Background

Small mammals (Soricomorpha and Rodentia) are important reservoirs for many zoonotic tick-transmitted pathogens [1]. Several species of the genus Ixodes may serve as bridge vectors for Borrelia spp. that allow their circulation among different hosts, including small mammals [2] that are common hosts for the immature tick. Synanthropic micromammal species have been widely investigated because they act as reservoirs in the natural transmission cycles of Borrelia spp. [3]. Moreover, they are considered epidemiological markers in evaluating the distribution of certain tick species [4,5,6,7]. Although the infection rate with Borrelia spp. in these hosts is usually low, their local abundance may confer them an important epidemiological role [8].

Romania has a wide range of habitats colonized by species of small mammals and ticks. The small mammal-vector associations in Romania have been investigated by Mihalca et al. [4], who showed that the majority of the ticks on these vertebrates were I. ricinus, the most important tick parasitizing humans [7] and the only known vector for Lyme borreliosis in Europe. People living in rural areas in Romania are in close contact with the habitats preferred by small mammals and their ticks. This promotes a serious risk by tick-borne infectious agents that have a major impact on public health [9]. The aim of this study was to provide epidemiological data from a survey of the presence of Borrelia spp. in small mammals from Romania thus complementing existing surveys in ticks and other vertebrates.

Methods

Sampling

All animals were caught in Romania during 2010–2011, as previously described [4]. Each captured rodent was identified to the species level. A total of 401 small mammals of 11 species (Apodemus agrarius, A. flavicollis, A. sylvaticus, A. uralensis, Micromys minutus, Microtus agrestis, M. arvalis, M. subterraneus, Myodes glareolus, Crocidura leucodon and C. suaveolens) were collected from 7 counties of Romania. Tissue samples were collected from the animals: both the heart and liver from 393 animals (98%), only the heart from 2 animals (0.5%) and only the liver from 6 animals (1.5%).

DNA extraction

Genetic material isolation was performed individually from tissues using a commercial DNA extraction kit (Isolate II Genomic DNA Kit; Bioline, London, UK) according to the manufacturer’s protocol. Extracted DNA was stored at − 20 °C for further analysis. For each extraction procedure, negative controls were used in order to identify possible cross-contamination.

qPCR and sequencing

Multiplex quantitative polymerase chain reaction (qPCR) was used for evaluating the presence of Borrelia spp. in tissue samples. For B. burgdorferi (s.l.) we targeted the gene encoding the outer surface protein A (ospA gene) and for B. miyamotoi a part of the flagellin B (flaB gene). The qPCR was performed as previously described [10,11,12] in a CFX96 Touch™ Real-Time PCR detection system (Bio-Rad, London, UK) in a final reaction volume of 20 μl, using IQ Multiplex Powermix (Bio-Rad).

All the qPCR positive samples were amplified by conventional PCR and sequenced (Macrogen Inc., Amsterdam, Netherlands). Conventional PCR was performed as described by Szekeres et al. [8], targeting the glycerophosphodiester phosphodiesterase gene (glpQ) of B. miyamotoi. For B. burgdorferi (s.l.) the 5S-23S rDNA intergenic spacer region (IGS) and the outer surface protein A (ospA) gene were targeted [12]. Nucleotide sequences were compared with those available in GenBank using the Basic Local Alignment Search Tool. In each PCR reaction set, positive and two negative controls were included.

Statistical analysis

Statistical analysis was performed by using Epi Info™ v.7.1.5 software. The frequency, infection prevalence and its 95% confidence interval were evaluated using a Chi-square independence test. A P-value of < 0.05 was considered statistically significant.

Results

Infection of small mammals with Borrelia spp

The overall prevalence of Borrelia spp. infection in micromammal tissues was 4.9% (95% CI: 3.2–7.5%). One out of two M. agrestis was infected (50%, 95% CI: 1.2–98.7%); infection prevalence was 14.2% in C. suaveolens (95% CI: 5.9–27.2%), 9.1% in Mi. minutus (95% CI: 0.2–41.2%), 6.2% in My. glareolus (95% CI: 0.7–20.8%), 5.6% in M. arvalis (95% CI: 1.1–15.6%), 4.7% in C. leucodon (95% CI: 0.1–23.8%), 4% in A. flavicollis (95% CI: 0.5–13.9%) and 3.4% in A. agrarius (95% CI: 0.7–9.7%). The infection prevalence was significantly different among micromammal species (χ2 = 23.6, df = 10, P < 0.01) suggesting a very different role in its intrinsic importance in the life-cycle of Borrelia spp. Three species, A. sylvaticus, A. uralensis and M. subterraneus, were not infected with Borrelia spp. (Table 1).

