Veterinary Microbiology 98 (2004) 313–322

Adhesion of outer membrane proteins containing tandemrepeats ofAnaplasmaandEhrlichia species

(Rickettsiales: Anaplasmataceae) to tick cells

José de la Fuentea,∗, Jose C. Garcia-Garciaa, Anthony F. Barbetb,Edmour F. Blouina, Katherine M. Kocana

a Department of Veterinary Pathobiology, College of Veterinary Medicine, Oklahoma State University, Stillwater, OK 74078, USAb Department of Pathobiology, University of Florida, PO Box 110880, Gainesville, FL 32611-0880, USA

Received 7 April 2003; received in revised form 3 November 2003; accepted 3 November 2003

Abstract

Infection of cells by tick-borne rickettsiae appears to be mediated by outer membrane proteins that allow pathogens to adhereto host cells. Major surface protein (MSP) 1a ofAnaplasma marginale, the type species for the genusAnaplasma, was shownpreviously to be an adhesin for tick cells. TheA. marginaleMSP1a has a variable number of tandem 28 or 29 amino acidrepeats located in the amino terminal region of the protein that contains an adhesion domain that is necessary and sufficient forinfection of tick cells. The MSP1a studies demonstrated the importance of combining structural and functional characteristics foridentification of adhesive proteins. In the present study other outer membrane proteins containing tandem repeats were selectedfrom organisms of the family Anaplasmataceae and studied for their adhesive properties to tick cells. The adhesive propertiesand protein characteristics were then analyzed in order to provide a predictor of the adhesion function of proteins identifiedfrom genome sequences. Proteins selected included theA. marginaleMSP1a,A. phagocytophilum100 and 130 kDa,Ehrlichiachaffeensis120 kDa,E. canis140 kDa andE. ruminantium“mucin”, which were all cloned and expressed inEscherichia coliand then tested as adhesins for cultured IDE8 cells. Of the proteins studied, theA. marginaleMSP1a and theE. ruminantium“mucin” were found to be adhesins for tick cells. Although all of these recombinant outer membrane proteins were glycosylated,theA. marginaleMSP1a andE. ruminantium“mucin” adhesins shared a common feature of having a high Ser/Thr content inthe tandem repeats. The results reported herein provide new information on the role ofE. ruminantium“mucin” as an adhesinfor tick cells and also suggest a role of glycans in adhesin molecules.© 2003 Elsevier B.V. All rights reserved.

Keywords: Anaplasmasp.;Ehrlichia sp.; Rickettsia; Outer membrane protein; MSP1a; Tick; Glycosylation

∗ Corresponding author. Tel.:+1-405-744-0372;fax: +1-405-744-5275.

E-mail address:josedelafuente@yahoo.com (J. de la Fuente).

1. Introduction

The organisms in the order Rickettsiales are classi-fied into two families, Anaplasmataceae and Rickettsi-aceae, based on genetic analyses of 16S rRNA,groESL

0378-1135/$ – see front matter © 2003 Elsevier B.V. All rights reserved.doi:10.1016/j.vetmic.2003.11.001

314 J. de la Fuente et al. / Veterinary Microbiology 98 (2004) 313–322

and surface protein genes (Dumler et al., 2001). Mostorganisms in the family Anaplasmataceae multiplyin both vertebrates and invertebrates (primarily ticksand trematodes). Organisms classified within the fam-ily Rickettsiaceae (GenusRickettsiaandOrientia) areobligate intracellular bacteria that grow freely withinthe cytoplasm of eukaryotic cells. Organisms placedin the family Anaplasmataceae (GenusAnaplasma,Ehrlichia, NeorickettsiaandWolbachia) are also obli-gate intracellular organisms, but they are found exclu-sively within membrane-bound vacuoles in the hostcell cytoplasm.

The genusAnaplasmaincludesA. marginale, thetype species that causes anaplasmosis in ruminants,andA. phagocytophilum(formerlyE. phagocytophila,E. equi and the HGE agent, recognized as synony-mous), which affects humans and horses producing agranulocytic ehrlichiosis (Dumler et al., 2001). ThegenusEhrlichia includesE. ruminantium(formerlyCowdria ruminantium), the tick-borne pathogen thatcauses heartwater disease in cattle, andE. canisandE. chaffeensis, which mainly affect canids and humansproducing a monocytic ehrlichiosis or subclinical in-fections (Dumler et al., 2001).

The role of outer membrane proteins in infectionand transmission of rickettsial pathogens has beendescribed (Lee and Rikihisa, 1998; Goodman et al.,1999; Herron et al., 2000; de la Fuente et al., 2001a,b).Ligand–receptor interactions have been demonstratedfor repeated regions in outer membrane proteins frompathogenic bacteria (Wren, 1991), including Anaplas-mataceae (Popov et al., 2000; de la Fuente et al.,2001a,b, 2003a). The 120 kDa outer membrane proteinof E. chaffeensisand theA. marginalemajor surfaceproteins (MSP) 1a and 1b were characterized as ad-hesins for mammalian cell receptors (McGarey et al.,1994; McGarey and Allred, 1994; Popov et al., 2000;de la Fuente et al., 2001a,b, 2003a). However, theA.marginaleMSP1a is the only outer membrane proteinfrom Rickettsiales shown to be an adhesin for culturedIDE8 and native tick gut cells, involved in adhesion,infection and transmission ofA. marginaleby Der-macentorspp. (de la Fuente et al., 2001a,b, 2003a).This protein is variable in molecular weight amonggeographic isolates ofA. marginalebecause of vary-ing numbers of tandem 28 or 29 amino acid repeatslocated in the amino terminal region of the protein(Allred et al., 1990; de la Fuente et al., 2001c, 2002)

and this region was shown to be necessary and suf-ficient for adhesion to tick cells (de la Fuente et al.,2003a).

