In Vitro and In Vivo Neutralization of the Relapsing Fever Agent Borrelia hermsii with Serotype-Specific Immunoglobulin M Antibodies

Abstract

The antigenic variation of the relapsing fever agent Borrelia hermsii is associated with changes in the expression of the Vlp and Vsp outer membrane lipoproteins. To investigate whether these serotype-defining proteins are the target of a neutralizing and protective antibody response, monoclonal antibodies were produced from spleens of infected mice just after clearance of serotype 7 cells from the blood. Two immunoglobulin M monoclonal antibodies, H7-7 and H7-12, were studied in detail. Both antibodies specifically agglutinated serotype 7 cells and inhibited their growth in vitro. Administered to mice before or after infection, both antibodies provided protection against infection or substantially reduced the number of spirochetes in the blood of mice after infection. Whereas antibody H7-12 bound to Vlp7 in Western blotting, enzyme-linked immunosorbent assay, and immunoprecipitation assays, as well as to whole cells in other immunoassays, antibody H7-7 only bound to wet, intact cells of serotype 7. Antibody H7-7 selected against cells expressing Vlp7 in vitro and in vivo, an indication that Vlp7 was a conformation-sensitive antigen for the antibody. Vaccination of mice with recombinant Vlp7 with adjuvant elicited antibodies that bound to fixed whole cells of serotype 7 and to Vlp7 in Western blots, but these antibodies did not inhibit the growth of serotype 7 in vitro and did not provide protection against an infectious challenge with serotype 7. The study established that a Vlp protein was the target of a neutralizing antibody response, and it also indicated that the conformation and/or the native topology of Vlp were important for eliciting that immunity.

To early immunologists, such as Ehrlich and Metchnikoff, the infection relapsing fever revealed the fine specificity of adaptive immunity (reviewed in reference (27)). During relapsing fever the immune response cleared the initial wave of spirochetes from the blood, yet the illness recurred with the proliferation of another antigenic type in the patient or experimental animal. Immunity developed to the second wave of bacteria in the blood, but the cumulative immune response did not prevent proliferation of a third serotype. Surviving untreated patients with relapsing fever seemed to recover only when the repertoire of serotypes was exhausted.

Humoral immunity is sufficient for clearing relapsing fever agents, such as Borrelia hermsii and B. turicatae, from the blood. In 1896 Gabritchewsky reported that serum from an infected patient lysed the spirochetes (15), and in 1906 Novy and Knapp showed that antiserum alone provided protection (22). Stoenner et al. confirmed and extended the findings of these and other early investigators by demonstrating that antiserum to a specific serotype would clear the blood of that particular serotype but not other serotypes (34). Newman and Johnson found that neither T cells nor the terminal components of complement were necessary for clearance of B. turicatae from the blood of mice (20, 21).

Early investigations of relapsing fever centered on immunity, but most of the recent research has focused on the pathogen. This more recent research has shown that B. hermsii can sequentially display up to about 30 different surface antigens during an infection's course (26, 34). These antigens are abundant lipoproteins and are of two major types: variable small proteins (Vsp) of about 22 kDa and variable large proteins (Vlp) of about 38 kDa (4, 9). As examples, serotype 7 of B. hermsii expresses Vlp7 on its surface, and serotype 26 expresses Vsp26 instead. Only one vlp or vsp gene is transcriptionally active at a time, a state that is determined by which one of the several different alleles is downstream of a particular promoter (17, 26). A vsp or vlp gene is replaced by another at this expression site by different kinds of recombination, most commonly a gene conversion from an archival site (25, 26).

While research on pathogenesis progressed, few investigations of immunity in relapsing fever were carried out. Mechanisms for Vsp and Vlp variation were better understood, but whether these proteins were actually the targets of a neutralizing response had not in fact been established. For the present study we obtained serotype-specific monoclonal antibodies from an early stage of the adaptive immune response to infection. These antibodies were then used in different functional and matrix-based immunoassays and in mouse infection experiments. The aims were to establish which antigen or antigens were the target of these antibodies and to characterize the state of the antigen under which the antibodies were neutralizing. Although this usage of “neutralization” is usually in the context of viruses, we think it also applies to bacteria when the endpoints are prevention against challenge, clearance of infection, and inhibition of growth.

MATERIALS AND METHODS

Organisms and culture conditions.

