Fermented Milk Starch and Milk Inulin Products as Vehicles for Lactic Acid Bacteria

Transcription

1 Plant Foods for Human Nutrition 59: , C 2004 Springer Science+Business Media, Inc. 155 Fermented Milk Starch and Milk Inulin Products as Vehicles for Lactic Acid Bacteria ANGELA ZULETA, 1 MARÍA I. SARCHI, 1 MARÍA E. RIO, 1, MARÍA E. SAMBUCETTI 1 MARCELO MORA, 2 SUSANA V. DE FABRIZIO 2 & JOSÉ L. PARADA 2 1 Facultad de Farmacia y Bioquímica, Universidad de Buenos Aires, Junín 956, (1113) Buenos Aires, Argentina; 2 Facultad de Ciencias Exactas y Naturales Universidad de Buenos Aires, Ciudad Universitaria, Pab. II. 1428, Buenos Aires, Argentina ( author for correspondence; merio@ffyb.uba.ar) Abstract. Formulations using cassava starch or inulin plus milk were fermented with three different lactic acid bacteria (LAB) strains: Lactobacillus plantarum D34, Lactobacillus sp. SLH6, and Streptococcus thermophilus ST4. Growth and acidification were followed in 3% powdered milk (M3), 3% milk 6% starch (M3-S6), and 3% milk 6% inulin (M3-In6). D34 and SLH6 growth was enhanced by starch in M3-S6, when compared to the count (CFU/ml) obtained in M3. Growth of all strains was promoted by inulin. All fermented products showed LAB counts of 8.0 log or higher. Carbohydrate utilization was in agreement with growth and acidification results. The highest increase in CFU in rat feces was observed in M3-S6 fermented with ST4; the D34 fermented product also increased CFU but SLH6 did not, either with starch or inulin. This suggests that ST4 and D34 strains provide a good choice to ferment the proposed formulations in order to obtain a marked improvement of natural intestinal flora. Key words: Fermented foods, Lactic acid bacteria, Milk, Prebiotics, Probiotics, Starch Introduction Lactic acid fermentation is a safe and simple way to preserve foods and has been used even before the Christian era. Moreover, the role played by lactic acid bacteria (LAB) in the prevention of diseases was originally suggested at the beginning of the 20th century when Metchnikoff postulated that LAB had beneficial effects on health and promoted longevity. Several traditional soured milks such as yoghurt, kefir, koumis, and leben, among others, have been used as food and also as food therapy, long before modern knowledge about the existence of probiotics [1]. At present, there is general agreement concerning the major role of the gastrointestinal microflora in health preservation and promotion [2]. Intestinal microflora represent a complex ecosystem, susceptible to be altered by several factors including food composition. During the last three decades several attempts have been made to modulate the indigenous intestinal flora by the intake of exogenous LAB [3, 4]. To increase the number of health-promoting bacteria in the intestinal ecosystem, two approaches have been proposed: oral administration of mono or mixed cultures of live microorganisms as fermented foods, dietary supplements or capsules and oral intake of bacterial stimulant substances, known as prebiotics, like inulin and its oligoderivatives [5, 6]. Both approaches have been useful to benefit the host by improving colon microflora composition. Some data suggest that better effects may be attained when LAB are administered in foods; however, it is not clear whether all kinds of food perform similarly [7]. In developed countries, fermented milks have usually been considered the traditional source of LAB. However, many lactic acid fermented foods, in addition to milk-based ones, have been available for centuries as local resources around the world [8]. Examples of fermented starchy products not containing milk include the Nigerian ogi, the Ethiopian kocho, and the Tanzanian togwa, all fermented with Lactobacillus plantarum, the starter most frequently used in lactic acid fermentation of plant materials [9]. Fermented cassava starch is favored in several Latin American countries [10]. It has both good taste and functional characteristics. Since some amylolytic LAB are able to accomplish its biotransformation in gelified starch suspensions obtained after thermal treatment, appropriate technologies could be set up to combine such raw material with reconstituted powdered milk in order to prepare lactic acid fermented products enriched with LAB and their metabolites. On the other hand, inulin, which has proven to stimulate the beneficial gut flora in detriment of harmful microorganisms [11], could be successfully used to enhance the potential of amylolytic LAB strains. In this paper, basic studies addressing the development of yoghurt-like healthy foods using cassava starch or inulin plus milk fermented with some selected strains of Lactobacillus and Streptococcus thermophilus are presented. Studies of LAB growth kinetics, fermentation properties, and utilization of starch and inulin were carried out on formulated products. Moreover, an experimental model, based on the principles of the method suitable for humans, was applied on young adult rats, to select those strains that were able to survive in the gastrointestinal tract and increase total LAB count in feces. Materials and Methods Materials and Strains. Powdered milk (fat content 1.5%, Nestlé, Argentina) reconstituted in distilled water, cassava

