|Home | About | Journals | Submit | Contact Us | Français|
Twenty Rhipicephalus sanguineus ticks collected in eastern Arizona were tested by PCR assay to establish their infection rate with spotted fever group rickettsiae. With a nested PCR assay which detects a fragment of the Rickettsia genus-specific 17-kDa antigen gene (htrA), five ticks (25%) were found to contain rickettsial DNA. One rickettsial isolate was obtained from these ticks by inoculating a suspension of a triturated tick into monolayers of Vero E6 monkey kidney cells and XTC-2 clawed toad cells, and its cell culture and genotypic characteristics were determined. Fragments of the 16S rRNA, GltA, rOmpA, rOmpB, and Sca4 genes had 100%, 100%, 99%, 99%, and 99%, respectively, nucleotide similarity to Rickettsia massiliae strain Bar29, previously isolated from R. sanguineus in Catalonia, Spain (L. Beati et al., J. Clin. Microbiol. 34:2688-2694, 1996). The new isolate, AZT80, does not elicit cytotoxic effects in Vero cells and causes a persistent infection in XTC-2 cells. The AZT80 strain is susceptible to doxycycline but resistant to rifampin and erythromycin. Whether R. massiliae AZT80 is pathogenic or infectious for dogs and humans or can cause seroconversion to spotted fever group antigens in the United States is unknown.
The genus Rickettsia contains a diverse and expanding number of obligately intracellular gram-negative bacteria (23, 43). These bacteria vary in their antigenic and microbiological characteristics, as well as their distribution, ecology, pathogenicity, and association with arthropod hosts including lice, fleas, ticks, and mites, as well as nonhematophagous hosts. Molecular methods are increasingly used to differentiate and define their taxonomic and phylogenetic relationships.
At least 14 well-characterized Rickettsia genotypes are recognized in northern America. In the classic, highly pathogenic typhus group, the widespread agent of murine typhus, Rickettsia typhi, is transmitted by fleas while Rickettsia prowazekii is maintained in a sylvatic cycle by the flying squirrels, Glaucomys volans, and its ectoparasites in the eastern United States. The ecology and pathogenicity of Rickettsia canadensis, which exhibits both typhus and spotted fever group characteristics, is poorly defined, but it has been isolated from Haemaphysalis ticks in Quebec and California. Rickettsia felis and Rickettsia akari are transmitted by fleas and mites, respectively, and have ecoepidemiological characteristics different from the other spotted fever group rickettsiae which are transmitted by ticks (13, 42). Rickettsia rickettsii, the causative agent of Rocky Mountain spotted fever (RMSF) is the best known species throughout the Americas. R. rickettsii is transmitted by the American dog tick Dermacentor variabilis and the wood tick Dermacentor andersoni in the eastern and western United States, respectively, and by Amblyomma cajennense in part of Texas and south of the United States. Rhipicephalus sanguineus, the brown dog tick, was recently implicated as a vector of R. rickettsii in eastern Arizona (21) and has been identified as a vector of RMSF in Mexico (15, 16). Rickettsia parkeri was recently described as a cause of spotted fever rickettsioses and is transmitted by the Gulf Coast tick, Amblyomma maculatum (41). Another spotted fever group rickettsia found in Amblyomma americanum ticks, referred to as “Rickettsia amblyommii,” has also been associated with a mild febrile illness and rash in the southern United States (19, 20, 58). Rickettsia montanensis and Rickettsia rhipicephali have been isolated from human-biting ticks, D. variabilis, D. andersoni, Dermacentor occidentalis, R. sanguineus, and Ixodes pacificus, but these rickettsiae have not been proven to be pathogenic for humans (8, 14, 31, 46, 47). Similarly, pathogenic potential has not been established for Rickettsia bellii which is found in a wide range of hard and soft ticks (45) and for a R. rickettsii-like rickettsia found in argasid ticks (36). The pathogenic potential is also unknown for the “Cooleyi” genotype of spotted fever group rickettsiae detected in the black-legged tick Ixodes scapularis from Texas (12). Similarly, the “Midichlorii” genotype found in I. scapularis and I. pacificus represents another rickettsia, but it is considered to be a transovarially maintained endosymbiont (51). The presence of apparently nonpathogenic rickettsiae whose vector host and geographic distributions overlap with those of pathogenic rickettsiae may interfere with the transmission of virulent species to humans, as has been proposed for Rickettsia peacockii and R. rickettsii in D. andersoni in the Bitterroot Valley of Montana (39).
