7
APPLIED AND ENVIRONMENTAL MICROBIOLOGY, 0099-2240/98/$04.0010 Nov. 1998, p. 4566–4572 Vol. 64, No. 11 Copyright © 1998, American Society for Microbiology. All Rights Reserved. Engineering of a Single-Chain Variable-Fragment (scFv) Antibody Specific for the Stolbur Phytoplasma (Mollicute) and Its Expression in Escherichia coli and Tobacco Plants FABRICE LE GALL, JOSEPH-MARIE BOVE ´ , AND MONIQUE GARNIER* Laboratoire de Biologie Cellulaire et Mole ´culaire, Institut National de la Recherche Agronomique et Universite ´ Victor Segalen Bordeaux 2, 33883 Villenave d’Ornon cedex, France Received 20 January 1998/Accepted 12 August 1998 From a hybridoma cell line (2A10) producing an immunoglobulin G1 directed against the major membrane protein of the stolbur phytoplasma, we have engineered scFv (single-chain variable-fragment) antibodies from the variable heavy (VH) and light (VL) domains of the immunoglobulin. The scFv gene was cloned and expressed in Escherichia coli. The expressed protein of 30 kDa could be recovered from the periplasmic fraction of the bacterial cells and was shown to be fully functional toward its phytoplasmal antigen, since enzyme-linked immunosorbent assay or immunofluorescence (IF) detection of the stolbur phytoplasma antigen by the scFv was identical to that of the native immunoglobulin. The scFv gene was then cloned in plasmid pBG-dAb-BIN of Agrobacterium tumefaciens to transform tobacco plants. The transformed plants were screened by PCR and Northern blotting for the presence and expression of the transgene, respectively, and by IF for expression of the scFv. One transgenic tobacco line, 1A6, was selected for challenge inoculation with the stolbur phyto- plasma. When grafted on a stolbur phytoplasma-infected tobacco rootstock, the transgenic tobacco shoots grew free of symptoms and flowered after 2 months, while normal tobacco shoots showed severe stolbur symptoms during the same period and eventually died. Expression in plants of antibodies (plantibodies) that are able to interfere with the multiplication of pathogens can pro- vide an efficient way to induce resistance to many diseases. Mollicutes are wall-less bacteria that infect humans, animals, and plants (4, 5, 20). They originated, in the course of evolu- tion, from low-G1C gram-positive bacteria by gene loss and genome reduction (regressive evolution); in particular, they have lost the genes responsible for the synthesis of a bacterial cell wall (5). Thus, mollicutes are limited by a single cytoplas- mic membrane. This is why metabolism and growth of molli- cutes are inhibited by antibodies directed against their mem- brane epitopes. These inhibitions are generally followed by lysis of the mollicute (18). Thus, plant-pathogenic mollicutes are, a priori, ideal candidates for a plantibody-controlled re- sistance strategy. Mollicutes (phytoplasmas and spiroplasmas) are responsible for more than 300 diseases of vegetable, ornamental, and pe- rennial plants (20). These agents are localized exclusively in the sieve tubes of the phloem tissue (9), into which they are inoculated by insect vectors (leafhoppers and psyllids). Phyto- plasmas, the largest group of plant-pathogenic mollicutes, can- not be grown in artificial media. As of today, diseases induced by plant mollicutes cannot be controlled. Recently, several groups were able to express mouse-derived antibodies in plants (1, 2, 8, 14, 17, 22, 24), and reduction of virus multiplication has been observed in transgenic plants expressing whole immunoglobulin G (IgG) or single-chain variable-fragment (scFv) antibody directed against the virus coat protein (27, 28). Since the growth of mollicutes is inhib- ited by antibodies, constitutive expression in a plant of an antibody specific for a given mollicute should prevent its mul- tiplication, especially since the number of mollicutes inocu- lated by an insect vector is small (,10 3 ). To evaluate the ability of antibodies to control mollicute diseases in plants, we have engineered and expressed in tobacco plants monoclonal anti- body 2A10, directed against the major membrane protein of the stolbur phytoplasma (16). The stolbur phytoplasma induces diseases in all solanaceous species worldwide, including to- bacco, a good model plant for transgenosis (15, 20), but also plants belonging to other species such as lavender and celery, where it induces decline and porcelain disease respectively. The stolbur agent has recently also been shown to be respon- sible for the following grapevine diseases: bois noir in France (3), Vergilbungskrankheit (VK) in Germany (25), and Austra- lian yellows in Australia (7). Since the stolbur phytoplasma is transmitted by the polyphagous leafhopper Hyalesthes obsol- etus (15), it is likely to be associated with other diseases as well. In this paper, we describe the engineering of monoclonal antibody 2A10 into an scFv, its expression in Escherichia coli cells and tobacco plants, and the ability of the expressed scFv to bind the phytoplasma antigen. An experiment in which one transgenic tobacco line has been challenged with the stolbur phytoplasma is also presented. MATERIALS AND METHODS Plant material. Healthy periwinkle (Catharantus roseus (L.) g. Don) and to- bacco (Nicotiana tabacum (L.) cv. P B D6) plants were grown from seeds. Seeds of the commercial P B D6 tobacco variety were kindly provided by Rene ´ Delon, Institut du Tabac, Bergerac, France. Stolbur phytoplasma-infected periwinkle plants were obtained as described previously (15). The stolbur phytoplasma was transmitted from periwinkle to tobacco via dodder (Cuscuta campestris L.) and then maintained in tobacco plants by graft inoculation of 1-month-old plants. The presence of the stolbur phytoplasma in graft-inoculated plants was assessed by symptom expression and confirmed by double-sandwich enzyme-linked im- munosorbent assay (DAS-ELISA) (see below). The plants were grown in a greenhouse at 25°C during the day and 20°C at night. Mouse hybridoma cell lines. Hybridoma 2A10 producing a monoclonal anti- body (MAb) (IgG1) against a membrane epitope of the stolbur phytoplasma (16) * Corresponding author. Mailing address: Laboratoire de Biologie Cellulaire et Mole ´culaire, INRA, BP 81, 33883 Villenave d’Ornon cedex, France. Phone: (33) (0) 5 56 84 31 49. Fax: (33) (0) 5 56 84 31 59. E-mail: [email protected]. 4566 on May 28, 2020 by guest http://aem.asm.org/ Downloaded from

