Rickettsia vini n. sp. (Rickettsiaceae) infecting the tick Ixodes arboricola (Acari: Ixodidae)

Background Recently, a new rickettsia named ‘Candidatus Rickettsia vini’ belonging to the spotted fever group has been molecularly detected in Ixodes arboricola ticks in Spain, the Czech Republic, Slovakia and Turkey, with prevalence reaching up to 100 %. The aim of this study was to isolate this rickettsia in pure culture, and to describe it as a new Rickettsia species. Methods A total of 148 ornitophilic nidicolous ticks Ixodes arboricola were collected in a forest near Breclav (Czech Republic) and examined for rickettsiae. Shell vial technique was applied to isolate rickettsiae in Vero cells. Rickettsial isolation was confirmed by optical microscopy and sequencing of partial sequences of the rickettsial genes gltA, ompA, ompB, and htrA. Laboratory guinea pigs and chickens were used for experimental infestations and infections. Animal blood sera were tested by immunofluorescence assay employing crude antigens of various rickettsiae. Results Rickettsia vini n. sp. was successfully isolated from three males of I. arboricola. Phylogenetic analysis of fragments of 1092, 590, 800, and 497 nucleotides of the gltA, ompA, ompB, and htrA genes, respectively, showed closest proximity of R. vini n. sp. to Rickettsia japonica and Rickettsia heilongjiangensis belonging to the spotted fever group. Experimental infection of guinea pigs and chickens with R. vini led to various levels of cross-reactions of R. vini-homologous antibodies with Rickettsia rickettsii, Rickettsia parkeri, ‘Candidatus Rickettsia amblyommii’, Rickettsia rhipicephali, Rickettsia bellii, and Rickettsia felis. Laboratory infestations by R. vini-infected I. arboricola larvae on chickens led to no seroconversion to R. vini n. sp., nor cross-reactions with R. rickettsii, R. parkeri, ‘Ca. R. amblyommii’, R. rhipicephali, R. bellii or R. felis. Conclusions Our results suggest that R. vini n. sp. is possibly a tick endosymbiont, not pathogenic for guinea pigs and chickens. Regarding specific phenotypic characters and significant differences of DNA sequences in comparison to the most closely related species (R. japonica and R. heilongjiangensis), we propose to classify the isolate as a new species, Rickettsia vini. Electronic supplementary material The online version of this article (doi:10.1186/s13071-016-1742-8) contains supplementary material, which is available to authorized users.


Background
Rickettsiae are Gram-negative coccobacilli belonging to the family Rickettsiaceae, order Rickettsiales in the alpha subdivision of the class Proteobacteria. Rickettsia spp. have small genomes (1.1-2.1 Mb) resulted from reductive evolution caused by their obligate endosymbiotic relationship to eukaryotic cells [1]. Their host diversity is remarkably high. Although all valid species are associated with arthropods, novel genotypes have also been identified in annelids, amoebae and plants [2,3]. A number of Rickettsia species can propagate in vertebrates, some of them cause diseases in humans and animals, to which they are transmitted by arthropod vectors such as fleas, lice, mites or ticks. Some species are considered non-pathogenic, and novel Rickettsia species reveal to be nearly cosmopolitan [4].
Originally, pathogenic rickettsiae used to be divided into two groups, the typhus group that included Rickettsia prowazekii and Rickettsia typhi, and the spotted fever group (SFG) composed by about 20 species [5]. The taxonomic position of other rickettsial species has remained unclear because of their genetic anomalies. Due to findings of intriguing variety of rickettsiae in arthropods and molecular analysis of rickettsial plasmids, the genus Rickettsia has been reclassified into SFG rickettsiae, typhus group rickettsiae, the transitional group, the Rickettsia bellii group, the Rickettsia canadensis group, and several basal groups [3,6]. However, some authors do not support the creation of the transitional group claiming that it is not monophyletic and is unhelpful as it does not take into account epidemiological criteria [1].
Tick-borne rickettsioses are caused by rickettsiae belonging to the SFG [4]. Rapid development of molecular methods brought reversed approach to tick-borne pathogen research, when disease cases are detected years after the tick-borne microorganism was first discovered [7]. There have been species of rickettisae detected in ticks years or decades before they became associated with human illness cases, e.g. Rickettsia monacensis, Rickettsia parkeri, Rickettsia massiliae and Rickettsia slovaca [4,8]. It is not clear if these novel tick-borne diseases were not noticed by physicians or whether they were absent. While it has been suggested that any novel described rickettsia from ticks should be considered a potential pathogen [5], many tick species just do not bite humans under natural conditions, or some rickettsial agents are just tick endosymbionts.
Recently, a novel SFG rickettsia has been found by molecular methods in bird-associated ticks. It was named 'Candidatus Rickettsia vini' and until now it has been detected in Spain, the Czech Republic, Slovakia and Turkey [9][10][11]. This bacterium has been molecularly detected mainly in Ixodes arboricola ticks, in which the prevalence is high (reaching 90-100 %) [11,12]. It has rarely been found in immature stages of Ixodes ricinus [9]. I. arboricola tick is widely distributed in the Palaearctic region. It lives in tree holes and nest boxes where it feeds on hole-breeding birds. Although this tick species does not represent a primary risk for humans, it shares several host species and overlaps in feeding period with Ixodes ricinus [13]. Therefore, tickborne microorganisms, including 'Ca. R. vini' , could be potentially bridged between these two tick species via cofeeding. Phylogenetic analysis based on partial sequences of four rickettsial genes (gltA, ompA, ompB, sca4) showed that 'Ca. R. vini' segregated closest to Rickettsia heilongjiangensis and Rickettsia japonica, supported by high bootstrap values [14]. The latter two species are causative agents of the Far East spotted fever (R. heilongjiangensis) and the Japanese spotted fever (R. japonica), and both have been reported in Asia [4].
In order to describe 'Ca. R. vini' as a new species, we isolated the bacterium in cell culture for the first time, and performed both molecular and phenotypical characterization of the isolates.

