Spatial patterns of Hyalomma marginatum-borne pathogens in the Occitanie region (France), a focus on the intriguing dynamics of Rickettsia aeschlimannii

ABSTRACT Hyalomma marginatum is an invasive tick species recently established in mainland southern France. This tick is known to host a diverse range of human and animal pathogens. While information about the dynamics of these pathogens is crucial to assess disease risk and develop effective monitoring strategies, few data on the spatial dynamics of these pathogens are currently available. We collected ticks in 27 sites in the Occitanie region to characterize spatial patterns of H. marginatum-borne pathogens. Several pathogens have been detected: Theileria equi (9.2%), Theileria orientalis (0.2%), Anaplasma phagocytophilum (1.6%), Anaplasma marginale (0.8%), and Rickettsia aeschlimannii (87.3%). Interestingly, we found a spatial clustered distribution for the pathogen R. aeschlimannii between two geographically isolated areas with infection rates and bacterial loads significantly lower in Hérault/Gard departments (infection rate 78.6% in average) compared to Aude/Pyrénées-Orientales departments (infection rate 92.3% in average). At a smaller scale, R. aeschlimannii infection rates varied from one site to another, ranging from 29% to 100%. Overall, such high infection rates (87.3% on average) and the effective maternal transmission of R. aeschlimannii might suggest a role as a tick symbiont in H. marginatum. Further studies are thus needed to understand both the status and the role of R. aeschlimannii in H. marginatum ticks. IMPORTANCE Ticks are obligatory hematophagous arthropods that transmit pathogens of medical and veterinary importance. Pathogen infections cause serious health issues in humans and considerable economic loss in domestic animals. Information about the presence of pathogens in ticks and their dynamics is crucial to assess disease risk for public and animal health. Analyzing tick-borne pathogens in ticks collected in 27 sites in the Occitanie region, our results highlight clear spatial patterns in the Hyalomma marginatum-borne pathogen distribution and strengthen the postulate that it is essential to develop effective monitoring strategies and consider the spatial scale to better characterize the circulation of tick-borne pathogens.


INTRODUCTION
Zoonoses are responsible for 60% of emergent diseases (WHO, 2023).Ticks play an important role in the spread of zoonotic diseases and are considered the primary vectors of pathogens in Europe (Boulanger et al. 2019).In France, there are about 45 tick species of interest for public and veterinary health, including Ixodes spp., the main vector of pathogens responsible for the Lyme disease and anaplasmosis (Pérez-eid 2007).
H. marginatum is a tick species that has recently become established in mainland southern France (Vial et al. 2016) although it has been established for decades on the French Mediterranean island of Corsica (Delpy 1946;Morel 1959;Grech-Angelini et al. 2020).H. marginatum is currently endemic in several countries from the Maghreb to the Iberian Peninsula and the eastern Mediterranean basin, including Turkey and the Balkans (ECDC 2023).With the increase of temperatures due to climate change, this tick species may become established in northern latitudes via animal movements and/or bird migrations (Gray et al. 2009;Vial et al. 2016;Fernández-Ruiz and Estrada-Peña 2021).It appears that H. marginatum become established in regions with specific climate features such as warm temperatures in summer and low precipitation which are typical of the Mediterranean climate (Bah et al. 2022;Estrada-Peña, Martínez Avilés, and Muñoz Reoyo 2011).In mainland France, H. marginatum establishment has been reported in several departments including Pyrénées-Orientales, Aude, Hérault, Gard, Var, Ardèche, Drôme (Bah et al. 2022).Since an invasion process seems to happen in the south of France, its adaptation abilities and its current expansion area in France are being closely monitored.
As many other tick species, H. marginatum harbours complex microbial communities, collectively known as the microbiota encompassing symbionts, commensals, environmental microbes and on the other hand, pathogens affecting vertebrate hosts.Among the pathogens, H. marginatum can carry a large diversity of bacteria, viruses and parasites, in Eurasia and Africa (Bonnet et al. 2023).In Afrotropical regions, Mediterranean basin and more particularly in France, H. marginatum is considered as the main candidate for the transmission of the deadly Crimean-Congo hemorrhagic fever virus (CCHFV) (Bernard et al. 2022;Perveen and Khan 2022).Besides, CCHFV was detected for the first time in France in ticks from this study and in ticks collected in 2023 (Bernard et al. 2024).
The risk represented by this tick requires an exhaustive identification of pathogens it carries and the characterization of their dynamics in both space and time.Based on previous studies, H. marginatum distribution in the Occitanie region (NUTS-1) seems to be clustered between two geographic areas in Gard and northern Hérault departments and on the other hand Pyrénées-Orientales and south Aude departments (Bah et al. 2022).No H. marginatum was found so far between these two geographic clusters for unclear reasons.In this context, we hypothesized that the spatial distribution of microbial communities particularly pathogens might vary depending on the geographic cluster.It is now accepted that spatial patterns at different scales impact tick distribution, density and their associated pathobiome, due to the of environmental characteristics such as abiotic factors (temperature, landscape, vegetation) and the presence of hosts in given areas forming specific environmental niches (Pollet et al. 2020;Bonnet and Pollet 2021).The small spatial scale is important to consider as the pathogen prevalence can vary among a given biotope, probably due to specific environmental factors (Sormunen et al. 2018).At a larger spatial scale, the type of habitat can also participate to the prevalence of pathogens as it was the case for A. phagocytophilum and Rickettsia sp. of the spotted fever group infections in Ixodes ricinus collected in Central France from two different habitats (pasture vs woodland) (Halos et al. 2010).Another study reported differences in Borrelia burgdorferi prevalences between urban/suburban habitats and natural/agricultural in Slovakia (Kazimírová et al. 2023).
While a previous study identified H. marginatum-borne pathogens in several sites of collection between 2016 and 2019 (Bernard et al. 2024), we propose in this study a focus on the spatial distribution of H. marginatum-borne pathogens with a focus on its clustered distribution in Occitanie, and by taking into account the influence of others factors such as the tick sex and the engorgement status (Sperling et al. 2020;Ahantarig et al. 2013;Dergousoff and Chilton 2012;Gern, Zhu, and Aeschlimann 1990).
To this purpose, a large-scale tick collection program was performed in the Occitanie region in May 2022 and resulted in the analysis of 510 ticks found in 27 sites across four departments.We characterized the influence of spatial patterns, tick sex and engorgement status on tick-borne pathogens infection rates and loads.

