Transmission of Bartonella henselae by Ixodes ricinus

The causative agent of cat-scratch disease in humans can be transmitted by this tick through saliva.

Bartonella spp. are facultative intracellular bacteria associated with several emerging diseases in humans and animals. B. henselae causes cat-scratch disease and is increasingly associated with several other syndromes, particularly ocular infections and endocarditis. Cats are the main reservoir for B. henselae and the bacteria are transmitted to cats by cat fl eas. However, new potential vectors are suspected of transmitting B. henselae, in particular, Ixodes ricinus, the most abundant ixodid tick that bites humans in western Europe. We used a membranefeeding technique to infect I. ricinus with B. henselae and demonstrate transmission of B. henselae within I. ricinus across developmental stages, migration or multiplication of B. henselae in salivary glands after a second meal, and transmission of viable and infective B. henselae from ticks to blood. These results provide evidence that I. ricinus is a competent vector for B. henselae. B artonella spp. are facultative intracellular bacteria associated with several emerging diseases in humans and animals (1). Domestic animals and wildlife represent a large reservoir for Bartonella spp., and at least 10 species or subspecies have been reported to cause zoonotic infections. B. henselae causes cat-scratch disease, possibly the most common zoonosis acquired from domestic animals in industrialized countries and is becoming increasingly associated with other syndromes, particularly ocular infections and endocarditis (2)(3)(4)(5)(6). Although cat fl eas are well-established vectors for B. henselae (7)(8)(9)(10), transmission by other arthropods, in particular ticks, has been suggested (11)(12)(13). Ixodes ricinus is the most widespread and abundant ixodid tick in western Europe and is frequently associated with bites in humans. It is a vector of emerging zoonotic pathogens including Borrelia burgdorferi sensu lato (14), Anaplasma phagocytophilum (15), and Babesia spp. (16).
Direct proof of transmission of Bartonella spp. by a tick was reported by Noguchi in 1926 (17), who described experimental transmission of B. bacilliformis (cause of Oroya fever) to monkeys by Dermacentor andersoni. In this study, ticks were allowed to feed on infected monkeys for 5 days. After removal, partially engorged ticks were placed on healthy monkeys in which disease then developed. This study showed that ticks could acquire and transmit the bacteria but did not demonstrate their vector competence or transtadial transmission throughout the tick's life cycle. Since this early study, the role of ticks in Bartonella spp. transmission has been strongly implied but never defi nitively demonstrated. Bartonella spp. DNA was detected in questing and engorged nymphs and adults Ixodes spp. collected in North America, Europe, and Asia (13,(18)(19)(20)(21)(22)(23)(24)(25)(26). If one considers that ixodid ticks feed only once per stage, Bartonella spp. DNA in questing ticks suggests transtadial transmission of these bacteria.
Other observations support Bartonella spp. transmission by ticks. Co-occurrence of Bartonella spp. with known tick-borne pathogens such as B. burgdorferi sensu lato, A. phagocytophilum, or Babesia spp. is not a rare event in ticks and hosts (13,19,24,27). A study conducted in a veterinary hospital in the United States (California) demonstrated that all dogs with endocarditis and infected with Bartonella spp. were also seropositive for A. phagocytophilum (28). In humans, several case studies have reported patients with concurrent Bartonella seropositivity and detection of Bartonella spp. DNA in their blood, along with B. burgdorferi infection of the central nervous system after tick bites (11,29). Moreover, Bartonella spp.

Transmission of Bartonella henselae by Ixodes ricinus
DNA has been detected in human blood cells after a tick bite (30), and 3 patients with B. henselae bacteremia who had no history of contact with cats but had sustained tick bites were reported in Texas (12). Finally, tick exposure was determined to be a risk factor associated with B. vinsonii seropositivity in dogs (31). Because Bartonella spp. are emerging human pathogens and Ixodes spp. can transmit a large spectrum of pathogens to humans, the capability of Ixodes spp. in transmitting human pathogenic Bartonella spp. should be determined. We used a membrane-feeding technique to infect I. ricinus with B. henselae, and investigated transtadial and transovarial transmission of viable and infective bacteria and putative transmission from tick saliva to blood during artifi cial blood meals.

