Leptospira spp. in Rodents and Shrews in Germany

Leptospirosis is an acute, febrile disease occurring in humans and animals worldwide. Leptospira spp. are usually transmitted through direct or indirect contact with the urine of infected reservoir animals. Among wildlife species, rodents act as the most important reservoir for both human and animal infection. To gain a better understanding of the occurrence and distribution of pathogenic leptospires in rodent and shrew populations in Germany, kidney specimens of 2973 animals from 11 of the 16 federal states were examined by PCR. Rodent species captured included five murine species (family Muridae), six vole species (family Cricetidae) and six shrew species (family Soricidae). The most abundantly trapped animals were representatives of the rodent species Apodemus flavicollis, Clethrionomys glareolus and Microtus agrestis. Leptospiral DNA was amplified in 10% of all animals originating from eight of the 11 federal states. The highest carrier rate was found in Microtus spp. (13%), followed by Apodemus spp. (11%) and Clethrionomys spp. (6%). The most common Leptospira genomospecies determined by duplex PCR was L. kirschneri, followed by L. interrogans and L. borgpetersenii; all identified by single locus sequence typing (SLST). Representatives of the shrew species were also carriers of Leptospira spp. In 20% of Crocidura spp. and 6% of the Sorex spp. leptospiral DNA was detected. Here, only the pathogenic genomospecies L. kirschneri was identified.


Introduction
Leptospirosis is a zoonotic disease of global importance caused by spirochaetes belonging to the genus Leptospira [1]. Leptospires are usually transmitted through contact with the urine of infected animals, either directly or through exposure to contaminated water or soil. The most common portal of entry is through abrasions or cuts in the skin or via the conjunctiva [2]. The spectrum of human disease is variable and can range from subclinical infection to severe signs of multi-organ dysfunction with high case fatality rates. The severity and clinical features of the disease vary according to the leptospiral serovar as well as the age and health status of the patient [3]. To date, 20 Leptospira species with more than 300 serovars, grouped in 20 serogroups, have been described. On the molecular level, Leptospira can be divided into six environmental non-pathogenic species, nine pathogenic species and five intermediate species, for which virulence has not been demonstrated experimentally [4].
Leptospirosis is a notifiable disease in Germany, with the annual human incidence ranging from 0.06 to 0.2 (mean 0.08) per 100,000 between 2001 and 2013 [5]. Historically, leptospirosis in Germany was mainly associated with agricultural exposure risks [6]. As a result of mechanization, improvements in sanitation, and better control in animal reservoirs, epidemic leptospirosis in Germany disappeared in the early 1960s, although sporadic outbreaks associated with agricultural activities still occur [7]. Nowadays, recreational activities linked to freshwater exposures, e.g., swimming in lakes and canoeing, and residential exposures such as gardening and owning pets were identified as major risk factors for disease. Further, travelling abroad to other European and non-European countries was another important exposure risk factor identified for Germany [6]. This is consistent with a similar trend observed in other industrialized countries, especially among participants in adventure travel and water sports in exotic locations [8,9].
Among wildlife species, rodents are the most important maintenance hosts for Leptospira spp. and may transfer infection to livestock, companion animals and humans [3,10]. Maintenance hosts usually have no or mild symptoms after infection and excrete leptospires in the urine constantly and over long periods of time. They form an infection reservoir through continuous cycles of transmission from parents to offspring [11]. In contrast, accidental hosts such as humans can suffer severe acute forms of the disease. Leptospiruria in accidental hosts is intermittent and of short duration [12].
As specific rodent species may be carriers of distinct leptospiral serovars in different geographic areas, the knowledge of the prevalent serovars and the maintenance host association is essential to understand the epidemiology of the disease in a region [3].
In Germany, representatives of several genera within the order Rodentia, which include murine, vole and rat species are found. Further, shrews of different genera within the order Soricomorpha have been described [13]. Little is known about the association of the Leptospira spp. with the majority of these rodent and shrew species in Germany. Our objective therefore, was to gain a better understanding of the occurrence of pathogenic leptospires in rodent and shrew populations in Germany and to identify the most common Leptospira genomospecies. To our knowledge this is the first nationwide study in Central Europe looking at leptospiral infection in such a broad range of rodents and shrews.