Serotypes 7, 21, 26, and 33 of strain HS1 of B. hermsii and strain B31 of B. burgdorferi (ATCC 35210) were used. Serotype 33 of B. hermsii was formerly called serotype C (4, 11). The spirochetes were cultivated in Barbour-Stoenner-Kelly medium (BSK) II broth, harvested by centrifugation, and stored frozen at −80°C in medium with 10% dimethyl sulfoxide as previously described (2). Spirochetes were counted in a Petroff-Hausser chamber by phase-contrast microscopy. In some experiments intact spirochetes were treated with proteinase K (Boehringer Mannheim, Indianapolis, Ind.) as described previously (5); for this study the cell concentration was 2 × 10⁹ cells/ml in phosphate-buffered saline (PBS) with 5 mM MgCl₂, the final proteinase K concentration was 200 μg/ml, and the incubation was for 40 min at 20°C.

Protein antigens.

The vlp7 gene was amplified by PCR from plasmid p7End3 (17). The forward primer was 5′ CCGAGATCTCATATGGCTGGACAACAACCAG3′, and the reverse primer was 5′ GATATCTAGATCTCACTTACTTGATTC3′. The underlined sequences correspond to the sequence of vlp7 at its 5′ and 3′ ends (9). The forward primer also includes BglII and NdeI restriction sites and codons for methionine and alanine, and the reverse primer also includes a BglII restriction site. Each PCR cycle with Taq polymerase (Perkin-Elmer) was 94°C for 1 min, 39°C for 1 min, and 72°C for 5 min. The product was first digested with BglII and cloned into the BamHI site of the pET11a vector; this plasmid was then digested with NdeI and religated without the small NdeI fragment at the cloning site. The construct was transformed first into Escherichia coli strain HMS174 and then into E. coli BL21(DE3). The insert was confirmed as vlp7 by sequencing. The insert encoded Vlp7 without the signal peptide and with a methionine-alanine replacing the N-terminal cysteine of the processed protein. The product was reactive with monoclonal antibodies H12915 and H12936 by Western blotting. A similar recombinant construct of the ospA gene of B. burgdorferi in pET9 has been described previously (13). E. coli containing the recombinant vectors was grown in Luria Bertani broth at 37°C with 100 μg of carbenicillin per ml. At an optical density at 600 nm of 0.5, isopropyl-β-d-thiogalactopyranoside (IPTG) was added, and the incubation continued for 12 h. At this point Vlp7 and OspA were estimated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) to be ∼50% of the total protein in the cells. The cells were harvested by centrifugation at 12,000 × g for 10 min and resuspended in PBS with phenylmethylsulfonyl fluoride (PMSF) at 25 μg/ml. The cell suspension was lysed in a French press, and the product was centrifuged at 48,000 × g at 10°C for 3 h. The supernatant was filtered through a 0.22-μm-pore-size membrane filter. Protein concentrations were determined by the Bradford method (Bio-Rad, Richmond, Calif.). By PAGE and Western blot analysis (see below), recombinant Vlp7 or OspA proteins were ∼90% of the total protein in the supernatant. Native Vlp7 protein was prepared from cultured B. hermsii by the method of Barstad et al. (7).

PAGE and Western blot analysis.

Spirochete lysates and other samples were subjected to SDS-PAGE and Western blot analysis as described previously (4). The acrylamide concentration of the separating gel was 12.5%. Molecular weight standards were obtained from GIBCO-BRL. Bound antibodies were detected with either alkaline phosphatase-conjugated protein A/G (Pierce, Rockford, Ill.) or alkaline phosphatase-conjugated goat anti-mouse immunoglobulin M (IgM) (Zymed, San Francisco, Calif.). The nitroblue tetrazolium–5-bromo-4-chloro-3-indolylphosphate (NBT/BCIP) substrate kit of Pierce was used for color development.

IFA.

For indirect immunofluorescence assays (IFA) of fixed cells, spirochetes were washed with PBS, and approximately 10⁴ cells in 10 μl were spotted within printed circles of glass slides (Roboz Surgical Instruments, Washington, D.C.). After the spots dried, the slides were fixed in methanol for 30 min, air dried, and then used immediately. The remainder of the procedure with undiluted hybridoma supernatant or a 1:100 dilution of ascitic fluid was as described previously (4). For IFA of live cells, spirochetes were washed with PBS and 10⁶ cells were resuspended in PBS with 2% bovine serum albumin (BSA; ICN, Costa Mesa, Calif.). At least 90% of the cells were motile under these conditions. Ascitic fluid under examination was added for a final concentration of 1:100 and a total volume of 0.5 ml. The suspension was gently rocked at room temperature for 1 h on a Labquake rotator (Lab Industries, Berkeley, Calif.). The suspension was then centrifuged at 8,000 × g for 10 min, washed once with PBS-BSA, and then resuspended in a 30-μl volume of a 1:15 dilution of fluorescein-conjugated, sheep anti-mouse IgG plus IgM plus IgA (Boehringer Mannheim) in PBS-BSA. The suspension was incubated for 30 min in the dark at 37°C. It was then diluted 1:10 by the addition of 270 μl of PBS-BSA, and 3 μl was immediately examined under fluorescent microscopy at 400× magnification.