Engineering of a Single-Chain Variable-Fragment (scFv ... · FABRICE LE GALL, JOSEPH-MARIE BOVE´, AND MONIQUE GARNIER* Laboratoire de Biologie Cellulaire et Mole´culaire, Institut

  • Upload
    others

  • View
    1

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Engineering of a Single-Chain Variable-Fragment (scFv ... · FABRICE LE GALL, JOSEPH-MARIE BOVE´, AND MONIQUE GARNIER* Laboratoire de Biologie Cellulaire et Mole´culaire, Institut

APPLIED AND ENVIRONMENTAL MICROBIOLOGY,0099-2240/98/$04.0010

Nov. 1998, p. 4566–4572 Vol. 64, No. 11

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

Engineering of a Single-Chain Variable-Fragment (scFv)Antibody Specific for the Stolbur Phytoplasma (Mollicute)and Its Expression in Escherichia coli and Tobacco Plants

FABRICE LE GALL, JOSEPH-MARIE BOVE, AND MONIQUE GARNIER*

Laboratoire de Biologie Cellulaire et Moleculaire, Institut National de la Recherche Agronomique et Universite VictorSegalen Bordeaux 2, 33883 Villenave d’Ornon cedex, France

Received 20 January 1998/Accepted 12 August 1998

From a hybridoma cell line (2A10) producing an immunoglobulin G1 directed against the major membraneprotein of the stolbur phytoplasma, we have engineered scFv (single-chain variable-fragment) antibodies fromthe variable heavy (VH) and light (VL) domains of the immunoglobulin. The scFv gene was cloned andexpressed in Escherichia coli. The expressed protein of 30 kDa could be recovered from the periplasmic fractionof the bacterial cells and was shown to be fully functional toward its phytoplasmal antigen, since enzyme-linkedimmunosorbent assay or immunofluorescence (IF) detection of the stolbur phytoplasma antigen by the scFvwas identical to that of the native immunoglobulin. The scFv gene was then cloned in plasmid pBG-dAb-BINof Agrobacterium tumefaciens to transform tobacco plants. The transformed plants were screened by PCR andNorthern blotting for the presence and expression of the transgene, respectively, and by IF for expression ofthe scFv. One transgenic tobacco line, 1A6, was selected for challenge inoculation with the stolbur phyto-plasma. When grafted on a stolbur phytoplasma-infected tobacco rootstock, the transgenic tobacco shoots grewfree of symptoms and flowered after 2 months, while normal tobacco shoots showed severe stolbur symptomsduring the same period and eventually died.

Expression in plants of antibodies (plantibodies) that areable to interfere with the multiplication of pathogens can pro-vide an efficient way to induce resistance to many diseases.Mollicutes are wall-less bacteria that infect humans, animals,and plants (4, 5, 20). They originated, in the course of evolu-tion, from low-G1C gram-positive bacteria by gene loss andgenome reduction (regressive evolution); in particular, theyhave lost the genes responsible for the synthesis of a bacterialcell wall (5). Thus, mollicutes are limited by a single cytoplas-mic membrane. This is why metabolism and growth of molli-cutes are inhibited by antibodies directed against their mem-brane epitopes. These inhibitions are generally followed bylysis of the mollicute (18). Thus, plant-pathogenic mollicutesare, a priori, ideal candidates for a plantibody-controlled re-sistance strategy.

Mollicutes (phytoplasmas and spiroplasmas) are responsiblefor more than 300 diseases of vegetable, ornamental, and pe-rennial plants (20). These agents are localized exclusively inthe sieve tubes of the phloem tissue (9), into which they areinoculated by insect vectors (leafhoppers and psyllids). Phyto-plasmas, the largest group of plant-pathogenic mollicutes, can-not be grown in artificial media. As of today, diseases inducedby plant mollicutes cannot be controlled.

Recently, several groups were able to express mouse-derivedantibodies in plants (1, 2, 8, 14, 17, 22, 24), and reduction ofvirus multiplication has been observed in transgenic plantsexpressing whole immunoglobulin G (IgG) or single-chainvariable-fragment (scFv) antibody directed against the viruscoat protein (27, 28). Since the growth of mollicutes is inhib-ited by antibodies, constitutive expression in a plant of anantibody specific for a given mollicute should prevent its mul-

tiplication, especially since the number of mollicutes inocu-lated by an insect vector is small (,103). To evaluate the abilityof antibodies to control mollicute diseases in plants, we haveengineered and expressed in tobacco plants monoclonal anti-body 2A10, directed against the major membrane protein ofthe stolbur phytoplasma (16). The stolbur phytoplasma inducesdiseases in all solanaceous species worldwide, including to-bacco, a good model plant for transgenosis (15, 20), but alsoplants belonging to other species such as lavender and celery,where it induces decline and porcelain disease respectively.The stolbur agent has recently also been shown to be respon-sible for the following grapevine diseases: bois noir in France(3), Vergilbungskrankheit (VK) in Germany (25), and Austra-lian yellows in Australia (7). Since the stolbur phytoplasma istransmitted by the polyphagous leafhopper Hyalesthes obsol-etus (15), it is likely to be associated with other diseases as well.