Field study in Breclav, Czech Republic
Free-living I. arboricola ticks were collected manually from nest boxes during after-breeding season in Breclav, Czech Republic (48°43'N, 16°54'E, 150 m above sea level, an oak-ash flood-plain forest), an area attractive to tourists. Nesting bird species had been previously identified during the breeding season using a bird guide book [15] and confirmed according to characteristic appearance of the nest during tick collecting. Ticks were identified to species according to Nosek & Sixl [16]. Collected ticks were brought alive to the laboratory and incubated at 12°C (relative humidity of 80 %) for 3 months and then at 24°C (relative humidity of 80 %) for 7 days before being subjected to the hemolymph test.

Hemolymph test and isolation of rickettsiae
Selected individuals were tested for the presence of Rickettsia-like structures using the hemolymph test [17]. Shortly, the distal part of a tick leg was cut, then a drop of hemolymph was dried on a microscope slide and stained using Gimenez method [18]. The whole-body remnants were immediately stored at -80°C to preserve living rickettsial organisms.
Isolation of rickettsiae from the tick samples was performed according to previous protocols [19] with some modifications. Briefly, ticks were surface-sterilized by immersion in iodine-alcohol for 10 min, washed in sterile water, macerated, and resuspended in 600 μl of brain heart infusion (BHI). For each tick sample, two shell vials with a confluent monolayer of Vero cells were each inoculated with 300 μl of the BHI suspension and then centrifuged for 1 h at 700× g and 22°C. The monolayers were washed and fed with 1 ml of minimal essential medium supplemented with 5 % of bovine calf serum (Hyclone, Logan, UT, USA) and 1 % of antibiotics and antifungal (penicillin, streptomycin and amphotericin B) and incubated at 28°C. Every 3 days, the medium was replaced by a new medium (without antibiotics and antifungal additives), and the aspirated medium was checked for the presence of Rickettsia-like organisms by Gimenez staining. If the result was positive, the monolayer of the shell vial was harvested and inoculated into a 25 cm 2 flask containing a monolayer of confluent uninfected Vero cells. Cells in the 25 cm 2 flask were checked by Gimenez staining until > 90 % of them were infected, when they were harvested and inoculated into 75 cm 2 flasks of Vero cells. The level of infection of cells was monitored by Gimenez staining of scraped cells from the inoculated monolayer. The rickettsial isolate was considered to be established in the laboratory after at least three passages through 75 cm 2 Vero cell flasks, each achieving a proportion > 90 % of infected cells [20].

Experimental infestations and inoculations
Selected larvae obtained from one egg cluster of I. arboricola were PCR-tested to confirm the presence of rickettsial DNA. Then, three tick naive chickens (denoted A, B and C) were each infested with 100 I. arboricola larvae from this cluster. Blood samples were collected from the chickens at the beginning of the infestation (day 0) and 21 days later. Two chickens (denoted D, E) and two male guinea pigs (denoted A, B), all tick naive, were each inoculated intraperitoneally with a suspension of ≈ 1 × 10 6 Vero cells infected with an I. arboricola rickettsial isolate derived from a fresh culture containing > 90 % infected cells. Blood samples were collected at 0 and 21 days after inoculation. The guinea pigs were examined daily for fever (if the rectal temperature was > 39.5°C) and scrotal reactions.

