JOURNAL OF CLINICAL MICROBIOLOGY, Mar. 2004, p. 1123–1128 0095-1137/04/$08.00⫹0 DOI: 10.1128/JCM.42.3.1123–1128.2004 Copyright © 2004, American Society for Microbiology. All Rights Reserved.
Vol. 42, No. 3
Isolation and Characterization of Borrelia hermsii Associated with Two Foci of Tick-Borne Relapsing Fever in California Curtis L. Fritz,1* Lawrence R. Bronson,1 Charles R. Smith,1 Martin E. Schriefer,2 James R. Tucker,1 and Tom G. Schwan3 California Department of Health Services, Sacramento, California1; Division of Vector-Borne Infectious Diseases, National Center for Infectious Diseases, Centers for Disease Control and Prevention, Fort Collins, Colorado2; and Laboratory of Human Bacterial Pathogenesis, Rocky Mountain Laboratories, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Hamilton, Montana3 Received 7 July 2003/Returned for modification 23 September 2003/Accepted 28 November 2003
Relapsing fever, caused by the spirochete Borrelia hermsii and transmitted by the soft tick Ornithodoros hermsi, is endemic in many rural mountainous areas of California. Between 1996 and 1998, 12 cases of relapsing fever associated with two exposure sites in northern California were investigated. Follow-up at exposure sites included collection of soft ticks and serum specimens from sylvatic rodents. Attempts to cultivate spirochetes were made through inoculation of patient blood into mice and by feeding Ornithodoros ticks on mice. Three isolates of B. hermsii were recovered from two blood specimens and one pool of ticks. The protein and plasmid profiles of the three isolates were comparable to those of previous B. hermsii isolates from the western United States. Western immunoblotting of patient sera demonstrated an expanding immunologic response to antigens within four distinct molecular weight regions by 3 to 4 weeks postonset. Antibody to B. hermsii was detected in sera from 4 of 11 yellow-pine chipmunks (Tamias amoenus); no other rodent species collected were seropositive. basis of the clinical presentation and a history of sleeping in a rural mountain dwelling within 2 weeks of onset. While the diagnosis may be confirmed by the observation of spirochetes in a patient’s blood smear, attempts to culture, isolate, and identify these spirochetes to the species level are rarely performed. In California, between 3 and 18 cases of RF were reported each year between 1985 and 2002 (California Department of Health Services, unpublished data). Few isolates of B. hermsii from California have been described, despite the frequent occurrence and wide distribution of human cases of B. hermsii infection reported in many areas of the state dating back to 1921 (3). The last reported isolation of B. hermsii from a California RF patient was in a resident of San Francisco who was infected while vacationing in Siskiyou County in 1964 (1). That isolate and two others recovered from California RF patients in the 1960s have been described previously (17, 18). Here we report on the results of investigations of recent RF outbreaks in California, describe the immunologic profiles of RF patients and rodents, and describe the isolation and characterization of the B. hermsii isolates from the RF patients and ticks associated with these outbreaks.
Tick-borne relapsing fever (RF) is a zoonotic disease endemic throughout much of Africa, central Asia, and the Americas (22). RF has been recognized as a cause of human morbidity in the western United States since the early 20th century (12). A recent epidemiologic review of RF in the United States indicated that 450 cases in patients with a broad age range (⬍1 to 81 years) and of both sexes (52% males) were reported in 11 western states between 1977 and 2000 (7). Individual cases of RF are most commonly reported, but large outbreaks also occur (2, 5, 24). RF is characterized by recurring episodes of fever and nonspecific symptoms (e.g., myalgia, headache, and gastrointestinal illness), separated by several days of apparent recovery (22). Although most patients recover promptly following initiation of antibiotic therapy, severe symptoms (e.g., Jarisch-Herxheimer reactions) and, rarely, death can occur. Three species of spirochete—Borrelia hermsii, Borrelia turicatae, and possibly, Borrelia parkeri—have been identified as agents of RF in the United States. The most important agent in the western United States is B. hermsii, which is transmitted through the bite of the argasid tick (Ornithodoros hermsi). B. hermsii is maintained in enzootic cycles that involve O. hermsi and small sylvatic rodents, principally chipmunks (Tamias spp.) and pine squirrels (Tamiascuirus spp.) above elevations of 1,000 m (6). Human infection typically occurs when infected ticks living in rodent nests within a human dwelling seek out an alternative host upon which to feed. Because O. hermsi is nocturnal and feeds quickly, in 20 to 90 min, RF patients rarely recall bites from these ticks. Most RF cases are diagnosed and treated empirically on the
MATERIALS AND METHODS Investigations of outbreaks. RF is a reportable disease in California. Between 1996 and 1998, the California Department of Health Services endeavored to investigate all outbreaks of RF occurring in California. For the purposes of this study, an RF fever outbreak was defined as two or more confirmed RF cases epidemiologically associated with a common place and a common time of exposure. A patient with a confirmed case of RF was defined as one who had a clinically compatible illness and for whom laboratory confirmation of RF could be obtained either by observation of spirochetes in a blood smear or by positivity by B. hermsii serology. A patient with suspected RF was defined as one who had a concurrent clinically compatible illness or a history of a clinically compatible illness that was epidemiologically linked to a confirmed case but that lacked laboratory confirmation.
