Isolation and identification of lactobacilli from novel-type probiotic and mild yoghurts and their stability during refrigerated storage

International Journal of Food Microbiology 47 (1999) 79–87

Isolation and identification of lactobacilli from novel-type probiotic and mild yoghurts and their stability during refrigerated storage U. Schillinger* Federal Research Centre for Nutrition, Institute of Hygiene and Toxicology, Haid- und Neustr. 9, D-76131 Karlsruhe, Germany Received 3 August 1998; received in revised form 4 December 1998; accepted 30 January 1999

Abstract A total of 26 Lactobacillus strains were isolated from various mild yoghurts and novel-type probiotic dairy products and from a starter culture preparation and were identified by using DNA–DNA hybridization technique. The species present in those products were found to be Lactobacillus acidophilus, L. johnsonii, L. crispatus, L. casei, L. paracasei and L. rhamnosus. DNA homology analysis revealed that some strains had been misclassified by their investigators. Three strains designated as L. acidophilus (L. acidophilus LA-1, L. acidophilus ATCC 43121 and the Lactobacillus strain from Biogarde  culture) were found to belong to L. johnsonii and L. acidophilus L1 to be L. crispatus. Strains designated as L. casei were found to be members of three separate species: L. casei, L. paracasei and L. rhamnosus. Viable numbers of lactobacilli in mild and probiotic yoghurts varied greatly including some products with very low Lactobacillus counts. The majority of the probiotic yoghurts, however, contained viable counts above 10 5 per g even at the end of the best before use period.  1999 Elsevier Science B.V. All rights reserved. Keywords: Lactobacilli; Probiotics; Yoghurts; DNA homology; Stability

1. Introduction Recently, a number of novel fermented dairy products have been developed and are being marketed under the name of probiotic products in Germany and other European countries. Due to their increasing popularity among consumers the number

*Tel.: 1 49-721-6625-152; fax: 1 49-721-6625-111. E-mail address: [email protected] (U. Schillinger)

of probiotic dairy products on the German market has increased tremendously during the last few years. A wide variety of new yoghurt-like products are available now. They are expected to exert a beneficial effect on the health status of the consumer as they contain selected lactic acid bacteria (LAB) isolated from the human intestine which are claimed to have probiotic properties. The application of LAB cultures of human origin in the manufacture of fermented milk products is not a new concept. More than 20 years ago, strains of Lactobacillus acidophilus and Bifidobacterium bifidum were introduced into dairy products because of the potential

0168-1605 / 99 / $ – see front matter  1999 Elsevier Science B.V. All rights reserved. PII: S0168-1605( 99 )00014-8

80

U. Schillinger / International Journal of Food Microbiology 47 (1999) 79 – 87

advantage of consuming active LAB adapted to the intestine and to produce mildly acidified yoghurts (Schuler-Malyoth et al., 1968). In Germany, those products became known as mild yoghurts or ‘bioyoghurts’. In the USA, acidophilus milk was developed. Also the majority of the new probiotic products contain bifidobacteria, strains of L. acidophilus or closely related species (so-called L. acidophilus group). Strains of the so-called L. casei group comprising the species L. casei, L. paracasei subsp. paracasei and subsp. tolerans and L. rhamnosus are also being increasingly applied in novel-type yoghurts. The taxonomy of both L. acidophilus and L. casei group has been subjected to considerable changes during recent years and may have caused some confusion (Collins et al., 1989; Fujisawa et al., 1992; Pot et al., 1993; Dicks et al., 1996; Mori et al., 1997). For instance, in 1989 the majority of strains classified as L. casei were separated from this species and received a new name, L. paracasei (Collins et al., 1989). In 1996, however, rejection of the name L. paracasei and revision of the species L. casei was proposed (Dicks et al., 1996) and the name L. zeae was introduced for strains formerly belonging to L. casei. Incorrect designations of the starter cultures by the manufacturers are still common and species names used often do not correspond to correct nomenclature. For instance on the labelling of some yoghurts the name Lactobacillus bifidus is still used for Bifidobacterium bifidum which is not a Lactobacillus species. Moreover, strains used in the manufacture of yoghurt are not always indicated on the label. For these reasons, it seemed to us that the identity and taxonomy of LAB used in the manufacture of novel-type probiotic yoghurts needed clarification. In order to get more information on the taxonomic status of LAB included in probiotic dairy products, strains of Lactobacillus were isolated from a number of yoghurt-like products and were identified using phenotypic properties and by DNA–DNA hybridization. Our studies also included a number of mild yoghurts for comparison. In order to get some information on the viability and stability of the strains in those products, viable counts of lactobacilli were determined at the day of purchase and at the end of the best before use period.

