B AS E Biotechnol. Agron. Soc. Environ. 2005 9 (3), 191–194

Variability of the microbial abundance of a kefir grainstarter cultivated in partially controlled conditionsVéronique Ninane (1), Gilbert Berben (1), Jean-Michel Romnee (1), Robert Oger (2)

(1) Département Qualité des Productions agricoles. Centre wallon de Recherches agronomiques. Chaussée de Namur 24.B–5030 Gembloux (Belgique). E-mail : ninane@cra.wallonie.be(2) Section Biométrie, Gestion des Données et Agrométéorologie. Centre wallon de Recherches agronomiques. Rue deLiroux 9. B–5030 Gembloux (Belgique).

Received 7 September 2004, accepted 11 May 2005

The variability of the abundance in lactobacilli, lactic acid streptococci and yeasts of a kefir grain starter cultivated in partiallycontrolled conditions – milk renewal at room temperature and incubation at defined temperature - was quantified. Expressedby the geometric relative standard deviations of counts repeated over time, it was respectively of 28%, 443% and 35% foreach of the three microbial groups. The origin of the microbial variability observed was ascribed to the heterogeneous mediumdeveloped around the grains during fermentation and, for the lactic acid streptococci, even more to the initial conditions offermentation, probably to parameters linked to milk renewal (room temperature, waiting time at room temperature beforeincubation).Keywords. Starter cultures, kefir, microorganims, lactic acid bacteria, yeasts.

Variabilité de l’abondance microbienne d’un ferment de kéfir traditionnel cultivé en conditions partiellementcontrôlées. La variabilité de l’abondance en lactobacilles, en coques lactiques et en levures de grains de kéfir cultivés dansdes conditions partiellement contrôlées – manipulations de renouvellement du lait à température ambiante et incubation enétuve contrôlée - a été quantifiée. Exprimée par l’écart type relatif géométrique d’énumérations répétées dans le temps, elleétait respectivement de 28 %, 443 % et 35 % pour chacun des trois groupes microbiens. L’origine de la variabilité microbienneobservée a été attribuée au développement d’un milieu hétérogène autour des grains pendant la fermentation et, pour lescoques lactiques, plus encore aux conditions initiales de fermentation, probablement des paramètres liés au renouvellementdu lait (température ambiante de la pièce, temps d’attente avant incubation).Mots-clés. Culture starter, kéfir, micro-organismes, bactérie lactique, levure.

1. INTRODUCTION

Kefir grains are those Caucasian originated dairystarters that look like cauliflower florets. These grainsconsist of a complex set of lactic acid bacteria, yeastsand, sometimes, acetic bacteria embedded in a specificpolysaccharide matrix. Kefir grains have the propertyto perpetuate, that is growing and multiply whilemaintaining a stable population. This constancy of thegrain microbial composition is a generally acceptedfeature (Garrote et al., 1998; Kuo, Lin, 1999; Simovaet al., 2002) that stay vague.

The aim of this work is to assess the standarddeviation around it for the main microbial groups of akefir grain starter cultivated in partially controlledconditions and to point out the sources of variation.

2. MATERIALS AND METHODS

2.1. Origin and maintenance of the kefir grainstarter

The kefir grain starter comes from the collection of theWalloon Agricultural Research Centre, DepartmentQuality of agricultural products (Gembloux, Belgium).It has been collected in the 1990’s from a Belgianhousehold. Since then, it has been propagated insterilised cow milk renewed daily during the workingdays. At every transfer in fresh tempered milk, thegrains are rinsed with sterile water. Handlings of milkrenewal occur at room temperature and grains are thenincubated without shaking at 22 °C. Once a week, thegrain:milk ratio is adjusted to 20% (w/v).

For the experiment, two sub-samples of the kefirgrain starter were cultivated in this way in standardisedcontainers. At the time of grain:milk ratio adjustment,the weight of the rinsed and dripped kefir grains wasreduced to 57 g ± 3 g. Care was taken to limit theweight difference between duplicates to less than 2 g.

2.2. Microbial composition of the kefir grain starter

Lactobacilli, lactic acid streptococci and yeasts in thetwo cultures were counted weekly for seven weeks.Grains were harvested for counting at the time ofgrain:milk ratio adjustments.

