Molecular characterization of a tsetse fly midgut proteolytic lectin that mediates differentiation...

9
Insect Biochemistry and Molecular Biology Insect Biochemistry and Molecular Biology 36 (2006) 344–352 Molecular characterization of a tsetse fly midgut proteolytic lectin that mediates differentiation of African trypanosomes Laila U. Abubakar a,b , Wallace D. Bulimo b , Francis J. Mulaa b , Ellie O. Osir a, a International Centre of Insect Physiology and Ecology (ICIPE), P.O. Box 30772, Nairobi, Kenya b Department of Biochemistry, University of Nairobi, P.O. Box 30197, Nairobi, Kenya Abstract Differentiation of bloodstream-form trypanosomes into procyclic (midgut) forms is an important first step in the establishment of an infection within the tsetse fly. This complex process is mediated by a wide variety of factors, including those associated with the vector itself, the trypanosomes and the bloodmeal. As part of an on-going project in our laboratory, we recently isolated and characterized a bloodmeal-induced molecule with both lectin and trypsin activities from midguts of the tsetse fly, Glossina longipennis [Osir, E.O., Abubakar, L., Imbuga, M.O., 1995. Purification and characterization of a midgut lectin–trypsin complex from the tsetse fly, Glossina longipennis. Parasitol. Res. 81, 276–281]. The protein (lectin–trypsin complex) was found to be capable of stimulating differentiation of bloodstream trypanosomes in vitro. Using polyclonal antibodies to the complex, we screened a G. fuscipes fuscipes cDNA midgut expression library and identified a putative proteolytic lectin gene. The cDNA encodes a putative mature polypeptide with 274 amino acids (designated Glossina proteolytic lectin, Gpl). The deduced amino acid sequence includes a hydrophobic signal peptide and a highly conserved N-terminal sequence motif. The typical features of serine protease trypsin family of proteins found in the sequence include the His/Asp/Ser active site triad with the conserved residues surrounding it, three pairs of cysteine residues for disulfide bridges and an aspartate residue at the specificity pocket. Expression of the gene in a bacterial expression system yielded a protein (M r 32,500). The recombinant protein (Gpl) bound D(+) glucosamine and agglutinated bloodstream-form trypanosomes and rabbit red blood cells. In addition, the protein was found to be capable of inducing transformation of bloodstream-form trypanosomes into procyclic forms in vitro. Antibodies raised against the recombinant protein showed cross-reactivity with the a subunit of the lectin–trypsin complex. These results support our earlier hypothesis that this molecule is involved in the establishment of trypanosome infections in tsetse flies. r 2006 Elsevier Ltd. All rights reserved. Keywords: Tsetse; Glossina; Lectin; Trypsin; Trypanosomes; Transformation 1. Introduction The midgut is an important site for bloodmeal digestion and pathogen transmission in all hematophagous insects. For example, in tsetse flies (Diptera: Glosinidiae), the ingestion of a bloodmeal initiates numerous structural, biochemical and developmental changes within the midgut, including the formation of a bilayered peritrophic mem- brane (Peters, 1992; Lehane and Msangi, 1991), secretion of proteases (Imbuga et al., 1992), lectins (Maudlin, 1991; Stiles et al., 1990), trypanolysins (Stiles et al., 1990; Osir et al., 1999) and other yet unknown factors. These factors, whose concentrations reach peak levels after 48–72 h post- bloodmeal, create a hostile environment for the ingested trypanosomes (Stiles et al., 1990; Onyango, 1993; Abuba- kar et al., 1995). Consequently, the majority of the ingested parasites, including the stumpy bloodstream forms that are believed to be preadapted for differentiation (Vickerman, 1985) are eliminated. The few surviving trypanosomes either circumnavigate or penetrate the peritrophic mem- brane to enter the ectoperitrophic space where they transform and establish themselves as procyclic (midgut) forms (Maudlin, 1985; Maudlin and Welburn, 1987). The successful establishment of trypanosome infections in tsetse flies is, therefore, greatly dependent on these factors. Recent studies on immune responsive genes showed that the tsetse immune system could discriminate not only ARTICLE IN PRESS www.elsevier.com/locate/ibmb 0965-1748/$ - see front matter r 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.ibmb.2006.01.010 Corresponding author. Tel.: +254 02 861173; fax: +254 02 803360. E-mail address: [email protected] (E.O. Osir).

Transcript of Molecular characterization of a tsetse fly midgut proteolytic lectin that mediates differentiation...

ARTICLE IN PRESS

InsectBiochemistry

andMolecularBiology

0965-1748/$ - se

doi:10.1016/j.ib

�CorrespondE-mail addr

Insect Biochemistry and Molecular Biology 36 (2006) 344–352

www.elsevier.com/locate/ibmb

Molecular characterization of a tsetse fly midgut proteolytic lectin thatmediates differentiation of African trypanosomes

Laila U. Abubakara,b, Wallace D. Bulimob, Francis J. Mulaab, Ellie O. Osira,�

aInternational Centre of Insect Physiology and Ecology (ICIPE), P.O. Box 30772, Nairobi, KenyabDepartment of Biochemistry, University of Nairobi, P.O. Box 30197, Nairobi, Kenya

Abstract

Differentiation of bloodstream-form trypanosomes into procyclic (midgut) forms is an important first step in the establishment of an

infection within the tsetse fly. This complex process is mediated by a wide variety of factors, including those associated with the vector

itself, the trypanosomes and the bloodmeal. As part of an on-going project in our laboratory, we recently isolated and characterized a

bloodmeal-induced molecule with both lectin and trypsin activities from midguts of the tsetse fly, Glossina longipennis [Osir, E.O.,