2. Materials and methods

2.1. Bacterial strains A total of 26 strains were isolated from various mild and probiotic yoghurt products and from a yoghurt starter culture (Table 1). L. acidophilus DSM 20079, L. amylovorus DSM 20531, L. casei DSM 20011, L. crispatus DSM 20584, L. gasseri DSM 20077, L. intestinalis DSM 6629, L. johnsonii DSM 10533, L. johnsonii DSM 20553, L. paracasei DSM 5622, L. plantarum DSM 20174 and L. rhamnosus DSM 20021 received from the DSM (German Collection of Microorganisms, Braunschweig) were used as authentic strains for comparison in the DNA homology studies. L. acidophilus ATCC 43121 (RP32) was also included in the DNA–DNA hybridizations. All strains were cultivated in MRS broth (De Man et al., 1960) at 378C.

2.2. Isolation and maintenance of strains Ten-gram samples of yoghurts were homogenised with 90 ml of quarter-strength Ringers solution by shaking for several minutes. From this suspension, serial dilutions were made in quarter-strength Ringers solution and plated by spreading 0.1 ml onto the surface of MRS agar (Merck, Darmstadt, Germany). After anaerobic incubation for 48 h at 378C colonies were enumerated and one colony of each type was isolated and tested for catalase reaction. Lactobacilli were distinguished from Streptococcus thermophilus and Bifidobacterium species by microscopic examination. Cells of all Lactobacillus strains isolated were rod shaped and therefore morphologically distinguishable from the spherical or ovoid cells of S. thermophilus and the typical bifid morphology of Bifidobacterium cells. Isolated strains were maintained as frozen cultures in MRS broth with 15% glycerol at 2 808C.

2.3. Phenotypic characterization Growth at 15 and 458C was tested in MRS broth after both 2 and 5 days of incubation, respectively. Gas (CO 2 ) production from glucose, fermentation of carbohydrates and the configuration of lactic acid

U. Schillinger / International Journal of Food Microbiology 47 (1999) 79 – 87

enantiomers produced were determined by the meth¨ ods of Schillinger and Lucke (1987).

2.4. Determination of mol.% G 1 C and DNA homology DNA of isolates and reference strains was isolated and purified according to the method of Marmur (1961) modified as described by Stackebrandt and Kandler (1979) . In some cases DNA isolation was performed by using an anion exchange column included in the NucleoBond kit for purifying genomic DNA (Clontech, Heidelberg, Germany). The DNA base composition was estimated from the thermal melting point of DNA (T M ) as described by Marmur and Doty (1962) using a Gilford Response spectrophotometer (Ciba Corning Diagnostics, Gilford). DNA homology was determined spectrophotometrically by using the modified (Huss et al., 1983) method of De Ley et al. (1970).

2.5. Enumeration of Lactobacillus viable counts In order to determine the viability of lactobacilli in mild and probiotic yoghurts during refrigerated storage, 21 different yoghurts (Table 5) were bought in supermarkets three or four times within the time period from 1995 to 1998. In 1995, determinations of Lactobacillus viable counts in yoghurts were started including yoghurts A, B, E, F, K, and L. Yoghurts B, E, F, K, and L were bought again at two different occasions in 1996 to assess the viability of lactobacilli in yoghurts during refrigerated storage and in the same year nine other yoghurts (C, H, I, J, M, N, Q, R, U) were investigated one time (yoghurt U), two times (C, I, J, R), three times (H, N) or four times (M, Q). In 1997, five novel probiotic yoghurts (S, T, W, X, Y) were analysed two times and additionally, viability of lactobacilli was assessed in yoghurts A, D, F, H, J, and U. In 1998, yoghurts A, B, C, D, I, N, R, S, T, U, W, X, and Y were assayed for Lactobacillus viable counts and in case of yoghurts A and D enumerations were done twice. Generally Lactobacillus viable counts were determined at the day of purchase and after storage at 48C till the end of the best before use period (about 2 or 3 weeks in most cases). For enumeration of viable lactobacilli, appropriate dilutions from yoghurts were

81

prepared as described above and spread on MRS agar. After anaerobic incubation at 378C for at least 48 h the different types of colonies on plates were counted. Lactobacilli were distinguished from bifidobacteria by differences in the colony morphology and by microscopic examination. In products containing more than one Lactobacillus strain such as yoghurt Q and Y, also differences in colony morphology allowed separate enumeration of the strains.