Molecular characterization
All whole-body remnants of the ticks used to inoculate shell vials and infected Vero cell 1st-4th passages were subjected to DNA extraction using the guanidine isothiocyanate technique, as described elsewhere [26], and DNA extracts were stored at -20°C until they were used as templates for polymerase chain reaction (PCR). DNA samples were tested by a battery of PCR protocols targeting fragments of four rickettsial genes: citrate synthase gene (gltA), the 190-kDa outer membrane protein gene (ompA), the 120-kDa outer membrane protein gene (ompB), and the 17-kDa protein gene (htrA) ( Table 1). Each PCR run included a negative control (2.5 μl of water) and a positive control (2.5 μl of DNA of Rickettsia parkeri strain NOD) samples. PCR products were purified by ExoSAP-IT® (USB), DNA-sequenced by Sanger dideoxy sequencing, and analyzed using BLAST [27] to determine similarities to other Rickettsia spp. available in GenBank, National Center for Biotechnology Information (NCBI) [28]. The DNA sequences obtained from the 4th passage-infected cells were submitted to the GenBank database (see below). Phylogenetic analyses were performed using the program MEGA version 6.06 in November 2015 [29]. The newly-generated partial DNA sequences (gltA, ompA, ompB, and htrA genes) were analyzed separately, and also concatenated. In both cases, nucleotides were aligned with the corresponding sequences of other Rickettsia species available in the GenBank database using MUSCLE algorithm as implemented in MEGA. The resulted alignment was checked and manually corrected. The evolutionary history was inferred using the Maximum Likelihood method based on the Tamura 3-parameter (I + G) model with 1000 replicates of random-addition taxa and tree bisection and reconnection branch swapping. All positions were weighted equally.

Morphology by light microscopy
Gimenez stained hemolymph smears were examined under oil immersion (at magnification of × 1000; 10× ocular and a 100× objective). Images of Rickettsia-like structures and adjacent Vero cells were captured using Leica Microscope DM4000-B. Etymology: The name vini has been proposed by Palomar et al. [9] who first detected molecularly this bacterium at La Rioja, a vineyard region in Spain. District of Breclav, the type-locality, is also an important area of vine production in the Czech Republic.

Description
Rickettsia vini n. sp. is a Gram-negative, nonmotile, obligately intracellular bacterium. The organism has a typical bacillary morphology with binary fission. It grows at 28°C on Vero cells in minimal essential medium with 5 % bovine calf serum supplement (Fig. 1). Sequencing of gltA, ompA, ompB, and htrA genes implies that this bacterium is significantly different from all recognized rickettsial species. It belongs to the SFG and is most closely related to R. japonica and R. heilongjiangensis. Rickettsia vini n. sp. is not pathogenic for chickens and guinea pigs through intraperitoneal inoculation, although it induces seroconversion in these animals (see below). The pathogenicity of this bacterium for vertebrate hosts, including humans, remains to be elucidated. Isolation of Rickettsia vini n. sp.
All 18 male ticks were subjected to the hemolymph test; of these three were found positive for Rickettsia-like organisms within their hemocytes, and subsequently subjected to the isolation of rickettsiae by the shell vial technique. Rickettsiae were successfully isolated from all three ticks and established in Vero cell culture (Fig. 1). The three isolates were designated as Rv-M1A, Rv-M2B, and Rv-M3B.

Experimental infestations, inoculations and serological tests
The chickens A, B and C that were infested with R. viniinfected I. arboricola larvae remained seronegative to all seven Rickettsia species, including R. vini antigens ( Table 2). A total of 15 to 17 engorged larvae were recovered from each chicken. Conversely, the chickens D, E that were inoculated with R. vini-infected Vero cells showed seroconversion for R. vini n. sp., R. rickettsii, and 'Ca. R. amblyommii' with titers ranging from 64 to 512 at 21 days after inoculation. Chicken E also seroconverted to R. bellii ( Table 2). None of the above five chickens showed apparent signs of disease during the present study.
The guinea pig A showed seroconversion to R. vini, R. rickettsii and 'Ca. R. amblyommii' with 512, 512 and 256 endpoint titers, respectively, 21 days after intraperitoneal inoculation of R. vini-infected Vero cells. The guinea pig B seroconverted only to R. vini n. sp., with a 128 endpoint titer ( Table 2). None of these guinea pigs developed fever, scrotal reactions or any other clinical alteration.
Phylogenetic analyses were inferred from the gltA, ompA, ompB, and htrA partial sequences, with each gene analyzed separately (Additional files 1, 2, 3 and 4). Then an analysis of a concatenated dataset was carried out on an alignment that included 2979 nt (1092, 590, 800, 497 nt for the gltA, ompA, ompB, and htrA genes, respectively). In all analyses, R. vini n. sp. segregated -, negative at the 1:64 serum dilution Abbreviation: NT not tested closest to R. japonica and R. heilongjiangensis cluster, which was supported by high bootstrap values (Fig. 2).

