Rickettsiae in Gulf Coast Ticks, Arkansas, USA

To determine the cause of spotted fever cases in the southern United States, we screened Gulf Coast ticks (Amblyomma maculatum) collected in Arkansas for rickettsiae. Of the screened ticks, 30% had PCR amplicons consistent with Rickettsia parkeri or Candidatus Rickettsia amblyommii.


The Study
We screened 112 A. maculatum ticks collected during March 2006-January 2008 from 22 dogs (Canis lupus familiaris) and 95 A. maculatum ticks collected during the 2008 hunting season from 52 white-tailed deer (Odocoileus virginianus Boddaert) for rickettsial DNA.
Collectors removed specimens; stored them in vials containing 100% ethanol; and recorded tick collection date, location, and host (6). Ticks were identifi ed by species, sex, life stage, and engorgement (9). Each sample was bisected longitudinally with a razor blade and subjected to the extraction procedure by using QIAGEN DNeasy (QIAGEN, Valencia, CA, USA) following the manufacturer's protocols.
Tick DNA extracts were screened for SFG Rickettsia spp. DNA by PCR by using genus-specifi c primers for the citrate synthase (gltA) (10) and rickettsial outer membrane protein B (rompB) (11) genes. Reaction products were analyzed (12), and positive amplicons for gltA (513 bp) and rompB (578 bp) were sent to the University of North Texas Health Science Center (Fort Worth, TX, USA) for sequence determination. At least 1 amplicon from each host was sequenced to determine the Rickettsia species identity. PCR products were hydrolyzed with ExoSAP-IT (USB Corporation, Cleveland, OH, USA), and sequence determination was performed by using a BigDYE Terminator v.3.1 Cycle Sequencing Kit (Applied Biosystems, Inc., Foster City, CA, USA) followed by capillary electrophoresis on an ABI PRISM 310 Genetic Analyzer (Applied Biosystems, Inc.) (13).
Sequences were edited, aligned, and analyzed with Sequencher 4.7 (Gene Codes Corporation, Ann Arbor, MI, USA) and compared with sequences in GenBank (National Center for Biotechnology Information, Bethesda, MD, USA). BEAST version 1.4.2 software (http://beast.bio. ed.ac.uk/Main_Page) was used to infer phylogenetic relationships and create dendrograms (14). The consensus tree ran for 10 6 generations with a burn-in of 2 × 10 4 . Established methods were used (12) to conduct parsimony bootstrap and maximum-likelihood analyses. Maximum-likelihood and unweighted parsimony analyses on the alignments were performed by using the branch and bound algorithm of PAUP* 4.0b10 (http://paup.csit.fsu.edu). Outgroup taxa were obtained from GenBank.
Of the 207 ticks, 62 were positive for Rickettsia spp. DNA by PCR. Nineteen ticks were positive for gltA only, 12 were positive for rompB only, and 31 were positive for both genes (Table). Of the ticks collected from white-tailed deer, 28 were positive, and those amplicons were 100% homologous with Candidatus Rickettsia amblyommii from GenBank (FJ455415, EU7228827, AY388899) (Table, Figure). Of the positive ticks collected from dogs, 3 had sequences with 100% similarity to either rompB (AF123717) or gltA (EF102236) of R. parkeri. A single tick (unengorged male) had a sequence 98% similar to GenBank sequences EF219464 (rompB) and EF451001 (gltA). The remaining 30 ticks collected from dogs that were positive all produced amplicons with 100% sequence identity to Candidatus R. amblyommii gltA (EF450708). However, rompB sequences  Figure). In total, 3 ticks collected from 3 different canine hosts produced sequences 100% identical to those of R. parkeri rompB (AF123717) and gltA (EF102236). Candidatus R. amblyommii sequences were identifi ed in 29 ticks collected from 13 dogs and 25 ticks collected from 25 deer. The resulting Bayesian tree showed weak support (consistency index 0.792, tree length 159) (Figure). Neighbor-joining and maximum-likelihood trees supported the GenBank homologies.

Conclusions
We report the identifi cation of SFG rickettsiae in A. maculatum ticks collected from Arkansas, specifi cally R. parkeri, Candidatus R. amblyommii, and an uncharacterized Rickettsia sp. sequence with high homology to GenBank sequence no. EF219464. Identifi cation of these rickettsiae may be a public health concern given their recent association with cases of spotted fever (4,7,8). The risk for spotted fever transmission to humans is unknown but may be of concern to public health offi cials in Arkansas because of canid-human relationships and habitat fragmentation that has moved deer ranges closer to human habitation. Additional investigations of the distribution of A. maculatum ticks, the pathogenesis of Rocky Mountain spotted fever, and the ticks' relationship to human disease should be conducted.

Acknowledgments
We thank the personnel associated with the Arkansas Game and Fish Commission and Arkansas Veterinary Medical Association involved with collecting tick samples.
This research was supported in part by the University of Arkansas, Arkansas Agricultural Experiment Station. Ms Trout is a PhD candidate in the Department of Entomology at the University of Arkansas. Her dissertation focuses on the spatial identifi cation and genetic characterization of ticks and their relationship with Borrelia and Rickettsia species within Arkansas.