The Genetic Diversity of Rickettsiella Symbionts in Ixodes ricinus Throughout Europe

Rickettsiella species are bacterial symbionts that are present in a great variety of arthropod species, including ixodid ticks. However, little is known about their genetic diversity and distribution in Ixodes ricinus, as well as their relationship with other tick-associated bacteria. In this study, we investigated the occurrence and the genetic diversity of Rickettsiella spp. in I. ricinus throughout Europe and evaluated any preferential and antagonistic associations with Candidatus Midichloria mitochondrii and the pathogens Borrelia burgdorferi sensu lato and Borrelia miyamotoi. Rickettsiella spp. were detected in most I. ricinus populations investigated, encompassing a wide array of climate types and environments. The infection prevalence significantly differed between geographic locations and was significantly higher in adults than in immature life stages. Phylogenetic investigations and protein characterization disclosed four Rickettsiella clades (I–IV). Close phylogenetic relations were observed between Rickettsiella strains of I. ricinus and other arthropod species. Isolation patterns were detected for Clades II and IV, which were restricted to specific geographic areas. Lastly, although coinfections occurred, we did not detect significant associations between Rickettsiella spp. and the other tick-associated bacteria investigated. Our results suggest that Rickettsiella spp. are a genetically and biologically diverse facultative symbiont of I. ricinus and that their distribution among tick populations could be influenced by environmental components.


Introduction
Ticks interact with multiple biotic and abiotic factors that modulate the success of their survival and development. Environmental factors -such as climate, habitat type, and host availability -are among the most limiting drivers for ticks, determining their geographical spread and abundance [1,2]. Moreover, ticks are involved in commensal, mutualistic, and parasitic relationships with a wide range of microbes, including bacteria, viruses, and protozoa, which can cause infectious diseases in humans and animals. Together, they compose the so-called tick microbiota [3].
Among members of tick microbiota, tick-borne pathogens have sparked the greatest interest, given their worldwide impact on public and veterinary health [4][5][6][7]. However, recent advances in molecular techniques enabled the identification of complex bacterial communities residing in ticks, including symbiotic bacteria that display an important role in tick ecology and physiology [8]. In fact, some of these symbiotic bacteria are considered essential for tick survival since their absence reduces the reproductive fitness of ticks [9]; others supplement the unbalanced diet of ticks by biosynthesizing vitamins that are scarce in vertebrate blood [10,11]. Microbial communities within ticks have been hypothesized to modulate the occurrence and diversity of tick-borne pathogens and consequently determine their transmission to vertebrate hosts [8].
Ixodes ricinus is the most widespread tick species in Europe and the primary vector of Borrelia burgdorferi sensu lato, the causative agent of Lyme borreliosis [12]. It also 1 3 transmits Borrelia miyamotoi, an emerging human pathogen belonging to the relapsing fever group borreliae [13]. In the past years, several studies have explored I. ricinus microbiota, identifying different heritable bacteria. Among these, Candidatus Midichloria mitochondrii (hereafter M. mitochondrii) is one of the most abundant symbionts and widespread among several tick species. It is mainly located within the mitochondria of host ovary cells [14,15], however, it has been also identified in salivary glands and Malpighian tubules, suggesting that it might play different roles in tick physiology [16]. Although there is evidence for the presence of M. mitochondrii DNA in vertebrate hosts of ticks, it remains unclear whether its presence in vertebrates may cause disease or may facilitate horizontal transmission [17].
Relatively few studies have so far reported on Rickettsiella infection in I. ricinus [20,[32][33][34][35], and its biological role in this important vector is entirely unknown. In the present study, we examined European Ixodes ticks for the presence of Rickettsiella spp. and determined its genetic diversity, with a focus on I. ricinus. Moreover, we evaluated possible associations of Rickettsiella spp. with M. mitochondrii and the pathogens B. burgdorferi s.l. and Borrelia miyamotoi.