The genomes of organisms in the genusAnaplasmaand Ehrlichia are being sequenced. However, func-tional analysis combined with the sequence is requiredto understand the mechanisms of pathogen infection,transmission and survival, and to provide informationfor design of better control strategies. The objective ofthis research was to characterize outer membrane pro-teins in Anaplasmataceae with tandem repeats as ad-hesins for tick cells. The data were analyzed togetherwith other characteristics of the proteins to identifyfeatures that could provide a predictor of tick adhesionmolecules for these species.

2. Materials and methods

2.1. Pathogens and cell lines

Isolates ofA. marginale, A. phagocytophilum, E.chaffeensis, E. canisandE. ruminantiumused in theseexperiments are described inTable 1. For adhesion tocultured cells,Ixodes scapularisIDE8 cells, used asa model for tick–pathogen interactions (as reviewedby Blouin et al., 2002), and human kidney 293T cells(Alberti et al., 2002) were cultured as previously re-ported (de la Fuente et al., 2003a).

2.2. Cloning and expression in Escherichia coli ofouter membrane proteins

The outer membrane proteins selected for the studyare summarized inTable 2. The coding region for theouter membrane proteins ofA. marginale, A. phago-cytophilum, E. canisandE. chaffeensiswas amplifiedby PCR using the oligonucleotide primers describedin Table 2 in a 50�l volume (0.2�M each primer,1.5 mM MgSO4, 0.2 mM dNTP, 1× AMV/Tfl reac-tion buffer, 5 UTfl DNA polymerase) employing theAccess RT-PCR system (Promega, Madison, WI).Reactions were performed in an automated DNAthermal cycler (Eppendorf Mastercycler® personal,Westbury, NY) for 35 cycles. After an initial denat-uration step of 30 s at 94◦C, each cycle consistedof a denaturing step of 30 s at 94◦C, annealing at50◦C for 30 s and an extension step of 2.5 min at

J. de la Fuente et al. / Veterinary Microbiology 98 (2004) 313–322 315

Table 1Anaplasmataceae outer membrane proteins included in the study

Outer membrane proteins GenBank accession number(reference)

PCR oligonucleotide primers (5′–3′) for cloningin E. colia

A. marginale(Oklahoma isolate) MSP1a AY010247 (de la Fuente et al.,2001c)

CCGctcgagATGTTAGCGGAGTATGTGTCCGAagatctCGCCGCCGCCTGCGCC

A. phagocytophilum(NY18 isolate) 130 kDa AF020522 (Storey et al., 1998) CCaagcttATGTTTGAACACAATATTCCTGATACCCgtcgacCAACGCGAGCACGTCATCATCAGA

A. phagocytophilum(NY18 isolate) 100 kDa AF020523 (Storey et al., 1998) CCgaattcCATGTATGGTATAGATATAGAGCTAAGCCgtcgacATAACTTAGAACATCTTCATCGTCAG

E. canis(Oklahoma isolate) 140 kDa AF112369 (Yu et al., 2000) CCgaattcCATGGATATTGATAACAATAATGTGACCCagatctTATTAAATCAACTGTTTCTTTGTTAG

E. chaffeensis(Wakula isolate) 120 kDa U49426 (Yu et al., 1997) CCgaattcCATGGATATTGATAATAGTAACATAAGCCagatctTACAATATCATTTACTACATTGTGAT

E. ruminantium(Highway isolate) “mucin” AF308673 (Barbet et al., 2001) GTTCATTCAACAGCAGAAGATTGGATGCCAGTAAATACTCTTTCACC

a Cloning sites (lowercase letters) were introduced in 5′ and 3′ oligonucleotide primers for cloning of PCR products intoE. coliexpression vectors, except for theE. ruminantium“mucin”, which was cloned directly into the pTrcHis2-TOPO vector.

68◦C (for E. canisand E. chaffeensisgenes) or anannealing-extension step of 2.5 min at 68◦C (for A.marginaleand A. phagocytophilumgenes). The pro-gram ended by storing the reactions at 4◦C. Theprimers introduced restriction sites for cloning intopFLAG-CTC (Sigma, St. Louis, MO) for expressionin theE. coli membrane as previously reported (de laFuente et al., 2001a,b, 2003a). The coding sequencesfor outer membrane proteins were fused to the se-quence of the FLAG peptide, except for the sequenceof theE. canis140 kDa protein, which was not fusedto the FLAG.

The open reading frame containing theE. ruminan-tium “mucin” was amplified by PCR of plasmid DNAfrom clone 26hw (Barbet et al., 2001) with oligonu-cleotide primers AB1048 (GTT· · · TTG) and AB1049(GAT · · · ACC) (Table 2) using the enzyme elongase(Invitrogen, Carlsbad, CA). Each cycle consisted of adenaturing step at 94◦C, annealing at 60◦C for 30 sand extension at 68◦C for 1 min. An amplified prod-uct of ∼1050 bp was gel purified and inserted intothe pTrcHis2-TOPO vector (Invitrogen). Sequencingof resultant clone DNA revealed the presence of thefull-length “mucin” gene and the expected myc epi-tope and His tags at the C-terminus.

The resulting plasmids (Table 3) were transformedinto JM109 and induced for expression of recombi-nant proteins as previously reported forA. marginaleMSP1a (de la Fuente et al., 2001a,b, 2003a). Control

E. coli was transformed with plasmid p33, whichcontains a clonedD. variabilis 16S rDNA frag-ment in the vector pGEM-T (de la Fuente et al.,2001a).

All constructs were sequenced at the Core Sequenc-ing Facility, Department of Biochemistry and Molec-ular Biology, Noble Research Center, Oklahoma StateUniversity using ABI Prism dye terminator cycle se-quencing protocols developed by Applied Biosystems(Perkin-Elmer, Foster City, CA).