2.3. Microbial counts

A first 1:9 (w/v) dilution was prepared by dispersing1 2 g ± 3 g of rinsed and dripped grains in theappropriate volume of Ringer solution with an Ultra-turrax T25 (IKA-Labortechnik, Germany). Subsequentserial dilutions were plated on MRS-Agar, on M 17-Agar and on YGC-Agar for enumeration ofrespectively the lactobacilli, the lactic acidstreptococci and the yeasts (all the media werepurchased from Merck). Plates were incubatedanaerobically during three days at 37 °C for thelactobacilli, aerobically during respectively two daysat 37 °C and five days at 22 °C for the lactic acidstreptococci and the yeasts. The exclusive presence ofyeasts on the yeast-mould specific medium wasexamined through colony morphology.

2.4. Accuracy of the microbial counting methods

Accuracy of the microbial counts was determined onseven replicates coming from a single batch of grains.

2.5. Statistical analysis

Standard deviations between duplicates (repeatabilityof measures between duplicates at the same time) andbetween weeks (reproducibility of measures overtime) were deduced from the variance componentscalculated with the VARCOMP procedure of the SASsoftware (SAS Institute, 1994).

2.6. Visualisation of microbial development with achromogenic substrate

Two grains of about 4 g each were cultivated inparallel during 24 hours in milk added with 5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside (X-gal) at0.4 mg ml-1. One of the cultures was kept unshakenduring the growing while the other was shaken at 140rotations per minute. They were both incubated at22 °C with a grain:milk ratio of 20% (w/v).

3. RESULTS AND DISCUSSION

The geometric means of the microbial counts,expressed in CFU* g-1 of grains, were of 1.4 × 108,3 . 9 × 1 04 and 1.1 × 1 07 respectively for thelactobacilli, the lactic acid streptococci and the yeasts.These figures of microbial abundance in kefir grainsare comparable to those published by Pintado et al.(1996).

The geometric relative standard deviations (GRS)of the counts, assigned to each of the two sources ofvariation: the duplicate cultures (GRSduplicates) andthe successive weeks (GRSweeks), are shown for eachmicrobial group in table 1. Due to the accuracy of thecounting methods, the deviation between the twocultures and, even more, between weeks wasundoubtedly biologically significant. The countingmethods gave indeed measurement errors that,expressed as GRS, were as low as 9%, 7% and 4% forlactobacilli, lactic acid streptococci and yeastsrespectively.

To our knowledge, the literature on the microbialcomposition of kefir grains refers only to occasionalmeasurements, not repeated over time, and only onestudy has been conducted with individual cultures ofthe same kefir grain starter (Simova et al., 2002).Calculated from the author’s data, the GRS for thecultures are 32%, 18% and 57% respectively forlactobacilli, lactic acid streptococci and yeasts. Thesefigures express a variation comparable to that observedin our study, but its biological significance isuncertain.

Ta b l e 1 reveals that the growth of the threemicrobial groups varied between the cultures. Thegrowth of the lactic acid streptococci and, to a lesserextent, the yeasts showed even more unstable overtime as the week-to-week deviation exceeds thatbetween duplicates.

192 Biotechnol. Agron. Soc. Environ. 2005 9 (3), 191–194 Ninane V., Berben G., Romnee JM., Oger R

CFU: Colony forming unit

Table 1. Variability of the measures expressed as geometricrelative standard deviations (%) between duplicates(GRSduplicates) and between weeks (GRSweeks) — Variabilitédes mesures exprimée par l’écart type relatif géométrique(%) entre les doubles (GRSduplicates) et entre les semaines(GRSweeks).

Microbial groups GRSduplicatesa GRSweeksa

Lactobacilli 28 28Lactic acid streptococci 29 443Yeasts 24 35a Values calculated from measures done seven times at intervalsof one week in two parallel kefir grain cultures.

In order to point out the origin(s) of the variationsobserved in microbial growth, the uncontrolledparameters were picked out and analysed.

Among them, appeared the manual rinsing of thegrains with a degree of rinsing being left to theoperator’s discretion. The subjectivity associated withgrain rinsing was however not a major source ofmicrobial variation, as omitting this step in a paralleltrial involving two other cultures did not in fact affectthe grain microbial composition (data not show).

Other possible sources of variation were the risingof grains to the surface and the syneresis of the milkthat occurred during fermentation. These eventsg e n e r a t e d a heterogeneous medium around the grainsthat was clearly variable: on removal from theincubator, some grains were surrounded by air whileothers by whey or curd and this varied from containerto container. The relation between this observation andmicrobial variation implies that the heterogeneousmedium generated during fermentation may influencethe grain microbial growth. This latter hypothesis wasconfirmed with the aid of a chromogenic substraterevealing a galactosidase activity. As lactose is themain available carbohydrate source, the activity of thisenzyme is indeed an indicator of the microbial growthfor the lactose-metabolising organisms, let say themajority of the kefir grain microflora. Grain comingfrom the heterogeneous medium showed an unevendistribution of microorganisms with a galactosidaseactivity (Figure 1a), while it was evenly distributedthroughout the all grain when it was incubated in amedium that was homogenised by shaking (Figure1b). That this discrepancy in growing behaviour is dueto an external origin was confirmed by giving evidencethat the microorganisms from the uncoloured parts ofthe grain may have a galactosidase activity. Spreadingsmear preparations from these areas on a culturemedium in fact produced colonies that all turned bluewhen drops of chromogenic substrate were added (datanot show).