In this paper, we describe the engineering of monoclonalantibody 2A10 into an scFv, its expression in Escherichia colicells and tobacco plants, and the ability of the expressed scFvto bind the phytoplasma antigen. An experiment in which onetransgenic tobacco line has been challenged with the stolburphytoplasma is also presented.

MATERIALS AND METHODS

Plant material. Healthy periwinkle (Catharantus roseus (L.) g. Don) and to-bacco (Nicotiana tabacum (L.) cv. P B D6) plants were grown from seeds. Seedsof the commercial P B D6 tobacco variety were kindly provided by Rene Delon,Institut du Tabac, Bergerac, France. Stolbur phytoplasma-infected periwinkleplants were obtained as described previously (15). The stolbur phytoplasma wastransmitted from periwinkle to tobacco via dodder (Cuscuta campestris L.) andthen maintained in tobacco plants by graft inoculation of 1-month-old plants.The presence of the stolbur phytoplasma in graft-inoculated plants was assessedby symptom expression and confirmed by double-sandwich enzyme-linked im-munosorbent assay (DAS-ELISA) (see below).

The plants were grown in a greenhouse at 25°C during the day and 20°C atnight.

Mouse hybridoma cell lines. Hybridoma 2A10 producing a monoclonal anti-body (MAb) (IgG1) against a membrane epitope of the stolbur phytoplasma (16)

* Corresponding author. Mailing address: Laboratoire de BiologieCellulaire et Moleculaire, INRA, BP 81, 33883 Villenave d’Ornoncedex, France. Phone: (33) (0) 5 56 84 31 49. Fax: (33) (0) 5 56 84 3159. E-mail: [email protected].

4566

on May 28, 2020 by guest

http://aem.asm

.org/D

ownloaded from

Page 2: Engineering of a Single-Chain Variable-Fragment (scFv ... · FABRICE LE GALL, JOSEPH-MARIE BOVE´, AND MONIQUE GARNIER* Laboratoire de Biologie Cellulaire et Mole´culaire, Institut

and hybridoma Myc-9E10-2 (ATCC CRL 1729) producing MAb 9E10 againstthe human c-myc protein (13) were used in this study.

E. coli strains and cloning vectors. The following E. coli strains and vectorswere used: E. coli XL1-Blue and plasmid BluescriptSK1 (pBS; Stratagene, LaJolla, Calif.), for cloning variable heavy (VH) and variable light (VL) sequences,and E. coli JM109 (Promega, Madison, Wis.) and plasmids pUC18 and pUC19(MBI Fermentas, Vilnius, Lithuania), for scFv expression. Plasmid pS8 (27) wasalso used for scFv constructions. Restriction enzymes SmaI, HindIII, XbaI, PstI,BstEII, EcoRI, SacI, SalI, and XhoI were purchased from MBI Fermentas. Allstandard techniques, if not described, were as described by Sambrook et al. (23).Plasmid pBG-dAb-BIN of Agrobacterium tumefaciens (27) was used for planttransformation.

mRNA isolation. Cells of hybridoma 2A10 producing the anti-stolbur phyto-plasma MAb were cultured in Iscove modified medium containing 20% (vol/vol)fetal bovine serum, 2% (vol/vol) glutamine (200 mM), and 1% (vol/vol) genta-micin (10 mg/ml) in a 5% CO2 humidified incubator. RNAs were prepared fromabout 109 hybridoma cells by the guanidine isothiocyanate method as describedby Chirgwin et al. (6). mRNAs were purified by affinity chromatography on anoligo(dT) cellulose column (Pharmacia, Uppsala, Sweden) as specified by themanufacturer.

cDNA synthesis and PCR amplification of Ig variable regions. First-strandcDNA was synthesized from the mRNA template with the First-Strand cDNAsynthesis kit (Pharmacia) with primers VH1FOR (59 TGAGGAGACGGTGACCGTGGTCCCTTGGCCCCAG 39) (21) and VK2FOR (59 CCGTTTGATCTCGAGCTTGGTGCC 39) (27) for amplification of the VH and VL regions, re-spectively. The VH regions were amplified with primers VH1FOR andVH1BACK (59 AGGTSMARCTGCAGSAGTCWGG 39) (21), in which S is Cor G, M is A or C, R is A or G, and W is A or T. Primers VK2FOR andVK2BACK (59 GACATCGAGCTCACTCAGTCTCCA 39) (27) were used foramplification of the VL regions. PCR was done for 35 cycles (1 cycle is 1 min at92°C, 1 min at 57°C, and 1 min at 72°C), in 50 ml of the following reactionmixture: 78 mM Tris-HCl (pH 8.8)–17 mM (NH4)2SO4–10 mM b-mercaptoetha-nol–2 mM MgCl2–0.05% W-1 detergent (Gibco BRL, Gaithersburg, Md.)–0.2mg of bovine serum albumin per ml–200 mM each dATP, dCTP, dGTP, anddTTP–1 mM each primer–10 ng of matrix–2.5 U of Taq DNA polymerase (GibcoBRL). The PCR products were analyzed on a 2% low-melting-point agarose-Trisacetate-EDTA (TAE) gel and visualized with ethidium bromide. PCR productsof the expected size were excised from the gel and purified with a Geneclean IIkit (Bio 101, Vista, Calif.) as specified by the manufacturer.

