media when an ultrafiltration unit was coupled to the bioreactor. However long-term stability of cultures with membrane reactors is limited by membrane fouling, particularly with whey-based media (Musale and Kulkarni, 1998). Whey permeate is the by-product of whey ultrafiltration used to concentrate whey proteins. Whey permeate, which is produced in large amounts and contains high concentrations of lactose (4.8%) and minerals (0.5%), is a good base medium for lactic acid bacteria culture (Lamboley et al., 1997).

Cell immobilization in natural polymers was previously developed for the production of biomass during continuous fermentations with major advantages such as high volumetric productivity, stable strain ratios in mixed immobilized cultures, prevention from washing-out during continuous cultures, reduction of susceptibility to contamination and bacteriophage attack, enhancement of plasmid stability, and protection of the cells from shear forces in the stirred reactor (Champagne et al., 1994; Lamboley et al., 1997; 1999; 2001 ; Macedo et al., 1999). However, very little data have been reported on bifidobacteria immobilization and utilization of immobilized cell technology for bifidobacteria biomass production. Bifidobacteria immobilized in κ-carrageenan/locust bean gum mixed gel beads have been tested for continuous fermentation of a supplemented milk (Ouellette et al., 1994). Immobilized cell growth and cell release from the gel beads into circulating milk allowed for a steady inoculation of the feed. However, B. longum is highly sensitive to K+ ions, which are required for the maintenance of κ-carrageenan/locust bean gum bead integrity during continuous fermentations (Paquin et al., 1990). Gellan gum gel was studied and used as an entrapment matrix for B. longum , for high mechanical bead stability, as well as high entrapped cell activity during continuous fermentation (Camelin et al., 1993).

In the present study, a continuous fermentation of MRS medium supplemented with whey permeate with cells of Bifidobacterium longum ATCC 15707 immobilized in gellan gum gel beads was studied for biomass production at different dilution rates. Preliminary experiments were conducted in free-cell batch fermentations to determine the effect of pH and whey permeate supplementation of the MRS medium on cell growth and maximal cell production.

Figure 2.1 shows the effect of pH on B. longum growth during free-cell batch fermentations at 37ºC, with or without whey permeate supplementation of MRS medium. All fermentations with or without pH control between 5.0 and 6.5 exhibited the same growth during the first 6 h of culture, which corresponded to non-limiting growth conditions. After this period, growth rates and biomass concentrations varied greatly depending on pH set points. Very high maximum populations were measured after 12 h during cultures in the MRS-whey permeate medium at pH=5.5 (1.7±0.5x1010 CFU/ml) and pH=6.0 (1.5±0.3x1010 CFU/ml), which were not significantly different (p>0.05). A significantly lower B. longum concentration (9.0±2.3x109 CFU/ml) was obtained during fermentation at pH=5.0 after 12 h culture. Biomass growth during 12 h cultures at pH=6.5 and without pH control was very similar, with maximum biomass concentrations of 3.5±0.5x109 and 3.3±0.3x109 CFU/ml, respectively.

Whey permeate supplementation of MRS medium did not change cell growth during the first 8 h of fermentations at pH=5.5, but a lower final biomass was reached with MRS (8.7±2.6x109 CFU/ml) compared with MRS-WP (1.7±0.5x1010 CFU/ml), which contained a higher total sugar concentration.

The highest acid concentrations in MRS-WP were obtained at pH=6.0 when sugars were completely metabolized, with lactic and acetic acid productions of 14.0±0.9 and 14.6±1.5 g/l after 12 h culture, respectively (Table 2.1). Significantly lower lactic and acetic acid productions (11.8±0.5 and 11.3±0.6 g/l, respectively) were found after 12 h culture at pH=5.5. Despite the complete hydrolysis of lactose, only 11.6±1.2 g/l of initial glucose was metabolized in this condition and residual glucose concentration was high (7.5±1.2 g/l). The ratios (w/w) of acetic to lactic acid at the end of cultures in MRS-WP for different pH were not significantly different and equal to 0.93±0.08 (Table 2.1). Galactose concentrations at the end of cultures for different pH (except for pH=6.5) were low (<2 g/l) and not significantly different from unfermented media (Table 2.1). However, residual galactose concentration at pH=6.5 was very high (8.2 g/l).

Glucose was completely consumed after 10 h culture carried out at pH=5.5 in MRS medium without whey permeate supplementation (Figure 2.2). The lactic and acetic acid productions were significantly lower than in MRS-WP at the same pH (7.2±0.8 and 9.6±0.9 g/l compared with 11.8±0.5 and 11.3±0.6 g/l for lactic and acetic acids, respectively). However, the acetic / lactic acid ratio (w/w) produced after 12 h culture in non-supplemented MRS (1.35±0.17) was significantly higher than for cultures in MRS-WP (Table 2.1).

