JOURNAL OF CLINICAL MICROBIOLOGY,0095-1137/98/$04.0010

Nov. 1998, p. 3278–3284 Vol. 36, No. 11

Copyright © 1998, American Society for Microbiology. All Rights Reserved.

Characterization of Monoclonal Antibodies to the 44-KilodaltonMajor Outer Membrane Protein of the Human Granulocytic

Ehrlichiosis AgentHYUNG-YONG KIM AND YASUKO RIKIHISA*

Department of Veterinary Biosciences, College of Veterinary Medicine, The Ohio State University,Columbus, Ohio 43210-1093

Received 23 February 1998/Returned for modification 14 May 1998/Accepted 4 August 1998

The major outer membrane proteins (OMPs) of the human granulocytic ehrlichiosis (HGE) agent, withmolecular sizes of 44 to 47 kDa, are immunodominant antigens in human infection. Monoclonal antibodies(MAbs) to the OMPs were made by immunizing BALB/c mice with the purified HGE agent and then by fusingspleen cells with myeloma cells. The immunologic specificities of three MAbs (3E65, 5C11, and 5D13) wereexamined with five human HGE agent isolates and one tick isolate. By Western blot analysis, all three MAbsrecognized the HGE agent but not Ehrlichia chaffeensis, Ehrlichia sennetsu, Ehrlichia canis, or their host cells.MAb 3E65 reacted with a 44-kDa protein in the homologous human isolate but not in the remaining fiveisolates. The two remaining MAbs recognized proteins with molecular sizes of 44 to 47 kDa in all six isolates.Western blot results with the OMP fraction of the six isolates were consistent with results with the whole HGEagent. Immunofluorescent-antibody staining and immunogold labeling with these MAbs showed that theseantigens were primarily present on the membrane of the HGE agent. MAbs 5C11 and 5D13 recognized therecombinant 44-kDa protein by Western immunoblot analysis, but MAb 3E65 did not. Passive immunizationwith MAb 3E65 was more effective in protecting mice from HGE agent infection than with MAbs 5C11 and5D13. These MAbs would be useful for analyzing the role of the major OMP antigens in HGE agent infectionand for serodiagnosis.

Human granulocytic ehrlichiosis (HGE) is an emerging tick-borne zoonosis characterized by fever, headache, myalgia, el-evated liver enzyme (serum aminotransferase) levels, leukope-nia, anemia, thrombocytopenia, and elevated C-reactiveprotein levels (2). The first 12 reported cases were in Wiscon-sin and Minnesota between 1990 and 1993. Subsequently,more than 400 cases have been identified in the upper Midwestand in the northeastern United States (1, 2, 4, 13, 15), andevidence of HGE has been also reported in Europe (3, 6, 10).The etiologic agent, currently referred to as the HGE agent, isan obligate intracellular gram-negative bacterium that repli-cates in the membrane-bound vacuoles of granulocytes. It isclosely related to the Ehrlichia equi-E. phagocytophila group ofthe tribe Ehrlichieae (4). The pathogen is transmitted mainly bythe deer tick, Ixodes scapularis, which is also the vector ofBorrelia burgdorferi, the agent of Lyme disease (9, 13). Humancoinfection with the HGE agent and B. burgdorferi has beenpreviously reported (8).

Diagnosis is often made by a retrospective immunofluores-cent antibody (IFA) test with the HGE agent or E. equi antigen(1–4, 6, 8, 11, 13, 14–17). Recently, we cloned and expressed a44-kDa major antigen of the HGE agent and demonstratedthat this recombinant antigen (rP44) may be more specific forserodiagnosis of HGE than whole organism-infected cells dueto the absence of heat shock proteins or other antigenicallycross-reactive proteins (17). PCR amplification of the 16SrRNA gene fragment of the HGE agent from peripheral bloodhas been gaining acceptance as a sensitive test at acute stagesof HGE. Microscopic examination of Romanowsky-stained pe-

ripheral blood smears may reveal the presence of ehrlichialmorulae in the neutrophils. Using five isolates of HGE agentsand a tick isolate, we reported that the major outer membraneproteins (OMPs) of the HGE agent, with molecular sizes be-tween 43 and 49 kDa, are immunodominant antigens in humaninfection (16, 17). Western blot analysis also revealed varia-tions in numbers and molecular sizes of the major antigenicOMPs of the six isolates. Since polyclonal antisera were usedfor that study, it was unclear whether the major OMPs ofsimilar sizes are common antigens among the six isolates.Monoclonal antibodies (MAbs) against a 44-kDa OMP of theHGE agent might be useful for clarifying the relationships ofthe OMPs of the HGE agent, for analyzing the antigenicepitopes, for understanding the immune responses of HGEagent infection, and for serodiagnosis. In this study, MAbsagainst a 44-kDa protein of HGE agent isolate 13 were pro-duced and characterized by using five HGE agent isolates, atick isolate, and rP44.