Table 1 The prevalence of Borrelia spp. in collected tissues
Full size table

The diversity of Borrelia spp. in small mammal tissues

Three species of the B. burgdorferi (s.l.) complex (B. afzelii, B. garinii/B. bavariensis and B. burgdorferi (s.s.) and one species of the relapsing fever group (B. miyamotoi) were identified. The most frequently detected species was B. afzelii (70%; 95% CI: 45.7–88.1%) followed by B. garinii/B. bavariensis, B. burgdorferi (s.s.) and B. miyamotoi (10%, 95% CI: 1.2–31.7%) (Table 2).

Table 2 The frequency and prevalence of Borrelia spp. in tissues
Full size table

Borrelia afzelii was detected in 8 species: A. agrarius (GenBank: KY038873), A. flavicollis (KY038874), C. leucodon (KY123664), C. suaveolens (KY123665), Mi. minutus (KY123663), M. agrestis (KY123654), M. arvalis (KY123655) and My. glareolus (KY123656). Borrelia garinii/B. bavariensis, B. burgdorferi (s.s.) and B. miyamotoi were detected in single individuals of two species each: B. garinii/B. bavariensis in A. agrarius (KY123657), B. garinii/B. bavariensis in M. arvalis (KY123658), B. burgdorferi (s.s.) in A. agrarius (KY123659), B. burgdorferi (s.s.) in M. arvalis (KY123662), B. miyamotoi in A. flavicollis (KY123660) and B. miyamotoi in My. glareolus (KY123661).

The infection rate in tissue samples

The overall infection rate was 3.2% in the heart and 2.7% in the liver, without significant difference between the tissue samples (P = 0.3) (Table 3). Four infections (20%; 95% CI: 5.7–43.7%) were detected in the heart and liver (two in A. agrarius and one each in C. sueveolens and M. agrestis), nine (45%; 95% CI: 23.1–68.5%) only in the heart (five in C. suaveolens, and one each in A. flavicollis, C. leucodon, Mi. minutus and My. glareolus) and seven (35%; 95% CI: 15.4–59.2%) only in the liver (three in My. glareolus and one each in A. agrarius, A. flavicollis, C. suaveolens and My. glareolus).

Table 3 The frequency of Borrelia spp. in the heart and liver of each animal
Full size table

From the total of 395 heart samples, 11 (95% CI: 1.6–4.9) were infected with B. afzelii, one (95% CI: 0.1–1.4) with B. garinii/B. bavariensis and B. burgdorferi (s.s.) In the liver, B. afzelii was detected in five samples (95% CI: 0.5–2.9), while B. garinii/B. bavariensis, B. miyamotoi and B. burgdorferi (s.s.) were each detected in two samples (95% CI: 0.1–1.8) (Table 2). None of the samples were co-infected.

Regarding the distribution of species by county, B. afzelii was found in Cluj (95% CI: 3.0–9.9), Constanţa (95% CI: 0.1–6.9) and Tulcea (95% CI: 0.1–26.0), B. garinii/B. bavariensis in Constanţa (95% CI: 0.1–6.9) and Bacău (95% CI: 0.8–90.6) counties, respectively B. burgdorferi (s.s.) in Constanţa (95% CI: 0.1–6.9) and Cluj (95% CI: 0.1–2.7) counties. Borrelia miyamotoi (95% CI: 0.1–3.5) was detected in A. flavicollis and My. glareolus trapped in Cluj (Table 4). Borrelia spp. DNA sequences detected in both tissues (heart and liver) in case of one specimen were 100% identical.