Some last parameters that determined the initialfermentation conditions i.e. just before incubationcould have been a source of variation. Indeed, the milkwas temperate and renewed at ambient temperature, letsay not regulated. Furthermore, due to the design ofthe experiment, cultures had to be allowed to stand forabout three-quarters of an hour before being put intothe incubator once the milk had been renewed.Although this duration is rather short if compared tothe length of the one day incubation, it might have asignificant influence on the evolution of the kefir grainmicroorganisms. This point was indirectly confirmedby figures obtained for the milk pH measured justbefore placing the cultures in the incubator. From oneday to another, they varied in a range between 5.3 and6.0, showing that the microbial activity was already

important and irregular. The microorg a n i s m sresponsible for primary acidification of the milk inkefir making are lactic acid streptococci (Koroleva,1988), precisely those that showed the greatestvariability of counts over time. Moreover, amongthem, lactococci are reputed to be heat-sensitive(Dellagio et al., 1994).

4. CONCLUSIONS

Significant variation in quantitative microbialcomposition of a kefir grain starter occurred inpartially controlled cultivation conditions. Variation inmicrobial abundance was observed for the three mainmicrobial groups of the kefir grain starter, but, amongthem, the lactic acid streptococci showed by far thegreatest variability.

Events that come with kefir fermentation - grainrising and milk syneresis - were pointed out for beingat the source of a microbial variability that affectedalmost equally the three microbial groups.

The lactic acid streptococci were however moreaffected by another source of variability that occurredat the early beginning of the fermentation. This sourcewas probably related to the room temperature and thewaiting time of the containers before placing into theincubator. Those latter parameters were indeed notregulated.

Kefir grain: variability of its microbial abundance 193

Figure 1. Visualisation of the grain microbial developmentwith a chromogenic substrate (X-gal). The distribution ofthe microorganisms able to metabolise the X-gal, the blueones, is heterogeneous after one day of grain cultivationwithout shaking (a) whilst it is homogeneous when theculture is shaken (b) — Visualisation du développementmicrobien du grain à l’aide d’un substrat chromogénique(X-gal). La répartition des micro-organismes capables demétaboliser le X-gal, colorés en bleu, est hétérogène aprèsun jour de culture sans agitation (a), tandis qu’elle esthomogène lorsque la culture est agitée (b).

Acknowledgements

We are grateful to Mohamed Echahraoui for his technicalassistance

Bibliography

D e l l a g i o F., de R o i s s a r t H., To r r i a n i S., Curk M C . ,J a n s s e n s D. (1994). Taxonomie, métabolisme,croissance et génétique des bactéries lactiques. In H. deRoissart, FM. Luquet ( Eds.). Bactéries lactiques.Uriage, France : Lorica vol. 1, p. 23–116.

G a r r o t e GL., A b r a h a m AG., De A n t o n i GL. (1998).Characteristics of kefir prepared with diff e r e n tgrain:milk ratios. J. Dairy Res. 65, p. 149–154.

Koroleva NS. (1988). Kefir and kumys starters. Bull. Int.Dairy Fed. 227, p. 35–40.

Kuo CY., Lin CW. (1999). Taiwanese kefir grains: theirgrowth, microbial and chemical composition offermented milk. Aust. J. Dairy Technol. 54, p. 19–23.

Pintado ME., Da Sylva LJA., Fernandes PB., Malcata X.,H o g g JA. (1996). Microbiological and rheologicalstudies on Portuguese kefir grains. Int. J. Food Sci.Technol. 31, p. 15–26.

SAS Institute Inc. (1994). SAS/STAT User’s Guide. Cary,NC, USA: Statistical Analysis System Institute.

S i m o v a E., Beshkova D., A n g e l o v A., Hristozova Ts . ,Frengova G., Spasov Z. (2002). Lactic acid bacteria andyeasts in kefir grains and kefir made from them. J. Ind.Microbiol. Biotechnol. 28, p. 1–6.

(7 ref.)

194 Biotechnol. Agron. Soc. Environ. 2005 9 (3), 191–194 Ninane V., Berben G., Romnee JM., Oger R