Potential of functional strains, isolated from traditional Maasai milk, as starters for the production of fermented milks

International Journal of Food Microbiology 107 (2006) 1 – 11 www.elsevier.com/locate/ijfoodmicro

Potential of functional strains, isolated from traditional Maasai milk, as starters for the production of fermented milks Francesca Patrignani a,*, Rosalba Lanciotti a, Julius Maina Mathara b,c, Maria Elisabetta Guerzoni a, Wilhelm H. Holzapfel c b c

a University of Bologna, Dipartimento di Protezione e Valorizzazione Agroalimentare, via Fanin, 46, 40127 Bologna, Italy Jomo Kenyatta University of Agriculture and Technology, Department of Food Science and Technology, P. O. Box 62000, Nairobi, Kenya Federal Research Centre for Nutrition and Food, Institute of Hygiene and Toxicology, Haid-und-Neu-Str. 9, D-76131 Karlsruhe, Germany

Received 4 November 2004; received in revised form 13 July 2005; accepted 7 August 2005

Abstract The purpose of this research was the evaluation of technological features and of the ability of functional LAB strains with desirable sensory characteristics, to produce fermented milk. Eight strains of Lactobacillus plantarum, Lactobacillus acidophilus, Lactobacillus paracasei and Lactococcus lactis, isolated from Maasai traditional fermented milk in Kenya and previously tested for their probiotic properties, were selected for this investigation. Technological features such as growth kinetics in fresh heat-treated whole milk medium and survival in the final product during storage at 4 -C, were studied. The strains Lb. acidophilus BFE 6059, Lb. paracasei BFE 5264 and Lc. lactis BFE 6049 showed the best potential and were thus selected for use as starter cultures in further trials with the objective to improve their technological performance and to optimise the sensory features of fermented milk obtained. The effects of fat ( F), non-fat milk solids (S) and fermentation temperature (T), modulated according to a Central Composite Design, on fermentation rates and viability losses during refrigerated storage of the chosen starters, and on product texture parameters, were studied. From the data analysis, it was possible to select optimum conditions for enhancing positive sensory traits of final products and for improving the survival of these potentially probiotic cultures. D 2005 Elsevier B.V. All rights reserved. Keywords: Probiotic strains; Fermented milk; Organoleptic properties; Texture; Technological features

1. Introduction Products containing probiotic strains and claimed to maintain health and well-being are becoming an increasingly important option for consumer choice. This resulted in rapid growth and expansion of the market for such products, in addition to increased commercial interest in exploiting their proposed health attributes (Stanton et al., 1998). Probiotic bacteria and their health effects are the focus of intensive international research. Evidence that documents probiotic, health-promoting effects for a few well-characterized lactic acid bacteria (LAB) strains (Lee and Salminen, 1995; Salminen et al., 1996, 1998; Mattila-Sandholm et al., 1999; Saarela et al., 2000) is accumulating. In order to exert positive

* Corresponding author. Tel.: +39 0547636132; fax: +39 0547382348. E-mail address: [email protected] (F. Patrignani). 0168-1605/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.ijfoodmicro.2005.08.004

health effects, the microorganisms need to be viable, active and sufficiently abundant in concentrations of at least 106 cfu/g in the product throughout the specified shelf life (Samona and Robinson, 1991; Vinderola et al., 2000). Focus has generally been on incorporation of selected strains of Lactobacillus spp. into milk and fermented milk products. For successful promotion of functional probiotic products, the food industry has to satisfy the demands and expectations of the consumer. All probiotic foods should be safe and have good sensory properties (Saarela et al., 2000). In particular, the sensory characteristics of fermented milks play an important role in product acceptance of consumers (Gardini et al., 1999). However, fermented milks obtained from the direct and sole use of probiotic strains are often characterised by the lack of desirable sensory features. In particular, structural defects and absence of aroma were reported for milk fermented by Bifidobacterium spp. and Lactobacillus acidophilus strains (Marshall and Cole, 1983).

4.58 T 0.02 4.60 T 0.01 4.65 T 0.03 5.47 T 0.06 5.52 T 0.04 – –

4.78 T 0.03 4.99 T 0.08 5.44 T 0.08 –

5.73 T 0.04 4.60 T 0.01 5.42 T 0.02 5.59 T 0.08 – 4.78 T 0.05 – –

4.64 T 0.02 4.76 T 0.05 4.78 T 0.03 4.60 T 0.02 5.56 T 0.02 5.36 T 0.09 4.60 T 0.02 – – 4.72 T 0.03 5.67 T 0.04 5.51 T 0.06 4.85 T 0.02 – –

5.93 T 0.02 5.14 T 0.06 5.87 T 0.01 5.91 T 0.02 6.04 T 0.02 5.45 T 0.04 6.00 T 0.06 6.00 T 0.02

5.14 T 0.05 5.77 T 0.07 5.77 T 0.05 5.28 T 0.02 5.93 T 0.10 5.92 T 0.04 5.63 T 0.01 5.96 T 0.02 6.06 T 0.04

29.5 h 25.5 h 12 h 11.5 h 11 h 10.5 h 9.5 h 8.5 h 7.5 h

5.86 T 0.02 5.00 T 0.04 5.56 T 0.04 5.81 T 0.02 Data not determined.

Basal medium was treated whole milk supplemented with tryptone and glucose. a Sterile skim powder was employed as not fat solid. b UHT cream was employed as fat.

a

0.75 0.75 0.75 2.25 2.25 2.25 2.25 1.50 1.50 1.50 1.50 1.50 1.50 1.50 0.00 3.00 1.50

5.71 T 0.02 –a – 4.60 T 0.01 – 5.78 T 0.02 – –

1.5 1.5 4.5 4.5 1.5 1.5 4.5 4.5 3.0 3.0 3.0 3.0 0.0 6.0 3.0 3.0 3.0

5.99 T 0.09 6.18 T 0.04 6.29 T 0.03 4.91 T 0.03 6.18 T 0.02 5.93 T 0.05 6.12 T 0.02 6.12 T 0.04

26.5 33.5 26.5 33.5 26.5 33.5 26.5 33.5 30.0 30.0 23.0 37.0 30.0 30.0 30.0 30.0 30.0