Expression of recombinant outer membrane pro-teins was assayed by SDS-PAGE and immunoblotting(de la Fuente et al., 2001a,b, 2003a). RecombinantE.coli cells were collected by centrifugation and resus-pended in TBS. Cells were disrupted by sonicationafter three cycles of 30 s, and the membrane fractioncollected by centrifuging at 12,000× g. Membraneproteins (5�g) were separated by SDS-PAGE andtransferred to a nitrocellulose membrane. Membraneswere probed using 10�g/ml solution of anti-FLAGM2 monoclonal antibody (Sigma) and anti-HIS mon-oclonal antibody (Amersham Pharmacia Biotech,Picataway, NJ) diluted 1:3000. Membranes werewashed and incubated with goat anti-mouse IgG al-kaline phosphatase conjugate (KPL, Gaithersburg,MD) diluted 1:10,000. After washing the membraneagain, color was developed using Sigma Fast BCIP/NBT alkaline phosphatase substrate tablets (Sigma,USA).

316J.

de

laF

ue

nte

et

al.

/V

ete

rina

ryM

icrob

iolog

y9

8(2

00

4)

31

3–

32

2

Table 2Characterization of the Anaplasmataceae outer membrane proteins used in the study

Characterization A. marginaleMSP1a

A. phagocytophilum130 kDa

A. phagocytophilum100 kDa

E. canis140 kDa E. chaffeensis120 kDa

E. ruminantium“mucin”

Number of amino acids (predictedmolecular weight)a

623 (63.5 kDa) 619 (66.1 kDa) 578 (61.4 kDa) 688 (73.7 kDa) 548 (60.9 kDa) 351 (35.7 kDa)

Adhesion to tick cellsb Yes No No No No YesAdhesion to mammalian cellsc Yes NRd NR NR Yes NRSurface protein Yes Yes Yes Yes Yes YesSize of tandem repeats (amino acids)a 29 33–34 89 36 80 9Portion of protein occupied by repeats

(amino acids/%)87/14 237/38.3 267/46.2 504/73.3 320/58.4 189/53.8

Protein/repeats region Ser+ Thr content (%)a 17.5/43.7 25.2/29.5 11.2/9.7 21.2/24.3 16.6/19.7 40/43.9Glycosylation Yes Yes Yes Yes Yes YesImmunodominant Yes Yes Yes Yes Yes YesVariable number of repeats between isolatese Yes NR NR NR Yes NR

Protein topologyf

a The predicted molecular weight, localization of repetitive structures and Ser/Thr content were determined using the Statistical Analysis of Protein Sequences (SAPS)(Brendel et al., 1992).

b Summarized from adhesion experiments toI. scapularisIDE8 cells (Table 3).c Adhesion to mammalian cells reported forA. marginaleMSP1a (McGarey et al., 1994; McGarey and Allred, 1994; de la Fuente et al., 2001a) and E. chaffeensis

120 kDa (Popov et al., 2000).d Not reported.e Variable number of repeats between different geographic isolates of the pathogen has been reported forA. marginaleMSP1a (Allred et al., 1990; de la Fuente et al.,

2001c, 2002; Palmer et al., 2001) and E. chaffeensis120 kDa (Chen et al., 1997; Cheng et al., 2003).f The topology of the protein was predicted using hydropathicity plot analysis (Kyte and Doolittle, 1982) and a transmembrane domain prediction model (Hofmann and

Stoffel, 1993). These proteins do not contain a signal peptide but an uncleaved internal secretion-anchor sequence was localized homologous to the MSP1a sequence(de la Fuenteet al., 2003a,b) using the program AlignX (Vector NTI Suite V 5.5, InforMax, North Bethesda, MD) with an engine based on the Clustal W algorithm (Thompson et al., 1994).

J. de la Fuente et al. / Veterinary Microbiology 98 (2004) 313–322 317

Tabl

e3

Adh

esio

nto

cultu

red

tick

cells

ofre

com

bina

ntE.

coli

expr

essi

ngA

napl

asm

atac

eae

oute

rm

embr

ane

prot

eins

Rel

evan

tpr

otei

nex

pres

sed

Pla

smid

carr

ied

byre

com

bina

ntE

.co

li

pFLC

1a(A

.m

arg

ina

leM

SP

1a)

p26H

WC

3(E

.ru

min

an

tium

“muc

in”)

pCA

F1

(E.

can

is14

0kD

a)

pCH

1(E

.ch

affe

en

sis

120

kDa)

pHG

EF

3(A

.p

hag

ocy

top

hilu

m13

0kD

a)

pFH

GE

100-

8(A

.p

hag

ocy

top

hilu

m10

0kD

a)

p33

cont

rola

(non

e)

Num

ber

ofco

lony

form

ing

units

(CF

U)

reco

vere

dfr

omID

E8

tick

cells

Exp

erim

ent

1(N

=3)

2060

±96

0bN

Dc

35±

118

±7

0±

0N

D22

0±

72E

xper

imen

t2

(N=

2)19

00±

141b

860

±14

1bN

DN

DN

DN

D13

0±

14E

xper

imen

t3

(N=

2)29

50±

778b

ND

ND

ND

ND

0±

029

±24

Exp

erim

ent

4(N

=2)

3500

±33

3b30

0±

11b

ND

ND

ND

ND

52±

12

aT

heco

ntro

lpl

asm

idp3

3co

ntai

nsa

clon

edD.

varia

bili

s16

SrD

NA

frag

men

tin

the

vect

orpG

EM

-T(

dela

Fue

nte

etal

.,20

01a

).R

ecom

bina

ntE

.co

litr

ansf

orm

edw

ithp3

3al

way

sga

vea

back

grou

ndad

hesi

onto

tick

cells

,w

hich

was

not

dete

cted

for

reco

mbi

nant

E.

coli

carr

ying

the

vect

oral

one.

bS

igni

fican

tlydi

ffere

ntfr

omco

ntro

l(P<

0.05

;S

tude

nt’s

t-te

st).

cN

otde

term

ined

.

