Case Report

A 74-year-old African American man was evaluated on 24 May 1999 at New York Medical College (NYMC; Valhalla) because of an abdominal skin rash that had been present for ∼4 days, as well as fatigue, cough, and right shoulder discomfort of ∼1 week’s duration. He denied having fever or headaches. The patient was a resident of southern Westchester County, New York. He traveled to Maryland on 6 May and then left Maryland on 6 May to travel to North Carolina, where he stayed from 7 May to 18 May. He returned to Maryland on 18 May and then left on 20 May to return to his home in Westchester County. He had walked in grassy areas in both Maryland and North Carolina. His medical history included a left nephrectomy in June 1998 for a renal cell carcinoma, prostate carcinoma diagnosed in 1997, colonic carcinoma resected in 1985, hypertension, and hypothyroidism

At examination, the patient had an oral temperature of 36.8°C. Results of musculoskeletal, neurologic, pulmonary, and cardiac examinations were normal. A 19×11-cm annular erythematous lesion surrounding an attached, engorged tick was present in the right lower quadrant of the abdomen, and a separate 4×3-cm erythematous lesion was present in the left upper quadrant (figure 1A1B). The patient first noticed the tick on 15 May but did not remove it because he thought it was a skin lesion. The tick was removed at NYMC on 24 May for taxonomic evaluation and microbiologic studies

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Figure 1

Erythema migrans of case patient and point of attachment of female Amblyomma americanum tick. A Skin lesion with central clearing, in right lower quadrant of abdomen. Raised lesion marks indicate from which area the attached tick was removed. B Skin lesion in left upper quadrant of abdomen, below surgical scar

Laboratory evaluation showed normal white blood cell count and differential, hemoglobin level, and platelet count. Results of analysis of serum chemistry were normal, except for an alkaline phosphatase level of 142 U/L (normal range, 35–110 U/L). An electrocardiogram showed premature atrial contractions

Beginning 24 May, the patient received a 14-day course of doxycycline (100 mg orally twice daily). The skin lesions resolved by 4 June. The patient reported that he had returned to his usual state of health by 16 June 1999

Materials and Methods

Skin biopsy and cultureOn 24 May 1999, a 2-mm punch biopsy sample was taken from the edge of the skin lesion in the right lower quadrant of the abdomen, farthest from the site of tick attachment. The skin sample was processed at NYMC for B. burgdorferi culture and PCR studies, as reported elsewhere [11]

Tick methodsThe removed tick was identified as a female of the species A. americanum and was sent to the Centers for Disease Control and Prevention (CDC) at Fort Collins, Colorado, for further study. At the CDC, the tick was cut in half longitudinally. One half was processed for PCR, as described below, and the other half was processed for coculture with Ixodes scapularis embryonic cells (cell line IDE2; a gift from T. Kurtii, University of Minnesota, Minneapolis). The coculture was prepared as described by Kurtii et al. [12], was incubated at 31°C, and was checked by dark-field microscopy every week for 2 months for the presence of spirochetes

PCRDNA from the skin specimen and the patient’s blood was extracted at NYMC by using an IsoQuick nucleic acid isolation kit (Orca Research). Because the skin specimen had been transported from the clinic in BSK-H medium (Sigma), DNA was prepared from both the skin itself and the medium fraction (“supernatant”). At the CDC, DNA from the tick was prepared by maceration of the tick in 1 mL of a guanidinium reagent (DNA STAT-60; Tel-Test) and was isolated according to the manufacturer’s instructions. DNA samples from the patient’s skin biopsy specimen, skin supernatant, and blood and from the tick were exchanged between the NYMC and CDC laboratories, and PCR was done at each site

Nested PCR amplification with broadly-reactive flagellin gene (flaB) primers was done at NYMC and the CDC by using primers FlaLL and FlaRL in the first amplification and primers FlaLS and FlaRS in the second amplification, after the amplicons in the first product were diluted 1:100 with water. Primers, reaction conditions, and thermal cycling were as described elsewhere [9]. Specimens also were tested for the presence of B. burgdorferi DNA by PCR, with species-specific primers directed at the 23S rDNA duplication [11], the 16S-23S rDNA spacer, and ospA [1]. In addition, amplification of the 16S rDNA sequence was done by using broad-range eubacterial primers [13] and primers based on the B. lonestari 16S rDNA sequence [9]