Experimental Section
The sample collection was part of a multiannual study to investigate pathogen occurrence and distribution in rodents and other small mammals in Germany [14][15][16][17][18][19][20][21][22]. In the frame of the national network on rodent-borne pathogens [23] between 2002 and 2010 mice, voles and shrews were trapped in 60 different regions (defined by postal codes) located in 11 of the 16 federal states in Germany. The trapping was performed according to the routine protocols established at the different facilities. All animals were frozen and sent to the Friedrich-Loeffler-Institut for further analyses. Here, species, weight, size, and sex were determined and animals were necropsied according to the protocols of the network rodent-borne pathogens [23]. Further, the genus and species of all Leptospira-positive animals was confirmed by a PCR-based cytochrome b analysis [24]. If the genus and/or species could not be identified, the animals were not included in the subsequent analyses. The animals where the geographical origin was unclear were included in the analysis of the association between leptospiral status and rodent species, but not in the geographical presentation.
One kidney of each animal was used during the analysis. Cross contamination between animals was avoided through sequential dissection of single animals and stringent disinfection measures after each dissection. DNA was extracted from 30 mg of kidney tissue with the QIAamp DNA Mini Kit (Qiagen, Hilden, Germany). The DNA samples were distributed among three different laboratories for further analysis. As the laboratories have different routine protocols, these were used during the study.
Laboratory A: One thousand three hundred seventy one kidneys sent to laboratory A were analysed by a duplex PCR adapted from Gravekamp et al. [25]. The PCR reaction was set up with 12.5 µL 2× Qiagen Multiplex PCR Mastermix, 0.5 µL primer (50 pmol/µL) and 2.5 µL template DNA. Thermal cycling was carried out with the 2720 Thermal Cycler (Applied Biosystems, Foster City, CA, USA). An initial denaturation step at 94 °C for 5 min was followed by template denaturation at 94 °C for 30 s, primer annealing at 55 °C for 30 s, a 30 s primer extension at 72 °C for a total of 35 cycles, with a subsequent final extension phase of 5 min at 72 °C.
Laboratory B: Four hundred forty seven samples were screened with a lipl32 PCR according to the protocol by Mayer-Scholl et al. [26]. All positive samples were then subjected to analysis with the duplex PCR used in laboratory A [25].
Laboratory C: One thousand one hundred fifty five samples were tested by lipl32 real-time PCR according to Stoddard et al. [27]. As in laboratory B, the less sensitive duplex PCR was used to further analyse all real-time PCR positive samples. The targets for the duplex PCR are the flagellin-encoding flaB gene (563 bp fragment) which only amplifies in L. kirschneri and the preprotein translocase-encoding secY gene (285 bp fragment), allowing the detection of genomospecies L. interrogans, L. borgpetersenii, L. weilii, L. noguchii, L. santarosai or L. meyeri [25].
Laboratories A, B, C: All samples yielding either the 285 bp fragment or both the 285 bp and 563 bp fragments were subsequently analysed by single locus sequence typing (SLST). The protocol was adapted from Slack et al. [28], substituting the mastermix in the protocol with the 2× Qiagen Multiplex PCR Mastermix. Due to sample size limitations only 93% of these samples could be analyzed by SLST. Samples in which the duplex PCR solely detected the 563 bp fragment were defined as genomospecies L. kirschneri.
To confirm the results of the duplex PCR, a selection of 28 DNA samples was further tested by sequencing of the secY gene [29].

Identification of Leptospira spp. in Trapped Rodents and Shrews
Ten percent (288) of all examined rodents and small mammals were Leptospira-positive in the lipl32 and/or the duplex PCR. Leptospira-positive rodents belonged to the genera Arvicola (25%), Microtus (13%), and Clethrionomys (6%) within the family Cricetidae and the genera Apodemus (11%) and Mus (6%) within the family Muridae. Animals belonging to the shrew family (Soricidae) were also identified as potential vectors for the transmission of Leptospira spp. In 20% of the Crocidura spp. and 6% of the Sorex spp. leptospiral DNA was detected. The regional distribution of Leptospira-positive animals in the different murine, vole and shrew species is shown in Figure 1.
In total, 257 samples were identified as Leptospira-positive by duplex PCR. In 148 (58 %) of the duplex PCR-positive samples the flagellin-encoding flaB gene was amplified, indicating infection with the genomospecies L. kirschneri. All samples further analysed by secY sequencing were confirmed as L. kirschneri.
One hundred and nine samples where the secY gene (285 bp fragment) was amplified, were further analysed by SLST. As the sensitivity of the SLST analysis was much lower than the duplex PCR, only 69% of the samples showed an amplicon that could be sequenced. Of these, 68 samples (91%) were identified as L. interrogans and 7 (9%) as L. borgpetersenii ( Table 2).  Notes: * The genus and/or species of 12 animals (all Leptospira-negative) could not be identified and therefore these animals were not included in the table; ** The column "Leptospira-positive animals" refers to the lipl32 and/or duplex PCR results; *** Identification of genotypes was performed by duplex PCR for L. kirschneri and SLST for L. interrogans and L. borgpetersenii. The different sensitivities of the assays must be taken into account.

Detailed Description of Leptospira Detection in Selected Murine and Vole Species
As 96% of all trapped animals were representatives of the genera Apodemus, Clethrionomys or Microtus all subsequent data analyses focussed on these genera. Eleven percent of the tested animals of the genus Apodemus spp. were carriers of different Leptospira species. Leptospira-positive animals belonged to the species A. flavicollis (59%), A. sylvaticus (22%) and A. agrarius (19%). Leptospira genomospecies causing infection in this genus were L. interrogans (53%), L. kirschneri (46%) and L. borgpetersenii (1%).
A difference in the distribution of trapped male and female animals was not found in the three genera and sex was not associated with detection of leptospiral DNA (data not shown). In our study, the average size of the animals did not differ significantly between infected and non-infected animals within any of the species (data not shown).

