Molecular and Biochemical Parasitology 104 (1999) 195–204

A single-chain antibody fragment specific for thePlasmodium berghei ookinete protein Pbs21 confers

transmission blockade in the mosquito midgut�

Shigeto Yoshida a,*, Hiroyuki Matsuoka a, Enjie Luo a, Kuni Iwai a,Meiji Arai a, Robert E. Sinden b, Akira Ishii a

a Department of Medical Zoology, Jichi Medical School, Tochigi 329-0498, Japanb Infection and Immunity section, Department of Biology, Imperial College of Science, Technology and Medicine,

London SW7 2AZ, UK

Received 2 March 1999; received in revised form 25 June 1999; accepted 26 July 1999

Abstract

Mouse monoclonal antibody 13.1 (mAb 13.1) directed against Pbs21, a 21-kDa sexual-stage surface protein ofPlasmodium berghei, is known to inhibit oocyst development from gametocytes and ookinetes in the mosquito midgut.To examine the properties and potential uses of a single-chain antibody fragment (scFv) for blocking transmission ofmalaria parasites to mosquitoes, we have cloned and sequenced the genes encoding variable regions of theimmunoglobulin heavy and light chains (VH and VL) of mAb 13.1. The VH and VL genes were assembled as an scFvgene, and expressed in a baculovirus expression system. Following purification of 13.1 scFv, Western blotting andinhibition ELISA assays confirmed that 13.1 scFv retained the binding specificity of the parent mAb 13.1 for Pbs21.Furthermore, 13.1 scFv bound to the surface of P. berghei ookinetes, and blocked oocyst development in themosquito midgut by at least 93%, as assessed by oocyst counts in mosquitoes. We suggest that the 13.1 scFv genecould be useful not only in studying the mechanism of transmission blockade, but also in generating, by mosquitogermline transformation, a model system to evaluate the production of mosquitoes refractory to malaria. © 1999Elsevier Science B.V. All rights reserved.

Keywords: Malaria; Plasmodium berghei ; Pbs21; Transmission blocking; Single-chain antibody

www.elsevier.com/locate/parasitology

Abbre6iations: VH, variable region of immunoglobulin heavy chain; VL, variable region of immunoglobulin light chain; scFv,single-chain antibody fragment; Bl, transmission blockade; mAb, monoclonal antibody; PfMSP-1, P. falciparum merozoite surfaceprotein 1; T. ni, Trichoplusia ni.� Note : The nucleotide sequence data of the VH and VL genes of mAb 5.2 have been deposited in the DDJB, EMBL, and

GenBank™ databases under the accession numbers AB028875 and AB028876, respectively.* Corresponding author. Tel.: +81-285-58-7339; fax: +81-285-44-6489.E-mail address: shigeto@jichi.ac.jp (S. Yoshida)

0166-6851/99/$ - see front matter © 1999 Elsevier Science B.V. All rights reserved.

PII: S 0166 -6851 (99 )00158 -9

S. Yoshida et al. / Molecular and Biochemical Parasitology 104 (1999) 195–204196

1. Introduction

Pbs21 is a protein expressed on the surface ofmacrogamete, zygote, ookinete, and oocyst stagesof the rodent malaria parasite, Plasmodiumberghei, as they develop in/on the mosquitomidgut [1,2]. It has been reported that anti-Pbs21mAbs effectively block the development of Plas-modium berghei ookinetes in vitro and oocysts inAnopheles stephensi mosquitoes [3,4]. The antigenhas thus been identified as an important target intransmission blocking immunity [3,5] and is amodel for development of transmission-blockingvaccines [6,7]. Ideally, transmission-blocking vac-cines should induce high-titre, long-lasting trans-mission-blocking antibodies after a singleimmunization. However, since the vaccine candi-date antigens of the Pfs25/28 family are expressedby parasites in the mosquito stage, but not in thevertebrate host [8], boosting of the immune re-sponse following a natural infection has neverbeen expected. Nonetheless, the induction oftransmission-blocking antibodies in patients incombination with anti-malarial drugs could be ofconsiderable importance to prevent the spread ofthe drug-resistant parasites.

A mouse mAb designated 13.1 recognizes theookinete surface antigen Pbs21 of P. berghei andeffectively blocks the transmission of the parasitesfrom mice to mosquitoes [3]. Therefore, a geneti-cally engineered mAb 13.1 and/or a gene encodinga single-chain antibody fragment (scFv) derivedfrom mAb 13.1 can be expected as a useful tool toprevent the transmission of malaria parasites inmosquitoes. For our interest, the use of an scFvcan offer attractive advantages over the use of amAb in terms of ease of manipulation at thegenetic level [9].