3. Results

3.1. Identification of Lactobacillus isolates Twenty-five Lactobacillus strains were isolated from eight different mild yoghurts and from 13 novel-type probiotic dairy products. A freeze-dried commercial starter culture preparation for mild yoghurts was also included in the investigations. Viable numbers of the lactobacilli present in the yoghurts were determined and are given in Table 1. All isolates were catalase-negative rods producing no gas from glucose. They were presumptively identified by determining growth behaviour at 158C and 458C, the enantiomer of lactic acid produced, the pattern of fermented carbohydrates (data not shown) and DNA base composition. On the basis of these properties, one strain (BFE 700) could be identified as L. plantarum (Table 2). Seventeen strains not growing at 158C and producing DL lactate from glucose were found to be members of the so-called L. acidophilus group (Fujisawa et al., 1992) (Table 2). Their G 1 C content of DNA was between 32.5 and 35.8. Sugar fermentation pattern did not allow a differentiation between L. acidophilus and the phenotypically very similar species L. johnsonii, L. crispatus, L. gasseri, L. amylovorus and L. gallinarum. Eight strains able to grow at 158C and producing either DL (one strain) or the L-enantiomer of lactic acid exclusively (seven strains) were identified as members of the L. casei group comprising the species L. paracasei, L. casei and L. rhamnosus. Their G 1 C content of DNA ranged from 45.0 to 47.0. DNA–DNA hybridizations with authentic strains of the relevant species enabled a clear identification

U. Schillinger / International Journal of Food Microbiology 47 (1999) 79 – 87

82

Table 1 Identification of Lactobacillus strains isolated from various mild and probiotic yoghurts or yoghurt-like products Isolate BFE no.

Origin (yoghurt code)

Viable counts per g

Declaration by the manufacturer

648 649 652 654 659 662 692 695 700 663 664 665 675 682 687 693 696 697 704 718 719 720 723 727 728 679

A B C D E F G (F)a H I K L L M N O P (NL)a Q Q R T U V (S)a W Y Y Starter culture

5.6 3 10 6 n.d. 3.4 3 10 3 1.6 3 10 4 n.d. n.d. n.d. 6.3 3 10 7 7.0 3 10 4 1.3 3 10 7 4.7 3 10 6 9.0 3 10 6 1.8 3 10 8 3.0 3 10 5 8.8 3 10 7 1.8 3 10 3 1.7 3 10 8 6.0 3 10 7 2.0 3 10 5 3.4 3 10 5 3.3 3 10 5 1.6 3 10 9 4.6 3 10 8 1.4 3 10 8 4.0 3 10 5 n.d.

Biogarde  culture Not specified Bioghurt  culture Biogarde  culture Biogarde  culture Bioghurt  culture L. acidophilus Not specified Not specified L. acidophilus LA-1 L. casei GG Not specified L. casei Actimel Bactolab culture L. casei Shirota L. acidophilus Gilliland L. acidophilus L. casei L. acidophilus LA-7 L. acidophilus Not specified L. acidophilus L. casei L. acidophilus LA-H3 L. casei LC-H2 L. acidophilus

a

Identification based on Phenotypic properties

DNA–DNA homology

L. L. L. L. L. L. L. L. L. L. L. L. L. L. L. L. L. L. L. L. L. L. L. L. L. L.

L. rhamnosus Lactobac. spec. L. johnsonii L. johnsonii L. rhamnosus L. johnsonii L. johnsonii L. acidophilus L. plantarum L. johnsonii L. rhamnosus L. acidophilus L. paracasei L. acidophilus L. paracasei L. crispatus L. acidophilus L. paracasei L. acidophilus L. johnsonii L. acidophilus L. acidophilus L. paracasei L. acidophilus L. casei L. acidophilus

casei group acidophilus group acidophilus group acidophilus group casei group acidophilus group acidophilus group acidophilus group plantarum acidophilus group casei group acidophilus group casei group acidophilus group casei group acidophilus group acidophilus group casei group acidophilus group acidophilus group acidophilus group acidophilus group casei group acidophilus group casei group acidophilus group

Letters in parentheses indicate the origin of the product: F, France; NL, Netherlands; S, Sweden.

