471

Nonspecific Proliferative Responses of Murine Lymphocytes to Borrelia burgdorferi Antigens Mark S. de Souza, Erol Fikrig, Abigail L. Smith, Richard A. Flavell, and S. W. Barthold

Section of Comparative Medicine; Section of Immunobiology; Division of Infectious Diseases, Department ofInternal Medicine; Howard Hughes Medical Institute, Yale University School of Medicine, New Haven. Connecticut

An inbred laboratory mouse model for the study of Lyme disease has recently been described. All mice tested became infected and developed arthritis, but disease severity was dependent on mouse strain, with resistant BALB/c mice developing mild disease and susceptible C3H/He mice developing severe disease [I]. The C3H/He mouse develops early spirochetemia after intradermal inoculation of Borrelia burgdorJeri, with subsequent polysynovitis, vasculitis, and carditis [2]. Mice are susceptible to infection and disease at all ages and develop disease as fully immunocompetent hosts. The mouse model will allow investigation of the pathogenesis of Lyme disease under carefully defined genetic and microbiologic conditions, Cases of human Lyme disease have been reported in the absence of detectable antibody production, but these patients had vigorous in vitro B. burgdorferi-specific peripheral blood lymphocyte proliferative responses [3]. A subsequent report found that the proliferative response by human lymphocytes to B. burgdorferi was nonspecific, since peripheral blood lymphocytes from healthy controls proliferated at levels comparable 'to lymphocytes derived from patients with Lyme disease [4]. . The present study addresses the effect of B. burgdorferi on lymphocytes derived from Lyme disease-susceptible and -resistant strains of mice in an attempt to clarify these findings under controlled laboratory conditions. The effect of both whole spirochetes and recombinant outer surface pro-

Received 19 August 1991; revised 12 November 1991, Grant support: AI-26815 (National Institute of Allergy and Infectious Diseases). Reprints or correspondence: Dr. Stephen W. Barthold, Section of Comparative Medicine, Yale University School of Medicine, 333 Cedar St., P.O. Box 3333, New Haven, CT 06510. The Journal of Infectious Diseases 1992;165:471-8 © 1992 by The University of Chicago. All rights reserved. 0022-\899/92/6503-0010$0 \.00

teins OspA and OspB on splenic lymphocyte function was assessed,

Materials and Methods Mice. Inbred C3H/HeNCrlBr (C3H) mice were purchased from Charles River Laboratories (Raleigh, NC). Inbred BALB/ cByJ (BALB) and C3H/HeJ mice were purchased from the Jackson Laboratory (Bar Harbor, ME). Specific pathogen-free mice were shipped in filtered crates, housed in microisolator cages (Lab Products, Maywood, NJ), and provided food (Agway, Syracuse, NY) and water ad libitum. Mice were 4-6 weeks old in all studies and were killed with carbon dioxide gas, followed by cardiac exsanguination. B. burgdorferi. Three strains of B. burgdorferi were used: N40 [I, 5], B31 [6], and 25015 [7]. All strains were low-passage (five passages or less) and grown in modified Barbour-StoermerKelly II medium [8] at 34°C, Spirochetes were grown to a concentration of 108 viable organisms/ml, pelleted by centrifugation at 10,000 g for 3 min, washed three times in PBS, and enumerated using darkfield microscopy. They were resuspended in RPM I 1640 medium (GIBCO BRL, Gaithersburg, MD) containing 5% fetal bovine serum (GIBCO), 0.05 mM 2-mercaptoethanol (Sigma Chemical, St. Louis), 2 mM L-glutamine (GIBCO), 100 units/rnl penicillin, and] 00 JLg/ml streptomycin (RPMI complete) at concentrations ranging from 2 X ] 0 4 to 2 X 107 organisrns/rnl before addition to lymphocyte culture. These counts represent protein concentrations of 0.0] and 10 JLg/ml, respectively, as determined by the Bradford assay (Bio-Rad Laboratories, Richmond, CA). For certain studies, spirochetes were inactivated in a 60°C water bath for I h. Unless otherwise stated, data shown represent organisms at a concentration of2 X 107/ml (10 JLg/ml protein), which gave the optimal response. Recombinant proteins. Recombinant OspA from B. burgdorjeri strain N40 was generated as a glutathione-transferase (GT) fusion protein as previously described [9]. The gene for OspB from B. burgdorferi strain B3 I was amplified from plasmid pTRH46 (provided by A. G. Barbour, University of Texas, San

Downloaded from http://jid.oxfordjournals.org/ at University of Rhode Island on April 4, 2015

Proliferative responses of naive splenocytes to Borrelia burgdorferi antigens from mice susceptible (C3H) and resistant (BALB) to Lyme borreliosis were investigated. B. burgdorferi spirochetes and recombinant outer surface proteins, OspA and OspB, were found to induce nonspecific proliferation of naive splenocytes from both strains of mice. Cell purification studies localized nonspecific proliferation to the B cell-enriched fraction. B. burgdorferi, OspA, and OspB were found to induce IgM and IgG synthesis in vitro. The mitogenic effect of B. burgdorferi was dissimilar to that of lipopolysaccharide (LPS), in that B cells from C3H/HeJ mice (LPSunresponsive) responded at levels comparable to those from C3H/HeNCrIBr mice. These results emphasize the need for caution in the study of antigen-specific proliferation for B. burgdorferi.

