Isolation, identification and characterisation of the dominant microorganisms of kule naoto: the Maasai traditional fermented milk in Kenya

International Journal of Food Microbiology 94 (2004) 269 – 278 www.elsevier.com/locate/ijfoodmicro

Isolation, identification and characterisation of the dominant microorganisms of kule naoto: the Maasai traditional fermented milk in Kenya Julius Maina Mathara a,*, Ulrich Schillinger a, Phillip Museve Kutima b, Samuel K. Mbugua c, Wilhelm H. Holzapfel a b

a Federal Research Centre for Nutrition, Institute of Hygiene and Toxicology, Haid-und-Neu-Str. 9, D-76131 Karlsruhe, Germany Department of Food Science and Technology, Jomo Kenyatta University of Agriculture and Technology, P.O. Box 62000 Nairobi, Kenya c Department of Food Technology and Nutrition, University of Nairobi, P.O. Box 29053 Nairobi, Kenya

Received 28 April 2003; received in revised form 2 November 2003; accepted 20 January 2004

Abstract From 22 samples of kule naoto, the traditional fermented milk products of the Maasai in Kenya, 300 lactic acid bacterial strains were isolated and phenotypically characterised by their ability to ferment different carbohydrates and by additional biochemical tests. Lactic acid bacteria (LAB), especially the genus Lactobacillus, followed by Enterococcus, Lactococcus and Leuconostoc, dominated the microflora of these samples. The major Lactobacillus species was Lactobacillus plantarum (60%), with a lower frequency of isolation for Lactobacillus fermentum, Lactobacillus paracasei and Lactobacillus acidophilus. Most strains produced enzymes such as h-galactosidase and peptidases, which are of relevance to cultured dairy product processing, and exhibited similar patterns of enzymatic activity between species. Enterobacteriaceae could not be detected in 15 out of 22 samples (detection level 102/ml). Conversely, yeasts (detection level 101/ml) were detected in those samples in which Enterobacteriaceae were not found. The pH values of all these samples were < 4.5. D 2004 Elsevier B.V. All rights reserved. Keywords: Maasai; Lactic acid bacteria; Traditional fermented milk; Kule naoto

1. Introduction The Maasai community is a Nilo-hamitic tribe living a nomadic life in the East Africa Rift-Valley of southern Kenya and northern Tanzania. The community has a conserved rich traditional heritage that revolves around cattle. They live in small settlements * Corresponding author. Tel.: +49-721-6625-465; fax: +49-7216625-453. E-mail address: [email protected] (J.M. Mathara). 0168-1605/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.ijfoodmicro.2004.01.008

of 8 – 15 huts per kraal, known as Manyatta. Kule naoto, the traditional lactic fermented milk product, is the major daily diet of the Maasai community, which rarely consume fruits or grains. On average, 2 –3 l of the fermented product is consumed per person per day. The product is produced from unpasteurised whole milk from the zebu breed of cows using centuries old practices. The product whose rich taste and consistency is similar to unsweetened commercial yoghurt or fresh cheese, is spontaneously fermented for at least 5 days, though a longer period is usually

270

J.M. Mathara et al. / International Journal of Food Microbiology 94 (2004) 269–278

preferred (Mathara et al., 1996; Mathara, 1999). Whole fresh raw milk is filled into a custom-made specially treated gourd made from the hollowed out dried fruit of the plant Lagenaria siceraria. The dried calabash, used as fermentation gourd, is gently rubbed with a burning end of a chopped stick from the tree Olea africana locally known as Enkidogoe, or also from other trees allowing charcoal to break inside (Mathara, 1999). This procedure is repeated at least three times. The gourd is filled with milk and then closed by a special cap obtained from the same gourd during its preparation. After fermentation, the product is gently shaken before consumption. These traditional lactic fermented milk products are popular among the Maasai community and people living in eastern Africa. Kenyans, in general, prefer it due to its excellent natural taste and aroma, among other functional benefits. Also, because of their cultural associations, the people believe in a therapeutic value towards curing or protection from ailments such as diarrhoea and constipation (Mathara, 1999). So far, there is limited scientific information to justify these claims. According to Holzapfel (1997, 2002), there is an increasing need to select microbial strains with functional properties for commercial production and for improvement of quality and safety of existing traditional fermented food products. This study was conducted against the background of the key role kule naoto plays in the nutrition of the Maasai people and the fact that large numbers of viable lactic acid bacteria (LAB) are consumed daily by the Maasai. The objectives of this study were to determine and study the predominant microbial groups in kule naoto and to identify and characterise the dominant (LAB) from representative product samples obtained from the Maasai community in Kenya.