Abubakar, L., Imbuga, M.O., 1995. Purification and characterization of a midgut lectin–trypsin complex from the tsetse fly, Glossina

longipennis. Parasitol. Res. 81, 276–281]. The protein (lectin–trypsin complex) was found to be capable of stimulating differentiation of

bloodstream trypanosomes in vitro. Using polyclonal antibodies to the complex, we screened a G. fuscipes fuscipes cDNA midgut

expression library and identified a putative proteolytic lectin gene. The cDNA encodes a putative mature polypeptide with 274 amino

acids (designated Glossina proteolytic lectin, Gpl). The deduced amino acid sequence includes a hydrophobic signal peptide and a highly

conserved N-terminal sequence motif. The typical features of serine protease trypsin family of proteins found in the sequence include the

His/Asp/Ser active site triad with the conserved residues surrounding it, three pairs of cysteine residues for disulfide bridges and an

aspartate residue at the specificity pocket. Expression of the gene in a bacterial expression system yielded a protein (Mr �32,500). The

recombinant protein (Gpl) bound D(+) glucosamine and agglutinated bloodstream-form trypanosomes and rabbit red blood cells. In

addition, the protein was found to be capable of inducing transformation of bloodstream-form trypanosomes into procyclic forms in

vitro. Antibodies raised against the recombinant protein showed cross-reactivity with the a subunit of the lectin–trypsin complex. These

results support our earlier hypothesis that this molecule is involved in the establishment of trypanosome infections in tsetse flies.

r 2006 Elsevier Ltd. All rights reserved.

Keywords: Tsetse; Glossina; Lectin; Trypsin; Trypanosomes; Transformation

1. Introduction

The midgut is an important site for bloodmeal digestionand pathogen transmission in all hematophagous insects.For example, in tsetse flies (Diptera: Glosinidiae), theingestion of a bloodmeal initiates numerous structural,biochemical and developmental changes within the midgut,including the formation of a bilayered peritrophic mem-brane (Peters, 1992; Lehane and Msangi, 1991), secretionof proteases (Imbuga et al., 1992), lectins (Maudlin, 1991;Stiles et al., 1990), trypanolysins (Stiles et al., 1990; Osiret al., 1999) and other yet unknown factors. These factors,

e front matter r 2006 Elsevier Ltd. All rights reserved.

mb.2006.01.010

ing author. Tel.: +254 02 861173; fax: +254 02 803360.

ess: [email protected] (E.O. Osir).

whose concentrations reach peak levels after 48–72 h post-bloodmeal, create a hostile environment for the ingestedtrypanosomes (Stiles et al., 1990; Onyango, 1993; Abuba-kar et al., 1995). Consequently, the majority of the ingestedparasites, including the stumpy bloodstream forms that arebelieved to be preadapted for differentiation (Vickerman,1985) are eliminated. The few surviving trypanosomeseither circumnavigate or penetrate the peritrophic mem-brane to enter the ectoperitrophic space where theytransform and establish themselves as procyclic (midgut)forms (Maudlin, 1985; Maudlin and Welburn, 1987). Thesuccessful establishment of trypanosome infections in tsetseflies is, therefore, greatly dependent on these factors.Recent studies on immune responsive genes showed thatthe tsetse immune system could discriminate not only

ARTICLE IN PRESSL.U. Abubakar et al. / Insect Biochemistry and Molecular Biology 36 (2006) 344–352 345

between molecular signals specific for bacteria andtrypanosome infections but also between different lifestages of trypanosomes (Hao et al., 2001). However,despite the critical role that differentiation plays intransmission of trypanosome infection, relatively littleinformation is available about the genes involved in theprocess and their regulation.

Initial evidence for the presence of a tsetse midgut factorthat stimulates the transformation of bloodstream-form trypanosomes was obtained by incubating parasitizedblood with crude midgut homogenates (Imbuga et al.,1992). These experiments also proposed the possibleinvolvement of trypsin or trypsin-like enzymes in thetransformation of bloodstream-form trypanosomesinto procyclic forms. In subsequent studies, a closecorrelation between the trypsin-like enzyme and lectinactivities was established (Osir et al., 1993). However, itwas not until we isolated a protein with two non-covalently linked subunits and both lectin and trypsin-likeproperties from tsetse midgut that the picturebecame clearer (Osir et al., 1995). The holoprotein(lectin–trypsin complex) comprised of two non-covalentlybound subunits (a and b) was later shown to be capableof stimulating transformation of bloodstream-formtrypanosomes in vitro (Osir and Abubakar, unpub-lished observation). Despite these findings, we could notestablish whether the lectin and trypsin activitieswere associated with one or both of the subunits. More-over, it remained unclear whether both subunits werenecessary for the transformation process. To gainfurther insight into the functions of the lectin–trypsin complex, we recently initiated studies aimed atcharacterizing the genes that encode the molecules. Thisreport deals with cloning and expression of the gene for alectin with proteolytic activity from Glossina fuscipes

fuscipes. We also show that the recombinant protein iscapable of stimulating the transformation of bloodstream-form Trypanosoma brucei brucei into procyclic formsin vitro.

2. Materials and methods

2.1. Experimental animals, insects and trypanosomes

Adult male tsetse flies, G. f. fuscipes (Diptera: Glossini-

diae), male Wistar rats and New Zealand white rabbitswere supplied by the Animal Breeding and QuarantineUnit (ABQU) at ICIPE. The flies were maintained aspreviously described (Abubakar et al., 1995). Bloodstreamforms of pleomorphic T. b. brucei ILTaT 1.4 (Miller andTurner, 1981) were used to infect male Wistar rats.Monoclonal antibodies (Mabs) specific to the majorprocyclin surface molecule (‘EP’ forms containing glutamicacid–proline repeats) were kindly supplied by InternationalLivestock Research Institute, Nairobi, Kenya, and Uni-versity of Victoria, Canada.

2.2. Construction of a tsetse midgut cDNA expression

library and screening with lectin–trypsin complex antibodies

Male teneral G. f. fuscipes (24 h after emergence) werefed twice on rabbit blood through an artificial siliconemembrane and subsequently starved for 72 h. Midgutswere carefully dissected from 20 flies and total RNA wasextracted using an RNA extraction kit (Promega, Madi-son, WI, USA). cDNA was prepared using a SMARTTM

cDNA library construction kit following the instructionssupplied by the manufacturer (Clontech, CA, USA). ThecDNA was digested using Sfi1 and ligated into thelTriplEx2 Vector (Clontech) followed by packaging usingGigapacks III Gold packaging extract according to themanufacture’s protocol (Stratagene, la Jolla, CA, USA).The lTriplEx2 library was converted to the plasmid

form (pTriplEx2) in Escherichia coli BM25.8 according tothe manufacturer’s protocols. Polyclonal antibodiesagainst G. f. fuscipes lectin–trypsin complex were preparedas previously described (Osir et al., 1986, 1995). The cDNAlibrary was screened with the polyclonal antibodiesaccording to Sambrook et al. (1989). Briefly, the serumwas preadsorbed with E. coli lysate at 1:10 dilution toreduce non-specific binding to bacterial proteins. Thepreadsorbed serum was further diluted to a workingdilution of 1:500. The library (in E. coli strain BM25.8)was plated out on an agar plate containing Ampicillin and�4000–5000 colonies were transferred onto nitrocellulosemembrane (Hybond C, Amersham, Cleveland, OH, USA).The filters were blocked with 5% non-fat dry milk andincubated overnight with antibodies against the lectin–trypsin complex. After washing, the filters were incubatedwith goat anti-rabbit conjugated alkaline phosphatasesecondary antibody and positive clones visualized byincubating with 5-bromo-4-chloro-3-indolyl-phosphate(BCIP) and nitroblue tetrazolium (NBT) substrates.

2.3. Plasmid preparation and sequencing

The positive colonies were selected and the plasmidsextracted from them using Qiagen Plasmid Miniprep Kitaccording to the manufacturer’s instructions and trans-formed into competent JM109 E. coli. Plasmid DNA wasextracted from the JM109 colonies growing on Agar/LB/Ampicillin plates and subjected to automated sequencingon an ABI Prism 371 DNA sequencer (Perkin Elmer,USA).

2.4. Subcloning of Gpl insert for protein expression

A full-length transcript (designated Glossina proteolyticlectin, Gpl) was subcloned into pET-28 bacterial expressionvector (Novagen, Madison, WI) to generate the recombi-nant protein. Using a PCR-based strategy, �3 mg of pET-28 plasmid DNA (containing 6�His-tags) was linearizedby digesting with BamH1 and HindIII restriction enzymes(3 h, 37 1C) and purified using the QIAquick Gel Extraction

ARTICLE IN PRESSL.U. Abubakar et al. / Insect Biochemistry and Molecular Biology 36 (2006) 344–352346

Kit (Qiagen). Similarly, two PCR primers that specificallyincorporated unique BamH1 and HindIII restriction sitesto the 50 and 30 ends, respectively, were used to amplifythe Gpl gene. The amplified fragment was ligated into thepET-28 vector, and transformed into competent E. coli

BL-21(DE3) cells. Positive clones growing on Agar/LB/Ampicillin plates were screened by PCR using the specificprimers.

2.5. Large-scale expression and purification of recombinant

protein

To prepare the recombinant protein for functionalstudies, positive Gpl clone was grown in LB/Ampicillinmedium to an A600 of 0.6 before the induction of proteinexpression with isopropylthio-b-D-galactosidase (IPTG;1mM final concentration). The cell lysates were centrifuged(10,000g, 15min, 4 1C) using a JA-25.5 rotor and aBeckman Avanti J-25 centrifuge and the resultant super-natant fraction was passed through a metal ion affinitycolumn (Talon IMAC resin, CLONTECH, Palo Alto, CA)that had been equilibrated with phospate buffered saline(PBS: 50mM NaH2PO4, pH 7, 700mM NaCl, 6mMguanidinium hydrochloride). The flow rate was maintainedat 10ml/h. The column was washed with 10 columnvolumes of PBS buffer and bound protein eluted in PBScontaining 0.2M imidazole. The eluted fractions wereconcentrated and dialyzed overnight against PBS. Proteinwere estimated using the bicinchoninic (BCA) protein assayreagent (Pierce, Rockford, IL, USA) with bovine serumalbumin as the standard. Electrophoresis of the samples onSDS-polyacrylamide gels was carried out on gradient gels(4–15%) according to Laemmli (1970).

2.6. Agglutination of trypanosomes and red blood cells by

the recombinant protein

The procedures for isolation of bloodstream-formT. b. brucei, rabbit red blood cells (RBC) and agglutinationassay have been previously described (Abubakar et al.,1995; Osir et al., 1995). Two-fold serial dilutions ofrecombinant E. coli extracts or bound column fractionwere prepared in phosphate buffered saline (PBS: 0.15Msodium phosphate/0.15M NaCl, pH 8.0), were mixed withan equal volume of freshly isolated T. b. brucei (�5� 106

parasites/ml) or 2% RBC (�107 cells/ml) and incubated(2 h, 27 1C). Agglutination titers were expressed as thereciprocals of the highest dilution, where completeagglutination of the parasites or RBCs was observed. Ina parallel experiment, 100mM D-glucosamine (D(+)-Glucosamine, Sigma, St. Louis, MO, USA) was mixedwith E. coli extracts or bound column fraction. Afterincubation (27 1C, 10min), bloodstream-form trypano-somes or RBC were added and agglutinations scored asdescribed above.

2.7. Enzyme and trypanosome transformation studies

Trypsin activity of the recombinant protein was deter-mined using a chromogenic substrate, carbo-benzoxy-val-gly-arg-4-nitranilide acetate (Chromozym-TRY, Boeh-ringer-Mannheim, FRG) (Imbuga et al., 1992). The in vitrotrypanosome transformation assay system used has alreadybeen described (Nguu et al., 1996; Imbuga et al., 1992). Nomedia were used in the incubation mixtures. Briefly,parasitized blood (�5� 106 trypanosomes/ml) was mixedwith the recombinant protein (�4.5 mg protein) in a totalvolume of 600 ml and incubated at 27 1C. At 4 h intervals,each mixture was vortexed and a 20 ml aliquot withdrawnfor preparation of thin wet smears. The wet smears wereair-dried, fixed in methanol before staining with Giemsa’sstain and examination using a Dilux compound microscope(Leitz Wetzlar, Germany). For each time point, 3–4 groupsof 100 trypanosomes each were counted from the smearsand classified according to their morphology as blood-stream, transition or midgut forms (Imbuga et al., 1992).The transformation process was further confirmed byindirect immunoflourescence assay using anti-procyclinantibodies.

2.8. Immunological studies

Antibodies against Gpl were raised in a New Zealandwhite rabbit using a previously described protocol (Osiret al., 1986). Isolated lectin–trypsin complex (Osir et al.,1995) was separated by SDS-PAGE (Laemmli, 1970) andimmunoblotting experiments performed using antibodiesagainst the recombinant protein (Osir et al., 1989).

2.9. Database search and sequence analysis

Sequence analyses against databases were performedusing the BLAST computer search program in theNational Center for Biotechnology Information (NCBI,Bethesda, MD, USA) (Altschul et al., 1997). Sequencealignment and statistical analyses of alignment wereperformed with MultAlign program (Corpet, 1988).

3. Results

3.1. Isolation of lectin cDNA from the midgut library

A full-length cDNA expression library containing�1.0� 107 independent clones was successfully constructedin lTriplEx2 phage vector from the total tsetse midgutRNA. The resultant amplified library yielded a titer ofX1010 plaques forming unit per milliliter (pfu/ml). Screen-ing the cDNA expression library (pTriplEx2) with apolyclonal antibody, raised against the native lectin–trypsin complex, yielded 12 positive colonies. To identifythe lectin activity insert(s), positive clones (transformed inE. coli strain JM109) were overexpressed and the recombi-nant lysates were assayed for agglutination activity using

ARTICLE IN PRESSL.U. Abubakar et al. / Insect Biochemistry and Molecular Biology 36 (2006) 344–352 347

both bloodstream-form trypanosomes and RBCs. Thelysates from three colonies agglutinated both bloodstreamtrypanosomes and RBCs. Plasmid DNA was preparedfrom the three positive clones and subjected to automatedDNA sequencing on both strands. Out of the three clonessequenced only one clone carried a full-length cDNAinsert. This insert (designated Glossina proteolytic lectin

Fig. 1. The nucleotide and deduced amino acid sequences o

Gpl-plasmid), comprising 933 bp, was selected for furtheranalysis.

3.2. Sequence analysis

The nucleotide sequence of Gpl-plasmid and the aminoacid sequence derived from it (GENBANK accession

f Gpl, isolated from G. f. fuscipes midgut cDNA library.

ARTICLE IN PRESSL.U. Abubakar et al. / Insect Biochemistry and Molecular Biology 36 (2006) 344–352348

number: AF525314) are shown in Fig. 1. The 933 bp cDNAcontained 35 nucleotides upstream of the initiation ATGcodon which represents the 50 untranslated region (UTR).The open reading frame (ORF) terminates with TGT, 823bases downstream from the initiation codon. Therefore, theinsert codes for a polypeptide of 274 amino acids. There are47 nucleotides after the stop codon, which represents the 30

UTR. Present within the 30 UTR is the putative poly-adenylation signal (aataaa), and the poly (A) tail occurs 14nucleotides downstream from the signal.

When the protein sequence was scanned to identifyfunctional domains using PrositeScan algorithm, severalstructural requirements pivotal to serine proteases wereidentified. A serine active site (GACNGDSGGPMV)belonging to the serine protease trypsin family wasidentified between amino acids 213 and 224 with arandomized probability of 7.319e�08. Similarly, a histidineactive site belonging to the serine protease trypsin familywas identified at amino acids 68–73 (LTAGHC) with arandomized probability of 2.601e�07. The calculated

Fig. 2. Comparative sequence analysis. The full sequence of Gpl was used to

alignments performed using MultAlign. Residues that are identical are shown i

m; the six conserved cysteine residues are indicated by ‘*’ and the substrate sp

Glossina serine protease 1 (AF252868); Gsp2—Glossina serine protease 2 (AF

serine protease 2 (AF074956).

molecular mass of the predicted mature protein sequenceis 29,179Da with an isoelectric point (pI) of 4.79.

3.3. Comparative sequence analysis

The full sequence of the putative protein encoded by Gpl-plasmid was used to search for similarity to other knowngenes in the GENBANK using BLASTP (Altschul et al.,1997) as well as the SWISS-PROT and TrEMBL Proteinknowledge bases at the Expert Protein Analysis System(ExPASy) Molecular Biology Server. The highest similarityscore was with chymotrypsin-like serine protease precursorfrom G. m. morsitans at 89% identities, 92% positives andan expectation of e�140. The second highest similarity scorewas with midgut-specific serine protease from Stomoxys

calcitrans at 58% identities, 69% positives with anexpectation of 8e�81.Multiple sequence alignment of the translated sequence

of Gpl with four homologous insect serine proteasesindicated that the active site residues, His-57, Asp-102

search for similar sequences in GenBank using BLASTP, and multiple

n red. The residues that are in the catalytic triad (H, D, S) are indicated by

ecificity sites by +. Gpl—Glossina proteolytic lectin (AF525314); Gsp1—

252869); Ssp1—Stomoxys serine protease 1 (AF074955); Ssp2—Stomoxys

ARTICLE IN PRESS

Table 1

Agglutination of bloodstream-form trypanosomes and rabbit red blood

cellsa

Samples Agglutination titersb Trypsin activity

(units� 10�4)

Bloodstream

parasites

RBCs

Control (PBS) 0 0 0

Control (E. coli JM109) 4 2 0.303

Gpl (unexpressed) 4 2 0.425

Gpl (expressed) 512 128 6.025

aDoubling serial dilutions of E. coli lysate with or without the insert

were incubated (27 1C, 1 h) with either bloodstream-form T. b. brucei

(�5� 106 parasites/ml) or 2% rabbit RBCs. Titers are expressed as

reciprocals of end point serial dilution that gave positive agglutination.

The samples were also assayed for trypsin activity using Chromozym-

TRY as substrate.bEnd points are expressed as reciprocals of the dilutions.

Table 2

Effects of carbohydrates on agglutination activity of Gpl1a

Type of sugar Agglutination titersb

Bloodstream parasites RBCs

Control 2048 1024

D-glucosamine 2 2

D-glucose 2048 1024

D-galactose 2048 1024

L.U. Abubakar et al. / Insect Biochemistry and Molecular Biology 36 (2006) 344–352 349

and Ser-195 (using the bovine chymotrypsinogen Anumbering system) (Brown and Hartley, 1966) and theregions around these residues were highly conserved(Fig. 2; denoted by m). In addition, the six cysteines thatare thought to contribute to the three disulfide bondscommonly found in invertebrate serine proteases are foundat conserved sites (Fig. 2; denoted by *). The Asp-189residue, which is diagnostic of tryptic specificity, wassimilarly conserved (Fig. 2; denoted by +) (Stroud et al.,1974). Likewise, the highly conserved N-terminal sequenceGRI(V)NG was present from position 30 to 35.

3.4. Analysis of recombinant Gpl

When the Gpl-plasmid was expressed in E. coli, a solubleprotein was obtained. In immunoblotting experimentsusing polyclonal antibodies against the lectin–trypsincomplex, a protein band (Mr�32,000) was detected in thelysates (Fig. 3, lane 2). When the lysate was applied onto aD(+)-glucosamine affinity column, the bound fraction wasfound to have a protein band of Mr�32,500 on SDS-PAGE (Fig. 3, lane 1, arrow).

3.5. Agglutination and trypsin activities

The ability of the recombinant Gpl to agglutinate rabbitRBCs and bloodstream-form T. b. brucei was assessed invitro. The RBCs and trypanosomes were agglutinated with

1 2 3 4 5 kDa

94

67

43

30

20

(B)(A)

14.4

Fig. 3. Dissociating gel electrophoresis and immunoblot analysis. Protein

samples were separated by gradient SDS-PAGE (4–20%) and then

transferred onto nitrocellulose paper. The blots were then reacted with

antiserum against lectin–trypsin complex. (A) Silver-stained gel: lane 1,

low molecular weight standards (Pharmacia); lane 2, recombinant E. coli

lysate (�17mg); lane 3, purified Gpl (�30mg). (B) Immunoblot: lane 4,

purified Gpl (�34mg); lane 5, E. coli lysate (�20mg).

N-acetyl-D-glucosamine 256 128

aA sample of Gpl (�5mg protein) was used to prepare serial dilutions in

PBS. To each dilution, 500mM of the different sugars were added.

Agglutination was assessed using bloodstream-form trypanosomes and

rabbit RBCs. Control consisted of the reaction mixture in the absence of

the sugars.bEnd points are expressed as reciprocals of the dilutions.

titers of 128 and 512, respectively (Table 1). No agglutina-tion was seen in the controls or the Gpl (unexpressed)samples (titers of 0 and 4, respectively). Agglutination oftrypanosomes and RBCs by Gpl was strongly inhibited byD(+)-glucosamine (Table 2). At 500mM, the agglutinationtiters were reduced by �90% in both cases. The sameconcentration N-acetyl-glucosamine reduced the titers by�30% and 42% in the case of trypanosomes and RBCs,respectively. When the substrate specificity of the recombi-nant molecule was assessed using a trypsin-specificchromogenic substrate, Chromozym-TRY, the recombi-nant lysate showed a higher trypsin activity (6.025units� 10�4) than the bacterial control lysate (0.303units� 10�4) (Table 1).

3.6. Functional studies

To assess the involvement of the Gpl in the transforma-tion of trypanosomes, the purified recombinant proteinwas incubated with bloodstream-form parasites and the

ARTICLE IN PRESSL.U. Abubakar et al. / Insect Biochemistry and Molecular Biology 36 (2006) 344–352350

parasites observed over time. The results showed a gradualstimulation of transformation process (Fig. 4A). After 4 h,�30% of the parasites were in the transition form, and themidgut forms were first detected after �8–12 h. The rates oftransformation by the recombinant Gpl and the nativelectin–trypsin complex were 0.078870.0048 (at the asymp-totic value of 98.94%) and 0.06670.0054 (at the asympto-tic value of 95.87%), respectively. There was no significantdifference between the two transformation rates (t-test,p40:0001). No transformation was observed in theabsence of the recombinant protein or the lectin–trypsincomplex, even after 24 h of incubation. The transformationprocess was further ascertained by indirect immunofluor-escence using anti-procyclin antibodies. In this case,fluorescence was only observed in trypanosomes treatedwith Gpl (Fig. 4B). In immunoblotting experiments,antiserum against Gpl reacted with the a subunit of thelectin–trypsin complex. No cross-reactivity was observedwith the b subunit of the complex (data not shown).

0

20

40

60

80

100

0 5 10 15 2Tim

% p

rocy

clic

try

pan

oso

mes

(A)

(B)

i

Fig. 4. (A) Transformation of bloodstream trypanosomes by lectin–trypsin c

(27 1C) with either lectin–trypsin complex (�4mg) [K] or Gpl (�4.5mg) [m

Transformation of the parasites was assessed as described in Section 2. Trend re

Immunoflourescence microscopy of procyclic T. b. brucei. The smears were pre

and fixed in 100% methanol (�20 1C, 20min). Transformed parasites expressin

and flourescein isothiocyanate-coupled anti-mouse secondary antibody. Non-t

using 1mg/ml DNA-binding dye 4060-diamidino-2-phenylindole (DAPI): (i) untr

with Gpl (�4.5 mg). DNA is stained blue and the green fluorescence is the pro

4. Discussion

We have characterized the cDNA of a novel proteolyticlectin (Gpl), which is synthesized and secreted in the midgutof the tsetse fly, G. f. fuscipes, in response to a bloodmeal.The 50 terminus of the deduced amino acid sequence of Gpl

was found to be rich in hydrophobic residues (e.g. G, A, V,L, F, P) and, therefore, a potential trypsin-specificactivation site. This indicated that Gpl is expressed asprepropeptide, and is thus secretory in nature. Expressionof the gene in E. coli yielded a proteolytic protein capableof agglutinating both bloodstream-form trypanosomes andrabbit RBCs. This activity was specifically inhibited byD(+)-GlcN. In addition, the recombinant Gpl was able tostimulate transformation of bloodstream-form trypano-somes into procyclic forms in vitro.A comparison of the amino acid sequences of the Gpl

with the protein database using BLASTP showed that itexhibits similarity to various serine proteases, most notably

0 25 30 35 40e (h)

ii

omplex and Gpl. Parasitized blood (5� 106 parasites/ml) was incubated

]. The control consisted of parasitized blood and BIS, pH 7.9 (&).

present mean values ðn ¼ 4Þ fitted in non-linear model y ¼ Að1� e�btÞ. (B)

pared for flourescence microscopy by cytocentrifugation (2000 rpm, 5min)

g procyclin were detected using monoclonal anti-procyclin antibody (EP)

ransformed bloodstream parasites were detected by DNA counterstaining

eated parasites after 24 h incubation (control) and (ii) after 12 h incubation

cyclin signal.

ARTICLE IN PRESSL.U. Abubakar et al. / Insect Biochemistry and Molecular Biology 36 (2006) 344–352 351

to the sequence of G. morsitans serine protease 1 (Gsp1)(Yan et al., 2001) that was reported while this work was inprogress. Multiple alignment showed that Gpl possess thecatalytic triad, His, Asp and Ser, and the highly conservedregions surrounding the His and Ser residues that aretypical of serine proteases (Fig. 2) (Kraut, 1977). They alsohave the highly conserved cysteine residues at positionsthat allow the formation of the three cysteine bonds typicalof invertebrate serine proteases and differentiating themfrom the vertebrate enzymes, which have four such bonds(Lehane et al., 1998). At the substrate specificity site, Gpl

contains the aspartate at position 212 (numbering as forGpl; Fig. 2), which is diagnostic of tryptic specificity(Stroud et al., 1974). The aspartic acid, which lies at thebottom of the substrate-binding pocket, is responsible fortrypsin’s preference for lysine and arginine (Hedstromet al., 1992), and occurs in all known trypsins with theexception of a blood meal-induced Aedes aegypti trypsin(Barillus-Mury et al., 1991). On the other hand, the glycineresidue in the substrate-binding pocket is replaced by aserine at position 251 (Fig. 2), which is characteristic ofchymotrypsin (Kraut, 1977). Chymotrypsin and trypsinhave very similar three-dimensional structures but differentsubstrate specificities. Chymotrypsin prefers bulky aro-matic side chains whereas trypsin preferentially interactswith positively charged side chains (e.g. Lys or Arg of asubstrate) at the bottom of the specificity pocket (Kraut,1977; Warshel et al., 1989). In this study, the substratespecificity of Gpl was determined empirically using atrypsin-specific chromogenic substrate, Chromozym-TRY. The observed trypsin activity suggests that theexpressed Gpl is more closely related to trypsin-likeenzymes than chymotrypsin-like proteases.

In order to assess the role of Gpl in tsetse–trypanosomeinteractions, the expressed protein was assayed for bothagglutination and transformation activities. Gpl aggluti-nated both RBCs and bloodstream-form trypanosomes,suggesting that the protein had lectin-like activity. In bothcases, agglutination was specifically inhibited by D(+)-glucosamine. Several proteins bearing carbohydrate bind-ing and serine protease motifs, with diverse biological rolesin development and immunity, have been reported (Ohashiet al., 1999; Vyse et al., 1994). In Anopheles gambiae, atrypsin-like serine protease (Sp22D) encompassing chitinbinding, low-density lipoprotein receptor and scavengerreceptor cysteine-rich domains has been identified (Danielliet al., 2000). In addition, lectins that have hydrolyticenzymatic activities have been reported in the mung beanspecies (Hankins and Shannon, 1978). A midgut defensinin association with serine protease was recently reported inStomoxys calcitran (Hamilton et al., 2002). In Glossina

species, a lectin with trypsin-like activity was isolated fromthe midgut of fed flies (Osir et al., 1995). The non-covalently linked a and b subunits of the holoprotein wereshown to be associated with the trypsin and glycosylresidues, respectively. The presence of glycosyl residues onthe b subunit was used as the basis for proposing that this

subunit was the lectin component. The fact that thesequence of Gpl did not reveal any potential glycosylationsites suggested its closeness to the a subunit of the complex.Indeed, antibodies against Gpl reacted only with the asubunit. The observation that Gpl had lectin activity is notsurprising since unglycosylated lectins such as calnexinhave already been reported (Lis and Sharon, 1998).Further work will be needed to discover the nature of theb subunit.An interesting observation in this study was the ability of

Gpl to stimulate transformation of bloodstream-formtrypanosomes into procyclic forms in vitro. In previousstudies, we concluded that this was a property of thelectin–trypsin complex (Osir and Abubakar, unpublishedobservation). The findings of this study suggest that Gpl

can by itself stimulate the transformation of bloodstream-form trypanosomes into procyclic forms. The role of thesecond subunit in this process deserves further detailedstudies.

Acknowledgments

The authors thank the ICIPE Director of Research andPartnerships for useful suggestions. This work wassupported by grants from the UNDP/World Bank/ WHOSpecial Programme for Research and Training in TropicalDiseases and the International Centre for ScientificCulture-World Laboratory (ICSC-WL) to EOO.

References

Abubakar, L., Osir, E.O., Imbuga, M.O., 1995. Properties of a bloodmeal-

induced midgut lectin from the tsetse fly, Glossina morsitans. Parasitol.

Res. 81, 271–275.

Altschul, S.F., Madden, T.I., Schaffer, A.A., Zhang, J., Zhang, Z., Miller,

W., Lipman, D.J., 1997. Gapped BLAST and PSI-BLAST: a new

generation of protein database search programs. Nucl. Acids Res. 25,

3389–3402.

Barillus-Mury, C., Graf, R., Hagedorn, H.H., Wells, M.A., 1991. cDNA

and deduced amino acid sequence of a blood meal-induced trypsin

from the mosquito, Aedes aegypti. Insect Biochem. 21 (8), 825–831.

Brown, J.R., Hartley, B.S., 1966. Location of disulphide bridges by

diagonal paper electrophoresis. The disulphide bridges of bovine

chymotrypsinogen A. Biochem. J. 101, 214–228.

Corpet, F., 1988. Multiple sequence alignment with hierarchical clustering.

Nucl. Acids Res. 16 (22), 10881–10890.

Danielli, A., Loukeris, T.G., Lagueux, M., Muller, H.M., Richman, A.,

2000. A modular chitin-binding protease associated with hemocytes

and hemolymph in the mosquito Anopheles gambiae. Proc. Natl. Acad.

Sci. 97 (13), 7136–7141.

Hamilton, J.V., Munks, R.J.L., Lehane, S.M., Lehane, M.J., 2002.

Association of midgut defensin with a novel serine protease in the

blood-sucking fly, Stomoxys calcitrans. Insect Mol. Biol. 11 (3),

197–205.

Hankins, C.N., Shannon, L.M., 1978. The physical and enzymatic

properties of a phytohemagglutinin from mung beans. J. Biol. Chem.

253, 7791–7797.

Hao, Z., Kasumba, I., Lehane, M.J., Gibson, W.C., Kwon, J., Aksoy, S.,

2001. Tsetse immune responses and trypanosome transmission:

implications for the development of tsetse-based strategies to reduce

trypanosomiasis. Proc. Natl. Acad. Sci. USA 98, 12648–12653.

ARTICLE IN PRESSL.U. Abubakar et al. / Insect Biochemistry and Molecular Biology 36 (2006) 344–352352

Hedstrom, L., Szilagyi, L., Rutter, W.J., 1992. Converting trypsin to

chymotrypsin: the role of surface loops. Science 255, 1249–1253.

Imbuga, M.O., Osir, E.O., Darji, N., Otieno, L.H., 1992. Studies on tsetse

midgut factors that induce differentiation of bloodstream T. brucei

brucei in vitro. Parasitol. Res. 78, 10–15.

Kraut, J., 1977. Serine proteases: structure and mechanism of catalysis.

Ann. Rev. Biochem. 46, 331–358.

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

of the head of bacteriophage T4. Nature 227, 680–683.

Lehane, M.J., Msangi, A.R., 1991. Lectin and peritrophic membrane

development in the gut of G. m. morsitans and a discussion of

their role in protecting the fly against infection. Med. Vet. Entomol. 5,

495–501.

Lehane, S.M., Assinder, S.J., Lehane, M.J., 1998. Cloning, sequencing,

temporal expression and tissue-specificity of two serine proteases from

the midgut of the blood-feeding fly Stomoxys calcitran. Eur.

J. Biochem. 254, 290–296.

Lis, H., Sharon, N., 1998. Lectins: carbohydrate-specific proteins that

mediate cellular recognition. Chem. Rev. 98, 637–674.

Maudlin, I., 1985. Inheritance of susceptibility to trypanosomes in tsetse

flies. Parasitol. Today 1 (2), 59–60.

Maudlin, I., 1991. Transmission of African trypanosomiasis: interaction

among tsetse immune system, symbionts and parasites. Adv. Dis.

Vector Res. 7, 117–148.

Maudlin, I., Welburn, S.C., 1987. Lectin mediated establishment of

midgut infections of Trypanosoma congolense and T. brucei in Glossina

morsitans. Trop. Med. Parasitol. 38, 167–180.

Miller, E.N., Turner, M.J., 1981. Analysis of antigenic types appearing in

first relapse populations of clones of Trypanosoma brucei. Parasitology

82, 63–80.

Nguu, E.K., Osir, E.O., Imbuga, M.O., Olembo, N.K., 1996. The effect

of host blood in the in vitro transformation of bloodstream

trypanosomes by tsetse midgut homogenates. Med. Vet. Entomol.

10, 317–322.

Ohashi, M., Kawamura, K., Fujii, N., Yibisui, T., Fujiwara, S., 1999. A

retinoic acid-inducible modular protease in budding Ascidians. Dev.

Biol. 214, 38–45.

Onyango, P., 1993. Relationship between lectins and trypsins in the

midgut of tsetse fly, Glossina morsitans morsitans. M.Sc. Thesis,

University of Nairobi.

Osir, E.O., Wells, M.A., Law, J.H., 1986. Studies of vitellogenin from the

tobacco hornworm, Manduca sexta. Arch. Insect Biochem. Physiol. 3,

217–223.

Osir, E.O., Labongo, L.V., Unnithan, G.C., 1989. A high molecular

weight diapause-associated protein from stem borer, Busseola fusca:

purification and properties. Arch. Insect Biochem. Physiol. 11,

173–187.

Osir, E.O., Imbuga, M.O., Onyango, P., 1993. Inhibition of Glossina

morsitans midgut trypsin activity by D-glucosamine. Parasitol. Res. 79,

93–97.

Osir, E.O., Abubakar, L., Imbuga, M.O., 1995. Purification and

characterization of a midgut lectin–trypsin complex from the tsetse

fly, Glossina longipennis. Parasitol. Res. 81, 276–281.

Osir, E.O., Abubakar, L.U., Abakar, M., 1999. The role of trypanolysin in

the development of trypanosomes in tsetse. In: Proceedings of the 25th

Meeting of the International Scientific Council for Trypanosomiasis

Research & Control (ISCTRC), Mombasa, KENYA. Publication

No. 120, pp. 417–421.

Peters, W., 1992. In Peritrophic Membranes: Zoophysiology, vol. 30.

Springer, Berlin.

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

Laboratory Manual, second ed. Cold Spring Harbour Laboratory

Press, New York.

Stiles, J.K., Ingram, G.A., Wallbanks, K.R., Molyneux, D.H., Maudlin,

I., Welburn, S.C., 1990. Identification of midgut trypanolysin and

trypanoagglutinin in G. pallidipes (Diptera: Glossinidae). Parasitology

101, 369–376.

Stroud, R.M., Kay, L.M., Dickerson, R.E., 1974. The structure of bovine

trypsin: electron density maps of the inhibited enzyme at 5 A and at

2.7 A resolution. J. Mol. Biol. 83, 185–208.

Vickerman, K., 1985. Developmental cycles and biology of pathogenic

trypanosomes. Br. Med. Bull. 41 (2), 105–114.

Vyse, T.J., Bates, G.P., Walport, M.J., Morley, B.J., 1994. The

organization of the human complement factor 1 gene (IF): a member

of the serine protease gene family. Genomics 24, 90–98.

Warshel, A., Naray-Szabo, G., Sussman, F., Hwang, J.K., 1989. How do

serine proteases really work? Biochemistry 28, 3629–3637.

Yan, J., Cheng, Q., Li, C., Aksoy, S., 2001. Molecular characterization of

two serine proteases expressed in gut tissue of the African trypanosome

vector, Glossina morsitans morsitans. Insect Mol. Biol. 10 (1), 47–56.