2.3. Binding assay for adhesion of recombinant E.coli to tick cells

The adhesion to cultured IDE8 tick cells of recom-binantE. coli expressing the outer membrane proteinswas evaluated inE. coli recovery adhesion assays asreported byde la Fuente et al. (2001a).

2.4. Competition of A. marginale MSP1a and E.ruminantium “mucin” for adhesion to tick IDE8 cells

RecombinantA. marginaleMSP1a andE. ruminan-tium “mucin” were affinity purified using anti-FLAGantibodies (Sigma) and histidine binding (Invitrogen),respectively, following manufacturer’s instructions.Recombinant MSP1a was biotinylated with d-BiotinN-hydroxysuccinimide ester (Sigma) (Harlow andLane, 1988). Cultured tick IDE8 and human 293Tcells were homogenized on ice with a polytron (Pow-erGen 125, Fisher, Pittsburgh, PA). Protein concen-trations were determined using the DC Protein Assay(Bio-Rad, Hercules, CA).

For binding assays, 1�g total protein of tick cell ex-tract (TCE) per well was used to coat microtiter plates(Immulon 2HB, Dynex Technologies, Chantilly, VA).One hundred microliters of each extract (10�g/ml) inPBS was used to coat the wells. As a control, plateswere coated (100�l/well) with 10�g/ml of humankidney 293T cell extract in a similar fashion. Plateswere incubated overnight at 4◦C and tightly coveredto prevent evaporation. Plates were washed three timeswith PBS with 0.05% Tween 20 (PBST). Non-specificsites were blocked by incubating the coated wells with2% skim milk in PBS for 2 h at 37◦C followed bywashing with PBST.

In the first series of competition experiments, thebiotinylated (Bt)-R2FL peptide, derived from MSP1awith adhesion to TCE (de la Fuente et al., 2003a), wasused as the target for competition with recombinantA.marginaleMSP1a,E. ruminantium“mucin” or BSA(competitors). Plates were incubated with 0, 8, 16,31, 62, 125, 250, and 500�g/well (in 100�l of 2%skim milk in PBS) of the competitor at 37◦C for 1 hand then with 0 and 25�g/well of the target Bt-R2FLat 37◦C for 1 h. The plates were washed three timeswith TBS. Binding of the target was detected us-ing horseradish peroxidase–streptavidin conjugate(Sigma) (1:20,000) as a secondary reagent and TMB

318 J. de la Fuente et al. / Veterinary Microbiology 98 (2004) 313–322

substrate (Sigma) in 0.05 M phosphate–citrate bufferof pH 5 containing 0.03% sodium perborate was usedfor color development. The OD was read at 450 nmafter stopping the reaction with 25�l 2 N H2SO4per well. The percent inhibition of specific Bt-R2FLbinding to tick cells was calculated for the [ODTCE −OD293T] values as [(ODTARGET/NO COMPETITOR −ODTARGET/COMPETITOR)/(ODTARGET/NO COMPETITOR–ODNO TARGET/NO COMPETITOR)] × 100. The experi-ment was repeated twice with two replicas per treat-ment and the percent inhibition between treatmentswas compared by Student’st-test.

In the second series of experiments, Bt-MSP1a (tar-get) and the “mucin” (competitor) competed directlyfor adhesion to tick cells. Zero, 2 and 10�g/well ofthe target were incubated with 0, 16, 31, 62, 125, and250�g/well of the competitor for 1 h at 37◦C in a fash-ion similar to the previous experiment. Zero, 16, 31,62, 125, and 250�g/well of non-biotinylated MSP1aand BSA were included as positive and negative con-trol competitors, respectively. Bound Bt-MSP1a wasquantified and the average±S.D. was plotted for eachtreatment(N = 2).

2.5. Analysis of glycosylation of recombinant outermembrane proteins

Glycosylation of the recombinant proteins ex-pressed inE. coli was determined by total carbo-hydrate labeling in solution using the Immun-BlotKit for glycoprotein detection (Bio-Rad, USA) andfollowing the manufacturer’s instructions. Briefly,20�g total protein of crude extracts of recombi-nant E. coli expressing the rickettsial proteins weredissolved in 20�l of 100 mM sodium acetate, pH5.5. Ovalbumin supplied with the kit was used as apositive control of the detection reaction. The car-bohydrates were oxidized with sodium periodate for20 min at room temperature (RT) in the dark. Sodiumbisulfite was added and samples incubated for 5 min.Biotin–hydrazide was then added to label the aldehy-des produced by oxidation of the carbohydrates andallowed to react for 60 min at RT. Sample loadingbuffer was finally added and the proteins denaturedfor 5 min at 100◦C. The samples were analyzed byWestern blot as described above using a 1:2000 so-lution of streptavidin–alkaline phosphatase conjugateand BCIP-TMB (Sigma) as substrate.

3. Results

3.1. Expression and characterization of outermembrane proteins in E. coli

Outer membrane proteins ofAnaplasma andEhrlichia species were selected to evaluate their ca-pacity to adhere to tick cells (Table 2). These immun-odominant proteins have been predicted to be exposedon the pathogen’s membrane and to contain tandemrepeated peptides in the extracellular region of the pro-tein (Table 2). Recombinant outer membrane proteinswere expressed inE. coli fused to the FLAG peptideor the His tag at the C-terminus, except for theE.canis140 kDa protein. Recombinant outer membraneproteins were expressed associated with theE. colimembrane at levels similar for all proteins, except theE. chaffeensis120 kDa, which was expressed at lowerlevels (Fig. 1A and B). Furthermore, all recombinantproteins were glycosylated, showing higher molecu-lar weights than those deduced from the amino acidsequence of the protein (Fig. 1CandTable 2).