ELISA.

For enzyme-linked immunosorbent assays (ELISA) with dried cells, washed spirochetes were suspended in PBS, and 10⁶ cells in a 50-μl volume were placed in wells of an Immulon 1 microtiter plate (Dynatech Laboratories, Chantilly, Va.). After incubation of the plates for 18 h at 37°C, 300 μl of 3% dried nonfat milk in PBS (milk-PBS) was added to each well. After 1 h, the wells were washed twice with PBS, and then 50 μl of antibody diluted in milk-PBS was added to each well. The plates were incubated for 2 h at room temperature and then washed three times with PBS. Bound antibody was detected with alkaline phosphatase-conjugated goat anti-mouse IgG plus IgM plus IgA in milk-PBS for 1 h at 37°C. The plates were washed twice with PBS, and the substrate p-nitrophenyl phosphate disodium in diethanolamine buffer (Sigma-Aldrich, St. Louis, Mo.) was added. For ELISA of recombinant proteins, 250 ng of protein in 100 μl of carbonate buffer (7.5 mM Na₂CO₃–17.5 mM NaHCO₃ [pH 9.6]) was dispensed into wells of Immulon 2 plates (Dynatech); the plates were then kept in a moist chamber at 4°C for 14 h. The liquid was aspirated, and then the assay proceeded as described above with blocking with milk-PBS. The positive control for this assay was recombinant OspA with monoclonal antibody H5332. For ELISA with wet cells, washed spirochetes were resuspended in carbonate buffer, and 10⁶ cells in 100 μl were loaded into wells of Immulon 1 plates, which were kept in a moist chamber at 4°C for 14 h. The suspension was removed, and the blocking step with milk-PBS proceeded. For assays with cells the hybridoma supernatants were undiluted, and the ascitic fluids were twofold serially diluted from 1:16. For assays with proteins the ascitic fluids were diluted 1:100. For both protein and cell assays the absorbance values of wells at 490 nm were recorded on an ELISA reader (Dynatech), and the cutoff point between a positive and negative titer was an optical density of 0.2. The assays were performed in duplicate.

Agglutination assays.

For the microagglutination assay, a 5-μl volume of undiluted hybridoma supernatant was mixed on a slide with 5 μl of a suspension of 4 × 10⁸ spirochetes per ml of RPMI 1640 (GIBCO-BRL, Gaithersburg, Md.). After incubation of the slide for 10 min at room temperature, a coverslip was applied and the slide was viewed by phase-contrast microscopy at 400× magnification for spirochete agglutination. Hybridoma supernatant containing antibody H5332 was used as a negative control. An assay for macroagglutination of spirochetes by antibodies in round-bottomed wells of polystyrene microtiter plates (Corning Lab Sciences, Corning, N.Y.) was carried out essentially as described elsewhere (30). The cells were suspended in BSK II medium with 6% heat-inactivated rabbit serum but without gelatin, neopeptone, yeastolate, or CMRL 1066 (BSK-B). Ascitic fluids were diluted in the same medium. The final concentration of cells was 10⁹/ml, and the ascitic fluids were at final dilutions of 1:10 or 1:100. Plates were incubated for 1 h at 37°C, and then well bottoms were examined for grossly visible clumping of cells. The assay was performed in duplicate.

Monoclonal antibodies.

The Vlp7-specific antibody H12915 (4, 7) a Vsp26-specific antibody (25), the Vsp33-specific antibody H4825 (3), and the OspA-specific antibody H5332 (5) have been described elsewhere. Additional monoclonal antibodies were produced for this study. Adult BALB/c mice were inoculated intraperitoneally with 10⁵ live, culture-grown serotype 7 cells on day 0, and infection was monitored by tail vein sampling. On day 4 the mice were euthanized, and their spleens were used for producing hybridomas as described previously (28). The supernatants were screened by IFA and ELISA with dried, whole cells of serotype 7 or 21, by microagglutination assay with serotype 7 cells, and by Western blot analysis with serotypes 7 and 21. Ascitic fluids were produced as described previously (28). Antibody isotypes were determined with the Immunotype kit of Sigma-Aldrich.

GIA and in vitro selection.