MATERIALS AND METHODS

Cultures and media. Five isolates of the HGE agent (isolates 13, 2, 11, 3, and6) (11, 16) and a tick isolate (USG) (5) were propagated in the human promy-elocytic leukemia cell line HL-60 (American Type Culture Collection [ATCC],Manassas, Va.) in RPMI 1640 medium supplemented with 5% fetal bovineserum (FBS) (Atlanta Biologicals, Norcross, Ga.), 1% minimal essential medium(MEM) nonessential amino acid mixture (GIBCO, Grand Island, N.Y.), 1 mMMEM sodium pyruvate (GIBCO), and 2 mM L-glutamine (GIBCO) (11, 16).Ehrlichia chaffeensis in THP-1 cells (ATCC) and Ehrlichia sennetsu in P388D1cells (ATCC) were maintained in RPMI 1640 medium supplemented with 10%FBS and 2 mM L-glutamine, while Ehrlichia canis in DH82 cells (12) and my-eloma cells (SP2/0-Ag14; ATCC) were maintained in Dulbecco’s modified Ea-gle’s medium (DMEM) (GIBCO) supplemented with 10% FBS and 2 mML-glutamine. All cultures were incubated at 37°C in a humidified 5% CO2–95%air atmosphere. Infectivities of all Ehrlichia spp. were determined by Diff-Quik(modified Giemsa; Baxter Scientific Products, Obetz, Ohio) staining, as de-scribed previously (11, 12).

* Corresponding author. Mailing address: Department of Veteri-nary Biosciences, College of Veterinary Medicine, The Ohio StateUniversity, 1925 Coffey Rd., Columbus, OH 43210-1093. Phone: (614)292-5661. Fax: (614) 292-6473. E-mail: rikihisa.1@osu.edu.

3278

on Septem

ber 24, 2020 by guesthttp://jcm

.asm.org/

Dow

nloaded from

Preparation of purified ehrlichiae and the outer membrane fraction. Infectedcells were weakly sonicated under predetermined conditions (16) to lyse infectedcells with minimum damage to ehrlichiae. After centrifugation to remove un-broken cells and nuclei of the host cells, the supernatants containing freedehrlichiae were size fractionated by Sephacryl S-1000 (Pharmacia, Uppsala,Sweden) chromatography, as previously described (12, 16). For sodium dodecylsulfate-polyacrylamide gel electrophoresis (SDS-PAGE), 25 mg of protein fromeach purified organism was aliquoted into 5 ml of 10 mM sodium phosphatebuffer (4 mM NaH2PO4 and 6 mM Na2HPO4, pH 7.4) containing 1 mM phe-nylmethylsulfonyl fluoride (PMSF) (Sigma, St. Louis, Mo.) and frozen at 285°C.The outer membrane fractions of the six isolates were prepared by a modifiedSarkosyl differential solubilization method (16). The outer membrane fractionwas washed once with 50 ml of 0.2% Sarkosyl in 10 mM sodium phosphate bufferby centrifugation at 10,000 3 g for 1 h. The final pellets were resuspended in 10mM sodium phosphate buffer containing 1 mM PMSF and frozen at 285°C.

Immunization. Eight 6-week-old male BALB/c mice (Harlan Sprague-Dawley,Indianapolis, Ind.) were injected intraperitoneally with a 0.2-ml mixture of equalvolumes of purified HGE agent 13 (200 mg/mouse) in PBS (2.7 mM KCl, 1.8 mMKH2PO4, 137 mM NaCl, and 10 mM Na2HPO4 [pH 7.4]) and Freund’s completeadjuvant (Sigma). Two weeks after the first immunization, each mouse wasboosted subcutaneously with the 0.2-ml mixture of equal volumes of 100 mg ofthe purified HGE agent in PBS and Freund’s incomplete adjuvant (Sigma). Themice were intraperitoneally injected with purified HGE agent in PBS (100mg/mouse) without adjuvant on the 7th day after the second immunization.Between 35 and 40 days after the first immunization, 0.5 ml of blood wascollected from the retroorbital venous plexus of immunized mice. Serum wascollected and antibody titers were determined by an IFA test using antigen slidesof HGE agent 13 prepared as previously described (11, 16). Four days beforesacrifice, the purified HGE agent (100 mg/mouse in 0.1 ml of PBS) was intrave-nously injected.

