Ixodid ticks naturally harbour a variety of bacterial symbionts that may be obligately or facultatively intracellular and are transovarially transmitted. These include species of the genera Rickettsia, Coxiella, Midichloria and Spiroplasma that occur with high frequency (1–4) and less common or well-characterised species of the genus Francisella (1, 3) and Occidentia (5). The insect symbionts Cardinium, Wolbachia, Arsenophonus and Rickettsiella have also been detected in or isolated from ticks (3, 6–10) but it is unclear whether or not their presence results from parasitism by insects such as the wasp Ixodiphagus hookeri (7, 9) or cohabiting mites (author's unpublished observations), and they are not known to be transovarially transmitted in ticks. Most studies of occurrence of bacterial symbionts in ticks are based on molecular detection in DNA extracted from individual or pooled ticks sampled directly from the field. Some recognised or putative tick symbionts have been isolated into culture, in either mammalian or tick cells; these include several species of Rickettsia (11–16), Francisella (17), several strains of Spiroplasma (10, 18–21) and one isolate each of Arsenophonus, Occidentia and Rickettsiella (5, 8, 10). In all cases, the unfed or partially-fed ticks had been collected from the field, and bacteria were isolated directly from homogenised/macerated whole ticks or aseptically-dissected internal organs, or from eggs laid by engorged female ticks.

Ixodes persulcatus, a tick species distributed widely from the eastern Baltic coast to Japan (22–25), has been reported to harbour symbionts including the Montezuma agent, now called Candidatus Lariskella arthropodarum (26–28), Coxiella and Spiroplasma spp. (29), as well as human and livestock pathogens including tick-borne encephalitis virus (TBEV), Kemerovo virus, Alongshan virus, Anaplasma phagocytophilum, Candidatus Neoehrlichia mikurensis, Ehrlichia muris, Rickettsia helvetica, Rickettsia heilongjiangensis, Candidatus Rickettsia tarasevichiae, Borrelia miyamotoi, Borrelia burgdorferi sensu lato, Theileria equi and several species of Babesia (25, 30–38). Co-infections with multiple pathogens and symbionts are common (29, 35, 39). There is a single report of isolation into culture of bacterial symbionts from Japanese I. persulcatus: R. helvetica and a Spiroplasma were isolated from field-caught adult male ticks into an Ixodes scapularis cell line (10).

Research on transmission of tick-borne pathogens of medical and veterinary interest depends largely on ticks maintained in laboratory colonies. However, few studies have assessed such ticks for presence of symbionts, despite the potential influence of the latter on the ability of ticks to harbour (40) and/or transmit pathogens. Prevalence of Candidatus Midichloria mitochondrii determined by molecular methods was found to be lower in Ixodes ricinus ticks from laboratory colonies than in field ticks, and to decrease (albeit in a small sample size) with increasing numbers of tick generations (2). A subsequent study, using a more sensitive assay, revealed the presence of extremely low levels of Ca. M. mitochondrii DNA in 60% of >10th generation laboratory colony I. ricinus (41). Ixodes arboricola were screened for bacterial symbionts by PCR and higher incidences were found in field-collected ticks than in laboratory colony ticks of three genera: Rickettsiella (28.0 vs. 0%), Midichloria (1.3 vs. 0%) and Spiroplasma (16.0 vs. 5.6%) (3). Both groups harboured similarly high levels of Rickettsia (96.0 vs. 100%), suggesting that transovarial transmission was highly efficient for Rickettsia, less efficient for Spiroplasma and might not occur for Rickettsiella. Both lower absolute numbers of bacteria including the symbionts Spiroplasma and Midichloria, and more limited diversity of bacterial species, were reported in midguts of I. ricinus ticks from a laboratory colony compared to wild-caught ticks, and extremely low numbers of bacteria (<100 organisms per midgut) were found in Rhipicephalus microplus ticks from a closed colony in Brazil (42). A bacterial symbiont, later identified as a Cardinium sp. (43), was isolated from first-generation adult I. scapularis reared in the laboratory from field-caught adults (6). Rickettsia raoultii was isolated from eggs laid by the first generation of adult Dermacentor reticulatus reared in the laboratory from field-caught ticks (21). However, we could not find any report of in vitro isolation of a bacterial symbiont from laboratory colony ticks maintained for additional generations.

