Microbial communities and symbionts in the hard tick Haemaphysalis longicornis (Acari: Ixodidae) from north China

Background Close relationships between ticks and microbial communities are important for tick fitness and pathogen colonization and transmission. Haemaphysalis longicornis, distributed widely in China, can carry and transmit various pathogens and pose serious damages to public health and economics. However, little is known about the broader array of microbial communities and symbionts in H. longicornis under natural conditions. In the present study, we investigated the composition of bacterial communities associated with H. longicornis and evaluated the putative symbionts. Methods The eubacterial 16S rRNA gene clone libraries of H. longicornis were constructed and analyzed by restriction fragment length polymorphism (RFLP) and DNA sequencing. In addition, diagnostic PCR was performed to assess the prevalence, vertical transmission and infection sites of the symbionts in H. longicornis. Results Vertically-transmitted symbionts, potential pathogens and allochthonous nonpathogenic bacteria were identified from the field-collected H. longicornis. Three types of symbionts (Coxiella-like, Arsenophonus-like and Rickettsia-like symbionts) were identified in a single host simultaneously. A series of analyses revealed the vertical transmission, prevalence, and infection sites of these symbionts. However, only Coxiella-like bacteria were transmitted stably in the laboratory-reared ticks. In addition, we identified a novel Coxiella-like agent with 95.31% sequence similarity to the taxon described previously. Conclusions The present study demonstrated that natural H. longicornis harboured a diverse array of microbial communities. Three types of symbionts were identified in a single host simultaneously. Moreover, high prevalence, vertical transmission and the infection sites supported an obligate symbiotic association between Coxiella symbiont and its host. The role of Coxiella symbiont in the host fitness and the interaction among microbial communities remained to be elucidated. Our investigation of microbial communities in the ticks revealed the complexity of ecological interactions between host and microbe and provided insight for the biological control of ticks.


Background
As obligate blood-feeding arthropods, ticks feed on terrestrial vertebrates and are one of the most important disease vectors worldwide. Ticks are involved in carrying and transmitting a diverse group of pathogenic bacteria, which exert severely negative impacts on human health, livestock production as well as wildlife [1][2][3]. Hard ticks (Ixodidae) usually have three life stages (larva, nymph, adult) in their lifespan and feed on distinct host species at different developmental stages. Thus, they are prone to acquiring various pathogens and spreading among vertebrate species during the whole life cycle [4].
In addition to pathogens, vertically-transmitted symbionts and foreign nonpathogenic bacteria also colonize the ticks [5][6][7]. Symbionts are closely associated with the host fitness [8,9] and transmission and virulence of pathogens [10,11]. To date, it is well-known that ticks harbour various symbionts, including Coxiella [12], Francisella [13], Wolbachia [14], Rickettsia [15], Arsenophonus [16], Candidatus Midichloria mitochondrii [17]. Moreover, some of them have crucial effects on the biological characteristics of ticks. For example, the obligate Coxiellalike symbiont manipulates the reproduction of the host Amblyomma americanum [18][19][20]. Ca. Midichloria mitochondrii, the symbiont of Ixodes ricinus, has the unique ability to enter and destroy mitochondria within ovarian cells of host [21,22], which is coupled with the process of engorgement and molt [23,24]. In addition, colonization of Dermacentor variabilis by R. peacockii can prevent secondary infection with other Rickettsia bacteria, including pathogenic species [25]. Clay [5] found that the infection frequency of Arsenophonus was negatively correlated to Rickettsia in A. americanum, indicating that infection with symbionts may affect the microbial community structure and their interaction.
Ticks could also obtain allochthonous nonpathogenic bacteria by sucking host blood and contact with the natural environment [26][27][28]. Although allochthonous bacteria are not directly related to the pathogenicity of hosts, they can reflect the habitats of hosts and even have the potential to impact on the density and transmission of host-borne agents [29]. Overall, diverse microbial communities can exert great impact on the fitness of ticks and the colonization and transmission of tick-borne pathogens.
In the present study, we investigated the composition of bacterial communities associated with H. longicornis using PCR-RFLP and DNA sequencing approaches, and evaluated the putative symbionts.

Sample collection and storage
H. longicornis samples were collected in Xiaowutai National Natural Reserve Area in China by flag dragging. A total of 500 ticks were collected and stored at −80°C. We also collected 1,000 ticks, which were reared on rabbits as described by Liu et al. [48]. Colonies of these ticks were reared on the ears of rabbits. Rabbits were maintained in a room with 50-55% relative humidity (RH) at 25-27°C and exposed to daylight. After detachment, ticks were collected and incubated in cottonplugged glass tubes filled with folded filter paper in an incubator with 75 ± 5% RH and 6/18 h of L/D cycle at 26 ± 1°C. The protocol of all animal experiments was approved by the Institutional Animal Care and Use Committee of Hebei Normal University.

