House Mice as a Real Sanitary Threat of Human and Animal Leptospirosis: Proposal for Integrated Management

Leptospirosis is a reemerging zoonosis and ranges in severity from benign to sometimes fatal. In cattle, infection may be responsible for abortion and infertility cases causing economic losses. Humans may be contaminated through direct contact with urine of infected animals or indirectly though interaction with urine-contaminated environment. Many wildlife species living close to cattle, especially commensal rodents may play a role in the transmission of leptospires. Because little is known on the epidemiology of nonmaintenance Leptospira serovars, appropriate management is still limited. On a French farm where human and cattle leptospirosis were detected, the transmission cycle was explored to propose appropriate mitigation measures. For that, commensal rodents present on the farm were trapped and their leptospires carriage was studied by molecular methods. Trapped mice were shown to carry two pathogenic Leptospira species (L. interrogans and L. kirschneri). Since these 2 serogroups were simultaneously detected in the trapped mice and in the cows of this farm, we suspected an initial Leptospira transmission from mice to cows requiring an effective management of mice on this farm. Because resistance to anticoagulant rodenticides due to Vkorc1 mutations has been largely described in rodents and first-generation anticoagulant rodenticides seemed to be inefficient in controlling mice on this farm, susceptibility of these mice to anticoagulants has been characterized by Vkorc1 sequencing. 50% of the trapped mice carried mutations in the Vkorc1 gene leading to severe resistance to first-generation anticoagulants. The management of such mice that are a real sanitary threat can be achieved only by using the most toxic second-generation anticoagulants or nonanticoagulant solutions.


Introduction
Leptospirosis is a worldwide zoonosis caused by pathogenic leptospires, bacteria from the spirochete order [1,2]. This disease is reemerging [3] affecting approximately 1 million people all over the world each year [4,5]. Over the last few years, the number of cases has increased in tropical regions [6]. Leptospirosis ranges in severity from benign in most cases to sometimes fatal with a mortality rate of around 10 percent and raises a real public health issue [7]. This disease can provoke a large variety of symptoms. Renal and hepatic failures are observed in human cases [8]. For treatment of human infections, antibiotics as penicillin and doxycycline can be used. In cattle, abortions and other reproductive problems are observed [9] causing economic losses for farmers. The World Organization for Animal Health (OIE) recommends streptomycin for farm animals [10]. Vaccines against some serovars are also available, depending on the country. The main route of contamination is through indirect contact with infected urine from carriers. Main carriers are rodents like rats and coypus but small mammals as mustelids can participate in leptospirosis transmission [11,12]. They accumulate bacteria in kidneys and excrete them by their urine. Excreted leptospires can thus survive in water and manure for several weeks [1,3], which may provide recontamination pathways. Because little is known on the epidemiology of leptospires, appropriate management is still limited. Nevertheless, as many rodents are main carriers of leptospires, efficient rodent management could prevent leptospirosis transmission. Rodent management may associate sanitary and architectural approaches as well as adapted farming practices. If insufficient, these approaches may be completed by the use of chemical rodenticides. Different classes of molecules are used in Europe (i.e., antivitamin K anticoagulants (AVKs), zinc phosphide, alpha-chloralose, etc.) with different usage restrictions. In France, AVKs are the main used rodenticides because of their delayed action avoiding food aversion. AVKs provoke haemorrhage by inhibiting the vitamin K epoxide reductase (VKORC1) involved in the vitamin K cycle limiting the activation of the vitamin K-dependent clotting factors II, VII, IX, and X.
To overcome resistance problems, a second-generation of AVKs including bromadiolone, difenacoum, brodifacoum, flocoumafen, and difethialone has been developed. This second-generation of AVKs is more toxic, more tissuepersistent, and thus more often associated with primary or secondary poisonings.
After the demonstration of animal and human leptospirosis cases in a cattle farm located in the Auvergne-Rhône-Alpes region of France, our study aims to identify the most likely transmission scheme(s) on this farm and the methods to be used to prevent any new cases.

