Abstract:
Midichloria mitochondrii is a tick-borne intracellular bacterium of the order Rickettsiales, found with high prevalence in the sheep tick (Ixodes ricinus). Midichloria mitochondrii is capable of vertical transmission in the tick, but recently evidence of potential horizontal transmission to the tick hosts through the blood meal has been reported. We investigated the presence of the bacterium in the blood of roe deer (Capreolus capreolus) collected from an area known to be highly infested with I. ricinus ticks. We collected blood and sera samples for 3 yr in Gardouch (Haute Garonne, France) and subjected them to molecular screening through PCR and to serological investigation using enzyme-linked immunosorbent assay and western blot. Bacterial DNA was detected in the blood of four of seven animals, but only at one or two points in time, whereas all sera were positive for M. mitochondrii antigens at all times. Our results indicated that the presence of the bacterium in the blood is transient, but the antibody response appeared to be long-lasting, possibly due to constant exposure to tick bites, and thus to repeated injection of bacteria. The role of M. mitochondrii in the mammalian host, and its interaction with other tick-borne bacteria, remains unknown.
European roe deer (Capreolus capreolus) are very common throughout Europe (Linnel and Zachos 2010). Based on their distribution and habits, roe deer offer a constant and reliable blood supply during the whole year for tick development and are thus regarded as a highly relevant host for ticks and as a reservoir for human tick-borne diseases (Vor et al. 2010; Vázquez et al. 2011). In addition, individual roe deer can transport ticks for long distances during dispersal (Debeffe et al. 2012), seasonal migrations (Cagnacci et al. 2011), and movements to search for mating partners (Richard et al. 2008). The most important tick associated with roe deer is Ixodes ricinus, a species responsible for the transmission of pathogenic bacteria, viruses, and protozoa to humans and animals. In addition to well-known pathogens, recent reports have shown the presence of several tick-associated bacteria in humans, dogs, sheep, sika deer (Cervus nippon), and goats (Parola et al. 2013), with roles that are still to be defined (Duron et al. 2017). Such a bacterium is Midichloria mitochondrii (Sassera et al. 2006), an intracellular bacterium of the recently described Midichloriaceae family of the order Rickettsiales (Montagna et al. 2013). These bacteria are located in the ovaries and salivary glands of I. ricinus (Mariconti et al. 2012; Epis et al. 2013). The latter localization suggests possible horizontal transmission as supported by the lack of cocladogenesis observed between Midichloria sp. bacteria and their tick hosts (Cafiso et al. 2016). Humans and dogs parasitized by I. ricinus are indeed seropositive to M. mitochondrii (Mariconti et al. 2012; Bazzocchi et al. 2013), and DNA of this bacterium was found in the blood of different mammalian species (Skarphédinsson et al. 2005; Bazzocchi et al. 2013). Our aim was to investigate the circulation of M. mitochondrii in roe deer, the host of choice for adult and nymphal stages of I. ricinus, and therefore considered an efficient animal sentinel for possible pathogens transmitted by this tick species (Wodecka et al. 2014).