2. Material and methods 2.1. Collection of samples Twenty-two traditional fermented milk samples were collected from the Maasai manyattas in the rural plains of Maasailand in Kenya. Samples were collected, both in their original fermentation vessels and in sterile sample bottles. Two sampling regimes were used, 1 during the rainy season (April and May) 2001,

with 13 samples and the second during the dry period (August and September) 2002, with 9 samples The pH of the samples was determined at the sampling site using a calibrated portable pH-meter (Messkoffer Qph 70, WWR-International, Germany). The samples were kept at 4 –8 jC and transported in special cool boxes within hours to the Federal Research Centre for Nutrition, Karlsruhe, Germany for analyses. 2.2. Microbial enumeration and isolation Ten grams of milk sample was transferred aseptically into 90-ml Ringer’s solution and mixed thoroughly. Serial dilutions (10 1 – 10 8) were made for each sample and 0.1 ml of the appropriate dilution spread plated on universal and selective media. Plate count agar (Merck, Darmstadt, Germany) was used for enumeration of aerobic mesophilic counts as described in the International Dairy Federation reference method (IDF 100B: 1991). MRS agar pH 6.4 (Merck) was used for enumeration of total LAB, Rogosa agar (Merck) for enumeration of lactobacilli and M17 (Merck) for enumeration of lactococci. Kanamycin aesculine azide agar (Merck) (KAA) was used for enumeration of enterococci and Violet Red Bile Dextrose Agar (VRBD) (Merck) for enumeration of Enterobacteriaceae. Acidified Potato Dextrose Agar (PDA) (Merck), pH 3.7, was used for enumeration of yeasts. MRS and M17 plates were incubated anaerobically at 30 jC using anaerobic jars together with the Merck Anaerocult A GasPak anaerobic system (Merck 1.13829). Representative colonies were isolated from MRS, M17, PDA and Kanamycin, aesculine azide agar and purified by streak plating using the same medium. Gram-positive, catalase-negative bacteria were purified by re-streaking on MRS agar pH 6.4 (Merck) and M17 (Merck), and were resuspended and preserved in the same medium containing 15% glycerol at 18 jC. Representative yeast colonies on PDA were examined by phase contrast microscopy, purified by successive streaking on PDA and stored on slants at 2 jC. 2.3. Phenotypic characterisation Initial characterisation of Gram-positive, catalasenegative bacterial isolates was by microscopy (cell morphology and arrangements) (Gerhardt et al.,

J.M. Mathara et al. / International Journal of Food Microbiology 94 (2004) 269–278

1981). Growth at 10, 15 and 45 jC in MRS broth was evaluated visually after 24, 48 and 72 h of incubation. Tests for presence of catalase, production of ammonia from arginine and type of fermentation were carried out as described by Harrigan and McCance (1976). Salt tolerance was determined using MRS broth containing 6.5% (w/v) NaCl with incubation for 72 h at 37 jC. Growth at pH 3.9 and 9.6 was determined in MRS with the pH adjusted by using HCl and NaOH, respectively. Dextran production was determined by roppy colony formation on MRS agar pH 6.5 supplemented with 10% sucrose (Merck, 1.05323). Approaches followed in the phenotypic differentiation were according to the information supplied by Wood and Holzapfel (1995) and Stiles and Holzapfel (1997). 2.4. Lactic acid configuration The type and amount of D( ) and L(+) isomers of lactic acid produced from glucose was assayed in modified MRS broth without beef extract and acetate and using D- and L-lactate dehydrogenase from a commercial kit (Hoffman La Roche Diagnostic Mannheim, Germany containing D-lactate dehydrogenase code no. 1585436, L-lactate dehydrogenase code no.

271

127221, glutamate pyruvate transaminase (GPT) code no. 737127 and nicotinamide adenine dinucleotide (NAD) code no. 127990 following instructions of the supplier). The modified MRS medium had the following composition in g/l: casein peptone 1.0, yeast extract 4.0, glucose 20, di-potassium hydrogen phosphate 2.0, Tween 80 1.0, di-ammonium hydrogen citrate 2.0, magnesium sulphate 0.2 and manganese sulphate 0.04. 2.5. Identification of meso-diaminopimelic acid (meso-DAP) Presence of meso-DAP in the bacterial cell wall was determined according to a modified method described by Schillinger and Lu¨cke (1987). Briefly, isolates were grown in 1.0 ml MRS broth (Merck), for 48 h, harvested and washed with 1 ml of distilled water. The sediment was re-suspended in 1.0 ml 6 N HCl and transferred to a capped test tube. The cells were hydrolysed overnight at 100 jC. The content of the tube was dried in an air stream. The sediment was resuspended in 1.0 ml of distilled water and dried in the same manner. Finally, the sediment was re-suspended in 0.1 ml distilled water and samples were spotted on

Table 1 Enzymatic profiles of lactic acid bacterial strains isolated from kule naoto, the Maasai traditional fermented milk product of Kenya Enzyme/genera