Tick Sampling
To evaluate the presence of Rickettsiella symbionts in I. ricinus populations, we studied a total of 4941 questing ticks from different latitudes across Europe (Appendix), which were morphologically identified to species level [36]. Northwestern and central Europe were represented by 2033 Ixodes ticks collected in green areas of Antwerp (Belgium; [37]), 999 I. ricinus from nine different forest sites across the Netherlands and 66 I. ricinus nymphs from Güterfelde, northeastern Germany, collected by dragging in April 2019. We also included 240 I. ricinus collected in two different areas from the southern United Kingdom. Northern latitudes were represented by tick populations collected in September 2018 in forested areas in Södermanland county, Sweden (n = 235). Finally, Southern Europe accounted for three mountain areas located in north-western and central Italy (n = 1,368) as previously described [38][39][40]. Furthermore, some specimens of questing Ixodes frontalis (n = 4) and Ixodes hexagonus (n = 6) ticks from Belgium were also included in this study.

Detection of Rickettsiella Symbionts and Other Tick-Borne Bacteria
DNA extraction from ticks of most locations was performed using ammonium hydroxide [41], except for the ticks from northwestern and central Italy from which tick DNA was extracted using DNAzol reagent (Life Technologies LTD, Warrington, UK) [40] and the QIAGEN DNeasy tissue kit (Qiagen, Hilden, Germany) [38], respectively.
All ticks were screened by multiplex polymerase chain reaction (qPCR) assays targeting the ospA and flaB genes of B. burgdorferi s.l. and the flaB gene of B. miyamotoi as previously described in [42,43]. To detect the tick symbionts Rickettsiella spp. and M. mitochondrii a multiplex qPCR protocol was designed using primers and probes described in Table 1. A sample was considered positive for Ricketsiella if at least one of the two targets, either gidA or sucB, was positive. Positive qPCR-Rickettsiella samples were subjected to end-point conventional PCR targeting a 786-bp fragment of gidA gene with primers previously described in [44].
All qPCRs were carried out on a LightCycler 480 (Roche Diagnostics Nederland B.V, Almere, the Netherlands) in a final volume of 20 μl with iQ multiplex Powermix, 3 μl of sample, primers with end concentration of 0.2 μM, and probes. Positive plasmid controls and negative water controls were used on every plate tested. Cycling conditions included an initial activation of the iTaq DNA polymerase at 95 °C for 5 min, followed by 60 cycles of a 5-s denaturation at 95 °C followed by a 35-s annealing-extension step at 60 °C (ramp rate 2.2 °C s −1 and a single point measurement at 60 °C) and a final cooling cycle of 37 °C for 20 s. To minimize contamination and false-positive samples, the DNA extraction, PCR mix preparation, sample addition, and qPCR analyses were performed in separate air-locked laboratories.

Statistical Analyses
Tick infection and coinfection prevalence and 95% binomial confidence intervals were calculated using RStudio software v1.1.463 [45]. Moreover, we compared the frequency of tick-borne bacterial infections and coinfections among the European countries investigated, including sampling areas within the single regions, as well as between tick life stages (nymphs and adults) using Pearson's chi-squared test. Finally, we used the Kappa (κ) statistic to assess the association by quantifying coinfections of Rickettsiella spp. with tick-borne pathogens (B. burgdorferi s.l. and B. miyamotoi) and M. mitochondrii [46].

Phylogenetic Analyses
Forward and reverse nucleotide sequences for each Rickettsiella gidA amplicon were accurately assembled and manually corrected into final nucleotide sequences of 723 bp using Geneious Prime (version 2020.0.4) software. These were aligned to available reference gidA sequences from different arthropod species, including ticks from North America, such as Ixodes angustus (GenBank accession n. HG792871), Ixodes kingi (HG792872), Ixodes sculptus (HG792873) and Ixodes woodi (JQ070345), other arachnid species (KU597421) and several insects belonging to the orders Coleoptera and Isopoda (JQ679309, JX406182, JN565687-89-91, JF288927). A Coxiella burnetii (AE016828.3) sequence was used as an outgroup given its evolutionary relationship with Rickettsiella spp. The nucleotide alignment was performed using ClustalW [47] respecting the coding frame. This method avoids nucleotide sequence misreading based on the conversion of nucleotide sequences into their corresponding peptide sequences which, in turn, will be eventually translated back into the nucleotide sequences they have originated from [48]. Nucleotide and amino acid diversities were calculated among all the analysed samples and expressed as percentage. Nucleotide alignment as well as the amino acid variable positions was evaluated in order to investigate the possible population structure. We assessed the phylogenetic relationships among Rickettsiella positive samples and reference sequences using a MCMC Bayesian approach by using the GTR substitution model [49] embedded in MrBayes software v3.2.7a [50]. The resulting consensus tree was viewed and edited using FigTree v1.4.4 [51].