3.2. Adhesion to tick cells of recombinant E. coliexpressing outer membrane proteins

Adhesion ofE. coli expressing recombinant outermembrane proteins to cultured IDE8 cells was evalu-ated in colony recovery assays.A. marginaleMSP1aandE. ruminantium“mucin” were strong adhesins forIDE8 cells (Table 3). TheE. canis140 kDa,E. chaf-feensis120 kDa andA. phagocytophilumproteins werenot adhesins to tick cells under these experimentalconditions (Table 3).

3.3. Competition between A. marginale MSP1a andE. ruminantium “mucin” for adhesion to IDE8 tickcells

Experiments were conducted to determine if theA.marginaleMSP1a and theE. ruminantium“mucin”competed for the same receptor on IDE8 tick cells.Indirect competition of recombinant adhesins with arepeated peptide from MSP1a, which was previouslyshown to bind to IDE8 cells, suggested thatE. rumi-nantium“mucin” and A. marginaleMSP1a competewith the R2FL peptide for adhesion to IDE8 tick cellsonly at low concentrations of the competitor (Fig. 2).

J. de la Fuente et al. / Veterinary Microbiology 98 (2004) 313–322 319

Fig. 1. Expression and characterization of recombinant outer mem-brane proteins inE. coli: (A) 10�g total protein were loaded perlane in an 8% SDS-polyacrylamide gel and stained with CoomassieBrilliant Blue. The gel was transferred to a nitrocellulose mem-brane for (B) Western blot analysis using as primary antibodiesa mixture of anti-FLAG M2 and anti-HIS monoclonal antibod-ies or (C) glycoprotein detection using periodate oxidation andbiotin–hydrazide conjugation. Plasmid carried by recombinantE.coli and relevant proteins expressed were as described inTable 3.Lanes 1, p33 (controlE. coli); 2, pFLC1a (A. marginaleMSP1a);3, pFLC1b.2 (A. marginaleMSP1b; de la Fuente et al., 2001a);4, pFHGE100-8 (A. phagocytophilum100 kDa); 5, pHGEF3 (A.phagocytophilum130 kDa); 6, pCH1 (E. chaffeensis120 kDa);7, pCAF1 (E. canis 140 kDa); 8, p26HWC3 (E. ruminantium“mucin”). Sigmamarker wide molecular range (Sigma) was usedas molecular weight markers (kDa) in the electrophoresis. Arrow-heads indicate the position of recombinant proteins.

However, direct competition betweenA. marginaleMSP1a andE. ruminantium“mucin” for adhesion tothe TCE demonstrated that the recombinant adhesinsdo not compete for the same receptor on IDE8 tickcells (Fig. 3).

4. Discussion

Tick-borne pathogens alternate in their life cyclebetween the mammalian host and the tick vector. Spe-cific ligands on the outer membrane proteins of thepathogen may adhere to specific receptors on tick cells,a prerequisite for infection of ticks. ForA. marginale,

0

20

40

60

80

100

120

Competitor

Per

cen

t in

hib

itio

n

MSP1a Mucin BSA MSP1a Mucin BSA

16 µg/well 32 µg/well

a

a

a,b

b

c

c

Fig. 2. RecombinantA. marginaleMSP1a andE. ruminantium“mucin” competed with the R2FL peptide for adhesion to IDE8 tickcells. Recombinant proteins were used to compete with biotinylatedR2FL peptide for binding to TCE. TCE and 293T-coated wellswere incubated with competitors (MSP1a, “mucin” and BSA) for1 h at 37◦C and then with the target Bt-R2FL peptide for 1 hat 37◦C. The percent inhibition of specific Bt-R2FL binding totick cells was calculated and the average± S.D. plotted for eachtreatment. The experiment was repeated twice with two replicasper treatment. Percent inhibition between treatments was comparedby Student’st-test. Differences were detected for 16 and 32�gcompetitor per well only. Different letters indicate statisticallysignificant differences(P < 0.05).

0

0.5

1

1.5

2

2.5

3

0 8 16 31 62 125 250

Competitor (µg/well)

Bo

un

d B

t-M

SP

1a (

µg/w

ell)

MSP1a

Mucin

BSA

Fig. 3. RecombinantA. marginaleMSP1a andE. ruminantium“mucin” did not compete for adhesion to IDE8 tick cells. Recom-binant proteins were used to compete directly for adhesion to tickcells. Two micrograms of the Bt-MSP1a target were incubatedwith the “mucin” competitor for 1 h at 37◦C. Non-biotinylatedMSP1a and BSA were included as positive and negative controlcompetitors, respectively. Bound Bt-MSP1a was quantified and theaverage± S.D. was plotted for each treatment(N = 2).

320 J. de la Fuente et al. / Veterinary Microbiology 98 (2004) 313–322

the type species for the genusAnaplasma, MSP1a hasbeen shown to be an adhesin for tick cells involved inadhesion, infection and transmission of the pathogenby Dermacentorspp. (de la Fuente et al., 2001a,b).Furthermore, the extracellular region of the proteincontaining the tandem repeated peptides was shownto be necessary and sufficient to mediate the adhe-sion and infection of tick cells (de la Fuente et al.,2003a). These results led to experiments demonstrat-ing that anti-MSP1a antibodies reduce the infectivityof A. marginalefor tick cells in vitro and in vivo,suggesting the inclusion of recombinant MSP1a invaccine preparations for the control of anaplasmosis(Blouin et al., 2003; de la Fuente et al., 2003b).

The experiments with MSP1a demonstrated the im-portance of functional analysis to identify tick ad-hesins in tick-borne rickettisiae. The goal of this re-search was to characterize the adhesion of other outermembrane proteins containing tandem repeats identi-fied in Anaplasmataceae to tick cells and to analyzeadhesive properties together with other characteristicsof the protein to identify features that may provide pre-dictors of tick cell adhesins from genome sequencespresently being determined.