Hybridoma production. When antibody titers were greater than 1:2,560, micewere sacrificed and spleen cells were fused with myeloma cells in either 50%polyethylene glycol 4000 (GIBCO) or 40% polyethylene glycol 4000 containing7.6% dimethyl sulfoxide (Sigma) in PBS, according to the procedure describedby Goding (7). Two milliliters of the fusion mixture (105 myeloma cells and 3 3105 to 4 3 105 spleen cells) in complete DMEM (cDMEM) (DMEM containing20% FBS, 1 mM MEM sodium pyruvate, 1% MEM nonessential amino acidmixture, 2 mM L-glutamine, and a 1:100 dilution of antibiotic-antimycotic mix-ture [GIBCO] [penicillin G, 10,000 U/ml; streptomycin sulfate, 10,000 mg/ml;amphotericin B, 25 mg/ml]) containing HAT supplement (diluted 1:100 from thestock solution of 10 mM Na-hypoxanthine, 40 mM aminopterin, and 1.6 mMthymidine [GIBCO]) was aliquoted into each well of 24-well culture plates withspleen feeder cells. All cultures were incubated at 37°C in a humidified 5%

CO2–95% air atmosphere. When colonies became visible, the medium was grad-ually replaced with cDMEM containing 100 mM Na-hypoxanthine (GIBCO) and16 mM thymidine (Sigma). For subcultures of positive clones, cDMEM was used.

Screening of candidate hybridoma clones. To screen clones producing MAbsagainst the HGE agent, both the IFA test and enzyme-linked immunosorbentassay (ELISA) were used. For the IFA test, 10 ml of undiluted supernatant wasadded to each well of HGE agent 13 slides prepared as previously described (11,16) and incubated at 37°C for 90 min. After a wash with 23 PBS (19 mMK2HPO4, 12 mM KH2PO4, and 300 mM NaCl [pH 7.4]) containing 0.05% Tween20 (Sigma), 10 ml of fluorescein isothiocyanate-conjugated rabbit anti-mouseimmunoglobulin G (IgG) 1 IgM (Jackson ImmunoResearch Laboratories, WestGrove, Pa.) at a 1:100 dilution was added and incubated at 37°C for 90 min. Allslides were counterstained with 0.1% Evans blue in 23 PBS and examined witha Nikon (Garden City, N.Y.) Microphot-FX epifluorescence microscope. Forphotography of labeled cells, infected cells were cytocentrifuged and fixed withDiff-Quik fixative containing methanol and incubated with antibodies as de-scribed above.

For ELISA, 96-well plates were coated with purified HGE agent 13 at 1 mg ofprotein/well in 50 mM Na-bicarbonate buffer (pH 9.6) and blocked with 5%nonfat dry milk in the same buffer. One hundred microliters of undiluted hy-bridoma supernatant was added to each well, including 100 ml each of both thepositive (polyclonal mouse anti-HGE agent 13) and negative (normal BALB/cmouse serum) controls at 1:10 dilutions in PBS containing 5% nonfat dry milk.All plates were incubated at 37°C for 1 h and rinsed three times with PBScontaining 0.05% Tween 20. Peroxidase-conjugated goat anti-mouse immuno-globulins (IgG 1 IgA 1 IgM) (ICN Pharmaceuticals, Aurora, Ohio) at a 1:500dilution in PBS containing 5% nonfat dry milk were added and incubated at 37°Cfor 1 h. ABTS [0.15% 2,29-azino-di-(3-ethylbenzthiazoline sulfonate)] (Sigma) in0.1 M sodium citrate buffer (pH 4.2) and 0.03% H2O2 were added and incubatedat room temperature for 15 min. The supernatants which had absorbances at 405nm of greater than 0.8 were again confirmed by the IFA test. The cutoff value of0.8 was chosen because negative control wells had absorbances of less than 0.5and 100% of wells with absorbances of above 1.0 were IFA positive; we chose 0.8to cover some weakly positive clones. Finally, all hybridomas identified as posi-tive clones by both tests were subcultured into six-well culture plates. When cellswere confluent, these clones were frozen in 90% FBS and 10% dimethyl sulfox-ide at 285°C. The wells identified as positive clones by primary screening weresubcloned three times via a limiting dilution technique (7).

Production of ascitic fluids containing MAbs. For rapid production of asciticfluids containing MAbs, 3 to 5 days prior to hybridoma injection, 0.5 ml of theinflammatory agent pristane (Sigma) was intraperitoneally injected into 6-week-old male BALB/c mice (10 mice per clone). At 12 days after intraperitonealinjection of actively growing hybridomas (3E65, 5C11, and 5D13) into the mice

FIG. 1. Western immunoblot analysis of HGE agent 13, E. chaffeensis, E. canis, E. sennetsu, and their uninfected host cells with three MAbs. (A) Proteins wereseparated on a 10% polyacrylamide gel and stained with Coomassie blue. (B) For Western blot analysis, hybridoma (3E65, 5C11, and 5D13) culture supernatants wereused at 1:10 dilutions. Molecular weight (MW) standards were from Bio-Rad.

TABLE 1. Characteristics of MAbs against the HGE agent

MAba IsotypeProtein(s) (kDa) recognized in HGE agent isolateb:

13 2 11 3 6 USG

3E65 IgG1 44, 110 110 110 110 110 1105C11 IgG2b 44, 89 41, 46, 47 47 47 45 44, 45, 465D13 IgG2b 44 41, 46, 47 47 47 45 44, 45, 46

a Subcloned three times by limiting dilution techniques.b Major antigens of the six HGE agent isolates are shown in boldface type.

