Purification and Characterization of Borrelia burgdorferi from Feeding Nymphal Ticks (Ixodes scapularis)

Abstract

Here we describe a protocol for purifying Borrelia burgdorferi from feeding ticks by velocity centrifugation and Percoll density gradient centrifugation. The purified spirochetes were motile and 10- to 20-fold purer than the bacteria in crude tick homogenates. The purified bacteria were present in sufficient quantity for protein and gene expression studies. In comparison to culture-grown bacteria, tick-borne spirochetes had several proteins that were upregulated and a few that were downregulated. When the levels of B. burgdorferi outer surface proteins OspA and OspC were measured, OspC protein and mRNA levels were lower in cultured bacteria than in bacteria purified from ticks. Although differences in OspA mRNA levels were observed between cultured and tick-borne bacteria, no differences were observed at the protein level. These experiments demonstrate that tick-transmitted borreliae display a gene expression and antigen profile different from that of spirochetes cultured in vitro.

Borrelia burgdorferi, a tick-borne spirochete, is the causative agent of Lyme disease, the most prevalent arthropod-borne disease in the United States (21). The bacteria, which cycle in nature between small mammals and Ixodes ticks, must not only survive in the mammalian host but also withstand the complicated changes that occur during tick feeding and digestion and the physiological changes associated with molting and quiescence (4). There is mounting evidence that borreliae alter the expression of surface molecules by transcriptional control as well as DNA recombination and rearrangement (18, 33). Many genes that appear to be selectively produced in the mammalian host have been identified, and they undoubtedly play a role in evasion of the host's immune system, dissemination in the host, and possibly invasion of particular organ systems (1, 7, 10–12). We are interested in studying the strategies used by B. burgdorferi to survive within ticks and to move from tick to host during feeding.

In unfed infected nymphal ticks, the spirochetes are present in the lumen of the gut. During tick feeding, spirochetes in the gut multiply and pass through the hemocoel to the salivary glands and enter the host through the salivary ducts (26). During transmission, the spirochetes have been shown to differentially produce two major outer surface proteins, designated OspA and OspC. Spirochetes in unfed ticks produce primarily OspA and no or very little OspC. During the blood meal, large numbers of the multiplying spirochetes induce expression of OspC (28, 29). OspA, which appears to be a receptor that mediates attachment to the tick gut (22), is cleared from the surface of some bacteria during the blood meal, while others continue to produce the protein (9, 20, 28). OspA and OspC are among over 100 lipoproteins encoded by the spirochete's genome (13). The totality of genes that are differentially expressed no doubt extends beyond ospA and ospC.

Many studies of B. burgdorferi pathogenesis have been performed with spirochetes grown in culture and by altering culture conditions to mimic conditions in vivo (17, 19, 25). More recently, cultured organisms were sealed in chambers, which were then implanted in host tissue (1, 8). These studies have been useful but cannot substitute for studies with spirochetes directly isolated from infected ticks and hosts. Here we describe the use of velocity and isopycnic centrifugation to successfully purify B. burgdorferi from ticks in the process of transmitting the infection to mice. Experiments were also done to study differences in gene expression and antigenic composition between spirochetes purified from ticks and those grown in culture.

MATERIALS AND METHODS

B. burgdorferi culture conditions.

Clonal populations of low-passage B. burgdorferi strains B31 and Westchester were grown in BSK II medium at 33°C to mid-log-phase density (1 × 10⁷ to 3 × 10⁷ cells/ml) and used in this study.

Animals.

Female C3H mice, 4 to 6 weeks old and free of B. burgdorferi infection (National Institutes of Health, Bethesda, Md.), were used for tick feeding experiments. The animals were caged individually and provided with antibiotic-free food and water ad libitum.

Ticks.

Ixodes scapularis nymphal ticks infected with the Westchester strain of B. burgdorferi (kindly provided by Durland Fish, Yale University, New Haven, Conn.) were used for most experiments. Nymphal ticks infected with the B31 strain (raised in our laboratory) and the N40 strain (kindly provided by John F. Anderson, Connecticut Agricultural Experiment Station, New Haven) were used in a few experiments to determine if the number of spirochetes recovered from feeding ticks was dependent on the strain. The B31 strain spirochetes were introduced into ticks as previously described (23).

Purification of B. burgdorferi from ticks.