* Corresponding author. Mailing address: Division of Communicable Disease Control, California Department of Health Services, MS 7307, P.O. Box 997413, Sacramento, CA 95899-7413. Phone: (916) 552-9730. Fax: (916) 552-9725. E-mail:
[email protected]. 1123
1124
FRITZ ET AL.
J. CLIN. MICROBIOL.
TABLE 1. Clinical and laboratory features of confirmed and suspect RF cases associated with two outbreak sites in Californiaa Patient no.
Mo/yr of onset
Age (yr)
Sex
Outbreak site
No. of febrile episodes
Blood smear result
1 2 3 4
8/96 8/96 4/97 8/97
42 7 41 38
M M F F
A A A A
3 2 1 4
ND ND Pos Neg
5
8/97
8
M
A
4
Neg
6 7 8 9 10 11 12
8/97 1995 1996 NI 8/98 8/98 8/98
53 11 39 7 48 2 4
M F M F M M M
A A A A B B B
4 2 1 0 3 1 1
Neg ND ND ND Pos Pos Neg
a
Serology result by:
No. of wk between onset and serum collection
EIA
Western bloting
3 3 2 4 9 4 9 4 ⬃100 ⬃52 NI ⬍1 ⬍1 ⬍1
Pos Equiv Pos Pos Pos Pos Pos Equiv Equiv Equiv Equiv Neg Neg Neg
Pos Pos Pos Pos Pos Pos Pos Pos Neg Neg Neg Neg Neg Neg
Abbreviations: M, male; F, female; NI, not ill; ND, not done; Pos, positive; Equiv, equivocal; Neg, negative.
Environmental evaluations were conducted at each exposure site. Implicated buildings were evaluated inside and outside for signs of rodent activity and ingress (e.g., fecal deposits and gnaw holes). Tomahawk (Tomahawk Live Trap, Tomahawk, Wis.) and Sherman (Sherman Trap Co., Tallahassee, Fla.) traps were set overnight to collect small rodents in and around buildings. Blood samples were collected by cardiac puncture, and the animals were then released. Carbon dioxide traps were set inside overnight to collect argasid ticks. Patient serology. All human sera were tested in the Diagnostic and Reference Laboratory (CLIA identification no. 06D0880233) of the Division of VectorBorne Infectious Diseases, Centers for Disease Control and Prevention, Fort Collins, Colo., by standardized diagnostic enzyme immunoassay (EIA) and Western blotting. Samples submitted for RF serology are routinely tested in parallel for antibodies against B. hermsii and Borrelia burgdorferi in order to document the specificity of the test. EIA was performed as described previously (24). Briefly, B. hermsii-coated wells (100 ng/well) were incubated with a 1:100 dilution of patient or control serum, and bound antibody was detected with goat antihuman immunoglobulin G (IgG) plus IgM conjugated with alkaline phosphatase. After substrate development, specimens with reactivities from ⬎1 to ⬍3 standard deviations above the mean for six negative controls were interpreted as equivocal, and specimens with reactivities ⬎3 standard deviations above the mean for the negative controls were interpreted as positive. Specimens with equivocal or positive EIA reactivities were further tested in separate IgM and IgG Western immunoblot assays. Western blots, as described below, were probed with human serum (1:100) for 30 min, and bound IgM or IgG was detected with alkaline phosphatase-labeled goat anti-human IgM or IgG conjugate and an appropriate substrate (Kirkegaard & Perry Laboratories). Western blots were considered positive if reactivity was observed in at least two of three regions (66, 41 to 39, and 21 to 18 kDa) for IgM and/or at least three of four regions (66, 41 to 39, 32 to 28, and 21 to 18 kDa) for IgG. Standardized positive and minimally reactive internal control sera were used to locate and score the diagnostic bands for each set of patient blots. Samples demonstrating equivocal or positive EIA results, followed