Discussion
This study described and characterized a new species of Rickettsia, R. vini n. sp., isolated from I. arboricola ticks collected in nest-boxes in the Czech Republic. This bacterium was first detected by PCR in I. arboricola and I. ricinus immature ticks collected from birds in La Rioja, a vineyard region in Spain [9] and named 'Ca. R. vini' [14]. To date, this bacterium has been detected molecularly in ticks in Europe and Turkey. The Palaearctic distribution of the type-species, the tick I. arboricola predicates possibly a similar wide occurrence of R. vini n. sp. Through molecular analyses (PCR detection), infection rates of R. vini in I. arboricola ticks up to 100 % have been reported [10,11,14]; however, we found only three out of 18 males positive for Rickettsia-like organisms using the hemolymph test. This may be caused by higher sensitivity of PCR detection, when compared to the hemolymph test, or/and because not all R. vini-infected ticks contain rickettsiae in their hemolymph. All PCRtested unfed larvae of I. arboricola from this study contained rickettsial DNA, indicating transovarial transmission of the rickettsial agent.
All animals inoculated intraperitoneally seroconverted after 21 days, sometimes with high homologous antibody titers to R. vini n. sp. ( Table 2). The guinea pig A showed cross-reactivity for R. rickettsii and 'Ca. R. amblyommii' antigens, while the guinea pig B reacted only to the homologous antigens. Both chickens D and E inoculated with R. vini-infected Vero cells showed homologous titers always equal or higher than heterologous titers. Cross-reactivities were observed with closely related species belonging to the SFG such as R. rickettsii and 'Ca. R. amblyommii' , although chicken E also reacted to R. bellii, a non-SFG agent. Cross-reactivity with lower titres for heterologous antigens has also been observed in experimental studies with guinea pigs intraperitoneally inoculated by Vero cells infected with R. monteiroi, R. bellii, R. rickettsii or R. canandensis [30]. Although R. felis is phylogenetically closer than R. bellii to R. vini, no crossreactivity with R. felis was observed. These results indicate that R. vini n. sp. possibly shares numerous antigenic constituents with other Rickettsia species, especially SFG members (Fig. 2). These findings are consistent with previous studies with mice, guinea pigs, dogs and opossums that were inoculated with different Rickettsia species [31][32][33][34].
Absence of clinical signs in R. vini-inoculated chickens D, E and guinea pigs A, B suggests that R. vini n. sp. is not pathogenic for these animals. None of the three chickens A, B, C infested by R. vini-infected larvae seroconverted, in contrary to chickens D, E that were inoculated with R. vini culture. While these results suggest a tick-symbiotical nature of R. vini, it is also possible that chickens (and guinea pigs) are just not susceptible to R. vini. If this is the case, the seroconversion of inoculated animals in the present study could be just a result of direct contact with bacterial antigens, rather than active infection. Such assumptions need to be confirmed in further studies. Finally, the non-susceptibility of chickens in the present study could be linked to the inoculation route (intraperitoneal inoculation), since other rickettsial agents were shown to cause skin lesions through intradermal inoculations of experimental animals, in contrast to no clinical alterations when the same agents were intraperitoneally inoculated [35,36]. Moreover, our results of animal inoculations do not exclude a possible susceptibility of the bird hosts of I. arboricola to R. vini under natural conditions.
The phylogenetic analyses of four rickettsial genes showed that R. vini n. sp. belongs to the SFG and is most closely related to R. japonica and R. heilongjiangensis, which is in compliance with previous studies [11,14]. In this study, we amplified different and longer fragments of the ompB and htrA genes of R. vini that have not been previously published. R. japonica and R. heilongjiangensis are associated with various tick vectors and mammal reservoirs [4]. It has been proposed that a new Rickettsia species should not show > 99.9 %, > 99.2 %, and > 98.8 % similarity for the gltA, ompB, and ompA genes, respectively, with the most homologous validated species [37]. Similiarity values of R. vini DNA sequences with the most closely related validated species are 99.7 % for the gltA gene of R. heilongjiangensis (CP002912), 96.8 % for the ompB gene of R. japonica (AP011533), and 97.1 % for the ompA gene of R. heilongjiangensis (CP002912). These comparison values also support the recognition of R. vini as a new species.