The entire procedure (Fig. 1) was performed at 4°C. Westchester strain-infected nymphal ticks that had fed for 48 h on individually caged C3H mice were collected with fine forceps. A small incision was made in the exoskeleton of the fed nymphal ticks to expose the internal organs. Groups of 20 ticks were homogenized in 400 μl of phosphate-buffered saline (PBS) in a Dounce tissue grinder with 2-ml working capacity. The ticks were homogenized using pestle A (clearance, 0.12 mm) with 20 strokes. Following this initial homogenization, the internal organs were released from the tick and partially homogenized while the hard exoskeleton remained mostly intact. The homogenate was then transferred to a new tissue grinder using a Pasteur pipette, leaving the exoskeleton in the first tissue grinder. The partial homogenate was again ground 15 times with the low-clearance pestle B (0.06 mm). Following the second homogenization, the tick tissues were fragmented into small particles while the spirochetes remained mostly intact in solution.

FIG. 1.

Diagrammatic representation of the purification methodology used to purify spirochetes from feeding ticks. DF, dark-field microscopy.

The homogenate was subjected to a low-speed spin at 84 × g for 1 min (Eppendorf 5415c centrifuge; Brinkmann Instruments Inc., Westbury, N.Y.), which resulted in the sedimentation of large tick tissues while the spirochetes remained in the supernatant (S1). The S1 from the low-speed centrifugation was centrifuged at 2,040 × g for 10 min, which sedimented the spirochetes with most of the soluble protein remaining in the supernatant (S2). The final pellet (P2) with the bacteria was resuspended in 1 ml of PBS.

Hanff et al. (16) used Percoll gradients to purify Treponema pallidum, the syphilis spirochete, from infected rabbit tissues. We used Percoll gradients to further purify spirochetes in the P2 fraction obtained from velocity centrifugation. The resuspended bacterial pellet was mixed with 5 ml of 60% Percoll solution (Amersham Pharmacia Biotech USA, Piscataway, N.J.). The Percoll solution with the bacteria was centrifuged in a 16- by 76-mm polycarbonate screw-capped tube (Nalgene Company, Rochester, N.Y.) at 30,000 × g for 30 min at 4°C (type 65 rotor, model L5.50B ultracentrifuge; Beckman Coulter, Inc., Fullerton, Calif.). Twenty 300-μl fractions were collected using a fraction collector. B. burgdorferi in each fraction was enumerated by dark-field microscopy using a Petroff-Hausser counting chamber. Protein concentration was measured by the Bradford assay, with modifications for the effect of Percoll on the assay as described previously (31). Nucleic acid concentration in each fraction was estimated by reading the absorbance at 260 nm in a spectrophotometer. A refractometer was used to determine the refractive index of each fraction, which correlates with the density of the gradient fractions. From the fractions containing spirochetes, Percoll was removed by diluting the 300-μl fractions with 1,700 μl of PBS and pelleting the bacteria by centrifugation at 12,000 × g for 10 min. The supernatant containing Percoll was removed, and the bacteria in the pellet were stored at −80°C until use.

Immunofluorescence assay (IFA).

Westchester strain spirochetes purified from infected ticks on gradients and from culture were pelleted and washed twice with PBS, and 10-μl fractions were spotted on silylated slides (PGC Scientifics, Frederick, Md.) to estimate the proportions of OspA- and OspC-producing spirochetes. The spots were air dried, acetone fixed, blocked with 5% fetal calf serum in PBS, and incubated with rabbit OspA antibodies (kindly provided by Erol Fikrig, Yale University School of Medicine, New Haven, Conn.) and rabbit OspC antibodies (kindly provided by Tom Schwan, Rocky Mountain Laboratories, Hamilton, Mont.) for 1 h at room temperature. Following incubation, the slides were washed thrice with PBS and incubated for 30 min with the secondary anti-rabbit antibodies conjugated with Texas red. Goat anti-Borrelia fluorescein isothiocyanate-conjugated antibodies (KPL Inc., Gaithersburg, Md.) were used along with the secondary antibodies to estimate the total number of spirochetes. The slides were then washed thrice with PBS, air dried, and mounted using Aqua-Polymount (Polysciences Inc., Warrington, Pa.).

Preparation of immune and infected mouse sera.

Serum was collected from mice infected or hyperimmunized with the Westchester strain of B. burgdorferi. To prepare infected mouse serum, two C3H mice were subcutaneously injected with 10⁴ live spirochetes. Infection of the mice was confirmed by Western blotting 3 weeks after injection of the spirochetes. Blood was collected at the end of the fourth week following euthanization of the mice; sera were separated and stored at −20°C until use.