Leuconostoc (n = 3)

Lb. casei group (n = 5)

Lb. rhamnosus (n = 6)

Lb. acidophilus group (n = 10)

Lb. plantarum (n = 10)

Alkaline phosphatase Esterase (C4) Esterase lipase (C8) Lipase (C14) Leucine arylamidase Valine arylamidase Cystine arylamidase Trypsin a-Chymotrypsin Acid phosphatase Naphthol-AS-BIphosphohydrolase a-Galactosidase h-Galactosidase h-Glucuronidase a-Glucosidase h-Glucosidase N-Acetyl-h-glucosaminidase a-Mannosidase a-Fucosidase

5 23 3.3 3.3 5 5 3.3 0 5 10 5 8.3 21.7 3.3 z 40 3.3 0 0 0

2 18 8 10 z 40 z 40 10 5 5 30 z 40 14 0 0 5 z 40 5 5 0

20 30 30 5 z 40 z 40 30 5 10 30 20 20 z 40 5 20 z 40 0 0 z 40

0 20 8.5 6.5 z 40 7.5 12 0 0 19 19 4.5 16.5 0 7.5 z 40 12.5 0 0

0 10 10 0 z 40 z 40 20 0 0 5 z 40 0 z 40 0 10 z 40 30 0 0

Results (nmol of chromophore released) are averaged and n is the number of strains tested. Enzyme activity is shown as nM of chromophore released after 6 h of incubation at 37 jC.

272

J.M. Mathara et al. / International Journal of Food Microbiology 94 (2004) 269–278

thin layer plates [DC Plastic Folien: Cellulose (20  20) Merck 5577]. Ascending one-dimensional chromatography was done in a modified solvent solution containing methanol/pyridine/10 N HCl/water in (32:4:1:7) v/v/v/v. The solvent was prepared one hour before use. After drying, the chromatogram was developed with ninhydrin solution and placed for 5 min in a 100 jC oven. Spots representing meso-diaminopimelic acid appeared dark green to grey and turned yellow within 24 h in the dark. 2.6. Carbohydrate fermentation assays The first determination of the carbohydrate fermentation profile of all the strains was done in MRS fermentation broth (Merck) without glucose and containing chlorophenol red (0.04 g/l) as pH indicator. The individual sugars were prepared as 2.5% (m/v) solutions and filter sterilised using a 0.5-Am filter. A micro-titre plate was used whereby 25 Al of sugar solution was added to 100 Al of basal medium containing the washed cell suspension. Strains were grown overnight at 30 jC in MRS and the cell pellet obtained by centrifugation at 10,000  g for 5 min. The pellet was washed twice in sterilised double distilled water before re-suspending it in basal medium. Subsequently, API 50CH galleries and API CHL medium (bioMerieux Marcy-I’Etoile, France) were used according to the manufacturer’s instructions for determining the sugar fermentation spectrum of some selected strains. APILAB PLUS V3.2.2 software database (bioMerieux) was used for interpretation of results. 2.7. Enzymatic profiles of predominant organisms The enzymatic profiles of the predominant LAB were assayed using API Zym galleries (bioMerieux) according to the manufacturer’s instructions. The activities of 19 enzymes were tested (see Table 1).

3. Results 3.1. pH of the fermented milk samples The average pH of the fermented milk samples was 4.4 with pH values ranging from 4.17 to 5.19 with

only four samples in which the pH exceeding 4.5 (Table 2). 3.2. Enumeration of microorganisms Table 2 summarises the microbial counts of the fermented milk samples obtained from 22 sampling regions (manyattas), and Fig. 1 shows the distribution of LAB and yeasts according to their levels of viable counts. LAB were the dominating organisms in all the samples with average values of 8.0 log10 cfu ml 1 on MRS and 7.9 log10 cfu ml 1 on M17 agar, with corresponding average mesophilic bacterial counts of 8.1 log10 cfu ml 1. In general, lactobacilli and lactococci were the predominant LAB and were mostly detected in a range from 107 to 109 ml 1, whereas enterococci were detected at a lower range with numbers >104/ml in 15 samples (Table 2). The mean Enteroccocus count was 5.5 log10 cfu ml 1 with a range of 3.3– 9.0 log10 ml 1, an indication that this group of LAB may contribute to some extent to the fermentation of these products. Only in seven samples Enterobacteriaceae were detected with counts ranging from 4.9 to 9.4 log10 cfu ml 1 (Table 2). In none of these seven samples, yeasts were detected. One sample had Enterobacteriaceae counts of 5.84 log10 cfu ml 1 and a mould count of 4.26 log10 cfu ml 1, but with no yeasts detected. The pH of these samples with detectable numbers of Enterobacteriaceae was z 4.5 with the exception of one sample. The remainder of the samples (15) had yeast counts with values ranging from 4.24 to 7.44 log10 cfu ml 1, with average of 6 log10 cfu ml 1. The pH of these samples was V 4.5. The observation that no Enterobacteriaceae were detected in samples with detectable numbers of yeasts is remarkable. This observation suggests a possible interaction between bacteria and yeasts involved in the Maasai milk fermentation. 3.3. Identification of lactic acid bacteria to genus level In total, 520 strains were isolated from the Maasai traditional fermented milk products and identified to genus level on the basis of cell morphology, gas production from glucose, growth behaviour at 10 jC, 45 jC and in presence of 6.5% NaCl and at pH 9.6 according to Wood and Holzapfel (1995) and