6.38 T 0.04 6.41 T 0.03 6.52 T 0.04 6.24 T 0.02 6.40 T 0.03 6.38 T 0.02 6.41 T 0.01 6.41 T 0.02

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

6.57 T 0.03 6.53 T 0.02 6.52 T 0.04 6.56 T 0.02 6.54 T 0.05 6.55 T 0.10 6.56 T 0.04 6.55 T 0.02

Milk fat (%)b [ F]

7h

Milk solids (%)a [S]

6h

Temperature (-C) [T]

3.5 h

Run

0h

Table 1 Temperature, not fat solid and fat percentages used in the experimental design

Strain BFE 5264 BFE 5259 BFE 5979 BFE 6049 BFE 5092 BFE 6059 BFE 5282 BFE 5878

Strains of Lb. acidophilus, Bifidobacterium and Lactobacillus paracasei have gained particular importance and are commonly employed in probiotic products on account of their health benefits. In addition to Lactobacillus spp. and Bifidobacterium spp., also other LAB genera, some non-lactics (Holzapfel et al., 1998) and also some yeasts (Psomas et al., 2001), especially Saccharomyces boulardii (Klein et al., 1994), are used in present-day commercial probiotic products. Although commercial probiotic strains are often derived from the gastrointestinal tract (GIT) of the adult human host, studies on characterisation of dominant LAB isolated from traditional fermented Maasai milk (kule nato) in Kenya, showed the occurrence of potentially probiotic Lactobacillus and Lactococcus strains (Mathara et al., 2004a). In fact, suitable safety and functional attributes have been demonstrated for selected strains, comprising acid and bile tolerance, cholesterol assimilation, adhesion to human cell line HT 29 MTX, and binding capacity to extracellular protein matrices (Mathara et al., 2004b). The main objective of this work was to select among potentially probiotic LAB isolated from Maasai milk strains that are suitable for use as starters for the production of fermented milk, and to optimise their technological performance. In particular, eight strains from Maasai fermented milk were selected on account of their promising probiotic features (Mathara et al., 2004b) and were investigated for relevant technological characteristics such as acidification ability in whole milk, surviving capacity during storage at 4 -C and the ability to produce exo-polysaccharides and to give textural characteristics similar to those of non-stirred commercial yoghurt. A three-factor five-level Central Composite Design (CCD) was used as an approach to improve the strain technological performances and the product textural features.

30.5 h

F. Patrignani et al. / International Journal of Food Microbiology 107 (2006) 1 – 11

Table 2 Acidification kinetics, expressed as reaching pH 4.6 in hours (h), at 37 -C for the tested strains inoculated into treated whole milk supplemented with tryptone and glucose

2

F. Patrignani et al. / International Journal of Food Microbiology 107 (2006) 1 – 11

3

Table 4 Viability of tested strains in fermented milk stored at 4 -C and produced by the strain fermentations at 37 -C

2. Materials and methods 2.1. Strains Lactobacillus acidophilus BFE 6059, Lactobacillus paracasei BFE 5979, BFE 5264, BFE 5259, Lactobacillus plantarum BFE 5092, BFE 5878, BFE 5282, and Lactococcus lactis BFE 6049, originating from fermented Maasai milk and previously tested for their probiotic features, were obtained from the culture collection of the Institute of Hygiene and Toxicology, BFEL, Karlsruhe, Germany.

Strain BFE 5264 BFE 5259 BFE 5979 BFE 6049 BFE 5092 BFE 6059 BFE 5282 BFE 5878

0 days

20 days

30 days

50 days

60 days

9.63 T 0.10 9.41 T 0.09 9.25 T 0.15 9.61 T 0.08 9.41 T 0.26 9.51 T 0.20 9.58 T 0.35 9.14 T 0.30

8.61 T 0.25 8.94 T 0.30 7.25 T 0.25 7.65 T 0.27 7.69 T 0.35 7.25 T 0.15 7.25 T 0.20 8.09 T 0.25

8.60 T 0.25 8.21 T 0.30 9.71 T 0.25 8.68 T 0.15 8.08 T 0.08 8.95 T 0.25 9.57 T 0.28 8.27 T 0.35

8.05 T 0.15 8.03 T 0.24 9.10 T 0.30 7.10 T 0.34 7.68 T 0.18 8.21 T 0.25 7.80 T 0.33 8.08 T 0.40

8.32 T 0.30 8.53 T 0.38 8.94 T 0.25 7.46 T 0.28 7.81 T 0.30 8.14 T 0.16 7.91 T 0.21 8.85 T 0.15

2.2. Preparation of inoculating culture 2.5. Textural measurement The Lactobacillus strains were cultured in MRS at 37 -C and the Lactococcus strain in M17 broth (Oxoid, Basingstone, UK) at 30 -C. 1 ml of overnight culture was transferred to reconstituted skim milk (Oxoid, Basingstone, UK) in tubes and incubated at 37 -C for 24 h. The inoculating culture counts were determined on MRS at 37 -C and M17 agar at 30 -C, under anaerobic conditions, using Anaerocult (Merck, Darmstadt, Germany). 2.3. Production of fermented milk For each of the eight strains tested, two flasks with 100 ml of fresh whole milk were heat treated at 105 -C for 7 min, followed by addition of tryptone (0.5%) and fructose (0.75%) solutions (Østlie et al., 2003), previously sterilised by Millipore 0.20-Am sterile filters (Billerica, MA, USA) and inoculated with 2% of a skim milk culture obtained after an incubation at 37 -C for 24 h. For every strain, samples were incubated at 30 -C and 37 -C. After reaching pH 4.6, the samples were stored at 4 -C for 60 days. For each strain, the fermentation processes (at 30- and 37 -C) were repeated three times, each time with a new sample of raw whole milk. 2.4. Viable microorganisms The growth of each strain in fermented milk was evaluated by plate counting on MRS (Lactobacillus strains) and M17 agar media (Lc. lactis).