The fragment ends were made blunt with Klenow DNA polymerase, clonedinto the SmaI site of pBS, and introduced into XL1-Blue E. coli competent cells.

Clones pBS::VH2A10 and pBS::VL2A10 were sequenced by the dideoxy chaintermination method with the T7 DNA polymerase sequencing kit (Pharmacia)and the universal forward and reverse primers.

Construction of pUC19::scFv-Secr[2A10] for expression in E. coli. A 900-bpHindIII-EcoRI fragment of plasmid pS8, containing a Shine-Dalgarno sequence,the pectate lyase signal peptide (pelB) of Erwinia carotovora (19), the c-mycpeptide tag sequence (13), and an scFv sequence, was cloned into theHindIII-EcoRI sites of pUC19 to give pUC19::scFv-Secr[S8]. The SalI-XhoIfragment encoding VL2A10 from pBS::VL2A10 and the PstI-BstEII fragmentencoding VH2A10 from pBS::VH2A10 were ligated into the SalI-XhoI and PstI-BstEII sites of pUC19::scFv-Secr[S8], respectively, to replace the original VHand VL fragments of the scFv contained in this plasmid. The resulting plasmid,pUC19::scFv-Secr[2A10], is shown in Fig. 1A. For control experiments, a similarconstruction in which the scFv gene was cloned in the reverse orientation wasmade in pUC18 and called pINV::scFv-Secr[2A10].

Construction of 35-pel-scFv[2A10] for expression in tobacco plants. TheHindIII-EcoRI fragment of pUC19 containing the scFv[2A10] construction wasmodified to introduce XbaI and SalI sites at the 59 and 39 ends, respectively. Forthat purpose, mutagenesis by PCR was performed with Pfu DNA polymerase(Stratagene) with primers XBA (59 TCTAGACTCGAAGCTTGCATGC 39)and SAL (59 GTCGACGAATTCGAGCTGG 39). Twenty-five cycles of PCRwere allowed to take place under the conditions described above. The PCRproducts were digested with XbaI and SalI and cloned into plasmid pBG-dAb-BIN that had been linearized with the same enzymes. This produced plasmid35-pel-scFv[2A10] (Fig. 1B), which was used for plant transformation.

Expression of scFv[2A10] in E. coli. E. coli JM109 cells transfected withpUC19::scFv-Secr[2A10] were grown at 30°C on a shaker overnight in 1 volumeof Luria-Bertani medium (23) containing 120 mM glucose and 100 mM ampicil-lin. The cells were pelleted, washed twice in 1 volume of 50 mM NaCl, resus-pended in 1 volume of Luria-Bertani medium containing 100 mM ampicillin and1 mM IPTG (isopropyl-b-D-thiogalactopyranoside), and incubated for 3 h at

FIG. 1. (A) Plasmid pUC19::scFv-Secr[2A10] used for expression of the scFv in E. coli. (B) Plasmid 35-pel-scFv[2A10] used to transform tobacco plants with A.tumefaciens. NOS terminator, nopaline synthase transcription terminator; Npt II, neomycin phosphotransferase II gene.

FIG. 2. Agarose gel electrophoresis of the DNA amplified with primers pairsVH1FOR and VH1BACK (lanes 1 and 2) and VK2FOR and VK2BACK (lanes3 and 4) from water (lanes 1 and 3) or cDNAs corresponding to the H (lane 2)and L (lane 4) IgG chains. M, 1-kb ladder (Gibco BRL).

FIG. 3. Western blot analysis with antibody 9E10 of the periplasmic (lanes 1through 3) and total (lanes 4 through 6) proteins of nontransformed E. coli cells(lanes 1 and 6), E. coli cells transformed with pINV::scFv-Secr[2A10] (lanes 2and 5), and E. coli cells transformed with pUC19::scFv-Secr[2A10] (lanes 3 and4). M, Rainbow protein molecular mass marker (Amersham).

VOL. 64, 1998 PHYTOPLASMA-DIRECTED PLANTIBODIES 4567

on May 28, 2020 by guest

http://aem.asm

.org/D

ownloaded from

Page 3: Engineering of a Single-Chain Variable-Fragment (scFv ... · FABRICE LE GALL, JOSEPH-MARIE BOVE´, AND MONIQUE GARNIER* Laboratoire de Biologie Cellulaire et Mole´culaire, Institut

30°C. The periplasmic and osmotic shock fractions of E. coli cells were obtainedby a method derived from that of Dubel et al. (11). The cells were pelleted at6,200 3 g for 10 min at 4°C, the pellet was resuspended in 1/10 volume of theoriginal culture in a buffer containing 50 mM Tris-HCl (pH 8.0), 20% (wt/vol)sucrose, and 1 mM EDTA and left for 30 min on ice with occasional shaking.After centrifugation, the supernatant representing the enriched periplasmic frac-tion was stored at 4°C. The bacterial pellet was resuspended by vortexing in 1/10of the original culture volume in a buffer containing 5 mM MgSO4 and incubatedfor 30 min on ice with occasional shaking. After centrifugation at 6,200 3 g for10 min at 4°C, the supernatant representing the osmotic shock fraction wasstored at 4°C. The E. coli periplasmic and osmotic shock fractions were analyzedby sodium dodecyl sulfate-polyacrylamide gel electrophoresis (15% acrylamide)with a 1.5-mm-thick gel in a V15.17 vertical gel electrophoresis apparatus (GibcoBRL). After separation, the protein bands were transferred to a nitrocellulosemembrane (Amersham, Little Chalfont, United Kingdom) with a horizontalelectrophoretic transfer system (Biolyon, Dardilly, France). The transblottedmembrane was probed with the 9E10 anti-c-Myc antibody, washed, and incu-bated with rabbit anti-mouse IgG labelled with alkaline phosphatase (Sigma,Saint Louis, Mo.) as specified by the manufacturer. Nitroblue tetrazolium and

5-bromo-4-chloro-3-indolylphosphate p-toluidine (Boehringer GmbH, Mann-heim, Germany) were used as substrates.