Study area and tick collection
A large-scale tick sampling was conducted during two weeks in May 2022.From the Stachurski and Vial (2018) study, we identified and visited 42 sites (riding schools and farms) from six French departments of the Occitanie region (Figure 1).Ticks were morphologically identified using a binocular loupe, sorted by sex (male/female) and engorgement status (fed, semi-engorged and unfed) and then stored at -80°C until further use.After the identification, a total of 510 H. marginatum ticks were analysed.They came from four departments of the Occitanie region: Hérault, Gard, Aude and Pyrénées-Orientales whereas it was not detected in Tarn and Haute-Garonne departments (Figure 1).We chose a maximum of 34 ticks per site, with an average of 18.8 ticks per site.When the number of ticks was lower than 34, all ticks were analyzed.When possible, an equal proportion of males and females were selected for each site.

DNA and RNA extraction
Tick crushing and nucleic acids extraction were done in a BSL3 facility.Ticks were washed in bleach for 30 seconds then rinsed three times for one minute in milli Q water to limit eliminate the environmental microbes present on the tick cuticle (Binetruy et al. 2019).Ticks were then cut using a scalpel blade and crushed individually in the homogenizer Precellys®24 Dual (Bertin, France) at 5,500 rpm for 40 sec, using three steel beads (2.8 mm, OZYME, France) in 400 µL of DMEM (Dulbecco's Modified Eagle Medium, Eurobio Scientific France) with 10% foetal calf serum.Total DNA and RNA was extracted using the NucleoMag VET extraction kit (Macherey-Nagel, Germany) as described by the manufacturer's instructions with the IDEAL TM 96 extraction robot (Innovative Diagnostics, France).Nucleic acids were eluted in 90 µL of elution buffer and stored at -20°C for DNA and -80°C for RNA until further analyses.

Tick-borne pathogens detection in tick DNA and RNA Microfluidic PCR detection Reverse transcription
Sample were retrotranscribed in cDNA using 1 µL of RNA with 1 µL of Reverse Transcription Master Mix and 3 µL of RNase-free ultrapure water provided with the kit (Standard Biotools, USA) using a thermal cycler (Eppendorf, Germany) with the following cycles: 5 min at 25°C, 30 min at 42°C and 5min at 85°C with a final hold at 10°C.

Targeted tick-borne pathogens
The tick-borne pathogens with a high probability to be carried by H. marginatum were targeted using 48 sets of primers (designs), according to data mining of the available literature (Bernard et al. 2024); Michelet et al. 2014;Gondard et al. 2018) (Table 1).On the 48 sets of designs, 14, 5 and 27 targeted bacteria, parasites and viruses respectively.Two designs were used for positive controls.The 48 designs were pooled to reach a 200 nM final concentration for each primer.CCHFV detection on samples was performed apart from this study (Bernard et al. 2024).