In this paper, we describe an scFv derived fromthe transmission-blocking mAb 13.1. We havecloned the DNA sequences encoding the variableregions of the immunoglobulin heavy and lightchains (VH and VL) of mAb 13.1. Using a bac-ulovirus expression system, we produced a func-tionally active 13.1 scFv, which retains thebinding specificity of the parent mAb 13.1. Fur-thermore, we demonstrate that 13.1 scFv signifi-cantly inhibits oocyst development in mosquitomidguts in transmission-blocking assays.

2. Materials and methods

2.1. Parasites, cell lines, and antibodies

The cloned line 2.34 of P. berghei ANKA strainwas maintained by cyclical passage through Balb/c mice and Anopheles stephensi (SDA 500 stain).Hybridoma clones 12.1 and 13.1 and mAbsagainst P. berghei Pbs21 have been described pre-viously [3]. Hybridoma clone 5.2 against P. falci-parum merozoite surface protein 1 (PfMSP-1) [10]was obtained from the American Type CultureCollection (Rockville, MD). All hybridoma cellswere maintained in RPMI 1640 supplementedwith 10% fetal calf serum. Antibodies werepurified from ascites using the E-Z Step™ (Phar-macia Biotech, Uppsala, Sweden). Sf9 cells weremaintained in Grace’s insect medium (Life Tech-nologies, Inc., Rockville, MD) supplemented with6% fetal calf serum. High Five™ cells, an insectcell line derived from Trichoplusia ni (T. ni ) eggcell homogenate (Invitrogen, San Diego, CA),were maintained in Ex-Cell™ 405 medium (JRHBioscience, Lenexa, KS).

2.2. Cloning and sequencing of the 13.1 and 5.2VH and VL genes

mRNA was extracted from 1×107 13.1 hybri-doma cells using the FastTrack™ 2.0 mRNAIsolation Kit (Invitrogen), and first-strand cDNAwas synthesized from the mRNA with reversetranscriptase using the First-Strand cDNA Syn-thesis Kit (Novagen, Madison, WI). Then, thecDNA was used as a template for PCR amplifica-tion of the 13.1 VH and VL genes, using Taq 2000DNA polymerase (Stratagene GmbH, Heidelberg,Germany) and the primer sets in the Mouse Ig-Prime Kit (Novagen). The resulting PCR prod-ucts were cloned into the pT7Blue plasmid vector(Novagen). The selected plasmid DNAs were se-quenced by the Dye Terminator Cycle SequencingFS Ready Reaction Kit (Applied Biosystems,Inc., Foster City, CA), and the products wereanalyzed on a 373A automated DNA sequencer(Applied Biosystems, Inc.). For the cloning of the5.2 VH and VL genes, mRNA was similarly ex-tracted from 1×107 5.2 hybridoma cells. The

S. Yoshida et al. / Molecular and Biochemical Parasitology 104 (1999) 195–204 197

construction and expression of the 5.2 scFv genewere performed as described for 13.1 scFv. 5.2scFv acted as a negative control and did not bindto Pbs21.

2.3. Construction of the 13.1 scF6 gene

VH and VL DNA were separately amplifiedusing cloned Pfu DNA polymerase (Stratagene)with specific sets of primers. The primers for VH