Study Area.
This study was conducted on a farm in the Auvergne-Rhône-Alpes region of France, in the Loire department. This farm was located in lowland surrounded by hills, served by a single road and divided into different agricultural buildings (two stalls, one animal feed storage shed and one vehicle shed) ( Figure 1). Two farmers were working on this farm to raise cattle for meat production. The herd was composed of 300 head of Charolais cattle with 130 adult cows. Animals were in the meadow and exclusively fed with grass in summer, and were in stabling and fed with hay and silage in winter. First-generation anticoagulant rodenticides were used by the farmer to control rodent infestation into agricultural buildings.

Rodent Trapping Method.
Mice were trapped using 40 mouse traps (Aegis Trap Souris, De Sangosse, France), set up where traces of rodents had been observed ( Figure 1) in the animal feed storage shed (referred to as site 1) and in the main stall (referred to as site 2). No rats and no trace of rats were observed in the different farm buildings. Baiting was done using peanut butter daily renewed during the capture period. Trapping was conducted during 24 consecutive days from November 17th, 2015, to December 10th, 2015. The trap effort (i.e., number of traps per site × number of days of trapping) was 480 trap-days in sites 1 and 2, respectively.
During the capture period, cows were mainly housed in the barn. Trapped animals were weighted, sexed, and assigned to one of two age categories: immature house mice (< 12g) and mature house mice (> 12g). Dead mice were autopsied in the laboratory. Kidneys were taken under sterile conditions. One part of the kidney was used immediately for culture, the other part (i.e., corticomedullary junction, 5 mg) was frozen at -20 ∘ C for the molecular detection and characterization of leptospires. A piece of tail of each rodent was stored in 70% ethanol as material for DNA extraction for Vkorc1 genotyping.

Rodent Species Identification.
Trapped animals were molecularly confirmed to be M. musculus domesticus (as opposed to a morphologically similar species) by sequencing a portion of the mtDNA cytochrome b gene. Briefly, genomic DNA was extracted from liver using silica columns (Macherey Nagel) according to the manufacturer's recommendations. Two microliters of genomic DNA were amplified by PCR using specific primers of cytochrome b [19]. The amplified product was sequenced on both strands; the resulting sequence was submitted to blast analysis (http:// blast.ncbi.nlm.nih.gov).

Serological Tests Using the Microscopic Agglutination Test (MAT).
MAT analyses were performed in the Laboratoire des Leptospires (VetAgro Sup, Marcy-l'Etoile, France) as described by [20] using a total of 23 Leptospira strains belonging to 14 serogroups. Blood samples were collected from seventy-six cows presenting reproductive disorders and centrifuged, and the serum obtained was stored at −20 ∘ C until MAT analyses. The screening was performed starting with a serum dilution of 1: 10 up to a dilution of 1: 800. The endpoint was the highest serum dilution showing 50% agglutination in free-moving leptospires. Positivity to a serovar was concluded when a MAT reaction was observed towards this unique serovar or when this serovar displayed a titer three times higher than the other serovars when cross reactions were observed. When cross-reactivities were observed between several serovars with similar titers, positivity was established towards cross-reacting serovars.

Leptospira Isolation.
Kidney from each mouse and 3 placental cotyledons from aborted cows were crushed and aseptically transferred to tubes containing Ellinghausen McCullough Johnson and Harris (EMJH) medium (Indicia, St Génis, France) [20]. Three serial dilution tubes were incubated at 30 ∘ C according to the protocol for pathogenic Leptospira isolation [20,21]. For a period of three months, the tubes were examined weekly using a dark-field microscope

16S Amplification for Leptospira Detection and Species
Identification. DNA extractions of mice kidneys were performed with QIAamp5 DNA Mini Kit, Qiagen, according to manufacturer's guidelines. Amplification of a part of 16S rDNA was performed to detect Leptospira and identify species as previously described [21,22] using LeptA primer 5'-GGCGGCGCGTCTTAAACATG-3' and LeptB primer 5'-TTCCCCCATTGAGCAAGATT-3' . PCR mix without the target DNA was included as a negative control, and PCR mix with a Leptospira strain was included as a positive control. PCR products were then sequenced by Genoscreen enterprise by a Sanger procedure. The resulting sequence was submitted to blast analysis ((http://blast.ncbi.nlm.nih.gov).