To prepare hyperimmune serum, two C3H mice were subcutaneously injected with 10⁷ heat-killed (60°C for 1 h) spirochetes in complete Freund's adjuvant. Boosters in incomplete Freund's adjuvant at the same dose were given at the end of the second and fourth weeks after the first dose. Blood was collected 2 weeks after the second booster; sera were separated and stored at −20°C until use.

SDS-PAGE and Western blotting.

A protein detector Western blot kit (KPL) was used to process the Western blots. B. burgdorferi proteins (5 × 10⁶ spirochetes/lane) were resolved on 12% minigels by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to nitrocellulose membranes. The membranes were treated with the blocking buffer provided in the kit and probed for 1 h with dilutions of mouse serum in blocking buffer. Westchester strain-hyperimmune and -infected mouse sera were used at 1:50 dilution; monoclonal antibodies to OspA, OspC, and flagellin at 1:50 dilution and to p66 (kindly provided by Alan G. Barbour, University of California, Irvine) at 1:10 dilution were also used. The secondary antibody was horseradish peroxidase-conjugated anti-mouse immunoglobulin at 1:1,000 dilution. Antibody reactivity was detected by the chemiluminescence method using LumiGlo substrate provided in the kit.

RNA extraction and purification; reverse transcription (RT)-PCR.

Total RNA was extracted from Westchester strain B. burgdorferi purified from ticks and from culture by using a Totally RNA kit (Ambion Inc., Austin, Tex.) according to the manufacturer's instructions. In brief, the spirochetes were suspended in 300 μl of guanidine thiocyanate denaturation buffer and subjected to two quick-freeze thaw cycles in ethanol and dry ice. The thawed suspension was transferred to a Dounce tissue grinder and homogenized 10 times each with large-clearance pestle A and low-clearance pestle B. The RNA was extracted from the homogenate once with phenol-chloroform and once with sodium acetate and acid phenol-chloroform. Following extraction, the RNA was precipitated with isopropanol, washed with 70% ethanol to remove salts, and resuspended in diethyl pyrocarbonate-2 treated distilled water.

The extracted RNA was treated with RNase-free DNase (Promega Corporation, Madison, Wis.) for 30 min at 37°C to remove any contaminating DNA. Aliquots of the treated RNA were reverse transcribed to obtain cDNA, using random primers and a Stratagene (La Jolla, Calif.) Prostar first-strand RT-PCR kit. A control reaction containing no reverse transcriptase was also performed for each sample to check for possible DNA contamination. Specific upstream and downstream primers were used to amplify flaB (5′-CGGCACATATTCAGATGCAGACAG-3′ and 5′-CCTGTTGAACACCCTCTTGAAGAACC-3′), ospA (5′-GGTCAAACCACACTTGAAGTT-3′ and 5′-GTCAGTGTCATTAAGTTCAAC-3′), ospC (5′-ATGAAAAAGAATACATTAAGTGC-3′ and 5′-TTAAGGTTTTTTTGGACTTTCTGC-3′), and bbk32 (5′-TGGTGAATTGGAGGAGCCTA-3′ and 5′-AAACGCCATTCTTGTCAATG-3′) with 5 μl of the synthesized cDNA as template. The PCR amplification program consisted of 35 cycles of denaturation at 94°C for 30 s, annealing at different temperatures (50°C for flaB, ospC, and bbk32; 53°C for ospA) for 1 min, and extension at 72°C for 1 min. Products were resolved on 1% agarose gels.

RESULTS

Purification of B. burgdorferi from partially engorged ticks.

When partially fed ticks were homogenized and subjected to velocity centrifugation, 92% of the spirochetes from the starting tick homogenate were recovered in the P2 fraction and were 2.8-fold pure with respect to the protein and nucleic acids in the starting material (Table 1). In different experiments, the purification of spirochetes by velocity centrifugation ranged from 2.2- to 3.3-fold for proteins and 1.3- to 2.8-fold for nucleic acids (data not shown).

TABLE 1.

Purification of B. burgdorferi from feeding ticks by velocity and density gradient centrifugation^a

Fractionb	No. of spirochetesc (104)	Yield (%)	Protein (μg)	Nucleic acids (μg)
Homogenate	119	100.0	339.10 (1.0d)	189.07 (1.0)
S1	119	100.0	308.43 (1.1)	96.22 (1.9)
P1	3.5	2.9	39.40 (0.25)	15.8 (0.35)
S2	0	0	159.58 (0)	75.95 (0)
P2	110	92.0	112.14 (2.8)	61.5 (2.8)
Percoll gradient
7–10	78	65.5	17.63 (12.6)	13.5 (9.18)
11–13	16.5	13.8	65.46 (0.72)	15.3 (1.71)

^aRepresentative data for protein and nucleic acid contents and distribution of B. burgdorferi in a single experiment with a homogenate prepared from 20 infected, feeding ticks. 