The results of adhesion assays with outer membraneproteins described herein demonstrated that only theA. marginaleMSP1a and theE. ruminantium“mucin”were strong adhesins for tick cells. However, differ-ences in the expression levels and/or the structure ofrecombinant proteins inE. coli could have affectedresults of adhesion experiments. Nevertheless, the re-sults of expression analysis suggested that the recom-binant outer membrane proteins were expressed on theE. coli membrane and the use of IDE8 tick cells hasbeen validated in various assays including theE. colirecovery adhesion assay (reviewed byBlouin et al.,2002).

The I. scapularis IDE8 cells, originally derivedfrom tick embryos (Munderloh et al., 1994), have beenused to propagateA. marginale (Munderloh et al.,1996a; Blouin and Kocan, 1998), A. phagocytophilum(Munderloh et al., 1996b, 1999), E. canis (Kocanet al., 1998) and E. ruminantium(Bell-Sakyi et al.,2000), and have proven to be a model for the studyof tick–pathogen interactions (reviewed byBlouinet al., 2002). Using this system we demonstrated thatA. marginaleMSP1a andE. ruminantium“mucin” donot compete for adhesion to IDE8 cells. Recombinant

MSP1a and “mucin” competed with the R2FL peptidebut did not compete for adhesion to IDE8 cells in directcompetition experiments. This difference could be ex-plained by the possibility of MSP1a and the “mucin”of sharing one component of the receptor, whichinteracts with the R2FL peptide, but having othercomponents of the receptor specific for each adhesin.Alternatively, the R2FL peptide could bind weakly toboth MSP1a and “mucin” receptors but the recombi-nant proteins can only bind to their respective receptor.Both possibilities provide an alternative explanationfor the specificity that likely governs tick–pathogeninteractions at the level of ligand–receptor interactionsbetween different pathogen and tick species. However,because of the embryonic origin of IDE8 cells and thecapacity of these cells to support the multiplicationof different pathogens not transmitted in nature byI.scapularis, these cells probably retain characteristicsof various cell types in adult ticks and properties notnecessarily identical to adult tick tissues and cells.

The recombinant outer membrane proteins wereglycosylated. The glycosylation ofE. chaffeensis120 kDa and E. canis 140 kDa proteins was re-ported before byMcBride et al. (2000). Recently,A. marginaleMSP1a was shown to be glycosylatedboth in its native form and when expressed inE. coli(Garcia-Garcia et al., unpublished data). The resultsof this study corroborated these findings and sug-gested that glycosylation of outer membrane proteinsin Anaplasmataceae may be a common characteristic.The mechanism by which the glycans are attached tothese proteins is unknown but they may be O-linkedto Ser/Thr glycosylation sites (McBride et al., 2000;Garcia-Garcia et al., unpublished data).

Analysis of the outer membrane proteins usedin this study correlating different characteristics ofthe protein with the capacity to adhere to tick cells,demonstrated that the only distinctive feature commonto A. marginaleMSP1a andE. ruminantium“mucin”was the higher than 43% Ser/Thr content of the regioncontaining the repeated peptides (Table 2). This valueis at least 1.5 times higher than the Ser/Thr content ofthe repeats region in the other outer membrane pro-teins analyzed. Local similarity analysis (Huang andMiller, 1991) between the MSP1a and “mucin” re-peats evidenced a 50% identity in 14 residues overlap(score 22; gap frequency 0%) (SSQSDQASTSSQLGin MSP1a and∗∗PEGA∗V∗∗∗PE∗ in “mucin”; aster-

J. de la Fuente et al. / Veterinary Microbiology 98 (2004) 313–322 321

isks represent identical amino acids) in the regioncontaining the Ser/Thr-rich neutralization epitope ofMSP1a (QASTSS;Palmer et al., 1987). The role ofcarbohydrate moieties in these proteins is unknownbut may play a role in the biological function ofadhesion.

The results of this study provided new infor-mation on the role ofE. ruminantium“mucin” asan adhesin for tick cells and suggested a role forglycans in the biological function of tick adhesionmolecules in Anaplasmataceae. This information maybe useful for searching open reading frames con-taining tandem repeats with a high Ser/Thr contentin the genome sequence databases of the Anaplas-mataceae for identification of putative tick adhesinproteins. Adhesion experiments with recombinantE.coli and tick IDE8 cells could then be done to testthe adhesion capacity of proteins before conductinginfection-inhibition experiments with anti-adhesinantibodies.

Acknowledgements

This research was supported by the project No.1669 of the Oklahoma Agricultural Experiment Sta-tion, the Endowed Chair for Food Animal Research(K.M. Kocan, College of Veterinary Medicine, Okla-homa State University), NIH Centers for BiomedicalResearch Excellence through a subcontract to J. de laFuente from the Oklahoma Medical Research Founda-tion, and the Oklahoma Center for the Advancement ofScience and Technology, Applied Research Grant No.AR00(1)-001, and a Grant-in-Aid of Research fromthe National Academy of Sciences through Sigma Xito J.C. Garcia-Garcia. J.C. Garcia-Garcia is supportedby a Howard Hughes Medical Institute PredoctoralFellowship in Biological Sciences. Dr. M. Palmarini(University of Georgia, Athens, GA) is acknowledgedfor providing the human 293T cells. Annie Moreland(University of Florida) is acknowledged for techni-cal assistance. Sue Ann Hudiburg and Janet J. Rogers(Core Sequencing Facility, Department of Biochem-istry and Molecular Biology, Noble Research Cen-ter, Oklahoma State University) are acknowledged foroligonucleotide synthesis and DNA sequencing, re-spectively. We thank Joy Yoshioka for editing themanuscript.