DNA extraction
The genomic DNA for constructing eubacterial 16S rRNA gene libraries were extracted from a group of adults (10 females and 10 males). To assess the prevalence, vertical transmission and infection sites, DNA was isolated from individual field-collected adults, and pooled ticks at different developmental stages (500 eggs, 200 larvae and 50 nymphs) and different tissues (ovaries, salivary glands, Malpighian tubules and midguts), respectively. Before DNA extraction, the whole tick samples were sterilized as described previously [5], and dissected tissues were washed in sterile phosphatebuffered saline (PBS) (137 mM NaCl, 2.7 mM KCl, 4.3 mM Na 2 HPO 4 · 7H 2 O, 1.4 mM KH 2 PO 4 , pH 7.4) for three times. Subsequently, all samples were snap-frozen in liquid nitrogen and purified using the DNeasy Tissue Kit (Qiagen, Germany) according to the manufacturer's instructions.

Eubacterial 16S rRNA gene library construction, RFLP analysis and sequencing
The eubacterial 16S rRNA gene clone library was constructed by amplifying about 1500 bp fragment of 16S rRNA gene using eubacterial universal primers 27 F/ 1492R [49]. Purified PCR products were cloned into pEASY-T1 vector (TransGen, China) and transformed into Escherichia coli TOP10 competent Cell (TransGen, China). The positive clones were subjected to restriction fragment length polymorphism (RFLP) analyses using both HaeIII and RsaI restriction endonucleases. Sequencing of the positive clones with different restriction fragment patterns were performed by Sangon Biotech Company (China).

Phylogenetic analyses
The 16S rRNA gene sequences of the most closely related species were retrieved from GenBank and then aligned. The similarity analyses were performed using the CLUSTAL_W program [50]. The phylogenetic tree was constructed using neighbour-joining method implemented in MEGA4.0 [51]. The stability of the tree clustering was evaluated by means of a bootstrap analysis of 1000 datasets [52].

Diagnostic PCR
To assess their prevalence, vertical transmission and infection site of symbionts, diagnostic PCR assay was performed with four sets of specific primers (

Diversity of bacterial communities from field-collected H. longicornis
Our results revealed diverse microbial communities associated with field-collected H. longicornis (Table 2). In the 16S rRNA gene clone libraries, the most abundant sequences (belonged to Coxiella and designated as CLS-Hl) in both males and females share 99.5% similarity to the symbiont of H. longicornis (GenBank: AB001519) [53]. In females, the second abundant sequence (designated as NCLS-Hl) belonged to Coxiella, sharing 95.3% similarity to Coxiella-like symbionts of Rhipicephalus sanguine (GenBank: D84559). Following NCLS-Hl, sequences belonged to Rickettsia (designated as RLS-Hl) and E. coli. Rickettsia bacteria shared the highest sequence similarity (99.8%) with symbiotic Rickettsia of D. varibilis (GenBank: U55820) [54]. In males, the second abundant microbes (designated as ALS-Hl) belonged to Arsenophonus, sharing over 99.5% and 97.9% similarity to symbionts of D. silvarum (GenBank: JN866582) [55] and D. variabilis (GenBank: AY265342) [56], respectively. Following ALS-Hl, there were RLS-Hl and NCLS-Hl. Interestingly, the top four abundant bacteria in females were consistent with those in males.

Symbionts
OmpA gene sequence analyses revealed that RLS-Hl belonged to Ca. Rickettsia hebeiii, with 99.1% sequence similarity [57]. To confirm vertical transmission of symbionts, eggs and the first generation of laboratory-reared ticks (larvae, nymphs and adults) was tested by diagnostic PCR assay. The putative symbionts were found in all samples detected, with the exception of NCLS-Hl. These results indicated that CLS-Hl, ALS-Hl and RLS-Hl were inherited from progeny to the next generation, while NCLS-Hl was not (Figure 1). To determine the stability of vertical transmission, we investigated the presence of each bacterium in the 7th generation of laboratory-reared ticks (F7). The results showed only CLS-Hl was detected in F7 ticks, while ALS-Hl and RLS-Hl were not, suggesting that CLS-Hl can be maintained stably in laboratory-reared ticks.

Discussion
In this study, we showed that the natural H. longicornis harboured a diverse group of bacterial species. Previous studies have shown that some tick species were colonized with complex microbial communities, including A. americanum, I. ricinus, I. scapularis and Rh. microplus [5][6][7][26][27][28]58,59]. Investigation of microbiota in the ticks revealed the complexity of ecological interactions between host and microbe and provided insight for the biological control of ticks.