Table 2 Grouping of Lactobacillus strains based on phenotypic properties and DNA base composition Lactobacillus BFE no.

Growth at 158C

Configuration of lactic acid

G 1 C content (mol.%)

Presumptive identification

649, 652, 654, 662, 663, 665, 682, 692, 693, 695, 696, 704, 718, 719, 727

2

DL

32.5–35.8

Lactobacillus acidophilus group

648, 659, 664, 675, 687, 697, 723, 728

1

L

45.0–47.0

Lactobacillus casei group

700

1

DL

43.0

Lactobacillus plantarum b

a b

( DL)a

All strains except BFE 697 produced L ( 1 ) lactic acid. Identification was based on sugar fermentation pattern.

of the lactobacilli of both the L. acidophilus and the L. casei group (Table 1) and confirmed the taxonomic position of strain BFE 700 classified as L. plantarum. Only one strain of the L. acidophilus group (BFE 649) did not show a high DNA homol-

ogy to all authentic strains included in this study and could not be identified to species level. Nine strains of the L. acidophilus group including L. acidophilus LA-H3 and LA-7 and the commercial starter culture showed a high DNA similarity with the type strain of

U. Schillinger / International Journal of Food Microbiology 47 (1999) 79 – 87

83

Table 3 DNA homology between strains of the L. acidophilus group % DNA homology with DNA from Strain BFE no.

L. acidophilus Ta DSM 20079

BFE 695

665 679 682 695 696 704 719 720 727 652 654 662 663 692 718 693 649

103 71 103 108 105 107

91 101

a T

102 100

L. acidophilus ATCC 43121

L. johnsonii Ta DSM 10533

95 30 38 45 33 76

42 16 34 37 44

29 31

24

108 79 82 22

12

L. crispatus Ta DSM 20584

28 46 34 44 25

28

108 64

L. johnsonii DSM 20553

74 68 79

103 84 109 100 101 85 15 33

L. gasseri Ta DSM 20077

L. amylovorus DSM 20531

L. gallinarum Ta DSM 10532

11 35

33

L. intestinalis Ta DSM 6629

20

45

80 42

20 0

43

Type strain.

L. acidophilus indicating a very close genetic relatedness with this species (Table 3). Some strains designated as L. acidophilus, however, were found to belong to the species L. johnsonii or L. crispatus. L. acidophilus LA-1, three strains (BFE 652, BFE 654, BFE 662) isolated from yoghurts containing Biogarde  cultures, strain BFE 718 from yoghurt T, BFE 692 from a French yoghurt (G) and L. acidophilus ATCC 43121 included as a reference strain showed a high degree of DNA similarity with the authentic strains of L. johnsonii (Tables 1 and 3). DNA of L. acidophilus Gilliland (L1 5 BFE 693) isolated from a probiotic drink was highly homologous to that of the type strain of L. crispatus (DSM 20584). DNA homology analysis showed that the eight strains (BFE 648, BFE 659, BFE 664, BFE 675, BFE 687, BFE 697, BFE 723, BFE 728) identified as members of the L. casei group belong to three different species. Four strains including L. casei Shirota belonged to the species L. paracasei and three strains including Lactobacillus GG were identified as L. rhamnosus (Table 4). The eighth isolate shared a high level of DNA homology with L. casei DSM 20011 which is the type strain of the species L. casei but was recently transferred to the new species L. zeae (Dicks et al., 1996).