Materials and Methods
Ticks I. ricinus ticks were collected from the forest of Gâvre (Loire-Atlantique, France) in 2006 by fl agging vegetation as described (32). Ticks were reared and maintained in chambers with a relative humidity of 80%-90% at 22°C before feeding on artifi cial skin. A total of 217 whole ticks were tested for Bartonella DNA.

Culturing of B. henselae
B. henselae (Houston-1 ATCC 49 882) were grown on 5% defi brinated sheep blood Columbia agar (CBA) plates, which were incubated at 35°C in an atmosphere of 5% CO 2 . After 10 days, bacteria were harvested, suspended in sterile phosphate-buffered saline (PBS), and used immediately for artifi cial feeding of ticks.

Housing of Cats
Three healthy male European cats were used (12-14 months of age, weight = 3.2-4.6 kg at the beginning of the study). These cats, which were bred by Harlan (Indianapolis, IN, USA), were imported at 9 months of age. Before the study, animals underwent clinical examinations and showed no signs of disease. Absence of Bartonella spp. was confi rmed by blood culture, tests for DNA of Bartonella spp., and serologic analysis. Animals were housed singly in cages in compliance with European guidelines (cage surface 6,000 cm 2 , cage height 50 cm). Animals were allowed to acclimate for 3 months to the facility, diet, and handling before the fi rst blood sample was taken and subsequent infection. Animals were fed ad libitum with standard feline maintenance diet (Harlan) and received water with no restrictions. Animal care was provided in accordance with the good animal care practice.

Feeding of I. ricinus Ticks with Ovine Blood
The general experimental framework of artifi cial feeding is shown in Figure 1. Ovine blood used in all experiments was obtained from 3 sheep reared at the National Veterinary School in Nantes, France. Absence of Bartonella spp. in the blood of these sheep was confi rmed by culture assay and tests for DNA of Bartonella spp. by PCR. Lithium heparin-coated vacutainer tubes (Becton Dickinson, Le Pont de Claix-Cedex, France) were used to draw blood by venipuncture. Blood was depleted of functional complement by heat treatment (incubation for 30 min at 56°C) before use. To avoid fungal and bacterial contamination during feeding experiments, decomplemented blood was supplemented with fosfomycin (20 μg/mL) and amphotericin B (0.25 μg/mL), which were previously determined to have no effect on B. henselae viability.

Feeding of Ticks with B. henselae-Infected Blood
The method of artifi cial feeding used in this study was adapted from the method of Bonnet et al. (33). Ticks were placed in 75-cm 2 tissue culture fl asks pierced at the top to accommodate a 4-cm diameter glass feeder. The feeder apparatus was closed with a parafi lm membrane at the top and nymphs, and 55 female adults) were engorged by feeding through artifi cial skin on B. henselae-infected blood for 5 days for larvae, 12 days for nymphs, and 21 days for adults. Larvae and nymphs were allowed to molt and engorged females were allowed to lay eggs. To evaluate transtadial and transovarial transmission, Bartonella spp. DNA was detected by PCR in salivary glands (SGs) and carcasses of 9 nymphs, 6 female adults, 9 pools of eggs, and resulting pools of larvae. Eighteen nymphs and 13 adult females fed on infected blood at preceding stages were refed for 84 h on noninfected blood. Bartonella spp. DNA was detected by PCR in SGs of 7 engorged nymphs and 3 engorged female adults. B. henselae colonies were isolated from SGs of 3 nymphs and 4 adults and from blood removed from feeders. Infectivity of B. henselae in SGs was tested by infecting 2 cats with 1 pair of SGs from a potentially infected nymph and 1 pair of SGs from a potentially infected adult, respectively. Second blood meal on uninfected blood meal (partial feeding during 84 h) a gerbil or rabbit skin membrane at the bottom. Gerbil skins were used for feeding larvae and rabbit skins for feeding nymphs and adults (33). To attract ticks, a constant temperature (37°C) was maintained by a water-jacket circulation system through the glass feeder. For blood infection, 10 μL of B. henselae suspension at a concentration of 10 9 CFU/ mL in PBS was added to 10 mL of blood supplemented with fosfomycin and amphotericin B. The culture box containing ticks was placed under the feeding apparatus, and 3 mL of B. henselae-infected ovine blood, changed twice a day, was introduced until the ticks were replete. To feed adult ticks, an equal number of males (for reproduction) and females were used. Separate apparatuses were used to engorge 200 larvae, 178 nymphs, and 55 female adults. As a control, 50 nymphs were fed under the same conditions on uninfected decomplemented ovine blood.
After feeding, larvae and nymphs were allowed to molt to nymphs and adults, respectively, and engorged females were allowed to lay eggs. Between each feeding on skin, ticks were starved for at least 2 months.