After 12 h storage at 4 -C, the different fermented milks were analysed for their textural features. Firmness, consistency, cohesiveness and viscosity index were evaluated using a back extrusion cell (A/AB) on a Texture Analyser TA DHI (Stable Micro System, UK) according to the manufacturer’s instructions. A solid rod (35 mm diameter) was thrusted into a cylindrical container (48 mm diameter) holding the sample using a 5 kg load cell. Two replicate tests were conducted for each yoghurt type obtained. The data reported were the means of 3 repetitions (3 different batches of raw milk) and two replicates. 2.6. Optimisation of product quality The technological performance of strains Lb. acidophilus BFE 6059, Lc. lactis BFE 6049, Lb. paracasei BFE 5264, characterised by fast acidification rates and high viability in fermented milks, were further improved by modulating compositional variables such as fat ( F), non-fat milk solids (S) and fermentation temperatures (T) according to three-factor five-level Central Composite Designs (CCD) (Box et al., 1978). The 17 combinations used for each strain are shown in Table 1. The basal medium used was the same employed for the selected strains and previously described. The different levels of full fat and non-fat milk solids were obtained by adding different concentrations of skim milk powder (Sacco srl, Cadorago, Italy) and UHT cream (Parmalat, Parma, Italy). The desired levels of skim milk were added before thermal

Table 3 Acidification kinetics, expressed as reaching pH 4.6 in hours (h), at 30 -C for the tested strains inoculated into treated whole milk supplemented with tryptone and glucose

Strain BFE 5264 BFE 5259 BFE 5979 BFE 6049 BFE 5092 BFE 6059 BFE 5282 BFE 5878 a

0h

3h

5h

7h

8h

9h

10 h

10.5 h

11 h

23.5 h

6.53 T 0.12 6.52 T 0.14 6.57 T 0.02 6.55 T 0.09 6.54 T 0.13 6.54 T 0.10 6.58 T 0.13 6.55 T 0.11

6.41 T 0.10 6.40 T 0.09 6.45 T 0.14 6.47 T 0.08 6.40 T 0.06 6.39 T 0.01 6.40 T 0.01 6.42 T 0.04

6.20 T 0.01 6.14 T 0.05 6.32 T 0.03 5.64 T 0.01 6.17 T 0.02 6.26 T 0.01 6.12 T 0.02 6.18 T 0.03

5.74 T 0.02 5.74 T 0.03 6.08 T 0.02 4.58 T 0.28 5.75 T 0.01 5.32 T 0.04 5.70 T 0.03 5.77 T 0.01

5.47 T 0.03 5.53 T 0.01 5.93 T 0.02

5.23 T 0.03 5.24 T 0.05 5.74 T 0.01

5.05 T 0.01 5.11 T 0.03 5.48 T 0.03

4.88 T 0.01 –a –

4.63 T 0.03 5.08 T 0.02 5.26 T 0.03

4.65 T 0.02 4.58 T 0.01

5.56 T 0.02 5.05 T 0.03 5.48 T 0.02 5.56 T 0.04

5.26 T 0.01 4.87 T 0.03 5.24 T 0.02 5.28 T 0.02

5.06 T 0.01 4.85 T 0.02 5.14 T 0.01 5.15 T 0.02

– 4.72 T 0.02 – –

4.65 T 0.01 4.60 T 0.08 4.62 T 0.03 5.10 T 0.02

4.59 T 0.02

Data not determined.

4

F. Patrignani et al. / International Journal of Food Microbiology 107 (2006) 1 – 11

Table 5 Viability of tested strains in fermented milk stored at 4 -C and produced by the strain fermentations at 30 -C Strain BFE 5264 BFE 5259 BFE 5979 BFE 6049 BFE 5092 BFE 6059 BFE 5282 BFE 5878

0 days

20 days

30 days

50 days

60 days

9.63 T 0.20 9.41 T 0.36 9.25 T 0.35 9.61 T 0.15 9.41 T 0.30 9.51 T 0.24 9.58 T 0.20 9.14 T 0.27

9.28 T 0.28 9.91 T 0.20 9.73 T 0.35 8.74 T 0.15 9.38 T 0.32 9.19 T 0.24 8.70 T 0.35 9.23 T 0.40

6.98 T 0.15 7.84 T 0.20 8.37 T 0.35 6.58 T 0.36 8.34 T 0.32 6.85 T 0.25 8.14 T 0.35 8.01 T 0.20

8.28 T 0.20 7.40 T 0.16 9.18 T 0.25 5.68 T 0.18 7.50 T 0.20 8.25 T 0.30 5.60 T 0.15 7.13 T 0.18

8.14 T 0.22 7.20 T 0.19 9.14 T 0.20 5.20 T 0.15 7.46 T 0.10 8.14 T 0.18 5.56 T 0.10 7.10 T 0.12

treatment, while the established UHT cream concentrations were aseptically supplemented before the inoculum of the potentially probiotic strains. The inoculum level of each run was about 7 log cfu/g. After inoculation, the samples were incubated at the temperatures established by CCD (see Table 1). For each run, the pH decrease during the fermentation was monitored. The samples were stored at 4 -C when the pH reached a value of 4.6. Immediately after fermentation and during the 30 days of refrigerated storage, the cell counts of each strain were monitored by plate counting. Moreover, after 12 h of storage at 4 -C, the rheological parameters and the exopolysaccharide concentrations (Dubois et al., 1956) were measured. The data recorded were modelled using a software Package (Statistica for Windows, Statsoft, Tulsa, USA) to fit the second order model to dependent variables, i.e. cell load, acidification rate, rheological indexes, and exo-polysaccharide production. The variables with a significance lower than 95% ( p > 0.05) were not included in the final models. Three-dimensional surface plots were drawn to illustrate the major and interactive effects of the independent variables on the dependent ones. These graphs were drawn imposing a constant value (i.e. the central points of the interval taken into consideration) to one independent variable. 2.7. pH measurements Acidification kinetics were followed using a pH-meter Hanna Instruments 8519 (INCOFAR, Modena, Italy). The data collected are the mean of three independent repetitions.