Multisequence Typing (MST) Analyses.
Positive samples for L. interrogans species were subjected to MST analyses allowing the sequencing of 3 intergenic regions, the MST1, MST3, and MST9 regions [23]. PCR mix without the target DNA was included as a negative control, and PCR mix with a Leptospira strain was included as a positive control. PCR was negative for samples tested; therefore sequencing was not necessary.

Vkorc1
Genotyping. Genomic DNA extractions from mice tails were performed with QIAamp5 DNA Mini Kit, Qiagen, according to manufacturer's guideline. Vkorc1 gene was amplified as previously described [25]. The amplified products were sequenced on both strands. Obtained sequences were subsequently compared with the published Vkorc1 gene sequences for Mus musculus domesticus (Genbank n ∘ GQ905710.1).

Ethics Statement.
This project was run in the framework of a rodent control program decided and organized by the farmer. Trapping was conducted by a private pest control firm during an eradication program conducted on the farm. Commercial mouse traps killing instantly mice were used. Traps were placed by the pest control operator where rodent tracks were detected. Daily control of the traps was done by the farmer. When the presence of rodent in the trap was noted by the farmer, the latter placed it at 4 ∘ C until its dissection the next day. This study was not considered to be an "experimental procedure" as defined by the French legislation (Rural Code, Article R214-89) and therefore was not subjected to an ethical committee approval in France. This study complied with the ethical standards of European regulations governing the care and use of animals in research and it did not involve any endangered or protected species, or protected areas.

Rodent Infestation on Farm.
A total of 12 house mice (Mus musculus domesticus) were caught, all in site 2 (about 2 every 2 days), and referred to as S1 to S12, including 100% of mature animals and 50% of males as shown in

Leptospiral Carriage in Cows.
Seventy-six cows presenting reproductive disorders and three aborted cows were subjected to serological analysis by MAT test. Reactivity towards the different tested serogroups with titers between 10 and 640 is presented in Table 2.
Among cows with reproductive disorders, forty-three were seropositive to a single serotype, either L. Sejroe (serovar Hardjo) or L. Grippotyphosa (Table 3) and twenty-three Total 43 23 66 * Positivity is ascertained when a MAT reaction is observed solely to this serovar or when this serovar displays a titer three times higher than the other serovars when cross reactions is observed. * * Cross reactions in which the previous parameters are not observed (in all these reactions, MAT titer against a member of the Sejroe serogroup was common). AUS: australis, GRIPP: grippotyphosa, IH: icterohaemorrhagiae, SJ: sejroe.
Out of the 3 aborted cows, 2 and 1 were demonstrated by MAT analysis to be infected by L. Sejroe and L. Grippotyphosa, respectively (Table 2). Unfortunately, all the attempts of cultures and PCR detection on aborted materials were unsuccessful.

Leptospiral Carriages in Trapped
Mice. Based on 16S amplification, 7 mice presented positive results for Leptospira presence. Three Leptospira genospecies were identified, including a saprophytic leptospire. L. biflexa was found in 2 mice, L. kirschneri in 3 mice, and L. interrogans in 1 mouse. One mouse was found carrying the two leptospires species L. interrogans and L. kirschneri ( Table 1). The prevalence of pathogen leptospiral renal carriage in trapped mice was 42%.
From the L. interrogans DNA sample extracted, MST and VNTR profiles could not be obtained because of unspecific amplifications. From the 4 L. kirschneri DNA samples extracted, complete VNTR profiles were obtained in 2 samples and corresponded to Leptospira belonging to serogroup Icterohaemorrhagiae as shown in Table 1 according to VNTR published database [24]. For the two other samples, VNTR profiles were incomplete because VNTR10 analyses could not be interpreted because of the presence of unspecific bands.
Four different Vkorc1 genotypes were observed in the trapped mice. One genotype with no Vkorc1 mutation was detected in 6 mice. One genotype with only 1 Vkorc1 mutation (g.2223A>G) leading to the Y139C protein mutation was detected in 3 mice. One genotype combining 2 missense mutations (g.34C>T and g.76G>T) leading to the R12W and A26S protein mutations and one silent mutation (g.111A>G) was found in only one mouse at the heterozygous state. The last genotype combining 4 missense mutations (g.34C>T, g.76G>T, g.142G>A, and g.976G>T) leading, respectively, to the R12W, A26S, A48T, and R61L protein mutations and one silent mutation (g.111A>G) was found in two mice, one at the homozygous state and one at the heterozygous state.