3.2. Viable numbers of lactobacilli at the end of best before use period During isolation procedure some Lactobacillus strains such as L. plantarum were detected in low numbers in the yoghurts (Table 1) indicating a low intake of viable bacteria by the consumer. During refrigerated storage of the yoghurts viability loss of the cultures may occur resulting in even lower numbers. Therefore the viable counts of Lactobacillus present in the yoghurts were determined at the day of purchase and after storage till the end of the best before use period. The data presented in Table 5 give the range of viable numbers detected during enumerations of yoghurts bought in the years between 1995 and 1998. In mild yoghurts (A–J) numbers tended to vary greatly and in most cases counts declined at the end of the storage period indicating a low stability of the strains in dairy products. Only two out of the nine mild yoghurts investigated contained Lactobacillus viable numbers above 10 6 per gram at every random sample. In four out of eight yoghurts viable lactobacilli detected at the end of the storage period were below 10 4 per gram. The 12 probiotic yoghurts studied (K–Y) tended to contain higher viable Lactobacillus counts as

U. Schillinger / International Journal of Food Microbiology 47 (1999) 79 – 87

84

Table 4 DNA homology between strains of the L. casei group and L. plantarum % DNA homology with DNA from Strain BFE no. 648 659 664 675 687 697 723 728 700 a T

L. casei DSM 20011 T a

L. paracasei DSM 5622 T a

34 23

L. plantarum DSM 20174 T a

91 108 104 42

4 111 88 96 73 24 5

48 48 43 38 106 23

L. rhamnosus DSM 20021 T a

94

Type strain.

Table 5 Lactobacillus viable counts of mild and probiotic yoghurts sold on the German market Yoghurt code

Type of yoghurt

Lactobacillus viable counts (log CFU / g)a Day of purchase

End of best before use period

A B C D E F H I J

Mild Mild Mild Mild Mild Mild Mild Mild Mild

6.7–7.7 3.6–4.9 4.6–5.5 4.2–7.0 5.5–6.4 3.0–6.1 7.1–7.8 3.0–4.5 5.1–6.9

6.0–7.1 2.0–3.8 , 3.0 3.0–6.0 4.0–4.8 , 3.0 5.1–7.4 , 3.0 3.0–6.1

K L

Probiotic Probiotic

M N Q

Probiotic Probiotic Probiotic

R S T U W X Y

Probiotic Probiotic Probiotic Probiotic Probiotic Probiotic Probiotic

7.1–8.1 8.0 (L. rhamnosus) 6.3–6.7 (L. acidophilus) 7.4–8.4 5.4–6.1 6.8–8.2 (L. acidophilus) 6.2–7.8 (L. paracasei) 3.9–5.9 4.7–6.5 5.5–6.8 5.3–5.5 8.6–8.7 7.1–7.8 5.8–8.4 (L. acidophilus) 4.7–5.3 (L. casei)

8.0 8.0 (L. rhamnosus) 6.1–6.6 (L. acidophilus) 7.4–7.6 , 3.0 7.3–7.7 (L. acidophilus) 6.8–7.1 (L. paracasei) 3.3–4.4 3.8–5.2 4.1–4.4 b 3.8–5.5 8.5–8.7 6.4–7.2 5.1–8.2 (L. acidophilus) 5.6 (L. casei)

a

Yoghurts were bought at different occasions within the time period from 1995 to 1998. Enumerations were done for each yoghurt at least three times both at the day of purchase and after storage at 48C till the end of best before use period. b In this yoghurt drink the bifidobacteria were considered to be the probiotic organisms and the Lactobacillus strain was used as additional culture in the product.

U. Schillinger / International Journal of Food Microbiology 47 (1999) 79 – 87

compared to the classical mild yoghurts (Table 5). Very high Lactobacillus viable numbers between 10 7 and 10 8 per g were detected in six probiotic products. In some yoghurt samples viable counts of L. acidophilus ( 5 L. johnsonii) LA-1, L. casei ( 5 L. rhamnosus) GG, L. paracasei Actimel from yoghurt M and L. acidophilus LA-H3 even exceeded 10 8 per gram. The lactobacilli remained at this high level till the end of the best before use period indicating a high stability of the cultures in those dairy products during refrigerated storage. Most other probiotic yoghurts contained lactobacilli in the range between 10 5 and 10 7 per gram. Lactobacillus numbers below 10 5 per gram were rarely found at the day of purchase. At the end of the best before use period, however, a reduction in viable counts below 10 4 per gram was observed in some cases.