472

de Souza et al.

resuspended at 10 7 cells/ml of magnetic beads coated with antimouse IgG (Collaborative Research, Bedford, MA). After a further 20-min incubation on wet ice, cells were applied to a BioMag separator (Collaborative Research). Cells not removed were processed for proliferative assays. B cell enrichment (for both BALB and C3H mice) was accomplished by negative selection using monoclonal antibodies TIB 210 and TIB 105 to Ly2, GK1.5 to L3T4, and Y19 to Thyl as previously described [12]. Briefly, cells were incubated with antibody for I h at 4°C and resuspended in RPMI serum-free medium containing Low-Tox-M rabbit complement (Accurate Chemical and Scientific, Westbury, NY) according to the manufacturer's instructions. Resulting viable cells were processed for proliferation assays. Purity of the cell populations was assessed using flow cytometry. Enriched cell populations (5-10 X 106/ml) were stained (4°C for 30 min) with anti-CD3 (C363.2 9B) and anti-Ia antibodies, followed by a second incubation with fluorescein isothiocyanate-labeled goat anti-rat IgG (Hyclone Laboratories, Logan, UT). Fluorescence was observed with a FACS Analyzer I (Becton-Dickinson, Mountain View, CA) at 488 nm. All monoclonal antibodies used in T and B cell enrichment experiments and flow cytometry were provided by K. Bottomly (Howard Hughes Medical Institute, Section of Immunobiology). Cytokine assays. Splenocytes from experimental mice were cultured in the presence of ConA, LPS, B. burgdorferi, recombinant proteins, or diluent and supernatants were harvested at 24 and 48 h for interleukin (IL)-2 and 11-4 assays, respectively. Supernatants were assessed for 11-2 using the CTLL cell line [13] and 11-4 using the CTAS cell line [14]. Supernatants (50 ~l) were added in twofold dilutions from 1:4 to 1:16 to 104 cells. Cells were incubated at 37°C for 24 h, labeled with 1 ~Ci of Tdr/well, and harvested 24 h later. Supernatants inducing proliferation were rescreened in the presence of culture medium, anti-I1-2 (S4B6.34), anti-I1-4 (llBll), orS4B6.34 plus IIBll. Standards used in each assay were 11-2-containing EI-4 cell supernatant and the 11-4-containing HeLa cell supernatant (11-4transfected HeLa cells were provided by T. Honjo, Kyoto University Faculty of Medicine, Japan). Determination of immunoglobulin secretion. Spleen or Benriched cells were cultured at 2 X 106/ml in the presence of diluent, ConA, LPS, B. burgdorferi, OspA, or OspB for 7 days at 37°C. Supernatants were removed and stored at -70°C for testing. Supernatants were diluted twofold beginning at 1:15 and tested for the presence of total IgM and IgG, as well as B. burgdorjeri-specific IgM and IgG. Total IgM and IgG were determined by coating IMMULON 2 96-well polystyrene plates (Dynatech Laboratories, Alexandria, VA) with 100 ~l of goat anti-mouse immunoglobulin ( 1 ~g/ml; Fisher Scientific, Orangeburg, NJ) overnight at 4°C. Wells were blocked with 200 ~l of 3% gelatin in PBS to decrease nonspecific binding and washed twice with PBS containing 0.05% Tween 20. Supernatant dilutions were applied in I OO-~l volumes and incubated at 37°C for 1 h. After three washes, 100 ~l of horseradish peroxidase-conjugated anti-IgM (Tago, Burlingame, CA) or anti-IgG (Kirkegaard & Perry Laboratories, Gaithersburg, MD) were added to each well and the plates incubated at 37°C for 1 h.

Downloaded from http://jid.oxfordjournals.org/ at University of Rhode Island on April 4, 2015