J.M. Mathara et al. / International Journal of Food Microbiology 94 (2004) 269–278

273

Table 2 Microbial numbers distribution (log10 cfu/ml) determined in Maasai traditional fermented milk (kule naoto) samples from different villages (Manyatta) in Kenya Sampling region (Manyatta)

pH

Mesophilic bacterial count

LAB count

Lactobacillus

Lactococcus

Enterobacteriaceaea

Yeastsb

Enterococcus

Lb. plantarumc (detected at levels z 106)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 Mean Standard deviation

4.4 4.4 5.0 4.8 4.7 4.5 4.3 4.2 4.3 4.3 4.4 4.4 4.4 4.3 4.3 5.2 4.3 4.3 4.2 4.2 4.3 4.5 4.4 0.3

8.4 7.4 9.2 9.2 8.7 8.4 8.5 6.5 7.5 6.1 8.6 8.0 8.3 8.3 7.8 9.0 7.9 7.9 7.8 7.8 8.1 8.6 8.1 0.8

8.2 7.5 8.5 9.2 8.8 8.0 8.2 6.6 7.4 6.6 8.3 7.3 8.0 7.9 8.0 8.9 8.1 8.2 7.9 7.6 8.0 8.5 8.0 0.6

8.3 5.1 8.5 9.0 8.5 6.6 7.3 6.5 5.5 5.5 7.6 5.4 7.0 7.5 7.7 8.7 8.9 7.8 7.8 7.7 8.0 7.8 7.4 1.2

6.2 7.3 9.4 9.0 8.5 8.7 7.2 7.4 6.2 7.3 8.2 7.2 8.3 8.2 7.7 9.4 7.7 7.7 7.7 7.7 7.4 8.5 7.9 0.9

< 2.0 < 2.0 8.4 8.2 6.8 6.4 < 2.0 < 2.0 < 2.0 < 2.0 < 2.0 < 2.0 < 2.0 < 2.0 < 2.0 9.4 < 2.0 4.9 < 2.0 < 2.0 < 2.0 5.8 7.1 1.6

6.6 4.2 < 1.0 < 1.0 < 1.0 < 1.0 6.5 6.1 5.2 5.4 6.3 5.0 6.2 5.3 7.4 < 1.0 6.9 < 1.0 6.3 6.5 6.2 < 1.0 5.9 0.9

5.6 3.6 9.0 9.0 4.5 6.2 3.9 4.6 3.6 4.2 3.7 3.3 3.4 5.7 5.8 7.4 7.2 6.4 3.7 6.0 7.4 7.2 5.5 1.8

+

+ + + + + + + + + + + + + + +

Enterobacteriaceae were only detected in samples with pH z 4.5 with exception of sample from region 18. a Detection level of 100 cfu/ml and 10 cfu/ml were used for detection of Enterobacteriaceae and yeasts, respectively. b Only moulds were detected in samples from the Manyattas 18 and 22, compromising a log10 6.5 and 4.3, respectively. c Positive sign indicates positive detection of Lb. plantarum strains at detection levels of z 106.

Stiles and Holzapfel (1997). Sampling was conducted in two seasons, during the rainy season April – May 2001, 13 samples, and during the dry season August – September 2002, 9 samples. All the LAB isolates were Gram-positive and catalase – negative and 55% were found to belong to the genus Lactobacillus, 14% to Lactococcus, 25% to Enterococcus and 6% to Leuconostoc. Strains of Lactobacillus plantarum constituted 60% of the lactobacilli. 3.4. Identification of the dominant lactic acid bacteria to species level In 20 samples, the lactobacilli were found to predominate the LAB. One hundred and thirty homofermentative Lactobacillus strains were identified as

plantarum, as shown in Fig. 2 and Table 3. They all produced DL-lactic acid, contained meso-DAP in the cell wall, and all could grow in MRS at pH 3.9 and in the presence of 6.5% NaCl. Lb. plantarum strains were the dominant lactobacilli in all the samples except those in which yeasts were not detected (Fig. 2 and Table 2). Lb. plantarum strains were obtained from the samples where Enterobacteriaceae were not detected with the exception of sample number 18. All Lb. plantarum strains were able to ferment ribose, galactose, fructose, mannose, mannitol, N-acetyl-glucosamine, amygdaline, esculin, cellobiose, maltose, melibiose, sucrose, trehalose, melezitose, raffinose and gentiobiose. Other sugars utilised by some Lb. plantarum strains were: glycerol (25%), L-arabinose (70%), rhamnose (5%), sorbitol (65%), methyl-D-man-