4. Discussion L. acidophilus is expected to be the main Lactobacillus species involved in the manufacture of mild and probiotic yoghurts. Our studies on Lactobacillus isolates from various dairy products using DNA–DNA hybridization, however, revealed that only one of the eight classical mild yoghurt products investigated contained a typical representative of L. acidophilus. On the other hand, the majority of the isolates from yoghurts claimed to be probiotic were identified as L. acidophilus. According to the labelling of the manufacturers, five out of eight mild yoghurts investigated in this study should contain Biogarde  or Bioghurt  cultures consisting of L. acidophilus and / or Bifidobacterium bifidum in addition to S. thermophilus (Tamime et al., 1995). The lactobacilli isolated from two of these products, however, showed the phenotypic characteristics of members of the L. casei group and were clearly identified as L. rhamnosus by DNA–DNA hybridizations. Obviously, Lactobacillus strains different from the Biogarde  or Bioghurt  cultures were used in the manufacture of these two yoghurts showing that correct labelling of the yoghurts was not used in all cases. Lactobacillus strains isolated from the other three products designated as containing Biogarde  cultures were phenotypically very similar with L. acidophilus. DNA

85

homology studies, however, revealed that the strain distributed under the name Biogarde  or Bioghurt  does not belong to the species L. acidophilus but to L. johnsonii. Some probiotic yoghurts also contained strains erroneously classified as L. acidophilus. DNA–DNA hybridization experiments revealed that L. acidophilus LA-1 from yoghurt K and L. acidophilus Gilliland ( 5 L1) from yoghurt P were L. johnsonii and L. crispatus, respectively. LA-1 is a strain with well documented probiotic properties such as the ability to inhibit the adherence of enteropathogens on human intestinal Caco-2 cells (Bernet-Camard et al., 1997) and with immune stimulatory properties (Schiffrin et al., 1995). L. acidophilus Gilliland is a human isolate studied by Gilliland and coworkers (Buck and Gilliland, 1994) which was selected for a probiotic product on the basis of a possible positive effect on cholesterol metabolism. L. acidophilus ATCC 43121 ( 5 RP32), another strain investigated by Gilliland and coworkers (Gilliland et al., 1985; Noh et al., 1997) to demonstrate a cholesterol lowering effect was included for comparison in our studies and was also shown not to belong to the species L. acidophilus as it showed a higher degree of DNA homology with the type strain of L. johnsonii. L. casei is another species increasingly being used in the manufacture of probiotic yoghurts. Six out of 13 probiotic products contained strains phenotypically classified as L. casei related. On the basis of DNA homology the majority of the strains including L. casei Shirota and L. casei Actimel were identified as members of the species L. paracasei according to current nomenclature. Dicks et al. (1996), however, proposed to reject the name L. paracasei, and to reclassify strains of this species as L. casei. This proposed change in nomenclature at the moment is still under discussion Lactobacillus GG, a well-known strain whose probiotic properties were extensively studied was called L. casei in most publications (Isolauri et al., 1991; Sarem-Damerdji et al., 1995). DNA homology analysis showed that this strain belongs to L. rhamnosus. This is in agreement with the results obtained by amplification of a 1.5 kb gene sequence of 16S ribosomal RNA (Saxelin, 1997). Strains differed in viability and ability to survive during refrigerated storage of the yoghurts. Lac-