DNA sequencingThe PCR products obtained in each laboratory from the skin-sample supernatant and the A. americanum tick were sequenced at the CDC after purification with a Qiagen PCR kit. Sequencing was done using a Taq DyeDeoxy terminator cycle-sequencing kit (Applied Biosystems) and was run on an Applied Biosystems Model 373A automated sequencer. Sequences were obtained from both DNA strands for each sample. The B. lonestari flaB sequences from skin and tick DNA have been deposited in GenBank (accession nos. AF273670 and AF273671)

Serologic testsSerum samples obtained from the patient during the acute (24 May) and convalescent (2 and 26 June) phases of illness were tested for antibodies to B. burgdorferi by an IgG/IgM ELISA (Wampole Laboratories). IgM and IgG antibodies to B. burgdorferi were measured separately by immunoblotting, using serum diluted 1:100. The B. burgdorferi antigens were from strain B31 (low passage; CDC reference collection) grown in BSK-H medium and were fractionated (15 μg per minigel) by discontinuous SDS-PAGE through 13.5% acrylamide gels; immunoblotting was done by standard procedures [1]

Results

No laboratory evidence of B. burgdorferi infection was found. Culture of the skin biopsy sample in BSK-H medium and coculture of the sample of tick cells failed to propagate B. burgdorferi. Moreover, B. burgdorferi sequences were not detected by PCR of DNA prepared from skin, skin supernatant, or blood, using 3 different sets of B. burgdorferi–specific primers. Acute- and convalescent-phase serum samples were negative for B. burgdorferi antibody by ELISA and immunoblotting. No IgM bands were found to be present on immunoblots for B. burgdorferi antigen; only a single IgG band at the position of FlaB, a 41-kDa antigen, was detected in both the acute- and convalescent-phase serum samples (data not shown)

In contrast, DNA evidence compatible with B. lonestari infection was obtained. Amplification of the flaB using Borrelia genus–wide PCR primers, resulted in products of ∼350 bp from both the patient (DNA from the supernatant of the skin sample) and the tick, at both the NYMC and the CDC (figure 2A). The positive control (figure 2A) was B. burgdorferi DNA from cultured spirochetes, not from skin. During 1991–1997, we detected B. burgdorferi DNA in 65% (192 of 294) of skin biopsy specimens from patients with erythema migrans. It was not uncommon for B. burgdorferi DNA to be detected in a skin-sample supernatant but not in the corresponding macerated skin extract (I.S., unpublished data). No amplicons were detected with DNA produced from the macerated skin sample, from the patient’s blood, or from the negative controls that lacked DNA

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Figure 2

Analysis of DNA from case patient and tick. A, flaB DNA amplified by polymerase chain reaction (PCR). Amplicons were separated by electrophoresis on a 1% agarose gel and were detected by ethidium-bromide staining. Lane M Molecular weight markers, in 100-bp intervals; lane 1 negative control (water); lane 2 case patient DNA from supernatant of a skin biopsy sample; lane 3 tick DNA; lane 4 positive control (Borrelia burgdorferi B31 DNA). B Deduced amino acid sequence of portion of flaB DNA amplified from both patient skin-sample supernatant and the attached tick. Sequences are compared with the equivalent regions of flaB from B. lonestari in ticks from New Jersey, B. miyamotoi HT31, and B. burgdorferi B31. GenBank accession nos. for the reference sequences are AAA99141, BAA07832, and CAA33696, respectively. *Amino acid position that showed sequence heterogeneity in the uncloned amplicons. Serine was deduced from the AGT sequence of tick DNA amplified at both New York Medical College (NYMC; Valhalla) and the Centers for Disease Control and Prevention (CDC; Fort Collins, Colorado) and from patient skin-sample supernatant DNA amplified at NYMC. Threonine was deduced from the ACT sequence of skin-sample supernatant DNA amplified at the CDC. Nos. correspond to the amino acid positions of FlaB in B. miyamotoi