by a positive IgM and/or IgG immunoblotting result and increased reactivities to B. hermsii compared to that to B. burgdorferi by both EIA and Western blotting, were interpreted as positive for exposure to RF. Due to the small numbers of serum samples from patients with confirmed RF available, the sensitivities of EIA and Western blotting have not been fully evaluated. Borrelia strains and cultivation. Adult mice were inoculated intraperitoneally with 0.25 ml of blood (in EDTA) from RF patients and examined daily for the presence of spirochetemia. For this, the tail of the mouse was nicked and a drop of blood was smeared on a microscope slide, air dried, stained with Giemsa, and examined by bright-field microscopy at ⫻970 with oil immersion. On the second or third day of spirochetemia, 0.25 ml of infected blood was collected by cardiac puncture and passaged into another mouse by intraperitoneal inoculation. Infected blood samples from the mice were collected by cardiac puncture, and 100 l was inoculated into a tube containing 9 ml of Barbour-Stoenner-Kelly-H (BSK-H) medium (Sigma-Aldrich Corp., St. Louis, Mo.) and incubated at 34°C. Spirochetes were harvested and examined after two to four passages. Spirochetes recovered from RF patients were compared to previously characterized Borrelia spp. B. hermsii DAH was isolated from the blood of an RF
patient in eastern Washington. B. parkeri and B. turicatae were isolated from Ornithodoros parkeri and Ornithodoros turicata, respectively, and are part of the bacterial reference collection of Rocky Mountain Laboratories. B. burgdorferi B31 was isolated from Ixodes scapularis ticks collected on Shelter Island, N.Y. (4). Escherichia coli pTA-1 harbors a recombinant plasmid containing the GlpQ gene of B. hermsii (20). This gene expresses an immunoreactive protein that is reactive with antibodies produced during RF infections but not Lyme disease (14, 20). Polyacrylamide gel electrophoresis. Whole-cell lysates of spirochetes were prepared as described previously (19). Sodium dodecyl sulfate-polyacrylamide gel electrophoresis with Laemmli buffer (10) and a vertical gel electrophoresis system (Bethesda Research Laboratories-GIBCO, Gaithersburg, Md.) were used to separate proteins according to the instructions of the manufacturer. Proteins were stained with Coomassie brilliant blue. Western blot analysis. Whole-cell lysates were electrophoresed in one-dimensional acrylamide gels and blotted onto nitrocellulose membranes by using Towbin buffer (23) and a Trans-Blot Cell (Bio-Rad Laboratories) according to the instructions of the manufacturer. The membranes were blocked overnight at room temperature with TSE-Tween (50 mM Tris [pH 7.4], 150 mM NaCl, 5 mM EDTA, 0.05% Tween 20) and were subsequently incubated either with monoclonal antibody H9826, which is specific for B. hermsii (18), or with serum samples (diluted 1:100) from patients or wild rodents. Bound antibodies were detected by 125I-labeled protein A autoradiography (19). DNA purification and analysis. Total DNA was purified from 500 ml of stationary-phase BSK-H cultures of spirochetes (21). DNA samples were examined by agarose gel electrophoresis with a Mini-Sub DNA Cell (Bio-Rad Laboratories). DNA samples were electrophoresed in 0.3% agarose gels with TBE buffer (90 mM Tris, 90 mM boric acid, 20 mM EDTA) to resolve the plasmids. Gels were run with ethidium bromide at 50 V for 5 min and then 12 V for 16 h, and the DNA was visualized by UV transillumination.