Pre-amplification
Each sample was pre-amplified using 1.25 µL of DNA mix (1:1 volume ratio of DNA and cDNA) with 1 µL of the PreAmp Master mix (Standard Biotools, USA), 1.5 µL of ultra-pure water and 1.25 µL of the pooled designs.PCRs were performed using a thermal cycler with the following cycles: 2 min at 95°C, 14 cycles at 95°C for 15 sec and 60°C for 4 min and finally a 4-min hold at 4°C.A negative control was used for each plate with ultra-pure water.Amplicons were diluted 1:10 with ultra-pure water and stored at -20°C until further use.

BioMark TM assay
The BioMark TM real-time PCR system (Standards Biotools, USA) was used for high-throughput microfluidic real-time PCR amplification using the 48.48 dynamic array (Standard Biotools, USA).The chips dispense 48 PCR mixes and 48 samples into individual wells, after which on-chip microfluidics assemble PCR reactions in individual chambers prior to thermal cycling resulting in 2,304 individual reactions.In one single experiment, 47 ticks and one negative control are being tested.For more details, please see (Michelet et al. 2014;Gondard et al. 2018a).

Validation of the results by PCR and sequencing
Conventional PCRs or qPCR were then performed for tick-borne pathogens positive-samples using different sets of primers than those used in the BioMark TM assay to confirm the presence of pathogenic DNA (Table S1).Amplicons were sequenced by Azenta Life Sciences (Germany) using Sanger-EZ sequencing and assembled using the Geneious software (Biomatters, New-Zealand).An online BLAST (National Center for Biotechnology Information) was done to compare results with published sequences in GenBank sequence databases.

Detection and quantification of T. equi and R. aeschlimannii by duplex Real-Time Fluorescence
Quantitative PCR Tick samples were also screened for the detection and quantification of T. equi and R. aeschlimannii using a second detection method by qPCR with primers and probes targeting different genes to those used in the BioMark TM assay (Table 2) (Kim et al. 2008;Rocafort-Ferrer et al. 2022).There are five known genotypes for T. equi (designated A-E) circulating in Europe, so we wanted to be sure that we could detect all of those five genotypes if present (Nagore et al. 2004;Bhoora et al. 2009;Salim et al. 2010;Qablan et al. 2013).The Takyon™ No ROX Probe 2X MasterMix Blue dTTP (Eurogentec, Belgium) was used with a final reaction volume of 20 µL containing 10 µL of Master Mix 2X (final concentration 1X), 5 µL of RNase free water, 1 µL of each primer (0.5 µM), probes (0.25 µM), and 2 µL of DNA template.The reaction was carried out using a thermal cycler according to the following cycles: 3 min at 95°C, 45 cycles at 95°C for 10 sec, 55°C for 30 sec and 72°C for 30 sec.Positive controls for both R. aeschlimannii and T. equi were prepared using a recombinant plasmid from the TA cloning® kit (Invitrogen, USA).A 10-fold serial dilution of the plasmid (from an initial concentration of 0.5x10 8 copy number/μL) was used to generate standard positive plasmids from 2.5x10 5 copy number/μL to 2.5x10 - 1 copy number/μL.Samples were detected in duplicates and quantified using the standard plasmids.
For R. aeschlimannii, we considered negative samples whose Cq number was higher than Cq 37.This detection limit was established regarding the last dilution of the standard curve that could be detected by qPCR.For T. equi, most samples were close to or below the detection limit established with the T. equi standard curve.Because the protozoan T. equi may be circulating at low levels, we included all positive samples.

Statistical analyses
Statistical analyses were performed with R software 4.2.0.Generalized linear mixed effect models (glmer, package lme4, (Bates et al. 2015)) were used to evaluate the effect of variables on the presence/absence (binomial distribution) of R. aeschlimannii, A. phagocytophilum, Francisella-LE (BioMark TM data) and T. equi (qPCR data).In addition, R. aeschlimannii and T. equi loads were analysed using a gamma distribution.Because a very few ticks were positive for A. marginale, only descriptive analyses were presented.We decided to include Francisella-LE in the analyses as a control, since this bacterium is definitely identified as a H. marginatum primary endosymbiont.
A first model was used on the whole dataset (n=510) to assess the influence of the geographic cluster (Aude/Pyrénées-Orientales) and Hérault/Gard), the tick sex (male and female) and the site of collection as a random effect.
A second model was used to evaluate the influence of the engorgement status, with a subset including females only (n=233), since the engorgement status of males could not be evaluated.This model included the engorgement status, the geographic cluster and the site of collection as a random effect.
Finally, a third model was used to assess the influence of the host on a subset corresponding to ticks located in the Aude/Pyrénées-Orientales cluster (n=323), since ticks were collected on both horses (n=233) and cattle (n=90), while they were only collected from on horses in Hérault/Gard.This model included the host, the tick sex and the site of collection as a random effect.
Significance of variables was assessed using the 'ANOVA' procedure within the package 'car' which performs a type III hypothesis (Fox and Weisberg 2018).Post hoc tests were conducted using the function 'emmeans' (Tukey HSD test).P-value associated to the random effect "site" was assessed by log-likelihood test in the model A.