3. Results All the tested strains, when inoculated into whole milk and incubated according the procedures previously described, were able to induce fast acidification independent on the temperature used within the range under study. In fact, almost all the samples reached pH 4.6 in time spans ranging between 7 and 30.5 h (Tables 2 and 3). However, for most of the strains, acidification rates were higher at 30 -C. On the other hand, all strains were characterised by mesophilic properties, or having an optimum growth ranging between 20 and 35 -C. Strains BFE 6059, 5264 and 6049 showed the best acidification performances both at 30 -C and 37 -C. Particularly, the strain 6049 was able to reach pH 4.6 in only 7 h from the inoculum in whole milk. The viable cell numbers showed that all strains, except BFE 6049 and BFE 5282 when inoculated in milk incubated at 30 -C, maintained levels higher than 107 cfu/g over 60 days of refrigerated storage, independently on the fermentation temperature (Tables 4 and 5). However, the use of a fermentation temperature of 30 -C, instead of 37 -C, resulted in higher viability loss of almost all the strains. In order to evaluate the ability of strains to give good rheological properties, the coagula obtained (both at 37- and 30 -C), after 12 h of storage at 4 -C, were subjected to texture analysis. The data were compared with those recorded on different samples of non-stirred commercial yoghurt (chosen as control) obtained by fermentation of Streptococcus thermophilus and Lactobacillus delbrueckii subsp. bulgaricus. The results shown in Table 6 evidenced that the strains BFE Lc. lactis 6049, Lb. acidophilus BFE 6059 and Lb. paracasei BFE 5264 were able to produce at 37 -C, with respect to the control, fermented milk with four times higher firmness, twice consistency and cohesiveness. On the contrary, these coagula, except for that obtained by the strain Lc. lactis BFE 6049, were characterised by much lower values of viscosity index. Higher firmness and consistency values could be observed, in relation to the control, also for the coagula obtained by test strains at 30 -C (Table 7). However, all the examined strains, except Lb. acidophilus BFE 6059 and Lb. paracasei BFE 5979, produced coagula with higher viscosity index values at 30 -C than at 37 -C. The strains Lc. lactis BFE 6049, Lb. acidophilus BFE 6059 and Lb. paracasei BFE 5264 were selected for further

Table 6 Texture parameters detected for fermented milks produced with the tested strains by using 37 -C as fermentation temperature Strain BFE 5092 BFE 5259 BFE 5264 BFE 5282 BFE 5878 BFE 5979 BFE 6049 BFE 6059 Controla a

Firmness (g)

Consistency (g s)

Cohesiveness (g)

Index of viscosity (g s)

15.03 T 0.10 24.93 T 0.66 138.95 T 2.50 77.58 T 0.78 43.95 T 0.57 66.75 T 0.70 131.95 T 1.07 113.74 T 1.01 33.93 T 0.52

248.59 T 1.29 312.62 T 0.91 1655.66 T 7.38 938.62 T 1.32 455.57 T 1.00 1055.21 T 3.95 1932.60 T 2.57 1209.77 T 3.98 905.03 T 1.03

9.36 T 0.25 7.73 T 0.11 47.73 T 0.82 21.49 T 0.36 8.32 T 0.24 30.27 T 0.19 53.89 T 0.92 34.53 T 0.78 25.63 T 0.15

15.48 T 0.08 8.72 T 0.10 37.84 T 0.56 35.91 T 0.72 12.34 T 0.15 58.49 T 0.43 60.68 T 0.68 22.39 T 0.16 66.31 T 0.44

Control was a commercial non-stirred yoghurt produced by fermentation of Streptococcus thermophilus and Lactobacillus delbrueckii subsp. bulgaricus.

F. Patrignani et al. / International Journal of Food Microbiology 107 (2006) 1 – 11

5

Table 7 Texture parameters detected for fermented milks produced with the tested strains by using 30 -C as fermentation temperature Strain BFE 5092 BFE 5259 BFE 5264 BFE 5282 BFE 5878 BFE 5979 BFE 6049 BFE 6059 Controla a

Firmness (g)

Consistency (g s)

Cohesiveness (g)

Index of viscosity (g s)

113.98 T 0.75 113.16 T 0.69 88.89 T 0.25 108.54 T 0.64 117.86 T 0.92 102.19 T 0.91 111.24 T 0.74 53.59 T 0.35 33.93 T 0.52

1951.11 T 2.41 1933.06 T 2.27 1387.27 T 1.97 1845.82 T 2.97 1963.17 T 3.02 1797.82 T 2.85 1896.30 T 2.47 916.88 T 1.98 905.03 T 1.03

41.32 T 0.15 45.03 T 0.38 33.04 T 0.21 42.90 T 0.29 41.97 T 0.20 41.53 T 0.19 54.98 T 0.23 20.99 T 0.10 25.63 T 0.15

58.23 T 0.35 65.85 T 0.48 59.78 T 0.39 63.88 T 0.95 62.49 T 0.80 50.91 T 0.45 81.70 T 0.78 26.96 T 0.27 66.31 T 0.24

Control was a commercial not stirred yoghurt produced by fermentation of Streptococcus thermophilus and Lactobacillus delbrueckii subsp. bulgaricus.

technical studies, with the purpose of finding optimal conditions for substrate composition and for fermentation. In fact, the chosen strains were characterised by fast acidification rates (less than 12 h to reach pH 4.6) and by the ability to give coagula with firmness, consistency and cohesiveness indexes higher than those of non-stirred commercial yoghurt. Moreover, these strains were characterised by limited viability losses during refrigerated storage. In particular, the strains BFE 6059 and BFE 5264 maintained viable cell numbers higher than 7 log cfu/g independently on fermentation temperature used within the range 26.5 and 37 -C, and during 60 days refrigerated storage. By comparison, viability of Lc. lactis strain BFE 6049 had a marked decrease during the storage, particularly when the fermentation had been performed at 30 -C. However, also at this last fermentation temperature, its viable numbers remained > 6 log cfu/g, even up to 30 days of storage, which may be considered as minimum threshold value required for probiotic strains in this kind of product. Moreover, Lc. lactis strain BFE 6049, although comparatively less resistant to extended refrigerated storage, was characterised in previous work both by desirable probiotic properties and by the highest acidification rate. In order to improve technical performances of strains and sensory features of the fermented milks, the modulating compositional variables such as fat ( F) and non-fat milk solids (S) and fermentation temperatures (T), a CCD was used for each strain as shown in Table 1.