Epidemiology of Nonmaintenance Leptospira Serovars on This
Farm. The high rate of abortion observed in the herd, whereas in previous years nothing abnormal had been reported, the detection of antibodies against leptospires in 66 cows out of 76 tested, and the fact that the farmer contracted leptospirosis at the same time allowed us to suspect the recent establishment of a leptospires transmission cycle on this farm.
Three serogroups, L. Sejroe (serovar Hardjo), L. Grippotyphosa, and to a lesser extent L. Icterohaemorrhagiae, were found in cows by MAT analyses. The three aborted cows were found to be infected either by L. Sejroe for two cows or by L. Grippotyphosa for the other one. These results suggest that both serogroups are responsible for abortions observed in this herd. Since cows are known to be natural hosts for serovar Hardjo [9,26] and the spirochete can survive in cow for years, it seems unlikely that the high rate of abortion and reproductive disorders observed at the time of the study was caused by the serovar Hardjo. Indeed, we can assume that in the absence of new animals entering the farm, the cows of this farm were already carriers of L. Sejroe (serovar L. Hardjo) in the previous years, while the reproductions proceeded normally. It is more likely that the high rate of abortions and the reproductive disorders suddenly observed was due to recent infection by the other serogroups L. Grippotyphosa and L. Icterohaemorrhagiae that are nonmaintenance Leptospira serovars in cows. Many wildlife species living in the cattle habitats may be suspected to have a role in the initiation of Leptospira transmission cycle. Nevertheless, since rodents are the natural carriers of these serogroups [27,28], it was normal to immediately consider the role of commensal rodents in this initiation even if rodent control operations have been regularly conducted by the farmer. The presence of rats has not been highlighted in the farm buildings despite a careful inspection, which seems to show the success of the solutions applied by the farmer to control rat infestations. On the contrary, trace of mice and mice themselves have been seen in two buildings of the farm, suggesting a failure of the solutions applied to control mice. Only 12 mice could be captured in site 2, while numerous mice and numerous traces of mice could be observed in sites 1 and 2. The strong food competition in the buildings has undoubtedly limited the success of trapping, preventing any conclusion on the size of the mice populations present on this farm. Nevertheless, our study highlights the presence of mice carrying two pathogenic Leptospira species (L. interrogans and L. kirschneri) (5/12 trapped mice) in the main stall of the farm. Within these 2 Leptospira species, 3 serovars were identified, one completely identified (i.e., Icterohaemorrhagiae) and the other two partially identified. We assume that the DNA of the L. kirschneri, partially typed by VNTR method, belongs to Grippotyphosa serogroup, the major representative of the L. kirschneri species in France, for which VNTR analysis has shown a weak ability to achieve its typing [29].
Since these 2 serogroups (Grippotyphosa and Icterohaemorrhagiae) were simultaneously detected in mice and cows, we could suspect an initial Leptospira transmission from mice to cows and then, hypothetically, from cows to the farmer who developed the disease at the period of cow abortions (Figure 2). Because mice share the same habitat as cows on this farm, initial transmission from mice to cows was most likely indirect through interaction with a urine-contaminated environment such as hay and grains. Indeed, contact between rodents and food in dairy cattle has been often described as a factor risk in a prevalence of Leptospira infection study [9,29]. After the initial transmission from mice to cows, we can hypothesize that the farmer acquired leptospirosis from cow through direct contact with the urine of infected animals or indirectly with a urine-contaminated environment such as manure. Even if the serogroup of Leptospira responsible for leptospirosis of the farmer is unknown, we can easily assume that the serogroup Grippotyphosa was also responsible for the leptospirosis of the farmer. In fact, in a recent epidemiologic study on human leptospirosis performed in the Franche-Comté department, located approximately 300 km at the North East of the farm, this serogroup was found to be the most often involved in human disease [8].
This farm, in which animal leptospirosis is concomitant to human leptospirosis, is clearly a public health concern. Indeed, leptospirosis is reported as fatal in about 10% of infected humans, with a possible fatality range of 5-40%. The context of this farm clearly shows that the importance of these 2 nonmaintenance serovars in bovine leptospirosis should not be underestimated in epidemiological survey studies. Radical measures must be implemented within the farm to stop the cycle. Different possible management points will be discussed above to limit and prevent Leptospira transmission.