Occurrence of Rickettsiella spp. in Different European Regions
Rickettsiella spp. was widespread in most of the I. ricinus populations, although its occurrence between geographic locations varied (Pearson's Chi-squared test, p < 0.001). Regions from western Europe showed the highest infection rates, except for the tick population from Germany, in which the bacterium was not detected at all. Tick populations from Belgium reached a prevalence of nearly 90%, followed by around 60% and 50% of infection observed in populations from the Netherlands and UK, respectively ( Table 2). We conversely recorded the lowest infection rates in I. ricinus ticks from northern and southern latitudes (Sweden and Italy), with less than 20% prevalence ( Table 2). This variability was also evident within the single

Rickettsiella spp. Through Ixodes ricinus Ontogeny
Adults (n = 389) and nymphs (n = 4185) were analysed, as well as a sample of 367 larvae from the Netherlands. Larvae, nymphs, and adult ticks harboured Rickettsiella spp. (Table 2), but the infection exhibited substantial changes according to tick life stage (p < 0.001) and the odds of infection for adults doubled those for nymphs (OR = 2.4, 95% CI = 1.7-3.4). Only in the Netherlands, infection in nymphs slightly prevailed over the adults' prevalence, but

Genetic Diversity of Rickettsiella spp.
We successfully obtained 263 gidA gene (723 bp) nucleotide sequences from I. ricinus from most European regions in the study, and one from I. frontalis collected in Belgium. Overall, gidA sequences showed a high level of similarity, with 85 of 100 sequences from Belgium, 65/98 from the Netherlands, 37/52 from the UK, 6/8 from Sweden, and 2/5 from Italy having 100% of identity. Given the amount of gidA sequences produced and the high degree of similarity observed between them, we selected 115 sequences comprising all nucleotide sequences with different levels of dissimilarity (n = 75), and a batch of those sharing 100% of identity (n = 40), representative for each European region. Phylogenetic tree topology showed a clustering of the Rickettsiella population into four well-defined clades ( Fig. 1) which, at time, exhibited different frequencies (p < 0.001). Rickettsiella-Clade I represented the largest group observed comprising 81.7% (95% CI = 76.5-86.2) of the entire population followed by Clade III (12.2%; 95% CI = 8.5-16.8).
Rickettsiella-Clades II and IV, on the other hand, respectively accounted for 2.7% (95% CI = 1.1-5.4) and 3.1% (95% CI = 1.3-5.9) of the symbiont population, thus representing the least frequent Rickettsiella Clades detected in I. ricinus ticks. We also identified a Rickettsiella spp. from a single I. frontalis nymph that clustered within Rickettsiella-Clade I and showed 99.6-100% of identity with Rickettsiella strains described from I. ricinus. Analysis of Rickettsiella sequences from North American hard ticks showed that only a Rickettsiella spp. recorded in I. angustus was closely related to those of I. ricinus and clustered within Clade III. Rickettsiella sequences identified in I. kingi, I. sculptus and I. woodi gathered within a single clade that was distantly related to those observed in I. ricinus (Fig. 1).
This classification relied upon the genetic structure observed in parallel at nucleotide and amino acids (AA) levels. Overall, Rickettsiella populations within the same clade consistently experienced a low degree of variability since the uncorrected p-distance (up) averaged around 0.77% (min.-max. = 0-2.68; Table S1, Supporting Information). For instance, Clade I constituted the most conserved clusters, despite being composed of Rickettsiella from different geographic areas and harbouring sequences obtained from different arthropod species. Clade I-gidA sequences are closely related to Candidatus Rickettsiella isopodorum reported in terrestrial isopods species (JQ679309; JX406182) from Germany, sharing a high level of identity (> 95%). Rickettsiella-clades differentiation was sustained not only by the genetic distance, which showed mean values below 13.6% (Table S1, Supporting Information) among groups, but also by the AA composition they encoded. Using a 241 AA sequence of Coxiella burnetii as a reference, we could identify a total of 28 variable sites in the Rickettsiella AA sequences, which enabled us to determine the protein sequence motifs that characterized the single clades (Table S2, Supporting Information). From a geographic point of view, different distribution patterns of each Rickettsiella clade were seen (Fig. 2). The most prevalent population, Clade I, was distributed among all I. ricinus investigated and was the only Rickettsiella clade detected in the British study sites. Rickettsiella-Clade III was detected in all study regions except for UK and prevailed in I. ricinus from Sweden. These two clades concurrently circulated with Rickettsiella-Clades II and IV in I. ricinus from closely related north-western regions. Clade II was only detected in ticks from Belgium study sites, whereas Clade IV was discovered in tick populations from the Netherlands and, to a lesser extent, Belgium. Rickettsiella spp. Fig. 1 Phylogenetic tree of gidA gene of Rickettsiella spp. obtained from questing Ixodes ricinus populations collected from different European regions. The tree was built by using GTR substitution model (γ = 0.05) and nodal support is given as a posterior probability for Bayesian Inference under percentage format; only probability values above 70 are shown next to the internal nodes. Coxiella burnetii was used as the outgroup. Reference sequences are identified by Gen-Bank accession number enclosed in parentheses. Amplicons obtained in the study are indicated with a black symbol representing the region of provenance: Belgium (■), Italy (▲), the Netherlands (•), Sweden (◆), and the UK (▼). A large gidA sequence set from Rickettsiella-Clade I was removed from the tree for readability purposes, containing a total of 64 sequences identified in I. ricinus from Belgium (n = 10), the Netherlands (n = 28), UK (n = 25), and Sweden (n = 1) from Clade IV were closely related (95% identity) with R. agriotidis (JN565687), R. costelytrae (JN565689), and R. pyronotae (JN565691) that were recorded in different beetle species from Germany and New Zealand.