Here we report isolation and preliminary genetic characterisation of a Spiroplasma from third-generation adult male and female I. persulcatus, originally collected in Siberia (Irkutsk Oblast, Russian Federation) and maintained in a laboratory colony for over 4 years.

When the tick cell cultures were examined by Giemsa-stained cytocentrifuge smear on day 53 p.i., bacteria resembling Spiroplasma were seen in cells that had received organs from 1/5 male and 5/5 female I. persulcatus ticks (Table 1). In all cases, the Spiroplasma were intracellular and concentrated in cytoplasmic vacuoles, but the appearance differed between the various tick cell lines (Figure 1). In the R. microplus cell lines (Figures 1A,B) and IDE2 (not shown), most vacuoles containing Spiroplasma also contained homogenous, light blue- or pink-staining background material, whereas in IRE11 cells (Figures 1C,D) such material was absent in most vacuoles containing Spiroplasma. It was not possible to determine whether this was due to differences between the cell lines or the Spiroplasma isolates, although in previous studies background material was visible in Spiroplasma-containing vacuoles in cells of the tick cell lines BME/CTVM23 and DALBE3 (21) but not of the tick cell lines IRE11, IRE/CTVM19 or IDE2 (20).

PCR amplification of fragments of the Spiroplasma-specific 16S rRNA and rpoB genes from DNA extracted on day 65 p.i. confirmed the presence of Spiroplasma in the six microscopically-positive cultures, and in both cases failed to amplify any products from DNA extracted from the four microscopically-negative cultures (Table 1). To determine the Spiroplasma infection rate in adult ticks of the parent colony, the 10 ticks remaining in Liverpool (four males, six females) and a further 41 ticks (12 males, 12 females fed as nymphs on guinea pigs and five males, 12 females fed as nymphs on gerbils) from the same generation of the BCCAS colony were screened using the Spiroplasma 16S rRNA and rpoB PCR assays. All of the ticks were positive for Spiroplasma by one or both assays, and amplification and sequencing of a 430 bp fragment of the tick 16S rRNA gene confirmed the species identity of the ticks tested in Liverpool as I. persulcatus (99.8% similarity to I. persulcatus from Omsk, Siberia, Russia, Genbank accession no. MH790201.1).

Sequence analysis revealed that, for the Spiroplasma 16S rRNA gene, all six culture isolates (designated Irkutsk1-6) and five representative tick samples screened in Liverpool were identical to each other, and identical to eight representative tick samples screened at BCCAS apart from one ambiguous nucleotide at position 105 (Table 2A). All sequences showed 99.8% similarity (99.5% query cover) to the only sequence from an I. persulcatus-derived Spiroplasma available in Genbank at the time of writing (LC388762.1) (10); interestingly, the only mismatch between our sequences and that of the Japanese isolate (10) was also at position 105 (Table 2A). The 16S rRNA sequence of the Irkutsk strains isolated from I. persulcatus was identical to several other Spiroplasma strains isolated from hard ticks: Spiroplasma sp. Bratislava 1 (KP967685, from Slovakian I. ricinus), Spiroplasma sp. 1033 (LC388770, from Japanese Haemaphysalis kitaokai) and Spiroplasma sp. 135 (LC388760 and LC388759, from Japanese Ixodes monospinosus) (10, 20) (Table 2A). Moreover, the Irkutsk 16S rRNA sequence showed 99.1–99.3% similarity to those of spiroplasmas isolated from North American Ixodes pacificus (Spiroplasma ixodetis, NR_104852), Spanish Dermacentor marginatus (Spiroplasma sp. strain DMAR11, MG859280) and Dutch D. reticulatus (Spiroplasma sp. strain DRET8, MG859282) (21, 55) (Table 2A). For the rpoB gene, all sequences obtained from the six culture isolates and 11 representative whole ticks were identical. At the time of writing, we could not find any published rpoB sequences from I. persulcatus-derived Spiroplasma for comparison, and most of those derived from other hard tick species were shorter than 600 bp. Considering the query cover higher than 99%, the rpoB sequence of the Irkutsk strains showed 99.9% similarity to the spiroplasmas isolated from I. ricinus (Spiroplasma sp. Bratislava1, KP967687) and Dermacentor spp. (Spiroplasma strain DMAR11, MG859278 and Spiroplasma strain DRET8, MG859277) (Table 2B), and 99.3% similarity to S. ixodetis (DQ313832). With a query cover of 43%, the sequences amplified in this study were identical to shorter sequences from spiroplasmas detected by PCR in other hard tick species (GenBank accession numbers MK267073-MK267077, MK267081-MK267085 and MK267097) (4), and also to Spiroplasma strain DMAR11 (MG859278) and Spiroplasma sp. Bratislava1 (KP967687) that showed polymorphisms in the longer gene fragment (Table 2B).

Phylogenetic analysis based on 16S rRNA sequences derived from Spiroplasma sp. strain Irkutsk1 and two representative whole ticks revealed that the I. persulcatus spiroplasmas clustered together with, but were not identical to, S. ixodetis (55) and most of the spiroplasmas from other hard ticks (Figure 2A). Similarly, the phylogeny obtained with the rpoB sequences showed tight clustering of the I. persulcatus Spiroplasma with most other tick-borne Spiroplasma sequences (Figure 2B).

The Spiroplasma 16S rRNA and rpoB gene sequences obtained in the present study were deposited in GenBank under accession numbers MW492370, MW498416, MW498417, MW528409-MW528411.

Colonisation in the laboratory has been previously reported to result in decrease or loss of the microbial symbiont Ca. M. mitochondrii in I. ricinus (2, 41), whereas Coxiella-like endosymbionts were detected at high prevalence in Ornithodoros rostratus, Amblyomma americanum, Dermacentor silvarum and R. microplus ticks maintained in laboratory colonies for unspecified numbers of generations (55). In the case of the I. persulcatus Spiroplasma in the present study, after three generations in the laboratory, 100% of whole adult ticks (21 males, 30 females) were PCR-positive for this endosymbiont. Moreover, 5/5 female ticks and 1/5 male ticks harboured sufficient levels of viable bacteria to allow in vitro isolation in tick cell lines. Admittedly, the sensitivity of this technique for detection of infection with Spiroplasma is unknown, so it is possible that the remaining four male ticks could also have harboured Spiroplasma but either at a level insufficient to allow isolation, or in an organ or tissue that was inadvertently not included in the inoculum, or in a state of viability not conducive to in vitro isolation. Tissue tropism of the symbiont Ca. M. mitochondrii in I. ricinus ticks was found to be highly specific to certain organs (56); further study is needed to determine the tissue tropism of Spiroplasma spp. in Ixodes spp. ticks.

The prevalence of Spiroplasma in the original Siberian field ticks from which the laboratory colony was initiated in 2015 is unknown, and therefore it is impossible to determine whether colonisation resulted in maintenance of, or increase in, the infection rate. However, it can be concluded that laboratory colonisation does not have a negative effect on occurrence of Spiroplasma in I. persulcatus, at least over three generations. There have only been two reports of detection of Spiroplasma in I. persulcatus ticks. Using 16S amplicon pyrosequencing, Spiroplasma were detected in salivary glands of at least 5/6 male and 5/6 female unfed I. persulcatus collected in the field in Japan (29), and Spiroplasma was successfully isolated into arthropod cell culture from 1/30 questing adult I. persulcatus collected in Japan (10). In contrast, no Spiroplasma or other mollicutes were recorded in questing I. persulcatus collected in the Novosibirsk area of Russia and examined by 16S metagenomic profiling (four pools of 87–120 ticks) (32), and Spiroplasma were not detected by species-specific 16S rRNA PCR in three questing I. persulcatus ticks collected in Finland (3).