RESULTS Between 1996 and 1998, 30 cases of RF were reported through the statewide surveillance system to the California Department of Health Services; 5 additional unreported cases were identified during the present study. Of the 35 total cases, 16 were associated with seven outbreaks. We describe here the results of investigations of three of these outbreaks at two different exposure sites. Site A. In August 1996, patients 1 and 2 (Table 1) developed febrile illnesses approximately 1 week after spending 5 days at site A. Both patients experienced recurrent episodes of fever up to 105°F for 2 to 3 days, separated by 3 to 6 days of apparent recovery. Patient 3 developed chills and fever to 104°F in April
VOL. 42, 2004
1997, 2 days after spending 5 days at site A. Patient 3 recovered within 48 h following initiation of antibiotic therapy and experienced a single febrile episode. In August 1997, three members of the same extended family (patients 4, 5, and 6) experienced the onset of similar febrile illnesses 3 to 6 days after visiting site A. In addition to these six acute RF cases, exposure to B. hermsii was identified retrospectively among three relatives (patients 7, 8, and 9) of patient 3, all of whom had visited site A. Patients 7 and 8 both experienced febrile illnesses approximately 1 week after visiting site A in 1995 and 1996, respectively. Patient 9 visited site A at the same time as patients 3, 7, and 8 but recalled no subsequent illness. Site A is a family-owned ranch in Siskiyou County approximately 15 mi south of the California-Oregon border, at the southwestern edge of Butte Valley. Located at an elevation of 1,300 m, the habitat is principally yellow pine-oak forest with big sagebrush, bitterbrush, and rabbitbrush mix. Visual inspections, rodent surveys, and tick collection were conducted at site A on three occasions in 1997. The initial investigation in April 1997 noted evidence of extensive activity by golden-mantled ground squirrels (Spermophilus lateralis) and yellow-pine chipmunks (Tamias amoenus) throughout the area surrounding the two guest cabins. Stick piles and feces of wood rats (Neotoma fuscipes) and deer mice (Peromyscus maniculatus) were observed in buildings and sheds adjacent to the cabins. Mouse feces were noted inside both cabins. Investigation of the sleeping quarters in one cabin revealed five dead deer mice and four nests within the box-spring mattress of one bed. Blood spots were identified on bedding from which two O. hermsi ticks, one fully engorged, were collected. Two additional O. hermsi ticks were found in the bedding of a second cabin. A second investigation was performed at the site in August 1997. Golden-mantled ground squirrels had access to the kitchen in one cabin, and chipmunks moved freely in and out of the attic of the second cabin. Ground squirrels, chipmunks, and a wood rat were collected; and serum samples were obtained. A third investigation was performed at the site in October 1997. Sherman traps and dry ice stations failed to collect rodents or argasid ticks in the main cabin. Two wood rats were captured and released following collection of blood samples. Two wood rat nests were noted in the attic above the master bedroom. A dead chipmunk was recovered from one nest. The rodent nests were dismantled, and the nest materials were thoroughly inspected. A total of 25 O. hermsi ticks were recovered from the nests by using a Berlese funnel. Site B. In August 1998, a 48-year-old male (patient 10) was hospitalized with a febrile illness following a 5-day vacation at site B. A high concentration of spirochetes was observed in a Wright’s-stained smear of his blood. The patient experienced a severe Jarisch-Herxheimer reaction following initiation of treatment with ceftriaxone. Seven and 12 days after the onset of illness in patient 10, two male siblings (patients 11 and 12; ages 2 and 4 years, respectively) who had accompanied patient 10 to site B were hospitalized with similar febrile illnesses. High concentrations of spirochetes were observed in the blood smear of patient 11, but no spirochetes were observed in the blood smear of patient 12. Site B was at a condominium complex in Placer County,
BORRELIA HERMSII ISOLATES IN CALIFORNIA
1125
approximately 6 mi west of Lake Tahoe, Calif. The condominium was one of six attached units in a single building. All three patients slept at least one night in a single bedroom at the condominium. Site B was inspected in August 1998. Carbon dioxide traps were placed under the foundation. The condominium had an open-beamed roof that may have allowed rodents to enter without obvious signs of access. All the rooms within the condominium were thoroughly inspected for rodent harborage and argasid ticks; however, no evidence of rodents or ticks was found. Prior to the evaluation by public health officials, rodent control efforts, laundering, and housecleaning had been done and may have contributed to the lack of any positive findings at this site. Serology. The antibody responses against B. hermsii among the patients evolved and peaked during the first few weeks postinfection, as evidenced by EIA and Western immunoblotting (Table 1). Reactivity