Relationship of Rickettsiella spp. with Other Tick-Borne Microorganisms
Both investigated tick-borne pathogens, B. burgdorferi s.l. and B. miyamotoi, were present in all study areas. The prevalence of B. burgdorferi s.l. varied among the studied tick populations and geographic locations (Pearson's chi-squared test, p < 0.001), with Sweden and Belgium showing the highest infection prevalence ( Table 2). Borrelia burgdorferi s.l. infection prevalence was associated with the tick life stage. It was significantly higher in adult ticks compared to nymphs (p < 0.0001). The odds of B. burgdorferi s.l. infection for adults, in fact, doubled those for nymphs (OR = 2.1; 95% CI = 1.6-2.7). Borrelia burgdorferi s.l. prevalence in adult ticks in Belgian and Italian tick populations (29.8% and 38.1%, respectively) was also significantly higher than the prevalence in nymphs (16.8% and 14.5%, respectively; p < 0.001).
Neither geographic location nor tick life stage was associated with the occurrence of B. miyamotoi in ticks. The infection prevalence of B. miyamotoi was low in all studied areas ( Table 2) and did not differ among regions (p = 0.07). The pathogen was observed in nymphs from all the investigated countries, but only in adults from Belgium (2.7%; 95% CI = 1.1-5.4) and Italy (1.7%; 95% CI = 0.06-12.6).
A generally high M. mitochondrii prevalence was detected (   (Table 2). In regions for which information on the sex of adults was available (Belgium, the Netherlands, UK, and Italy), the odds of M. mitochondrii infection were significantly higher in female ticks compared to that in males (OR = 18.0; 95% CI = 7.8-48.4).
We detected the highest rates of bacterial coinfections in ticks from Belgium (around 75%), followed by the UK and Sweden (Table 4 and 5). While around 70-80% of coinfected ticks from Belgium and UK exclusively involved symbionts, over 75% of coinfections from the other countries involved at least one pathogen and one symbiont (Table 4). In Germany, only two out of 27 infected ticks were coinfected, with B. burgdorferi s.l. and M. mitochondrii.
We detected all possible combinations of coinfections. Coinfections with two bacteria were the most frequently observed in all regions. Belgium recorded the highest rates of Midichloria-Rickettsiella coinfections, affecting around 75% of the tick population. In countries where coinfections in ticks predominantly involved pathogens (B. burgdorferi s.l. and B. miyamotoi) in combination with symbionts (M. mitochondrii and Rickettsiella spp.), B. burgdorferi s.l. played the major role since it was far more prevalent than B. miyamotoi. Around 60-70% of the coinfections observed in Italy, Netherlands and Sweden were constituted by B. burgdorferi s.l. and at least one symbiont. As regards multiple coinfections, with more than two bacteria, the combination B. burgdorferi s.l.-M. mitochondrii-Rickettsiella spp. was the most prevalent. Coinfections with all four investigated bacteria were recorded in ticks from Belgium, Italy, the Netherlands, and Sweden.
The life stage of I. ricinus was not associated with the frequency of coinfections by two bacterial species, while multiple coinfections, with three bacteria or more, markedly prevailed in adult ticks compared to nymphs (p < 0.001).
Our results did not show significant associations between Rickettsiella spp. and B. burgdorferi s.l.,