^bObtained as outlined in Fig. 1 and Materials and Methods. 

^cDetermined by dark-field microscopy. 

^dFold purification with respect to the starting homogenate. 

Spirochetes were centrifuged with Percoll solutions that ranged in concentration from 43 to 65%. Optimal separation was obtained with 60% Percoll (data not shown). In multiple experiments, when the P2 fraction was centrifuged at 30,000 × g for 30 min in 60% Percoll, B. burgdorferi sedimented in fractions 7 to 13 (Fig. 2). The spirochetes were separated away from the protein peak by two fractions and from the nucleic acid peak by five fractions (Fig. 2). The results from a representative experiment are presented in Table 1 and Fig. 2. The spirochetes in fractions 7 to 10 were highly pure (12.6- and 9.18-fold pure for proteins and nucleic acids, respectively), whereas the trailing edge of bacteria in fractions 11 to 13 was less pure (0.72- and 1.71-fold pure for proteins and nucleic acids, respectively). Sixty-six percent of the bacteria in the starting material were present in the highly pure fractions, and 14% were present in the less pure fraction. More than 95% of the spirochetes collected in the fractions were observed to be motile by dark-field microscopy.

FIG. 2.

Distribution of protein, nucleic acids, and B. burgdorferi on 60% Percoll density gradients. One milliliter of the tick homogenate was mixed with 60% Percoll and centrifuged at 30,000 × g for 30 min. Twenty 300-μl fractions were collected from the bottom of the gradient, protein and nucleic acid contents were assayed, and borreliae were enumerated by dark-field microscopy.

Antigenic profile of B. burgdorferi purified from feeding ticks.

Experiments were performed with bacteria purified ticks and cultured bacteria to directly compare antigen profiles. Protein extracts from tick-derived (Westchester strain) and cultured (Westchester and B31 strains) B. burgdorferi were analyzed by Western blotting (Fig. 3). The blots were probed with (i) hyperimmune mouse serum raised by immunizing mice with 10⁷ heat-killed bacteria, (ii) sera from mice that were infected with a low dose of B. burgdorferi (10⁴ bacteria) (infected mouse serum), and (iii) a mix of monoclonal antibodies directed against B. burgdorferi proteins OspA, FlaB, OspC, and p66. The hyperimmune serum should recognize antigenic molecules expressed by spirochetes grown in culture, while the infected mouse serum will recognize antigens that are expressed in the mouse during an active infection.

FIG. 3.

Western blot analysis of spirochetes purified from ticks and grown in culture. Blots were probed with hyperimmune mouse serum (a), with infected mouse serum (b), and with monoclonal antibodies to OspA, OspC, FlaB, and p66 (c). Lanes: 1, Percoll-purified Westchester strain spirochetes; 2, cultured Westchester strain spirochetes; 3, cultured B31 strain spirochetes. Large and small arrows represent protein bands that were upregulated and downregulated, respectively, by spirochetes in ticks during feeding in comparison to culture-grown spirochetes.

When hyperimmune serum was used to probe the blots, bands of approximately 97, 65, 61, 37, 27, and 22 kDa were upregulated in tick-derived bacteria in comparison to culture-grown organisms (Fig. 3a). We are uncertain about the identities of the 97-, 65-, 61-, 37-, and 27-kDa proteins, but the 22-kDa protein and the band above the 29-kDa molecular weight marker may be OspC and OspA, respectively, since their size ranges are approximately those of OspC and OspA. A single band of 75 kDa was downregulated in ticks in comparison to cultured bacteria. With the hyperimmune serum, no marked differences were observed between cultured Westchester and B31 strains.

Western blots were also probed with sera from mice infected with B. burgdorferi (Fig. 3b). Four proteins (65, 47, 25, and 18 kDa) were downregulated and three proteins (97, 68, and 17 kDa) were upregulated in tick-derived bacteria in comparison to the cultured Westchester strain. The protein band of 97 kDa, which was upregulated in tick-derived bacteria, appears to be the same as the similarly sized protein which was also upregulated when probed with hyperimmune serum (Fig. 3a and b).

Blots containing proteins from tick-derived and cultured bacteria were also probed with a mix of monoclonal antibodies to FlaB, OspA, OspC, and p66 (Fig. 3c). The p66 protein was produced by the spirochetes in culture and in feeding ticks. Similarly, OspA was produced by both cultured and tick-transmitted bacteria. The OspC monoclonal antibody used here was raised against the B31 strain, and the antibody appears not to recognize the OspC protein from the Westchester strain (Fig. 3c and unpublished data).

OspA and OspC production by individual spirochetes within feeding ticks and in culture.

Indirect immunofluorescence studies were carried out to analyze the production of OspA and OspC by individual Westchester strain spirochetes from ticks and culture. Similar proportions of spirochetes from ticks and from culture (67% versus 63%) produced OspA. In contrast, 10% of the bacteria purified from ticks stained with a polyclonal OspC antibody, whereas only 0.4% of the bacteria grown in culture stained with this antibody.

Expression of selected B. burgdorferi genes in tick-borne and cultured spirochetes.

Having compared the differential production of proteins by cultured and tick-derived bacteria, we performed RT-PCRs using primers to ospA, ospC, flaB, and bbk32 to see if the regulation of these genes was evident at the transcriptional level. Total RNA was prepared from tick-borne and cultured spirochetes, and cDNA was synthesized. In RT-PCRs, the amount of cDNA template used was derived from similar numbers (2 × 10⁶) of cultured and tick-borne bacteria. The RT-PCR signals for flaB, a gene constitutively expressed by B. burgdorferi, were similar in the tick-derived and cultured bacterial samples, confirming that the cDNA templates used in all reactions were derived from similar numbers of bacteria (Fig. 4). As expected, ospA mRNA was detected in both populations of bacteria. However, when 20-fold less cDNA was used, signal for ospA was detected in cultured but not tick-derived bacteria. B. burgdorferi increases the transcription of ospC during transmission, and ospC mRNA was detected in tick-derived bacteria even when 20-fold less cDNA was used (Fig. 4), indicating an increase in the ospC mRNA levels during tick feeding in comparison to cultured bacteria. Bbk32, a 47-kDa fibronectin binding protein of B. burgdorferi, is selectively expressed in vertebrate hosts (10). bbk32 expression was observed only in feeding ticks, indicating that expression of bbk32 in the mammal, begins early during transmission from the tick to the host (Fig. 4).

FIG. 4.

Semiquantitative RT-PCRs for comparison of B. burgdorferi flaB, ospA, ospC, and bbk32 expression levels in ticks during feeding and in culture. RT-PCR was performed with total RNA from spirochetes purified from ticks (T) or grown in BSK II medium (C). For each of the primers used, PCR was performed with genomic DNA (D) as a positive control. RT reactions were performed with (+) or without (−) reverse transcriptase to rule out DNA contamination of the RNA samples. The amount of cDNA used was derived from 10⁴ or 2 × 10⁵ spirochetes.

Spirochete strain influences the yield of bacteria from ticks.

In addition to the Westchester strain, experiments were performed to purify bacteria from the more commonly used N40 and B31 strains of B. burgdorferi. The experimental protocol followed here was the same as that described for Westchester strain-infected nymphal ticks. When homogenates were prepared from groups of 20 ticks each infected with the Westchester, N40, or B31 strain of B. burgdorferi, the B31 and N40 strains consistently yielded fewer spirochetes than the Westchester strain. The yields of spirochetes from 20 infected ticks were estimated to be 1 × 10⁶ to 2 × 10⁶ for the Westchester strain, 8 × 10⁴ to 2 × 10⁵ for N40, and 5 × 10⁴ to 7 × 10⁴ for B31.

DISCUSSION

Previous investigators have used Percoll density gradients to purify syphilis spirochetes from infected rabbit testes (16). In the present study, we developed a method using velocity and Percoll density gradient centrifugation methods to purify Lyme disease spirochetes from groups of feeding ticks. The whole experimental procedure took approximately 2 h from the start of homogenization to collection of the pure spirochete fractions. Approximately 1 × 10⁶ to 2 × 10⁶ spirochetes were recoverable on the gradients from a homogenate of 20 partially fed ticks infected with the Westchester strain of B. burgdorferi sensu stricto. The purified bacteria were motile and present in sufficient quantity to detect mRNA and proteins produced during transmission.