Plant transformation. A. tumefaciens LB4404 was electroporated with 35-pel-scFv[2A10] plasmid as specified previously (12). Transformation of leaf disks ofP B D6 tobacco plants and regeneration were done as described previously (10).

Analysis of transgenic plants. Genomic DNA was extracted from fresh leaveswith a DNeasy plant mini kit (Qiagen S.A.). The presence of the transgene wasdemonstrated by PCR amplification with BIS (59 CTGAGCGGATAACAATTTCAC 39) and BIR (59 GACCGGCAACAGGATTCAATC 39) primers for 35cycles (1 cycle is 1 min at 93°C, 1 min at 58°C, and 1 min at 72°C) from 400 ngof plant DNA. Amplified products were analyzed on 1% agarose gels.

Northern blots. Total RNA was isolated from tobacco leaf tissue with theRNeasy mini kit (Qiagen S.A.). RNA (10 mg) was separated in 1% formalde-hyde–agarose gel and blotted onto a nitrocellulose membrane (Hybond-C extra;Amersham). The blot was hybridized with the 902-bp HindIII-EcoRI fragmentfrom pUC19::scFv-Secr[2A10], purified by the Geneclean II kit (Bio 101), andlabelled with [a-32P]dATP by the random-priming procedure (random primersDNA labeling system; Gibco BRL).

FIG. 4. IF reactions obtained with the scFv 2A10 produced by E. coli cells (A and B), tobacco leaves (D), or native hybridoma-produced IgG (C) on healthy (B)or stolbur phytoplasma-infected (A, C, and D) midrib sections. P, phloem; X, xylem. Magnification, 31,075.

4568 LE GALL ET AL. APPL. ENVIRON. MICROBIOL.

on May 28, 2020 by guest

http://aem.asm

.org/D

ownloaded from

Page 4: Engineering of a Single-Chain Variable-Fragment (scFv ... · FABRICE LE GALL, JOSEPH-MARIE BOVE´, AND MONIQUE GARNIER* Laboratoire de Biologie Cellulaire et Mole´culaire, Institut

Immunofluorescence. The immunofluorescence (IF) method described previ-ously (16) for the detection of the stolbur phytoplasma on plant sections was usedwith the following modifications to test the reaction of the scFv produced by E.coli or tobacco plants with the phytoplasma.

For the E. coli-produced scFv, sections from healthy or stolbur phytoplasma-infected plants were incubated for 30 min at room temperature with the periplas-mic or osmotic shock fractions obtained as described above. After being washedwith phosphate-buffered saline containing 0.05% Tween 20 (PBS-Tween), theywere incubated with hybridoma supernatant 9E10 for 30 min at room tempera-ture. After a second wash with PBS-Tween, the sections were incubated withanti-mouse IgG-fluorescein isothiocyanate conjugate (Sanofi Diagnostic Pasteur,Marnes la Coquette, France) as specified by the manufacturer.

For scFv produced by tobacco plants, 1 g of a 2-month-old tobacco leaf wasground into 2 ml of PBS-Tween and filtered through cheesecloth. The filtrate wasincubated for 30 min with sections of healthy or stolbur phytoplasma-infectedplants and then processed as described above. Similar extracts from normaltobacco leaves were used as controls. IF reactions in which the scFv or MAb9E10 was omitted were also performed.

ELISA. DAS-ELISA for the detection of the stolbur phytoplasma with MAb2A10 was done as described previously (15).

For scFv produced by E. coli, the microplates were coated with a twofolddilution of the E. coli periplasmic fraction in sodium carbonate buffer (pH 9.5)at 37°C for 4 h. After being washed three times with PBS-Tween, the plates wereincubated overnight at 4°C with the plant extracts obtained as described previ-ously (15). After being washed with PBS-Tween, the plates were incubated with2 mg of alkaline phosphatase-labelled 2A10 IgG (Sanofi Diagnostic Pasteur) perml for 4 h at 37°C. The alkaline phosphatase activity was revealed with p-nitrophenyl phosphate at 1 mg/ml in substrate buffer. The optical density at 405nm was measured after a 1-h incubation at room temperature.

RESULTS

Amplification of the VH and VL regions of IgG 2A10 andconstruction of the scFv. When the cDNA synthesized fromhybridoma 2A10 mRNAs was amplified by PCR with universalprimers for mouse IgG VH and VL regions, a 340-bp band wasobtained for each PCR (Fig. 2, lanes 2 and 3). These amplifiedDNAs were cloned, sequenced, and shown to correspond tothe VH and VL regions of mouse Igs.

The VH and VL DNA fragments were then linked with theflexible peptide (Gly4 Ser)3 sequence as described in Materialsand Methods. At each step, the sequences of the various cloneswere determined and compared to the initial sequences ob-tained for the VH and VL fragments to verify that the con-struction was in frame.

Cloning and expression of the scFv gene in E. coli. The scFvconstruct was cloned in pUC19 between the leader sequencepelB of Erwinia carotovora and a tag sequence coding for the11-amino-acid product of the c-myc oncogene, under the con-trol of the lacZ promoter. The expression of the construct isshown in Fig. 3, where E. coli total proteins (lanes 4 to 6) orperiplasmic proteins (lanes 1 to 3) have been separated andprobed with MAb 9E10, which is specific for the tag peptide. Aprotein of about 30 kDa, in agreement with the size of the scFvDNA, could be seen in lanes 3 and 4 corresponding to theperiplasmic and total protein fractions of transformed E. colicells. No proteins could be revealed by MAb 9E10 in nontrans-

formed E. coli cells (lanes 1 and 6) or in E. coli cells trans-formed with a plasmid in which the scFv gene was cloned in thereverse orientation (lanes 2 and 5).

Antigen-binding activity of the secreted scFv. To verify thereactivity of the E. coli-expressed scFv versus the stolbur phy-toplasma antigen, IF and ELISA reactions have been carriedout. As shown in Fig. 4, a strong green fluorescence was ob-served in the phloem tissue of stolbur phytoplasma-infectedperiwinkle plants incubated with the E. coli-produced scFv(Fig. 4A), while only the yellow-green autofluorescence of thexylem tissue, but no fluorescence in the phloem, was observedon the healthy sections (Fig. 4B). The intensity of the scFv-induced IF reaction was similar to that obtained with the nativeIgG from hybridoma supernatant 2A10 (Fig. 4C). No fluores-cence was observed with fractions obtained from nontrans-formed E. coli cells or when MAb 9E10 was omitted.

Similarly, when E. coli-expressed scFvs were used to coatELISA plates (Table 1), the OD405 obtained with stolbur phy-toplasma-infected periwinkle extracts was high (OD405 51.102) and similar to that obtained with the native IgG (OD4055 1.0). No reaction (OD405 5 0.06) was obtained with extractsprepared from healthy periwinkle plants or when the plantextract was replaced by PBS buffer (OD405 5 0.120).

Cloning of the scFv in A. tumefaciens, transformation, andanalysis of tobacco plants. The scFv construction from E. coliwas cloned into plasmid pBG-dAb-BIN of A. tumefaciens un-der the control of the 35S cauliflower mosaic virus promoter.The construct was sequenced to verify that no modification hadoccurred and was used to transform tobacco P B D6 leaf discs.

TABLE 1. DAS-ELISA detection of the stolbur phytoplasmaantigen with scFv 2A10 produced by E. coli cells

Antibody usedfor coating Antigen AP-labeled

antibody OD405

scFv 2A10 PBS buffer 2A10 IgG 0.120Stolbur-infected

periwinkle extract2A10 IgG 1.102

Healthy periwinkle extract 2A10 IgG 0.006

IgG 2A10 Stolbur-infectedperiwinkle extract

2A10 IgG 1.0

Healthy periwinkle extract 2A10 IgG 0.02

TABLE 2. Analysis of kanamycin-resistant tobacco lines by PCR,Northern blotting, and IF for the presence of the scFv gene,

mRNA, and protein

Tobacco transformantscFv present by:

PCR (BIS/BIR)a Northern IF

Nontransformed P B D635-Pel-scFv-1A1 2 2 235-Pel-scFv-1A2 2 2 235-Pel-scFv-1A3 2 2 235-Pel-scFv-1A4 2 2 235-Pel-scFv-1A5 2 2 235-Pel-scFv-1A6 1 1 135-Pel-scFv-1A7 2 2 235-Pel-scFv-1A8 1 1 135-Pel-scFv-1A9 1 1 135-Pel-scFv-1A10 2 2 235-Pel-scFv-1A11 2 2 235-Pel-scFv-5B2 1 1 135-Pel-scFv-1B1 2 2 235-Pel-scFv-1B2 2 2 235-Pel-scFv-1B3 1 1 135-Pel-scFv-1B4 1 1 135-Pel-scFv-1C1 1 1 135-Pel-scFv-1C2 1 1 135-Pel-scFv-1C3 1 1 135-Pel-scFv-1C4 1 1 135-Pel-scFv-1D2 1 1 135-Pel-scFv-2B1 1 1 135-Pel-scFv-2B2 1 1 135-Pel-scFv-2C1 1 1 135-Pel-scFv-2D1 1 1 135-Pel-scFv-4B1 1 1 135-Pel-scFv-4B2 1 1 135-Pel-scFv-5A1 1 1 1

a Primers used for detection of the transgene.

VOL. 64, 1998 PHYTOPLASMA-DIRECTED PLANTIBODIES 4569

on May 28, 2020 by guest

http://aem.asm

.org/D

ownloaded from

Page 5: Engineering of a Single-Chain Variable-Fragment (scFv ... · FABRICE LE GALL, JOSEPH-MARIE BOVE´, AND MONIQUE GARNIER* Laboratoire de Biologie Cellulaire et Mole´culaire, Institut

Twenty-eight kanamycin-resistant tobacco plants were ob-tained and studied. Table 2 shows a summary of the results ofPCR, Northern blotting, and IF reactions, which are meant todetect the transgene, its mRNA, and the expressed scFv, re-

spectively, in the kanamycin-resistant tobacco lines. The genecould be found by PCR in 18 of 28 plants. Examples of scFvgene amplification by PCR are illustrated in Fig. 5A for to-bacco transformants 1A1 to 1A8 (lanes 1 to 8) and 1B1, 1B2,and 1B3 (lanes 9 to 11). The corresponding mRNAs could bedetected in the 18 PCR-positive tobacco plants by Northernblotting. Figure 5B illustrates the Northern blots carried out onthe same tobacco plants as in Fig. 5A. Expression of a func-tional scFv protein was demonstrated by IF in all 18 tobaccoplant extracts in which the scFv gene and mRNA were de-tected. The positive IF reaction given by an extract of tobaccotransformant 1A6, observed on a transverse section of a stol-bur phytoplasma-infected tobacco plant, is illustrated in Fig.4D. Ten tobacco transformants and the normal P B D6 controlplant were negative in the three tests.

