Background

While Mycobacterium tuberculosis and Mycobacterium bovis are the classic agents causing human and bovine tuberculosis (TB) respectively, there have been instances of the isolation of M. tuberculosis as well as M. bovis from bovine tissues and milk [1–3]. There is widespread infection with bovine tuberculosis in Ethiopian cattle [2, 4, 5]. These two tubercle bacilli can be transmitted between humans and animals in two-way fashion directly (i.e., via aerosol inhalation) or indirectly from animals to humans (i.e., by consumption of raw contaminated dairy products and meat) [6, 7]. Milking cows with TB of the udder may also act as amplifiers of TB by rendering the whole milk supply infective in places where milk pooling is practiced. Consumption of such foods is likely to cause extrapulmonary TB (EPTB).

Neither the diagnosis of EPTB nor the distinction whether TB is caused by M. tuberculosis or M. bovis is possible by chest radiography, sputum smear microscopy and sputum culture. Misdiagnosis of M. bovis infection for M. tuberculosis or vice versa can have important consequences, including the wrong treatment with pyrazinamide in the case of the former. Reports of the World Health Organization continue to show that the notification rate of EPTB in Ethiopia is very high (35% of new cases) [8].

Lactic acid bacteria (LAB) are indigenously found in various habitats including milk, meats and the mucosal surfaces of animals [9, 10]. LAB have been used for centuries for preservation and bioconversion of various foods and drinks (e.g., production of yogurt, wine, cheese, beverages, sauerkraut, etc). They also have important industrial applications in the making of fermented foods and drinks. Many of the LAB species are nonpathogenic and safe, at least in healthy and immune-competent people, and some have been designated as probiotics. These species have several claimed benefits, one of which is their potential for use as anti-infective agents [11–13].

Several reports indicate the potential of selected LAB species, among their other beneficial effects, in inhibiting pathogenic bacteria such as Helicobacter pylori[14–18], Listeria monocytogenes[19, 20], Salmonella enterica[21] and Clostridium species [20, 22, 23] by various mechanisms including inhibition of bacterial adhesion, enhancement of host immunity and production of antimicrobial substances.

In Ethiopian country households, fresh milk is traditionally allowed to ferment at room temperature. Fermented milk can be stored and consumed for up to 20 days. Moreover, raw fresh and fermented milk is consumed widely [24]. Consumption of raw, contaminated milk is a risk factor for TB [25, 26].

This study aimed both to identify and evaluate the survival of mycobacteria in milk fermented in the laboratory under conditions simulating traditional household milk fermentation. The study involved milk culture in which the fate of inherently-present mycobacteria could be assessed by periodic plating on defined, selective solid medium.

The study site and milk collection

The site from which the milk sample used in this experiment was obtained is located 30 kms west of the Ethiopian capital, Addis Ababa. It is a State Dairy Farm with herd size of 200 cows. The farm serves as one of the major suppliers of milk to the city, to which milk from all lactating cows is pooled into large containers and sent. A one-liter pooled milk sample was obtained from 30 tuberculin skin-test positive cows from the farm. The milk was collected hygienically into sterile containers and transported to AHRI within a cold chain. The milk was maintained within dark bottle in a cardboard box at room temperature for the duration of the experiment.

Antibiotics for selective inhibition of LAB

To inhibit the growth of LAB during enumeration of mycobacteria, an antibiotic cocktail was incorporated into Middlebrook 7H10 agar medium as described previously [27].

Enumeration of mycobacteria

Enumeration of mycobacteria at time intervals 0, 3, 7 and 14 days was performed as described previously [27].

Identification of mycobacteria from pooled milk

Tests used for identification of unknown bacteria from the milk culture included acid-fast staining, catalase test and PCR. For PCR, a 2X master mix from Qiagen was used. The PCR reactions were multiplex PCR using three sets of primers: Mycgen-F, Mycgen-R, Mycav-R; RD4F, RD4R, RD4intF; and RD10F, RD10R, RD10intR [28]. Mycgen-F and Mycgen-R are specific for the genus Mycobacterium and give a product size of 1030 bp. Mycgen-F together with Mycav-R is specific for Mycobacterium avium with a product size of 180 bp. The cycling conditions were an initial denaturation at 95°C for 15 min followed by 35 cycles of denaturation at 95°C for 1 min, annealing at 61°C for 30 s and extension at 72°C for 2 min with final extension at 72°C for 7 min. With the RD4 multiplex, a product size of 446 bp is expected if it is M. bovis and 335 bp if it is M. tuberculosis. With RD10, a product size of 202 bp is expected if it is M. bovis and 308 bp if it is M. tuberculosis. The cycling conditions for both RD4 and RD10 primers were initial denaturation at 95°C for 15 min followed by 35 cycles of denaturation at 95°C for 1 min, annealing at 55°C for 1 min and extension at 72°C for 1 min with final extension at 72°C for 10 min.

Bacterial growth from pooled milk culture and pH of fermented milk

From the pooled milk culture, growth was obtained on 7H10 plates. It was later found to be acid-fast-positive. Tests conducted to assess the catalase activity of this bacterial growth yielded catalase-positive results. There were 4.7 ± 4.4 and 3.9 ± 3.7 Log CFU mL^-1 on days zero (start of the milk culture) and 3 respectively. This mycobacterial growth was subject to elimination by indigenous LAB after 7 days.

The pH of milk was measured by saving 50 mL of the milk sample for this purpose. The pH changed from ~ 7.00 on day zero to 4.59 on day 8, with the day zero value being dependent on how fast after milk collection the pH was measured.

Identification of bacterial growth from pooled milk culture by PCR

The isolate from the milk culture, following molecular characterization by PCR, gave a product size of 1030 bp with the Mycobacterium genus-specific primers (Figure 1a). The species-level identification with the RD4 deletion primers gave a product size of 335 bp while with the RD10 deletion primers it gave a product size of 308 bp, both typical of M. tuberculosis DNA amplification products (Figure 1b), indicating the presence of M. tuberculosis in the pooled milk sample. The PCR results with the genus-specific and deletion primers also confirmed that the milk isolate was neither Mycobacterium avium nor Mycobacterium africanum.