The data recorded for each strain were modelled according to the polynomial quadratic equations in order to identify the independent variables that significantly affected the fermentation rate, the cell load during the storage, the exopolysaccharide production as well as rheological parameters of final products. The best-fit equations obtained, reported in Tables 8– 10, allowed us to evaluate the effects of linear, quadratic and interactive terms of the independent variables temperature (T), skim milk (S) and UHT cream ( F) on the chosen dependent variables. In order to better understand the interactive effects of the independent variables, surface plots were drawn. As shown in Fig. 1a, b and c, the maximum viable cell numbers of Lb. acidophilus BFE 6059, after 30 days of storage at 4 -C, were associated with a fermentation temperature of 30 -C, and with added skim milk and UHT cream, ranging between 3% and 4.5% and 1.5% and 2.25%, respectively. The minimal coagulation time for this strain was recorded in the presence of 6% skim milk and with a fermentation temperature of 30 -C. As evidenced by Fig. 2, drawn from Eq. (1), Table 10, with UHT cream at its central value (1.5%), the maximum coagulation times for the strain Lb. acidophilus were observed both at lowest temperature and highest skim milk concentrations as well as at highest temperature and lowest skim milk values. By contrast, the lowest coagulation times were observed for the highest temperature and skim milk concentrations as well as for the lowest values of these independent variables. Similar results

Table 8 Best-fit equations relative to the effects of the different CCD variables on the fermentation time and viability for the strain Lactobacillus paracasei BFE 5264 as well as exo-polysaccharide content and rheological parameters of fermented milk obtained Equationsa Eq. Eq. Eq. Eq. Eq. Eq. Eq. a b c d e

(1) (2) (3) (4) (5) (6) (7)

(fermentation time) (cell loads at 30 days) (exo-polysaccharide content) (firmness) (cohesiveness) (consistency) (index of viscosity)

5.018[T] 0.472[T][ F] 0.616[T] 0.010[T 2] 9.515[T] 0.158[T 2] 5.529[T] 2.288[T] 82.602[T] 0.058[T 2]

0.236[T][S] + 4.754[ F][S]

[T] = temperature (-C); [ F] = UHT cream (%); [S] = skim milk (%). Regression coefficient. F-value. Only terms with p < 0.05 were included. Standard error.

2

0.113[T]

Rb

Fc

Pd

S.E.e

0.982 0.999 0.999 0.868 0.859 0.869 0.835

68.424 24 377 15 118 48.783 45.152 49.783 38.835

<0.000 <0.000 <0.000 <0.000 <0.000 <0.000 <0.000

6.398 0.169 3.358 98.556 42.384 1458.0 36.604

6

F. Patrignani et al. / International Journal of Food Microbiology 107 (2006) 1 – 11

Table 9 Best-fit equations relative to the effects of the different CCD variables on the fermentation time and viability for the strain Lactococcus lactis BFE 6049 as well as exo-polysaccharide content and rheological parameters of fermented milk obtained

Eq. Eq. Eq. Eq. Eq. Eq. Eq. a b c d e

(1) (2) (3) (4) (5) (6) (7)

(fermentation time) (cell load at 30 days) (exo-polysaccharide content) (firmness) (cohesiveness) (consistency) (index of viscosity)

Equationsa

Rb

Fc

Pd

S.E.e

0.317[T] + 7.9494[S] 0.263[T][S] 0.574[T] 0.012[T][ F] + 0.095[ F][S] 0.001[T]2 9.475[T] 0.157[T]2 4.429[T] + 0.306[T][S] 41.589[ F] 1.330[T][ F] + 0.182[T][S] + 0.61[T]2 68.668[T] + 12.603[T][S] 37.516[S]2 3.858[T] + 15.419[ F][S] 23.414[ F]2 4.564[S]2

0.986 0.999 0.999 0.996 0.997 0.997 0.939

164.20 2304.7 19436 1017.4 516.43 852.42 24.637

<0.000 <0.000 <0.000 <0.000 <0.000 <0.000 <0.000

1.829 0.374 2.956 14.806 6.8823 229.25 31.511

[T] = temperature (-C); [ F] = UHT cream (%); [S] = skim milk (%). Regression coefficient. F-value. Only terms with p < 0.05 were included. Standard error.