Challenge inoculation of normal and transgenic 1A6 to-bacco plants. Five shoots (3 cm long) produced from axillarybuds of the tobacco transformant 1A6 were excised and top-grafted onto stolbur phytoplasma-infected tobacco plants ofthe same variety, P B D6. As controls, five similar shoots froma normal tobacco plant also were top-grafted onto infectedplants. The tobacco plants used as rootstocks were tested in-dividually by ELISA for the presence of the phytoplasma be-fore grafting. The OD405 of the ELISA of each plant wasbetween 1.5 and 2. After 2 months, all the plants from thetransgenic tobacco shoots had grown as well as the uninfectedplants, were symptomless, and had developed flower buds.However, all the plants from the normal, nontransgenic to-bacco shoots were severely stunted, had developed typical stol-bur symptoms, including short internodes and leaf crinkle, anddid not flower. This is illustrated in Fig. 6, where plants thatdeveloped from two transgenic shoots (Fig. 6C and D) are

FIG. 5. PCR (A) and Northern (B) analysis of kanamycin-resistant tobaccotransformants. Lanes: 1 to 8, tobacco transformants 1A1 to 1A8; 9 to 11, 1B1 to1B3; NT, nontransformed tobacco plant; P, plasmid 35-pel-scFv[2A10]; M, 1-kbladder (Gibco BRL).

FIG. 6. (B to D) Normal (B) and transgenic (C and D) tobacco 1A6 plants 2 months after top-grafting onto stolbur phytoplasma-infected PBD6 rootstocks. Theyellow arrow indicates the grafting point. (A) Uninfected PBD6 tobacco plant (control).

4570 LE GALL ET AL. APPL. ENVIRON. MICROBIOL.

on May 28, 2020 by guest

http://aem.asm

.org/D

ownloaded from

Page 6: Engineering of a Single-Chain Variable-Fragment (scFv ... · FABRICE LE GALL, JOSEPH-MARIE BOVE´, AND MONIQUE GARNIER* Laboratoire de Biologie Cellulaire et Mole´culaire, Institut

compared to one obtained from a nontransgenic shoot (Fig.6B). A normal, ungrafted tobacco plant (Fig. 6A) is shown asa control.

DISCUSSION

For the first time, we have engineered, cloned, and ex-pressed in E. coli and tobacco plants a functional scFv specificfor a mollicute, i.e., the stolbur phytoplasma.

Mollicutes are strictly restricted to the sieve tubes within thephloem tissues. The cauliflower mosaic virus promoter that wehave used is known to result in expression in all plant tissues,including phloem (26). Even though further experiments areneeded to analyze our transgenic tobacco plants, the protec-tion against the phytoplasma infection witnessed in the trans-genic tobacco transformant 1A6 is an indirect indication thatthe scFv molecule is indeed present in the phloem tissue andmore precisely in the sieve tube sap, where the phytoplasmasare located. Since our scFv construction includes the leadersequence pelB, the scFv is likely to be expressed through thesecretory pathway. A construction in which the scFv is clonedunder the control of a “phloem-specific” promoter (26) willalso be undertaken.

Our attempts to detect the scFv in tobacco protein extractsby Western blotting have failed (data not shown). Generally,an immunoaffinity chromatography step is required to purifyand concentrate the scFv before Western blotting or ELISAanalysis (27). This step is time-consuming, because manyplants must be tested. In this work, we were able to rapidlyscreen the transgenic tobacco plants producing antibodies byperforming a simple IF reaction involving stolbur-phyto-plasma-infected leaf sections incubated with crude leaf extractsfrom the transformed tobacco plants. This allowed us to selectthe scFv-producing tobacco plants even before the scFv geneand mRNA were detected.

In the experiments reported here, shoots from transgenictobacco plant 1A6 expressing the stolbur phytoplasma-specificscFv were challenged with the stolbur phytoplasma being top-grafted onto tobacco plants heavily infected by the phyto-plasma. The tobacco plants from the transgenic shoots grew aswell as uninfected plants did and were symptomless in spite ofthe large phytoplasma inoculum used for this experiment. Thisindicates that the plantibody strategy is likely to provide a wayto control phytoplasma diseases. However, the experimentmust be repeated on a larger scale, which could not be donewith the F0 parental line, since only a few shoots were suitablefor grafting. The homozygote tobacco lines, obtained afterautopollenization of the F0 generation, produced in this work,will be used for such an experiment. Indeed, in this case, alarge number of tobacco plants can be inoculated by the insectvector H. obsoletus or by side-grafting. The multiplication ofthe phytoplasma in transgenic or normal plants will be easilymonitored during plant development.

ACKNOWLEDGMENTS

This work was supported by grants from the French Ministry ofEducation Research and Technology and by the “Pole Genie Bi-ologique et Medical” d’Aquitaine. Fabrice Le Gall was supported by athesis fellowship from INRA (Institut National de la RechercheAgronomique) and SEITA (Societe d’Exploitation Industrielle desTabacs et Allumettes).