Maternal transmission of R. aeschlimannii
To assess the potential maternal transmission of the bacteria R. aeschlimannii, five fed females collected on the field were individually placed in 50 mL Falcon tubes in an incubator at a temperature of 27°C for several weeks in the BSL3 facility.About 100 eggs belonging to each of five females were isolated and frozen at -80°C for subsequent analyses while the rest of the eggs were left into the incubator.Only a few eggs (from one adult female) hatched into larvae (n=12).The eggs and larvae were placed separately in a 0.2 mL microtube with 100 µL of DMEM and crushed against the bottom of the tube using a sterile needle.DNA was extracted using 100 µL of homogenate with the Genomic DNA tissue kit (Macherey-Nagel, Germany).Detection and quantification of R. aeschlimannii was performed by targeting the ompB gene by qPCR (Table 2).

Genotyping of microbes detected in H. marginatum
On the 510 ticks analysed using the BioMark TM assay, 445 samples were positive for R. aeschlimannii.The sequences obtained were blasted and all allowed the identification of R. aeschlimannii; the longest sequence (GenBank Accession Number: PP236764) showed 100% identity with R. aeschlimannii collected from a H. marginatum from England in 2019 (AN: MT365092.1).The comparison of the obtained sequence with another one from a tick collected on a horse in 2019 in the Gard department of the Occitanie region showed 99% query cover and 98.76% identity (Bernard et al. 2024) (AN: PP379722).The sequencing results allowed the identification of R. aeschlimannii and these results were extrapolated to the other samples exhibiting identical amplification patterns.Two species of Anaplasma were detected: eight samples for A. phagocyotophilum and four for A. marginale.Three sequences from samples positive for A. phagocytophilum were blasted and all allowed the identification of A. phagocytophilum; the longest sequence (AN: PP265050) showed 99% identity with a sequence isolated from an Ixodes ricinus tick in Italy in 2008 (JQ669948.1).Finally, all four sequences of positive samples for A. marginale allowed the identification of A. marginale, of with the longest sequence (AN: PP218690) showed 100% identity with a sequence obtained from an infected cattle in Iran (GenBank AN : MK016525.1).
T. equi was detected in nine samples.One sequence was obtained (AN: PP227163) and blasted resulting in 100% identity with a sequence of T. equi from an infected Rhipicephalus bursa (MK732476.1) in Corsica, France.T. orientalis was detected in one tick (AN: PP358744), the sequence showed 100% identity with a sequence from a Haemaphysalis longicornis tick in China in 2014 (MH208633.1).
Re-evaluated R. aeschlimannii and T. equi infection rates by qPCR were respectively 89.4% [86.7 -92.1%] and 9.2% [6,7 -11.7%].R. aeschlimannii infection rate was similar with respect to the detection method used whereas the re-evaluation of T. equi infection rate using the qPCR was about five times higher.
Only one co-infection was reported between A. marginale and T. orientalis in one tick whose infection rate was 0.2% [0 -0.6%].Francisella-LE was not considered in co-infections neither R. aeschlimannii whose pathogenic status will be discussed.
T. equi, Francisella-LE and A. phagocytophilum infection rates were not significantly influenced by the geographic cluster (Table 3).The number of positive ticks to A. marginale (n=4) was too low to be statistically analysed but its infection rate was 0% in HG and 1.2% [0 -2.5%] in APO.

According to the collection sites
The collection site had a significant influence on R. aeschlimannii infection rate (p-value = 3.88x10 -29 ) as illustrated by a variability ranging from 29% to 100% between one site to another in Occitanie.In the cluster HG, 50% of the sites had an infection rate up to 100% compared to 84% of the sites in the cluster APO.Interestingly, infection rates for sites located in the geographical cluster HG were more variable than those from the cluster APO (Figure 3A, 3B).The collection site did significantly influence T. equi infection rate (p-value= 1.58x10 -9 ), ranging from 0% to 41% between one site to another in Occitanie (Figure S1A).Finally, it did not have a significant influence on Francisella-LE (p-value= 0.70) and A. phagocytophilum (p-value= 0.0557) infection rates although they varied respectively from 80% to 100% and 3.3% to 20% (Figure S1B, S1C).Finally, A. marginale infected rates were quite similar between the two sites of collection where it was detected (3.1% and 8.8%) (Figure S1D).

The engorgement status
The infection rate of T. equi in female ticks was significantly influenced by the engorgement status (χ²= 8.8679; df= 2; p-value= 0.01187).It was higher in fed females compared to the unfed ones with values reaching 1.8% in average [0 -4.4%] for unfed females and 14.9% in average [6.6 -23.2%] for fed females.
Finally, Francisella-LE, R. aeschlimannii and A. phagocytophilum infection rates as well as R. aeschlimannii and T. equi loads were not significantly influenced by the engorgement status (Table 3  and 4).

DISCUSSION
Recently established in the south of France and known to potentially transmit human and animal pathogens, Hyalomma marginatum might represent a future problem of public and animal health.A better assessment of the risk linked to this tick firstly requires an exhaustive identification of H. marginatum-borne pathogens and a better characterization of their dynamics.In this context, the objective of this study was to characterize the influence of spatial patterns on both the infection rates and loads of H. marginatum-borne pathogens in the region Occitanie in France.
In our study, we detected R. aeschlimannii, T. equi, T. orientalis, A. phagocytophilum and A. marginale in adult H. marginatum ticks.The global infection rates of these pathogens across the Occitanie region were 87.3%, 9.2%, 0.2%, 1.6% and 0.8% respectively.Most of the detected pathogens corresponded to microbes known to have circulated in this geographical area over the last five years (Bernard et al. 2024) which would suggest a certain stability in the circulation of the main H. marginatum-borne pathogens in these departments.At first sight, this would suggest identifying sentinel sites with regular monitoring of pathogens transmitted by H. marginatum.However, we did not detect the bacteria Ehrlichia minasensis previously reported at very low prevalence (in a single tick collected from a horse in the Occitanie region (Bernard et al. 2024).On the other hand, we detected Theileria orientalis, responsible for benign theileriosis in cattle (Watts, Playford, and Hickey 2016), in a fed female collected on a bovine in a single site in Pyrénées-Orientales.This detection was particularly unexpected as this parasite was not previously reported in the area.Furthermore, pathogen infection rates were variable across the sites.In epidemiological terms, these results underline the regular monitoring (e.g.multiyear surveys) of H. marginatum-borne pathogens at fine spatial scales in order to detect pathogens circulating insidiously in the studied area.Please refer to Bernard et al. 2024 for more details on the vectorial competence discussion of the detected pathogens in H. marginatum.
Excluding R. aeschlimannii due to the very high infection rate estimated in ticks for this bacterium (see below for specific discussion on R. aeschlimannii), 11.8% of ticks were positive for at least one of the other tested pathogens and only one co-infection was identified between A. marginale and T. orientalis (0.2%).This low number of co-infection contrasts with Ixodes ricinus which has up to five different co-infections (Moutailler et al. 2016).Co-infection are known to enhance disease severity as this is the case for babesiosis and Lyme disease (Grunwaldt, Barbour, and Benach 1983).Overall, while H. marginatum is known to carry a wide range of pathogens (Bonnet et al. 2023), we detected a low diversity of pathogens circulating in ticks in the region Occitanie with two Anaplasma species, two Theileria species, one species of Rickettsia and only one co-infection.H. marginatum carries fewer pathogens than other tick species of public and veterinary importance like Ixodes ricinus, due to differences in their life cycle such as the number and the host spectrum (Lejal et al. 2019;Rizzoli et al. 2014).However, this substantial number of positive ticks for pathogens known to potentially affect both human and animal health, does incites to accentuate the prevention and maintain a regular surveillance of H. marginatum-borne pathogens in this area.