were p13.1VH-1 (5%-GGGGGATCCGGACTA-CAAGGACGACGATGACAAGATCTCAGA-GGTTCAGCTGCAGCAGTCTGGGGCAGAG-CTT-3%) and p13.1VH-2 (5%-GCCACCGCCA-GAGCCACCTCCGCCTGAACCGCCTCCAC-CTGTCGACGAGACGGTGACTGAGGTTCCTCGACCCCAGTA-3%); the primers for VL werep13.1VLG9-1 (5%-GGCGGTTCAGGCGGAG-GTGGCTCTGGCGGTGGCGGATCGGATA-TCGATGTTTTGATGACCCAAACTCCAC-3%)and p13.1VLG9-2 (5%-GGAGGAGGCCCC-CTGGGCCCCGTTTTATTTCCAGTTTGGTC-CCCCCTCCGAACG-3%). The FLAG-encodingsequence in p13.1VH-1 is underlined. The overlap-ping complementary sequences encoding the(Gly4Ser)3 linker peptide in p13.1VH-2 andp13.1VLG9-1 are double-underlined. BamHI andSfiI sites in p13.1VH-1 and p13.1VLG9-2, respec-tively, are in italics. The VH and VL genes ob-tained after 30 cycles of PCR (94°C 1 min, 58°C 1min, 72°C 1.5 min) were gel-purified; next, over-lapping PCR was performed to construct the scFvgene under the same conditions, but with thep13.1VH-1 and p13.1VLG9-2 primers. The 13.1scFv DNA was gel-purified, digested successivelywith BamHI and SfiI, and inserted into BamHI/SfiI-digested pBACgus-scFv, which contains thefirst 21 codons of the melittin gene signal se-quence [11] (5%-AGATCTACGCGTATGAAA-TTCTTAGTCAACGTTGCCCTTGTTTTTAT-GGTCGTATACATTTCTTACATCTATGCGG-ATCCGGAAATAAAACGGGGCCCAGGGGG-CCCCGCGGCCGC-3%: the melittin gene signalsequence is underlined and newly created BglII,BamHI, SfiI, and NotI sites are in italics) inBamHI/NotI-digested pBACgus-1 (Novagen).The resulting transfer vector was designatedpBACgus-13.1 scFv.

2.4. Expression and purification of 13.1 scF6 ininsect cells

Sf9 cells were co-transfected with 250 ng of thepBACgus-13.1 scFv and 50 ng of BacVector-3000DNA (Novagen) [12], and then recombinantviruses were isolated by plaque assay according tothe manufacturer’s protocol.

For protein production, T. ni insect cells weregrown to a density of 1−1.5×106 cells ml−1 inEx-Cell™ 405 medium. Cells were infected withrecombinant baculovirus at a multiplicity of infec-tion of 3 and incubated at 27°C for 3 days. Afterharvesting, the cell pellet was resuspended in alysis buffer (6 M guanidine hydrochloride, 150mM NaCl, 50 mM Tris-HCl, pH 7.5) at a concen-tration of 2×107 cells ml−1 and centrifuged at16 000×g for 25 min at 4°C. The supernatantwas incubated with the Ni-NTA Superflow (Qia-gen GmbH, Hiden, Germany) overnight at 4°C.The mixture was then poured into a 10×1.5-cmcolumn and washed with a buffer (6 M urea, 20mM imidazole, 150 mM NaCl, 50 mM Tris-HCl,pH 7.5). Proteins immobilized on the Ni-NTAmatrix were renatured using a 6 M–0 M urealinear gradient in TBS buffer (100 mM NaCl, 50mM Tris-HCl, pH 7.5) by HPLC, and then elutedwith 100 mM glycine, pH 3.5, containing 150 mMNaCl. The eluted fractions were immediately neu-tralized by addition of 1 M Tris-HCl, pH 8.0.

The purified 13.1 scFv was run on a 12%SDS-PAGE, and then transferred to a polyvinyli-dene difluoride membrane. The N-terminalamino-acid sequence was analyzed on a PPSQ-10protein sequencer (Shimadzu, Kyoto, Japan).Western blotting was performed to assess theexpression and purification of 13.1 scFv usingbiotinylated mouse anti-FLAG M2 mAb (East-man Kodak Co., Rochester, NY) [13].

2.5. Binding specificity of 13.1 scF6 and 5.2 scF6to Pbs21

2.5.1. Western blottingProtein extracts of hemolymph from Bombyx

mori (silkworm larvae) infected with recombinantbaculovirus expressing the Pbs21 gene and of invitro cultured ookinetes of P. berghei were pre-

S. Yoshida et al. / Molecular and Biochemical Parasitology 104 (1999) 195–204198

pared as described previously [3,7]. The proteinsamples were separated by 15% SDS-PAGE andtransferred to Immobilon™ Transfer Membrane(Millipore, Bedford, MA). Following blockingwith BSA, the filters were incubated with thepurified 13.1 scFv at a concentration of 20 mgml−1. 13.1 scFv bound to the antigen was subse-quently detected by incubating the filters withbiotinylated mouse anti-FLAG M2 mAb, fol-lowed by treatment with streptavidin-alkalinephosphatase. The filters were stained with NBT/BCIP (Life Technologies, Inc.) and comparedwith conventional Western blotting, using the par-ent mAb 13.1 as a primary antibody and biotiny-lated goat anti-mouse IgG (Vector Laboratories,Inc., Burlingame, CA) as a secondary antibody.