Spiroplasma infection rates determined by molecular analysis vary widely in other Ixodes spp. ticks collected from the field in different geographical areas. Prevalence of Spiroplasma in I. ricinus nymphs and adults ranged from 0–0.3% in UK [(21); author's unpublished data] through 5–6% in Hungary and Czech Republic (57, 58) to 23–30% in The Netherlands, France, Switzerland and Spain (3, 59). The bacterium was detected in 10% of Ixodes uriae from Russia (3), 14–16% of I. arboricola from Belgium (3, 60) and 100% of Ixodes ovatus from Japan (29). The type species S. ixodetis was isolated from 7/30 pools representing 600 I. pacificus from USA, suggesting a prevalence between 1.2 and 23% (61). Considering this level of variation between species and geographical location, the infection rates of 100% in whole ticks and 60% following in vitro isolation in the present study suggest that Spiroplasma survives well-under laboratory colony conditions, in both male and female I. persulcatus ticks.

Presence of Spiroplasma in laboratory colony ticks could affect their ability to be infected experimentally with, and/or transmit, tick-borne pathogens, and therefore their use in this context. A recent study examined correlations between presence of Spiroplasma in field-collected I. ricinus in Switzerland, and presence in these ticks of bacterial pathogens and symbionts (40). Negative correlations were found between Spiroplasma and the pathogens Rickettsia spp. and Borrelia valaisiana in individual I. ricinus, but positive correlations were found between Spiroplasma and the symbionts Lariskella and Rickettsiella at the population level. Further studies are needed to examine whether presence of Spiroplasma in Ixodes spp. ticks has any effect on acquisition, replication or transmission of tick-borne arboviruses such as TBEV or protozoa such as Babesia spp., or indeed any effect on the viability of the ticks themselves.

The molecular analysis revealed almost no differences between the Spiroplasma isolated from colonised I. persulcatus of Russian origin and cultured for 2.5 months in cell lines derived from heterologous tick species (I. ricinus, I. scapularis and R. microplus), Spiroplasma DNA detected in whole ticks from the same colony and the Spiroplasma isolated into I. scapularis cells from Japanese I. persulcatus (10). Ambiguity was seen in a single nucleotide in the ~476 bp fragment of the 16S rRNA gene amplified in the present study, and the same nucleotide showed a difference when compared with the sequence from the Japanese isolate. The rpoB gene fragment analysed in our study was longer than the 16S rRNA sequences, providing more phylogenetic information, although the shorter length of most of the published sequences from other tick species (4) reduced the coverage available for comparison. The overall topology of the tree and the relationship between strains in the tick-borne Spiroplasma branch were very close to the results based on the 16S rRNA gene, although neither of these gene fragments are sufficient to confidently separate Spiroplasma strains or species. Nevertheless, as reported previously (21) it is clear that the spiroplasmas harboured by different Ixodes spp. ticks are not identical, and also differ from those harboured by Dermacentor spp. ticks from broadly contiguous geographic regions.

In conclusion, our study has shown that efficient vertical transmission of Spiroplasma can be maintained in I. persulcatus ticks under laboratory colony conditions for at least three generations, and has confirmed that co-cultivation of internal organs with tick cell lines is a simple and effective technique for in vitro isolation of intracellular tick symbionts such as Spiroplasma spp. Further molecular analysis of the cultured Spiroplasma strains derived from I. persulcatus, either by Sanger sequencing of additional genes or by whole genome sequencing, is required to clarify the phylogenetic relationships between them and Spiroplasma harboured by I. persulcatus of different geographical origins and by other tick species, and to facilitate an accurate taxonomic classification of these genotypes.

The original sequences generated for this study are publicly available in the NCBI Genbank repository under accession numbers MW492370, MW498416, MW498417, MW528409-MW528411.

The animal study was reviewed and approved by Czech Academy of Sciences (approval no. 161/2010).

AB, VH, JE, TV, MP, and LB-S carried out the experimental work. AB, VH, AP, and LB-S analysed the data. JC, IK, and DR carried out the field work. LB-S conceived the study and drafted the manuscript. AB, VH, JC, DR, and AP revised the manuscript. All authors reviewed and agreed to the final version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.