2 156 Table 1. Bacterial strains used as single starters to obtain fermented products Bacterial strains Culture characteristics Description Source Lactobacillus plantarum D34 Lac+, St+ M/T Isolated from traditional cassava starch CIRAD Culture Collection, France fermentation in Colombia Lactobacillus sp. SLH6 Lac+, St+ M/T Isolated from commercial lactic beverage Food Microbiology Culture Collection Lab. FCEN-UBA, Argentina Streptococcus thermophilus ST4 Lac+, St T/M Isolated from commercial starter Food Microbiology Culture Collection Lab. FCEN-UBA, Argentina Note. Lac+, ability to ferment lactose in agar BCP; St+, amylolytic strain developed in starch agar; St, nonamylolytic strain; M/T, good growth at mesophilic or thermophilic temperatures; T/M good growth at mesophilic but better at thermophilic temperatures, in MRS agar. starch (Tres Armas, Argentina) and inulin (Raftiline, ORAFTI, Belgium) were used. LAB strains used in this work were L. plantarum D34, Lactobacillus sp. SLH6, and S. thermophilus ST4, whose characteristics are shown in Table 1. During this study, lactic bacteria were subcultured in MRS (De Man, Rogosa, Sharpe) broth or agar, Merk KGaA, Darmstadt, Germany [12]. MRS-starch was prepared by substituting the glucose in the original composition for 2% starch as only a source of sugar. Strain amylolytic activity was developed by exposing agar plates to iodine vapours [10, 13] and lactosefermenting ability was determined in BCP agar (Lactose Broth with Bromcresol Purpre) Merk KGaA, Darmstadt, Germany [14]. Experimental LAB-Fermented Products. After preliminary experiences with milk and starch mixtures (15), a basic formulation containing 3% powdered milk and 6% cassava starch (M3-S6) was chosen because it provided good bacterial growth and lactic acid production for LAB. An alternative product was prepared using 6% inulin instead of starch (M3-In6). Reconstituted milk was prepared from 3 and 9% powdered milk (M3 and M9) and used to obtain referential data under the same conditions. To reduce natural contaminants and to accomplish starch gelatinization, M3-S6 was subjected to thermal treatment at quite drastic conditions, that is, 20 min of shaking (120 strokes/min) at 90 C. Additionally, all preparations were autoclaved at 120 C for 15 min to completely remove thermally resistant spores, mostly coming from starch. Lactic fermentation was carried out with the corresponding single starter strain at 37 C. Mixtures were inoculated with an overnight pure culture in milk in order to achieve roughly 10 3 CFU/ml and the evolution of growth was followed up by ph and microbial enumeration. Changes in ph were recorded every 3 hours for 10 hours and then at 24 hours. Nonfermented controls were prepared and evaluated in the same manner. LAB count was performed in triplicate using MRS agar incubated anaerobically at 37 C for hours. Enumeration values are reported in log CFU/ml. Utilization of Carbohydrates. Lactose, starch, and inulin were analyzed in the final products by high performance liquid chromatography (HPLC) [16, 17], using Waters Associates, Inc., Milford, USA, equipment with an R-40 refractive index detector and integrator. Chromatographic test were performed using an Aminex HPX-87C (Bio-Rad Hercules, U.S.A.) anion exchange column, and deionized water at 85 C as the mobile phase used at a rate of 0.6 ml/min. Reference sugars including glucose, fructose, galactose, lactose, and inulin were from Sigma Chemical Co. (St. Louis). Samples (10 g) of each product were treated with ca. 100 ml boiling water, ph 7, and kept at 85 C with continuous magnetic stirring on a hot plate for 15 min and then centrifuged. The supernatant was filtered through a 0.20-µm membrane before injection. Inulin and free sugars, except lactose, were determined directly. To assess starch and oligosaccharide derivatives, samples were treated with amylase and amyloglucosidase at 60 C for 30 min, at ph 6 and 4.5 respectively. The glucose formed was calculated as starch by using a factor of 0.9. As the above Aminex column fails to resolve completely maltose from lactose, enzyme treatment breaks down maltose into glucose, disclosing the lactose peak. Utilized carbohydrates were calculated by subtracting values found in fermented products from those in control products. Analysis was performed in triplicate. Results were compared by one-way analysis of variance (ANOVA). A Tukey Kramer test for multiple comparison was applied a posteriori (18). Viability of LAB in Rat Feces. Groups of three rats of the Wistar strain, of either sex, mean weight of 180 ± 20 g, were used. Rats were lodged individually in stainless steel cages with wire mesh bottoms. As a first step, acceptability of fermented products was evaluated under conditions of ad libitum product intake in order to establish the test dose to be used along the experimental feeding period; the observed mean intake was 50 ± 3 g/rat/day, so that 50 g were considered adequate. Experimental design: Rats were offered daily one test dose of each fermented product containing viable CFU/ml or an equivalent amount