The scheming dynamics of Rickettsia aeschlimannii
Rickettsia aeschlimannii is known as the agent of spotted fever, a human infection that mainly occurs in North and South Africa, in Greece, Italy and Germany.This bacteria was detected for the first time in H. marginatum in 1997 in Morocco (Beati et al. 1997).In our study, R. aeschlimannii infection rates were very high whatever the detection method and the targeted gene, the citrate synthase gene through the BioMark TM assay (87.3%) or with the ompB gene through qPCR (89.4%).These results are consistent with previous studies performed on H. marginatum in southern France which reported high infection rates; from 75% in individual ticks in H. marginatum collected around the French Mediterranean sea between 2016 and 2019 (Bernard et al. 2024) to 100% of the pools of H. marginatum in Corsica (Grech-Angelini et al. 2020).Such high infection rates observed in our study and many others (Maitre et al. 2023;Bernard et al. 2024) raise the question of the pathogenic status of this detected Rickettsia, since a high number of human patients suffering from rickettsiosis should be observed, but reported cases remain extremely rare and only imported cases have been reported in France to date.While Grech-Angelini et al. ( 2020) suggested that human exposure to R. aeschlimannii infection in Corsica is high and human cases of tick-borne spotted fever acquired in Corsica could often be due to R. aeschlimannii, there is no information available for its vectorial competence based on experimental approaches, showing that this assumption must be nuanced.Further studies are required to characterize the full genome of R. aeschlimannii found in H. marginatum ticks in order to compare it with the complete genome of R. aeschlimannii isolated from a human case.This will help to determine whether the bacterium encountered in ticks is the one responsible for human rickettsiosis cases or whether it is a different strain, potentially involved in a symbiotic relationship with the tick.
Several of our results raise the question of whether R. aeschlimannii is a symbiotic bacterium in H. marginatum.Indeed, the transovarian transmission of R. aeschlimannii in H. marginatum paired with its high infection rates both support this assumption (Parola, Paddock, and Raoult 2005;Azagi et al. 2017).In the literature, such hypothesis have already been suggested for H. marginatum (Maitre et al. 2023;Boularias et al. 2021).Tick symbionts are typically classified as primary or secondary symbionts, depending on their role in the tick's survival.In the present case, R. aeschlimannii infection rate is high (87.3%),but still not reaching typical infection rate of primary symbiont close to 100% like this is the case for Francisella-LE (96.7%).In addition, its status as a primary symbiont seems unlikely because of the significant spatial variability in terms of infection rates and loads, where the infection rate of Francisella-like endosymbiont is very stable regardless of spatial patterns.Therefore, R. aeschlimannii would rather be a secondary endosymbiont.Overall, this hypothesis contrasts with the postulate that Rickettsia symbionts are most common in ticks of the genera of Ixodes, Amblyomma, and Dermacentor, whereas it has been less frequently found in Rhipicephalus, Haemaphysalis, and Hyalomma ticks (Burgdorfer 1981;Hussain et al. 2022;Nováková et al. 2018).
Regarding the functions of symbiotic Rickettsia in tick physiology, it was already demonstrated that they can be involved in the nutrition, such as the Rickettsia buchneri endosymbiont in Ixodes scapularis or I. pacificus (Tokarz et al. 2019;Benson et al. 2004), as it harbours all required genes for folate biosynthesis (Hunter et al. 2015).This nutritive role is unlikely in the case of H. marginatum, insofar as R. aeschlimannii do not have complete biosynthesis pathways for B vitamins, as opposed to the two primary symbionts Francisella-LE and Midichloria that possess either complete biotin, riboflavin or folic acid pathways (Buysse et al. 2021).
Other function of Rickettsia in arthropods in the literature mention defence (Łukasik et al. 2013) and reproductivity (Engelstädter and Hurst 2009).Our results indicate similar infection rates whatever the tick sex assuming that the function of R. aeschlimannii appears to benefit both male and female ticks (whether fed or not).Its role in reproductivity seems unlikely since higher proportion should be observed in females.In regard of the present information, we suggest that R. aeschlimannii would be involved in tick defence against abiotic factors and further investigations for example using transcriptomics approaches should be conducted in the future to confirm such a hypothesis.
In the present study, strong spatial patterns were observed for R. aeschlimannii, whether it be at a large spatial scale (geographic cluster) or at a small one (from one site to another, especially in Hérault/Gard).Some of the infection rates variability between the sites might be explained by unequal tick numbers per sites.It is also possible that while R. aeschlimannii is present in ticks since it is maternally transmitted, the quantities are too low to be detected efficiently, contributing to this variability.In this study, we thus propose two hypotheses to explain this uneven spatial distribution of R. aeschlimannii infection rates.
The first is to assume that R. aeschlimannii infection rate can be influenced by environmental characteristics (humidity, temperature, vegetation, host variability).Rickettsia aeschlimannii would replicate preferentially when environmental conditions threaten the tick and this would be in Aude/Pyrénées-Orientales rather than in Hérault/Gard.In the literature, it was reported that infection by Rickettsia sp. of the spotted fever group in Ixodes ricinus is influenced by the geographic location and their environmental characteristics (forest fragmentation, vegetation, hosts) (Halos et al. 2010;Narasimhan et al. 2021).Another study reported significant difference in loads of Rickettsia phylotype G021 between Ixodes pacificus from different collection sites and vegetation habitats (Cheng et al. 2013).In order to explore this hypothesis, it would be interesting to evaluate the infection rate by R. aeschlimannii and the loads from ticks experimentally maintained in microcosms located in different environmental conditions (vegetation, temperature) in the field (Rahajarison et al. 2014), in order to evaluate the influence of these variables.
The second hypothesis is that the presence and loads of R. aeschlimannii would vary from one H. marginatum population to another.It is interesting to note that, the spatial distribution of R. aeschlimannii in the two geographical areas (Hérault/ Gard and Aude/Pyrénées-Orientales) is consistent with two genetically differentiated populations of H. marginatum characterised using mitochondrial markers clusters main haplotypes (Giupponi et al., personal communication).The presence of structured genetic differentiation could be explained by the introduction of tick populations by migratory birds or horses and cattle from different geographical origins.These introduced tick populations could harbour microbial communities with different compositions, which could explain the spatial patterns of R. aeschlimannii in French ticks.