The growth inhibition assay (GIA) was carried out with B. hermsii or B. burgdorferi as described elsewhere (30). There were 10⁶ spirochetes in each well, and plates were incubated with plastic seals (Sensititre Microbiologic Systems, Westlake, Ohio) at 34°C for 72 h before gross visual examination for color change and microscopic examination for immobility of cells and for clumping and blebbing of cells. In some experiments 2 hemolytic units of guinea pig complement (Sigma-Aldrich) were added to each well. Selection with monoclonal antibodies in medium was carried out by adding dilutions of ascitic fluid to 6 ml of BSK II medium containing 10⁶ spirochetes in tubes. The cultures were incubated at 34°C for 7 to 10 days and then passed at a 1:30 dilution to another tube of medium with the same antibody. The spirochetes that grew in the second culture were harvested and examined by PAGE and Western blotting with antibodies to serotypes 7, 21, 26, and 33.

Immunoprecipitation.

Spirochetes were centrifuged (12,000 × g for 3 min) and resuspended in 180 μl of BSK-B at a density of 10⁹ cells/ml. To the suspension was added 20 μl of ascitic fluid, and this was incubated for 1 h at 37°C. The cells were centrifuged and resuspended in 1% octyl glucopyranoside (Boehringer Mannheim) in 50 mM Tris (pH 7.5)–150 mM NaCl–5 mM EDTA (TSE) with 25 μg of PMSF/ml. The lysate was incubated at 37°C for 1 h and then centrifuged. To the supernatant was added 50 μl of goat anti-mouse IgM-coupled microspheres (Kirkegaard & Perry, Gaithersburg, Md.) which had been blocked for 2 h with 5% nonfat milk in TSE. The suspension was incubated for 2 h at 37°C on a Labquake rotator. The beads were pelleted by centrifugation, washed four times with TSE, resuspended in a 0.1× volume of PAGE sample buffer, and then subjected to PAGE and Western blot analysis with antibody H12915.

Mouse infections.

Female, 4- to 6-week-old BALB/c or C57BL/6 mice (Jackson Laboratories, Bar Harbor, Maine) were infected by intraperitoneal (i.p.) inoculation with 10³ spirochetes. Male, adult nude (nu/nu) BALB/c mice or their nu/+ littermates (Jackson Laboratories) were inoculated i.p. with 10⁵ spirochetes in 0.1 ml of PBS. Daily assessments of spirochetemia were made by obtaining blood from a tail vein, mixing it with an equal volume of PBS, and then examining it under a coverglass by phase-contrast microscopy. Forty 400× magnification fields were examined, and the following scoring system for the spirochetemia was used: −, 0; +, 1 to 5; ++, 6 to 39; and +++, 40 to 200 spirochetes/40 fields. Citrated blood was cultured in broth medium as described previously (2), and the first-passage cultures were examined by SDS-PAGE and Western blot analysis with antibodies. Identity of spirochetes in the blood was also determined using PCR with primers for the expressed vlp7 gene as described previously (24).

Passive transfer and treatment with antibodies.

For passive immunization, BALB/c mice which had received 650 rads from a ¹³⁷Cs gamma radiation source were inoculated i.p. with 0.5 ml of ascitic fluid diluted in PBS, 1 h before the injection of 10³ spirochetes. For the treatment experiment, irradiated BALB/c mice were inoculated with 10³ spirochetes and tail vein blood was monitored for the amount of spirochetemia. When the spirochetemia had reached its peak, i.e., a spirochetemia score of +++, the mice were infused through the tail vein with 100 μl of ascitic fluid diluted in PBS.

Immunization and infectious challenge.

Adult female BALB/c or C57BL/6 mice were immunized with recombinant protein emulsified with complete Freund's adjuvant by subcutaneous injection at four sites on day 0 and then boosted once or twice with either the same adjuvant in PBS or incomplete Freund's adjuvant. Mice were bled for immunoassays and then challenged with 10³ serotype 7 spirochetes i.p. at different intervals after the last immunization. Infection was assessed by microscopic examination of the blood, by PCR of the plasma, and by culture.

RESULTS

Monoclonal antibodies.

