A novel Ehrlichia transmitted by Amblyomma americanum (lone star ticks) was recently discovered in Panola Mountain State Park, Georgia, USA. The "Panola Mountain Ehrlichia" (PME), which is closely related to E. ruminantium, caused transient febrile illness, followed by chronic latent infection, in a goat [1]. This agent was also associated, using PCR and sequencing, with a case of human illness following a bite from a nymphal Amblyomma acquired at Panola Mountain State Park [2]. White-tailed deer (Odocoileus virginanus) are a probable vertebrate reservoir for PME in the United States; deer are susceptible to infection, are naturally exposed to the agent, and are competent reservoirs for tick transmission of this agent [3].

Ehrlichia ruminantium is endemic in southern Africa and the Caribbean, and its pathogenicity in cattle, sheep, and goats varies from mild febrile illness to fatal heartwater [4-7]. Several species of Amblyomma vector E. ruminantium in Africa and A. variegatum is the vector in the Caribbean [4]. This exotic disease has not been detected in the United States, but it could be introduced into the country by the importation of animals and tick vectors from endemic areas [8,9]. White-tailed deer are also susceptible to infection with E. ruminantium [10], and could provide a sylvatic reservoir for E. ruminantium in the United States. Thus, the Ehrlichia discovered at Panola Mountain State Park could be a divergent strain of E. ruminantium that was recently introduced to this country. This park is located within the Atlanta, Georgia, metropolitan area and is less than twenty miles from the Hartsfield International Airport, a port of entry for imported animals and animal products.

To evaluate the geographic distribution and public health risk posed by this emerging agent, we developed a sensitive nested PCR assay based on the gltA (citrate synthase) gene, which is conserved within species of Ehrlichia [11], and screened DNA extracts from A. americanum collected from vegetation and submitted through four different human-biting tick surveillance programs. All PCR amplicons were confirmed by sequencing the gltA fragment, and the genetic diversity of this novel agent was evaluated by amplification and sequencing of the variable Major Antigenic Protein 1 gene (map1) [12].

Amblyomma americanum were collected from 203 counties in 23 states, representing much of the geographic range of this tick (Figure 1).

Questing ticks were collected from vegetation at 32 sites in 8 states. Of 1964 questing adult ticks, including 867 adult males and 1097 adult females, 14 ticks (10 males, 1.15%, and 4 females, 0.36%) were positive for PME (Table 1). The overall prevalence of infection in questing ticks was 0.71%, and the prevalence of infection in sites with infected ticks ranged from 1.32–2.88%.

Human-biting ticks were obtained, via surveillance programs, from 21 states. Of 1835 A. americanum obtained from surveys of human-biting ticks, 22 DNA samples from 5 states were positive (Table 2). Six of these positive samples were pools of 2–9 ticks, with the remainder representing individual ticks. Given the low prevalence of infection, we assumed that only one tick in a positive pool was likely to be positive; the infection rate among human-biting ticks was therefore 22/1835, or 1.20%. Eleven of the positive samples were from nymphal ticks (1.12%), three from males (0.71%), five from females (1.17%), and five from pools containing more than one life stage.

Overall, 35 positive ticks were obtained from 18 counties in ten states: Florida, Georgia, Kentucky, Maryland, Missouri, New Jersey, New York, Ohio, Oklahoma, and Virginia (Figure 1). Four of these ticks were co-infected with E. chaffeensis and/or E. ewingii (Table 3). None of the positive ticks were co-infected with "Borrelia lonestari". All of the positive ticks were collected between the months of April and September. Positive ticks were identified from collections made in 1998 through 2006; on Fire Island, in Suffolk Co., New York, individual positive ticks were collected from vegetation in both 1998 and 2003, suggesting that the agent was maintained in this area over several years. Similarly, positive human-biting ticks in Burlington Co., New Jersey were collected in 2001 and 2006, and ticks collected from Scioto Co., Ohio in 2000 and 2001 were positive.

The nested gltA amplicon was sequenced from all positive ticks. All positive amplicons had sequences consistent with PME; E. chaffeensis or E. ewingii were not amplified. The gltA sequence was highly conserved and was identical to the reported PME sequence (GenBank: DQ363995) in 35/36 ticks; one sequence included a single nucleotide polymorphism (EU272407).

A portion of map1 was successfully amplified and sequenced from 31 of the 36 positive tick DNA samples; the outer PCR amplicon, when available, produced 730 bp of sequence, and the inner amplicon produced 384 bp. Two map1 genogroups were identified and were seen both in questing ticks and in human-biting ticks (Table 3): one group was 100% homologous to the sequence reported from PME (GenBank:DQ324368), and the other group was 89.5% similar to PME, with two polymorphic loci identified within this group of sequences. There were five base pair mismatches between the sequences in the latter genogroup and the outer pair of PCR primers previously developed for the detection of the map1 gene from this agent (Pmap-38F/Pmap-581R, [1]). The two different PME clades were not geographically isolated; both clades were detected in ticks from Jones County, Georgia and Monmouth County, New Jersey. Additionally, one pool of ticks from Ocean County, New Jersey, contained both genotypes. The amplicons produced from this mixed template pool were cloned and sequenced. Sequence analysis indicated representatives of both genotypes were present, as well as several sequences that appear to be combinations of these two genotypes and may indicate a crossover event in map1 (GenBank: EU272356–EU272360).