VOL. 36, 1998 MAbs TO HGE AGENT 44-kDa PROTEIN 3279

on Septem

ber 24, 2020 by guesthttp://jcm

.asm.org/

Dow

nloaded from

(2 3 106 to 3 3 106 cells/mouse in 0.2 ml of sterile PBS), ascitic fluids containingMAbs were harvested and centrifuged at 10,000 3 g for 10 min. After the asciticfluid for each clone was pooled and inactivated by incubation at 56°C for 30 min,antibody titers were determined by the IFA test with an antigen slide of HGEagent 13. Ascitic fluid containing MAbs was diluted at 1:10, filtered through a0.45-mm-pore-size filter, and frozen at 220 or 285°C.

Isotyping of MAbs. Using either ascitic fluids or culture supernatants, isotypesof MAbs were determined by the Mouse Monoclonal Typing Kit (The BindingSite, Birmingham, United Kingdom), which is based on the Ouchterlony immu-nodiffusion technique, and by the Mouse MAb ID/SP Kit (Zymed Laboratories,Inc., South San Francisco, Calif.), which is based on a streptavidin-biotin ampli-fication system in an HGE agent antigen-dependent ELISA.

SDS-PAGE and Western blot analysis. Twenty-five micrograms of proteins ofpurified isolates (isolates 13, 2, 11, 3, 6, and USG) and the OMP fraction of eachisolate; purified E. chaffeensis, E. sennetsu, and E. canis; uninfected host cells(HL-60, THP-1, and DH82); and 10 mg of rP44, affinity purified as describedelsewhere (17), were dissolved in sample buffer (5% 2-mercaptoethanol, 10%glycerol, 2% SDS, and 0.08% bromophenol blue in 62.5 mM Tris buffer [pH6.8]). To examine whether the antigenic epitope is trypsin sensitive, 100 mg ofpurified HGE agent was incubated in 0.5 ml of 0.25% trypsin (Sigma) in PBS for15 min at room temperature. Samples, including 10 ml of diluted (1:20) broad-range molecular weight standards (Bio-Rad, Hercules, Calif.), were boiled at

100°C for 5 min. SDS-PAGE and Western blotting were performed as describedelsewhere (12, 16). Protein blots were immersed in blocking buffer (5% nonfatdry milk in 23 PBS) at 4°C overnight and incubated with culture supernatants (ata 1:10 dilution in blocking buffer) or ascitic fluids (at a 1:50 dilution) of hybrid-oma clones 3E65, 5C11, or 5D13 at room temperature for 2 h. After three rinseswith TNTT buffer (50 mM Tris-HCl, 150 mM NaCl, 0.02% Tween 20, and 0.01%thimerosal [Sigma] [pH 7.4]), the blots were incubated at room temperature for2 h with peroxidase-conjugated goat anti-mouse immunoglobulins (IgG 1 IgA 1IgM) (ICN Pharmaceuticals) at a 1:2,000 dilution in blocking buffer. The blotswere washed three times with TNTT buffer for 5 min each. The peroxidase-positive bands were detected by immersing the blots in a developing solution (73mM sodium acetate, pH 6.2) containing 0.3% diaminobenzidine tetrahydrochlo-ride (Nacalai Tesque, Inc., Kyoto, Japan) and 0.04% H2O2 at room temperaturefor 5 min. The enzyme reaction was terminated by washing the blots in 0.1 MH2SO4.

Immunogold labeling. HGE agent 13-infected cells (2 3 107) were fixed in afixative (2.5% paraformaldehyde, 0.5% glutaraldehyde, 0.03% trinitrophenol,and 0.03% CaCl2 in 0.05 M cacodylate buffer [pH 7.4]) at room temperature for1 h and en bloc stained with 1% uranyl acetate in 0.1 M maleate buffer (pH 5.2)at room temperature for 1 h. The specimen was dehydrated in a series of gradedethanols (50 to 90%) and embedded in LR gold (Polysciences, Inc., Warrington,Pa.) which had been polymerized at 225°C in a low-temperature UV curing unit(Polysciences). Ultrathin sections (800 Å) were cut with a diamond knife andmounted on 300-mesh nickel grids. The grids were incubated for 1 h with eachMAb in ascitic fluid or positive and negative control sera diluted 1:50 withPBS-GT (0.1% gelatin and 0.01% Tween 20 in PBS, pH 7.4). The grids wererinsed in PBS-GT and incubated at room temperature for 1 h with goat anti-mouse IgG 1 IgM conjugated with 10-nm gold particles (Amersham, ArlingtonHeights, Ill.) diluted 1:20 with PBS-GT. The grids were gently rinsed in PBS-GTand distilled water. The sections were stained with 2% uranyl acetate in waterand observed with a Philips 300 transmission electron microscope at 60 kV.