References

Alberti, A., Murgia, C., Liu, S.-L., Mura, M., Cousens, C., Sharp,M., Miller, A.D., Palmarini, M., 2002. Envelope-induced celltransformation by ovine beta-retroviruses. J. Virol. 76, 5387–5394.

Allred, D.R., McGuire, T.C., Palmer, G.H., Leib, S.R., Harkins,T.M., McElwain, T.F., Barbet, A.F., 1990. Molecular basisfor surface antigen size polymorphisms and conservation of aneutralization-sensitive epitope inAnaplasma marginale. Proc.Natl. Acad. Sci. USA 87, 3220–3224.

Barbet, A.F., Whitmire, W.M., Kamper, S.M., Simbi, B.H., Ganta,R.R., Moreland, A.L., Mwangi, D.M., McGuire, T.C., Mahan,S.M., 2001. A subset ofCowdria ruminantiumgenes importantfor immune recognition and protection. Gene 275, 287–298.

Bell-Sakyi, L.M., Paxton, E.A., Munderloh, U.G., Sumption, K.J.,2000. Growth ofCowdria ruminantium, the causative agent ofheartwater, in a tick cell line. J. Clin. Microbiol. 38, 1238–1240.

Blouin, E.F., Kocan, K.M., 1998. Morphology and developmentof Anaplasma marginale(Rickettsiales: Anaplasmataceae) incultured Ixodes scapularis(Acari: Ixodidae) cells. J. Med.Entomol. 35, 788–797.

Blouin, E.F., de la Fuente, J., Garcia-Garcia, J.C., Sauer, J.R.,Saliki, J.T., Kocan, K.M., 2002. Use of a cell culture systemfor studying the interaction ofAnaplasma marginalewith tickcells. Anim. Health Res. Rev. 3, 57–68.

Blouin, E.F., Saliki, J.T., de la Fuente, J., Garcia-Garcia, J.C.,Kocan, K.M., 2003. Antibodies toAnaplasma marginalemajorsurface proteins 1a and 1b inhibit infection for cultured tickcells. Vet. Parasitol. 111, 247–260.

Brendel, V., Bucher, P., Nourbakhsh, I., Blaisdell, B.E., Karlin, S.,1992. Methods and algorithms for statistical analysis of proteinsequences. Proc. Natl. Acad. Sci. USA 89, 2002–2006.

Chen, S.M., Yu, X.J., Popov, V.L., Westerman, E.L., Hamilton,F.G., Walker, D.H., 1997. Genetic and antigenic diversity ofEhrlichia chaffeensis: comparative analysis of a novel strainfrom Oklahoma and previously isolated strains. J. Infect. Dis.175, 856–863.

Cheng, C., Paddock, C.D., Reddy Ganta, R., 2003. Molecularheterogeneity ofEhrlichia chaffeensisisolates determined bysequence analysis of the 28-kilodalton outer membrane proteingenes and other regions of the genome. Infect. Immun. 71,187–195.

de la Fuente, J., Garcia-Garcia, J.C., Blouin, E.F., Kocan, K.M.,2001a. Differential adhesion of major surface proteins 1a and1b of the ehrlichial cattle pathogenAnaplasma marginaletobovine erythrocytes and tick cells. Int. J. Parasitol. 31, 145–153.

de la Fuente, J., Garcia-Garcia, J.C., Blouin, E.F., Kocan, K.M.,2001b. Major surface protein 1a effects tick infection andtransmission of the ehrlichial pathogenAnaplasma marginale.Int. J. Parasitol. 31, 1705–1714.

de la Fuente, J., Van Den Bussche, R.A., Kocan, K.M., 2001c.Molecular phylogeny and biogeography of North Americanisolates ofAnaplasma marginale(Rickettsiaceae: Ehrlichieae).Vet. Parasitol. 97, 65–76.

322 J. de la Fuente et al. / Veterinary Microbiology 98 (2004) 313–322

de la Fuente, J., Van Den Bussche, R.A., Garcia-Garcia, J.C.,Rodriquez, S.D., Garcia, M.A., Guglielmone, A.A., Mangold,A.J., Friche Passos, L.M., Barbosa Ribeiro, M.F., Blouin, E.F.,Kocan, K.M., 2002. Phylogeography of New World isolates ofAnaplasma marginale(Rickettsiaceae: Anaplasmataceae) basedon major surface protein sequences. Vet. Microbiol. 88, 275–285.

de la Fuente, J., Garcia-Garcia, J.C., Blouin, E.F., Kocan, K.M.,2003a. Characterization of the functional domain of majorsurface protein 1a involved in adhesion of the rickettsiaeAnaplasma marginaleto host cells. Vet. Microbiol. 91, 265–283.

de la Fuente, J., Kocan, K.M., Garcia-Garcia, J.C., Blouin, E.F.,Halbur, T., Onet, V., 2003b. Immunization againstAnaplasmamarginalemajor surface protein 1a reduces infectivity for ticks.J. Appl. Res. Vet. Med., in press.

Dumler, J.S., Barbet, A.F., Bekker, C.P.J., Dasch, G.A., Palmer,G.H., Ray, S.C., Rikihisa, Y., Rurangirwa, F.R., 2001.Reorganization of genera in the families Rickettsiaceae andAnaplasmataceae in the order Rickettsiales; Unification ofsome species ofEhrlichia with Anaplasma, Cowdria withEhrlichia, and Ehrlichia with Neorickettsia; Descriptions offive new species combinations; and designation ofEhrlichiaequi and “HGE agent” as subjective synonyms ofEhrlichiaphagocytophila. Int. J. Syst. Evol. Microbiol. 51, 2145–2165;corrigendum Int. J. Syst. Evol. Microbiol. 52 (2002) 5–6.

Goodman, J.L., Nelson, C.M., Klein, M.B., Hayes, S.F.,Weston, B.W., 1999. Leukocyte infection by the granulocyticehrlichiosis agent is linked to expression of a selectin ligand.J. Clin. Invest. 103, 407–412.