The biomass yield after 12 h free-cell batch cultures at different pH varied between 2.3±0.3x108 and 5.7±1.7x108 CFU/g sugar consumed. The organic acid (sum of lactic and acetic acids) yields from sugar for different pH and media were not significantly different and averaged 0.76±0.08 g/g sugar consumed.

Table 2.2 shows residual sugars and biomass and organic acid productions determined in the outflow of continuous fermentations with cells immobilized in gellan gum gel beads, as a function of dilution rate. Cell count was highest (4.9±0.9x109 CFU/ml) at D=0.5 h-1 and it did not change significantly with dilution rate in the range from 1.0 to 2.0 h-1 (3.7±0.2x109 CFU/ml). However, the highest biomass volumetric productivity increased from 2.5±0.4x109 CFU•ml-1•h-1 to 6.9±0.4x109 CFU•ml-1•h-1 with dilution rate increasing in the range from 0.5 to 2.0 h-1. B. longum counts in beads determined at the end of experiment were high, equal to 6.8±1.9x1010 CFU/g.

Glucose and lactose concentrations (approximately 20 g/l of each sugar) in the MRS-WP were not limiting factors for the continuous culture. Sugar utilization decreased from 4.4±1.1g/l and 8.5±0.5 g/l to 2.0±0.5 g/l and 1.9±0.6 g/l, for glucose and lactose, respectively, for D increasing from 0.5 to 2.0 h-1 (Table 2.2). However, the lactose intake rate was not significantly different for all dilution rates tested, averaging 4.1±0.3 g•l-1•h-1. The glucose intake rate was significantly higher at high D of 2.0 h-1 (4.0±0.5 g•l-1•h-1) than for low D of 0.5 and 1.0 h-1 (Table 2.2). Galactose concentrations remained low between 0.6±0.2 and 1.7±0.4 g/l, for D decreasing from 2.0 to 0.5 h-1. Lactic and acetic acid productions were highest at low dilution rate (0.5 h-1), equal to 9.3±0.7 and 8.0±1.5 g/l, respectively. However the organic acid productivities increased with the dilution rate to maximum values of 11.9±1.0 and 10.5±1.5 g•l-1•h-1 at D=1.5 h-1 for lactic and acetic acids, respectively, and remained constant for a higher D of 2.0 h-1. The ratio (w/w) of acetic and lactic did not show a defined trend with D, varying between 0.85±0.10 and 0.99±0.03 in the tested range (Table 2.2).

This study is the first to report supplementation of MRS medium with whey permeate for production of bifidobacteria. The MRS medium is a complex broth medium (de Man et al., 1960) that is commonly used to grow fastidious strains, such as lactobacilli and bifidobacteria. However its relatively low sugar content (2%, w/w) combined with its high cost may limit industrial applications. Whey-based media have already been used to grow bifidobacteria (Corre et al., 1992). In this study whey supplemented with casamino acids or yeast extract and a reducing agent, such as ascorbic acid or cysteine, appeared to be a suitable medium for B. bifidum growth during batch cultures without pH regulation. However, the maximum cell production (9x108 CFU/ml) was lower than in MRS medium (1.4x109 CFU/ml). Supplementation of MRS medium with whey permeate was investigated in our study to increase MRS characteristics for bifidobacteria growth. Indeed, whey permeate represents a low-cost source of lactose and other components such as minerals and nitrogen compounds that can be limiting factors for growth of bifidobacteria.

Free-cell batch fermentations were carried out in MRS-WP with pH control in the range from 5.0 to 6.5 to determine B. longum optimal growth conditions. Very high biomass concentrations of 1.7±0.5x1010 and 1.5±0.3x1010 CFU/ml were obtained at pH=5.5 and 6.0, respectively. This is in agreement with the fact that optimal growth of bifidobacteria occurs within pH interval of 5.5-6.0 (Scardovi, 1986). Organic acid productions were higher in cultures at pH=6.0 than at pH=5.5. Available sugars were completely consumed at pH=6.0, contrary to fermentations at pH=5.5 where approximately 8.5 g/l glucose remained in the medium after the 12 h culture. This data can be explained by the higher concentration of un-dissociated lactic and acetic acids at low pH, which are the most inhibitory forms (Gätje and Gottschalk, 1991). HPLC analyses clearly demonstrated that B. longum preferentially metabolized lactose to glucose during free-cell batch fermentations of a medium containing the two sugars (Figure 2.2). Galactose produced from lactose hydrolysis was also readily consumed, as shown by the very low galactose concentrations measured after 12 h cultures (< 2 g/l), except for pH=6.5 (8.2 g/l) (Table 2.1).

B. longum production was significantly lower during free-cell batch fermentations carried out in non-supplemented MRS medium with optimal pH conditions (pH=5.5) (8.7±2.6x109 CFU/ml) compared with MRS-WP at the same pH (1.7±0.5x1010 CFU/ml). The same trend was observed for organic acid productions. The complete glucose intake after 10 h culture clearly shows that sugar was the limiting factor in MRS fermentation (Figure 2.2). However other components than lactose in whey permeate, such as minerals and some nitrogen compounds, could also promote the growth of B. longum in MRS-WP.