were also obtained for the coagulation time of BFE 6049 (data not shown). The fermentation temperature of 30 -C was optimal for Lc. lactis strain BFE 6049 for maintaining a high viability during storage (Fig. 3b). The viable cell numbers of this strain were significantly affected also by skim milk and UHT cream. In fact, the highest viability was obtained with the highest levels of these two independent variables (Fig. 3a). However, in absence of UHT cream, the increase of skim milk concentration did not result in an increase of viability of the considered strain. On the other hand, without skim milk added, the supplementation of UHT cream had a negative effect on viability. This last variable, after 30 days of refrigerated storage for strain Lb. paracasei BFE 5264, was significantly affected only by fermentation temperature (Eq. (2), Table 8), and reached an optimum at 30 -C. The coagulation time of samples inoculated with Lb. paracasei BFE 5264 was the fastest at the highest fermentation temperature studied, and for the CCD central levels of UHT and skim milk added (1.5% and 3%, respectively) (figures not shown). By contrast, when the temperature was at its CCD central level (30 -C), the addition of 6% of skim milk and 3% of UHT cream determined a coagulation time longer than 40 h (figures not shown). Consistency and cohesiveness of samples produced by fermentation with Lc. lactis BFE 6049, considering the CCD

central level of UHT cream (Fig. 4a and b) increased when fermentation temperature and skim milk concentration increased. As shown in Fig. 4c, increasing the fermentation temperature and the addition of skim milk up to 3% had positive effects also on product viscosity. By contrast, the firmness of the coagulum, obtained by the same strain, was not influenced by skim milk levels (not shown). For same samples, cohesiveness and viscosity were affected by UHT cream concentrations as Fig. 4d, e and f show. In particular, for fermentation temperature lower than 30 -C, cohesiveness of the product was positively affected by increasing UHT cream concentrations. As shown in Fig. 4f, obtained maintaining skim milk at its central value (3%), the addition of cream had a positive influence on viscosity only up to levels of 1.5%, independently on fermentation temperature. In fact, viscosity reached a maximum for samples fermented by strain Lc. lactis BFE 6049 for intermediate levels of cream added and other conditions being equal. However, the increase of fermentation temperature, independently on UHT cream level, had a positive effect on viscosity of the product. Fat concentration did not exhibit significant influence on the rheological parameters of fermented milk produced by Lb. acidophilus strain BFE 6059. In fact, firmness, cohesiveness, consistency and viscosity were effected only by temperature

Table 10 Best-fit equations relative to the effects of the different variables on the fermentation time and viability for the strain Lactobacillus acidophilus BFE 6059 as well as exo-polysaccharide content and rheological parameters of fermented milk obtained Equationsa Eq. Eq. Eq. Eq. Eq. Eq. Eq. a b c d e

(1) (2) (3) (4) (5) (6) (7)

(fermentation time) (cell load at 30 days) (exo-polysaccharide content) (firmness) (cohesiveness) (consistency) (index of viscosity)

2

40.225[S] 1.309[T][S] + 0.031[T] 0.580[T] + 0.377[ F] + 0.246[S] 0.009[T]2 9.277[T] 0.147[T]2 240.911[S] 8.370[T][S] + 0.297[T]2 0.111[T]2 250.124[T] + 10.375[T][S] + 11.351[T]2 109.791[S] + 4.160[T][S]

[T] = temperature (-C); [ F] = UHT cream (%); [S] = skim milk (%). Regression coefficient. F-value. Only terms with p < 0.05 were included. Standard error.

0.114[ F]2

0.037[S]2

Rb

Fc

Pd

S.E.e

0.971 0.999 0.999 0.962 0.973 0.995 0.783

77.159 11376 7792.1 49.656 249.91 441.88 10.290

<0.000 <0.000 <0.000 <0.000 <0.000 <0.000 <0.002

8.769 0.146 4.764 75.350 24.714 460.23 60.426

F. Patrignani et al. / International Journal of Food Microbiology 107 (2006) 1 – 11

7

and non-fat dry matter added, while the highest fat concentrations did not affect these parameters. Both temperature and non-fat dry matter positively influenced product viscosity and

Fig. 2. Three-dimensional contour plot concerning the effects of the interactions [T]  [S] on Lactobacillus acidophilus BFE 6059 coagulation time.

a

b

Fig. 1. Three-dimensional contour plots concerning the effects of the interactions [ F]  [S] (a), [T]  [S] (b), [T]  [ F] (c) on Lactobacillus acidophilus BFE 6059 cell loads in fermented milk after 30 days of storage at 4 -C.

Fig. 3. Three-dimensional contour plots concerning the effects of the interactions [S]  [ F] (a) and [T]  [S] (b) on Lactococcus lactis BFE 6049 cell loads in fermented milk after 30 days of storage at 4 -C.

8

F. Patrignani et al. / International Journal of Food Microbiology 107 (2006) 1 – 11

Fig. 4. Three-dimensional contour plots concerning the consistency, cohesiveness and index of viscosity of fermented milk obtained by Lactococcus lactis BFE 6049 fermentation. The figures are relative to the effects of interaction [T]  [S] (a) on consistency, [T]  [S] (b) on cohesiveness, [T]  [S] and [T]  [ F] (c), (f) on viscosity index, [T]  [ F] (d) and [S]  [ F] (e) on cohesiveness of final products.

consistency. The highest values for these parameters were obtained for samples fermented at 37 -C and containing 6% of skim milk (Fig. 5a). These conditions determined a maximum

level for the viscosity index (Fig. 5b). This last index was positively affected by fermentation temperature increase, in particular when skim milk concentrations were higher than 3%.

F. Patrignani et al. / International Journal of Food Microbiology 107 (2006) 1 – 11

a

b

Fig. 5. Three-dimensional contour plots concerning the effects of interaction [T]  [S] (a) on consistency and [T]  [S] (b) on viscosity of fermented milk obtained by Lactobacillus acidophilus BFE 605 9 fermentation.

