Borrelia, Ehrlichia, and Rickettsia spp. in Ticks Removed from Persons, Texas, USA

Some tick-borne agents may pose yet-unknown public health risks.

D ata concerning the full distribution of tick-borne agents and their potential relationship to both emerging and characterized illnesses in the southern United States are not widely available. Persons who become ill after a tick bite may be at increased risk because a tick bite may not be considered as the source of the pathogen and because of the length of time that febrile illness may elude effective treatment. Detailed knowledge of the causative agents, their distribution, and their relationship to potential vectors is also lacking. Most tick survey data for microorganisms in the genera Borrelia, Rickettsia, and Ehrlichia have been collected in areas where the associated diseases are considered endemic. Lyme disease, Rocky Mountain spotted fever, or human monocytotrophic ehrlichiosis are not considered to be endemic to Texas. Studies of microorganisms carried in ticks in non-disease-endemic areas might provide information about potentially pathogenic organisms, their vectors, and reservoirs. These data might also provide an opportunity to examine the ecology of emerging zoonoses for which different ecologic determinants for disease transmission may be present.
In 2000, the 77th Texas Legislature Subcommittee on Administration prepared a report addressing the potentially severe nature of tick-borne disease in Texas. As of October 1, 2004, the Tick-Borne Disease Research Laboratory at the University of North Texas Health Science Center (UN-THSC) became the primary facility for testing ticks submitted to the Texas Department of State Health Services (TX DSHS).

Methods
From October 1, 2004, through September 30, 2008, tick specimens were submitted to UNTHSC through the Zoonosis Control Program of the TX DSHS. Only ticks that had been attached to a person were submitted to UNTHSC, where they were screened for the genera Borrelia, Rickettsia, and Ehrlichia with genus-specifi c PCRs. Ticks were identifi ed to the species level by TX DSHS entomologists before being transferred to UNTHSC (1)(2)(3). Poor condition of some specimens made identifi cation by morphologic examination diffi cult. Unidentifi ed ticks were conclusively identifi ed by molecular methods developed at UNTHSC, which used amplifi cation of 12S rDNA (Table 1) and sequence determination (data not shown). Additionally, the identity of any tick containing an organism not previously reported in that species was also confi rmed by the same molecular methods. Of all tick specimens, 10% were screened Borrelia, Ehrlichia, and Rickettsia spp. in Ticks Removed from Persons, Texas, USA by the same molecular identifi cation technique to verify the accuracy of morphologic identifi cation. Ticks were bisected laterally by using aseptic technique and a sterile scalpel blade. For independent verifi cation of results, half of each tick was stored in 100% ethanol at -80°C. For larvae and nymphs, the entire tick was used for DNA extraction. Total DNA was isolated from the second half by using an E.N.Z.A. Mollusc DNA Isolation Kit (Omega Bio-Tek, Inc., Norcross, GA, USA) according to manufacturer's recommended protocol. Extracted DNA was subjected to PCR that used primers for the amplification of the tick's 12S rDNA or Borrelia spp., Ehrlichia spp., and Rickettsia spp. genes ( Table 1).
The locations of PCR setup and PCR product handling were physically separated. Reaction setup was performed in a class II type B2 biological safety cabinet that had been cleaned with 0.6% sodium hypochlorite daily and UV ir-radiated for 30 min before and after each use. To minimize risk for contamination, pipettor sets were dedicated to specifi c functions, i.e., reagent dispensing, template isolation, PCR setup, and template handling. Certifi ed DNA/RNasefree fi lter barrier tips were used to prevent aerosol contamination. PCR setup was never performed in the presence of isolation materials, and reagent handling was separated both physically and temporally from templates. PCR assays were performed in duplicate with appropriate controls.
A typical, initial PCR was performed in a 25-μL reaction volume by using 5 pmol/L of each appropriate primer in conjunction with a fi nal reaction concentration of 1× Ge-neAmp PCR Buffer II (Applied Biosystems, Foster City, CA, USA), 160 ng/μL bovine serum albumin, 1.0 mmol/L MgCl 2 , 200 μmol/L of each dNTP, 1.25 U Amplitaq (Applied Biosystems), and 10 μL of template. To establish the species of the tick specimen, we amplifi ed 12S rDNA with Verifi cation of amplifi cation was performed by agarose gel electrophoresis, followed by staining with 1X SYBR Green I (BioWhittaker Molecular Applications ApS, Rockland, ME, USA). Amplicons were examined with a UVP EC3 Imaging System (UVP, LLC, Upland, CA, USA) and subsequently analyzed by VisionworksLS Image Acquisition and Analysis Software (UVP, LLC). Secondary PCR systems (Table 1) were used to confi rm positive results and did not contain primers that would amplify control DNA commonly used in the laboratory. Unincorporated primers were removed from samples producing amplicons before sequence determination by using ExoSAP-IT (USB Corporation, Cleveland, OH, USA).
DNA sequencing was performed for both strands of the PCR amplicons by using a BigDye Terminator Cycle Sequencing Kit, version 3.1 (Applied Biosystems). Unincorporated dye terminators were removed before electrophoresis by using Performa DTR Gel Filtration Cartridges (Edge BioSystems, Gaithersburg, MD, USA). Capillary electrophoresis was performed by using an ABI PRISM 310 Genetic Analyzer (Applied Biosystems). Final sequence analysis and editing was performed by using Sequencer 4.7 (Gene Codes Corporation, Ann Arbor, MI, USA). Using BLASTN, version 2.2.10 (www.ncbi.nlm.nih.gov/blast/ Blast.cgi), we then compared edited sequence data with genetic sequences from characterized examples of Borrelia spp., Rickettsia spp., and Ehrlichia spp. published in Gen-Bank.
DNA sequences consistent with those of Borrelia spp. were derived from A. americanum, A. cajennense, D. variabilis, and I. scapularis ticks. The most commonly encountered Borrelia genetic material demonstrated at least 99% sequence identity or was identical to that of previously sequenced Candidatus Borrelia lonestari isolates (AY850063, AF538852) (19). Additionally, a borreliae fl aB sequence was generated from 1 D. variabilis tick, which had 94% sequence similarity with that of Candidatus Borrelia texasensis (AF264901) (20) and sequences amplifi ed from an uncultured Borrelia sp. from the bat tick Carios kelleyi (EF688577, EF688579) (21) and (EU492387). The fl aB sequence contained 11 single-nucleotide polymorphisms relative to the corresponding section of AF264901. The Borrelia sp. 16S rDNA sequence generated from the same D. variabilis tick was also identical to that published for Candidatus B. texasensis (AF467976) (20,22). This tick was submitted from Webb County, the same Texas county from which the borreliae that produced GenBank sequence AF264901 were isolated. A single I. scapularis specimen produced the fl aB sequence, which had 99% identity with B. burgdorferi (AE000783) (23).
Genetic data consistent with those from Ehrlichia spp. were observed for A. americanum, A. cajennense , and A. maculatum ticks. Amplicons produced from A. americanum and A. maculatum ticks were 99% similar to the homologous region of the E. chaffeensis disulfi de oxidoreductase gene (dsb) sequences in GenBank (CP000236) (24). A single sample from A. cajennense ticks produced a DNA sequence that was 97% similar to that of the CP000236 se-quence and contained 8 single-nucleotide polymorphisms relative to the similar sequence. Several single-nucleotide polymorphisms locations are at the same position as nucleotide differences identifi ed between the dsb gene of E. ewingii (AY428950) (25) and E. canis (AF403710) (26). The nucleotide polymorphisms found within the dsb gene did not change the predicted amino acid sequence in relation to E. chaffeensis (data not shown).

Discussion
By screening a diverse group of Texas tick species for a range of microorganisms and potential pathogens, we identifi ed several novel associations: Candidatus B. lonestari in A. cajennense ticks, E. chaffeensis in A. cajennense ticks, and A. maculatum, and R. parkeri in D. variabilis ticks (Table 3). Because the geographic distribution of diseases caused by the agents is generally characterized by the distribution of the tick vectors, these fi ndings provide insights regarding the distributions and endemicity of several potential emerging tick-borne agents.
SFGR were the most commonly observed agents in this survey. Both Candidatus R. amblyommii and Candidatus R. cooleyi are not well studied and are of undetermined pathogenicity. Current average SFGR seropositivity in Texas residents is also unknown, yet prior estimates indicate that it is higher than would be assumed from cases of Rocky Mountain spotted fever reported to the TX DSHS (27). Transmission through blood products has been noted previously (28,29). Unreported subclinical infections might cause concern about local blood products and could potentially compromise immunodefi cient transfusion recipients. Additionally, detection of R. amblyommii in questing A. americanum larvae suggests transovarial transmission of the microbe, and the likelihood of pathogen transmission   A. cajennense ticks have been associated with E. ruminantium (33) and spotted fever group Rickettsia spp. (34). According to seropositivity in a human population in Argentina, these ticks have also been suspected of transmitting ehrlichiosis (35). The presence of E. chaffeensis in an A. cajennense tick seems novel. Long et al. (13) suggest an E. ewingii MIR of 7.6% in southcentral Texas A. americanum ticks. Similar results for Ehrlichia spp. in A. cajennense tick populations may be plausible.
Screening ticks for a range of bacterial agents has provided several additional associations. These fi ndings provide insights regarding the distributions and endemicity of potentially pathogenic and emerging tick-borne agents. Some of these tick-borne agents may pose an unknown health risk. Because of the wide distribution of these ticks, accurate assessments of the frequency of bacterial agents in these tick populations, their potential for causing human disease, and the ability for these tick species to act as competent vectors are warranted. Continued study and monitoring will play a vital role in public health assessment for related disease risks.