274

J.M. Mathara et al. / International Journal of Food Microbiology 94 (2004) 269–278

Fig. 1. Microbial distribution in 22 Maasai traditional fermented milk (kule naoto) samples from different villages (Manyatta) in Kenya.

noside (50%), methyl-D-glucoside (5%), turanose (70%), D-arabitol (90%) and gluconate (80%). Erythritol, D-arabinose, D-xylose, L-xylose, adonitol, hmethyl-xyloside, dulcitol, inositol, inulin, amidon, glycogen, xylitol, D-lyxose, D-tagatose, D-fucose, Darabitol, 2-ceto-gluconate and 5-ceto-gluconate were not fermented. Six Lactobacillus trains were identified as Lactobacillus rhamnosus. They all produced L(+) lactic acid, were DAP negative and were able to ferment rhamnose. Ten Lactobacillus strains were identified as Lactobacillus acidophilus. They all produced DL-lactic acid, they were DAP negative, grew well at 45 jC and were arginine negative. Tests using API 50CHL sugar profiles indicated with >90% identification probability of these strains to be as Lb. acidophilus. Fourteen Lactobacillus strains were as

Lactobacillus paracasei collectively grouped here as Lb. casei group. They produced exclusively L(+)-lactic acid and had no DAP in the cell wall. Seventy-eight heterofermentative Lactobacillus strains were identified as Lactobacillus fermentum on the basis of the sugar fermentation pattern and biochemical properties. Among the coccoid LAB, over 100 isolates were classified as Enterococcus. Sixty of which were identified to species level and were found to be Enterococcus faecium. They fermented arabinose, grew well at 45 jC and in MRS adjusted to pH 9.6 and in MRS broth containing 6.5% NaCl. Sixty coccoid LAB strains belonged to Lactococcus and 35 of them were identified as Lactococcus lactis. Six Leuconostoc strains were identified as Leuconostoc mesenteroides subsp. dextranicum. They pro-

Fig. 2. Distribution of 238 Lactobacillus strains isolated from the Maasai traditional fermented milk products in Kenya. For sampling region (Manyatta), see text.

J.M. Mathara et al. / International Journal of Food Microbiology 94 (2004) 269–278 Table 3 Identity and distribution of 339 lactic acid bacteria strains isolated from Maasai traditional fermented milk in Kenya Species

Number of strains identified

Level of isolation (per ml)

Lactobacillus plantarum Lactobacillus fermentum Lactobacillus casei Lactobacillus rhamnosus Lactobacillus acidophilus Leuconostoc mesenteroides Lactococcus lactis Enterococcus faecium

130 78 14 6 10 6 35 60

107 – 108 107 – 108 107 – 108 107 – 108 106 – 107 106 – 107 107 – 108 106 – 107

duced D( )-lactic acid and produced dextran from glucose. One strain of Leuconostoc did not produce dextran from sucrose and could be identified as either Weissella paramesenteroides or Leuconostoc mesenteroides subsp. mesenteroides. 3.5. Enzymatic profiles of the lactic acid bacteria isolates The activity of the 19 enzymes of technological importance were tested. Results indicate a high hgalactosidase activity among the Lb. plantarum and Lb. rhamnosus strains. But none for the Lb. paracasei strains Relatively low h-galactosidase activity was observed among the Leuconostoc strains. Lb. rhamnosus strains had high esterase and a-fucosidase activity compared to other groups of strains tested. Leuconostoc strains had high a-glucosidase activity compared to other groups. The Lb. casei group, Lb. rhamnosus and Lb. plantarum strains had high leucin and valine arylamidase activities.