86

U. Schillinger / International Journal of Food Microbiology 47 (1999) 79 – 87

tobacillus viable counts obtained from mild and probiotic yoghurts bought three or four times during the time period from 1995 to 1998 showed that some strains such as L. johnsonii LA-1 or Lactobacillus GG were present in high numbers even at the end of the best before use period and therefore survived well in dairy products. Nighswonger et al. (1996) and Saxelin (1996) also found a good survival of Lactobacillus GG in yoghurt-like products. In some products, however, Lactobacillus numbers varied greatly and declined markedly to low numbers at the end of the best before use period indicating a poor viability of the probiotic culture. Problems with the stability of strains of L. acidophilus and bifidobacteria in fermented milk products have been reported (Gilliland and Speck, 1977; Medina and Jordono, 1994; Shah et al., 1995). Amongst others viability of probiotic LAB is dependent on the strain of associative yoghurt organism (Nighswonger et al., 1996; Dave and Shah, 1997). In such yoghurts containing the probiotic strain as a supplement, the production of hydrogen peroxide by the yoghurt starter culture may adversely affect the viability of the added L. acidophilus (Gilliland and Speck, 1977; Hull et al., 1984). Most of the probiotic products investigated in this study, however, did not contain L. delbrueckii subsp. bulgaricus, whereas S. thermophilus was detected in high numbers in all yoghurts (data not shown). It seems reasonable to assume that adequate numbers of the probiotic bacteria need to be consumed to exert a health-promoting effect for the consumer and it has been suggested that to have any therapeutic effect, the minimal number of probiotic bacteria in a product should be above 10 5 or 10 6 per gram (Speck, 1978; Kim, 1988). Moreover, it is important that the lactobacilli remain viable during refrigerated storage of the product for a certain period as yoghurts may be consumed after storage in the refrigerator for several weeks. Our results indicate that not all products available on the market contain probiotic bacteria in high numbers and that not all of them show good survival characteristics. The majority of the probiotic yoghurts investigated in this study, however, contained the probiotic strain in numbers markedly higher than the suggested minimum level of 10 5 per gram even after storage till the end of best before use period.

Acknowledgements The author gratefully acknowledges Mrs. Heike ¨ Schafer for excellent technical assistance.

References Bernet-Camard, M.-F., Lievin, V., Brassart, D., Neeser, J.-R., Servin, A.L., Hudault, S., 1997. The human Lactobacillus acidophilus strain LA1 secretes a nonbacteriocin antibacterial substance(s) active in vitro and in vivo. Appl. Environ. Microbiol. 63, 2747–2753. Buck, L.M., Gilliland, S.E., 1994. Comparisons of freshly isolated strains of Lactobacillus acidophilus of human intestinal origin for ability to assimilate cholesterol during growth. J. Dairy Sci. 77, 2925–2933. Collins, M.D., Phillips, B.A., Zanoni, P., 1989. Deoxyribonucleic acid homology studies on Lactobacillus casei, Lactobacillus paracasei sp. nov., subsp. paracasei and subsp. tolerans, and Lactobacillus rhamnosus sp. nov., comb. nov. Int. J. Syst. Bacteriol. 39, 105–108. Dave, R.I., Shah, N.P., 1997. Viability of yoghurt and probiotic bacteria in yoghurts made from commercial starter cultures. Int. Dairy J. 7, 31–41. De Man, J.C., Rogosa, M., Sharpe, M.E., 1960. A medium for the cultivation of lactobacilli. J. Appl. Bacteriol. 23, 130–135. De Ley, J., Cattoir, H., Reynaerts, A., 1970. The quantitative measurement of DNA hybridization from renaturation rates. Eur. J. Biochem. 12, 133–142. Dicks, L.M.T., Du Plessis, E.M., Dellaglio, F., Lauer, E., 1996. Reclassification of Lactobacillus casei ATCC 393 and Lactobacillus rhamnosus ATCC 15820 as Lactobacillus zeae nom. rev., designation of ATCC 334 as the neotype of L. casei subsp. casei, and rejection of the name Lactobacillus paracasei. Int. J. Syst. Bacteriol. 46, 337–340. Fujisawa, T., Benno, Y., Yaeshima, T., Mitsuoka, T., 1992. Taxonomic study of the Lactobacillus acidophilus group with recognition of Lactobacillus gallinarum sp. nov. and Lactobacillus johnsonii sp. nov and synonymy of Lactobacillus acidophilus group A3 (Johnson et al., 1980) with the type strain of Lactobacillus amylovorus (Nakamura, 1981). Int. J. Syst. Bacteriol. 42, 487–491. Gilliland, S.E., Speck, M.L., 1977. Instability of Lactobacillus acidophilus in yogurt. J. Dairy Sci. 60, 1394–1398. Gilliland, S.E., Nelson, . C.R., Maxwell, C., 1985. Assimilation of cholesterol by Lactobacillus acidophilus. Appl. Environ. Microbiol. 49, 377–381. Huss, V., Festl, H., Schleifer, K.H., 1983. Studies on the spectrophotometric determination of the DNA hybridization from renaturation rates. System. Appl. Microbiol. 4, 184–192. Hull, R.R., Roberts, A.V., Mayes, J.J., 1984. Survival of Lactobacillus acidophilus in yogurt. Aust. J. Dairy Technol. 39, 164–166. Isolauri, E., Juntunen, M., Rautanen, T., Sillanaukee, P., Koivula,