Discussion
This study contributes to the limited available knowledge about Rickettsiella symbionts in Ixodes ticks throughout Europe. Our results confirm the broad distribution of this heritable bacteria among different tick populations, encompassing a wide array of habitat types and climatic environments. In particular, we observed Rickettsiella spp. in geographically distant populations of I. ricinus, in which they displayed significantly different infection rates and highly diverse genetic features.
Intracellular bacteria of the genus Rickettsiella seem common in Ixodes ticks, though they have been also detected in metastriate tick species such as Amblyomma spp. and Haemaphysalis spp. [52], and soft ticks [32]. Rickettsiella spp. have been reported in Ixodes ticks from all over the world, including I. kingi, I. sculptus, I. angustus, I. woodi, and I. scapularis from North America [18,21,53]; I. tasmani from Australia [19]; and I. uriae from different locations of the northern and southern hemispheres [22]. In addition, Rickettsiella spp. have recently been detected in the tree-hole tick I. arboricola [54,55]. To date, however, there are relatively few records of Rickettsiella infection in I. ricinus in Europe [20,[32][33][34][35]. The prevalence of Rickettsiella spp. in I. ricinus in our study varied across its geographic range. This finding is in line with previous large-scale studies concerning the polar seabird tick, I. uriae, in which Rickettsiella prevalence ranged from 3 to 24% [22], and the Canadian I. sculptus, where the infection rate averaged from 2 to 42%, according to the geographic location [21]. We recorded a higher prevalence in I. ricinus from western Europe, where Rickettsiella infection rate fluctuated between 48 and 89%. Lower prevalence rates were recorded for tick populations collected from the more northern and southern latitudes here investigated (16% and 3%, respectively). However, sampling differed in the studied countries, including different areas within the Netherlands, UK, and Italy versus unique sampling locations in Germany, Sweden and Belgium, and this could have influenced the detected prevalence.
Our results suggest that Rickettsiella spp. is a facultative symbiont of I. ricinus, and its infection rate varies depending on the environmental context. Also, the variable Rickettsiella infection rates in different tick species and populations might result from species-specific ecology and behaviour. For instance, I. uriae and I. sculptus generally live near their vertebrate hosts [36,56], which might ensure the successful development and survival of ticks. In these circumstances, the cost of bearing a bacterial symbiont could outweigh the benefits potentially provided by the symbiont, leading to a low infection rate. Conversely, I. ricinus is exposed to a constantly changing environment [57]. The exophilic lifestyle hypothetically requires/facilitates relationships with facultative symbionts, which depending on environmental conditions, affect a variety of phenotypes of their hosts. Consequently, this could be one of the explanations for a significantly higher infection rate of Rickettsiella spp. observed in I. ricinus than in other tick species.
Moreover, environmental factors have been reported to affect the composition of microbial communities residing in I. ricinus [34,35,58]. Constant interactions between ticks and the immediate environment could lead to bacteria acquisition directly from the soil and vegetation [34,58,59], where they spend most of their life. Thus, differences in habitat characteristics might determine either the presence or absence of certain microbial taxa in different tick populations [20]. Accordingly, we may hypothesize that climate and habitat type could influence the spatial distribution of Rickettsiella spp., limiting its occurrence to local environments. For example, we observed remarkable differences in Rickettsiella distribution in mountain areas from the northwestern Alps: despite having similar habitat characteristics, we only detected the symbiont in one out of the two populations examined, in which 11.8% of the ticks were infected. Local microclimatic habitats might justify these differences. In fact, recent studies carried out in areas of the Swiss Alps reported a mean Rickettsiella prevalence of 63% in I. ricinus, uncovering a good correlation between tick abundance and precipitation with the occurrence of the bacterium [35]. However, we cannot rule out that the observed differences in Rickettsiella prevalence and diversity are stochastic events.
Essential biological processes of ticks, such as feeding and moulting, also affect the composition of tick microbiota [60] and may result in changes in microbial richness [58,61]. In this study, we detected Rickettsiella spp. in I. ricinus larvae, nymphs, and adults, suggesting that the symbiont is maintained through successive life stages. Although no differences related to sex were observed, the prevalence was significantly higher in adults. As it has been indicated for other tick species [54], Rickettsiella sp. infection is possibly advantageous to ticks, increasing the survival success of infected individuals to the following life stage. Alternatively, the increased prevalence of Rickettsiella spp. in adults might result from an occasional horizontal transmission occurring, for example, via co-feeding of infected and uninfected ticks. Therefore, multiple feedings could increase the possibility of acquiring the symbiont. Nevertheless, in both scenarios, the subsequent transmission from adults to the offspring is hampered to some extent as we observe lower prevalence in larvae. To understand the transmission dynamics of Rickettsiella spp., future studies should investigate Rickettsiella sp. prevalence on a temporal scale and the role of vertebrates in propagating the symbiont in tick populations.