Dynamics of other H. marginatum microbes
As expected, Francisella-LE was unsurprisingly detected in almost all ticks (96.7%).Francisella-LE is a known primary symbiont vital for tick survival and reproduction by assuring the production of B vitamins production (Duron et al. 2018;Azagi et al. 2017).Our results are consistent with previous studies in which infection rates by Francisella-LE reached 97% in H. marginatum ticks collected in southern France (Bernard et al. 2024) and 90% of H. marginatum pools collected in Corsica (Grech-Angelini et al. 2020).The absence of influence of spatial groups on the rate of infection is not surprising, as it is a primary symbiont necessary for the tick's physiology.On contrast, the sex of ticks had a slight but significant influence on the presence of Francisella-LE with higher infection rates in females than in males.It is interesting to note that this observation has already been reported in Dermacentor ticks (Sperling et al. 2020;Dergousoff and Chilton 2012).Although Francisella-LE quantification was not estimated in our study, this difference may also be explained by the fact that Francisella-LE loads might be lower in males and therefore make the detection more difficult, as already demonstrated in D. variabilis ticks (Chicana et al. 2019).This might be linked to the functional role of this symbiont in the reproduction, which is probably less necessary for males than for females (Duron et al. 2017).Neither the engorgement status of females nor the host of the tick had any influence on the presence of Francisella-LE.
Unlike Francisella-LE, A. phagocytophilum was detected in only 1.6% of the collected ticks.A. phagocytophilum is the etiological agent of human granulocytic anaplasmosis (HGA), a tick-borne fever in ruminants and equine granulocytic anaplasmosis (Chen et al. 1994;Karshima et al. 2023;Stannard et al. 1969).Anaplasma phagocytophilum is maintained in wild ruminants (roe-deer, white-tails deer, white-footed mice), domestic animals (cattle, sheep) and birds in an enzootic cycle which is consistent with our results since 7/8 of A. phagocytophilum infected ticks were collected from cattle.Since A. phagocytophilum is considered an emerging pathogen of horses (Dzięgiel et al. 2013), it is not surprising to find a tick collected from a horse infected by this bacterium.Finally, most of the infected ticks were females (6/8) that were either semi-engorged or fed although no significant influence of the tick sex or the engorgement status was observed.This result would indicate that ticks become infected after a blood meal suggesting that the tick infection is a reflect of the host infection.However, such a hypothesis might imply that more ticks should be infected.One thing is for sure that the role of H. marginatum as a vector of this pathogen is questioned and further studies need to address its vectorial competence (Bernard et al. 2024) Interestingly, another Anaplasma species, Anaplasma marginale, was detected in ticks with an infection rate of 0.8%.Anaplasma marginale is responsible for fever, anaemia, weight loss and abortions in cattle.The main reported vectors of this bacteria are Rhipicephalus spp.and Dermacentor spp. in tropical and subtropical regions that can be found in Europe.Statistics were not applicable due to the low number of A. marginale infected ticks (n=4) but descriptive data showed that all infected ticks were engorged females (semi or fully fed).All the ticks were collected in two sites of Aude/Pyrénées-Orientales, on cattle.The most likely hypothesis is that the four female fed ticks infected with A. marginale became infected by blood feeding on the same infected animal bovine host.
Finally, the equine piroplasmosis agent T. equi parasite was detected in our study.We first estimated T. equi infection rate at 1.8% with the BioMark TM assay.However, this approach made us aware of the probable underestimation of the infection rate as we used primers that did not allow the detection of all genotypes potentially circulating in France.We re-evaluated T. equi infection rates using another gene (18S) that can target several genotypes of T. equi by qPCR an infection resulting in a rate of 9.2%.Interestingly, this infection rate was very low compare to a study that showed 43% of H. marginatum infected with T. equi collected from horses in Camargue, next to the Occitanie region (Rocafort-Ferrer et al. 2022).This large difference can probably be explained by the design of the tick collection that aimed to target stables where cases of piroplasmosis are frequently reported.Most of T. equi infected ticks were collected on horses (33/47).As horses are known to be reservoirs of T. equi and a high circulation rate of piroplasmosis has been described in France, it is not surprising that the majority of infected ticks were collected on horses (Nadal, Bonnet, and Marsot 2022).However, it should be noted that 14/47 of the infected ticks were collected from cattle, which is more surprising since cattle are not known to be susceptible to infection by T. equi but rather to T. annulata and T. parva.Such a result could suggest maternal transmission of this parasite in ticks but this has not yet demonstrated and should be tested experimentally in H. marginatum.It could also suggest that the adult tick could have been infected during its immature stages on a susceptible host.However, it is currently unknown if the hosts of immature stages (birds and lagomorphs) can carry and be infected with T. equi.
While T. equi infection rates were not influenced by the geographic cluster, significant differences were observed on T. equi loads, with a lower number of genomes.µL - in Hérault/Gard than in Aude/Pyrénées-Orientales.However, this result has to be nuanced regarding the fact that a tick collected on cattle in a site of Aude/Pyrénées-Orientales presented very high T. equi loads compared to other infected ticks.
Finally, although T. equi infection rates and loads were not influenced by the sex of the tick, T. equi infection rate was significantly higher in fed females compared to unfed ones, suggesting that H. marginatum ticks tend to be infected by T. equi via the blood meal on a parasitemic horse rather than being responsible for the transmission of T. equi to an uninfected horse.