4. Discussion With pH values ranging from 4.17 to 5.16, the acidity of Maasai traditional fermented milk products is comparable with that of most fermented dairy products. However, those with pH values >4.5 contained moderate to high numbers of Enterobacteriaceae and therefore do not appear to be sufficiently safe. An example is the sample from region 16 with a pH value of 5.2 and an Enterobacteriaceae mean count of 9.4 log10 cfu/ml. The lactic acid, together

275

with other metabolites, and the strong competitive effects of the LAB population, may be mainly responsible for the extended shelf life of up to 3 months of fermented products with sufficiently low pH. The LAB dominated the microbial population and viable counts of 6.6– 9.15 log10 cfu ml 1, with mean vales of 8.0 log10 cfu/ml, were recorded. The counts compared with findings of similar studies on fermented milks within the East African region. According to Isono et al. (1994), counts of LAB and lactobacilli reach 108 –109/g and 107 – 109/g, respectively. Miyamoto et al. (1986) and Mathara (1999) reported similar range of bacterial counts. In this study, the dominant Lactobacillus strains apparently contributed mainly to the technological quality attributes of the fermented product. Also, Lb. plantarum strains clearly seem to dominate lactobacilli in Maasai fermented milk. This observation may be related to the fact that the fermentation of the milk is exclusively carried out in a plant gourd previously prepared by treatment inside with splints of pre-heated smoking wood (Mathara, 1999) and to the adaptation of the particular Lb. plantarum strains to milk. According to Stiles and Holzapfel (1997), and Wood and Holzapfel (1995), Lb. plantarum strains are known to be commonly associated with plant based food fermentations. From cultured milk in Cameroon, Jivoua and Milliere (1990) identified 47 out of 426 isolates as Lb. plantarum. From a total of 26 Lactobacillus isolates from fermented milk in northern Tanzania, Isono et al. (1994) identified only four strains as Lb. plantarum. Out of 21 isolates from naturally fermented milk in Zimbabwe (Matukumira, 1996), only 3 strains were identified as Lb. plantarum, whilst only 3 out of 336 bacteria isolates from South African traditional fermented milks were identified as Lb. plantarum (Beukes et al., 2001). Medina et al. (2001) identified 92% of all lactobacilli in ewe’s milk and cheese from Northern Argentina as Lb. plantarum. Olasupo et al. (2001) reported that, all the Lactobacillus isolates from four different fermented products in Nigeria, including the fermented milk products wara and nono, were found to belong to Lb. fermentum. This is the first report on the occurrence of high levels of Lb. plantarum strains associated with Maasai traditional fermented milk products in Kenya. All the Lb. plantarum strains tested were able to ferment raffinose. This property indicates good bacterial tech-

276

J.M. Mathara et al. / International Journal of Food Microbiology 94 (2004) 269–278

nological potential associated with consumption of the fermented milk products together with legume based products like beans high in flatulence causing sugars. It also shows potential for use as a starter culture in development of soy based fermented milk products (Xanthopoulos et al., 2000). Yeasts appear to be commonly associated with traditional fermented dairy products and have been reported in several studies (Isono et al., 1994; Gadaga et al., 2001; Beukes et al., 2001). Isono et al. (1994) reported occurrence of yeasts in seven of 10 samples of traditional fermented milk in northern Tanzania with the mean counts ranging from 6.0 to 8.0 log10 cfu ml 1. In the current study, yeast counts ranged from 4.3 to 7.4 log10 cfu ml 1. However, a definite correlation was observed in all the samples with regard to the occurrence of yeasts and Enterobacteriaceae. No Enterobacterieaceae were detected in any of samples where yeasts were detected. The pH values of these samples were less or equal to 4.5. However, one sample (pH 4.5) had Enterobacteriaceae alongside with moulds but no yeasts were detected. This observation indicates a possible interaction of yeasts and the bacterial flora in the fermentation of fermented Maasai milk and may suggest a relation to the pH. Enterobacteriaceae, especially coliforms, are associated with poor hygiene and their occurrence in the product may indicate a potential health risk (Beukes et al., 2001). However, there are no reported cases of diarrhoea associated with the consumption of Maasai traditional fermented milk products (Mathara et al., 1996; Mathara, 1999; Nakamura et al., 1999). This is the first report indicating a possible interaction of Enterobacteriaceae and yeasts in the Maasai traditional fermented milk products. In this report, practically no Lb. plantarum strains were detected in those samples where Enterobacteriaceae were found. Further investigation is necessary to establish the extent and possible mechanisms of these interactions which most probably involve the LAB, and their particular domination relative to metabolic activity and acid production. In addition to Lb. plantarum, Lb. acidophilus and Lb. rhamnosus were also found among the lactobacilli in Maasai traditional fermented milk products. Strains of Lb. acidophilus have been reported only for a few other traditional fermented milk products. They comprised 8.5% of the bacterial flora isolated from the