The yield of spirochetes was dependent on the strain of B. burgdorferi sensu stricto used in the experiment. Ticks infected with B31 and N40 strains of B. burgdorferi yielded fewer bacteria than Westchester strain-infected ticks. The Westchester strain has been passaged in a tick-rodent cycle in the laboratory, whereas the N40 and B31 strains have been cloned and passaged in culture. The passage history of the bacteria may account for the greater numbers of spirochetes within ticks infected with the Westchester strain.

Antigenic differences between culture-grown and tick-borne bacteria.

Previous studies have examined gross antigenic differences between cultured and tick-borne bacteria by culturing borreliae from ticks in BSK-II medium and comparing their antigenic profile to that of bacteria that have been maintained for a long time in culture (30). The main problem with this approach is that tick-borne spirochetes grown in BSK II medium, even for a short time, may not retain the bacterial antigenic profile expressed within the tick.

Studies point to tick-borne spirochetes having protein and mRNA profiles that are different from those of culture-grown organisms (14, 27). The technical advances reported here have allowed us to directly compare the antigens produced within ticks to those produced in culture. The results have demonstrated that several antigens are either up- or downregulated in ticks in comparison to cultured bacteria. The identities of these interesting proteins remain to be established. The large number of proteins that were differentially produced within ticks is consistent with the hypothesis that multiple bacterial proteins regulate transmission from tick to host.

OspA and C are the best-studied B. burgdorferi antigens with respect to expression within ticks and in culture. Unfed ticks produce primarily OspA and little or no OspC. During the blood meal, spirochetes upregulate OspC and 30 to 60% of bacteria downregulate OspA (28). In cultured bacteria, the levels of OspA and OspC are variable and influenced by variables such as temperature, pH, and growth stage (5, 14, 19, 25, 32). However, under standard laboratory conditions for growing B. burgdorferi, prototype strains such as B31 and N40 consist of bacteria that produce mostly OspA and little or no OspC. The Westchester strain used in the studies described here also followed that paradigm. Higher levels of ospC mRNA were detected in ticks than in culture. In IFA studies, only 0.4% of the cultured bacteria produced OspC whereas 10% of the bacteria in feeding ticks produced the protein. This number is not as high as the 40 to 60% OspC induction that has been reported for other strains such as N40 and B31 and may be related to strain differences. In the case of OspA, lower mRNA levels were detected in Westchester strain spirochetes within ticks than in culture. However, no differences were observed at the protein level between tick-borne and cultured bacteria both by IFA and by Western blotting. Although OspA transcription may be downregulated early during tick feeding, the protein already on the surface of the spirochete may be stable and not cleared as early as 2 days into the blood meal. In studies with the B31 strain, Schwan and Piesman reported that 2 days into the blood meal, 89.6% spirochetes produced OspA and the peak of OspA clearance was observed only 4 days into the blood meal (28). In preliminary studies with the Westchester strain, a greater proportion of bacteria were observed to clear OspA as the blood meal progressed beyond 2 days (data not shown).

Here we also report on the production Bbk32 and p66 by spirochetes within ticks and in culture. Probert and Johnson (24) demonstrated that B. burgdorferi Bbk32 is a fibronectin binding protein that may play a role in attachment of the spirochetes to the extracellular matrix. Moreover, it has also been suggested that coating the spirochete with fibronectin may mask it from the host's immune system (15). bbk32 is expressed by spirochetes in the host but not by Borrelia grown in culture (10). Using RT-PCR, we compared levels of bbk32 mRNA produced by spirochetes within feeding ticks and in culture. Our results demonstrating that bbk32 is expressed in ticks but not in culture are in agreement with the results of another recent study (12).

B. burgdorferi p66 is a molecule that is expressed in the mammalian host (3). p66 probably function as a bacterial adhesin because it binds to integrins (6). Bunikis and Barbour recently demonstrated that although organisms grown in culture produce p66, OspA masks p66 and prevents surface exposure (2). They speculated that one function of OspA, which is abundantly produced within the tick, maybe to protect surface molecules such as p66 from tick proteases. Our observation that p66 was indeed produced within the tick is consistent with the hypothesis of Bunikis and Barbour that OspA may serve a protective role in the tick. The downregulation of OspA that occurs as spirochetes move from the vector to the host may be required to expose p66 so that the protein can interact with integrins in the host.

In summary, we have developed a method for purifying spirochetes which permits direct biochemical characterization of spirochetes within feeding ticks. The bacteria purified from ticks had an antigen composition distinct from that of cultured bacteria. These results underscore the importance of using tick challenge instead of syringe inoculation of cultured bacteria in Lyme disease pathogenesis and vaccine studies.