Feeding with Uninfected Blood of Ticks Infected with B. henselae at Preceding Stages
Nymphs and adults fed on blood infected with B. henselae at preceding stages were fed with uninfected, decomplemented, ovine blood in 2 glass feeders as described above. After 84 h of refeeding, nymphs and females attached to the skin were removed and dissected. Bartonella spp. DNA was detected every 24 h in blood from the fi rst 48 h of attachment onward. At each time point, 3 mL of blood was removed and centrifuged for 30 min at 3,000 × g. The supernatant was aspirated, and the pellet (200 μL) was used to detect bacterial DNA. After 84 h of feeding, 10 μL of blood was used for B. henselae culture.

Tick Dissection
Salivary glands (SGs) from I. ricinus adults and nymphs were dissected under a magnifying glass in sterile PBS. All dissection material was cleaned with DNAoff (Eurobio, Courtaboeuf, France) and rinsed with sterile water between each sample. Individual pairs of tick SGs and the remaining tick carcasses were suspended in 150 μL of PBS before culture and used to infect cats or frozen at -80°C until DNA extraction.

DNA Extraction
Carcasses (ticks without SGs), entire ticks, and pools of 50 larvae were mechanically disrupted as described (34). Eggs were crushed by using a microtissue grinder and dissected SGs were directly used for DNA extraction. DNA was extracted from all tick samples by using the Nucleo-spin Tissue kit according to the manufacturer's instructions (Macherey-Nagel, Duren, Germany). DNA was extracted from blood samples by using the Nucleospin Blood Quick Pure kit (Macherey-Nagel). For carcasses and entire ticks, the fi nal elution volume was 100 μL for adults and 30 μL for nymphs. For DNA from SGs, the fi nal elution volume was 50 μL for adults and 20 μL for nymphs. Pools of eggs, larvae, and blood DNA were eluted in volumes of 50 μL.
The second seminested amplifi cation was performed with 5 μL of a 100-fold dilution of the fi rst PCR product and primers pc535 and bsp16F (5′-TCTCTACGGAATAACACAGA-3′) a 16S rRNA Bartonella spp.-specifi c primer, and resulted in amplifi cation of a 337-bp product. The PCR cycle was identical for both amplifi cation reactions: an initial denaturation step for 8 min at 94°C; 35 cycles of denaturation for 1 min at 94°C, annealing for 1 min at 54°C, and extension for 1 min at 72°C; and a fi nal extension step at 72°C for 10 min. Each reaction was conducted in a total volume of 25 μL with 0.5 μmol/μL of each primer, 2.5 mmol/L of each dNTP, 2.5 μL of 10× PCR buffer, and 1 U of Taq DNA polymerase (Takara Biomedical Group, Shiga, Japan). Negative (ticks fed on uninfected ovine blood) and positive (B. bacilliformis DNA to easily detect any cross-contamination) controls were included in each assay. All PCRs were performed in a thermocycler MyCycler (Bio-Rad, Strasbourg, France).

Sequencing and Sequence Analysis
The expected 337-bp PCR product was isolated by agarose gel electrophoresis, excised from the gel, purifi ed by using NucleoSpin Extract II (Macherey-Nagel), and sent to QIAGEN (Hilden, Germany) for direct sequencing. Sequences were compared with known sequences listed in the GenBank nucleotide sequence databases by using the BLAST search option of the National Center for Biotechnology Information (www.ncbi.nlm.nih.gov/BLAST).