2.5.2. Inhibition ELISAMicrotiter plates coated with 1 mg ml−1 of

affinity-purified recombinant Pbs21 (BmrPbs21)[7] were incubated with different dilutions of anti-bodies, mAb 13.1, mAb 12.1, or mAb 5.2 for 1 hat room temperature. Plates were washed withPBS containing 0.05% Tween 20, and a fixedsaturating amount (determined by ELISA titra-tion) of 13.1 scFv (1.25 mg ml−1) was added andincubated for 1 h at room temperature. Bound13.1 scFv was detected using biotinylated mouseanti-FLAG M2 mAb and streptavidin-horseradish peroxidase (DAKO, Glostrup, Den-mark). The data are presented as percentinhibition of maximal ELISA absorbance (deter-mined by 13.1 scFv binding in the absence ofblocking Ab).

2.6. Transmission-blocking assay

Batches of at least 25 female A. stephensimosquitoes were placed in pots and starved for 24h. Mice, infected with P. berghei 3 days previouslyand exhibiting gametocytemias, were anesthetizedwith Rompun-Vetalar (1:2 ratio; Bayer andParke-Davis Veterinary, respectively). For firstfeeding, mosquitoes were allowed to feed for 30min and were then incubated at 21°C at 80%relative humidity. For second feeding, each mousewas intravenously administered 50 mg of 13.1scFv, 5.2 scFv, mAb 13.1, or mAb 5.2 in 100 ml of

PBS, and 1 h later, a batch of mosquitoes wasallowed to feed on each treated mouse and the fedmosquitoes were maintained as above. After afurther day’s starvation, the surviving blood-fedmosquitoes were collected and reared in pots at21°C with 5% fructose as a food source. Fourteendays after blood feeding, mosquito midguts weredissected and the number of oocysts was deter-mined. Infection rate of mosquitoes (prevalence),geometric mean number of oocysts on the midgut(intensity), and percent transmission blockade (%Bl) were calculated as described previously [3,6].Data were analyzed by Student’s t-test.

3. Results

3.1. Cloning of V region cDNAs of mAb 13.1and 5.2

The cDNA encoding VH and VL regions ofmAb 13.1 were cloned by reverse transcriptasePCR from mRNA isolated from cloned 13.1 hy-bridoma cells. The nucleotide sequences of the VH

and VL cDNAs were determined. Comparison ofthe nucleotide sequences in the GenBank databaserevealed the 13.1 VH and VL genes were identicalto the sequence previously deposited by directsubmission (Nardone et al., Accession NumbersU31740 and U31741). The VH and VL cDNAswere connected using a DNA linker encoding fora (Gly4Ser)3 peptide to assemble the 13.1 scFvgene, and cloned into a baculovirus transfer vec-tor. The resulting plasmid, designated pBACgus-13.1 scFv, codes for the melittin signal sequencefollowed by the FLAG epitope tag sequence atthe N-terminus and a histidine tag sequence at theC-terminus of 13.1 scFv, under the control of thepolyhedrin promoter (Fig. 1). 5.2 scFv was clonedand analyzed similarly (data not shown).

3.2. Production and purification of 13.1 scF6

Recombinant baculovirus expressing the 13.1scFv gene was constructed by co-transfection ofSf9 cells with pBACgus-13.1 scFv and baculovirusDNA. T. ni insect cells were infected with therecombinant baculovirus to produce 13.1scFv,

S. Yoshida et al. / Molecular and Biochemical Parasitology 104 (1999) 195–204 199

and the cell pellet was analyzed by SDS-PAGEand Western blotting. The cell pellet contained adominant protein with an apparent molecularmass of 34 kDa (Fig. 2A, lane 2), similar topolyhedrin of the wild-type virus (Fig. 2A, lane1). 13.1 scFv containing a histidine tail at theC-terminus was purified from the cell lysate as asingle band on a Ni-NTA column (Fig. 2A, lane3). Using this purification procedure, about 12mg of pure 13.1 scFv was obtained from 1×109

T. ni. insect cells representing some 10–15% ofthe total protein. Western blotting shows thatthe purified 13.1 scFv was detected by biotiny-lated anti-FLAG M2 mAb, which recognizedthe FLAG sequence at the N-terminus of therecombinant antibody (Fig. 2B, lane 3). 5.2 scFvwas similarly produced in the baculovirus ex-pression system and fully retained the antigen-binding properties for the PfMSP-1 molecule(unpublished data). In this study, 5.2 scFvserved as a negative control scFv, because al-though it shares 62% identity at the amino-acidlevel with 13.1 scFv, including melittin signal se-quence, FLAG, the flexible linker (Gly4Ser)3,

and His-tag amino-acid sequences, but dose notbind to Pbs21.

