In Europe, ticks are the most important vectors of human and animal infectious diseases, and transmit more pathogens than any other arthropod (Jongejan and Uilenberg, 2004; Colwell et al., 2011). These diseases are normally maintained in stable natural cycles involving ticks, wildlife, and/or domestic animals, whereas humans are accidental hosts (De La Fuente et al., 2008). Ixodes ricinus is the most widespread and abundant European tick species capable of transmitting several diseases of both medical and veterinary importance (Heyman et al., 2010). Ixodes ricinus has the greatest impact on human public health by transmitting Lyme borreliosis etiological agents, caused by at least four Borrelia genospecies in Europe: Borrelia burgdorferi sensu stricto, Borrelia garinii, Borrelia afzelii, and Borrelia spielmanii. The relapsing fever spirochete, Borrelia miyamotoi, is transmitted by the same Ixodes species and has recently been described in ticks as well as in a human case from the Netherlands (Hovius et al., 2013). In addition to Borrelia transmission, Ixodes ricinus can transmit many other pathogens, including: Anaplasma spp. such as Anaplasma phagocytophilum, Rickettsia spp. from the spotted fever group, Candidatus Neoehrlichia mikurensis, Ehrlichia spp., Bartonella spp., Francisella tularensis, and Coxiella burnetii (Parola and Raoult, 2001; Cotte et al., 2008; Fertner et al., 2012). Ticks can also transmit Babesia genus protozoa, such as Babesia divergens or the newly described Babesia venatorum (sp. EU1) and Theileria spp. (Bishop et al., 2004; Bonnet et al., 2007a).

Increased human travel, animal transport, and environmental changes are responsible for the emergence and/or spread of numerous tick-borne pathogens (TBPs) in Europe (Dantas-Torres et al., 2012). Therefore, effective tick-based surveillance is essential for monitoring human and/or animal disease emergence (Diuk-Wasser et al., 2014). Ticks harbor a variety of pathogens, some of which are obligate intracellular organisms and/or are impossible to artificially culture. Consequently, molecular approaches are thus indispensable for TBP identification. In conventional amplification-based assays, TBP detection occurs for a restricted number of target pathogens known to be transmitted by certain tick species collected at particular sites (Cotte et al., 2010). The main disadvantage of this approach is the limited number of different targets that can be tested, given the quantity of DNA required for one PCR. To improve surveillance of human and animal diseases, new investigative tools are required which perform high-throughput testing of a wider panel of TBPs.

Therefore, the aim of this study was to conduct high-throughput monitoring of tick-borne human and animal pathogens in Europe. Accordingly, we developed a novel high-throughput epidemiological surveillance method to identify both major and neglected European TBPs (bacteria and parasites). This tool utilizes a microfluidic system (BioMark™ dynamic array system, Fluidigm) that is capable of performing parallel real-time PCRs using either 96.96 chips or 48.48 chips resulting in either 9216 or 2304 individual reactions, respectively (Liu et al., 2003). In a single experiment, 94 ticks or pools of ticks can be tested for the presence of 25 bacteria and 12 parasites, as well as confirmation of the tick species. As only a few microliters of sample are required for each test, this system can also be used in conjuction with the typically low-volume DNA extracts prepared from ticks. Then we applied this method to screen 7050 Ixodes ricinus collected from three European countries; France, Denmark, and the Netherlands. We demonstrated increased surveillance efficiency of major and neglected TBPs, and improved monitoring of the emerging diseases important to public and animal health.

In this study, we implemented a method using multiple primers/probe sets able to perform high-throughput detection of TBPs on an unprecedented scale. This large-scale investigation has (i) enabled the detection of rare pathogens such as Bartonella henselae and (ii) generated prevalence estimations for frequent, rare, or unexpected pathogens, thus creating a comprehensive overview of the epidemiological situation for 37 bacteria and parasites present in I. ricinus, in six European sites (two in France, Denmark, and the Netherlands).

Initial testing of the BioMark™ system showed that some pathogens could not be detected. Indeed, assessment was performed on positive DNA controls extracted from cultures, animal blood, ticks, or plasmids, therefore DNA quality and concentration were highly variable between samples. An initial step of pre-amplification was therefore added to specifically amplify targeted pathogen sequences. Regarding the three non-specific assays, two hypotheses can be made: either lack of specificity or potential co-infection of the DNA samples. As positive controls were isolated from pure bacterial cultures, only non-specific cross-reaction explains the lack of specificity. The set of primers and probe designed against B. burgdorferi s.s. cross-reacted with B. garinii and B. valaisiana. However, this cross-reaction did not occur for every field sample. Cross-reactions were also observed between R. conorii and R. slovaca. There was a difference of approximately 10 cycles between the CP values for the expected Rickettsia species and the cross-reacting species. It will be interesting to test both sets of primers and probe on DNA extracted from ticks uniquely infected with each of the Rickettsia. However, this issue is not likely to arise with field samples, as R. slovaca is transmitted by Dermacentor marginatus and R. conorii by Rhipicephalus sanguineus. In conclusion, the primers and probe sets for B. burgdorferi s.s., R. conorii, and R. slovaca need further optimization, so the current results obtained for these species should be interpreted with care. Several of the targeted pathogens cannot be cultured, or are rare and consequently unavailable from field samples, therefore plasmids containing target sequences were used as positive controls. For these pathogens and associated primers/probe sets, further evaluation of specificity is required. This tool was developed for epidemiologic rather than diagnostic purposes, therefore detection limits and sensitivity have not been experimentally determined. These experiments are somewhat difficult to implement and require a gold standard for each pathogen and consistent positive controls, which are not available for all TBPs.