Mouse protection assay. Five groups of three mice each (C3H/HeN, 4-week-old females; Harlan Sprague-Dawley) were inoculated intraperitoneally withHGE agent 13-infected HL-60 cells (greater than 90% infected cells, 106 cells/mouse) which had been preincubated at 37°C for 30 min with 0.3 ml of each MAbin ascitic fluid or positive or negative control mouse sera. MAbs and the positivecontrol serum were diluted in PBS so that the final IFA titer against HGE agent13 was 1:2,000. One day after infection, mice were injected intraperitoneally with0.3 ml of each MAb or control sera. At 5 days post-HGE agent inoculation, micewere sacrificed for collection of blood specimens. DNA was extracted from theblood with the QIAamp blood kit (Qiagen Inc., Chatsworth, Calif.). A pair ofprimers, 497-521 (59-TAGGCGGTTCGGTAAGTTAAAG-39) and 747-727 (59-GCACTCATCGTTTACAGCGTG-39) (13), was used for amplification in athermal cycler (model 480; Perkin-Elmer) with 5 min of denaturation at 94°C,followed by 40 cycles of denaturation at 94°C, annealing at 57°C, and extensionat 72°C for 1 min each. After the last cycle, extension was continued for 7 min at72°C. Each 50-ml PCR mixture contained 1 mg of template DNA, 5 ml of 103reaction buffer, 0.2 mM concentrations of deoxynucleoside triphosphates, 2.5mM MgCl2, 1.25 U of Taq polymerase, and 16 pmol of each primer. PCRproducts (10 ml each) were electrophoresed in 1.5% agarose gels containing 0.5mg of ethidium bromide at 95 V for 1 h and photographed under UV illumina-tion with a gel video system (Gel Print 2000i; Biophotonics Corporation, AnnArbor, Mich.).

FIG. 2. Effect of PMSF on the molecular size of major antigens recognizedby MAb 5C11. Uninfected HL-60 cells, purified HGE agent 13, 1 mM PMSF-treated HGE agent, and the OMP fraction of the HGE agent were separated ona 10% polyacrylamide gel, transferred to a nitrocellulose membrane, and incu-bated with a 1:50 dilution of MAb 5C11 in ascitic fluid. Molecular weight (MW)standards were from Bio-Rad.

FIG. 3. Comparisons of six purified HGE agent isolates with three MAbs on a 10% polyacrylamide gel (A) and by Western blot analysis (B). (A) Proteins frompurified HGE agent isolates (13, 2, 11, 3, 6, and USG) were separated on a 10% polyacrylamide gel and stained with Coomassie blue. (B) For Western blot analysis,hybridoma (3E65, 5C11, and 5D13) culture supernatants were used at 1:10 dilutions. Molecular weight (MW) standards were from Bio-Rad.

3280 KIM AND RIKIHISA J. CLIN. MICROBIOL.

on Septem

ber 24, 2020 by guesthttp://jcm

.asm.org/

Dow

nloaded from

RESULTS

Production of hybridomas and MAbs. By both IFA test andELISA, 16.7% of colonies (127 of 761) were positive at the firstscreening. Subsequently, 39 hybridomas with strong antibodyproduction were subcloned three times and 20 hybridomaswere identified as positive by Western blot analysis. Threeclones (3E65, 5C11, and 5D13) which reacted to the approxi-mately 44-kDa major antigen of HGE agent 13 were selectedfor this study. Ascitic fluid was produced with each MAb, andMAb isotypes were determined (Table 1). IFA titers of asciticfluids and culture supernatants of the three MAbs used were1:20,480 and 1:320, respectively. These three clones have beenstable for more than 1 year.

Western immunoblotting. All three MAbs recognized theHGE agent but not E. chaffeensis, E. canis, E. sennetsu, or theiruninfected host cells (HL-60, THP-1, and DH82) (Fig. 1B).Although band intensities and molecular sizes were slightlyvariable among HGE isolates, the major antigens recognizedby the three MAbs were around 44 kDa (Fig. 1B and Table 1).

In our previous study, with polyclonal antibodies, multiplebands of approximately 44 kDa were seen in whole HGE agentantigen, but fewer bands were seen in the OMP fraction (16).Since we had used 1 mM PMSF for preparation of the OMPfraction of HGE agents but not for purification of whole HGEagent organisms, we examined the effect of 1 mM PMSF on themolecular size of proteins recognized with MAb 5C11 by West-ern blot analysis. We found that in the presence of PMSF, thedensity of the 47-kDa band was reduced and that of the 44-kDaband became thicker, suggesting that a 47-kDa protein is theprecursor of the 44-kDa protein and that a protease whichcleaves the 47-kDa protein to a 44-kDa protein might be de-graded in the absence of PMSF (Fig. 2). Therefore, 1 mMPMSF was added to all specimens throughout this study. Nodifference was detected if PMSF was added at the beginningor just after purification of the HGE agent, suggesting thatthis change primarily occurs after purification of the HGEagent.

Trypsin treatment of HGE agent 13 completely destroyedthe reactivity of all three MAbs, indicating that these MAbsrecognize trypsin-sensitive epitopes (data not shown).