Discussion and Conclusions
To date, little is known about the occurrence and distribution of pathogenic leptospires in the most common rodent and shrew species in Germany and other European countries. Most studies have described the occurrence and identity of Leptospira spp. in different rat species. In Germany, Runge et al. [30] found leptospiral DNA in 21% of the examined Norway rats (Rattus norvegicus), and in France, leptospiral renal carriage was shown in 34.7% of the studied Norway rats [31]. In Poland, renal carriage in Norway rats and black rats (Rattus rattus) ranged between 2 and 40%, depending on the region [32].
Our study indicates that, albeit with a lower abundance, pathogenic leptospires are also widely distributed in murine, vole and shrew species throughout Germany. The occurrence of Leptospira spp. in small rodents and shrews in Europe has been investigated in earlier, less extensive studies. In a serological study in southern Germany, 8% of the 266 rodents were seropositive, primarily to the serovars Grippotyphosa and Australis [33]. Europe's largest rodent species, the beaver (Castor fiber), was recently identified as a potential reservoir host for zoonotic leptospires in Germany. L. interrogans multi locus sequence type 24 was identified as the infecting species [34].
In a Swiss study 12.6% of the 190 examined rodents and other small mammals caught in the city of Zurich carried leptospiral DNA [35]; while in a Croatian study 7% of the 227 mice and vole species were Leptospira-positive in a PCR [36]. In a recent study in rodents from Lower Austria, L. kirschneri was detected in three Apodemus mice, but genomospecies L. interrogans, L. borgpetersenii, L. weilii, L. noguchii, L. santarosai or L. meyeri were found in three voles [37].
In our study Microtus spp. were most often carriers of leptospires (13%), followed by representatives of the genus Apodemus (11%). This is in contrast to a study from Croatia where house mice from an urban area showed the highest carrier rate and confirmed its role as a major reservoir in this area [36]. This difference could be due to the much lower number of house mice investigated in our study compared to the other species.
Leptospirosis is usually perceived as a zoonotic disease mainly restricted to rural areas. Yet, in a recent German study, at least 12% of the reported human leptospirosis cases were contracted in urban areas and were primarily related to residential activities, such as gardening and working on private ponds [6]. In two studies in England, Apodemus spp. were nearly exclusively found in residential gardens [38,39]. Also in rural areas, the wood mouse (Apodemus sylvaticus) is often associated with habitats close to humans [40]. As this rodent species and humans share their environment, the wood mouse may be important in the epidemiology of human leptospirosis.
In this study, L. kirschneri was the most prevalent Leptospira genomospecies. L. kirschneri was most common in the vole species Microtus arvalis and the murine species Apodemus agrarius. Previously published data indicate that the serovar Grippothyphosa, which belongs to the genomospecies L. kirschneri, is most often associated with mice of the genus Apodemus in Europe [3]. Interestingly, in our study Apodemus flavicollis and A. sylvaticus were more often infected with L. interrogans.
The public health relevance of the serovar Grippothyphosa is well documented [41]. At the beginning of the previous century, infections with the serovar Grippotyphosa were known as "mudfever" in Germany and associated with a variety of fieldwork activities [42]. In an outbreak of leptospirosis in strawberry harvesters in Germany in 2007, the common vole (M. arvalis) was identified as the most likely reservoir for the infection of the field workers. Leptospires isolated from the animal kidneys were identified by DNA sequencing and by monoclonal antibodies as Leptospira kirschneri serovar Grippotyphosa and serovar Vanderhoedeni [7].
In contrast to another study [43], the average body size, which is used in some rodents as an approximation of age and sexual maturity [44] did not differ significantly between infected and non-infected animals. This could be due to differing rodent species examined in the two studies.
A limitation to the current study is that the serovars of the infecting leptospires were not determined. Serological classification by microagglutination is the gold standard method performed to confirm human leptospirosis in clinical cases. As a multitude of serovars are associated with a single genomospecies [3], it is difficult to infer from the molecular identification of the genomospecies to the serologically detected infecting serovar. This makes the comparison of clinical and epidemiological studies based solely on molecular methods difficult. An additional limitation was that the data origin and quality were heterogeneous, as trapping and laboratory analyses were performed according to the protocols established at the different facilities. The screening for Leptospira spp. in 46% of the kidney samples was performed with the duplex PCR, which is less sensitive than the lipl32 conventional PCR [26]. The analytical sensitivity of the lipl32 conventional PCR for rodent kidney samples is 10 2 leptospires/mL [26]. The sensitivity of the real-time PCR is expected to be comparable, as the analytical sensitivity in spiked blood and urine samples ranged from 10 1 -10 2 leptospires/mL [27]. The specificity of all three assays was 100% [26,27]. Therefore, the proportion of Leptospira-positive animals is most likely underestimated.
In conclusion, this study was based on a very large panel of rodent species, a wide geographical coverage of Germany and a high sample size. Leptospiral DNA was detected not only in different murine and vole species, but also in some shrew species. SLST analysis revealed first insights into a host association of Leptospira genomospecies.
Future studies should aim to investigate the leptospiral status of rodent species by molecular and serological assays in association with their natural habitat, environmental conditions and population dynamics in order to better understand the epidemiology of Leptospira in these animal populations.