Harlow, E., Lane, D., 1988. Antibodies: A Laboratory Manual.Cold Spring Harbor Laboratory, Cold Spring Harbor, p. 341.

Herron, M.J., Nelson, C.M., Larson, J., Snapp, K.R., Kansas, G.S.,Goodman, J.L., 2000. Intracellular parasitism by the humangranulocytic ehrlichiosis bacterium through the P-selectinligand, PSGL-1. Science 288, 1653–1656.

Hofmann, K., Stoffel, W., 1993. TMbase—a database of membranespanning proteins segments. Biol. Chem. 347, 166.

Huang, X., Miller, W., 1991. A time-efficient, linear-space localsimilarity algorithm. Adv. Appl. Math. 12, 337–357.

Kocan, K.M., Munderloh, U.G., Ewing, S.A., 1998. Developmentof the Ebony isolate ofEhrlichia canis in cultured Ixodesscapularis cells. In: Proceedings of the 79th Conference ofResearch Workers in Animal Diseases, Chicago (Abstract 95).

Kyte, J., Doolittle, R.F., 1982. A simple method for displaying thehydropathic character of a protein. J. Mol. Biol. 157, 105–132.

Lee, E.H., Rikihisa, Y., 1998. Protein kinase A-mediated inhibitionof gamma interferon-induced tyrosine phosphorylation of Januskinases and latent cytoplasmic transcription factors in humanmonocytes byEhrlichia chaffeensis. Infect. Immun. 66, 2514–2520.

McBride, J.W., Yu, X.J., Walker, D.H., 2000. Glycosylation ofhomologous immunodominant proteins ofEhrlichia chaffeensisand Ehrlichia canis. Infect. Immun. 68, 13–18.

McGarey, D.J., Barbet, A.F., Palmer, G.H., McGuire, T.C., Allred,D.R., 1994. Putative adhesins ofAnaplasma marginale: majorsurface polypeptides 1a and 1b. Infect. Immun. 62, 4594–4601.

McGarey, D.J., Allred, D.R., 1994. Characterization ofhemagglutinating components on theAnaplasma marginaleinitial body surface and identification of possible adhesins.Infect. Immun. 62, 4587–4593.

Munderloh, U.G., Wang, Y.L.M., Chen, C., Kurtti, T.J., 1994.Establishment, maintenance and description of cell lines fromthe tick Ixodes scapularis. J. Parasitol. 80, 533–543.

Munderloh, U.G., Blouin, E.F., Kocan, K.M., Ge, N.L.,Edwards, W., Kurtti, T.J., 1996a. Establishment of the tick(Acari: Ixodidae)-borne cattle pathogenAnaplasma marginale(Rickettsiales: Anaplasmataceae) in tick cell culture. J. Med.Entomol. 33, 656–664.

Munderloh, U.G., Madigan, J.E., Dumler, J.S., Goodman, J.L.,Hayes, S.F., Barlough, J.E., Nelson, C.M., Kurtti, T.J.,1996b. Isolation of the equine granulocytic ehrlichiosis agent,Ehrlichia equi, in tick cell culture. J. Clin. Microbiol. 34, 664–670.

Munderloh, U.G., Jauron, S.D., Fingerle, V., Leitritz, L., Hayes,S.F., Hautman, J.M., Nelson, C.M., Huberty, B.W., Kurtti, T.J.,Ahlstrand, G.G., Greig, B., Mellencamp, M.A., Goodman, J.L.,1999. Invasion and intracellular development of the humangranulocytic ehrlichiosis agent in tick cell culture. J. Clin.Microbiol. 37, 2518–2524.

Palmer, G.H., Waghela, S.D., Barbet, A.F., Davis, W.C., McGuire,T.C., 1987. Characterization of a neutralization sensitiveepitope on the AM 105 surface protein ofAnaplasmamarginale. Int. J. Parasitol. 17, 1279–1285.

Palmer, G.H., Rurangirwa, F.R., McElwain, T.F., 2001. Straincomposition of the EhrlichiaAnaplasma marginalewithinpersistently infected cattle, a mammalian reservoir for ticktransmission. J. Clin. Microbiol. 39, 631–635.

Popov, V.L., Yu, X.J., Walker, D.H., 2000. The 120 kDaouter membrane protein ofEhrlichia chaffeensis: preferentialexpression on dense-core cells and gene expression inEscherichia coliassociated with attachment and entry. Microb.Pathogen. 28, 71–80.

Storey, J.R., Doros-Richert, L.A., Gingrich-Baker, C., Munroe,K., Mather, T.N., Couglin, R.T., Beltz, G.A., Murphy,C.I., 1998. Molecular cloning and sequencing of threegranulocyticEhrlichia genes encoding high-molecular-weightimmunoreactive proteins. Infect. Immun. 66, 1356–1363.

Thompson, J.D., Higgins, D.G., Gibson, T.J., 1994. CLUSTALW: improving the sensitivity of progressive multiple sequencealignment through sequence weighting, positions-specific gappenalties and weight matrix choice. Nucl. Acids Res. 22, 4673–4680.

Wren, B.W., 1991. A family of clostridial and streptococcalligand-binding proteins with conserved C-terminal repeatsequences. Mol. Microbiol. 5, 797–803.

Yu, X.J., Crocquet-Valdes, P., Walker, D.H., 1997. Cloning andsequencing of the gene for a 120-kDa immunodominant proteinof Ehrlichia chaffeensis. Gene 184, 149–154.

Yu, X.J., McBride, J.W., Diaz, C.M., Walker, D.H., 2000.Molecular cloning and characterization of the 120-kilodaltonprotein gene of Ehrlichia canis and application of therecombinant 120-kilodalton protein for serodiagnosis of canineehrlichiosis. J. Clin. Microbiol. 38, 369–374.