We detected an extensive genetic diversity of Rickettsiella spp. infecting I. ricinus. Four different clades composed the genetic structure of this symbiont, with some of them being associated with specific geographic areas. The diversity of Rickettsiella spp. was reported in other Ixodes species; for instance, Duron et al. [22] reported up to 12 genetically distinct strains in I. uriae, while three different 16S rRNA haplotypes were observed to coexist within I. tasmani populations [19]. On the other hand, only one Rickettsiella strain was reported for I. angustus, I. sculptus, and I. kingi, but in this study, the sample size was limited [21].
We acknowledge some limits with regard to gidA gene marker employed for this study. Relatively few gidAreference sequences can be gathered from the GenBank database, thus preventing comparisons with Rickettsiella sequences already detected in other I. ricinus populations and in different arthropod species. Notwithstanding, gidA gene has been reported as a good marker to infer in the phylogeny of this genus, in particular at the infrageneric level [62]. Rickettsiella strains here identified, besides being genetically diverse, were also biologically different, as supported by the protein sequence motifs that defined each clade. These results may indicate that different Rickettsiella strains could have different roles in the life cycle of ticks. For example, it was suggested that Rickettsiella infection may affect tick reproduction in lab-reared colonies of I. woodi [18], while such effects were not related to the occurrence of Rickettsiella spp. in I. arboricola [55]. Moreover, some Rickettsiella strains have been historically considered entomopathogenic [23,62], whereas other strains are facultative mutualist in their hosts [31,[63][64][65]. Whole-genome sequencing could possibly help in the identification and taxonomic classification of tick symbionts and in the evaluation of their biological relationship with ticks. For instance, this method could particularly help in clarifying the taxonomic attribution of Diplorickettsia, which was described as a new bacteria genus of I. ricinus [66], but is nested within the Rickettsiella genus [22,67]. However, difficulties in isolation and cultivation still hamper the study of these microorganisms.
The close phylogenetic relations between Rickettsiella strains of I. ricinus and of other arthropods denotes the possibility of horizontal transmission among different arthropod species. Duron et al. [22] disclosed similar results in Rickettsiella strains identified in I. uriae, arguing that the symbiont might combine both vertical and horizontal transfers to ensure its maintenance and global distribution within arthropod communities. On the other hand, the specific natural environment, together with factors intrinsic to the tick hosts, might have significantly contributed to the evolution of Rickettsiella spp. In fact, isolation patterns were evident for some Rickettsiella populations: specific evolutionary forces experienced by I. ricinus and its symbionts may support the unexpected high variability observed in Rickettsiella populations from certain locations of central-western Europe. Consequently, further studies should investigate whether the genetic diversity of Rickettsiella spp. in I. ricinus accounts for adaptive selection in specific environments, while assessing potential co-phylogenetic associations between the symbiont and their tick hosts, as observed for other heritable endosymbionts [32,68,69].
Ticks are commonly co-infected with tick-borne pathogens, which have implications for public health [70]. The success of pathogen transmission, however, may be affected by their co-occurrence with symbiotic bacteria, leading to direct facilitation or competition effects among microbes [8]. Our results did not disclose any significant association between Rickettsiella symbionts and pathogens, which might entail that these symbionts do not interfere with the presence of B. burgdorferi s.l. and B. miyamotoi. These results are in line with previous studies in which a greater selection of tick-borne pathogens and symbionts was assessed [70]. By contrast, Aivelo et al. [35] reported strong associations between Rickettsiella spp. and different Lyme spirochetes. However, these positive correlations may be influenced by the greater proportion of adult ticks studied by Aivelo and colleagues, in which co-infections are more likely to occur compared to nymphs. Notwithstanding, our results do not exclude the possibility of associations between symbionts and pathogens. The heterogenicity of our results may not only be biased by differences in the epidemiological context, but also influenced by the diverse sample size tested per each geographic location.

