Interaction between Lactic Acid Bacteria and Mycobacterium bovis in Ethiopian Fermented Milk: Insight into the Fate of M. bovis

Mycobacterium bovis causes tuberculosis in animals and humans.Infected cows can transmit the bacillus to humans via milk.Milk also contains lactic acid bacteria (LAB). LAB isolatedfrom milk were put in milk cultures together with spiked M.bovis. Different LAB had different abilities to reduce M. boviscounts, as M. bovis was undetectable in some while it persistedin two of the cultures.

INTRODUCTION

Top

INTRODUCTION

Mycobacterium bovis causes tuberculosis (TB) at both pulmonaryand extrapulmonary sites in animals and humans (3, 4, 12). Oneway that transmission of M. bovis from animals to humans occursis via consumption of contaminated milk. In Ethiopia, >80%of the population lives in the countryside, farming and rearingcattle. Traditionally, milk is fermented at household temperatures,which can range from 15 to 24°C depending on the locality,in clay pots that are first cleaned and smoked to add flavorand aroma to the milk. Smoking slows the souring process, asdo low temperatures (5). Fermented milk can be stored at householdtemperatures and consumed for up to 20 days (5).

Information on the prevalence of M. bovis infection with animalorigin in humans in Ethiopia is scarce. One study (G. Ameni,P. Bonnet, and M. Tibbo, unpublished report) found a high infectionrate in cattle from 12 dairy farms. In Ethiopia and elsewherein Africa, M. bovis has been isolated from raw milk (3). InTB-afflicted children in Ethiopia, the proportion of extrapulmonaryTB was found to be high in those who frequently drink raw milk(6), and reports from the WHO also showed that, overall, thenotification rate of extrapulmonary TB was high (35% of newcases) (13).

Lactic acid bacteria (LAB) are found in various habitats, includingmilk (10, 11). Several reports on the enumeration and identificationof LAB from different African fermented milk samples showedthat mean LAB counts range between 10⁷ to 10⁹ CFU/ml of fermentedmilk (1, 5, 7, 9). LAB may inhibit pathogens in foods by producingantimicrobial peptides and/or acid, in addition to the competitiveexclusion of pathogens (2, 8). This study made use of milk cultureswith or without LAB, into which M. bovis was spiked, and wasintended to give insight about the fate of M. bovis in milkduring the natural fermentation process, as determined by plating.

Isolation of LAB, collection of milk, and growth of M. bovis.

Top

Isolation of LAB, collection...

LAB were isolated earlier from fermented milk on LAB media (MRSagar, plate count agar [PCA], and M17 agar; Oxoid Ltd., Basingstoke,England), as summarized in Table 1, and were stored in a freezerfor future use in milk cultures. Milk to be used for cultureswas collected hygienically from a skin test-negative cow andplaced into sterile containers. To produce bacterial suspensionfor spiking, M. bovis ATCC 19210 was grown in Middlebrook 7H9broth containing 0.05% Tween 80, 10% oleic acid-albumin-dextrose-catalase,and 0.4% pyruvate. This laboratory-grown strain may not behaveexactly like M. bovis from the milk of infected cattle, butthis experiment serves as a starting point to assess the invivo situation.

View this table:
[in this window]
[in a new window]

TABLE 1. Media and conditions for LAB isolation from fermented milk and presumptive identification^a

Milk culture setup.

Top

Milk culture setup.

Preliminary experiments showed that heating milk at 80°Cfor 10 min kills all resident microflora. A summary of the milkculture setup, including whether heat treatment was used, LABwas added, and M. bovis was spiked, is given in Table 2. Milkwas divided into 1-liter portions each in 2-liter presterilizeddark glass, screw-cap bottles, heat treated, and cooled. Onemilliliter of M. bovis (optical density at 580 nm, 0.700) culturewas spiked into milk and gently shaken. Then, 100 µl ofindividual (milk cultures 1 to 5) or mixed (milk culture 6)LAB (optical density at 600 nm, 0.600) was added. The bottles,along with the controls (milk cultures 7 and 8), were kept atroom temperature (20 to 22°C) in a cardboard box withinthe P3 laboratory.

View this table:
[in this window]
[in a new window]

TABLE 2. Milk culture setup showing which was heat treated and to which LAB and/or M. bovis was added

Enumeration of M. bovis.

Top

Enumeration of M. bovis.

