Isolation of Brucella microti from soil.

To the Editor: Brucella microti is a recently described Brucella species () that was isolated in 2000 from systemically infected common voles (Microtus arvalis) in South Moravia, Czech Republic. The organism is characterized by rapid growth on standard media and high metabolic activity, which is atypical for Brucella (). The biochemical profile of B. microti is more similar to that of Ochrobactrum spp., of which most species are typical soil bacteria.

On the basis of the close phylogenetic relationship of Brucella spp. and Ochrobactrum spp. and the high metabolic activity of B. microti, we hypothesized that this Brucella species might also have a reservoir in soil. To test this hypothesis, we investigated 15 soil samples collected on December 11, 2007, from sites in the area where B. microti was isolated from common voles in 2000 (). Ten of the samples were collected from the surface and at a depth of up to 5 cm near different mouse burrows 5 m apart. The remaining 5 samples were collected from an unaffected area without clinical cases of vole infection. The pH of soil samples ranged from 5.9 to 6.3. No frosts were recorded before the time of collection.

To specifically detect B. microti in soil samples, we have developed a PCR that targets a genomic island of 11 kb (H.C. Scholz et al., unpub. data) that is unique for B. microti. Briefly, primers Bmispec_f (5′-AGATACTGGAACATAGCCCG-3′) and Bmispec_r (5′-ATACTCAGGCAGGATACCGC-3′) were used to amplify a 510-bp fragment of the genomic island. PCR conditions were denaturation at 94°C for 5 min, followed by 29 cycles at 94°C for 30 s, 60°C for 30 s, and 72°C for 30 s. Total DNA was prepared from 0.5 g of each soil sample by using the MO BIO Ultra Clean Soil DNA Kit (Dianova, Hamburg, Germany). DNA was eluted with 50 μL of double-deionized water of which 2 μL was used in PCRs. Template DNA of B. microti CCM 4915T was used as a positive control. Type strains of all recognized Brucella species, 1 strain of each biovar of all species, and type strains of 11 Ochrobactrum species were used as negative controls.

In this PCR, 5 of 15 soil samples and the positive control were positive for the 510-bp fragment; other Brucella spp. and Ochrobactrum spp. were negative. Of the 5 positive samples, 3 were collected from surface soil collected near mouse burrows. However, the remaining 2 positive samples were collected from the unaffected and supposedly negative-control area.

For direct cultivation of Brucella spp. from soil, 2 g each of 2 selected PCR-positive samples with the highest amplification rate (both from the affected area) were thoroughly homogenized in 5 mL of phosphate-buffered saline (PBS), pH 7.2, in 50-mL tubes. Of a serial dilution in PBS (100–10–4), 100 μL was plated onto Brucella agar (Merck, Darmstadt, Germany) supplemented with 5% (vol/vol) sheep blood (Oxoid, Wesel, Germany) and Brucella selective supplement (Oxoid) and incubated at 37°C. Twenty suspicious colonies from the 100 dilution plate of 1 soil sample were subcultivated on Brucella selective agar. Two of the subcultivated bacteria (BMS 17 and BMS 20) reacted positively with monospecific anti-Brucella (M) serum. Both isolates were positive in the B. microti–specific PCR. Sequencing of the 510-bp fragments from both strains (GenBank accession nos. AM943814 and AM943815) and comparison with the known nucleotide sequence of B. microti showed 100% identity.

To confirm that strains BMS 17 and BMS 20 were B. microti, these strains were subjected to multilocus sequence analysis and multilocus variable number of tandem repeat analysis (MLVA) as described (,–). Multilocus sequence typing profiles of these strains were identical to the type strain B. microti CCM 4915T and strain CCM 4916. MLVA showed that these strains also clustered with B. microti strains CCM 4915T and CCM 4916, with identical panel 1 and panel 2A genotypes but a different panel 2B genotype.

In summary, we successfully isolated B. microti from soil samples collected at the same site 7 years after primary isolation of this novel species from common voles. B. microti could still be isolated from the same soil samples 6 months after storage at 4°C. This finding indicates long-term survival of B. microti in soil; thus, soil might function as a reservoir of infection. Identification of B. microti as a potential soil bacterium is consistent with Brucella spp. whole genome sequencing data, in particular with the genome sequence of B. suis, which exhibits fundamental similarities with plant pathogens such as Agrobacterium spp. and Rhizobium spp. (). Whether soil is the primary habitat of B. microti or other vectors, such as nematodes, remains to be investigated.