Analogously, the skim milk addition determined an increase of firmness when the fermentation temperature was lower than 30 -C (not shown). The rheological indexes for fermented samples produced by Lb. paracasei strain BFE 5264, were affected only by fermentation temperature (Eqs. (4), (5), (6) and (7), Table 8). For all strains, exo-polysaccharide production was positively influenced by temperature as single parameter, but negatively by temperature as quadratic term. Consequently, based on these equations, for each strain tested, it was possible to identify the optimal temperature for the production of exopolysaccharides. These temperatures were about 30 -C for the strains BFE 6049 and BFE 5264, and 32 -C for the strain BFE 6059. 4. Discussion The health benefits from milk co-fermented with probiotic Lactic Acid Bacteria (LAB) are well documented, and depend on high viability of the probiotic microorganisms (Gardini et al., 1999). To produce a satisfactory fermented probiotic milk

9

product, the viable cell count at the time of consumption should be above 6 log cfu/g in order to comply with the FIL-IDF standard (IDF, 1992) and to supply a sufficient ‘‘daily dose’’ of viable bacteria (Samona and Robinson, 1991; Vinderola et al., 2000). Therefore, it is essential that selected strains, with proven probiotic activity, are able to maintain high viability in refrigerated product. This condition could represent a principle for the selection of probiotic strains to employ in fermented milk products as adjuncts. In fact, cases of marked loss of viability in refrigerated dairy products have been reported for Bifidobacterium spp. and strains of Lb. acidophilus, the most commonly used probiotic bacteria (Gomes and Malcata, 1999). The need for careful strain selection, particularly when other LAB species are employed as probiotic cultures, was thus pointed out. In fact, also strains belonging to Lactobacillus casei (Lb. paracasei), Lactobacillus rhamnosus and Lb. plantarum are commonly applied in probiotic products or detected in novel types of yoghurt (Suskovic et al., 1997; Holzapfel et al., 1998; Bernardeau et al., 2001). These strains are well studied for their probiotic features but only few reports focused simultaneously on their technological properties (Olasupo et al., 2001). Technological aspects to be considered in probiotic LAB selection for fermented milk include the phage resistance, viability throughout processing and storage, ability to give rise to fast fermentation in an improper system such as milk, and to impart good sensory properties (Saxelin et al., 1999; Mattila-Sandholm et al., 2002). This latter aspect plays an important role in consumer acceptance (Gardini et al., 1999). The results obtained in this investigation have permitted the selection of some potentially probiotic strains endowed with interesting technical properties. In particular the strains Lc. lactis BFE 6049, Lb. acidophilus BFE 6059, and Lb. paracasei BFE 5264 showed high viability during refrigerated storage, following fast acidification both at 30 - and 37 -C. Moreover, these strains gave rise to products characterised by rheological indexes such as firmness, cohesiveness and consistency higher than those of commercial yoghurt obtained with Lb. delbrueckii subsp. bulgaricus and S. thermophilus, also when used in pure culture. The ability to impart good rheological properties is an important selection criterion for starter cultures. In order to improve texture and body of fermented milks and yoghurts, several strategies have been proposed. In particular, the application of hydrostatic or dynamic high pressure (Lanciotti et al., 2004) and the use of strains able to produce exopolysaccharides have been proposed as alternatives to the use of additives, which can adversely affect the yoghurt taste, flavour, aroma and mouth-feel (Marshall and Rawson, 1999; De Ancos et al., 2000). To improve the textural and aromatic features of this kind of products, also co-inocula of probiotic strains with Lb. delbrueckii subsp. bulgaricus and S. thermophilus have been proposed (Gomes and Malcata, 1999). Moreover, S. thermophilus is applied to sustain the mild acid fermentation typical of yoghurt containing probiotic lactobacilli (Holzapfel et al., 1998). However, the use of a combination of usual starter and probiotic bacteria necessitates

10

F. Patrignani et al. / International Journal of Food Microbiology 107 (2006) 1 – 11

the study of strain interactions for accurate selection of compatible strains able to confer good properties to the product, and maintain a high viability during the storage, as well as for the identification of the appropriate operative conditions (Ishibashi and Shimamura, 1993; Samona et al., 1996). Moreover, the use of co-cultures complicates the traceability of the probiotic strains. In fact, to exert beneficial effects on the host, the probiotic strains have to maintain their viability also when subjected to competition by other strains during fermentation. The use of pure probiotic starter cultures such as the strain in this investigation, endowed with appropriate characteristics, could represent an interesting and practicable alternative to the use of mixed cultures consisting of both probiotic and starter strains. A further tool to improve technological strain features and sensory properties of fermented milks can be identified as the modulation of physico-chemical variables commonly used at industrial level (Torriani et al., 1996; Gardini et al., 1999). In fact, the use of CCD and the modelling of the data obtained, allowed to identify the effect of the compositional variables and their interactions, both on the viability of the strains in the product, as well as on technological properties of the samples. In particular, the polynomial models obtained and the relative response surfaces allowed, for each strain considered, the identification of: (a) the levels of the three independent variables able to minimise both the fermentation time and loss of viability during refrigerated storage; (b) the optimal fermentation temperature for the exo-polysaccharide production; and (c) the levels of the three independent variables able to impart specific rheological features to the final products. Since the conditions assuring the achievement of major consistency do not necessarily coincide with the assuring of best viscosity or cohesiveness, the choice of the levels of the three independent variables is subordinate to the desired rheological traits of final product. For this reason, the response surface methodology can represent a useful tool also to differentiate the probiotic fermented milk for rheological properties, expanding the product gamma in order to satisfy the heterogeneous consumer demand. In fact, the probiotic fermented milk sector is characterized, in comparison with the traditional yoghurt one, by a quantitative wide offering of products but very homogeneous for sensorial features. References Bernardeau, M., Vernoux, J.P., Gueguen, M., 2001. Probiotic properties of two Lactobacillus strains in vitro. Milchwissenschaft 56 (12), 663 – 667. Box, G.E.P., Hunter, W.G., Hunter, J.S., 1978. Statistics for Experimenters. An Introduction to Design Data Analysis and Models Building. John Wiley & Sons, New York. De Ancos, B., Cano, M.P., Gomez, R., 2000. Characteristics of stirred lowfat yoghurt as affected by high pressure. International Dairy Journal 10, 105 – 111. Dubois, M., Gilles, K.A., Hamilton, J.K., Roberts, P.A., Smith, F., 1956. Colorimetric determination of sugars and related substances. Analytical Chemistry 28, 350 – 356. Gardini, F., Lanciotti, R., Guerzoni, M.E., Torriani, S., 1999. Evaluation of aroma production and survival of Streptococcus thermophilus, Lactobacil-