When the six isolates (13, 2, 11, 3, 6, and USG) were com-pared, MAb 3E65 reacted strongly with a 44-kDa protein inHGE agent 13 but not in the remaining five isolates. It alsoreacted weakly with a 110-kDa protein in all isolates (Fig. 3B).

These two proteins appear to have a common epitope in HGEagent 13, since it is unlikely that, after subcloning three timesby the limiting dilution technique, two hybridomas would becocloned in MAb 3E65. Although MAbs 5C11 and 5D13 werederived from different master plates prepared from the samemouse spleen, they reacted in almost identical manners byWestern blot analysis. Both recognized proteins that are ap-proximately 44 kDa in all six isolates (Fig. 3B). The molecularsizes and numbers of the major antigens recognized by MAbs5C11 and 5D13 varied slightly, from 44 to 47 kDa, among thesix isolates (Fig. 3B and Table 1). MAbs 5C11 and 5D13recognized antigens of the same molecular sizes both in wholeorganisms and in the OMP fractions of the six isolates (Fig.4B). Thus, all of these antigenic proteins common within eachisolate and among the six isolates appear to be present in theouter membrane.

MAb 3E65 did not recognize the rP44 antigen, while MAbs5C11 and 5D13 did (Fig. 5). These results show that theepitope of the 44-kDa protein recognized by MAb 3E65 isdifferent from those recognized by MAbs 5C11 and 5D13.

Immunofluorescence and immunogold labeling. Polyclonalanti-HGE agent mouse serum strongly labeled entire HGE

FIG. 4. Western blot analysis of the OMP fractions from six HGE agent isolates with three MAbs. (A) The OMP fractions from the six HGE agent isolates (13,2, 11, 3, 6, and USG) were separated on a 10% polyacrylamide gel and stained with Coomassie blue. (B) For Western blot analysis, hybridoma (3E65, 5C11, and 5D13)culture supernatants were used at 1:10 dilutions. Molecular weight (MW) standards were from Bio-Rad.

FIG. 5. Western blot analysis of rP44 with three MAbs. (A) Purified HGEagent 13 and affinity-purified rP44 were separated on a 10% polyacrylamide geland stained with Coomassie blue. (B) For Western blot analysis, hybridoma(3E65, 5C11, and 5D13) culture supernatants were used at 1:10 dilutions. Mo-lecular weight (MW) standards were from Bio-Rad.

VOL. 36, 1998 MAbs TO HGE AGENT 44-kDa PROTEIN 3281

on Septem

ber 24, 2020 by guesthttp://jcm

.asm.org/

Dow

nloaded from

agent 13 organisms, while the negative control mouse serumdid not label any structure at all in the IFA test. All threeMAbs labeled both intracellular and extracellular HGE agentsbound to HL-60 cells in a ring-like pattern (Fig. 6). Immuno-

gold labeling results show that a 44-kDa OMP antigen is con-centrated on the membrane of the HGE agent; with polyclonalantiserum and MAb 3E65, some labeling appeared to bepresent on inclusion membranes, too (Fig. 7).

FIG. 6. IFA staining of HGE agent 13 in HL-60 cells with three MAbs. Infected HL-60 cells were fixed in Diff-Quik fixative and incubated with MAbs (3E65, 5C11,and 5D13) in ascitic fluid at 1:100 dilutions and with positive and negative control (NS) mouse sera at 1:50 dilutions. Note the ring-like labeling of the HGE agent withall three MAbs. Magnification, 31,600.

FIG. 7. Transmission electron micrograph of HGE agent 13 in HL-60 cells immunogold labeled with MAbs. Infected cells were embedded in LR Gold, and ultrathinsections on grids were incubated with MAbs (3E65 and 5C11) and positive and negative control mouse sera and with goat anti-mouse IgG 1 IgM conjugated with 10-nmgold particles (Amersham). Bars, 0.4 mm.

3282 KIM AND RIKIHISA J. CLIN. MICROBIOL.

on Septem

ber 24, 2020 by guesthttp://jcm

.asm.org/

Dow

nloaded from

Mouse protection test. Passive immunization with polyclonalmouse anti-HGE agent serum and MAb 3E65 consistentlyshowed superior protection, with 67% (2 of 3 mice) protectedfrom HGE agent 13 infection; with MAbs 5C11 and 5D13,33% (1 of 3 mice) were protected (Fig. 8). Although the micedid not show any clinical signs during the 5-day infection pe-riod, all negative control mice were positive for HGE agentDNA by PCR, indicating subclinical infection. The detectionlimit of PCR was 0.15 pg of DNA in blood.