Two sites per country were studied for the field investigation. The technique permitted the detection of 10 bacterial species; B. burgdorferi s.s., B. garinii, B. afzelii, B. valaisiana, B. spielmanii, B. miyamotoi, A. phagocytophilum, Candidatus N. mikurensis, R. helvetica, B. henselae, and two parasitic species; B. divergens, and B. venatorum, with variable prevalences according to the site of collection. Taken together, the estimated prevalences for all pathogens obtained on pools of 25 nymphs in this study are mostly consistent with European published data. For future studies, it will be fascinating to investigate smaller nymph pools to obtain more accurate estimations of TBP prevalences. The prevalence of B. miyamotoi is reported for the first time in Denmark at two sites and is quite similar between the three European countries in our study. Borrelia miyamotoi is transmitted by the same Ixodes species as the etiologic agents of European Lyme borreliosis, and has been detected in Ixodes ticks in Europe (Richter et al., 2003). Up until now no human cases have been reported in France or Denmark, but our data and the recent case of human infection described in the Netherlands (Hovius et al., 2013) suggest that surveillance needs to be improved. Candidatus N. mikurensis was detected in all three countries, with the highest prevalence in the Netherlands. Several human cases have been reported over the past decade in Europe (Maurer et al., 2013). However, clinical symptoms are not pathognomonic, suggesting the existence of unreported cases due to reduced awareness of symptoms by public health professionals (Jahfari et al., 2012). As this emerging human pathogen is widespread in Europe, it requires careful monitoring. Rickettsia helvetica was described as the most prevalent pathogen in all three countries. Even if its pathogenicity remains unclear, R. helvetica has been implicated in the development of fatal perimyocarditis (Sprong et al., 2009). Isolation of the bacterium from a patient is needed to definitely confirm R. helvetica as a human pathogen; however, R. helvetica already represents an excellent candidate for future emergence (Parola, 2004). Over the last few years, I. ricinus has been identified as a competent vector for Bartonella henselae (Cotte et al., 2008). Little data are available on its prevalence in ticks; and it has been estimated at between 11 and 40% in Europe (Dietrich et al., 2010). Bartonella henselae has never been reported in Danish ticks, but two variant types were detected in cats and mice (Engbaek and Lawson, 2004). Its presence in French ticks could be linked to the presence of wild cats in eastern France compared to the other countries. Babesiosis can be a variable but potentially severe disease, and is best known as an animal affliction. However, increasing numbers of human cases have refocused epidemiological attention on this emerging zoonosis (Hildebrandt et al., 2013). Our study demonstrates that these hemoparasites are widely present in European ticks, and were observed for the first time in Danish ticks. Babesia microti was not encountered in this study but has previously been detected in the Netherlands at a prevalence ranging from 0.1 to 9% (Wielinga et al., 2009; Tijsse-Klasen et al., 2011).

Interestingly, among the targeted pathogens, 15 bacterial species and 10 parasitic species were not detected in any country, leading us to conclude that they are not present in I. ricinus from those European sites. Indeed, these TBPs are either very rare (Parola and Raoult, 2001; Sprong et al., 2012b) or have never been previously detected in the sampled regions, or are transmitted by other stage or other tick species (Parola and Raoult, 2001). Francisella tularensis and Coxiella burnetii are linked to important human and veterinary public health problems that require surveillance (Sprong et al., 2012b; Carvalho et al., 2014); however, the role of ticks in the transmission of these pathogens is nonetheless debated. Their apparent absence across the three European countries in I. ricinus ticks suggests that the risk of acquiring tularemia or Q fever from questing ticks could be negligible.

This new screening approach based on microfluidic systems allowing multiple parallel real-time PCRs, is a powerful tool for TBP surveillance in Europe. This study demonstrates the technique's capacity for large-scale studies utilizing the unique ability to simultaneously analyze large numbers of samples and multiple target pathogens. As demonstrated for babesiosis, vector surveillance could be very useful for monitoring disease emergence (Diuk-Wasser et al., 2014). Compared to an array with fixed panels of probes, this new tool presents the major advantage that it can be easily adapted to new situations, as it is entirely possible to add or remove primers/probe sets in order to modify the panel of targeted pathogens and tick species. Further studies will indeed confirm if this approach heralds the necessary breakthrough in epidemiological surveillance of vector-borne pathogens, broadening the monitoring of human and animal diseases.

In conclusion, our study clearly demonstrates the utility of a fast tool that allows comprehensive testing of high numbers of TBPs in ticks, and can be easily customized to fit regional demands or to screen tick or host samples for new or emerging diseases.