lus delbrueckii subsp. bulgaricus and Lactobacillus acidophilus in fermented milks. International Dairy Journal 9, 125 – 134. Gomes, A.M.P., Malcata, F.X., 1999. Bifidobacterium spp. and Lactobacillus acidophilus: biological, biochemical, technological and therapeutical properties relevant for use as probiotics. Trends in Food Science and Technology 10 (1999), 139 – 157. Holzapfel, W.H., Haberer, P., Snel, J., Schillinger, U., Huis in’t Veld, J.H.J., 1998. Overview of gut flora and probiotics. International Journal of Food Microbiology 41, 85 – 101. IDF, 1992. General Standard of Identification for Fermented Milks. Standard 163. International Dairy Federation, Brussels. Ishibashi, N., Shimamura, S., 1993. Bifidobacteria: research and development in Japan. Food Technology, 126 – 135 (June). Klein, S.M., Elmer, G.W., McFarland, L.V., Surawier, C.M., Levy, H.R., 1994. Recovery and elimination of the biotherapeutic agent Saccharomyces boulardii in healthy human volunteers. Pharmacological Research 10, 1615 – 1619. Lanciotti, R., Vannini, L., Pittia, P., Guerzoni, M.E., 2004. Suitability of highdynamic-pressure-treated milk for the production of yoghurt. Food Microbiology 21 (6), 753 – 760. Lee, Y.-K., Salminen, S., 1995. The coming of age of probiotics. Trends in Food Science & Technology 6, 241 – 245. Marshall, V.M., Cole, W.M., 1983. Threonine aldolase and alcohol dehydrogenase activities in Lactobacillus bulgaricus and Lactobacillus acidophilus and their contribution to flavour production in fermented milks. Journal of Dairy Research 50, 375 – 379. Marshall, M.V., Rawson, H.L., 1999. Effects of exo-polysaccharideproducing strains of thermophilic lactic acid bacteria on the texture of stirred yoghurth. International Journal of Food Science & Technology 34, 137 – 143. Mathara, J.M., Schillinger, U., Kutima, P.M., Mbugua, K.S., Holzapfel, W.H., 2004a. Isolation, identification and characterisation of the dominant microorganisms of kule naoto: the Maasai traditional fermented milk in Kenya. International Journal of Food Microbiology 94, 269 – 278. Mathara, J.M., Schillinger, U., Kutima Museve, P., Mbugua, S.K., Holzapfel, W.H., 2004b. Functional characteristics of lactic acid bacteria isolated from traditional Maasai fermented milk products in Kenya. Book of Abstract of 19th International ICFMH Symposium, Food Micro 2004, New Tools for Improving Microbial Food Safety and Quality, p. 382. Mattila-Sandholm, T., Matto, J., Saarela, M., 1999. Lactic acid bacteria with health claims interference and interactions with gastrointestinal flora. International Dairy Journal 9, 25 – 35. Mattila-Sandholm, T., Myllarinen, P., Crittenden, R., Mogensen, G., Fonde´n, R., Saarela, M., 2002. Technological challenger for future probiotic foods. International Dairy Journal 12, 173 – 182. Olasupo, N.A., Shillinger, U., Holzapfel, W.H., 2001. Studies on some technological properties of predominant lactic acid bacteria isolated from Nigerian fermented foods. Food Biotechnology 15 (3), 157 – 167. Østlie, H., Helland, M.H., Narvhus, J.A., 2003. Growth and metabolism of selected strains of probiotic bacteria in milk. International Journal of Food Microbiology 87, 17 – 27. Psomas, E., Andrighetto, C., Litopoulou-Tzanetaki, E., Lombardi, A., Tzanetakis, N., 2001. Some probiotic properties of yeast isolates from infant faeces and feta cheese. International Journal of Food Microbiology 69, 125 – 133. Saarela, M., Mogensen, G., Fonde´n, R., Matto, J., Mattila-Sandholm, T., 2000. Probiotic bacteria: safety, functional and technological properties. Journal of Biotechnology 84, 197 – 215. Salminen, S., Isolauri, E., Salminen, E., 1996. Clinical uses of probiotics for stabilizing the gut mucosal barrier: successful strains and future challenges. Antonie van Leeuwenhoek 70, 347 – 358. Salminen, S., von Wright, A., Morelli, L., Marteau, P., Brassart, D., de Vos, W.M., Fonde´n, R., Saxelin, M., Collins, K., Mogensen, G., Birkeland, S.E., Mattila-Sandholm, T., 1998. Demonstration of safety of probiotics—a review. International Journal of Food Microbiology 44, 93 – 106. Samona, A., Robinson, R.K., 1991. Enumeration of bifidobacteria in dairy products. Journal of the Society of Dairy Technology 44, 64 – 66.

F. Patrignani et al. / International Journal of Food Microbiology 107 (2006) 1 – 11 Samona, A., Robinson, R.K., Marakis, S., 1996. Acid production by bifidobacteria and yoghurt bacteria during fermentation and storage of milk. Food Microbiology 13, 275 – 280. Saxelin, M., Grenov, B., Svensson, U., Fonden, R., Reniero, R., MattilaSandholm, T., 1999. The technology of probiotics. Trends in Food Science & Technology 10, 387 – 392. Stanton, C., Gardiner, G., Lynch, P.B., Collins, J.K., Fitzgerald, G., Ross, R.P., 1998. Probiotic cheese. International Dairy Journal 8, 491 – 496. Suskovic, J., Brkic, B., matosic, S., Maric, V., 1997. Lactobacillus acidophilus M92 as potential probiotic strain. Milchwissenschaft 52 (8), 430 – 435.

11

Torriani, S., Gardini, F., Guerzoni, M.E., Dellaglio, F., 1996. Use of response surface methodology to evacuate some variables affecting the growth and acidification characteristics of yoghurt cultures. International Dairy Journal 6, 625 – 636. Vinderola, C.G., Bailo, N., Reinheimer, J.A., 2000. Survival of probiotic microflora in Argentinean yoghurts during refrigerated storage. Food Research International 33, 97 – 102.