Management Implications. Management of mice in buildings: Until now, scientific studies have mainly shown that
Leptospira is frequently carried in rats and studies reporting carriage of Leptospira in mice are few [30]. This could explain why the farmer was not worried about the presence of mice in these buildings. Because leptospiral-infected mice share the same habitat as cows on this farm, they represent a real threat even if infestation was limited. The first action to be implemented is to effectively manage rodents in livestock buildings, but also in food and materials storages. The farmer had applied first-generation anticoagulants. Clearly this method has been effective only in controlling rats. The genetic analysis carried out on the trapped mice enabled us to understand the failure of this treatment. In fact, 50% of the trapped mice carried mutations in the Vkorc1 gene. Mutations of this gene have been widely studied in Europe and France because a number of them have been shown to be responsible for resistance to AVKs which are the most frequently used molecules in Europe to manage rodents. The mutations found on this farm have already been described in Europe and France [31,32]. In the South East of France, its frequency has been reported to be close to 20% [32] in mice trapped from 21 different sites. It is therefore not surprising to find this mutation in this site. The set of mutations (R12W, A26S, A48T, R61L) detected on this farm has also been previously described in Germany [31,33] and then in France [25,32]. This set of 3 or 4 mutations has been reported to be the consequence of the introgression of Vkorc1 gene from the Algerian mouse (Mus spretus) in the genome of Mus musculus domesticus. The Y139C mutation and the set of 3 or 4 mutations have been shown to be responsible for severe resistance to anti-vitamin K anticoagulants, particularly to first-generation molecules. The failure of the treatments carried out by the farmer is therefore understandable, especially since this resistance is compounded by the presence in the buildings of easily accessible food stocks and certainly responsible for a low consumption of baits by mice. The management of such mice can be achieved only by using secondgeneration anticoagulants such as difenacoum, difethialone, brodifacoum, or flocoumafen. These anticoagulants are more effective in such strains, but they are more persistent and more often associated with secondary poisoning of wildlife due to the ingestion of intoxicated rodents. Their use must therefore be limited to the area around a building and it is better to use them in baiting boxes.
Limiting access to incoming mice: The presence of several leptospiral strains and several mutations of the Vkorc1 gene within the same farm is quite surprising. Indeed, generally within the same site and the same rodent family, a single leptospiral strain or a unique Vkorc1 gene mutation is found. For example, a recent study conducted in the city of Lyon (France) showed the presence of a unique strain of leptospires in rats caught in different sites of the same city [22]. Similarly, a study conducted in a Paris park of approximately 80 ha reported the presence of a unique Vkorc1 gene mutation in a sample of more than 80 rats trapped at different park locations [34]. On this farm, 2 species and at least 3 leptospires serovars were detected with individuals potentially carrying 2 species of leptospires, as well as 4 different Vkorc1 genotypes. Two hypotheses are possible: (1) Within the farm several micropopulations of mice forming a metapopulation could be present, nesting in different places of the farm/building and crossing very rarely, possibly in places where they eat, leading potentially to a transfer of pathogens or even resistance alleles; a suitable management as explained previously must be done by considering all the potential sites of breeding and feeding. (2) Within the farm a single micropopulation is present, and the diversity in pathogens and Vkorc1 genotypes is due to new entrants. Transportation of calves, slaughter, and feedstuffs certainly fosters the exchange of mice by transportation, which can result in the variety of genotypes and L. species/serovars in one location. New entrants introduced by transport or from the surrounding area could exchange genes and pathogens with local individuals when animals meet; control measures should be combined with architectural and sanitary measures to prevent the introduction of new entrants or the dispersion of outward bounds. Indeed, the presence of rodents carrying pathogenic leptospires in the farm poses a risk for all stakeholders exchanging products with this farm. Nevertheless, on a farm, control measures are extremely difficult or impossible to implement. Moreover, the plurality of Vkorc1 gene mutations found in a homozygous state on the same site could favor recombination between genotypes that could generate resistance to the first-generation AVKs, but also to the second-generation AVKs as already described by [32].
Prophylaxis measures: On this farm, to reduce the risk of pathogenic leptospires contamination, prophylaxis measures are also important in order to reduce farm and domestic animals and human contamination, but also contamination of new entrant mice. Containment of seeds to prevent them to be contaminated by infected mice urine and disinfection of equipment and surfaces which could have been in contact with the pathogen must be done as far as possible. If possible calving boxes and quarantine zones must be set up for cattle, before spreading manure in pasture, a storage period higher than 8 days should limit risk of contamination because of the short survival period of leptospires in manure [35].
Human vaccine could be preconised against serovar Icterohaemorrhagiae. The infecting strain of the farmer could not be characterized, but symptoms were febrile illness associated with mild hepatitis and this form is usually known to be caused by Grippotyphosa, a serovar considered to be less severe than clinical forms caused by Icterohaemorrhagiae. In France, only vaccine against serovar Hardjo is available for bovine. In this case, using this vaccine solely would not be efficient unless it is associated with Grippotyphosa antigen as the vaccines available in the US. Our study points out that Hardjo is not the only serovar responsible for bovine abortion and the vaccine to be used should therefore associate antigens other than the only Hardjo antigen which is currently available for controlling bovine leptospirosis in France.