On days 0, 3, 7, and 14, 10 ml from each milk culture was withdrawnafter gentle agitation of the bottle, mixed with 5 ml phosphate-bufferedsaline (PBS) to aid in homogenization, and centrifuged at 1,000x g for 90 min at 20°C using a Beckman GS-6 centrifuge andBeckman GH-3.8 rotor. The pellet and cream were resuspendedin 2 ml PBS, 10-fold serial dilutions (1 ml into 9 ml PBS) weremade, and 2 ml of the first and second dilutions treated byadding 7.2 µl of an antibiotic cocktail (consisting ofchloramphenicol at 4 µg/ml, trimethoprim at 8 µg/ml,polymyxin B at 100 U/ml, ampicillin at 2 µg/ml, gentamicinat 0.5 µg/ml, and ketoconazole at 0.4 µg/ml [SigmaChemical Co.]) and left standing for 1 h. Then, portions werespread onto Middlebrook 7H10 agar plates containing the antibioticcocktail at the concentrations described above in triplicate.Plates were placed in a 37°C incubator. Counting of thecolonies was done between the fourth and sixth weeks and wasrepeated up to six times. Plots of mean M. bovis counts weregenerated using GraphPad Prism version 4.01 (GraphPad Software,Inc.).

In milk cultures 1, 2, and 3, M. bovis was undetectable by day14, while it persisted in milk cultures 4 and 5 (Fig. 1A). Inmilk culture 6, which contained LAB mix, M. bovis was undetectableby day 7 (Fig. 1A). M. bovis continued to grow in milk culture7, to which no LAB were added, whereas it was undetectable inmilk culture 8, which contained indigenous LAB, by day 14 (Fig.1B). All those plates without growth were inspected for a further4 weeks (beyond 6 weeks). No growth was observed, indicatingthat LAB eliminated M. bovis. For these milk cultures, the lowestdetectable limit of M. bovis was estimated to be 1 to 2 CFU/ml.At the concentrations of the antibiotic cocktail used, therewas no significant difference in CFU between M. bovis platedwith or without antibiotics (P > 0.05).

View larger version (18K):
[in this window]
[in a new window]

FIG. 1. M. bovis counts enumerated by plating dilutions of pellet and cream fractions of milk cultures from the LAB-M. bovis interaction study. Data presented at each time point are means ± standard deviations, where n is 3. (A) M. bovis-spiked cultures with LAB added. Symbols: ${blacklozenge}$ , LAB1 (milk culture 1); ${blacksquare}$ , LAB2 (milk culture 2); ${blacktriangleup}$ , LAB3 (milk culture 3); x, LAB4 (milk culture 4); +, LAB5 (milk culture 5); •, LAB mix (milk culture 6). (B) M. bovis-spiked cultures without LAB added. Symbols: ${Delta}$ , milk culture 7; ${square}$ , milk culture 8 (contains indigenous LAB).

The M. bovis strain used for spiking the milk cultures was ableto grow in milk in the absence of LAB, as evidenced by colonyappearance in milk culture 7 and also as determined by plating2 µl of the M. bovis suspension used for spiking ontoMiddlebrook 7H10 plates, which gave viable counts of 4.83 x10³ CFU/ml and 5.76 x10⁴ CFU/ml on days 3 and 7, respectively.

To assess if LAB were still present in the milk cultures, 100µl from some of the milk cultures (2, 6, and 8) was platedonto MRS agar and PCA on day 21 of culture. Plates were incubatedat 37°C for 48 h. Thus, milk cultures 2, 6, and 8 containedLAB at levels of 7.20 x 10⁴, 4.80 x 10⁴, and 1.60 x 10⁴ CFU/ml,respectively. However, these counts were probably underestimates,since different LAB species have different growth requirementsfor temperature and aerobiosis or anaerobiosis. The eliminationof M. bovis by LAB might be quickened if the diversity of LABspecies increases.

It was beyond the scope of this work to determine the actualmechanisms of inhibition for M. bovis by LAB. However, the resultsof this experiment are consistent with previous experimentsconducted in this laboratory with the same objective.

ACKNOWLEDGMENTS

I thank Susanna Grundstrom, Morten Harboe, and Sven Hoffnerfor material assistance and Liya Wassie for technical assistance.

This work was supported by the AHRI core budget. AHRI is supportedby NORAD and Sida.

FOOTNOTES

* Mailing address: AHRI, P.O. Box 1005, Addis Ababa, Ethiopia. Phone: 251 11 321 1332. Fax: 251 11 321 1563. E-mail: solomonic2000{at}hotmail.com

Published ahead of print on 23 January 2009.

REFERENCES

Top