We previously obtained serotype-specific antibodies by infecting mice, treating them with antibiotics at the height of spirochetemia, and then boosting them about 1 month later with the same serotype (4, 7). By this method we obtained IgG monoclonal antibodies H12915 and H12936, among others. For the present study, we sought monoclonal antibodies that were representative of the immune response that cleared the blood of the first wave of spirochetes. Accordingly, mice were infected and spleen cells were obtained on the first day that spirochetes were undetectable in the blood. From two mouse spleens 60 hybridomas that produced antibodies to serotype 7 but not to serotype 21 were identified by screening with IFA, ELISA, microagglutination, and Western blot analysis; 90% of the hybridomas produced IgM antibodies. Only one hybridoma in the screenings bound to serotype 21 as well as to serotype 7. Western blot screening also revealed a single hybridoma that produced antibody that bound to a small amount of Vsp33 in the predominantly serotype 7 population (6). This antibody was designated H33-5 and served as a control.

Among the 60 serotype 7-specific antibodies, five different patterns of reactivity were observed in the assays: (i) 11 antibodies were positive by IFA, ELISA, Western blotting, and agglutination; (ii) 31 antibodies were positive by IFA, ELISA, and Western blotting but did not agglutinate cells; (iii) 9 antibodies bound to serotype 7 cells by IFA and ELISA but were negative by Western blotting and did not agglutinate cells; (iv) 8 antibodies were positive by IFA, ELISA, and agglutination assays but were negative by blotting; and (v) 1 antibody agglutinated cells but was negative by IFA, ELISA, and Western blotting. There were no antibodies that were positive by blotting but negative by IFA, ELISA, or agglutination assay.

Three serotype 7-specific antibodies were selected for further study: H7-7, which agglutinated cells but was unreactive by Western blotting; H7-12, which agglutinated cells and also bound to Vlp7 in blots; and H7-9, which bound to Vlp7 in blots but did not agglutinate cells. The H7-7, H7-9, H7-12, and H33-5 antibodies were monoclonal and were typed as IgM. Ascitic fluids were produced with each of the hybridomas, and these were used in subsequent, more detailed assays of antibody reactivities.

Western blot assay, ELISA, and IFA.

The Western, ELISA, and IFA assays measure binding of antibody to antigen on a matrix or in suspension and do not assess function. We used whole cells of serotypes 7, 21, and 33 for these experiments. The results of these and other assays are summarized in Table 1. By Western blotting, antibodies H7-12 and H7-9, as well as the positive control IgG_2a antibody H12915, bound to Vlp7 in the serotype 7 lysate but not to Vlp21 or Vsp33 in the other lysates. Antibodies H33-5 and H4825 bound to Vsp33 in the blots. Antibody H7-7 did not detectably bind to Vlp7 or any other component of B. hermsii on the blots, even when the PAGE samples were not heated and were prepared without reducing agent (data not shown).

TABLE 1.

Summary of results of different immunoassays with monoclonal antibodies to B. hermsii

Assay	Antigen (serotype, lipoprotein)	Reciprocal titer for monoclonal antibody:
		H7-7	H7-12	H7-9	H12915	H33-5	H4825
Western blota	7, Vlp7	−	+	+	+	−	−
	21, Vlp21	−	−	−	−	−	−
	33, Vsp33	−	−	−	−	+	+
ELISAb	7	<16	1,024	1,024	NDc	32	32
	7 (wet)	2,048	8,192	ND	ND	<16	ND
	21	<16	32	16	ND	32	32
	33	32	32	16	ND	4,096	1,024
	Recombinant Vlp7	−	+	+	+	−	ND
	Native Vlp7	−	+	+	+	−	ND
IFAd	7	−	+	+	+	+	+
	7 (wet)	+	+	ND	+	ND	−
	21		−	−	−	−	−
	33	−	−	−	−	+	+
Agglutination	7	≥100	≥100	<10	<10	<10	<10
	21	<10	<10	ND	<10	<10	<10
	33	<10	<10	<10	<10	≥100	≥100
GIA	7	131,072	1,024	<16	<16	<16	<16
	33	<16	<16	<16	<16	524,288	32,768

^aWhole-cell lysates were subjected to Western blot analysis. +, antibody bound to the homologous Vlp or Vsp; −, antibody did not bind to any component in the lysate. 

^bFor assays with cells with the various serotypes, limiting dilutions of serum from 1:16 were carried out. Vlp7 assays were performed at a serum dilution of 1:100, and the results are reported as + or −. 

^cND, not done. 

^dIFA with cells and a serum dilution of 1:100. Results are reported as + or −. 

The same patterns of results were obtained by ELISA with dried cells adherent to the plastic and by IFA with cells fixed to glass (Table 1). H7-12, H7-9, and H12915 bound to recombinant Vlp7 and native Vlp7 in the ELISA, but H7-7 bound to neither form of Vlp7 under the same conditions. Antibody H7-7 but not antibody H33-5 bound to serotype 7 cells that remained unfixed and wet for IFA and ELISA. Antibody H33-5 did bind to serotype 33 cells under these conditions.