Continuous fermentation with B. longum immobilized in gellan gum gel beads was carried out to study the effect of dilution rate, in the range from 0.5 to 2.0 h-1, on biomass production. Very high cell concentrations were continuously produced in the effluent, with maximum at D=0.5 h-1 (4.9±0.9x109 CFU/ml). The fermentation medium composition was not apparently a limiting factor, as high sugar concentrations remained in the effluent of the continuous culture. In the only other work on continuous production of bifidobacteria with immobilized cells, a pure culture of B. infantis immobilized in κ-carrageenan/locust bean gum gel beads was used to continuously ferment skim milk (10% solids) supplemented with 1% yeast extract (Ouellette et al., 1994). Cell counts in the cultured milk increased from 1.0 to 2.2x109 CFU/ml for dilution rates decreasing in the range from 1.0 to 0.5 h-1. Compared with our study, the lower cell production could be due to different growth rates for the two strains and to the fact that the MRS-WP might be better suited for growth of bifidobacteria than milk-based media (Desjardins et al., 1991). B. bifidum maximal productivity in a whey-based medium using a continuous stirred tank reactor attached to an ultrafiltration device was approximately 35-fold lower (2x108 CFU•ml-1•h-1) than in our study (6.9±0.4x109 CFU•ml-1•h-1 at D=2.0 h-1), for similar cell concentrations in the ultrafiltration retentate (5x109 CFU/ml) and in the effluent of the immobilized cell reactor (3.5±0.2x109 CFU/ml) (Corre et al., 1992). Maximal volumetric productivity at pH=5.5 during 12-h free-cell batch cultures, considering one 12 h experiment per working day, reached 7.2x108 CFU•ml-1•h-1, which is approximately 9.5-fold lower than for continuous fermentations.

Lactic and acetic acid concentrations in the effluent of the continuous fermentations were much lower for all dilution rates tested than after 12 h of free-cell batch fermentations carried out with the same pH and medium. However organic acid productivities were approximately 4.5- (D=0.5 h-1) to 11.5-fold (D=2.0 h-1) higher during continuous fermentation with immobilized cells compared with free-cell batch fermentations (12 h culture).

Very high B. longum counts were determined in beads at the end of the experiment (6.8±1.9x1010 CFU/g). Similar bead colonizations were reported for the same culture immobilized in κ-carrageenan/locust bean gum gel beads during fermentations in MRS (Doleyres et al., 2002) or in MRS-WP (Camelin et al., 1993), with 9.6x1010 and 7.0x1010 CFU/g, respectively. However, cell concentration in beads was much lower (2*1010 CFU/g) during continuous fermentation of skim milk supplemented with yeast extract with immobilized B. infantis (Ouellette et al., 1994).

Our study showed that MRS medium supplementation with whey permeate is a very suitable fermentation medium for B. longum production and increased cell production by 2-fold compared with MRS. In addition, the immobilized cell technology can be used to continuously produce fastidious micro-organisms, such as bifidobacteria, with a high volumetric productivity and high biomass concentrations in the outflow of the continuous fermentation, even at high dilution rates exceeding the specific growth rate.

Table 2. 1 : Residual sugar concentrations and biomass and acid productions after 12 h culture during B. longum ATCC 15707 pH-controlled batch fermentations in MRS medium with or without whey permeate supplementation.

Means and standard deviations are calculated from triplicate experiments. Means with the same letter (by columns) are not significantly different (P >0.05).

1 Initial glucose, galactose and lactose concentrations were 19.5±1.1, 0.8±0.5, 0 g/l in MRS and 19.5±0.9, 1.4±0.4, 20.9±0.8 g/l in MRS-WP.

2 Final pH was 4.5

n.a. = Only one repetition was available for these data.

Table 2. 2 : Residual sugar concentrations and biomass and acid productions during continuous fermentation of MRS-WP at pH=5.5 with immobilized cells of B. longum ATCC 15707 in gellan gum gel beads as a function of dilution rate.

Means and standard deviations are calculated from duplicate experiments. Means with the same letter (by columns) are not significantly different (P >0.05). 1 Initial glucose, galactose and lactose concentrations were 18.9±0.2, 0.5±0.3, 20.3±0.6 g/l.

Figure 2. 1 : Viable cell counts during free-cell batch fermentations of B. longum ATCC 15707 in MRS-WP at pH=5.0 (▲), 5.5 ( × ), 6.0 (♦), 6.5 (∆) and without pH control (◊), or in unsupplemented MRS medium at pH=5.5 (□).

Figure 2. 2 : Lactose (▲), glucose (■), galactose (♦), lactic (□) and acetic (◊) acid concentrations determined by HPLC analysis during free-cell pH-controlled (pH=5.5) batch fermentations of B. longum ATCC 15707 in MRS medium (a) or in MRS-WP (b).