Conclusions
Our study demonstrates the widespread distribution of Rickettsiella symbionts among I. ricinus populations in Europe. Overall, Rickettsiella spp. displayed a high genetic variability according to the geographic locations, which highlights the importance of small-scale factors in shaping the distribution of this symbiotic bacterium. The in-depth evaluation of micro and macroclimatic variables affecting the occurrence of Rickettsiella spp. among different tick populations, as well as further phylogenetic studies, will hopefully elucidate co-evolution patterns between Rickettsiella spp. and ixodid ticks, and the specific geographic distribution of Rickettsiella strains. Moreover, our preliminary data suggest that Rickettsiella symbionts may also be present in other Ixodes species in Europe, but further research is needed to improve the knowledge about the symbiont's ecology.
Funding Open access funding provided by Università degli Studi di Torino within the CRUI-CARE Agreement. This research was financially supported by the Dutch Ministry of Health, Welfare and Sport (VWS) and a grant from the European Interreg North Sea Region program, as part of the NorthTick project. AMN received financial support from the Federal Ministry of Education and Research (BMBF) under project number 01KI1720 as part of the 'Research Network Zoonotic Infectious Diseases'. DH is funded by the Marie Sklodowska-Curie Actions (EU-Horizon 2020, Individual Global Fellowship, project no 799609).

Data Availability
Representative sequence data are available on NCBI with Accession numbers: MW876489-MW876509.

Declarations
Ethics Approval and Consent to Participate Informed consent was obtained from all individual participants included in the study.

Conflict of Interest
The authors declare no competing interests.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.