3 157 of nonfermented controls during the four initial days of the experiment according to the following schedule: pots containing 50 g of each fermented or control experimental product were offered at 10 PM and left overnight. At 8 AM pots were removed and product intake controlled. Rats were then allowed to feed on stock diet ad libitum up to 4 PM. At this time, any residual feed was removed in order to allow for feces sample collection. At 10 PM experimental products were reintroduced to start a new daily cycle. After the fourth cycle, animals were allowed to feed only on stock diet ad libitum for an additional 14-day period. A group receiving stock diet along the entire 18 days of the experiment was taken as control. Samples of fresh, recently defecated units of feces were collected between 4 and 10 PM at 0 time and after 1, 2, 9, 11, and 18 days in order to evaluate early (day 1 to 3), intermediate (day 9 to 11) and late (day 18) effects of experimental products on bacterial counts. Feces were collected in 1 ml saline and unit weight was recorded by difference. Viability of LAB was evaluated after separately homogenizing feces samples in 1-ml sterile saline. Convenient dilutions were plated in triplicate on MRS plates incubated anaerobically as described above and CFUs were expressed per gram of wet feces. Colony-forming units data in rat feces were analysed with SP.SS statistical software, SSPP Inc., Chicago [19]. Statistical significance for combined effects of diet and day was assessed by two-factor analysis of variance with repeated measures of one factor and homogeneous variances according to Winer et al. [16]. Significant inter se differences in the effects of bacterial counts in die on feces at each day of sampling were determined by Newman Keuls procedures of a posteriori comparisons. Results and Discussion Carbohydrate Utilization, lactic Acid Bacteria Growth, and Acid Production. Interrelationships among these parameters of Lactobacillus D34 and SLH6, as well as by Streptococcus ST4 in fermented products, M3, M9, M3-S6, and M3-In6, were as follows: Carbohydrate utilization in the fermented products correlated with the LAB features shown in Table 1, since strains utilized lactose or starch as expected (Figure 1). Three and 9% reconstituted milk preparations promoted good bacterial growth, with mean values between 7.6 and 8 log CFU/ml, respectively. When the 3% solid content of the formulations was increased by addition of 6% starch or inulin, the growth of D34 and SLH6 was improved. Starch was used by D34 and SLH6 in M3S6 in similar amounts (P > 0.05) (Figure 1B) but D34 reached the greatest population, 9 log CFU/ml, in agreement with its marked amylolytic ability. In the case of SLH6, the level of growth reached was 8.7 log CFU/ml. Streptococcus ST4 was unable to utilize starch (Figure 1B); consequently the number Figure 1. Utilization of carbohydrates by L. plantarum D34, Lactobacillus sp. SLH6 and S. thermophilus ST4 in fermented products with M3 (A), M3 S6 (B), and M3-In6 (C). Carbohydrates: lactose; starch; inulin. Bars for each carbohydrate not sharing a common letter are significantly different (P < 0.05). of microorganisms attained with M3S6 was the same as with M3 (Figure 1A). Inulin was metabolized by SLH6 and ST4 in similar quantities (P > 0.05), which proved higher than the corresponding value obtained for strain D34 (P < 0.05) (Figure 1C); however, all three strains reached populations about 9 log CFU/ml. Acid production monitored by means