Conclusions
Reproductive disorders and abortions suddenly observed on this farm could be associated with leptospirosis. Since L. Sejroe (serovar Hardjo), L. Grippotyphosa, and L. Icterohaemorrhagiae have been detected in cows, clinical symptoms may be due to one or more of these serogroups and appropriate measures should be taken. The presence of L. Sejroe (serovar Hardjo) would justify the vaccination of the herd against this serovar. However, this vaccination will not prevent contaminations of cows by L. Grippotyphosa or L. Icterohaemorrhagiae carried by mice that could also contribute to clinical signs observed on this farm.
The fact that several mice present on this farm carried L. Grippotyphosa or L. Icterohaemorrhagiae demonstrates the relevance of mice control on this farm to mitigate the risk of Leptospira transmission from wildlife to livestock and/or humans even if the number of mice captured was limited. Nevertheless, management of such mice carrying zoonotic agent is challenging because of their severe resistance to AVK rodenticides. The low sampling of captured mice did not allow us to compare the prevalence of leptospiral carriage between resistant mice and nonresistant mice. A more complete study should be conducted to evaluate a potential link between resistance and carriage of leptospires. A biological cost of resistance could possibly lead to a more frequent contamination of resistant mice. Once the transmission cycle is established in livestock, rodent control must be combined with appropriate environmental and medical prophylactic measures. It would have been interesting to be able to repeat MAT analysis after the implementation of all the recommended measures to know the evolution of the situation, but unfortunately this was not possible.
Such surveillance studies are essential to identify circulating serovars that are essential for the development of appropriate vaccines. Because of the small number of trapped and analyzed mice, this study may be considered as a preliminary report on Leptospira carriage by house mice. More studies will be needed to assess Leptospira prevalence in mice.

Data Availability
All the data used to support the findings of this study are included within the article.

Disclosure
The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.