Sequencing of 4 PCR products, from amplicons produced independently in each laboratory from DNA of the skin-sample supernatant and of the tick, revealed that this portion of flaB was identical in host and vector samples, with the possible exception of the amino acid marked by an asterisk in figure 2B. Sequences differed only slightly from those obtained elsewhere for B. lonestari in A. americanum ticks from New Jersey and Texas [9]. All sequences in this study contained an extra AGA at positions 330–332, compared with the published B. lonestari sequences (GenBank U26704-5) and 2 nucleotide differences near the 3′ end of the amplicons. The deduced amino acid sequence contains an extra glutamic acid as a consequence of the AGA insertion at amino acid position 211 (figure 2B). In this study, we found 2 conservative amino acid changes near the carboxy terminus, an isoleucine followed by serine, compared with valine-threonine in the B. lonestari sequences in ticks from New Jersey. Alignment of the deduced amino acid sequences obtained from the skin biopsy sample and the tick with the equivalent regions of FlaB from other Borrelia species confirms that the sequence does not differ significantly from the B. lonestari sequence (figure 2B)

Discussion

The clinical and laboratory data from this study strongly suggest that A. americanum ticks transmit B. lonestari to humans and cause an illness that is indistinguishable clinically from early stage Lyme disease. First, there was a temporal relationship between the bite of an A. americanum tick and erythema migrans at the site of the tick bite and at a remote region. Second, we identified B. lonestari flaB sequences in both the tick and a skin sample obtained from the periphery of the skin lesion. Third, laboratory evidence (culture, PCR, and serologic tests) does not support a diagnosis of infection by B. burgdorferi although the skin lesion resolved after standard antibiotic therapy for Lyme disease

Curiously, our repeated attempts to amplify 16S rDNA fragments both from the biopsy specimen and from the tick extract, using broad-range eubacterial and B. lonestari–specific primers, were unsuccessful. The reason for this failure is unclear, although the relatively larger PCR product expected from such amplification (∼1400 bp vs. 350 bp) and a low number of Borrelia genome equivalents in the samples may account for this observation

The validity of using flaB sequences to establish Borrelia taxonomy has been demonstrated elsewhere [14]. Furthermore, in numerous instances, sequence-based identification of a microbe that had not been cultured provided the initial evidence for disease causation. Another tickborne illness, human monocytic ehrlichiosis, is a pertinent example [15]. The seminal observations from gene sequencing of Ehrlichia chaffeensis provided the impetus for the eventual development of culture methods and specific serum antibody tests

An important question is how many of the other reported cases of erythema migrans or erythema migrans–like illness in southern states and elsewhere in the United States are due to B. lonestari. A Lyme disease–like illness after a bite from an A. americanum tick has been reported in New Jersey [8]. Of 3 series of patients with an illness involving a rash who were from the southern United States, 9%–14% recalled a preceding tick bite that was consistent with that of A. americanum ticks [3, 4, 6]. Because only adult female A. americanum ticks are likely to be identified by patients, these proportions should be regarded as minimum estimates. Masters et al. [5] have described 17 selected patients with erythema migrans–like illness who were from Missouri, all of whom had a preceding bite from an A. americanum tick. Theoretically, B. lonestari–infected A. americanum may be sympatric with B. burgdorferi–infected I. scapularis particularly in states along the eastern seaboard. This coexistence has been documented in New Jersey [2,8]. In such areas, 2 different Borrelia species may cause erythema migrans, and not all erythema migrans will indicate Lyme disease

A specific serologic test for evidence of exposure to B. lonestari is not yet available. We are working to develop and validate such a test, because it will be an important tool in determining the etiology of erythema migrans in individual patients and in understanding the epidemiology of B. lonestari infections

In the United States, B. lonestari may be a new tickborne pathogen of humans. This case report suggests that efforts to assess the public health importance of B. lonestari infection are warranted. This assessment should include concerted clinical and epidemiologic studies and field investigations of vector ticks and their hosts