B. henselae Culture from Tick SGs and Blood
After refeeding and dissection, 4 pairs of SGs from 4 potentially infected adults females and 3 nymphs were incubated in 1 mL of Schneider Drosophila medium (Invitrogen, Cergy-Pontoise, France) at 35°C in an atmosphere of 5% CO 2 . After 6 days of incubation, 10-μL samples were placed on CBA plates. Ten microliters of blood was removed from the glass feeder after 84 h of refeeding with adults suspected of having B. henselae and incubated for 6 days in Schneider Drosophila medium before being placed on CBA plates.

Infection of Cats with Adult and Nymph SGs Potentially Infected with B. henselae
One cat was intravenously infected with 150 μL of PBS containing SGs from refed adult ticks suspected of having B. henselae; another cat was infected with 150 μL of PBS. Before handling, cats were anesthetized with ketamine (10-15 mg/kg bodyweight) and diazepam (100 μg/kg bodyweight) given intravenously. One milliliter of blood was obtained from the jugular vein at days 7, 14, and 21 postinfection, and 100 μL of blood dilutions (1:10, 1:100, and 1:1,000) was directly placed on sheep blood CBA plates. Colonies were counted and CFU/mL was estimated after incubation for 10 days at 35°C in an atmosphere of 5% CO 2 .

Detection of Bartonella spp. DNA in Questing I. ricinus Ticks
We tested 217 ticks collected in the forest of Gavre to determine prevalence of Bartonella spp. DNA. None of the 98 nymphs, 49 female ticks, and 70 male ticks collected in the forest showed amplifi cation of the Bartonella spp. 16S rRNA gene.

Transstadial Transmission of B. henselae DNA by I. ricinus Ticks
A total of 433 I. ricinus ticks were used in 2 independent experiments with feeding on artifi cial skin. Of these ticks, 169 (84.5%) of 200 larvae, 111 (62.3%) of 178 nymphs, and 19 (34.5%) of 55 female adults were successfully engorged and spontaneously detached. After feeding and detachment, all ticks were maintained in humidity chambers to enable molting or laying of eggs. After 3 months, 47 larvae (27.8%) and 58 nymphs (30.1%) molted into nymphs and adults (28 males and 30 females), respectively. Of the 19 engorged females, 9 laid eggs. Nine nymphs exposed to infection as larvae were tested for B. henselae DNA. All carcasses showed amplifi cation of the expected 337-bp DNA fragment, and no amplifi cation product was obtained from SGs. Four carcasses of 6 females exposed to B. henselae as nymphs showed the 337-bp DNA fragment, and no amplifi cation product was obtained with SGs (Table, Figure 2). No amplifi ed fragment was detected with DNA extracts from SGs or carcasses of control ticks fed on uninfected blood. All PCR products were sequenced and shared 100% identity with the 16S rRNA gene of B. henselae Houston-1 (Genbank accession no. BX897699).

Transmission of Viable and Infective B. henselae by I. ricinus Ticks
Eighteen nymphs and 13 female adults fed on B. henselae-infected blood at preceding life stages were refed for 84 h with uninfected blood. DNA extracts were prepared from SGs and carcasses of 7 partially engorged nymphs and 3 partially engorged females. Bartonella spp. DNA was amplifi ed from carcasses and SGs of 4 nymphs and 1 female. For 1 nymph and 1 female, the specifi c 337-bp DNA fragment was amplifi ed only in SG DNA extracts but not in carcass extracts (Table; Figure 3, panel A). Four pairs of SGs from 4 partially engorged females and 3 pairs of SGs from 3 partially engorged nymphs were incubated separately in Schneider Drosphila medium for 6 days before being placed on sheep blood agar. B. henselae colonies appeared after 7 days for all SGs tested, which indicated the presence of viable bacteria in SGs from adults and nymphs.
To determine whether B. henselae in SGs of nymphs and adults were infective, we infected 2 healthy cats with 1 pair of SGs from a nymph and 1 pair of SGs from a female adult. The cat infected with SGs from an adult became bacteremic 7 days after infection (5 × 10 5 CFU/mL at day7, 6 × 10 6 CFU/mL at day 14, and 1 × 10 6 CFU/mL at day 21), and the cat infected with SGs from a nymph became bacte-Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 14, No. 7, July 2008 1077 remic after 14 days of infection (8 × 10 5 CFU/mL at day 14 and 2 × 10 7 CFU/mL at day 21). Blood samples from feeders were obtained every 12 h for the fi rst 48 h to detect B. henselae. In the fi rst 60 h, B. henselae DNA or viable B. henselae was not detected in blood samples. After 72 h of refeeding of infected ticks, Bartonella spp. DNA was successfully amplifi ed ( Figure 3, panel B), and B. henselae colonies were isolated from these blood samples after preincubation for 6 days in Schneider Drosophila medium, which indicated viability of the transmitted bacteria.