Blood and sera samples were collected from seven roe deer in a 15 ha enclosure (9 ha woodlot and 6 ha of pasture) at the French National Institute for Agricultural Research experimental station at Gardouch in France (Haute Garonne). Each animal was sampled three times, in April of 2014, 2015, and 2016, with the exception of one deer that died before the third sampling. Ticks from each roe deer were collected, morphologically identified through standard dichotomous keys (Manilla 1998), and confirmed to be I. ricinus. Blood samples were withdrawn from the jugular vein; a portion was centrifuged for serum collection and another was used for DNA extraction. A qualitative PCR on the 12S ribosomal RNA gene of C. capreolus was carried out to check the quality of the DNA extraction (Fajardo et al. 2008). Detection of circulating M. mitochondrii DNA was performed using a nested PCR approach. External primers GYRB-RT-F1 (5′ AAGCTAAGAATTTGGCGTGATG 3′) and GYRBR6 (5′ GTTTTGGCTTCATTTGGATTTTC 3′) were designed on the gyrB gene sequence of M. mitochondrii and used to amplify a fragment of 776 base pairs (annealing temperature 57 C; final primers concentration 1 μM). The resulting products were subjected to a real time PCR using primers gyrBRT-f (5′ CTTGAGAGCAGAACCACCTA 3′) and gyrBRT-r (5′ CAAGCTCTGCCGAAATATCTT 3′) described in Sassera et al. (2008) and amplifying a fragment of 146 base pairs internal to the previous amplified region. The second PCR was carried out using a real-time PCR approach to maximize the sensitivity and specificity of bacterial DNA detection. Blood samples were considered positive for the presence of M. mitochondrii DNA when a clear peak was present, the melting temperature of the reamplified samples corresponded to that of a positive control, and the resulting Sanger sequences showed 100% identity with a fragment of the published gyrB sequence of M. mitochondrii (accession no. AM159536.1; nucleotide position from 565 to 710). In parallel, the recombinant flagellar protein FliD (rFliD) of M. mitochondrii was used to determine anti-rFliD antibody levels in deer sera with an enzyme-linked immunosorbent assay (ELISA). The negative controls used to establish a cut-off value were the sera of 10 red deer from a virtually tick-free area (Merler et al. 1996) and the sera of two roe deer newly-born in an enclosure. The cut-off of the ELISA reaction was established at an optical density (OD) of 0.35 (mean OD on negative controls plus three times standard deviation) determined at a wavelength of 450 nm. All parasitized C. capreolus samples showed an average OD greater than 0.35, whereas the OD of all the samples obtained from tick-free animals were under the cut-off value. The ELISA results were confirmed by western blot (Mariconti et al. 2012).
Three deer were negative to M. mitochondrii DNA, two showed positivity at one of the three sampling times, and the remaining two were positive at two of three times (Table 1). These results confirmed that M. mitochondrii was released with the tick saliva during the blood meal. However, the negative results from the second sampling in some individuals that had been positive in the first one, and specifically the temporal pattern of roe deer 145 (positive–negative–positive), indicated that the presence of M. mitochondrii in roe deer blood was transient, and that this bacterium likely did not multiply heavily in the host blood. The two possible explanations for this finding are the presence of a transient infection or the movement of the bacteria to other body tissues. Interestingly, the pattern of immunization of the roe deer to M. mitochondrii was different compared to the results concerning the presence of the bacterium in blood samples. The ELISA and western blot serological assays showed seropositivity against M. mitochondrii in all samples from all 3 yr (Table 1). This result suggested that roe deer, when continuously parasitized by a large number of ticks, were continuously exposed to ongoing injection of M. mitochondrii. In this context, the presence of M. mitochondrii in tick saliva might play a role in the immune response and immune-modulation, possibly important for infection by tick-borne pathogens.
Results of molecular (nested PCR for the amplification of a fragment of gyrB gene) and serological analyses (enzyme-linked immunosorbent assay and western blot) for the presence of Midichloria mitochondrii performed on roe deer (Capreolus capreolus) blood and sera collected in Spring 2014, 2015, and 2016 in France. Results are expressed as: + = positive sample; − = negative sample; NA = sample not available.

Because roe deer are the host of choice for nymphal and adult stages of I. ricinus, they are a good species for the study of some tick-borne pathogens, but not for all of them. Roe deer play an important role as reservoirs of Anaplasma phagocytophilum, two Bartonella species, Babesia divergens, and Theileria spp., but not for Borrelia spp. (Skotarczak et al. 2008; Chastagner et al. 2014). Deer are considered as dilution hosts for the agent of Lyme disease, because they maintain high tick intensities, which perpetuate tick populations, but they do not support transmission. The transient bloodstream presence of M. mitochondrii observed with our molecular analysis could be due to the function of roe deer as a dead-end transmission host, a situation similar to that of Borrelia spp. (Jaenson and Talleklint 1992; Nelson et al. 2000). Future studies could be focused on the role of roe deer in the circulation of M. mitochondrii, biological interactions of this bacterium with the host or with other bacteria transmitted by ticks (through the immune reaction of the host against bacteria), the possible replication of this bacterium inside the host, and the role of the bacterium in possible pathological alterations after its transmission.
The authors wish to thank the people working in Gardouch involved in roe deer capture and blood sampling and the Climate Change—Emerging Infectious Diseases project (Metaprogramme French National Institute for Agricultural Research—Adaptation of Agriculture and Forests to Climate Change) for providing tick density data. The work was partially supported by research grant 2016-1012 RST–Cariplo Regione Lombardia to Davide Sassera.