Symbionts
Interestingly, some sequences were phylogenetically related to previously reported tick-associated bacteria, including (1) two types of Coxiella-like bacteria, CLS-Hl and NCLS-Hl, (2) Arsenophonus-like symbiont, ALS-Hl, (3) Rickettsia-like symbiont, RLS-Hl. It is now widely accepted that the co-infection with multiple symbionts  occurs in many arthropod hosts [60]. However, coinfection with symbionts has been demonstrated only in a few tick species, such as A. americanum [5], Ornithodoros moubata [61], D. silvarum [55] and Rh. sanguineus [62]. We reported here for the first time that Coxiella-like, Arsenophonus-like and Rickettsia-like symbionts were detected simultaneously in wild H. longicornis. Interestingly, we identified two different types of Coxiella. CLS-Hl, with 95.1% similarity to the C. burnetii, shared 99.5% sequence identify with the symbionts of H. longicornis previously reported [39,53]. However, NCLS-Hl, with 95.1% and 95.3% similarity to the C. burnetii and symbionts of Rh. sanguine, respectively, was regarded as a novel Coxiella-like agent and has never been described before. Phylogenetic analyses revealed (Figure 3) that CLS-Hl and NCLS-Hl form an individual clade, although both of them belong to the cluster of hard tick species (Ixodidae). CLS-Hl was clustered with symbionts of Haemphysalis (H. longicornis and H. shimoga), while NCLS-Hl was grouped together with the symbionts of Rhipicephalus (Rh. sanguineus and Rh. turanicus). However, the determination of vertical transmission does not support that NCLS-Hl was a symbiont. Further studies are needed to confirm the pathogenicity of this novel Coxiella agent to vertebrate hosts. In ticks, Coxiella bacteria are the most common agent, exhibiting diverse genotypes in different tick species [5,12,[63][64][65][66][67][68]. By comparing the phylogeny between the tick mitochondrial 16S rRNA gene and the Coxiella agents carried by ticks, Almeida [12] showed that ticks were phylogenetically associated with their Coxiella agents, suggesting that Coxiella symbionts and ticks shared a long co-evolution process. In H. longicornis, Coxiella symbionts from different areas, such as Japan [53] and Korea [39], shared identical genotypes. This wide geographical distribution indicates that Coxiella symbionts form ancient symbiotic associations with their host before spreading.  Previous studies suggested that Coxiella symbionts in A. americanum, with 100% infection, vertical transmission, reduced-genome, regulation of the reproduction of the host, was closely related to the fitness of the ticks. Besides, in O. rostratus [12], D. silvarum [55] and Rh. sanguineus [62], Coxiella-like symbionts exhibited nearly perfect prevalence and were presumed as obligate symbionts. In this study, 100% infection rate, perfect vertical transmission, colonization in specialized tissues, CLS-Hl was regarded as obligate and contributed to the fitness of ticks.
Arsenophonus is one of the four major inherited symbionts of arthropods with an infection rate of 5% [60]. Arsenophonus has been reported to infect A. americanum [5], D. variablilis [16], D. andersoni [16] and D. silvarum [55]. Interestingly, the sequence of ALS-Hl was identical to that of D. silvarum collected from the same place as the H. longicornis, possibly suggesting horizontal transmission of Arsenphonus-like symbionts between different genera of ticks. Previous studies also provided evidence of horizontal transmission of symbionts between D. variabilis and D. andersoni [16]. Therefore, Arsenophonus possessed not only intragenus, but also intergenus horizontal transmission.
Rickettsia, with both symbiotic and pathogenic lifestyle, inhabits multiple tick species. Pathogenic rickettsiae have been extensively investigated, while only few studies on symbiotic rickettsiae have been reported. To date, Rickettsia-like symbionts have been identified in D. variabilis [54], D. andersoni [69], D. silvarum [55] I. scapularis [15] and A. americanum [5]. In this study, we showed molecular evidence for the vertical transmission of the Rickettsia-like symbiont (Ca. Rickettsia hebeiii) in H. longicornis. Moreover, high prevalence in the natural environment suggested interaction between RLS-Hl and their hosts.
However, RLS-Hl and ALS-Hl were not detected in F7 laboratory-reared ticks, suggesting that they cannot be maintained stably from one generation to the progeny in the laboratory. This phenomenon was also observed in I. ricinus, whose symbionts IricES1 were lost from female adults to progeny in the laboratory [70]. The transmission efficiency of symbionts decreased significantly in the laboratory, which was partially due to the competition among symbionts and bottlenecks during vertical transmission [71]. Alternatively, transmission reduction may be due to temperature differences between natural and laboratory environments [72].

Conclusions
A diverse array of microbial communities were identified from field-collected H. longicornis. Three types of symbionts were detected in a single host simultaneously. Moreover, analysis on the prevalence, vertical transmission and infection sites supported obligate symbiotic association between Coxiella symbionts and its host. The role of Coxiella symbionts in the host fitness and the interaction among microbial communities remains to be elucidated. Our investigation of microbial communities in the ticks revealed the complexity of ecological interactions between host and microbe and provided insight for the biological control of ticks.

Competing interests
The authors declare that they have no competing interests.

Authors' contributions
Liu L-M and Liu J-Z conceived and designed the study, constructed 16S rRNA gene libraries and performed RFLP analyses, drafted the manuscript, and critically revised the manuscript. Liu J-N carried out RFLP analyses and diagnostic PCR. Liu Z, Yu Z-J, Xv S-Q and Yang X-H, participated in sample collection, study implementation and data collection and helped to revise the manuscript. Li T, Li S-S and Guo L-D participated in sample collection and tick feeding. All authors read and approved the final manuscript.