Discussion
This study demonstrated transmission of B. henselae by I. ricinus ticks across different developmental stages, migration and multiplication of viable and infective B. henselae in SGs after a second blood meal, and transmission of B. henselae from ticks to blood. These fi ndings indicate that I. ricinus is a competent vector for B. henselae.
Vector biologists and epidemiologists have suggested that ticks may play a role in transmission of Bartonella spp. (11,12,19,23,25,28,29). This suggestion was based on indirect data for detection of bacterial DNA in ticks (18,19,24), humans exposed to tick bites (30), or serologic evidence of co-infection of humans with pathogens known to be trans-mitted by ticks (11,36). Diffi culties in rearing I. ricinus and lack of a rodent model for B. henselae infection may explain the absence of data demonstrating the role of this tick as a vector of B. henselae. Recent development of an artifi cial method suitable for feeding ticks (33) enabled us to study experimental infection of ticks with blood containing B. henselae, to monitor B. henselae through various tick stages, and to evaluate putative transmission of bacteria from the tick to blood.
To select a tick population with the lowest Bartonella spp. DNA prevalence, we estimated the prevalence of Bartonella spp. DNA in questing I. ricinus collected in different areas in France. The lowest Bartonella spp. DNA prevalence was in Loire-Atlantique (19 and unpub. data). We thus used ticks collected in this area for our study.
We detected B. henselae DNA in 100% of carcasses from nymphs and 67% of carcasses from adults fed on ovine blood containing B. henselae at their preceding stages. No B. henselae DNA was amplifi ed in corresponding SGs in a nested PCR, which is more sensitive than amplifi cation of the classic citrate synthase gene. This result demonstrated that bacteria could be ingested by I. ricinus larvae and nymphs during feeding on artifi cial skin and that bacterial DNA was maintained in the tick after molting. However, no or undetectable numbers reached the SGs.
Although bacterial DNA was detected in eggs laid by females fed on blood containing B. henselae, larvae obtained from these eggs were PCR-negative for B. henselae. This fi nding suggests external contamination of eggs with DNA rather than transovarial persistence of bacteria.
When molted nymphs and female ticks potentially contaminated with B. henselae at their previous developmental stage were refed on uninfected blood, viable B. henselae were detected in SGs after 84 h of engorgement. Two hypotheses could explain the absence of detectable bacterial DNA in SGs after an infected blood meal and molting, when it becomes detectable after a partial refeeding blood meal. The fi rst hypothesis is that the 84-h refeeding period may act as a stimulus and enable migration of bacteria from the gut to SGs of the tick as previously described for B. burgdorferi sensu lato (14). The second hypothesis is that this refeeding period may stimulate multiplication of bacteria already present in SGs, but at undetectable levels. More investigations are needed to validate one of these hypotheses. Bacteria located in SGs of nymphs and adults are infective because injection of 1 pair of infected SGs into cats induced high levels of bacteremia. Cats became bacteremic in the fi rst 2 weeks after injection, as described for cat infection with B. henselae by fl eas, and bacteremia levels were similar to those observed in cats infected by fl ea bites (7,8 In conclusion, we demonstrated by using feeding on artifi cial skin that B. henselae, the cause of cat-scratch disease in humans, could be transmitted by ticks through saliva. Although further investigations are needed to clarify the epidemiology of such transmission, health authorities must take into account the possibility of bartonellosis in persons exposed to tick bites, and B. henselae must be identifi ed as a tick-borne pathogen. A B