Antonio), cloned, expressed, and purified using similar techniques. Briefly, published nucleotide sequences of each gene [ 10] were used to construct oligonucleotide primers, which were flanked by 5' BamHI and 3' EcoRI restriction enzyme digestion sites to facilitate cloning. B. burgdorferi N40 or B31 genomic DNA (lOng) was used as a template for polymerase chain reaction. Amplified segments were isolated by agarose gel electrophoresis and purified by electroelution on a DEAE membrane. Purified OspA-N40 and OspB-B31 DNA were digested with EcoRI and BamHI, then ligated into plasmid pGEX-2T (Pharmacia Fine Chemicals, Piscataway, NJ) in frame with GT. Escherichia coli strain DH5a was transformed with the plasmids containing either OspA or OspB by electroporation. Recombinant fusion protein production was induced in the log phase of culture by I mM isopropyl-Bvo-thiogalactoside for 3 h. Cells were then washed in PBS, lysed by sonication, and centrifuged at 14,000 g. Supernatants containing GT fusion proteins were placed over a glutathione-sepharose 4B column (Pharmacia), then eluted with 5 mM glutathione. Recombinant GT prepared in the same way as the recombinant B. burgdorferi proteins was used in proliferation assays as a negative control. Detection ofendotoxin. The endotoxin content of the recombinant proteins was determined using the limulus amebocyte lysate (LAL) assay (Whittaker Bioproducts, Walkersville, MD) according to the manufacturer's instructions. This LAL assay detected a minimum of 0.006 ng/rnl E. coli 055:B5 endotoxin. Pyrogen-free water and RPMI complete medium served as negative controls. Spleen cell proliferation assays. Pooled spleens from 3-10 mice were suspended in PBS (5 mljspleen) and processed as described [11]. Briefly, splenocytes were triturated using Ten Broeck tissue grinders, washed, and resuspended in RPMI complete medium. Cells were adjusted to 2 X 10 6 viable cells/ml, and 1OO-~l volumes were added to 96-well cluster dishes. B. burgdorferi, recombinant protein, concanavalin A (ConA; Sigma), or lipopolysaccharide (LPS; Sigma) was added at the time of culture initiation at concentrations of I0 ~g/ml for B. burgdorferi and recombinant proteins, 4 ~g/ml for con A, and 50 ~g/ml for LPS. Polymyxin B (Sigma) was used at a concentration of 30 ~g/ml. Final culture volumes were 200 ~l. Cultures were labeled with 1 ~Ci of[ 3H]thymidine (Tdr, 80 Ci/rnmol; Du Pont NEN, Boston) 18 h before harvesting. Cells were harvested with a semiautomated collector (Cambridge Technology, Watertown, MA), and Tdr incorporation was determined by liquid scintillation spectroscopy. Cells were harvested at 24-h intervals beginning 48 h after culture initiation. Results are expressed as the mean (±SD) counts per minute (cpm) for stimulated cells minus mean cpm for diluent or GT-treated cells (three replicates per treatment). Enrichment for Band T cells. Disaggregated spleen cells from 7-:-10 mice were applied to lymphocyte separation medium (Organon Teknika, Durham, NC). After centrifugation at 1500 g for 15 min, lymphocytes were collected and washed three times in PBS. For T cell enrichment, cells were resuspended to 10 7 cells/ml in monoclonal antibodies to Ia with specificity for b, d, p, q, r, and v (212) for BALB cells and TIB 92 with specificity to la k for C3H mice. Cells were incubated on ice for I hand

lID 1992;165 (March)

lID 1992;165 (March)

B. burgdorferi Stimulation of Lymphocytes

After three washes, a substrate of 3,3',5,5'-tetramethylbenzidine and H 20 2 (Kirkegaard & Perry) was added for 20 min. The reaction was stopped using 100 ~l/well 1 N HC" and the plates were analyzed at 450 nm using an MR 600 microplate reader (Dynatech). B. burgdorferi-specific antibody was assayed using an identical system, except that plates were coated overnight with a preparation of B. burgdorferi N40 antigen at 0.125 #Lg/well and air dried at 37°C [15]. The antibody titer was determined by the highest dilution of supernatant that gave an absorbance reading of ~0.4 at A 4 50 • Statistical analysis. Differences between mean values were analyzed with Student's two-tailed paired t test for treatments within one mouse strain and Student's unpaired t test for comparison of means between mouse strains.

30

473

C3H -.... BALB -a-

A

15

o L.==~::::::::::::::::----, • 1

· 2

50

a.

Splenocytes from both naive C3H and BALB mice proliferated in response to stimulation with B. burgdorferi N40 and recombinant proteins OspA and OspB in a dose-dependent manner as measured at 72 h (figure 1). Cultures treated with 10 ~g/ml GT showed similar proliferation results to diluenttreated splenocytes (4467 ± 188 and 3467 ± 636, respectively). Results using heat-inactivated B. burgdorferi N40 were nearly identical to those using untreated N40 spirochetes (17,819 ± 736 and 17,469 ± 1777, respectively, for naive C3H splenocytes). Subsequent studies used untreated B. burgdorferi and recombinant proteins OspA and OspB at 10 ~g/ml. A consistent finding over four experiments was that the proliferative response to strain N40 was similar for splenocytes from C3H and BALB mice (figure 1), but the response to recombinant proteins OspA and OspB was higher for C3H cells (figure 1). Extending the study to different strains of B. burgdorferi indicated that the stimulatory effect was not confined to strain N40 (figure 2). Kinetic studies for the time intervals examined indicated that the response was maximal at 48 h (figures 2 and 3). Subsequent studies measured Tdr uptake at 48 h. Cytokine analyses using ConA-stimulated splenocyte supernatants as a positive control revealed that neither B. burgdorferi nor its recombinant proteins stimulated either 11-2 or 11-4 production (data not shown), implying that the response was not T cell-mediated. Flow cytometry revealed that the T cell-enriched populations for BALB and C3H mice contained