to immunodominant antigen regions with molecular masses of 66, 41 to 39, 32 to 28, and 21 to 18 kDa was observed in Western blots. GlpQ was shown to comigrate with flagellin at 40 kDa (Fig. 1), and patient antibody against this antigen was demonstrated in recombinant immunoblots (data not shown). Blood samples were collected from nine patients during the acute phase of infection, within 1 month of disease onset. Samples collected from patients 10 to 12 within the first week of disease onset were negative by EIA and displayed minimal banding in immunoblots (e.g., Fig. 1, lane 1). Samples collected from patients 1 to 6 between 1 and 4 weeks after onset demonstrated increased reactivities by EIA (equivocal or positive results) and increased numbers of bands reactive for IgM and IgG on immunoblots (e.g., Fig. 1, lanes 2, 3, and 4); during this collection interval, immunoblots of samples from all patients were positive. The specificities of the antibodies from RFseropositive patients were demonstrated by increased reactivities against B. hermsii compared to that against B. burgdorferi by EIA and Western blotting (data not shown). Acute- and convalescent-phase sera were collected from two patients (patients 4 and 5). Convalescent-phase samples, collected approximately 9 weeks postonset, demonstrated sustained IgM and IgG reactivities in Western blots (e.g., Fig. 1, lane 5). Antibody levels were greatly diminished in samples collected at later time points; samples collected a year or more after the onset of disease (patients 7 and 8; Fig. 1, lanes 7 and 6, respectively) were equivocal by EIA and displayed strong reactivities to a single band at 20 kDa and faint reactivities to several other bands in IgG immunoblots. Blood samples were obtained from 33 rodents at site A: 14 S. lateralis, 11 T. amoenus, 5 P. maniculatus, 2 N. fuscipes, and 1 Tamias senex rodent species. Western blot analysis showed that 4 (36%) of the 11 T. amoenus chipmunks had antibodies to B. hermsii antigens and to heterologous GlpQ expressed in E. coli (Fig. 2). No antibodies to B. hermsii were detected in sera from the other rodent species. Identification and characterization of B. hermsii isolates. Two isolates (isolates RAL and WAD) of B. hermsii were recovered from the blood of patients (patients 4 and 11, respectively), and one isolate (isolate SIS) originated from ticks. The two isolates from the human patients were established in pure culture only after laboratory mice (Mus musculus) were first inoculated with the spirochetemic blood, as all attempts to
1126
FRITZ ET AL.
J. CLIN. MICROBIOL.
FIG. 1. IgM and IgG Western blot reactivities of sera from patients with RF to B. hermsii. Samples are arranged in ascending time intervals from the time of onset to the time of sample collection. Lanes: 1, patient 11 (⬍1 week); 2, patient 1 (3 weeks); 3, patient 4 (4 weeks); 4, patient 6 (4 weeks); 5, patient 5 (9 weeks); 6, patient 8 (52 weeks); 7, patient 7 (100 weeks). Immunodominant regions are indicated in the left margin. The location of GlpQ is indicated in the right margin. MW, molecular mass marker (the numbers next to the blots are in kilodaltons).
isolate the organisms in culture medium directly from the patients’ blood failed. SIS was isolated from the blood of a mouse on which 22 O. hermsi ticks collected from the wood rat nest at site A had fed. The protein profiles of whole-cell lysates were typical for B. hermsii, but the profiles for isolates from patients infected in eastern Washington were different (Fig. 3A). Differences were present in the mid-20-kb range, in which the variable small proteins migrate. The isolates were identified as B. hermsii by reactivity with monoclonal antibody H9826, an antiflagellin antibody specific to this species (Fig. 3B). The plasmid profiles of the new isolates were also consistent with those for plasmids from other isolates of the same species originating from Washington (Fig. 4). Some variability in the sizes of the linear plasmids in the 22- to 26-kb range was evident. DISCUSSION This report describes the first isolation in more than 30 years of B. hermsii spirochetes from RF outbreaks in California. Until recently, isolation of Borrelia was hampered by lack of an effective cultivation protocol. In 1971, Kelly (9) described a liquid medium that supported the continuous cultivation of B. hermsii in vitro. In 1982, Burgdorfer and coworkers (4) reported on a Borrelia spirochete that had been detected in ticks, isolated in Kelly’s medium, and bound to antibodies from Lyme disease patients. Although interest in research on this new spirochetal disease spread rapidly throughout the biomedical community, the medium remains difficult to make and maintain. Hence, culture of blood samples from suspect bor-
FIG. 2. Immunoblots showing the reactivities of antibodies in the sera of four chipmunks to whole-cell lysates of B. hermsii DAH and recombinant GlpQ in E. coli TA-1 (A to D) and no reactivity in a seronegative (Neg) sample (E). Lysates of the E. coli vector only lack B. hermsii GlpQ. The arrow on the left shows the position of GlpQ, and molecular mass standards (MMS; in kilodaltons) are shown on the right.
VOL. 42, 2004
FIG. 3. (A) Protein profiles of whole-cell lysates of B. hermsii isolates from California (isolates WAD, RAL, and SIS) compared to those of isolates from eastern Washington (isolates DAH and FRO) and other Borrelia species. (B) Immunoblot analysis of whole-cell lysates with monoclonal antibody H9826, which is specific for B. hermsii. The arrow on the left shows the position of flagellin to which the monoclonal antibody reacts. Molecular mass standards (MMS; in kilodaltons) are shown on the left.
reliosis patients is limited to a few specialized clinical or research laboratories. In our experience, the resultant shipping delays reduce the viability of the spirochetes in the samples to the extent that inoculation directly into liquid medium fails. However, inoculation of blood samples into mice can amplify existing spirochete numbers and improve the success of subsequent in vitro cultivation. The paucity of isolates of RF spirochetes is exacerbated by the difficulty of finding infected O. hermsi ticks in nature. Unlike the ixodid (hard) tick vectors of Lyme disease spirochetes, which can be collected in large numbers from vegetation, the argasid (soft) ticks which transmit B. hermsii confine themselves to rodent burrows, nests, and hard-to-find refugia except during brief nocturnal feedings. The collection of ticks during investigations of RF cases is rarely successful, even when a specific exposure location (e.g., the master bedroom) is known. Traps baited with CO2 provide a theoretical means of enticing feeding behavior in soft ticks (13), but in practice these efforts are rarely rewarding. Ticks have occasionally been found in unlaundered bed linens which have been left in situ subsequent to the RF patients’ exposure. Also, as in site A in the present study, ticks may be recovered from their usual harborage in a rodent nest when the nest can be located. Whole blood from RF patients is the most likely source of new isolates of B. hermsii. However, if spirochete densities are less than 105 cells per ml, the bacteria can elude microscopic examination, the diagnosis may be missed, and culture will not be attempted. Spirochetemia is maximal during the febrile periods of RF; few to no spirochetes are observed in blood specimens collected during afebrile intervals. Smears of blood collected during inappropriate periods in the illness cycle offer little diagnostic benefit and, because they yield false-negative
BORRELIA HERMSII ISOLATES IN CALIFORNIA
1127
results, can delay accurate diagnosis and antimicrobial treatment. Serology is an ancillary diagnostic option for patients with RF-compatible illness who have recovered, who are epidemiologically linked to a confirmed case or implicated exposure site, or for whom the only available specimen is blood collected during an afebrile period. The sensitivities of serologic assays for RF are dependent upon the kinetics of the immunologic response following infection. In the present study, antibody was not detected in patients 10, 11, and 12 during the first week of disease. (Although RF in patient 12 was not laboratory confirmed, his symptomatology, rapid response to antibiotic treatment, and strong epidemiologic link to patients 10 and 11 strongly suggested that he experienced a contemporaneous B. hermsii infection.) Sera collected from patients 1 to 6 at 2 to 4 weeks postonset were equivocal or positive by EIA and positive by Western immunoblotting. At least eight immunoreactive antigens were commonly observed in IgG immunoblots from samples from patients 4, 5, and 6 collected between 4 and 9 weeks postonset. Sera collected from patients 7 and 8 a year or more after disease displayed equivocal reactivities by EIA and negative reactivities by immunoblotting. While Western immunoblotting for RF is available only through selected research laboratories at present, the results of the present study indicate that it has promise for confirmation of the diagnosis of RF in patients with early disease (⬎1 week postonset) and treated and/or resolved disease (ⱕ9 weeks postonset). Antibody cross-reactivity against numerous proteins of B. hermsii and B. burgdorferi occurs (11, 15, 20) and may lead to false-positive laboratory results when the clinical and epidemiologic pretest assessment is inadequate. In the present study the pretest likelihood of RF among ill patients was high and
FIG. 4. Plasmid profiles of B. hermsii isolates from California (isolates WAD, RAL, and SIS) compared with those of two isolates from eastern Washington (isolates DAH and FRO) and two other species of RF spirochetes. Plasmids were resolved in a 0.3% agarose gel stained with ethidium bromide and visualized by UV transillumination. Molecular size standards (in kilobases) are shown on the left.