traditional fermented milk product rob from Sudan (Abdelgadir et al., 2001). Matukumira (1996) isolated Lb. acidophilus from amasi, a Zimbabwean naturally fermented milk product. Eight percent of the Lactobacillus isolates in ewe’s milk and cheese from Northern Argentina (Medina et al., 2001) were identified as Lb. acidophilus. This is the first report on the occurrence of Lb. rhamnosus and Lb. acidophilus strains (with isolation range of 107 – 108/ml) in Maasai traditional fermented milk. The isolated Lb. rhamnosus strains are well adapted to the milk as indicated by the high h-galactosidase activity. According to Isono et al. (1994), Weissella confusa (Lactobacillus confusus) was the dominant LAB species isolated from the Maasai fermented milk products in northern Tanzania. Out of 32 Lactobacillus isolates, 26 were identified as W. confusa and, of the remaining 6, 1 as W. viridescens, 1 strain as Lb. brevis and 4 strains as Lb. plantarum. In our study, 78 strains out of 130 heterofermentative Lactobacillus strains were identified as Lb. fermentum. Most of the other heterofermentative Lactobacillus strains may be Weissella species according to preliminary data from molecular confirmatory tests. Lb. fermentum and species of Weissella are occasionally found in raw milk, but their technological role in fermentation of milk has not been widely reported. Only 6% of the isolated strains were identified as Leuconostoc species. Both dextran forming and nondextran forming strains were identified. Leuconostoc strains may contribute to the development of flavour quality attributes of fermented Maasai milk. The low percentage of Leuconostoc strains isolated from the fermented milk samples could partly be explained by their complex nutritional requirements (Medina et al., 2001), but probably also by their lower adaptation to milk. Leuconostoc species generally show a weak competitive ability during fermentation of milk (Wood and Holzapfel, 1995). The absence of proteases (trypsin and chymotrypsin), the high activities of peptidases (leucine, valine and cystine-aminopeptidase) and the low esteraselipase (C4 and C8) activities among the strains tested appear desirable traits for flavour and texture development in milk fermentation. Starters with low proteinase and strong peptidase activities are useful in reducing bitterness and improving body and texture defects (Arora et al., 1990) The high h-galactosidase

J.M. Mathara et al. / International Journal of Food Microbiology 94 (2004) 269–278

and h-glucosidase activities among the Lb. plantarum and Lb. rhamnosus strains suggest their general ability to ferment lactose. These strains are probably mainly responsible for the characteristic flavour, body and taste of Maasai traditional fermented milk products. All strains showed variable activities of lipase and esterase enzymes. Such strains could be involved in removal of free fatty acids due to their high esterase and lipase activities. No strain tested for its enzymatic profiles from wara and nono (Olasupo et al., 2001) showed any h-glucosidase or lipase activities. Similar enzymatic profiles among LAB isolated from soft variety Chhurpi, a traditional cheese typical of Sikkim Himalayas, was reported by Tamang et al. (2000).

277

desirable technological properties, show potential for use in small scale milk fermentation.

Acknowledgements The authors wish to thank the German Academic Exchange Services (DAAD) for financial support of this study. Thanks also to all the Maasai families in Kenya for providing the research samples. The technical staff at the Institute of Hygiene and Toxicology, Federal Research Centre for Nutrition, Karlsruhe, Germany is also gratefully acknowledged for their concerted effort and support during the experimental work.

5. Conclusions In this study on traditional Maasai fermented milk products, the domination of Lb. plantarum and Lb. fermentum among the associated LAB was established. Other strains occurring in relatively high numbers were identified as representatives of Lactococcus lactis, E. faecium, Leuc. mesenteroides, Lb. paracasei, Lb. rhamnosus and Lb. acidophilus. Data from this study suggest a possible interaction between Enterobacteriaceae; yeasts and the pH during the fermentation of Maasai milk. Enterobacteriaceae were not detected simultaneously with yeasts. This may reflect the inhibitory effect of reduced pH values (to which yeast are more resistance) and/or a probable antagonistic effect of yeasts and LAB against Enterobacteriaceae, thereby contributing to the quality and safety of the fermented milk products. The occurrence of a wide diversity of LAB in Maasai traditional fermented milk could in part be responsible for the desirable quality attributes associated with the products. Most representatives are well adapted to the milk environment as is evidenced by their enzymatic profiles. According to Rolle and Satin (2002), traditional small scale fermentation technologies offer considerable potential for stimulating development in the food industry of developing countries in light of their low cost, scalability, minimal energy and the infrastructure requirements and the wide consumer acceptance of such traditional fermented products. LAB strains associated with Maasai traditional fermented milk products, and with

References Abdelgadir, W.S., Hamad, S.H., Mo¨ller, P.L., Jakobsen, M., 2001. Characteristics of the dominant microbiota of Sudanese fermented milk Rob. International Dairy Journal 11, 63 – 70. Arora, G., Lee, B.H., Lamoureux, M., 1990. Characterisation of enzyme profiles of Lactobacillus casei species by rapid API ZYM system. Journal of Dairy Science 73, 264 – 273. Beukes, E.M., Bester, B.H., Mostert, J.F., 2001. The microbiology of South African traditional fermented milks. International Journal of Food Microbiology 63, 189 – 197. Gadaga, T.H., Mutukumira, A.N., Narvhus, J.A., 2001. The growth and interaction of yeasts and lactic acid bacteria isolated from Zimbabwean naturally fermented milk in UHT milk. International Journal of Food Microbiology 68, 21 – 32. Gerhardt, P., Murray, R.G.E., Costilow, R.N., Nester, E.W., Wood, W.A., Krig, N.R., Phillips, G.B., 1981. Manual of Methods for General Bacteriology. American Society for Microbiology, Washington, DC. Harrigan, W.F., McCance, M.E., 1976. Laboratory methods in Food and Dairy Microbiology. Academic Press, London. Holzapfel, W., 1997. Use of starter cultures in fermentation on a household scale. Food Control 8 (5/6), 158 – 241. Holzapfel, W.H., 2002. Appropriate starter culture technologies for small-scale fermentation in developing countries. International Journal of Food Microbiology 75 (3), 197 – 212. International Dairy Federation, 1991. Milk and Milk Products-Enumeration of Microorganisms. Colony Count at 30 jC (IDF Standard 100b: 1991). IDF, Brussels. Isono, Y., Shingu, I., Shimizu, S., 1994. Identification and characteristics of lactic acid bacteria isolated from Maasai fermented milk in northern Tanzania. Bioscience, Biotechnology, and Biochemistry 58 (4), 660 – 664. Jivoua, C., Milliere, J.B., 1990. Lactic flora and enterococci in cultured milk (Pindidam) manufacture in Adamaoua, Cameroon. Lait 70, 475 – 486 (abstract).

278

J.M. Mathara et al. / International Journal of Food Microbiology 94 (2004) 269–278

Mathara, J.M., 1999. Studies on Lactic acid producing microflora in Mursik and Kule naoto, traditional fermented milks from Nandi and Maasai communities in Kenya. MSc. Thesis. University of Nairobi, Kenya. Mathara, J.M., Miyamoto, T., Koaze, H., Kiiyukia, C., Yoneya, T., 1996. Production of traditional fermented milk in Kenya. Review. Shizuoka Prefectural Univ. Publication, Japan, pp. 257 – 265. Matukumira, A.N., 1996. Investigation of some prospects for the development of starter cultures for industrial production of traditional fermented milk in Zimbabwe. Doctor Scientiarum Thesis. Department of Food Science, Agricultural University of Norway. Medina, R., Katz, M., Gonzalez, S., Oliver, G., 2001. Characterisation of the lactic acid bacteria in ewe’s milk and cheese from Northwest Argentina. Journal of Food Protection 64 (4), 559 – 563. Miyamoto, T., Gichuru, S.G.G., Akimoto, T., Nakae, T., 1986. Identification and properties of lactic acid bacteria isolated from fermented beverages in East Africa. Japanese Journal of Zootechnical Sciences 57 (3), 265 – 276. Nakamura, T., Sugai, M., Ozawa, A.N., Ariga, H., Koaze, H., Kiiyukia, C., Arai, I., Urashima, T., 1999. Microbiological properties of maziwa lala, a Kenyan traditional fermented milk of Maasai community in Kenya. Milk Science 48 (1), 9 – 14.

Olasupo, N.A., Schillinger, U., Holzapfel, W.H., 2001. Studies on some technological properties of predominant lactic acid bacteria isolated from Nigeria fermented foods. Food Biotechnology 15 (3), 157 – 167. Rolle, R., Satin, M., 2002. Basic requirements for the transfer of fermentation technologies to the developing countries. International Journal of Food Microbiology 75 (3), 181 – 187. Schillinger, U., Lu¨cke, F.K., 1987. Identification of lactobacilli from meat and meat products. Food Microbiology 4, 199 – 208. Stiles, M.E., Holzapfel, W., 1997. Lactic acid bacteria of food and their current taxonomy. Review. International Journal of Food Microbiology 36, 1 – 27. Tamang, J.P., Dewan, S., Thapa, S., Olasupo, N.A., Schillinger, U., Wijaya, A., Holzapfel, W.H., 2000. Identification and enzymatic profiles of the predominant lactic acid bacteria isolated from soft-variety Chhurpi, a traditional cheese typical of the Sikkim Himalayas. Food Biotechnology 14 (1&2), 99 – 112. Wood, B.J.B., Holzapfel, W.H., 1995. The Genera of Lactic Acid Bacteria, vol. 2. Blackie Academic & Professional, Glasgow. Xanthopoulos, V., Hatzikamari, M., Adamidis, T., Tsakalidou, E., Tzanetakis, N., Tzanetaki, E.L., 2000. Heterogeneity of Lactobacillus plantarum isolated from Feta cheese throughout ripening. Journal of Applied Microbiology 88, 1056 – 1064.