Phylogenetic reconstructions based on the map1 amino acid sequences were attempted (Figure 2) using all of the DNA sequences that were 730 bp in length. The sequences from the ticks formed two distinct clades, with 100% bootstrap support for each clade (1000 psuedoreplicates), and they formed a cluster within the larger taxonomic group of E. ruminantium strains in 82% of the replicates. The PME/E. ruminantium clade was a sister taxa to E. chaffeensis and E. canis in 100% of the replicates. Bootstrap support for separation of PME and E. ruminantium into separate taxa was weak (< 500/1000).

We detected the novel Panola Mountain Ehrlichia sp. (PME), recently discovered in A. americanum from north central Georgia, USA, in 36 ticks from 18 counties in ten states. DNA from PME was detected in ticks from the traditional southeastern range of A. americanum as well as from recently established populations in the northeastern USA. The overall prevalence of infection with PME, approximately 1–3%, was similar to that previously reported from Panola Mountain [1]. Similar prevalences of E. ewingii and E. ruminantium are seen in populations of Amblyomma infected with these agents [14,17-19].

Earlier assays for PME were based on the gene encoding the Major Antigenic Protein 1 (map1) [1]; this gene can be highly variable [12], and polymorphisms in the primer annealing sites could prevent the detection of some genotypes of PME. We therefore developed a sensitive and specific nested PCR assay based on the conserved gltA gene. As expected, gltA amplicons obtained from PME-infected ticks were 99.5–100% identical to each other and to the sequence previously reported from this agent. Significant mismatches were seen between the previously reported map1 primers and the map1 sequences from PME-infected ticks. Additionally, map1 could not be amplified from all of the gltA-positive ticks, presumably due to primer site mismatches, confirming that detection of PME using map1 assays is limited by the variability of this gene.

This gene is, however, a valuable tool to assess genetic variability in PME across the geographic range of A. americanum. Using phylogenetic reconstruction of map1 sequences, we identified two clades, one of which includes the genotype previously reported from Panola Mountain. Both PME clades were closely associated with each other and were more closely related to E. ruminantium than to any other species of Ehrlichia, similar to previous reports. Sequences that suggest the possibility of crossover between these two genotypes were also obtained from a few clones from a single DNA sample; collection and analysis of other specimens, to determine if this is an isolated finding, are ongoing.

The extensive geographic distribution of PME and the presence of genetic variability within the species suggest that this agent was not recently introduced to the United States. A recent point source introduction should have resulted in a limited geographic distribution of a single genotype or possibly of closely related genotypes with minimal divergence. Although two distinct genotypes of PME were detected, less genetic diversity was seen than is reported to occur between African strains of E. ruminantium, suggesting that the map1 gene of PME might evolve more slowly than that of its African counterparts, that the diversity of susceptible Amblyomma and vertebrate species in Africa might have contributed to the diversity of E. ruminantium on that continent, or that the introduction of E. ruminantium-like bacteria to North America might have been more recent than the introduction into Africa.

Finally, the lack of reports of heartwater, or heartwater-like disease, in ruminants, wild or domestic, from areas in which PME was detected provides indirect support that this agent has low pathogenicity for these animals. However, this emerging ehrlichiosis has been associated with a human case of illness following the bite of a nymphal A. americanum from Panola Mountain State Park. Amblyomma americanum is an abundant and aggressive human-biting tick throughout its geographic range [20]. Our detection of PME in 22 human-biting ticks underscores the potential public health risk of this emerging tick-transmitted disease. The broad geographic range and patchy distribution of ticks harboring PME suggests that human cases could occur throughout the eastern United States but be sporadic in nature and therefore difficult to diagnose. Sensitive and specific PCR assays, performed on whole blood samples collected during the acute febrile period, might assist in the rapid clinical diagnosis of undifferentiated ehrlichial infections in people with a history of tick bite by A. americanum.

The authors declare that they have no competing interests.

AL conceived of the study, carried out portions of the molecular genetic studies, participated in the sequence alignment and analysis, performed statistical analysis, and drafted the manuscript. TM, ES, and KB participated in the design of the study, collected tick samples, carried out portions of the molecular genetic studies, and helped to draft the manuscript. PW participated in the design of the study, collected tick samples, carried out portions of the molecular genetic studies, participated in the sequence analysis, and helped to draft the manuscript. MY and RF carried out portions of the molecular genetic studies, participated in the sequence alignment, and helped to draft the manuscript. LG, PF, and DK participated in the design of the study, collected tick samples, and helped to draft the manuscript. All authors read and approved the final manuscript.

The pre-publication history for this paper can be accessed here: