A cement protein of the tick Rhipicephalus appendiculatus, located in thesecretory e cell granules of the type III salivary gland acini, induces strong

antibody responses in cattleq

Richard Bishop*, Bronwen Lambson, Clive Wells, Pratibala Pandit, Julius Osaso,Catherine Nkonge, Subhash Morzaria, Antony Musoke, Vishvanath Nene1

International Livestock Research Institute (ILRI), P.O. Box 30709, Nairobi, Kenya

Received 10 December 2001; received in revised form 18 February 2002; accepted 18 February 2002

Abstract

Protein components of the cement cone of ixodid ticks are candidates for inclusion in vaccines against tick infestation, since they are

essential for tick attachment and feeding. We describe here the cloning of a cDNA encoding a 36 kDa protein, designated Rhipicephalus

Immuno-dominant Molecule 36 (RIM36), present in salivary glands and the cement cone material secreted by Rhipicephalus appendiculatus.

The 334-amino-acid sequence of RIM36 has a high content of glycine, serine and proline. The protein contains a predicted N-terminal signal

peptide and two classes of glycine-rich amino acid repeats, a GL[G/Y/S/F/L] tripeptide and a GSPLSGF septapeptide. Comparison of

genomic and cDNA sequences reveals a 597 bp intron within the 3 0 end of the RIM36 gene. Immuno-electron microscopy demonstrates that

RIM36 is predominantly located in the e cell granules of the type III salivary gland acini. An Escherichia coli recombinant form of the

proline-rich C-terminal domain of RIM36 reacts with antisera from Bos indicus cattle, either experimentally infested with R. appendiculatus,

or exposed to ticks in the field. The 36 kDa protein is strongly recognised on Western blots of salivary gland lysates and soluble extracts of

purified R. appendiculatus cement cones by polyclonal antibodies generated against recombinant RIM36, and by antisera from cattle

experimentally infested with ticks. The data indicate that this tick cement component is a target of strong antibody responses in cattle

exposed to feeding ticks. q 2002 Australian Society for Parasitology Inc. Published by Elsevier Science Ltd. All rights reserved.

Keywords: Rhipicephalus appendiculatus; Tick cement protein; Glycine-rich repeat; e cell; Antibody response

1. Introduction

Ticks surpass all other arthropods in the number and

variety of pathogens they transmit to domestic animals,

and rank second only to mosquitoes as vectors of human

pathogens. In addition to transmission of viral, bacterial,

protozoan, fungal and nematode diseases, they are also

directly responsible for damage to hides and lost production

in livestock. The economic impact of ticks and tick-borne

diseases, together with costs of control measures, has been

estimated at 7 billion dollars globally in the livestock sector

(McKosker, 1979). Existing control measures involving

application of chemical acaricides to control livestock

ticks are becoming less sustainable due to acaricide resis-

tance among the ticks (Nolan and Schnitzerling, 1986),

increasing costs, and concerns relating to toxic residues in

milk and meat (reviewed by Willadsen, 1997). Alternative

control options such as vaccines designed to reduce tick

infestation, which are deployed as part of integrated control

measures involving reduced acaricide usage, are therefore

being actively explored (progress reviewed by Opdebeeck,

1994; Willadsen, 1997).

The only currently available commercial vaccine against

tick infestation is based on BM86 a ‘concealed’ gut antigen

of the one host tick Boophilus microplus (Riding et al., 1994;

Willadsen et al., 1995). However, naturally acquired resis-

tance to ticks is based on a complex combination of host

immune responses to a variety of antigens, including a strong

cutaneous cell-mediated component against secreted sali-

vary gland peptides (reviewed by Willadsen, 1980). A strik-

ing feature of the tick vector/host interface is that the tick

secretes a variety of pharmacologically active salivary gland

molecules, which are designed to increase the efficiency of

feeding, into the mammalian host. One such category of

secreted molecules are the components of tick attachment

cement which enable ixodid ticks to remain attached to the

International Journal for Parasitology 32 (2002) 833–842

0020-7519/02/$20.00 q 2002 Australian Society for Parasitology Inc. Published by Elsevier Science Ltd. All rights reserved.

PII: S0020-7519(02)00027-9

www.parasitology-online.com

q Nucleotide sequence data reported in this paper are available in the

GenBank database under accession number AY045671.

* Corresponding author. Tel.: 1254-2-630743; fax: 1254-2-631499.

E-mail address: r.bishop@cgiar.org (R. Bishop).1 Present address: The Institute for Genomic Research, 9712 Medical

Centre Drive, Rockville, MD 20850, USA.

host during the prolonged feeding period of 4–8 days and

prevent host immune response molecules from coming into

contact with the tick proboscis (reviewed by Binnington and

Kemp, 1980; Sonenshine, 1993). The tick cement cone is a

layered structure comprising two major types of cement. The

first type of cement is produced minutes after tick attachment

and hardens rapidly to form a rigid central core of the cone.

The second type of cement is secreted later from approxi-

mately 24 h after attachment and hardens gradually to form a

relatively flexible outer cortex. Cement production typically

continues until the third or fourth day after attachment

(Moorhouse and Tatchell, 1966; reviewed by Binnington

and Kemp, 1980). Tick cement is primarily proteinaceous

(Kemp et al., 1982), but also contains some carbohydrate

and lipid. The processes by which cement hardens are

currently unknown and the individual proteins comprising

tick cement are poorly characterised. Cement cone proteins

represent candidates for inclusion in vaccines against ticks,

or the pathogens they transmit, since formation of the cone is

essential for the tick to attach and feed. It has been shown that

un-purified cement components (Brown and Askenase,

1986), and also a purified 90 kDa cement protein (Shapiro

et al., 1987, 1989) can induce a measure of host resistance to

tick infestation when used as experimental vaccines in

laboratory animal models. A recombinant form of an extra-

cellular-matrix-like protein present in the salivary gland of

Haemaphysalis longicornis, which is hypothesised to be a

cement component, has also been shown to generate anti-

body responses in rabbits, which induce mortality in feeding

nymphal and larval ticks (Mulenga et al., 1999).

The African brown ear tick, Rhipicephalus appendicula-

tus is the main vector of the protozoan parasite Theileria

parva, which induces East Coast fever, a rapidly fatal

disease in cattle in sub-Saharan Africa. The disease is parti-

cularly severe in exotic Bos taurus animals, with major

economic impact, in eastern central and southern Africa

(Norval et al., 1992). Rhipicephalus appendiculatus is also

the vector of several viral diseases of livestock and humans

(Nuttall et al., 1994) and has been used extensively in

studies of promotion of tick-borne virus transmission by

salivary gland products (Nuttall, 1998). We describe herein

the cloning of an R. appendiculatus cement protein (RIM36)

containing glycine-rich repeats, which induces strong anti-

body responses in cattle in association with tick feeding.

Immuno-electron microscopy indicates that RIM36 appears

to be expressed predominantly in the e cells of the type III

salivary gland acinus, in which the mammalian-infective

sporozoites of T. parva are also located.

2. Materials and methods

2.1. Tick material and salivary gland dissection

Rhipicephalus appendiculatus tick material was derived

from the Muguga stock that has been maintained as an

experimental colony at ILRI for over 20 years. The colony

is maintained on cattle and rabbits. For salivary gland

dissection, ticks were allowed to feed on a clean rabbit for

4 days, plucked off the rabbit and washed with a mild deter-

gent, prior to pinning onto a dissection dish, coated with

wax impregnated with Sudan black. The immobilised

ticks were then flooded with sterile PBS (pH 7.4) and the

salivary glands excised under a dissection light microscope.

The dissected glands were kept frozen at 270 8C prior to

use.

2.2. Cement collection and processing for SDS–PAGE

analysis

Ticks were allowed to feed on clean rabbits for 4 days,

removed and placed in a tube for a period of 1–2 days to

allow the cement to dislodge from the mouthparts. Cement

that was still attached was removed using forceps. The

cement was stored at 270 8C prior to use. The cement

was homogenised in PBS in a tissue grinder and mixed

with sample buffer (0.5 M Tris (pH 6.8) containing 10%

SDS, b-mercaptoethanol 3% and glycerol 30%) in the

following ratio: two volumes of cement suspension plus

one volume of sample buffer. The mixture was boiled for

5 min, spun at 14 000 £ g for 10 min and the supernatant

analysed on SDS–PAGE gels.

2.3. Experimental R.appendiculatus infestation of cattle

Theileria parva-infected R. appendiculatus adult ticks

were fed artificially on four Boran (Bos indicus) cattle in

the ILRI tick unit, as described by Bailey (1960). Animals

BK 222 and BK 232 were each infested with 10 ticks (five

males and five females) and BK 233 and BJ 410 were

infested with 100 ticks (50 males and 50 females). The

ticks were allowed to feed on the cattle for 10 days. In

addition, animals Tm194 and Tm 196–Tm199, exposed to

natural tick challenge at Kakuzi, in Central Province of

Kenya, for approximately 2 months, and VP10 and VP17

exposed to field tick challenge for 43 days at Kilifi, in the

Coast province of Kenya, were also used. Sera were

collected before and after exposure on a weekly basis and

stored at 220 8C prior to use.

2.4. Production of experimental antiserum in cattle and

rabbits

The anti-T. parva sporozoite serum used was C16, raised

by immunising cattle with dissected T. parva infected R.

appendiculatus salivary glands, enriched for sporozoite

material using DEAE cellulose chromatography. Genera-

tion of C16 is described in detail in Iams et al. (1990).

Antisera against a bacterially expressed recombinant protein

encoding the C-terminal section of RIM36 was raised in

rabbits. The rabbits were immunised with 100 mg of recom-

binant antigen, emulsified in Freund’s complete adjuvant

and then boosted twice, at 2-week intervals, with an equiva-

R. Bishop et al. / International Journal for Parasitology 32 (2002) 833–842834

lent amount of protein in Freund’s incomplete adjuvant. Ten

days after the final boost sera was obtained from the animals

and checked for reactivity by Western blot analysis using a

dilution of 1:100. Rabbits with strong antibody reactivity

were bled out under anaesthesia 14 days after the last

boost, and the antisera stored at 2208 C prior to use.

2.5. Isolation of genomic and cDNA clones

Preparation of an oligo-dT primed lgt11 cDNA library

from total RNA isolated from purified T. parva sporoblasts,

together with the immunoscreening of this library with anti-

body C16, which resulted in isolation of a cDNA encoding

the RIM36 tick protein has been described previously (Nene

et al., 1992). Affinity purification of specific antibodies

corresponding to the RIM36 protein expressed in lgt11

was performed according to Ozaki et al. (1986). The

cDNA clone of RIM36 isolated by immunoscreening of

the purified sporoblast library with antibody C16 was

subcloned into a pGEM T vector (Promega). Production

of a second cDNA library in lgt11 from poly(A)1 RNA

of T. parva-infected R. appendiculatus salivary glands is

also described in Nene et al. (1992). Isolation of additional

cDNA clones from the infected salivary gland library using

a PCR-amplified RIM36 probe used standard methodolo-

gies (Sambrook et al., 1989). The 5 0 end of the RIM36

coding sequence was obtained by a 5 0 RACE (rapid ampli-

fication of cDNA ends) procedure using the SMART RACE

cDNA Amplification kit (Clontech) with an internal primer

ILO 8416 (5 0-GAG TGG ACT GCC GTA TCC ACC GAG

CAG A-3 0) derived from a sequence close to the 5 0 end of

the original cloned RIM36 sequence.

2.6. Northern and Southern blotting

Preparation of tick genomic DNA from immature eggs

used the method of Barker (1998). Agarose gel electrophor-

esis, Southern blotting onto Hybond N1 membranes (Amer-

sham) and hybridisation employed standard procedures

(Sambrook et al., 1989). Probe labelling used a Megaprime

kit (Amersham) and 200 mCi of high specific activity (6000

Ci/mmol) [a-32P]dCTP radiolabel. Northern blotting of total

uninfected and T. parva-infected total RNA (20 mg) used

the rapid method described by Pelle and Murphy (1993).

Membranes were washed in 0:5£ SSC/0.1% SDS, or 0:1£

SSC/0.1% SDS at 65 8C.

2.7. Nucleotide sequencing

The sequences of the original RIM36 cDNA clones were

determined manually using the ‘fmol’ DNA sequencing

system (Promega), using a combination of vector primers

and internal oligonucleotide primers based on acquired

sequence. Subsequent sequences were derived using the

Dye Terminator cycle sequencing system (Perkin Elmer

Applied Biosystems) and an automated sequencer (Perkin

Elmer Applied Biosystems, ABI 377).

2.8. Expression of RIM36 in Escherichia coli

A PCR using high-fidelity Thermococcus littoratus (Tli)

thermostable DNA polymerase (Promega), with cloned

RIM36 cDNA as the template, was performed using cycling

conditions of 1 min each at 94 8C, 55 8C and 72 8C for 30

cycles. Oligonucleotides ILO 3983 (5 0-CGC GGA TCC

AGG AGA TTC CCC GGC GAC-3 0) and ILO 3984 (5 0-

CGG GGT ACC TTA GAT TGC AAC GTG TTC CTG-3 0),

which respectively incorporated synthetic BamHI and KpnI

sites, were used as primers. The 336 bp PCR product corre-

sponding to the C-terminal 112 amino acid residues of

RIM36 was cloned into BamHI/KpnI double-digested

pQE30 expression vector (Qiagen). The construct was

checked by restriction analysis and nucleotide sequencing.

Expression was performed using the M15 E. coli strain after

induction with 1 mM isopropyl thiogalactoside for 4 h, as

described in the QIA expressionist product information

(Qiagen). After lysis of the E. coli cells, the expressed

protein incorporating six amino-terminal histidines was

denatured in 8 M urea/0.1 M phosphate buffer (pH 8.0),

purified by a batch affinity chromatography procedure, on

Ni–NTA agarose (Qiagen) and eluted using 100 mM EDTA,

pH 8.0. The urea was removed by step dialysis against PBS

(pH 8.0), prior to vaccination of experimental animals.

2.9. Polyacrylamide gel electrophoresis and Western

blotting

Purified recombinant protein was size fractionated by

12.5% or 7.5–17.5% gradient SDS–PAGE and the protein

transferred onto Nytran or nitrocellulose (Schleicher &

Schuell) filters following standard procedures (Laemmli,

1970) and blocked in 5% milk powder. The filters were

reacted overnight at 4 8C with a 1:2000–1:4000 dilution of

sera from cattle experimentally exposed to tick infestation,

or a 1:100 dilution of C16-anti-sporozoite bovine serum, or

1: 4000 dilution of rabbit anti-RIM36 antiserum. After

washing the bound antibody was detected, either by horse-

radish-conjugated goat anti-rabbit whole immunoglobulins

using 3 0,3-diaminobenzidine as chromogen, or protein G

labelled with 125I (Amersham).

2.10. Immuno-electron microscopy (Immuno-EM)

For immuno-EM, T. parva-infected and uninfected

dissected salivary glands, from ticks, which had been fed

for 4 days, were fixed in 4% paraformaldehyde containing

0.1% glutaraldehyde and 0.2% picric acid in 0.1 M phos-

phate buffer. Lowicryl K4M-embedded specimens were

prepared essentially following the enhanced membrane

contrast method (Berryman and Rodewald, 1990) as modi-

fied by Burleigh et al. (1993). Ultra-thin sections (60 nm)

were incubated with the primary antibody and then labelled

with goat anti-rabbit IgG conjugated to 10 nm gold

(Biocell). Sections were stained with aqueous uranyl acetate

R. Bishop et al. / International Journal for Parasitology 32 (2002) 833–842 835

and Reynolds lead citrate, and examined in a JEOL 1010

transmission electron microscope.

3. Results

3.1. Isolation of a cDNA encoding a 36 kDa salivary gland

protein from R. appendiculatus

C16, a bovine antiserum raised against T. parva sporo-

zoites enriched from infected R. appendiculatus salivary

gland material by chromatography on DEAE cellulose has

been used to clone DNA encoding sporozoite antigens (Iams

et al., 1990; Nene et al., 1992). However, this antiserum

contains substantial reactivity with tick antigens (Fig 1A).

Affinity purified antibodies from C16 specific for fusion

protein expressed by Isg16, one of the lgt11 clones isolated

by immunoscreening with C16, predominantly recognised a

protein of approximately 36 kDa present in total uninfected

tick salivary gland lysates (Fig. 1B), suggesting that the

encoded protein might be of tick salivary gland origin.

Hybridisation of a PCR product derived from the lgt11

insert sequence to a Southern blot of R. appendiculatus

genomic DNA confirmed that the sequence was present

within the tick genome, and the pattern of fragments

detected was consistent with a single copy of the gene

(data not shown). Hybridisation was also observed to a

single restriction fragment in Boophilus decoloratus geno-

mic DNA, but not in Amblyomma variegatum genomic

DNA. The insert sequence did not hybridise to T. parva

genomic DNA (data not shown).

3.2. Sequence of the gene encoding RIM36

The nucleotide sequence of a 1214 bp consensus cDNA

sequence (GenBank accession number AY045761) encoding

the RIM36 protein contained a 9 bp poly(A) tail, a putative

AATAAA polyadenylation signal 13 bp upstream and an 88

bp 3 0 untranslated region (3 0 UTR). Northern blotting

revealed an approximately 1300 bp transcript in both unin-

fected and T. parva-infected R. appendiculatus salivary

gland RNA, which was consistent with the size of the

cDNA (data not shown). The cDNA contained a 334 codon

open reading frame from bp 62 to 1066. A signal peptide was

located at the N terminus of the protein with cleavage

predicted between residues 18 and 19 when using neural

network-based programs (Nielsen et al., 1997; SignalP-NN

http://www.cbs.dtu.dk), and residues 16 and 17 using when

using hidden Markov models (SignalP-HMM http://

www.cbs.dtu.dk), or weight matrix analysis (Von Heijne,

1986; http://www.hgmp.mrc.ac.uk). The signal peptide is

shown heavily underlined in Fig. 2. The overall amino acid

composition of the protein was rich in glycine (24.5%),

proline (11.4%), leucine (12.5%) and serine (13.2%), but

there were no cysteine residues. A notable feature of the

sequence was the presence of domains containing amino

acid repeats. Nineteen copies of a GL dipeptide, associated

with either G (seven copies), Y (four copies), S (four copies),

F (two copies) or L (two copies) in position three of an amino

acid triplet were located between amino acid residues 31 and

119. Five copies of GLG followed by a GLS were repeated

directly in tandem. The GL[G/Y/S/L/F] tripeptides within

the N- terminal domain are highlighted by boxes in Fig. 2.

In addition 12 tandemly repeated copies of a septapeptide

repeat GSPLSGF (shown underlined in Fig. 2) were located

directly C-terminal to the glycine-rich domain. A search of

the NCBI patent database revealed 100% amino acid iden-

tity, spanning 143 codons, with two non-contiguous regions

of sequence 13 (accession number AX003897) within UK

patent number WO9924567. The patent ‘Tissue cement

proteins from Rhipicephalus appendiculatus’ was granted

in 1999 to two inventors, P. Nuttall and G. Paesen. Sequence

13 (shown aligned with RIM36 in Fig. 2) was isolated by

screening of tick salivary gland cDNA libraries with antisera

raised against purified R. appendiculatus cement proteins

(Patent WO9924567).

3.3. The RIM36 gene contains an intron

Comparison of the RIM36 cDNA sequence with a partial

sequence generated by PCR amplification from tick geno-

R. Bishop et al. / International Journal for Parasitology 32 (2002) 833–842836

Fig. 1. Recognition of a 36 kDa polypeptide in lysates of Rhipicephalus

appendiculatus salivary glands by bovine antisera C16 and components of

C16 specific for lgt11 clone Isg16. The left panel (A) shows a Western blot

of uninfected R. appendiculatus salivary glands, from female ticks fed for 4

days and probed with bovine C16 anti-sporozoite serum. The right panel

(B) shows a similar blot, reacted with C16 components; affinity purified

using the fusion protein expressed by lgt11 clone Isg16.

mic DNA revealed the presence of a 597 bp intron close to

the 3 0 end of the coding sequence (Fig. 2). The intron exhib-

ited the 5 0 GT and 3 0 AG, which are conserved intron/exon

junction sequences in eukaryotes. The G/C content of the

intron is 34%, whereas that of the flanking exons is 63%. In

the phylum arthropoda introns average 39% G/C and exons

52% G/C. According to the ISIS intron sequence database

(http://www.introns.com), the average length of more than

R. Bishop et al. / International Journal for Parasitology 32 (2002) 833–842 837

Fig. 2. Full-length cDNA and deduced amino acid sequence of RIM36 compared with a cloned Rhipicephalus appendiculatus cement protein (SEQ13) and a

partial genomic DNA sequence (G-DNA). An alignment is shown of a full-length RIM36 cDNA (GenBank accession no. AY045761), with sequence 13

(GenBank accession no. AX003897) of an R. appendiculatus cement protein, described within UK patent W09924567, and a genomic PCR product derived

from the 3 0 end of the RIM36 gene. GLX repeats within the deduced polypeptide are highlighted by boxes and seven amino acid GSPLSGF tandem repeats are

underlined. The predicted N-terminal signal peptide is underlined with a thick line. The start of the C-terminal RIM36 domain expressed in Escherichia coli is

indicated by a vertical arrow. Nucleotide substitutions between the RIM36 cDNA and sequence 13 are highlighted in bold in sequence 13. Sequences present in

RIM36 but absent from sequence 13 are indicated by a broken line.

2000 introns so far characterised in the arthropoda is , 200

bp and few described arthropod introns exceed 500 bp.

3.4. Sequence polymorphism within the RIM36 coding

sequence

Comparison of RIM36 and sequence 13 revealed five

nucleotide sequence differences, none of which resulted in

amino acid substitutions. However, DNA encoding a 59-

amino-acid domain present in RIM36 was absent from the

corresponding region of sequence 13 (indicated by the

broken line in Fig. 2). Four additional RIM36 cDNA clones

were isolated by library screening, using the original clone

as a probe. Comparison of the five RIM36 cDNA sequences,

revealed that four, including the original sequence, were

identical at the nucleotide level. The fifth sequence exhib-

ited six nucleotide differences, three within the 3 0 UTR and

three within the C-terminal region of the coding sequence.

All three of the nucleotide sequences within the coding

region also resulted in amino acid substitutions. These

data were consistent with allelic polymorphism, assuming

a single copy of the gene in the haploid genome, since like

most ixodid ticks, R. appendiculatus is diploid.

3.5. Reactivity of recombinant RIM36 with antisera from

cattle infested by ticks

The C-terminal 112 amino acids of RIM36 were

expressed in E. coli. Attempts were also made to express

full-length RIM36, but these were unsuccessful. The puri-

fied recombinant protein migrated with an Mr of approxi-

mately 22 kDa when analysed by SDS–PAGE gels (Fig. 3A,

lane 2), which was larger than the predicted size of 12–13

kDa. The aberrant migration may be attributable to the

unusual amino acid composition of this domain, which

comprised 18.8% proline. The purified recombinant

RIM36 protein reacted strongly with bovine C16 antisera

after Western blotting (Fig 3B, lane 2). Replicate strips of

the same Western blot used with C16 were reacted with

antisera from four B. indicus cattle, which had been experi-

mentally infested with R. appendiculatus. Representative

results using sera from one of the four cattle are shown in

Fig 3C. There was no apparent signal after 3 days, a visible

signal 14 days after initial exposure to ticks and a strong

reaction with recombinant RIM36 after 49 days (Fig. 3C,

lanes 1–3). Comparable Western blot results were obtained

from sera collected from the other three cattle experimen-

tally infested with 10–100 R. appendiculatus ticks (data not

shown). Antisera collected from seven cattle exposed to

field tick challenges in the Central Province and the Coast

Province of Kenya were also tested for reactivity with

recombinant RIM36 on Western blots. Antisera from one

animal challenged in the Central Province reacted at a

level comparable with the experimentally exposed cattle,

two other cattle from the Central Province and two from

the Coast Province gave weaker positive reactions and

two exhibited no apparent reaction. The result of the

Western blot using the most strongly reacting of these

bovine sera, which was taken from animal Tm194 after

exposure to field ticks for 2 months, is shown in Fig. 4A

(the signal is highlighted by an arrow). The blot result is

compared with replicate Western blot strips reacted with

antibodies taken at days 14, 21 and 35 from the experimen-

tally exposed animal BK233 (Fig. 4B). Additional signals

on the Western blot in Fig. 4 from bands of 44–45 kDa and

27–28 kDa are probably attributable to reactivity of the

antisera with contaminating E. coli proteins (Hengen, 1995).

3.6. Anti-RIM36 antibodies and sera from cattle infested

with ticks recognise a 36 kDa protein in R. appendiculatus

salivary gland and purified cement lysates

Antisera from rabbits immunised with purified recombi-

nant RIM36 recognised a protein of approximately 36 kDa

in Western blots of lysates of R. appendiculatus salivary

glands (Fig. 5A, lane 1) and also reacted strongly with puri-

fied recombinant RIM36 (Fig. 5A, lane 2). There were also

weaker reactions with salivary gland proteins of approxi-

mately 24, 38–40 and 100 kDa. The rabbit antisera produced

similar reactivity with a soluble extract of R. appendiculatus

cement proteins, although the cross-reactivity with the 100

kDa protein was relatively stronger (Fig. 5A, lane 3). Pre-

immunisation sera from the rabbits did not recognise either

the recombinant or native RIM36 on Western blots, and

there was also no reactivity with equal quantities of an

irrelevant antigen expressed in the same E. coli plasmid

R. Bishop et al. / International Journal for Parasitology 32 (2002) 833–842838

Fig. 3. Reactivity of recombinant RIM36 with bovine anti-sporozoite serum

C16, and antisera from cattle experimentally exposed to Rhipicephalus

appendiculatus tick infestation. Panel (A) (lane 2) shows a Coomassie

blue-stained SDS–PAGE gel containing recombinant C-terminal RIM36,

purified on nickel agarose (size markers are shown in lane 1). Panel (B)

shows a Western blot of 5 mg of recombinant RIM36 (lane 2) and equivalent

quantities of Theileria parva p32, a control antigen, expressed in the pQE30

system (lane 1) reacted with C16 antibody. Panel (C) shows a Western blot

of recombinant RIM36 incubated with bovine serum samples collected

from animal BK233, experimentally exposed to R. appendiculatus infesta-

tion, after 3 days (lane 1), 14 days (lane 2) and 49 days (lane 3).

vector and purified using the same procedure (data not

shown). Bovine sera collected from five different cattle

experimentally exposed to infestation with R. appendicula-

tus exhibited a strikingly similar pattern of reactivity with

salivary gland lysates and purified cement cone components

to that of the rabbit anti-recombinant RIM36 sera (represen-

tative results from one animal are shown in Fig. 5B, lanes 1

and 3). Bovine antisera collected prior to tick infestation

(day 0) did not exhibit detectable reactivity on Western

blots of R. appendiculatus salivary gland lysates, recombi-

nant RIM36, or purified cement extracts (data not shown).

3.7. Immuno-electron microscopy demonstrates localisation

of RIM36 to the secretory granules of the e cell of R.

appendiculatus type III salivary gland acini

Immuno-electron microscopy using the rabbit antisera

generated against recombinant C-terminal RIM36, resulted

in strong labelling of the secretory granules of e cells in

dissected salivary glands from female R. appendiculatus,

which had been fed for 4 days (Fig. 6). Similar labelling

was also apparent in e cells within salivary glands derived

from 4-day fed T. parva sporozoite-infected R. appendicu-

latus, but there was no labelling of the sporozoites (data not

shown). However, the granules from e cells from infected

salivary glands containing T. parva sporozoites were not

visible, due to complete occlusion of the cells by sporo-

zoites. Weak labelling, whose significance was unclear

since it was only marginally above background labelling

from pre-immune sera, was also observed in d-cell granules

(data not shown) but not in any of the other salivary gland

secretory cell types, such as the f-cell granules (see Fig. 6),

present in the type II and type III salivary gland acini.

4. Discussion

Our data provide strong evidence that the tick salivary

gland sequence we have designated RIM36 encodes a

protein present within the R. appendiculatus cement cone.

Since we did not test other tissues in this study, it is possible

the sequence might also be expressed elsewhere in the tick.

In addition to the recognition of a 36 kDa protein in extracts

of the cement cone by antisera raised against recombinant

RIM36, two regions of the predicted protein are identical in

amino acid sequence to an R. appendiculatus cement protein

described within a UK patent. The RIM36 protein is rich in

glycine, serine and leucine, which are prominent in the over-

R. Bishop et al. / International Journal for Parasitology 32 (2002) 833–842 839

Fig. 5. Reactivity of rabbit antisera specific-for recombinant RIM36 and

antisera from cattle exposed to tick infestation with salivary gland lysates

and extracts of purified cement material from Rhipicephalus appendicula-

tus. Panel (A) shows rabbit anti-RIM36 antisera reacted with blots of lysates

from dissected salivary glands of 4-day fed adult female R. appendiculatus

ticks (lane 1); recombinant C-terminal RIM36 (lane 2); a soluble extract of

purified R. appendiculatus cement cones collected from cattle infested with

equal numbers of male and female ticks (lane 3). Panel (B) shows a repli-

cate Western blot to that shown in panel (A), reacted with bovine antisera

from animal BK233 taken at day 21 after experimental infestation with R.

appendiculatus ticks.

Fig. 4. Reactivity of recombinant RIM36 with antisera from cattle exposed

to field tick challenge. Panel (A) shows a Western blot of recombinant

RIM36 reacted with bovine antisera from animal TM194 which had been

exposed to a field tick challenge for 60 days at Kakuzi farm Central

Province. Panel (B) shows replicate Western blot strips of recombinant

RIM36 reacted with bovine sera collected from experimentally infested

animal BK233 on days, 14 (lane 1), 21 (lane 2) and 35 (lane 3).

all amino acid content previously determined for the cement

substance of the related ixodid tick B. microplus (Kemp et

al., 1982). A 35 kDa protein, which is strongly recognised

by antisera from guinea pigs resistant to infestation with R.

appendiculatus, has been shown to be a major component of

both purified cement and salivary gland extracts of R.

appendiculatus (Shapiro et al., 1986). It seems probable

that this 35 kDa protein corresponds to RIM36. Features

of the RIM36 amino acid sequence, including the signal

peptide and the presence of glycine-rich repeats, are consis-

tent with a role as an extracellular matrix protein. A high

glycine content is a feature of vertebrate extracellular matrix

proteins, including keratin and collagen, due to the ability of

glycine to adopt a wide range of chain conformations.

RIM36 contains both a GLX triplet repeat domain at the

N-terminus and the amino acid triplet GSP within the

central septapeptide repeat block. Glycine triplet repeats,

and in particular glycine/proline triplet repeats, are typical

of both vertebrate and invertebrate collagens (Mariyama et

al., 1994; Kramer, 1994; Sicot et al., 1997). It has even been

suggested that tick cement components may mimic compo-

nents of vertebrate skin in order to use host-derived

enzymes during the cement hardening process. A recently

described extracellular matrix-like protein present within

the salivary gland of the ixodid tick Haemaphysalis long-

icornis also contains glycine triplet repeats (GXX) reminis-

cent of those present in collagen (Mulenga et al., 1999). The

tick cement cone is primarily proteinaceous, but also

contains some carbohydrate and lipid. There were no

predicted N-linked glycosylation sites within RIM36, but

the septapeptide repeat contains multiple copies of the

motif SGFG which is associated with post translational

additional of glycosaminoglycans, such as chondroitin

sulphate (Bourdon et al., 1987). However the additional

criterion for this motif, of acidic residues being located

between positions 22 and 24 (Bourdon et al., 1987) was

not satisfied. Further data will therefore be required to deter-

mine whether these putative glycosaminoglycan sites are

functional.

Immuno-electron microscopy data localise RIM36 predo-

minantly to the secretory granules of the e cells of the type III

acinus of the tick salivary gland. Disappearance of cytoplas-

R. Bishop et al. / International Journal for Parasitology 32 (2002) 833–842840

Fig. 6. Localisation of RIM36 within the e cell granules (E) of Rhipicephalus appendiculatus salivary glands by immuno-electron microscopy. Granules in the

adjacent f cell (F) are unlabelled. Dissected salivary glands from uninfected female ticks after 4 days of feeding are shown. Scale bar: 0.5 mm.

mic secretory granules in correlation with cement deposition

in B. microplus, and comparative histochemical data from H.

spinigera have implicated the secretory a cells of acinus type

II, and d and e cells of acinus type III as the salivary gland

precursor cells of cement (Binnington, 1978, reviewed by

Binnington and Kemp, 1980). Our immuno-electron micro-

scopy data (Fig. 6) provide the most direct evidence to date

that one function of the e cell is deposition of components of

tick cement. The localisation of RIM36 to e cells differs from

that of a 90 kDa polypeptide of R. appendiculatus, which is a

conserved component of cement in several species of ixodid

ticks (Jaworski et al., 1992). The 90 kDa molecule was found

using protein A-gold immuno-cytochemistry to be present in

the a and d cells, as well as the e cell granules of adult feeding

R. appendiculatus (Venable et al., 1986). The fact that RIM36

was still expressed in significant quantities in 4-day fed ticks

may indicate that it is a component of the external cortex of

the cement, which is secreted for up to 96 h after attachment

(Moorhouse and Tatchell, 1966). Theileria parva sporozoites

are located predominantly in the secretory granules of the e

cell of R. appendiculatus and production of mature T. parva

sporozoites is maximal in the salivary glands of 3–4-day fed

ticks (reviewed by Shaw and Young, 1994). It is tempting to

speculate T. parva sporozoites and RIM36 could be co-

secreted into the host, since there is some evidence to suggest

that sporozoites are released gradually from an individual

infected cell in a manner comparable to the release of secre-

tory granules by apocrine secretion (Shaw and Young, 1994).

Our data demonstrate that the C-terminal domain of the

RIM36 protein, which lacks the glycine-rich repeat regions,

evokes a strong antibody response in four different B. indi-

cus cattle experimentally infested with R. appendiculatus,

and also in cattle exposed to ticks in the field. Previous

research demonstrated that B. microplus cement apparently

did not provoke a detectable host immune response (Tatch-

ell and Moorhouse, 1968). The polymorphism which we

have observed in the RIM36 gene, involving a deletion of

coding sequence between two R. appendiculatus stocks,

might be due to selection imposed by the host immune

response. Our experiments suggest that the cement protein

RIM36 is an immunodominant molecule in cattle. The

immune response engendered may be a consequence of

the fact that B. indicus, which evolved in Asia and entered

Africa in the last few thousand years (reviewed by Hanotte

et al., 2000), is a relatively recently acquired host for this

African tick species. The strong antibody responses induced

by this molecule suggest that it may have potential as a

diagnostic marker for detection of cattle which have been

exposed to feeding R. appendiculatus ticks.

Acknowledgements

We are grateful for the excellent technical assistance of

the ILRI tick unit staff. This is ILRI publication number

200154.

References

Bailey, K.P., 1960. Notes on the rearing of Rhipicephalus appendiculatus

and their infection with Theileria parva for experimental transmission.

Bull. Epizoot. Dis. Africa 8, 33–43.

Barker, S.C., 1998. Distinguishing species and populations of rhipicepha-

line ticks with ITS2 ribosomal RNA. J. Parasitol. 84, 887–92.

Berryman, M.A., Rodewald, R.A., 1990. An enhanced method for post-

embedding immunocytochemical staining which preserves cell

membranes. J. Histochem. Cytochem. 38, 159–70.

Binnington, K.C., 1978. Sequential changes in salivary gland structure

during attachment and feeding of the cattle tick Boophilus microplus.

Int. J. Parasitol. 8, 97–115.

Binnington, K.C., Kemp, D.H., 1980. Role of tick salivary glands in feeding

and disease transmission. Adv. Parasitol. 18, 316–40.

Bourdon, M.A., Krusius, T., Cambell, S., Schwarz, N.B., Ruoslahti, E.,

1987. Identification and synthesis of a recognition signal for the attach-

ment of glycosaminoglycans to proteins. Proc. Natl. Acad. Sci. USA 84,

3194–8.

Brown, S.J., Askenase, P.W., 1986. Amblyomma americanum: physico-

chemical isolation of a protein derived from the tick salivary gland

that is capable of inducing immune resistance in guinea pigs. Exp.

Parasitol. 62, 40–50.

Burleigh, B.A., Wells, C.W., Clarke, M.W., Gardiner, P.R., 1993. An inte-

gral membrane glycoprotein associated with an endocytic compartment

of Trypanosoma vivax: identification and partial characterization. J.

Cell Biol. 120, 339–52.

Hanotte, O., Tawah, C.L., Bradley, D.G., Okomo, M., Verjee, Y., Ochieng,

J., Rege, J.E.O., 2000. Geographic distribution and frequency of a taur-

ine Bos taurus and an indicine Bos indicus Y specific allele amongst

sub-Saharan cattle breeds. Mol. Ecol. 9, 387–96.

Hengen, P.N., 1995. Purification of His-Tag fusion proteins from Escher-

ichia coli. Trends Biochem. Sci. 20, 285–6.

Iams, K.P., Hall, R., Webster, P., Musoke, A., 1990. Identification of (gt11

clones encoding the major antigenic determinants expressed by Thei-

leria parva sporozoites. Infect. Immun. 58, 1828–34.

Jaworski, D.C., Rosell, R., Coons, L.B., Needham, G.R., 1992. Tick (Acari:

Ixodidae) attachment cement and salivary gland cells contain similar

immunoreactive polypeptides. J. Med. Entomol. 29, 305–9.

Kemp, D.H., Stone, B.F., Binnington, K.C., 1982. In: Obenchain, F.D.,

Galun, R. (Eds.). Physiology of Ticks, Pergamon, Oxford/New York,

pp. 119–68.

Kramer, J.M., 1994. Structures and functions of collagens in Caenorhab-

ditis elegans. FASEB J. 8, 329–36.

Laemmli, U.K., 1970. Cleavage of structural proteins during the assembly

of the head of the bacteriophage T4. Nature 227, 680–5.

Mariyama, M., Leoinen, A., Mochizuki, T., Tryggvason, K., Reeders, S.T.,

1994. Complete primary structure of the human alpha 3 (IV) collagen

chain. Co expression of the alpha 3 (IV) and alpha 4 (IV) collagen

chains in human tissues. J. Biol. Chem. 269, 23013–7.

McKosker, P.J., 1979. Global aspects of the management and control of

ticks of veterinary importance. In: Rodriguez, J. (Ed.). Recent Advances

in Acarology, Vol. 2. Academic Press, London, pp. 45–53.

Moorhouse, D.E., Tatchell, R.J., 1966. The feeding processes of the cattle

tick Boophilus microplus (Canestrini): A study in host-parasite rela-

tions. Part I. Attachment to the host. Parasitology 56, 623–32.

Mulenga, A., Sugimoto, C., Sako, Y., Ohashi, K., Musoke, A., Morzaria, S.,

Onuma, M., 1999. Molecular characterization of a Haemaphysalis long-

icornis tick salivary gland-associated 29-Kilodalton protein and its effect

as a vaccine against tick infestation in rabbits. Infect. Immun. 67, 1652–8.

Nene, V., Iams, K.P., Gobright, E., Musoke, A.J., 1992. Characterisation of

the gene encoding a candidate vaccine antigen of Theileria parva spor-

ozoites. Mol. Biochem. Parasitol. 51, 17–28.

Nielsen, H., Englebrecht, J., Brunak, S., von Heijne, G., 1997. Identification

of prokaryotic and eukaryotic signal peptides and prediction of their

cleavage sites. Protein Eng. 10, 1–6.

R. Bishop et al. / International Journal for Parasitology 32 (2002) 833–842 841

Nolan, J., Schnitzerling, H.J., 1986. Drug resistance in arthropod parasites.

In: Cambell, W.C., Rew, R.W. (Eds.). Chemotherapy of Parasitic

Diseases, Plenum, New York, pp. 603–20.

Norval, R.A.I., Perry, B.D., Young, A.S., 1992. The Epidemiology of Thei-

leriosis in Africa, Academic Press, London.

Nuttall, P.A., 1998. Displaced tick–parasite interactions at the host inter-

face. Parasitology 116, S65–72.

Nuttall, P.A., Jones, L.D., Labuda, M., Kaufman, W.R., 1994. Adaptations

of arboviruses to ticks. J. Med. Entomol. 31, 1–9.

Opdebeeck, J.P., 1994. Vaccines against blood-sucking arthropods. Vet.

Parasitol. 54, 205–22.

Ozaki, L.S., Mattei, D., Jendoubi, M., Druhle, P., Blisnick, T., Guillote, M.,

Puijalon, O., Pereira da Silva, M., 1986. Plaque antibody selection:

rapid immunological analysis of a large number of phage clones posi-

tive to sera raised to Plasmodium falciparum antigens. J. Immunol.

Methods 89, 213–9.

Pelle, R., Murphy, N.B., 1993. Northern hybridization: rapid and simple

electrophoretic conditions. Nucleic Acids Res. 21, 2783–4.

Riding, G.A., Jarmey, J., Mckenna, R.V., Pearson, R., Cobon, G.S., Will-

adsen, P., 1994. A protective concealed antigen from Boophilus micro-

plus: purification, localization and possible function. J. Immunol. 153,

5158–66.

Sambrook, J., Fritsch, E.F., Maniatis, T., 1989. Molecular cloning: A

Laboratory Manual, 2nd Edition. Cold Spring Harbor Laboratory

Press, Cold Spring Harbor, NY.

Shapiro, S.Z., Voigt, W.P., Fujisaki, K., 1986. Tick antigens recognised by

serum from a guinea pig resistant to infestation with the tick Rhipice-

phalus appendiculatus. J. Parasitol. 72, 454–63.

Shapiro, S.Z., Buscher, G., Dobbelaere, D.A.E., 1987. Acquired resistance

to Rhipicephalus appendiculatus (Acari: Ixodidae): identification of an

antigen eliciting resistance in rabbits. J. Med. Entomol. 24, 147–54.

Shapiro, S.Z., Voigt, W.P., Ellis, J.A., 1989. Acquired resistance to Ixodid

ticks induced by tick cement antigen. Exp. Appl. Acarol. 7, 33–41.

Shaw, M.K., Young, A.S., 1994. The biology of Theileria species in ixodid

ticks in relation to parasite transmission. Advances in Disease Vector

Research, Springer, New York, pp. 23–63.

Sicot, F.X., Exposito, J.Y., Masselot, M., Garrone, R., Deutsch, F., 1997.

Cloning of an annelid fibrillar collagen gene and phylogenetic analysis

of vertebrate and invertebrate collagens. Eur. J. Biochem. 246, 50–56.

Sonenshine, D.E., 1993. Biology of Ticks, Vol. 2, Oxford University Press,

Oxford.

Tatchell, R.J., Moorhouse, D.E., 1968. The feeding processes of the cattle

tick Boophilus microplus (Canestrini). II. The sequence of host-tissue

changes. Parasitology 58, 441–59.

Venable, J.H., Webster, P., Shapiro, S.G., Voight, W.P., 1986. An immu-

nocytochemical marker for complex granules of tick salivary glands

which traces e-granule shedding to interstitial labyrinthine spaces.

Tissue Cell 18, 765–81.

Von Heijne, G., 1986. A new method for prediction of signal cleavage sites.

Nucleic Acids Res. 14, 4683–90.

Willadsen, P., 1980. Immunity to ticks. Adv. Parasitol. 18, 293–313.

Willadsen, P., 1997. Novel vaccines for ectoparasites. Vet. Parasitol. 71,

209–22.

Willadsen, P., Bird, P., Cobon, G.S., Hungerford, J., 1995. Commercialisa-

tion of a recombinant vaccine against Boophilus microplus. Parasitol-

ogy 110, S43–50.

R. Bishop et al. / International Journal for Parasitology 32 (2002) 833–842842