To investigate whether 13.1 scFv was present inthe culture supernatants, the growth medium wascollected. No detectable amounts of 13.1 scFvwere obtained, whereas 5.2 scFv was continuouslysecreted into the culture supernatant, as judged bysubsequent SDS-PAGE or immunoblotting analy-ses (data not shown). To determine, therefore, ifunexpected cleavage within the melittin signal se-quence occurred, N-terminal amino-acid sequenc-ing was performed. The N-terminal amino-acidsequence was DPDYKDDDDKISEVQ, whichmatched the predicted amino-acid sequence ofresidues 22–34 of 13.1 scFv (Fig. 1), indicatingthat the 13.1 scFv precursor was cleaved correctlyafter the melittin signal sequence. The reason forthe lack of 13.1 scFv secretion is unknown. Itmight be due to membrane anchoring, becauseflow cytometry analysis of live cells and an indi-rect fluorescent antibody test using anti-FLAGM2 mAb revealed that 13.1 scFv is localized onthe surface of the baculovirus-infected insect cells(data not shown).

Fig. 1. Schematic representation of a construct encoding 13.1 scFv. Construction of the pBACgus-13.1 scFv is described in Materialsand methods. The junction sequences between the gene segments of the 13.1 scFv construct are shown. P, polyhedrin promoter; SP,melittin signal sequence; F, FLAG epitope tag sequence; L, linker peptide (Gly4Ser)3; 6H, six histidine residues. Dashes indicateextensive sequences of the VH and VL regions omitted for clarity presentation.

S. Yoshida et al. / Molecular and Biochemical Parasitology 104 (1999) 195–204200

Fig. 2. Analysis of 13.1 scFv expressed in insect cells. (A)SDS-PAGE analysis of the cell pellets of insect cells infectedwith the wild-type baculovirus (lane 1) and the recombinantbaculovirus harboring the 13.1 scFv gene (lane 2). 13.1 scFvprotein was purified from recombinant baculovirus-infectedcells by a Ni-NTA column as described in Materials andmethods. Purified protein (3 mg) was subjected to 12% SDS-PAGE and then stained with Coomassie Brilliant Blue R-250(lane 3). (B) Western blotting analysis of 13.1 scFv. The cellextracts (lanes 1–3) were subjected to 12% SDS-PAGE (as inPanel A) and transferred to Immobilon™ membrane for West-ern blotting. The membrane was incubated with biotinylatedmouse anti-FLAG M2 mAb, followed by treatment withstreptavidin-alkaline phosphatase, as described in Materialsand methods.

when the parent mAb 13.1, but not mAb 12.1 ormAb 5.2, was used for pre-incubation (Fig. 4).These results demonstrated that 13.1 scFv notonly bound to Pbs21, but also recognized thesame epitope on the Pbs21 molecule, as did theparent mAb 13.1.

In addition, the binding of 13.1 scFv to nativePbs21 on the surface of ookinetes was tested byindirect immunofluorescent staining of liveookinetes. Whereas there was only faint fluores-cence staining with 5.2 scFv, strong staining wasobserved on the entire surface of the ookineteswith 13.1 scFv (data not shown). These resultsabove taken together demonstrate that purified13.1 scFv fully retains the antigen-binding proper-ties of the parent mAb 13.1.

3.4. Biological acti6ity of 13.1 scF6 on oocystde6elopment in 6i6o

We used passive immunization methods to de-termine whether 13.1 scFv can function as anactive ligand for Pbs21 with transmission-block-ing activity in vivo. 13.1 scFv, 5.2 scFv, mAb

Fig. 3. Comparison of specific binding of mAb 13.1 and 13.1scFv. Recombinant Pbs21 (lanes 1 and 3) and P. bergheiookinete (lanes 2 and 4) were prepared for SDS-PAGE underreducing conditions and run on a 15% gel. Proteins weretransferred to Immobilon™ membranes, probed with eithermAb 13.1 (lanes 1 and 2) followed by biotinylated goatanti-mouse IgG antibody or 13.1 scFv (lanes 3 and 4) followedby biotinylated mouse anti-FLAG M2 mAb. The membraneswere then treated with streptavidin-alkaline phosphatase asdescribed in Materials and methods.

3.3. Binding specificity of purified 13.1 scF6 toPbs21

The binding specificity of purified 13.1 scFvwas demonstrated by Western blotting. Thepurified 13.1 scFv specifically reacted with nativePbs21 (Fig. 3, lane 4) from P. berghei ookinetes,as well as recombinant Pbs21 (Fig. 3, lane 3),produced in hemolymph of B. mori infected withrecombinant baculovirus. The specificity of 13.1scFv is indistinguishable from that of the parentmAb 13.1 (Fig. 3, lanes 1 and 2).

The reactivity of purified 13.1 scFv with Pbs21was also determined by an inhibition ELISA.Bound 13.1 scFv was detected with biotinylatedmouse anti-FLAG M2 mAb followed by strep-tavidin-horseradish peroxidase. The binding of13.1 scFv to BmrPbs21 was completely inhibited

S. Yoshida et al. / Molecular and Biochemical Parasitology 104 (1999) 195–204 201

Fig. 4. Reactivity of 13.1 scFv with recombinant Pbs21 in inhibition ELISA. Microtiter plates were coated with the affinity-purifiedrecombinant Pbs21. mAb 13.1 (), mAb 12.1 (), and mAb 5.2 (�) were used at increasing concentrations to inhibit for thebinding with 13.1 scFv to recombinant Pbs21. Then, a saturating amount of 13.1 scFv was added, and bound 13.1 scFv wasquantitated by ELISA using biotinylated mouse anti-FLAG M2 mAb followed by streptavidin-horseradish peroxidase, as describedin Materials and methods. Values are expressed as percent inhibition of maximal absorbance, determined by adding 13.1 scFvwithout prior blocking mAb.

13.1, mAb 5.2, and PBS were compared for theirtransmission-blocking potential against P. bergheiin Anopheles stephensi. Infection rates ofmosquitoes (prevalence), geometric mean numberof oocysts on the midgut (intensity), and % Bl areshown in Table 1. Compared with pre-administra-tion transmission values, administration of either13.1 mAb or 13.1 scFv dramatically decreased theprevalence of infection in mosquitoes fed on miceNos. 2, 3, 6, and 7 (10–30%), whereas the preva-lence of infection in mosquitoes fed on mice Nos.1, 4, and 5 treated with PBS, negative controlmAb 5.2 or 5.2 scFv remained high (82–93%).Intensity of infection was also significantly re-duced. A statistically significant (PB0.01, Stu-dent’s t-test) Bl of oocyst development wasachieved both by administration of 13.1 scFv(95.9% and 93.4%) and by mAb 13.1 (90.2% and92.3%). Despite retaining high prevalence afteradministration in the control groups, Bl was sur-prisingly high (39.2, 40.1, and 42.8, respectively).It may be partially because the density of gameto-cytes in mice was reduced following sample ad-ministration (100 ml).

4. Discussion

We report here the cloning and expression of anscFv gene encoding transmission-blocking anti-body specific for Pbs21 of P. berghei. A bac-ulovirus expression system was utilized to produce13.1 scFv in serum-free culture medium. Ten tofifteen percent of the total cellular protein pro-duced in T. ni. insect cells infected with the re-combinant virus was 13.1 scFv.

13.1 scFv was purified and renatured from thecell lysate as a single protein by a Ni-NTAcolumn. Western blotting and inhibition ELISAassays showed that the purified 13.1 scFv fullyretains the binding specificity of the parent mAb13.1 for the Pbs21 molecule. The epitope recog-nized by mAb 13.1 is located in the second epider-mal growth factor-like domain of Pbs21 as alinear sequence [14]. Our observation that 13.1scFv recognized the same epitope as did mAb13.1, but not mAb 12.1, is consistent with aprevious study indicating the epitopes recognizedby mAb 12.1 and mAb 13.1 are distinct [15].

To elicit effective and long-lasting transmission-blocking antibodies, several attempts have been

S.

Yoshida

etal./

Molecular

andB

iochemical

Parasitology

104(1999)

195–

204202

Table 1Effects of 13.1 scFv on oocyst development in mosquito midguts

AdministrationPre-administration

Mouse Treatment Infected/dis- Prevalence of in-Geometric mean Geometric meanPrevalence of % transmissionInfected/dis-sectedsected infection (%) (intensity) blockadeafection (%) (intensity)

Exp. 11 909/10 5.83 39.290 9.59 PBS 9/10

30 0.71 90.263/1013.1 mAb2 7.22866/713.1 scFv 2/11 18 0.30 95.9b3 6/7 86 7.33

Exp. 282 3.34 40.14 19/20 95 5.58 5.2 mAb 9/1193 6.41 42.814/155.2 scFv5 11.2110022/22

13.1 mAb 3114 21 0.54 92.3b18J186 100 7.0310 0.24 93.43/299.15 13.1 scFv977 31/32

a Percent transmission blockade (% Bl)=100−Geometric mean number of oocysts in the administered group

Geometric mean number of oocysts in the pre-administered group×100.

b Significantly less (PB0.01) than value for pre-administration group as calculated by Student’s t-test.

S. Yoshida et al. / Molecular and Biochemical Parasitology 104 (1999) 195–204 203

made not only to express antigens of the appropri-ate conformation, but also to optimize the deliveryof the transmission-blocking target proteins [16–19]. In this study, significant (93%) inhibition ofoocyst development in the mosquito midguts wasachieved by passive administration of 13.1 scFv.

In addition to production of functionally active13.1 scFv, our future objectives are not only todetermine the functions of Pbs21 molecules, butalso to generate transgenic mosquito vectors ex-pressing genes that confer resistance to malariaparasites. Recently, transgenic plants expressingan scFv gene were successfully generated [20], andthe protein expressed was effective in neutralizingthe target virus. The application of this technologyawaits the development of secure methods for theroutine transfection of Anopheline vectors ofPlasmodium [21].

To model this novel concept for malaria con-trol, the 13.1 scFv gene could be used to generatea refractory mosquito against malaria parasite byintroducing the gene into the germline of Anophe-line mosquito vector. We are nonetheless awarethat a mosquito expressing an scFv recognizing asingle antigenic epitope is not in itself going to bea practicable method of malaria control. Furtherexperiments are therefore underway to determineif the introduction of antimalarial lytic peptide,such as Shiva-1 [22], to 13.1 scFv can be a dramat-ically enhanced cytocidal effect on malaria para-site in mosquito vector.

Acknowledgements

We are grateful to Dr. A. Kobayashi for analyz-ing an amino acid sequence. This work was sup-ported in part by grants from Jichi Medical SchoolYoung Investigator Award and Grant-in-Aid forScientific Research on Priority Areas (08281104)from the Ministry of Education, Science, Cultureand Sports.

References

[1] Sinden RE, Winger LA, Hartley RH, Carter HE,Tirawanchai N, Davies CS, Sluiters JG. Ookinete anti-

gens of Plasmodium berghei : a light and electron micro-scopic immunogold study of the 21-kD determinantrecognized by transmission blocking antibodies. Proc RSoc (Lond) 1987;230:443–58.

[2] Paton MG, Barker GC, Mastuoka H, Ramesar J, JanseCJ, Water AP, Sinden RE. Structure and expression of aconserved and post-transcriptionally regulated gene en-coding a surface protein of the sexual stages from malariaparasite Plasmodium berghei. Mol Biochem Parasitol1993;59:263–76.

[3] Winger LA, Tirawanchai N, Nicholas J, Carter HE,Smith JE, Sinden RE. Ookinete antigens of Plasmodiumberghei. Appearance on the zygote surface of an Mr 21kD determinant identified by transmission-blocking mon-oclonal antibodies. Parasite Immunol 1988;10:193–207.

[4] Ranawaka GRR, Fleck SL, Alejo-Blanco AR, SindenRE. Characterization of the modes of action of anti-Pbs21 malaria transmission-blocking immunity: ookineteto oocyst differentiation in vivo. Parasitol 1994;109:403–11.

[5] Tirawanchai N, Winger LA, Nicholas J, Sinden RE.Analysis of immunity induced by the affinity-purified 21kD zygote-ookinete surface antigen of Plasmodiumberghei. Infect Immun 1991;59:36–44.

[6] Margos G, Kurtenbach K, Posnett E, Barker GC, Mat-suoka H, Paton MG, Sinden RE. Expression of thePlasmodium berghei ookinete protein Pbs21 in a bac-ulovirus-insect cell system produces an efficient trans-mission blocking immunogen. Parasite Immunol1995;17:167–76.