4 158 Figure 2. Viability of LAB in feces of rats fed 4 days with the following: Control M3 S6; M3 S6 fermented with D34; M3 S6 fermented with SLH6; M3 S6 fermented with ST4; Control M3-In6; M3-In6 fermented with SLH6; Stock diet. Values bearing different letters for a particular day are significantly different (P < 0.01). Values are not significantly different (P > 0.01). of ph change showed that values depended not only upon strain metabolic features but also upon the kind of substrate used. Thus, in M3-S6, L. plantarum D34 acidified more than SLH6 and ST4 strains; at the end of fermentation (24 hours) the final phs were 3.40, 4.15, 4.35 for D34, SLH6, and ST4, respectively. In the case of the inulin-containing product, M3In6, all three strains had similar acidification values, resulting in 4.25, 4.00, and 4.00 ph values for D34, SLH6, and ST4, respectively. For all these products initial ph values were around 6.6. Galactose was not found in any of the fermented products except in the fermented milk M9 formulation, in amounts that represented 29 30% of metabolized lactose. In regard to lactose utilization by LAB, some authors (19 20) have demonstrated that several strains, such as L. bulgaricus, L. lactis, L. acidophilus, and S. thermophilus, use only the glucose moiety and release galactose totally or partially into the growth medium. The reason why galactose was not found in products fermented with 3% milk is that, under limited lactose concentrations, some strains of LAB seem to be able to consume all the galactose [21]. Substrates were not exhausted at any time, even when highly utilized, remaining from 30 to 89% in M3 fermented with D34 and M3 S6 with ST4, respectively. The essential condition for beneficial strains to be delivered by foods is that they are alive in substantial amounts in the supplied products. This condition is fulfilled by all the strains used in the milk starch or milk inulin fermented products tested in this study with populations around the value of viable LAB required by the National Yoghurt Association (USA) to be recognized as a live active culture [22]. Viability of LAB. In addition to the essential condition of being alive in substantial amounts in the supplied products, the beneficial effects of LAB strongly depend upon the ability to survive in the intestinal tract. The first condition was accomplished by all strains used in assayed fermented products. To test the second condition, evolution of total LAB in gut was followed by colony counts in fresh feces from rats fed fermented products and compared to the respective nonfermented ones as well as to stock diet (Figure 2). As expected, CFU values in feces from rats fed stock diet remained fairly constant throughout the experimental period, between 7 and 9.5 log CFU/g. A similar behaviour was observed when rats received milk starch or milk inulin nonfermented products. On the other hand, CFU values showed a response depending on the LAB used in fermentation and on sampling day; no significant increase in fecal count was observed during the first 2 days after feeding 50 g of each experimental product. However, differences became significant at day 9, when changes depending on the strain utilized in fermentation were observed. Thus, in the group receiving M3 S6 fermented with S. thermophilus ST4, fecal bacterial counts increased up to a level of 10 log CFU/g, significantly different from all other groups (P < 0.01). On day 18, bacterial

5 159 count attained with D34 was also significantly higher, 10 log CFU/g, than for all other groups (P < 0.01), except for ST4, 11 log CFU/g, which remained significantly higher than all the rest. Therefore, early effects of LAB supplement were not apparent for any strain, but both ST4 and D34 effects were strongly dependent on time, with ST4 increasing CFU faster than D34. Moreover, even when both strains attained bacterial counts significantly higher than all other groups, the effect of ST4 still remained significantly greater than D34 (P < 0.01) when late effects were investigated on day 18. In contrast, changes in SLH6 counts were negligible all along the experiment, even at the end. These results show that some of the used lactic acid bacteria ST4 and D34 allow for survival and population increase of total lactic bacteria in fecal samples all along the 18-day experimental period, although rats were only fed for 4 days with these products. Conversely, products fermented with SLH6 failed to show any beneficial effect on the CFU count. The prebiotic effect of inulin has been intensively studied in experimental models and in humans. In those studies, inulin proved to be very effective in increasing bifidobacteria levels in feces and also, to some extent, in lactic acid bacteria [11]. On this basis, we speculated that SLH6 could be also stimulated in intestine as it was in the fermented product. However, the experimental trial did not confirm our hypothesis, since substitution of inulin for starch (M3- In6) and fermentation with SLH6 did not produce the expected increase in feces CFU levels during the 18-day experimental period, proving to be statistically similar to control or stock group values (P > 0.01). As the only strain used with inulin as substrate was SLH6, which consumed about 70% of the original inulin during the fermentation process (Figure 1), the lack of response to the prebiotic might be attributed to the scarce amount remaining in the product. However, as SLH6 showed poor survival in feces with either substrate, the possibility of increasing fecal counts