1128
FRITZ ET AL.
immunospecificity was documented by differential antibody reactivities against B. hermsii. In addition, reactivity to GlpQ has been shown to be a discriminator of these related spirochetal infections (14, 20); to date, GlpQ and its enzymatic activity have been found only in RF Borrelia (16). Use of this diagnostic approach provided further evidence of RF exposure in rodents collected in the field; 4 of 11 (36%) serum samples collected from chipmunks (T. amoenus) from site A were seropositive for lysates of B. hermsii and the heterologously produced GlpQ. Given the ephemeral presence of detectable spirochetes in the blood of infected animals, as in humans, examination of rodent sera for anti-GlpQ antibodies should offer a more specific surveillance tool for RF Borrelia ecology. Removal and exclusion of rodents are the principal means of preventing additional RF cases at a building implicated as a site of RF transmission. At site A, T. amoenus chipmunks were frequently observed to enter and exit the attic of one cabin through holes at the roofline and to nest within the walls and attic. The same attic space contained at least two wood rat nests, suggesting that O. hermsi populations were sustained in the cabin by two active rodent species with substantially different ecologies and behaviors. Resident rodents can be eliminated through the placement of traps throughout the building, particularly near areas containing food sources. The exterior of the building should be examined thoroughly; and holes, cracks, and other areas of ingress should be identified and sealed. Because O. hermsi ticks live in or near rodent nests, elimination of rodents without removal of the nests may actually increase the risk of transmission to humans, as ticks seek out an alternative host upon which to feed. However, rodent nests are often in inaccessible spaces in the building and cannot be readily located. In the present study, the wood rat nest at site A was not discovered until the third site investigation and after at least seven cases of RF had been acquired. RF remains a cause of morbidity for visitors to the mountains of the western United States. Because of the often selflimited clinical course and rapid response to empirical antibiotics, many cases of RF may go undiagnosed and unreported. However, RF can be severe, with neurologic, respiratory, and cardiovascular complications and death occasionally reported (8, 22). Therefore, health care providers should be familiar with the characteristic clinical presentation and pattern of RF and solicit travel histories from patients who present with compatible symptoms. Identification and reporting of RF cases to appropriate public health authorities are critical for appropriate follow-up, as buildings currently or previously infested with rodents can serve as refuges for ticks and a continued source of infection. Recent developments in immunodiagnostics, including Western immunoblotting and GlpQ serology, may provide additional means by which to further define the epidemiologic scope and ecologic distribution of RF. ACKNOWLEDGMENTS We thank Merry Schrumpf for technical assistance and Gary Hettrick for graphic arts.