Immunoprecipitation.

Having shown that H7-7's binding was specific for serotype 7, we next used immunoprecipitation to assess the binding of H7-7 to Vlp7. The antibodies were allowed to bind to the cells under conditions in which H7-7 bound to wet cells. The cells were then lysed with a nondenaturing detergent before precipitation with anti-IgM antibody on beads. The proteins in the gel were then transferred to a blot, and this was probed with anti-Vlp7 antibody (Fig. 1). Antibody H7-12, but not antibodies H7-7 or H33-5, immunoprecipitated Vlp7.

FIG. 1.

Western blot analysis of immunoprecipitation products of B. hermsii strain HS1. Serotype 7 cells were incubated with IgM monoclonal antibodies H33-5, H7-7, or H7-12. After lysis in nondenaturing detergent the antigen-antibody complexes were precipitated with anti-IgM beads, and these were subjected to PAGE (12.5% acrylamide) and Western blot analysis with anti-Vlp7 monoclonal antibody as the probe. The positive control lane contained serotype 7 cells, and the negative control lane contained serotype 33 cells. On the left are the migrations of molecular weight standards (MWS) in thousands.

Agglutination and growth inhibition assays.

The preceding experiments demonstrated that H7-7 bound to serotype 7 cells but only under what were close to live culture conditions for the cells. We further investigated the functional effects of this antibody and the other antibodies on viable cells. The results are summarized in Table 1. Antibodies H7-7 and H7-12 agglutinated serotype 7 cells but not serotype 21 or 33 cells. Antibodies H33-5 and H4825 agglutinated serotype 33 cells but not cells of the other two serotypes. The antibodies H7-9 and H12915 were serotype-specific in other assays but did not agglutinate cells of any serotypes in this assay. If serotype 7 cells were first treated with proteinase K, neither H7-7 nor H7-12 agglutinated the cells (data not shown).

While H7-7 had a 4-fold lower titer than H7-12 by wet ELISA, H7-7's titer was 128-fold higher than H7-12's titer by GIA. Addition of complement to the medium at the start of incubation did not change the minimal inhibitory titer of H7-7 but did increase by eightfold the inhibitory titer of H7-12. Neither H33-5 nor H4825 inhibited serotype 7 spirochetes but these ascitic fluids did inhibit the growth of serotype 33 spirochetes at high dilutions.

In vitro selection with antibodies.

To cultures of serotype 7 or serotype 33 was added ascitic fluid with antibody H7-7 at a final dilution of 1:1,000 or antibody H33-5 at a 1:125 dilution. The populations that grew to stationary phase were passed to fresh media with the same antibody. The second-passage cultures were analyzed by PAGE and Western blotting (Fig. 2). Antibody H33-5 prevented any growth in the serotype 33 culture; it had no effect on the growth of the serotype 7 culture (data not shown). On the other hand, antibody H7-7 did not inhibit the growth of the serotype 33 cells, but it did delay by 3 days the stationary growth phase of the culture inoculated with serotype 7 cells. After the second passage of this culture in H7-7 antibody, the cells in the population predominantly expressed Vsp33; Vlp7 in the H7-7-exposed population was undetectable by Western blotting.

FIG. 2.

Western blot analysis of B. hermsii populations after in vitro selection with monoclonal antibodies H33-5 or H7-7. Cultures of serotype 7 were incubated without antibodies, with anti-serotype 33 antibody H33-5, or with anti-serotype 7 antibody H7-7 for two passages and the harvests were subjected to PAGE (12.5% acrylamide) and Western blot analysis with anti-Vlp7 antibody H12915 or anti-Vsp33 antibody H4825. Serotype 33 cells were in the right-most lane of each blot. On the left are the migrations of molecular weight standards (MWS) in thousands.

Passive transfer and treatment with antibodies.

For the in vivo experiments we used antibodies H7-7 and H7-12 because they agglutinated serotype 7 cells and inhibited growth in vitro. These are characteristics that were associated with protective immunity in studies of B. burgdorferi (18, 29). In the first experiment, groups of five irradiated mice were administered the antibodies H7-7, H7-12, or H33-5 before the infection, and then the blood was examined on days 4, 7, and 10. All five mice that received antibody H33-5 were spirochetemic on day 4. None of the five mice that received antibody H7-7 had detectable spirochetes in the blood on days 4, 7, or 10 (P < 0.01, by two-tailed Fisher's exact test). Of the five mice that received antibody H7-12, only one was spirochetemic on day 7 (P < 0.05). Culture of the blood revealed a population that was predominantly serotype 7. Two of the five H33-5 mice were then treated intravenously with antibody H7-7, and another two received H33-5. On examination the next day both mice that had been treated with H33-5 remained spirochetemic, while the mice that received H7-7 were not.