U. Schillinger / International Journal of Food Microbiology 47 (1999) 79 – 87 T., 1991. A human Lactobacillus (Lactobacillus casei sp strain GG) promotes recovery from acute diarrhea in children. Pediatrics 88, 90–97. Kim, H.S., 1988. Characterization of lactobacilli and bifidobacteria as applied to dietary adjuncts. Cult. Dairy Prod. J. 23, 6–9. Marmur, J., 1961. A procedure for the isolation of deoxyribonucleic acid from microorganisms. J. Mol. Biol. 3, 208–218. Marmur, J., Doty, P., 1962. Determination of the base composition of deoxyribonucleic acid from its thermal denaturation temperature. J. Mol. Biol. 5, 109–118. Medina, L.M., Jordono, R., 1994. Survival of constitutive microflora in commercially fermented milk containing bifidobacteria during refrigerated storage. J. Food Prot. 56, 731–733. Mori, K., Yamazaki, K., Ishiyama, T., Katsumata, M., Kobayashi, K., Kawai, Y., Inoue, N., Shinano, H., 1997. Comparative sequence analyses of the genes coding for 16S rRNA of Lactobacillus casei-related taxa. Int. J. Syst. Bacteriol. 47, 54–57. Nighswonger, B.D., Brashears, M.M., Gilliland, S.E., 1996. Viability of Lactobacillus acidophilus and Lactobacillus casei in fermented milk products during refrigerated storage. J. Dairy Sci. 79, 212–219. Noh, D.O., Kim, S.H., Gilliland, S.E., 1997. Incorporation of cholesterol into the cellular membrane of Lactobacillus acidophilus ATCC 43121. J. Dairy Sci. 80, 3107–3113. Pot, B., Hertel, C., Ludwig, W., Descheemaeker, P., Kersters, K., Schleifer, K.H., 1993. Identification and classification of Lactobacillus acidophilus, L. gasseri and L. johnsonii strains by SDS PAGE and rRNA-targeted oligonucleotide probe hybridization. J. Gen. Microbiol. 139, 513–517.

87

Sarem-Damerdji, L., Sarem, F., Marchal, L., Nicolas, J.-P., 1995. In vitro colonization of human colon mucosa by exogenous Lactobacillus strains. FEMS Microbiol. 131, 133–137. Saxelin, M., 1996. Colonization of the human gastrointestinal tract by probiotic bacteria. Nutr. Today Suppl. 31, 5S–8S. Saxelin, M., 1997. Lactobacillus GG — a human probiotic strain with thorough clinical documentation. Food Rev. Int. 13, 293– 313. Schiffrin, E.J., Rochat, F., Link-Amster, H., Aeschlimann, J.L., Donnet-Hughes, A., 1995. Immunomodulation of human blood cells following the ingestion of lactic acid bacteria. J. Dairy Sci. 78, 491–497. ¨ Schillinger, U., Lucke, F.-K., 1987. Identification of lactobacilli from meat and meat products. Food Microbiol. 4, 199–208. ¨ Schuler-Malyoth, R., Ruppert, A., Muller, F., 1968. Die Mikroorganismen der Bifidusgruppe (syn. Lactobacillus bifidus). 2. Mitteilung: Die Technologie der Bifiduskultur im milchverarbeitenden Betrieb. Milchwissenschaft 23, 554–558. Shah, N.P., Lankaputhra, W.E.V., Britz, M.L., Kyle, W.S.A., 1995. Survival of Lactobacillus acidophilus and Bifidobacterium bifidum in commercial yoghurt during refrigerated storage. Int. Dairy J. 5, 515–521. Speck, M.L., 1978. Acidophilus food products. Dev. Ind. Microbiol. 19, 95–101. Stackebrandt, E., Kandler, O., 1979. Taxonomy of the genus Cellulomonas, based on phenotypic characters and deoxyribonucleic acid–deoxyribonucleic acid homology and proposal of seven neotype strains. Int. J. Syst. Bacteriol. 29, 273–282. Tamime, A.Y., Marshall, V.M.E., Robinson, R.K., 1995. Microbiological and technological aspects of milks fermented by bifidobacteria. J. Dairy Res. 62, 151–187.