J. CLIN. MICROBIOL. REFERENCES 1. Bissett, M. L., and W. Hill. 1987. Characterization of Borrelia burgdorferi strains isolated from Ixodes pacificus ticks in California. J. Clin. Microbiol. 25:2296–2301. 2. Boyer, K. M., R. S. Munford, G. O. Maupin, C. P. Pattison, M. D. Fox, A. M. Barnes, W. L. Jones, and J. E. Maynard. 1977. Tick-borne relapsing fever: an interstate outbreak originating at Grand Canyon National Park. Am. J. Epidemiol. 105:469–479. 3. Briggs, L. H. 1922. Relapsing fever in California. JAMA 79:941–944. 4. Burgdorfer, W., A. G. Barbour, S. F. Hayes, J. L. Benach, E. Grunwaldt, and J. P. Davis. 1982. Lyme disease—a tick-borne spirochetosis? Science 216: 1317–1319. 5. Centers for Disease Control and Prevention. 2003. Tickborne relapsing fever outbreak after a family gathering—New Mexico, August 2002. Morb. Mortal. Wkly. Rep. 52:809–812. 6. Dworkin, M. S., D. E. Anderson, Jr., T. G. Schwan, P. C. Shoemaker, S. N. Banerjee, B. O. Kassen, and W. Burgdorfer. 1998. Tick-borne relapsing fever in the northwestern United States and southwestern Canada. Clin. Infect. Dis. 26:122–131. 7. Dworkin, M. S., P. C. Shoemaker, C. L. Fritz, M. E. Dowell, and D. E. Anderson, Jr. 2002. The epidemiology of tick-borne relapsing fever in the United States. Am. J. Trop. Med. Hyg. 66:753–758. 8. Fuchs, P. C., and A. A. Oyama. 1969. Neonatal relapsing fever due to transplacental transmission of Borrelia. JAMA 208:690–692. 9. Kelly, R. 1971. Cultivation of Borrelia hermsi. Science 173:443–444. 10. Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680–685. 11. Magnarelli, L. A., J. F. Anderson, and R. C. Johnson. Cross-reactivity in serological tests for Lyme disease and other spirochetal infections. J. Infect. Dis. 156:183–187. 12. Meador, C. N. 1915. Five cases of relapsing fever originating in Colorado, with positive blood findings in two. Colo. Med. 12:365–368. 13. Miles, V. I. 1968. A carbon dioxide bait trap for collecting ticks and fleas from animal burrows. J. Med. Entomol. 5:491–495. 14. Porcella, S. F., S. J. Raffel, M. E. Schrumpf, M. E. Schriefer, D. T. Dennis, and T. G. Schwan. 2000. Serodiagnosis of louse-borne relapsing fever with glycerophosphodiester phosphodiesterase (GlpQ) from Borrelia recurrentis. J. Clin. Microbiol. 38:3561–3571. 15. Rath, P. M., G. Rogler, A. Schonberg, H. D. Pohle, and F. J. Fehrenbach. 1992. Relapsing fever and its serological discrimination from Lyme borreliosis. Infection 20:283–286. 16. Schwan, T. G., J. M. Battisti, S. F. Porcella, S. J. Raffel, M. E. Schrumpf, E. R. Fischer, J. A. Carroll, P. E. Stewart, P. Rosa, and G. A. Somerville. 2003. Glycerol-3-phosphate acquisition in spirochetes: distribution and biological activity of glycerophosphodiester phosphodiesterase (GlpQ) among Borrelia species. J. Bacteriol. 185:1346–1356. 17. Schwan, T. G., K. L. Gage, and B. J. Hinnebusch. 1995. Analysis of relapsing fever spirochetes from the western United States. J. Spirochetal Tick-Borne Dis. 2:3–8. 18. Schwan, T. G., K. L. Gage, R. H. Karstens, M. E. Schrumpf, S. F. Hayes, and A. G. Barbour. 1992. Identification of the tick-borne relapsing fever spirochete Borrelia hermsii by using a species-specific monoclonal antibody. J. Clin. Microbiol. 30:790–795. 19. Schwan, T. G., K. K. Kime, M. E. Schrumpf, J. E. Coe, and W. J. Simpson. 1989. Antibody response in white-footed mice (Peromyscus leucopus) experimentally infected with the Lyme disease spirochete (Borrelia burgdorferi). Infect. Immun. 57:3445–3451. 20. Schwan, T. G., M. E. Schrumpf, B. J. Hinnebusch, D. E. Anderson, Jr., and M. E. Konkel. 1996. GlpQ: an antigen for serological discrimination between relapsing fever and Lyme borreliosis. J. Clin. Microbiol. 34:2483–2492. 21. Simpson, W. J., C. F. Garon, and T. G. Schwan. 1990. Analysis of supercoiled circular plasmids in infectious and non-infectious Borrelia burgdorferi. Microb. Pathog. 8:109–118. 22. Southern, P. M., and J. P. Sanford. 1969. Relapsing fever: a clinical and microbiological review. Medicine 48:129–149. 23. Towbin, H., T. Staehelin, and J. Gordon. 1979. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc. Natl. Acad. Sci. USA 76:4350–4354. 24. Trevejo, R. T., M. E. Schriefer, K. L. Gage, T. J. Safranek, K. A. Orloski, W. J. Pape, J. A. Montenieri, and G. L. Campbell. 1998. An interstate outbreak of tick-borne relapsing fever among vacationers at a Rocky Mountain cabin. Am. J. Trop. Med. Hyg. 58:743–747.