In the second treatment experiment, antibodies H7-7, H7-12, or H33-5 were given to irradiated mice after the infection had reached its peak in the blood. Antibodies H7-12 and H33-5 were diluted 1:3 in PBS, and antibody H7-7 was diluted 1:20. The effectiveness of the antibodies in reducing the density of spirochetes in the blood was then assessed. The results are shown in Table 2. The infected mice that were treated with either antibody H7-7 or H7-12 were significantly more likely than mice that received antibody H33-5 to have undetectable (0) or low numbers (+ score) of spirochetes in the blood on day 5 (P < 0.01). The H7-12-treated mice were more likely than the H7-7-treated mice to have detectable spirochetes in the blood on day 5 (P = 0.06) and heavy infections (+++) on day 6 (P < 0.01).

TABLE 2.

Treatment of B. hermsii-infected, irradiated mice with monoclonal antibodies

MAba	Mouse no.	Spirochetemia scoreb on:
		Day 4	Day 5	Day 6
H7-7	1	+++	−	+
	2	+++	−	+
	3	+++	−	+
	4	+++	−	−
	5	+++	−	+
	6	+++	−	+
H7-12	7	+++	−	+++
	8	+++	+	+++
	9	+++	+	+++
	10	+++	−	+++
	11	+++	+	+++
H33-5	12	+++	+++	+++
	13	+++	+++	+++
	14	+++	+++	+++
	15	+++	+++	+++
	16	+++	+++	+++

^aMAb, monoclonal antibody. 

^bScoring: −, 0; +, 1 to 5; ++, 6 to 39; +++, 40 to 200 spirochetes/40 fields at 400× magnification. 

With the exception of one of the H7-7-treated mice, the mice had detectable spirochetes in the blood on day 6 in the second experiment. Blood cultures were obtained from three of the five spirochetemic mice that had been treated with H7-7, five of the five mice that had received antibody H7-12, and five of the five mice that had received antibody H33-5. All 10 of the H7-12-treated or H33-5-treated mice had serotype 7 in the blood cultures. In contrast, two of the three H7-7-treated mice had serotype 26 in the blood instead of serotype 7 (P < 0.05). Antibody H7-7 selected against cells expressing Vlp7 in vivo as well as in vitro.

Immunization with recombinant proteins.

Groups of 5 BALB/c mice were immunized with 20 μg of Vlp7 or OspA with complete Freund's adjuvant on day 0, boosted with the same dose of antigen but in incomplete Freund's adjuvant on days 14, 34, and 64, and then bled and challenged on day 88 with 10³ cells of serotype 7. By day 94 all the mice were infected by direct examination and culture of the blood; PCR confirmed that the populations were predominantly serotype 7. By dry whole-cell ELISA, the five mice vaccinated with Vlp7 had reciprocal titers to serotype 7 cells of 2,048 (n = 3) or 4,096 (n = 2). By GIA without added complement, mice immunized with OspA inhibited B. burgdorferi growth at titers of 1:256 or greater. At dilutions of 1:16 or higher the prechallenge sera from mice immunized with Vlp7 did not inhibit serotype 7 growth. In another experiment, five C57BL/6 mice received 20 μg of recombinant Vlp7 with complete Freund's adjuvant on day 0 and were boosted on days 14 and 28, bled on day 42, and then challenged with serotype 7 on day 60. All mice became infected, and none had growth-inhibiting antibodies to serotype 7 as detected by GIA.

Nude mouse infections.

Having observed the effectiveness of IgM antibodies for neutralization and passive protection, we determined whether infection with serotype 7 could elicit a neutralizing immune response from T cell-deficient mice. Groups of 4 nu/nu or nu/+ mice were infected on day 0, and the infection was monitored by microscopic examination of the blood. On day 3 the nude mice and their immunocompetent littermates were spirochetemic, but by day 5 spirochetes were not detectable in the blood in any of the mice. Both groups of mice had relapses of spirochetes in the blood by day 8.