DISCUSSIONOur previous study showed that multiple bands between 43

and 49 kDa of HGE agents are recognized in both whole cellsand OMP fractions by polyclonal antisera against the six strainsof the HGE agent. We also found that although these antigenicepitopes were cross-reactive with antisera to other isolates, theHGE agent has a strain polymorphism in its major antigenicproteins (16). The present study, using MAbs against OMPs ofthe HGE agent, suggests two major reasons for the presence ofthese multiple major antigenic proteins. One is the degrada-tion of a protease which probably cleaves a 47-kDa protein toa 44-kDa protein during HGE agent storage or during SDS-PAGE. One possibility is that this 3-kDa molecular size dif-ference is due to the cleavage of an N-terminal signal peptide,termed a leader sequence. Our DNA sequence data of the44-kDa major antigen gene shows an amino acid sequence(VRA) recognizable by signal peptidases, and the chemicallydetermined N-terminal amino acid sequence showed that thissite is actually cleaved in a native mature 44-kDa protein (17).The bacterial signal peptidase is not inhibited by PMSF (14,18). Since PMSF treatment of the HGE agent prevented adegradation-induced artifact of multiple bands of major anti-gens, we recommend the use of PMSF for HGE agent SDS-PAGE and Western blot analysis. Although there is no N-glycosylation site based on the predicted amino acid sequencefrom the 44-kDa protein gene (17), whether this protein is Oglycosylated or whether any glycosylation may contribute tomultiple bands is unknown. The second and more significantreason for the presence of multiple bands may be the presence

of multiple cross-reactive major antigenic proteins of differentmolecular sizes, since even in the presence of PMSF thesemultiple bands were seen in some isolates. This reason is sup-ported by our recent study which revealed the presence ofmultiple homologous gene copies of the major 44-kDa OMPsin the HGE agent genome (17).

Our rP44 antigen approximately corresponds to the N-ter-minal half of the 44-kDa protein and was commonly recog-nized by all HGE patient sera tested and by anti-E. equi serum(17). Therefore, MAbs 5C11 and 5D13 appear to recognize anantigenic epitope present at the N-terminal half of the 44-kDaprotein. Since MAb 3E65 did not recognize rP44, it may bereacting to the C-terminal half of the 44-kDa protein. If this isthe case, the C-terminal half of the 44-kDa protein of HGEagent 13 may contain an antigenic epitope in common with the110-kDa protein, but homologous 44-kDa proteins of the re-maining isolates may lack this epitope. MAbs 5C11 and 5D13appear to behave in almost identical manners. An additionalstudy is required to determine whether the two MAbs recog-nize the same epitope.

The three MAbs protected mice from HGE agent 13 infec-tion, suggesting that the 44-kDa protein has a significant role inestablishing HGE agent infection in mice. MAb 3E65 is aseffective as polyclonal antiserum in protecting mice from HGEagent infection and is more effective than MAbs 5C11 or 5D13,suggesting that the epitope recognized by MAb 3E65 may bemore accessible to the antibody or may have a more criticalrole in HGE agent infection than those recognized by the othertwo MAbs. These MAbs may serve as useful reagents in ana-lyzing the interaction of the HGE agent with host cells and mayalso be useful for purifying major OMPs and in HGE serodi-agnosis.

ACKNOWLEDGMENTS

This research was supported by grant AI40934 from the NationalInstitutes of Health.

We thank Evelyn Handley, of the EM laboratory of VeterinaryBiosciences, The Ohio State University, for technical assistance. Ourcolleagues Norio Ohashi and Ning Zhi are appreciated for helpful

FIG. 8. PCR detection of the HGE agent 16S rRNA gene in the blood of HGE agent-inoculated mice passively immunized with MAbs. Three mice each wereinoculated with HGE agent 13-infected HL-60 cells and each MAb or positive or negative control mouse serum. Each group of mice was inoculated again on the 2ndday with the same antibody. DNAs were prepared from the blood of all mice at 5 days post-HGE agent inoculation. One microgram of DNA was used as the templatefor PCR. As a positive control, 10 pg of DNA extracted from purified HGE agent 13 was used, while distilled water without template was used as a negative control.The arrow shows the HGE agent-specific 16S rRNA gene fragment (287 bp) obtained by PCR amplification. N1 to N3, negative control mouse serum-treated mice;P1 to P3, positive control (polyclonal anti-HGE agent) mouse serum-treated mice; E1 to E3, MAb 3E65-treated mice; C1 to C3, MAb 5C11-treated mice; D1 to D3:MAb 5D13-treated mice. Numbers on the left indicate molecular sizes of fX174 RF DNA/HaeIII fragments (GIBCO). The experiment was repeated twice.

VOL. 36, 1998 MAbs TO HGE AGENT 44-kDa PROTEIN 3283

on Septem

ber 24, 2020 by guesthttp://jcm

.asm.org/

Dow

nloaded from

advice in purifying rP44. We also thank Lawrence Dearth and Ray-mond Mankoski for editorial assistance.

REFERENCES

1. Aguero-Rosenfeld, M. E., H. W. Horowitz, G. P. Wormser, D. F. McKenna,J. Nowakowski, J. Munoz, and J. S. Dumler. 1996. Human granulocyticehrlichiosis: a case series from a medical center in New York State. Ann.Intern. Med. 125:904–908.