CONCLUSION
This study characterized the spatial distribution of H. marginatum-borne pathogens in the region Occitanie where the tick has recently become established.At the scale of the whole region, we reported that (i) 11.8% of H. marginatum ticks were infected with at least one species of Theileria or Anaplasma; (ii) the infection rate of R. aeschlimannii was very high (87.3%),questioning its pathogenic status regarding of the low number of human cases; (iii) R. aeschlimannii is hypothesized to be a secondary symbiont of H. marginatum due to its high infection rates and its maternal transmission; (iv) infection rates of all detected pathogens were quite variable from one site of collection to another, demonstrating the importance of the sampling effort for pathogen circulation surveillance.The particularly marked spatial patterns for R. aeschlimannii presence and loads might be linked to the tick population.Beyond the spatial scale, pathogens dynamic can also fluctuate depending on the temporal scale, even at a monthly scale (Pollet et al. 2020;Lejal et al. 2019).It is necessary to highlight temporal patterns that might affect pathogens and symbionts in H. marginatum.To do this annual and monthly monitoring of ticks collected in a given location would help us identify temporal dynamics.

Figure 1 :
Figure 1: Map of the region Occitanie where the tick sampling was performed in May 2022.The Occitanie region is delimited by a black line.White circles indicate H. marginatum free sites and blue crosses indicate H marginatum positive sites where ticks were collected and selected for analyses in 2022.Blue-coloured departments represent the geographic clusterAude/Pyrénées-Orientales and the green-coloured departments represent the geographic cluster Hérault/Gard.Greycoloured departments were visited but no H.marginatum were found.Uncoloured areas correspond to other departments that were not visited for the tick collection since H. marginatum introduction/installation was not reported (ECDC).

Figure 2 :
Figure 2: Infection rate and bacterial load of R. aeschlimannii in H. marginatum ticks collected in the two geographic clusters Hérault/Gard and Aude/Pyrénées-Orientales.The number of ticks per cluster is indicated above the x axis legends.Significant differences are represented by asterisks (*: p-value < 0.05).A. R. aeschlimannii presence (1) or absence (0) for each individual tick examined by the BioMark TM assay is summarized by the mean (infection rate in %) and errors bars represent 95% confidence interval.B. Bacterial loads expressed in genome.µL - obtained by qPCR assay are represented by boxplot summarizing the median, 1st and 3rd quartiles.Each grey dot represents the loads of R. aeschlimannii for one tick.Mean is symbolized by the black dot.

Figure 3 :
Figure 3: A. R. aeschlimannii Infection rate in H. marginatum for each site of collection in the two geographic clusters Hérault/Gard and Aude/Pyrénées-Orientales.The number of sites is indicated above the x axis legends.Infection rates for each site were determined with the BioMark TM assay and are represented by black dots.B. R. aeschlimannii spatial distribution and its infection rates for each site; represented by a red circle when the infection is 100% and in white when it is less than 100%.Precise infection rates are indicated next to each site.

Figure 4 :
Figure 4: Infection rate and bacterial load of R. aeschlimannii in H. marginatum males and females.The number of males and females is indicated above the x axis legends.Significant are represented by asterisks (*** : p-value < 0.001)."ns" : nonsignificant.A. R. aeschlimannii presence (1) or absence (0) for each individual tick examined by the BioMark TM assay is summarized by the mean (infection rate in %) and errors bars represent 95% confidence interval.B. Bacterial loads expressed in genome.µL - were obtained by qPCR assay and presented by boxplot summarizing the median, 1st and 3rd quartiles.Each grey dot represents the loads of R. aeschlimannii for one tick.Mean is symbolized by the black dot.