DISCUSSION

Variable antigens had been identified in B. hermsii and other relapsing fever agents, and the Vlp and Vsp lipoproteins in the outer membrane had been shown to define each of the several serotypes that a single strain can manifest, but the evidence that the Vlp and Vsp proteins were the targets of a protective or neutralizing immune response had heretofore been indirect. The present study established that at least of one these proteins, Vlp7, is the target of protective and neutralizing antibody. Monoclonal IgM antibody H7-12 specifically bound to native and recombinant Vlp7 in Western blotting, ELISA, or immunoprecipitation assays. In functional in vitro assays this antibody agglutinated spirochetes and inhibited the growth of cells of the homologous but not heterologous serotypes. In mouse infection experiments, passively administered H7-12 antibody protected mice from challenge and substantially and rapidly reduced the number of serotype 7 cells in the blood of infected mice.

Another serotype-specific IgM antibody, H7-7, was even more efficacious in protecting against infection and treating infected mice, but this antibody did not bind to Vlp7 in ELISA, Western blotting, or immunoprecipitation assays. The antibody bound only to wet, intact cells in the assays. The lack of binding of H7-7 to protease-treated cells suggests that its antigen is a protein. Other evidence indicates that the protein is Vlp7. Firstly, the binding and neutralization effects of H7-7 were specific for serotype 7. The antibody did not affect either serotype 21 or 33. Secondly, under both in vitro and in vivo conditions antibody H7-7 selected against serotype 7 cells, an indication that the antibody selection was directed against Vlp7. It is possible that serotype 7 differs from other isogenic serotypes in another protein besides Vlp, but this has not been noted to date (4, 9, 11, 17, 26).

Ascitic fluid with antibody H7-7 was one-fourth as active as H7-12 ascitic fluid by wet cell ELISA, but in the functional in vitro assays and the mouse challenge experiments antibody H7-7 retained its activity at 10- to 100-fold higher dilutions than that for H7-12 activity. The conditions under which H7-7 did bind to serotype 7 cells suggest that the antibody-epitope interaction either depends on the conformation of Vlp7 or requires a certain topological arrangement of Vlp7 within the membrane. The latter may include multimerization of Vlp7 or association with another protein. In B. burgdorferi certain outer membrane lipoproteins hinder access of antibodies to another outer membrane protein (8).

Further evidence of the importance of the conformation and in situ characteristics of Vlp7 came from the active immunization studies. Using a protocol similar to one with a recombinant OspA that produced protective immunity against B. burgdorferi (14, 32), we found that immunization with recombinant Vlp7 did not provide protection against infection with serotype 7 B. hermsii in three different experiments. In the present study immunization with unlipidated OspA with adjuvant produced antibodies that bound to OspA in blots and B. burgdorferi cells in vitro as expected. The OspA immunization also produced antibodies that inhibited the growth of B. burgdorferi at high titers. By the same procedure, recombinant Vlp7 elicited antibodies that bound to Vlp7 in blots and to serotype 7 cells by ELISA, but there was not a neutralizing response detected by in vitro assay. The lack of binding of H7-7 to either native or recombinant Vlp7 indicates that lipidation or other possible posttranslational modifications in B. hermsii do not account for the results of infectious challenge of mice vaccinated with recombinant Vlp7. It is possible that there is a second, yet-to-be-identified antigen that codefines serotype 7 and that this hypothetical antigen is the target of protective antibodies such as H7-7. However, we think that this is a less parsimonious explanation than one that invokes the critical importance of Vlp conformation. A study of B. burgdorferi demonstrated the role of the conformation of the OspC protein in providing a protective immune response as an immunogen (16).

The present study confirmed that T-cell help is not a requirement for clearance of B. hermsii from the blood of mice (21) and that IgM antibodies can be protective (1, 19, 35). Nude mice cleared spirochetes from the blood as effectively as wild-type mice, and IgM monoclonal antibodies were neutralizing in the presence or absence of complement. Nude mice are also able to control infections by African trypanosomes, such as Trypanosoma brucei (10, 23). African trypanosomes manifest a form of antigenic variation that shares several features with the antigenic variation of relapsing fever (12).

Although they are neither carbohydrates nor polypeptides with repetitive epitopes (7, 9), the polymorphic Vlp and Vsp proteins have other features that may explain the T-cell-independent response to these antigens. Firstly, they are lipoproteins, the lipid moiety of which may directly and indirectly stimulate cytokine help for the humoral response to T-cell-independent antigens (33). Secondly, the abundance of Vlp and Vsp proteins on the spirochete's surface may be akin to the tight packing of identical capsid proteins in virus particles. The capsid proteins may function as T-cell-independent antigens in arrays but not as monomers (36). Thirdly, some of the Borrelia spp. variable proteins are resistant to proteases (37). Consequently, they may persist in infected animals in an undegraded form, a state that has been associated with a T-cell-independent response (31).