2. Bakken, J. S., J. Krueth, C. Wilson-Nordskog, R. L. Tilden, K. Asanovich,and J. S. Dumler. 1996. Clinical and laboratory characteristics of humangranulocytic ehrlichiosis. JAMA 275:199–205.

3. Bakken, J. S., J. Krueth, R. L. Tilden, J. S. Dumler, and R. E. Kristansen.1997. Serologic evidence of human granulocytic ehrlichiosis in Norway. Eur.J. Clin. Microbiol. Infect. Dis. 15:829–832.

4. Chen, S. M., J. S. Dumler, J. S. Bakken, and D. H. Walker. 1994. Identifi-cation of a granulocytotrophic Ehrlichia species as the etiological agent ofhuman disease. J. Clin. Microbiol. 32:589–595.

5. Coughlin, R. T., C. Gingrich-Baker, T. Mather, and D. Fish. 1996. Trans-mission, isolation, and cultivation of granulocytic Ehrlichia resulting frominfection of dogs by adult Ixodes scapularis collected from Eastern UnitedStates, abstr. 52. In Abstracts of the 12th Sesquiannual Meeting of theAmerican Society for Rickettsiology and Rickettsial Disease. American So-ciety for Rickettsiology and Rickettsial Disease, Albany, N.Y.

6. Dumler, J. S., L. Dotevall, R. Gustafson, and M. Granstrom. 1997. A pop-ulation-based seroepidemiologic study of human granulocytic ehrlichiosisand Lyme borreliosis on the west coast of Sweden. J. Infect. Dis. 175:720–722.

7. Goding, J. W. 1996. Monoclonal antibodies: principles and practice, p. 116–191. Academic Press, San Diego, Calif.

8. Nadelman, R. B., H. W. Horowitz, T. C. Hsieh, J. W. Wu, M. E. Aguero-Rosenfeld, I. Schwartz, J. Nowakowski, S. Varde, and G. P. Wormser. 1997.Simultaneous human granulocytic ehrlichiosis and Lyme borreliosis. N. Engl.J. Med. 337:27–30.

9. Pancholi, P., C. P. Kolbert, P. D. Mitchell, K. D. Reed, Jr., J. S. Dumler, J. S.

Bakken, S. R. Telford III, and D. H. Persing. 1995. Ixodes dammini as apotential vector of human granulocytic ehrlichiosis. J. Infect. Dis. 172:1007–1012.

10. Petrovec, M., S. L. Furlan, T. A. Zupanc, F. Strle, P. Brouqui, V. Roux, andJ. S. Dumler. 1997. Human disease in Europe caused by a granulocyticEhrlichia species. J. Clin. Microbiol. 35:1556–1559.

11. Rikihisa, Y., N. Zhi, G. P. Wormser, B. Wen, H. W. Horowitz, and K. E.Hechemy. 1997. Ultrastructural and antigenic characterization of a granulo-cytic ehrlichiosis agent directly isolated and stably cultivated from a patientin New York State. J. Infect. Dis. 175:210–213.

12. Rikihisa, Y. 1991. Cross-reacting antigens between Neorickettsia helmintho-eca and Ehrlichia spp. shown by immunofluorescence and Western immu-noblotting. J. Clin. Microbiol. 29:2024–2029.

13. Telford, S. R., III, J. E. Dawson, P. Katavolos, C. K. Warner, C. P. Kolbert,and D. H. Persing. 1996. Perpetuation of the agent of human granulocyticehrlichiosis in a deer tick-rodent cycle. Proc. Natl. Acad. Sci. USA 93:6209–6214.

14. Tschantz, W. R., and R. E. Dalbey. 1994. Bacterial leader peptidase 1.Methods Enzymol. 244:285–301.

15. Wormser, G., D. McKenna, M. Aguero-Rosenfeld, et al. 1995. Human gran-ulocytic ehrlichiosis—New York. Morbid. Mortal. Weekly Rep. 44:593–595.

16. Zhi, N., Y. Rikihisa, H. Y. Kim, G. P. Wormser, and H. W. Horowitz. 1997.Comparison of major antigenic proteins of six strains of the human granu-locytic ehrlichiosis agent by Western immunoblot analysis. J. Clin. Microbiol.35:2606–2611.

17. Zhi, N., N. Ohashi, Y. Rikihisa, H. W. Horowitz, G. P. Wormser, and K.Hechemy. 1998. Cloning and expression of the 44-kilodalton major outermembrane protein gene of the human granulocytic ehrlichiosis agent andapplication of the recombinant protein to serodiagnosis. J. Clin. Microbiol.36:1666–1673.

18. Zwizinski, C., T. Date, and W. Wickner. 1981. Leader peptidase is found inboth the inner and outer membranes of Escherichia coli. J. Biol. Chem.256:3593–3597.

3284 KIM AND RIKIHISA J. CLIN. MICROBIOL.

on Septem

ber 24, 2020 by guesthttp://jcm

.asm.org/

Dow

nloaded from