[7] Matsuoka H, Kobayashi J, Barker GC, Miura K, ChinzeiY, Miyajima S, Ishii A, Sinden RE. Induction of anti-malarial transmission blocking immunity with a recombi-nant ookinete surface antigen of Plasmodium bergheiproduced in silkworm larvae using the baculovirus expres-sion vector system. Vaccine 1996;14:120–6.

[8] Carter R, Graves PM, Quakyi IA, Good MF. Restrictedor absent immune responses in human populations toPlasmodium falciparum gamete antigens that are targets ofmalaria transmission-blocking antibodies. J Exp Med1989;169:135–47.

[9] Raag R, Whitlow M. Single-chain Fvs. FASEB J1995;9:73–80.

[10] Siddiqui WA, Tam LQ, Kan S-C, Kramer KJ, Case SE,Palmer KL, Yamaga KM, Hui GSN. Induction of protec-tive immunity to monoclonal-antibody-defined Plasmod-ium falciparum antigens requires strong adjuvant in Aotusmonkey. Infect Immun 1986;52:314–8.

[11] Tessier DC, Thomas DY, Khouri HE, Laliberte F, VetnetT. Enhanced secretion from insect cells of a foreignprotein fused the honeybee melittin signal peptide. Gene1991;98:177–83.

[12] Kitts PA, Possee RD. A method for producing recombi-nant baculovirus expression vectors at high frequency.Biotechniques 1993;14:810–7.

[13] Knappik A, Pluckthun A. An improved affinity tag basedon the FLAG peptide for the detection and purification

S. Yoshida et al. / Molecular and Biochemical Parasitology 104 (1999) 195–204204

of recombinant antibody fragments. Biotechniques1994;17:754–61.

[14] Spano F, Matsuoka H, Ozawa R, Chinzei Y, Sinden RE.Epitope mapping on the ookinete surface antigen Pbs21of Plasmodium berghei : identification of the site of bind-ing of transmission-blocking monoclonal antibody 13.1.Parassitologia 1996;38:559–63.

[15] Tirawanchai N, Sinden RE. Three non-repeated transmis-sion blocking epitopes recognized in the 21 kD surfaceantigen of zygotes-ookinetes of Plasmodium berghei. Para-site Immunol 1990;12:435–46.

[16] Kaslow DC, Isaacs SN, Quakyi IA, Gwadz RW, Moss B,Keister DB. Induction of Plasmodium falciparum trans-mission-blocking antibodies by recombinant vacciniavirus. Science 1991;252:1310–3.

[17] Margaret M, Gozar G, Price VL, Kaslow DC. Saccha-romyces cerevisiae-secreted fusion proteins Pfs25 andPfs28 elicit potent Plasmodium falciparum transmission-blocking antibodies in mice. Infect Immun 1998;66:59–64.

[18] Quakyi IA, Miller LH, Good MF, Ahlers JD, Isaacs SN,Nunberg JH, Houghten RA, Keister DB, Coligan JE,

Moss B, Alling D, Berzofsky JA, Kaslow DC. Syntheticpeptides from P. falciparum sexual stage 25-kDa proteininduce antibodies that react with the native protein: therole of IL-2 and conformational structure on immuno-genicity of Pfs25. Peptide Res 1995;8:335–44.

[19] Rawlings DJ, Kaslow DC. Adjuvant-dependent immuneresponse to malaria transmission-blocking vaccine candi-date antigens. J Exp Med 1992;176:1483–7.

[20] Tavladoraki P, Benvenuto E, Trinca S, De Martinis D,Cattaneo A, Galeffi P. Transgenic plants expressing afunctional single-chain Fv antibody are specifically pro-tected from virus attack. Nature 1993;366:469–72.

[21] Jasinskiene N, Coates CJ, Benedict MQ, Cornel AJ,Salazar Rafferty C, James AA, Collins FH. Stable trans-formation of the yellow fever mosquito, Aedes aegypi,with the Hermes element from the housefly. Proc NatlAcad Sci USA 1998;95:3743–7.

[22] Jaynes JM, Burton CA, Barr SB, Jeffers GW, Julian GR,White KL, Enright FM, Klei TR, Laine RA. In vitrocytocidal effect of novel lytic peptides on Plasmodiumfalciparum and Trypanosoma cruzi. FASEB J1988;2:2878–83.

.