We report here the first detection and isolation of Rickettsia massiliae from R. sanguineus ticks collected in Arizona in an area where R. rickettsii infection is endemic (21).
In June 2004, ticks were collected by flagging in a peridomestic environment in an area of eastern Arizona as previously described (21). Ticks were surface disinfected using a series of washes with 10% bleach for 1 to 2 min, 70% ethanol for 3 to 5 min, and three times with distilled water, and excess water was blotted with filter paper. One half of each tick was processed for DNA extraction using a QIAamp DNA Mini Kit (QIAGEN, Valencia, CA). The other half was placed in a sterile vial containing 0.1 ml of sucrose glutamate buffer (0.22 M sucrose, 0.1 M potassium phosphate, 0.005 M sodium l-glutamate, pH 7.0) supplemented with 5 mM MgCl2 and 1% Hypaque-76 (Nycomed, Inc., Princeton, NJ) (SRM buffer) and frozen at −80°C.
Frozen tick samples were thawed and triturated, the volume was increased to 0.5 ml with SRM buffer, and 0.15 ml was used to inoculate a 25-cm2 flask containing a confluent monolayer of Vero E6 cells (CRL 1587/Vero 76; American Type Culture Collection, Manassas, VA). Cells were maintained in RPMI 1640 cell culture medium (GIBCO-Invitrogen Corp., Grand Island, NY) supplemented with 2% fetal bovine serum (HyClone Laboratories Inc., Logan, UT), 5% tryptose phosphate broth, and 1 mM l-glutamine at 34°C in a CO2 incubator. One milliliter of a tissue culture suspension that contained rickettsial DNA as determined by PCR was used to infect a confluent monolayer of XTC-2 cells (49). The flasks were incubated for 1 h at room temperature by rocking, supplemented with Leibovitz L-15 cell culture medium (Invitrogen) containing 2% fetal bovine serum, 5% tryptose phosphate broth, and 1 mM l-glutamine; the culture was transferred into a CO2 incubator at 28°C. The presence of rickettsiae in the cell cultures was first detected in XTC-2 cells following staining with acridine orange (35). Infected XTC-2 cultures were further passaged onto VERO cells at 34°C until sustained abundant growth of the rickettsiae was established.
Rickettsiae were propagated in VERO cell monolayers, the heavily infected cells were harvested with glass beads, and the suspension was pelleted by centrifugation at 10,000 rpm (17,000× g) for 20 min. The pellet was resuspended in 100 ml of K-36 buffer (0.05 M potassium phosphate, 0.1 M KCl, 0.15 M NaCl, pH 7.0) (67). It was centrifuged for 10 min at 1,000 rpm (200 × g) in a tabletop centrifuge (Beckman GRP, Hamburg, Germany), and the supernatant containing rickettsiae was filtered through an AP-25 glass fiber filter (Millipore Corp., Bedford, MA). Rickettsiae were concentrated by centrifugation for 20 min at 10,000 rpm; the pellet was resuspended in 10 ml of SRM buffer and frozen in aliquots at −80°C. The viable titer of purified rickettsiae was determined by plaque titration on nonirradiated L929 mouse fibroblast cells (ATCC CCL-1) grown in Eagle's minimal essential medium (Invitrogen) and M199 medium (Invitrogen) mixed at a ratio of 1:1 (vol/vol), containing 2% fetal bovine serum, 5% tryptose phosphate broth, and 1 mM l-glutamine as described previously (68). A second overlay of agarose containing 0.5 M NaF was added 3 days after primary inoculation to inhibit metabolism and growth of the host cells (48).
To quantify growth characteristics of the new isolate, confluent monolayers of Vero cells and XTC-2 cells were grown in 33-mm dishes and infected with rickettsiae at a multiplicity of infection (MOI) of 0.1 PFU per cell as previously described (25). Infected cells were harvested from duplicate wells immediately following inoculation and on days 1, 3, 5, 7, 9, and 11 after inoculation. The cell culture medium and infected cells were combined and centrifuged in an Eppendorf vial for 10 min at 14,000 × g, and the resulting pellet was used for DNA extraction using a QIAamp DNA Mini Kit (QIAGEN). DNA was eluted with 200 μl of AE (10 mM Tris-Cl, 0.5 mM EDTA, pH 9.0) buffer (QIAGEN) and stored at 4°C. Rifampin (0.5 to 2 μg/ml), erythromycin (1 to 8 μg/ml), and doxycycline (0.03 to 0.125 μg/ml) (Sigma, St. Louis, MO) were added in serial twofold dilutions, and their effect was evaluated on day 5 following inoculation. VERO cells infected with R. rickettsii strain Bitterroot under the same conditions were used as a control; preparation of R. rickettsii seeds has been described previously (25). The effect of antibiotics was determined by comparing the amounts of rickettsial DNA detected and normalized to the total DNA quantity recovered in each sample. The DNA concentration was measured using a PicoGreen double-stranded DNA quantitation kit according to the manufacturer's instructions (Molecular Probes, Eugene, OR). Statistical significance was assessed by a Student's t test.
No work was done with European isolates of R. massiliae in the same facility during the time this work was done. A nested PCR assay to amplify a fragment of the 17-kDa antigen gene (htrA) was used to detect the presence of rickettsiae in ticks as described previously (1, 64). A rompA SYBR green quantitative PCR assay was used to measure the amount of rickettsial DNA in cell culture samples on an i-Cycler (Bio-Rad Laboratories Inc., Hercules, CA) (24). PCR amplification of the 16S rRNA, GltA, rOmpA, rOmpB, and Sca4 genes from the isolated rickettsiae for DNA sequencing was performed using QIAGEN Master Mix reagents. The oligonucleotide primers used are shown in Table Table1;1; rompB and sca4 were characterized using primers published previously (54, 62). The primers were made by the CDC Core Facility (Atlanta, GA) and were used at a final concentration of 1 μM unless otherwise specified. Restriction endonucleases, RsaI, and PstI (New England BioLabs, Beverly, MA), were used for restriction fragment polymorphism (RFLP) analysis (26). RFLP fragments were resolved on 2% agarose gels.
Sequence reactions were performed using an ABI PRISM 3.0 BigDye Terminator Cycle Sequencing kit as recommended by the manufacturer (Applied BioSystems, Foster City, CA). The sequenced products were purified with a QIAGEN Dye Removal Kit and run on an Applied BioSystems 3100 Nucleic Acid Sequence Analyzer.
The nucleotide sequences generated during this study were deposited in the NCBI GenBank under the following accession numbers: DQ517444 for the 17-kDa antigen gene fragment, DQ212705 for the gltA fragment, DQ212706 for the 16S rRNA gene, DQ212707 for the rompA fragment, DQ503428 for rompB, and DQ503429 for sca4.
Twenty questing adult R. sanguineus ticks (8 males and 12 females) were collected by flagging around a single control household evaluated as a part of an investigation into cases of RMSF in Arizona (21). By nested PCR assay, which amplifies a 220-bp fragment of the 17-kDa antigen gene of Rickettsia, five ticks (three males: AZT68, AZT69, and AZT82; two females: AZT80 and AZT81) were found positive for the presence of rickettsial DNA. All five amplicons had identical nucleotide sequences that shared 99% sequence homology with the 17-kDa protein gene fragment from R. rhipicephali and the R. rhipicephali-like rickettsiae, ARANHA and ATT (Fig. (Fig.1).1). A 602-bp fragment from the 5′ region of the Rickettsia rompA was amplified only from the DNA of the two samples from AZT68 and AZT80. The PstI and RsaI restriction profiles of the two fragments were each the same, and they were identical to those previously published for the strain Bar29 of R. massiliae (Fig. (Fig.22).
Following primary inoculation of VERO cells with suspensions of the five ticks found positive by PCR assay (AZT68, AZT69, AZT80, AZT81, and AZT82), the cultures were maintained for 4 weeks, but no obvious cytotoxic effect was observed, and inconclusive results were obtained upon examination of slides stained with acridine orange. To verify the presence of any rickettsial DNA, an aliquot of each of the inoculated cell cultures was tested by PCR of the gltA gene fragment. Samples AZT68 and AZT80 tested positive for rickettsial DNA. Because the poor growth in VERO cells suggested that growth of the rickettsiae might be temperature sensitive, suspensions of these infected cells were inoculated onto confluent monolayers of XTC-2 cells and incubated at 28°C. Changes in morphology of the infected monolayers were noticed, and the presence of rickettsia-like organisms in the cell culture medium infected with only AZT80 cell suspension was detected 10 days following inoculation. The cells in the supernatant were used to infect another flask of XTC-2 cells, and fresh medium was added to the attached cells in the original culture. The presence of rickettsial DNA was detected in both of these cultures by PCR 3 weeks following the inoculation. Another passage was performed in XTC-2 cells followed by inoculation of VERO cells with the infected cell culture supernatant. An isolate of rickettsiae designated AZT80 was obtained that could be then stably maintained in VERO cells at 34°C.
The VERO-passaged AZT80 rickettsiae grew well in both XTC-2 and VERO cells (Fig. (Fig.3).3). The VERO cell monolayer infected with AZT80 underwent senescence at 3.5 to 4 weeks following inoculation. Flasks of XTC-2 monolayers inoculated with AZT80 were maintained for 5.5 months as a persistent infection by changing the cell culture medium every 10 days to 2 weeks.
The ability of the AZT80 isolate to form lytic plaques in confluent monolayers of VERO E6 and L929 cells was evaluated. No lytic plaques were produced in L929 cells infected without NaF treatment or in VERO E6 cells with or without NaF. Small 1-mm diameter plaques were observed in L929 cultures treated with sodium fluoride after 10 days. The PFU titer of the semipurified suspension of AZT80 isolate used for this experiment was estimated to be 107 PFU per ml.
When XTC-2 cells were infected at an MOI of 0.1 rickettsiae per cell, the amount of DNA fluctuated greatly initially, but AZT80 DNA did not accumulate in a significant amount per plate of XTC-2 cells during days 5 to 11 of observation (P = 0.55) and did not exhibit a logarithmic increase typical for spotted fever group rickettsiae (Fig. (Fig.4A).4A). In contrast, the amount of AZT80 DNA increased nearly 10-fold (P = 0.0075) in VERO cells infected under the same conditions (Fig. (Fig.4B4B).
VERO cells infected with AZT80 and treated with 0.03 to 0.125 μg/ml of doxycycline for 5 days had a significantly lower quantity of rickettsial DNA compared to untreated cells infected with isolate AZT80 (P = 0.007 to 0.01) (Fig. (Fig.5).5). Infected cells grown in the presence of 1 μg/ml erythromycin had rickettsial DNA similar to that in untreated infected culture (P = 0.55); treatment with 2, 4, and 8 μg/ml of erythromycin caused a reduction in rickettsial DNA (P = 0.016, P = 0.005, and P = 0.007, respectively) compared to that detected in untreated culture infected with AZT80. While cultures supplemented with 0.5 to 1 μg/ml of rifampin had a similar quantity of rickettsial DNA copy numbers as untreated infected controls (P = 0.47 and P = 0.75, respectively), cells infected with AZT80 and treated with 2 μg/ml of rifampin contained significantly reduced rickettsial DNA compared to the untreated control (P = 0.01) but still a greater amount than AZT80-infected cells treated with the highest dose of doxycycline (P = 0.006) and erythromycin (P = 0.04). As a control for the antibiotics, VERO cells were infected with R. rickettsii at the same MOI as AZT80 and treated with 0.125 μg/ml doxycycline, 4 μg/ml erythromycin, and 1 μg/ml rifampin for 5 days (Fig. (Fig.5,5, inset). In these cultures, a significant decrease in rickettsial DNA was detected in infected cells treated with both doxycycline and rifampin compared to untreated infected controls (P = 0.023 and P = 0.024, respectively). However, there was no significant reduction in rickettsial DNA in cultures treated with erythromycin (P = 0.6) compared to untreated VERO cells infected with R. rickettsii.
Partial sequences of the 16S rRNA gene, gltA, and rompA genes were determined for the AZT80 isolate. The 905-bp fragment of the 16S rRNA gene had 100% sequence identity with the homologous gene fragment of R. massiliae strain Bar29 isolated from R. sanguineus ticks in Catalonia, Spain (7). The 334-bp fragment of gltA differed by 1 nucleotide from gltA of Bar29 with 99% nucleotide sequence and 100% amino acid sequence homology. The 578-bp fragment of the rompA gene had 99% sequence homology with R. massiliae rompA sequences (Fig. (Fig.6A).6A). Up to four nucleotide pairwise differences were detected in this fragment of rompA among the three spotted fever group rickettsiae referred to as R. massiliae, including isolates Bar29, Mtu1, and GS, and the new isolate AZT80. Only one of these differences, T359, resulted in a change of amino acid sequence (Fig. (Fig.6B6B).
The 4,864-bp fragment of AZT80 rompB differed by only 1 nucleotide from the homologous gene of Bar29 and that caused a single amino acid substitution. Two nucleotide differences were identified when the 2,259-bp fragment of AZT80 sca4 was compared to sca4 of Bar29.
We describe the first detection and isolation of R. massiliae from R. sanguineus ticks in the United States. For many years, only R. rhipicephali was known to be associated with brown dog ticks in the continental United States (14, 31). R. sanguineus was implicated as a vector for R. rickettsii infection in Mexico (15, 16), but precise identification of these rickettsiae has not been confirmed using contemporary molecular tools. In contrast, recent studies focused on an RMSF outbreak investigation in eastern Arizona clearly demonstrated the presence of R. rickettsii in brown dog ticks (21). It was also suggested that this organism has been present in the area for an extended time, based on retrospective serological analysis of dog serum samples collected since 1995 (38). However, we report here the detection and isolation of another spotted fever group rickettsia in this site that could also elicit spotted fever group antibody responses in dogs. R. massiliae has not previously been found in the United States or elsewhere in the Western Hemisphere, thus extending the geographic distribution of this rickettsial species to the New World.
Before this study was initiated, the known presence of R. massiliae was restricted to several countries of southern Europe and from Africa (2, 4-7, 11, 63). In these regions, strains Mtu1 and Mtu5 were isolated from Rhipicephalus bursa and R. sanguineus from southern France, respectively, (5), strain GS from R. sanguineus collected in central Greece (2), and strain Bar29 from R. sanguineus collected in Catalonia, Spain (7). R. massiliae was also detected by PCR/RFLP in R. sanguineus found in Portugal (4). The presence of R. massiliae has been detected in several other ticks of the so-called Rhipicephalus spp. complex including the Mtu5 genotype in Rhipicephalus senegalis and Mtu1 genotype in Rhipicephalus sulcatus, Rhipicephalus lunulatus, and Rhipicephalus mushamae in Africa (63); and the Bar29 genotype in R. sanguineus and Rhipicephalus turanicus in Switzerland (11). While R. massiliae AZT80 possesses unique genetic characteristics that allow its simple identification and differentiation from other spotted fever agents found in the United States (6, 26, 30, 53-56, 62), the antigenic properties of AZT80 are currently being evaluated. R. massiliae strain Mtu1 has a unique subset of antigenic epitopes that are not expressed by Bar29 or GS strains (69), thus allowing its differentiation.
In Europe and Africa Rhipicephalus ticks are known to vector other spotted fever group rickettsiae including Rickettsia conorii, the etiologic agent of boutonneuse fever or Mediterranean spotted fever (MSF) (18, 27, 43, 44). In addition, R. rhipicephali, a close phylogenetic relative of R. massiliae (30, 53-56, 62) has also been detected in R. sanguineus ticks in Spain (33), Portugal (4), France (22), and the African continent (63). In these countries, particularly in Greece and Spain, it has been noted that a discrepancy exists between the seroprevalence of antibodies to spotted fever group rickettsiae in local human populations, the incidence of reported rickettsioses (3, 17, 28, 32, 60), and the occurrence of severe and fatal forms of MSF (28, 29, 34, 57, 62). Similarly, there is a noticeable discrepancy in the prevalence of antibodies reactive with virulent R. conorii in dog populations and the occurrence of disease in different locations in Spain (32, 34, 60, 61). These observations support the idea that both virulent and avirulent rickettsiae may be responsible for seroreactivity to spotted fever group agents in human and canine populations (9, 10, 17).
The AZT80 isolate does not elicit pronounced cytotoxic effects in VERO cells like R. rickettsii and R. rhipicephali (25, 46). European isolates of R. massiliae, including Mtu1 and Bar29, persist in HELA and VERO cells without visible changes in the morphology of the host cells (5-7) and exhibit a limited ability to produce lytic plaques in L929 and VERO cells (52), features often attributed to nonpathogenic rickettsiae. In addition, R. massiliae appears to establish persistent infection in its tick vector, R. turanicus (37), while both R. rickettsii and R. conorii infections are very detrimental for their respective vectors, D. variabilis (40) and R. sanguineus (59).
Until recently, an association of R. massiliae with illness in humans in Europe had not been demonstrated by isolation or PCR detection of the agent in clinical specimens (44). However, Catalan patients often present with mild forms of MSF compared to cases reported from other regions of Spain (10, 60). Furthermore, children diagnosed with MSF in Catalonia showed unusual unresponsiveness to treatments to rifampin, which has been used as an alternative antibiotic when tetracycline is contraindicated (9). R. conorii and R. rickettsii do not exhibit resistance to rifampin (52). However, in vitro evaluation of antibiotic susceptibility of the R. massiliae and Catalan isolate Bar29 demonstrated resistance to rifampin (7, 52), again suggesting that this rickettsia may be responsible for mild rickettsioses in the Catalan region. The single confirmed case involving infection with an R. massiliae isolate occurred in Sicily and was initially described as MSF with rash and eschar but low antibody titer to R. conorii (65). Whether R. massiliae contributes to clinical manifestations of spotted fever rickettsiosis or human and canine seroreactivity to spotted fever group rickettsiae in Arizona needs further evaluation. The AZT80 strain of R. massiliae is susceptible to doxycycline but resistant to rifampin and erythromycin.
Our findings demonstrate that the noncytotoxic AZT80 strain of R. massiliae is present in R. sanguineus ticks in the same geographic area where pathogenic R. rickettsii has been identified as a cause of fatal RMSF (21). There is no information available now on the prevalence of R. massiliae in Arizona or other areas of the United States where R. sanguineus ticks can be found. Since R. massiliae shares the tick vector R. sanguineus with R. rickettsii in eastern Arizona and since this tick can bite humans, its presence has to be taken into consideration when epidemiological surveillance for RMSF based on serology is conducted.
The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the Department of Health and Human Services.