Abstract
From 1997 to 2009, the Tick-Borne Disease Laboratory of the U.S. Army Public Health Command (USAPHC) (formerly the U.S. Army Center for Health Promotion and Preventive Medicine) screened 5286 Dermacentor variabilis ticks removed from Department of Defense (DOD) personnel, their dependents, and DOD civilian personnel for spotted fever group rickettsiae using polymerase chain reaction and restriction fragment length polymorphism analysis. Rickettsia montanensis (171/5286 = 3.2%) and Rickettsia amblyommii (7/5286 = 0.1%) were detected in a small number of samples, but no ticks were found positive for Rickettsia rickettsii, the agent of Rocky Mountain spotted fever (RMSF) until May 2009, when it was detected in one D. variabilis male removed from a child in Maryland. This result was confirmed by nucleotide sequence analysis of the rickettsial isolate and of the positive control used in the polymerase chain reaction, which was different from the isolate. Lethal effects of rickettsiostatic proteins of D. variabilis on R. rickettsii and lethal effects of R. rickettsii infection on tick hosts may account for this extremely low prevalence. Recent reports of R. rickettsii in species Rhipicephalus sanguineus and Amblyomma americanum ticks suggest their involvement in transmission of RMSF, and other pathogenic rickettsiae have been detected in Amblyomma maculatum. The areas of the U.S. endemic for RMSF are also those where D. variabilis exist in sympatry with populations of A. americanum and A. maculatum. Interactions among the sympatric species of ticks may be involved in the development of a focus of RMSF transmission. On the other hand, the overlap of foci of RMSF cases and areas of A. americanum and A. maculatum populations might indicate the misdiagnosis as RMSF of diseases actually caused by other rickettsiae vectored by these ticks. Further studies on tick vectors are needed to elucidate the etiology of RMSF.
Introduction
Materials and Methods
Ticks
Each year from 1997 to 2009, D. variabilis removed from DOD personnel were sent to the HTTK from ∼100 participating military medical treatment facilities in the continental United States. Ticks were identified to species using standard taxonomic keys (Keirans and Litwak 1989).
DNA isolation
From 1997 until 2004
DNA was isolated using the IsoQuick nucleic acid extraction kit (ORCA Research, Bothell, WA) as described in Stromdahl et al. (2001). Beginning in 2005, a new kit, Zymo Genomic DNA II Kit (Orange, CA), was used that was less laborious, and produced DNA of equivalent quality and quantity. Individual ticks were placed in 250 μL of Genomic Lysis Buffer (guanidine thiocyanate), bisected with a sterile 18-gauge hypodermic needle, and incubated at room temperature for 15 min. After vortexing for 30 s and centrifuging at 14,000 g for 5 min, the supernatant was placed into a Zymo-Spin column and centrifuged again for 1 min at 14,000 g, and the flow-through was discarded. The column was then washed twice using g-DNA wash buffer (25% ethanol and 25% isopropanol): 250 μL was added to the column and centrifuged at 14,000 g for 1 min, the flow through was discarded, and the step was repeated exactly. Finally, 35 μL of highly purified water was added to the column, incubated for 5 min at room temperature, and centrifuged at 14,000 g for 30 s to elute the DNA. Samples were tested immediately by PCR for relevant pathogens, and archived at −70°C.
Spotted fever group rickettsia PCR
1997–1999
D. variabilis were screened for spotted fever group rickettsiae as previously described (Stromdahl et al. 2001) using rickettsial outer-membrane protein B gene (ompB) primers BG1-21 and BG2-20 and speciated with RsaI RFLP (Eremeeva et al. 1994). Samples positive in this PCR were reconfirmed using primer pair Rr190.70p and Rr190.602n targeting the rickettsial outer-membrane protein A (ompA) gene of R. rickettsii and speciated with PstI RFLP (Regnery et al. 1991). The PstI RFLP discriminates among R. amblyommii, R. montanensis, R. parkeri, and R. rickettsii, and others (Eremeeva et al. 1994). The Rr190.70p/Rr190.602n/PstI digest fragment pattern of R. amblyommii was described and pictured in Stromdahl et al. (2008).
2000–2005
D. variabilis were screened using primer pair Rr190.70p and Rr190.602n and speciated with PstI RFLP. Samples positive in this PCR were reconfirmed using primers BG1-21 and BG2-20 and speciated with RsaI.
2005–2009
We moved from conventional to real-time PCR to improve the turn-around time of tick testing and to the enhance sensitivity of our assays. D. variabilis were screened by a quantitative real-time PCR (qPCR) assay in a Roche LightCycler (Roche Applied Science, Indianapolis, IN) using primers and probe targeting a conserved range of ompB of the tick-borne spotted fever group rickettsiae (Jiang et al. 2005b). Ready-To-Go beads were reconstituted in a 20 μL volume containing 2 μL of tick DNA as template, 0.5 μM of primer RR1595F, 0.5 μM of primer RR1722R, 0.5 μM of probe RR1654, and an additional 3.2 mM MgCl2 to bring the final reaction volume of MgCl2 to 5.0 mM. Cycling conditions involved an initial 3 min denaturation at 94°C, and then 50 amplification cycles of 94°C for 5 s (20°C/s slope), and 60°C for 30 s (20°C/s slope, single acquisition mode). Any samples positive in this reaction were reconfirmed using primer pair Rr190.70p and Rr190.602n.
Positive control for all reactions above was nucleic acid of R. rickettsii, kindly provided by Abdu Azad (University of Maryland School of Medicine).
Nucleotide sequencing and phylogenetic analysis
To confirm and further characterize the R. rickettsii found in the D. variabilis tick (Tick 090589), standard PCR and nested PCR were performed to amplify ompB, ompA, and the surface cell antigen 4 gene (sca4) using Platinum PCR SuperMix High Fidelity (Invitrogen) and run on a T-Gradient Thermocycler (Biometra, Goettingen, Germany) along with the positive control. PCR amplicons were purified and sequenced in both directions on an automated ABI Prism 3130xl genetic analyzer (Applied Biosystems, Foster City, CA) similar to that described previously (Jiang, et al. 2005a). Amplicon of ompB for positive control was also sequenced to compare with the sample in case of contamination. Primers used for PCR amplification and sequencing are listed in Table 1. The sequences were assembled using Vector NTI advance 10 software (Invitrogen), and compared with the sequences of those R. rickettsii isolates in Genbank. Phylogenetic analyses were performed with MacVector 8.1 software (Accelrys, San Diego, CA).
Primers used for PCR amplification.
Primers used for sequencing.
Primers used for nested PCR amplification.
ORF, open-reading frame; PCR, polymerase chain reaction.
Results
Ticks
From 1994 until May 2009, 2490 (2490/5285 = 47.1%) male, 2795 (2795/5285 = 52.9%) female, and 1 nymphal D. variabilis were received from locations described in Table 2. R. montanensis (171/5286 = 3.2%) and R. amblyommii (7/5286 = 0.1%) were detected in a small number of samples. Engorgement status was estimated when the ticks were received; 4.4% fully engorged, 10.3% appeared to be partially engorged, and 85.3% unengorged.
Participation in the HTTK is voluntary; therefore, fluctuations in tick abundance and geographic distribution are often the result of soldier training schedules and installation participation instead of tick biology. Low tick numbers from a region may indicate lack of military installations there, or lack of participation in the HTTK.
HTTK, Human Tick Test Kit Program.
On May 19, 2009, a male D. variabilis (Tick 090589) removed from a U.S. Navy dependent (2-year-old girl) was submitted to the HTTK by the Naval Clinic, Annapolis, MD. The girl was asymptomatic; no serologic or molecular testing was done. The tick was positive in the R. rickettsii ompB qPCR assay (ct 28.68) and positive in the subsequent Rr190.70 PCR, and fragments identified in the PstI RFLP matched the positive control and appeared to indicate R. rickettsii.
Nucleotide sequencing and phylogenetic analysis
The sequences of 3136 bp of ompA, 4830 bp of ompB, and 1784 bp of sca4 from R. rickettsii isolate T090589 (from Tick 090589), and 4217 bp of ompB from the positive control DNA sample were obtained (Table 3). BLAST searches (
R. rick SS, R. rickettsii “Sheila Smith.”
The phylogenic trees were constructed using ompA (Fig. 1) and ompB (Fig. 2) by Neighbor-Joining Best tree method. R. rickettsii isolate T090589 was found to be in the same clade as other R. rickettsii isolates with a bootstrap value of 100% (data not shown).
Phylogenetic best tree based on 3135 bp ompA sequence. The tree was constructed using Neighbor Joining method Best Tree mode using MacVector 8.1 after multiple alignment of Rickettsia rickettsii T090589 with 5 R. rickettsii strains and 20 other Rickettsia species.
Phylogenetic analysis of 4830 bp ompB sequence. The tree was constructed using Neighbor Joining method Best tree mode using MacVector 8.1 after multiple alignment of R. rickettsii T090589 with the positive control and 3 R. rickettsii strains and 20 other Rickettsia species.
Nucleotide sequence accession numbers
The sequences of R. rickettsii reported from D. variabilis Tick 090589 have been deposited in GenBank with accession numbers GU395292 for ompA, GU395293 for ompB, and GU395294 for sca4.
Discussion
The low prevalence (1/5286) of R. rickettsii in D. variabilis ticks continues to be intriguing and implies that encounter with the American dog tick does not pose a significant risk for RMSF. However, human cases continue to be reported from the geographical range of D. variabilis. From 2001 to 2005, 4567 cases, including 140 from Maryland, where the infected tick described here was acquired, were reported to the U.S. Centers for Disease Control and Prevention (CDC) via Case Report Forms from state health departments (Adjemian et al. 2009). Although many of these cases could be attributed to serological crossreaction with less pathogenic or nonpathogenic rickettsiae (Apperson et al. 2008, Paddock et al. 2008), 1089 of these 4567 cases were classified as severe, and 23 were fatal, implying that some patients were acquiring virulent strains of R. rickettsii.
The extremely low prevalence of R. rickettsii in D. variabilis may be a result of the lethal effects of R. rickettsii infection on Dermacentor tick hosts. Nymphs of Dermacentor andersoni experimentally infected as larvae survived as infected nymphs, but 94% of the nymphs infected as larvae failed to survive the molt to adults, and 88% of the adult females infected as nymphs died before feeding (Niebylski et al. 1999). If D. variabilis is similarly affected by R. rickettsii, reducing the adult stage of infected ticks is key in the role of transmission to humans because only the adult stage of the tick bites humans; of the 5286 D. variabilis removed from humans and sent to the HTTK, only 1 has been an immature.
Conversely, the low prevalence of R. rickettsii in D. variabilis may also be a result of the immune response that the tick mounts to counteract the lethal effect of rickettsial colonization. Two isoforms of the antimicrobial defensin and a kunitz-type serine protease inhibitor (DvKPI) have been detected in D. variabilis, and interactions between these and R. montanensis (surrogate for R. rickettsii) have been studied in a series of in vitro and in vivo experiments that demonstrate that R. montanensis invasion can be limited by D. variabilis immune response (Ceraul et al. 2007, 2008, 2010).
Despite the adverse effects of the rickettsiae on the tick and the success of the tick's immune system in response against rickettsiae, some infected ticks do survive and maintain the pathogen cycling through mammal hosts. Indeed, D. variabilis rickettsiostatic agents may preserve the vector competency of D. variabilis by reducing the level of infection so that not all infected ticks succumb to the lethal effects of rickettsial colonization (Ceraul et al. 2010). Additionally, Dermacentor ticks have advantages that may contribute to the rapid focal magnification of the pathogen in ticks and vertebrate hosts. They are able to acquire rickettsial infection by cofeeding on a host with a R. rickettsii–containing tick, and infected adults have the ability to attach but only partially feed, and then attach a second time and transmit rickettsiae to a new host (Niebylski et al. 1999, Parola et al. 2005). Thus, a cohort of commensal ticks and hosts can amplify the pathogen to produce a nidus of rickettsial transmission. If humans are in proximity (e.g., interacting with tick-infested dogs), they too can become infected.
Perhaps rickettsial infection was present in some of the 5286 negative ticks, but at a level too low to detect. However, the assays used in the HTTK are very sensitive; for example, the ompB qPCR assay can detect consistently three genomic equivalents per reaction (Jiang et al. 2005b). Furthermore, these ticks were removed from humans; 10.3% appeared to be partially engorged, 4.4% fully engorged, and even the ticks that appeared unengorged (85.3%) had begun feeding, initiating the physiological effect of the bloodmeal on rickettsial development. Multiplication of R. rickettsii increases during tick feeding (Wike and Burgdorfer 1972), and reactivation of virulence in R. rickettsii has been shown to begin after only 10 h of feeding (Hayes and Burgdorfer 1982), at which time D. variabilis adults do not exhibit signs of engorgement (Lisa Coburn, Manager, Tick Rearing Facility, National Tick Research & Education Resource, Oklahoma State University, personal communication). Therefore, it is probable that any R. rickettsii infection would be present at detectable levels in these ticks removed from humans.
R. rickettsii infection has been recently identified in additional tick species at rates far higher than typically detected in D. variabilis; these species may be contributing to the human case numbers of RMSF. In 2005, a cluster of human RMSF cases in Arizona was associated with Rhipicephalus sanguineus ticks; 2/70 (3%) of the ticks collected were infected with R. rickettsii, and R. sanguineus was the only tick species found in the area (Demma et al. 2005). R. rickettsii has also been detected in populations of R. sanguineus and Amblyomma americanum in areas where there are sympatric populations of D. variabilis: in Georgia, Garrison et al. (2007) detected R. rickettsii in 1/8 (12.5%) R. sanguineus; in California, Wickso et al. (2007) detected R. rickettsii in 1/62 (2%) R. sanguineus; Berrada et al. (in press) detected R. rickettsii in 4/271 (1.5%) questing unfed adult A. americanum ticks from a single site in Kansas; and Heise et al. (2010) identified sequences closely resembling R. rickettsii in questing unfed A. americanum ticks from Georgia and Oklahoma.
R. sanguineus collected in the United States have been shown to transmit R. rickettsii under experimental conditions (Parker et al. 1933), and this tick is recognized as a vector of R. rickettsii in Mexico (Burgdorfer et al. 1975, Demma et al. 2005); therefore, it was not surprising to also find R. rickettsii naturally infecting U.S. ticks. A. americanum has been incriminated as a vector of R. rickettsii for nearly 100 years (Cooley and Kohls 1944), and can also transmit R. rickettsii under experimental conditions (Parker et al. 1933). However, virulent strains of R. rickettsii also have lethal effects on this tick (Burgdorfer and Brinton 1975), and contemporary techniques of analysis have identified very high prevalences R. amblyommii, not R. rickettsii in A. americanum (Mixson et al. 2006, Apperson et al. 2008, Stromdahl et al. 2008, Smith et al. 2010, Jiang et al. 2010b). Because maintenance of multiple infections of Rickettsia spp. in ticks is believed to be prevented by transovarial interference (the resistance of the tick ovaries to infection with more than one rickettsial species), A. americanum is not currently regarded as a vector of RMSF (Goddard and Norment 1986, Weller et al. 1998, Childs and Paddock 2003). However, recent evidence suggests that rickettsial coinfection of ticks may occur: a single D. variabilis infected with three species of rickettsiae (Rickettsia bellii, R. montanensis, and R. rickettsii) was collected in Ohio (Carmichael et al. 2010), and two of the four Kansas A. americanum infected with R. rickettsii were also carrying R. amblyommii (Berrada et al. 2009). Even if R. rickettsii infects A. americanum only infrequently, the vector potential of this tick is magnified because of its abundance and wide distribution in the United States, and the frequency with which it attacks humans.
Inclusion of these tick species into the etiology of RMSF could offer an answer to one of the more perplexing questions surrounding human cases. Why do RMSF patients fail to recall tick bites? Masters et al. (2003) review of the literature reports that ∼40% of cases have no history of tick bite. Only the adults of D. variabilis attack humans; D. variabilis adults are large ticks, not easily missed, whereas nymphal and larval R. sanguineus and A. americanum are tiny and frequently overlooked. Immatures of A. americanum are notorious man biters, and many tickbite victims experience multiple concurrent bites. Immatures of R. sanguineus will also feed on humans, and a recent study of experimentally infected R. sanguineus larvae revealed that infection with R. rickettsii had no significant lethal effect on engorged larvae, and the infected larvae successfully molted to infected nymphs (Labruna et al. 2008). Each year, a small number of R. sanguineus are submitted to the HTTK (Table 4). Of the 55 ticks received, 42% were adults and 58% were immatures, including 13 R. sanguineus nymphs removed at one time from a 2-year-old boy. Unlike other anthropophilic ticks in the U. S., R. sanguineus is adapted to survive and reproduce in dry indoor environments (Uspensky and Ioffe-Uspensky 2002, Yoder at al. 2006). This tick is often associated with small children (Goddard 1989, Carpenter et al. 1990), 57% of the tick bite victims whose age is reported in Table 3 are under 5 years old, and in the USAPHC Tick-Borne Disease Laboratory it has been given the epithet “crib tick.” Perhaps the ticks that go unrecognized by RMSF patients are immature stages of R. sanguineus and A. americanum.
The areas of the U.S. hyperendemic for RMSF are those where D. variabilis ticks exist in sympatry with populations of A. americanum and Amblyomma maculatum, a newly recognized vector of rickettsial pathogens (Jiang et al. 2010a, Paddock et al. 2010), but RMSF is rarely reported from areas with of D. variabilis but no Amblyomma populations (Fig. 3). Each year military installations in Minnesota, Wisconsin, and Pennsylvania submit to the HTTK hundreds of D. variabilis, but almost no A. americanum and A. maculatum (Stromdahl et al. 2001). Despite the large numbers of D. variabilis actively attacking humans for long seasonal periods, these locations report no or few cases of RMSF (Adjemian et al. 2009). It could be that interactions among the sympatric species of ticks, for example, cofeeding, are involved in the development of a focus of RMSF transmission. On the other hand, the overlap of foci of RMSF cases and areas of Amblyomma populations might indicate the misdiagnosis as RMSF of diseases actually caused by other rickettsiae vectored by A. americanum or A. maculatum (Paddock et al. 2008). In areas of sympatric D. variabilis and A. americanum populations, ehrlichiosis vectored by A. americanum, even fatalities due to Ehrlichia chaffeensis, might possibly be attributed to RMSF. A. americanum infected with E. chaffeensis and Ehrlichia ewingii can be coinfected with R. amblyommii (Mixson et al. 2006), and a tick bite victim infected with an Ehrlichia might also be infected with R. amblyommii. RMSF and ehrlichiosis cases can have similar acute clinical appearances, and if RMSF is suspected and ehrlichiosis is overlooked, crossreaction in the RMSF serology could lead to a diagnosis of RMSF, not ehrlichiosis (Carpenter et al. 1999, Apperson et al. 2008).
Approximate distribution of Amblyomma americanum (Childs and Paddock 2003), Amblyomma maculatum (Paddock et al. 2008), Dermacentor variabilis (Centers for Disease Control and Prevention:
Although the detection of one R. rickettsii–positive D. variabilis after 12 years and 5285 ticks reconfirms its status as a vector of the most lethal tick-borne disease in the United States, the infrequency of infection implies that other factors must account for the number of cases of RMSF reported from the range of D. variabilis: many of the cases could be caused by other species of Rickettsia crossreacting in diagnostic tests, and other species of ticks might be transmitting R. rickettsii. To reflect this new paradigm of rickettsial disease, in 2010, the CDC case reporting category will be changed from “RMSF” to “Spotted Fever Rickettsiosis (including RMSF) (Openshaw et al. 2010). Many basic questions about RMSF remain unanswered. Further studies are needed on R. rickettsii, its tick vectors, and their hosts to elucidate the etiology and identify the cast of characters and cascade of events contributing to the development of a focus of virulent RMSF.
Footnotes
Acknowledgments
We thank SPC Roshar Pompey, Danielle Thomas, PFC John Torres, and SPC Sean Welch for technical assistance. The views expressed in this article are those of the author and do not reflect the official policy or position of the Department of the Army, Department of the Navy, Department of Defense or the U.S. Government. This work was partially supported by the Global Emerging Infections Surveillance and Response System, a Division of the Armed Forces Health Surveillance Center, work unit number 0000188M.0931.001.A0074. While an employee of the U.S. Government (EYS, MV, ALR) this work was prepared as part of my official duties. Title 17 U.S.C. §105 provides that Copyright protection under this title is not available for any work of the United States Government.
Disclosure Statement
No competing financial interests exist for the authors.