Table 4 The infection prevalence of Borrelia spp. in tissues of mammals captured in each county (see [3] for a review)
Full size table

Discussion

Apodemus flavicollis, A. sylvaticus and My. glareolus are the most common rodent species in Europe [8]. Their role in the circulation of Borrelia spp. is well acknowledged [9, 13, 14]. The present study shows that DNA of Borrelia spp. is prevalent in A. agrarius, A. flavicollis, C. leucodon, C. suaveolens, Mi. minutus, M. agrestis, M. arvalis and My. glareolus with a variable prevalence. Previous studies from European countries reported a slightly higher level of infection with B. burgdorferi (s.l.) in small mammal tissues. The prevalence reported in different countries is concurrent with the patchy distribution of Borrelia spp. For some species our data on prevalence was in line with other European countries, e.g. Croatia [15] and France [16] (from 7 to 7.5%) for My. glareolus.

The patterns of host association of different Borrelia spp. species are reported to differ remarkably [16]. The life-cycle of B. afzelii is dependent on multi-trophic interactions driven by the aggregation of ticks on rodents [17]. Borrelia afzelii has a large range of reservoirs, with small mammals commonly the most important reservoirs [14]. Bank voles and wood mice are considered preferred hosts for larvae of I. ricinus and may promote a high infection with B. afzelii in the resulting nymphs [18]. Moreover, B. afzelii has also been shown as the most common and widely distributed species of the Lyme group in questing [19,20,21,22] and engorged ticks [23,24,25] attached to humans, and in tissues of other vertebrates [26,27,28,29] in Romania. In this study, the majority of the small mammal species were infected with B. afzelii (70%). Similar prevalence rates were found in A. agrarius from Croatia (1.9%) and in A. flavicollis (1.5%). Studies from Hungary, Lithuania and Poland reported a slightly higher prevalence in A. agrarius and A. flavicollis [8, 30,31,32]. However, we consider that the inter-country comparison of prevalence is meaningless in this context, since the conditions of parasitism by ticks were different, as well as the environment and the period of the year for the surveys.

Small rodents have been found to be reservoirs of B. garinii/B. bavariensis and B. valaisiana. Overall, we found relatively low (10%) prevalence of B. garinii/B. bavariensis infection in small mammals. Different subtypes of B. garinii have been shown to be transmitted by birds. The distinct ecotype of B. garinii OspA serotype strains that actually corresponds to B. bavariensis utilizes rodents as reservoir hosts and has been associated with high pathogenicity in humans [33]. Our molecular detection protocol does not differentiate between B. garinii and B. bavariensis, thus further typing and/or multi-locus sequence analyses (MLSA) should be performed to delineate between B. garinii and B. bavariensis [34].

Borrelia miyamotoi is a tick-borne relapsing bacterium transmitted by I. ricinus in Europe. Reports from Hungary indicate the presence of B. miyamotoi in A. flavicollis (with infection prevalence of 4.8% in ticks, 0.3% in skin, 0.5% in the spleen) [8]. The presence of B. miyamotoi has been reported only in Asia (A. argenteus), North America (Peromyscus leucopus) and in Hungary (My. glareolus) [8, 33, 35, 36].

In the present study, shrews (C. leucodon and C. suaveolens) were found to have a higher infection rate (11.4%) in comparison to mice (3%) and voles (4.5%). Experimental studies on the reservoir role of the two insectivore species have not been performed so far. Here, we report the presence of Borrelia spp. in tissues of these species: B. afzelii was detected in the heart (C. leucodon, 1/21; C. suaveolens, 6/49) as well as in the liver (C. suaveolens, 1/48; liver and heart in a single case). As the number of the investigated shrews was relatively low, we are not able to conclude their relative contribution to the enzootic cycle of tick-borne pathogens.

The variety of natural habitats in Romania shapes a wide distribution of small mammals, the country being a natural focus for tick-borne rodent-associated zoonotic pathogens. These data, together with previous studies on the distribution of ticks and the prevalence of Borrelia spp. within them will contribute a general picture of the risk in the country.

Conclusions

The pathogen-vector-reservoir interaction is fundamental to understand the epidemiology and prevent tick-borne diseases. Our analyses demonstrated that voles, mice and shrews are carriers of Borrelia spp. Consequently, the presence of different Borrelia species in the tissues of small mammals is an accurate marker of their circulation. These results are of importance for public health. To our knowledge, this is the first report of Borrelia spp. DNA in C. leucodon and C. suaveolens, and the presence of B. miyamotoi DNA in the liver of My. glareolus. Future studies are necessary to evaluate the contribution of shrews to the spread and maintenance of pathogens.