with inulin when associated to other LAB strains cannot be ruled out. Although the survival of lactobacilli in the gut depends mainly upon bacterial features, it may also be affected by the physicochemical characteristics of the food, perhaps protecting the microorganism in some particular way related to the starch in the food used for administration [23]. In this study, the ST4 strain survives and increases total LAB count in feces from rats fed with the starchy-milk products. The S. thermophilus, a traditional yogurt culture starter, has usually been considered as one that survives poorly in the human gastrointestinal tract [9]; however, our results concerning S. thermophilus are in agreement with those obtained recently in humans receiving a pharmaceutical product [24]. The conclusions we could draw from the study, on the basis of the obtained results, confirms that the fermentable substrate utilization pattern of LAB is not useful as a measure of survival in the gastrointestinal tract. Furthermore, S. thermophilus ST4 and L. plantarum D34 strains offer a good choice for the fermentation of the proposed formulations in order to make an improvement in natural intestinal flora. Therefore, it would be worthwhile to test in humans the viability of those lactic acid bacteria delivered in a milk starch fermented food. Acknowledgments This work was carried out with the financial support of the University of Buenos Aires (Integrated Project IX22, UBACYT). The authors thank G. Larumbe for technical assistance. References 1. Fuller R (1989) Probiotics in man and animals. J Appl Bacteriol 66: Shanahan F (2000) Therapeutic manipulation of gut flora. Science 289: Goldin BR (1998) Health benefits of probiotics. Br J Nutr 80 (Suppl. 2): S203 S Guililand SE (1990) Health and nutritional benefits from lactic acid bacteria. FEMS Microbiol Rev 87: Gibson GR, Roberfroid MB (1995) Dietary modulation of the human colonic microbiota: Introduction to the concept of prebiotic. J Nutr 132: Charalampopoulos D, Wang R, Pandiella SS, Webb C (2002) Application of creals and cereal components in functional foods: A review. Int J Food Microbiol 79: Vanderhoof JA (2001) Probiotics: future directions. Am J Clin Nutr 73(Suppl): 1152S 1155S 8. Young J (1998) European market development in prebiotic and probiotic containing foodstuffs. Bri J Nutr 80(Suppl 2): S231 S Molin G (2001) Probiotics in foods not containing milk or milk constituents, with special reference to Lactobacillus plantarum 299v. Am J Clin Nutr 73(Suppl): 380S 385S. 10. Parada JL, Zapata LE, de Fabrizio SV, Martinez A (1996) Microbiological aspects of cassava starch fermentation. World J Microbiol Biotechnol 12: Kolida S, Tuohy K, Gibson GR ( 2002) Prebiotic effects of inulin and oligofructose. Br J Nutr 87(Suppl 2): S193 S De Man JD, Rogosa M, Sharpe ME (1960) A medium for the cultivation of Lactobacilli. J Appl Bacteriol 23: Giraud E, Gosselin L, Marin B, Parada JL, Raimbault M (1993) Purification and characterization of an extracellular amylase activity from Lactobacillus plantarum strain A6. J Appl Bacteriol 75: Huggins AR, Sandine WE (1979) Differentiation of fast and slow milk-coagulating isolates of lactic streptococci. J Dairy Sc 70(Suppl 1): Sambucetti ME, Zuleta A, Kafka A, de Fabrizio SV, Parada JL (1998) Estudio de las características y comportamiento de diferentes almidones en alimentos fermentados. Memorias del Seminario del Proyecto de Investigación Precompetitiva XI.8 (CYTED). Baños, Ecuador, September 2 4, pp Scott F, Hatina G (1988) HPLC Determination of sugars, starch and oligosaccharides in infant formulas using resin-based, fixed-ion column. J Food Sci 53: Zuleta A, Sambucetti ME ( 2001) Inulin determination for food labeling. J Agric Food Chem 49:

6 Winer BJ, Brown DR, Michels KM (1991) Statistical principles in experimental design, 3rd ed. New York: Mc Graw-Hill. 19. SSPP (1999) Advanced Models Chicago: SPSS Inc. 20. Hickey MW, Hillier AJ, Richard Jago G (1986) Transport and metabolism of lactose, glucose and galactose in homofermentative Lactobacilli. Appl Environ Microbiol 51: Benthin S, Nielsen J, Villadsen J (1994) Galactose expulsion during lactose metabolism in Lactococcus lactis subsp. cremoris FD1 due to dephosphorylation of intracellular galactose 6-phosphate. Appl Environ Microbiol 60: Sanders ME (1998) Development of consumer probiotics for the US market. Br J Nutr 80(Suppl 2): S213 S Crittenden R, Laitila A, Forsell P, Matto J, Saartela M, Mattila- Sandholm T, Myllarinen P (2001) Adhesion of bifidobacteria to granular starch and its implications in probiotic technologies. Appl Environ Microbiol 67: Brigidi P, Swennen E, Vitali B, Rossi M, Matteuzzi D (2003) PCR detection of bifidobacterium strains and Streptococcus thermophilus in feces of human subjects after oral bacteriotherapy and yogurt consumption. Int J Food Microbiol 81: