Detection of Borrelia lonestari, Putative Agent of Southern Tick-Associated Rash Illness, in White-Tailed Deer (Odocoileus virginianus) from the Southeastern United States

Abstract

To determine if white-tailed deer may serve as a reservoir host for Borrelia lonestari, we used a nested PCR for the Borrelia flagellin gene to evaluate blood samples collected from deer from eight southeastern states. Seven of 80 deer (8.7%) from 5 of 17 sites (29.4%) had sequence-confirmed evidence of a B. lonestari flagellin gene by PCR, indicating that deer are infected with B. lonestari or another closely related Borrelia species. Our findings expand the known geographic range of B. lonestari and provide the first evidence of this organism in a vertebrate other than humans.

In the northeastern United States, Borrelia burgdorferi sensu stricto, the causative agent of Lyme disease, is maintained in nature through a cycle involving the white-footed mouse (Peromyscus leucopus) and other rodents as primary reservoir hosts and the black-legged tick, Ixodes scapularis, as vector (9, 20, 33). Although they mount an antibody response following infection, white-tailed deer appear unable to infect ticks with B. burgdorferi and thus do not serve as a reservoir host for this organism (26, 30, 34). The primary role of deer in the natural history of B. burgdorferi is that of serving as a major host for the adults of I. scapularis, allowing vector tick populations to flourish and spread (29, 34).

In the southeastern and south central United States, Lyme disease caused by B. burgdorferi sensu stricto occurs much less commonly than it does in the northeastern United States (3, 29). However, a Lyme disease-like illness that develops following the bite of the Lone Star tick, Amblyomma americanum, has been described (2, 7, 12, 17, 27, 28). Individuals affected with this illness, termed “southern tick-associated rash illness” (STARI), commonly develop a localized expanding circular skin rash (erythema migrans) at the site of the tick bite similar to that seen with classic Lyme disease. A mild illness characterized by generalized fatigue, headache, stiff neck, and occasionally fever and other constitutional signs accompanies the rash (7). The disease appears to respond to antibiotic treatment (7, 28).

STARI, which is also referred to as “Masters' disease” in recognition of the physician who first described its clinical presentation, has been attributed to infection with an as-yet-uncultivated spirochete tentatively referred to as Borellia lonestari (14). Cases consistent with this clinical presentation have been reported from several southeastern and south central states, including Missouri, Maryland, Georgia, South Carolina, and North Carolina (2, 7, 12, 17, 27, 28). The majority of patients with STARI do not have laboratory evidence of infection with B. burgdorferi sensu stricto (7).

Because the etiologic agent has not yet been isolated in culture, PCR of 16S ribosomal DNA (rDNA) and the flagellin gene has been used to identify the presence of B. lonestari in humans and ticks. Molecular evidence of B. lonestari has been found in wild-caught A. americanum ticks from Alabama, Tennessee, Texas, and New Jersey, as well as in a tick removed from an individual who had traveled to North Carolina and Maryland before presentation (4, 6, 14; T. L. Stegall-Faulk et al., unpublished data). Despite relatively widespread documentation of this organism in ticks, a vertebrate reservoir host that could be responsible for maintaining infection in the tick population has not yet been identified.

White-tailed deer are a preferred host for larvae, nymphs, and adults of A. americanum (19). Because of their intimate association with Lone Star ticks and because of their important role in the natural history of Ehrlichia chaffeensis, which is also vectored by Lone Star ticks, we decided to evaluate the infection status of white-tailed deer for B. lonestari from areas where A. americanum ticks are known to occur on deer (22).

Archived blood samples collected in the course of other research were identified from white-tailed deer from several populations in the southeastern United States where deer are known to be parasitized by Lone Star ticks. For collection of samples, white-tailed deer were killed by cervical gunshot, and whole blood was obtained by sterile heart aspiration. For all but two of the sites (Hilton Head Island and Kiawah Island, S.C.), each deer was examined, and the ticks found were identified and recorded; ticks were not collected from the deer sampled at Hilton Head Island and Kiawah Island, although A. americanum is known to be present at both locations (W. R. Davidson, unpublished data). After collection, either whole blood was aliquoted and stored at −20°C, or erythrocytes were lysed and an erythrocyte-free fraction was collected by centrifugation and stored at −20°C (22).

Blood samples obtained from deer collected from 17 locations in eight different states were used in this study. Samples were collected from deer from Arkansas (Cache River National Wildlife Refuge [NWR], Felsenthal NWR, Pea Ridge National Military Park, Shirey Bay/Rainey Brake Wildlife Management Area [WMA], and White River NWR), Florida (St. Vincent NWR), Georgia (St. Catherine's Island), Kentucky (Ballard WMA, Fort Knox, and West Kentucky WMA), Louisiana (Tensas River NWR), Mississippi (Dahomey NWR and Hillside NWR), North Carolina (Cape Hatteras National Seashore and Mattamuskeet NWR), and South Carolina (Hilton Head Island and Kiawah Island). Samples from 2 to 10 deer from each site were tested.

DNA was extracted from whole blood or from erythrocyte-free preparations by using the GFX genomic blood DNA purification kit (Amersham Biosciences Corp., Piscataway, N.J.) according to the manufacturer's instructions.

Samples were tested for the presence of Borrelia DNA by using a nested PCR designed to amplify a 330-bp region of the flagellin gene (flaB) (4). Primers FLALL (5′-ACATATTCAGATGCAGACAGAGGT-3′) and FLARL (5′-GCAATCATAGCCATTGCAGATTGT-3′) were used in the primary reaction, and primers FLALS (5′-AACAGCTGAAGAGCTTGGAATG-3′) and FLARS (5′-CTTTGATCACTTATCATTCTAATAGC-3′) were used in the secondary reaction. Both primary and secondary reactions were assembled in 25-μl volumes containing 5 U of Taq DNA polymerase (Promega, Madison, Wis.), 50 pmol of each primer, 200 μM deoxynucleoside triphosphate (dNTP), 50 mM KCl, 10 mM Tris-HCl (pH 9.0), 0.1% Triton X-100, and 2.5 mM MgCl₂. Ten microliters of template DNA was added to the primary reaction mixture, and 1 μl of primary PCR product was added as a template for the secondary reaction. The thermal cycling conditions for both reactions were 95°C for 3 min, followed by 40 cycles of 95°C for 1 min, 55°C for 1 min, and 75°C for 1 min. Five microliters of the secondary PCR was loaded onto a 2% agarose gel with Tris-acetate-EDTA buffer, separated for 25 min at 90 V, and visualized with a UV transilluminator. Negative controls using water as the template material were included in each step of the assay (extraction, primary PCR, and secondary PCR). Positive controls, consisting of deer blood samples PCR positive for the B. lonestari flagellin gene as confirmed by sequence analysis (described below), were also included with each set of reactions. To reduce the possibility of contamination of samples, DNA extraction and each of the two amplification cycles were performed in distinct, dedicated areas.

Amplicons from all samples positive by PCR were purified and concentrated with a Microcon 100 microconcentrator (Amicon, Inc., Beverly, Mass.) according to the manufacturer's directions. Purified amplicons were submitted to the core facility at The University of Georgia for sequencing according to the Applied Biosystems, Inc., protocol for the ABI 373A automated sequencer (Perkin-Elmer, Foster City, Calif.). The resultant sequences were aligned with DNAsis Mac v 2.0 and compared with all published Borrelia sp. flagellin gene sequences available in the GenBank database.

A. americanum was present on at least one deer from each location from which deer samples were examined for ticks (15 of 15 locations). Other ticks found on these deer included Amblyomma maculatum (9 of 15 locations), I. scapularis (3 of 15 locations), and Ixodes affinis (1 of 15 locations). Evidence of B. lonestari infection was found in 7 of the 80 deer (8.7%) tested from a total of 5 of the 17 sites (29.4%). Positive animals were detected at Sea Pines Plantation, Hilton Head Island, Beaufort County, S.C. (SP041); Pea Ridge National Military Park, Benton County, Ark. (588-3 and 588-6); Bodie Island, Cape Hatteras National Seashore, Dare County, N.C. (591-1 and 591-2); Mattamuskeet National Wildlife Refuge, Hyde County, N.C. (599-2); and St. Catherine's Island, Liberty County, Ga. (600-2).

Sequence analysis of the amplicons from each of the seven positive deer samples revealed 100% identity with B. lonestari flagellin gene (flaB) sequences available in GenBank (Table 1). The B. lonestari flagellin sequences generated from three of our deer samples were identical to sequences of that gene reported from ticks in Texas (U26704), New Jersey (U26705), and Alabama (AF298653 and AF298654) (4, 6). Sequences from the other four positive deer samples contained a 3-bp insert (5′-AGA-3′) at positions 330 to 332 and were identical to sequences reported in ticks from Tennessee (AF408410) and from a tick (AF273671) removed from the skin of a patient (AF273670) who had traveled to North Carolina and Maryland prior to presentation (14; Stegall-Faulk et al., unpublished).

TABLE 1.

Comparison of B. lonestari flagellin B gene sequences from GenBank with those amplified from white-tailed deer blood samples

Sequence name	GenBank accession no.	Source or reference	Sequence sourcea	Nucleotide difference at positionb:
				330	331	332	348
TX	U26704	4	LST	*	*	*	G
NJ	U26705	4	LST	*	*	*	A
NC/MD 1	AF273670	14	Human	A	G	A	A
NC/MD 2	AF273671	14	LST	A	G	A	A
AL 1	AF298653	6	LST	*	*	*	A
AL 2	AF298654	6	LST	*	*	*	A
TN	AF408410	Stegall-Faulk et al., unpublished	LST	A	G	A	A
SP041	AF538846	This study	WTD	*	*	*	A
588-3	AF538847	This study	WTD	A	G	A	A
588-6	AF538848	This study	WTD	A	G	A	A
591-1	AF538849	This study	WTD	*	*	*	A
591-2	AF538850	This study	WTD	A	G	A	A
599-2	AF538851	This study	WTD	*	*	*	A
600-2	AF538852	This study	WTD	A	G	A	A

^aLST, Lone Star tick. WTD, white-tailed deer.

^bNumbers correspond to nucleotide positions of B. lonestari flaB gene sequence U26705. An asterisk indicates no corresponding base at this position.

The causative agent of STARI has not yet been isolated in culture; thus, a full comparison of the organism from human cases and the organism from deer is not possible (32). However, the flagellin gene sequences we amplified from deer blood are identical to those amplified from Lone Star ticks described as harboring B. lonestari and are also identical to the single sequence reported from a patient exhibiting evidence of disease, indicating that these deer were infected with B. lonestari or another closely related Borrelia species (4, 6, 14). Although other animals may also be infected with this organism, this is the first report of evidence of B. lonestari in a vertebrate other than a human.

This study evaluated samples collected only from deer in areas where Lone Star ticks are known to occur. Other ticks, including I. scapularis, I. affinis, and A. maculatum, were found on deer in some of these areas. Of the five areas with deer samples that tested positive for B. lonestari, three had evidence of I. scapularis, one had evidence of I. affinis, and two had evidence of A. maculatum infesting deer. The reported distribution of each of these tick species overlaps with that of A. americanum (1, 5, 8, 15, 36). Although PCR evidence of B. lonestari has not been obtained from ticks other than A. americanum, further investigation of additional tick species as potential vectors of B. lonestari may be warranted (6).

Spirochetes resembling Borrelia spp. have been described in A. americanum from Alabama, Missouri, New Jersey, North Carolina, Oklahoma, and Texas (4, 11, 18, 21, 24, 25, 31). Molecular evidence of B. lonestari has been reported in A. americanum from Alabama, New Jersey, Tennessee, and Texas, as well as in a patient who apparently acquired infection in either North Carolina or Maryland (4, 6, 14). Suspected cases of disease have also been reported from Missouri, Georgia, South Carolina, and Maryland, but molecular assays to ascertain the presence of B. lonestari in those patients were not performed (2, 7, 12, 17, 27, 28). Our data confirm that B. lonestari or a very closely related Borrelia spp. is present in nature in areas of Georgia, South Carolina, Arkansas, and North Carolina.

The Lone Star tick feeds on white-tailed deer in all three parasitic stages of its life cycle, an association that has been shown to be important in the natural history of E. chaffeensis in the southeastern United States (10, 22, 23). A similar situation also may exist with Ehrlichia ewingii (37). Our positive PCR results suggest white-tailed deer are infected with B. lonestari, but the actual role of deer as a reservoir host is unclear. Although we used blood to evaluate the infection status of deer by PCR, skin is a more commonly used tissue for amplification and isolation of Borrelia spp. and could potentially yield a higher percentage of positive samples (13, 16, 35). Additional studies are needed to verify infection in deer and transmission of B. lonestari by feeding ticks.

Nucleotide sequence accession numbers.

The GenBank accession numbers for the B. lonestari flagellin gene (flaB) from the deer reported in this paper are AF538846, AF538847, AF538848, AF538849, AF538850, AF538851, and AF538852.