We thank Rene Delon and Francois Dorlhac de Borne (Institut duTabac, Bergerac, France) for providing the P B D6 tobacco seeds andfor their contributions during transformation and regeneration of to-bacco plants. We thank Patrizia Galeffi (ENEA, Rome, Italy) forproviding us with plasmids for scFv cloning and for helpful discussions.

REFERENCES

1. Baum, T. J., A. Hiatt, W. A. Parrott, L. H. Pratt, and R. S. Hussey. 1996.Expression in tobacco of a functional monoclonal antibody specific to styletsecretions of the root-knot nematode. Mol. Plant-Microbe Interact. 9:382–387.

2. Benvenuto, E., R. J. Ordas, R. Tavazza, G. Ancora, S. Biocca, A. Cattaneo,and P. Galeffi. 1991. “Phytoantibodies”: a general vector for the expressionof immunoglobulin domains in transgenic plants. Plant Mol. Biol. 17:865–874.

3. Boudon-Padieu, E. 1996. Jaunisses a phytoplasmes de la vigne. Diagnostic,epidemiologie, et developpement des recherches. C. R. Acad. Agric. France82:5–21.

4. Bove, J. M., P. Carle, M. Garnier, J. Renaudin and C. Saillard. 1989.Molecular and cellular biology of spiroplasmas, p. 244–364, In C. Calavanand J. M. Bove (ed.), The mycoplasmas, vol. V. Academic Press, Inc., NewYork, N.Y.

5. Bove, J. M. 1993. Molecular features of mollicutes. Clin. Infect. Dis. 17:10–31.

6. Chirgwin, J. M., A. E. Przybyla, R. J. McDonald, and W. J. Rutter. 1979.Isolation of biologically active ribonucleic acid from sources enriched inribonuclease. Biochemistry 18:5294–5299.

7. Davis, R. E., E. L. Dally, D. E. Gundersen, I.-M. Lee, and N. Habili. 1997.“Candidatus Phytoplasma australiense,” a new phytoplasma taxon associatedwith Australian grapevine yellows. Int. J. Syst. Bacteriol. 47:262–269.

8. De Neve, M., M. De Loose, A. Jacobs, H. Van Houdt, B. Kaluza, U. Weilde,M. Van Montagu, and A. Depickerl. 1993. Assembly of an antibody and itsderived antibody fragment in Nicotiana and Arabidopsis. Transgenic Res.2:227–237.

9. Doi, Y., M. Teranaka, K. Yora, and H. Asuyama. 1967. Mycoplasma or PLTgroup-like microorganisms found in the phloem elements of plants infectedwith mulberry dwarf and potato witches’ broom. Ann. Phytopathol. Soc. Jpn.33:259–266.

10. Dorlhac de Borne, F., M. Vincentz, Y. Chupeau, and H. Vaucheret. 1994.Cosuppression of nitrate reductase host genes and transgenes in transgenictobacco plants. Mol. Gen. Genet. 243:613–621.

11. Dubel, S., F. Breitling, R. Kontermann, T. Schmidt, A. Skerra, and M. Little.1995. Bifunctional and multimeric complexes of streptavidin fused to singlechain antibodies (scFv). J. Immunol. Methods 178:201–209.

12. Dulk-Ras, A. D., and P. J. J. Hooykaas. 1995. Electroporation of Agrobac-terium tumefaciens. Methods Mol. Biol. 55:63–72.

13. Evan, G. J., G. K. Lewis, G. Ramsay, and J. M. Bishop. 1985. Isolation ofmonoclonal antibodies specific for human c-myc proto-oncogene product.Mol. Cell. Biol. 5:3610–3616.

14. Firek, S., J. Draper, M. R. L. Owen, A. Gandecha, B. Cockburn, and G. C.Whitelam. 1993. Secretion of a functional single-chain Fv protein in trans-genic tobacco plants and cell suspension cultures. Plant Mol. Biol. 23:861–870.

15. Fos, A., J. L. Danet, L. Zreik, M. Garnier, and J. M. Bove. 1992. Use of amonoclonal antibody to detect the stolbur mycoplasmalike organism inplants and insects and to identify a vector in France. Plant Dis. 76:1092–1096.

16. Garnier, M., G. Martin-Gros, M. L. Iskra, L. Zreik, J. Gandar, A. Fos, andJ. M. Bove. 1990. Monoclonal antibodies against the MLOs associated withtomato stolbur and clover phyllody, p. 263–269. In G. Stanek, G. H. Cassel,J. G. Tully, and R. F. Whitcomb (ed.), Recent advances in mycoplasmology.Fischer Verlag, Stuttgart, Germany.

17. Hiatt, A., R. Cafferkey, and K. Bowdish. 1989. Production of antibodies intransgenic plants. Nature 342:76–78.

18. Kenny, G. E. 1979. Antigenic determinants, p. 351–380. In M. F. Barile andS. Razin (ed.), The mycoplasmas, vol. I. Academic Press, Inc., New York,N.Y.

19. Lei, S. P., H. C. Lin, S. S. Wang, J. Callaway, and G. Wilcox. 1987. Char-acterization of the Erwinia carotovora pelB gene and its product pectatelyase. J. Bacteriol. 169:4379–4383.

20. McCoy, R. E., A. Caudwell, C. J. Chang, T. A. Chen, L. N. Chiykowski, M. T.Cousin, J. L. Dale, G. T. N. deLeeuw, D. A. Golino, K. J. Hackett, B. C.Kirkpatrick, R. Marwitz, H. Petzold, R. C. Sinha, M. Sugiura, R. F. Whit-comb, I. L. Yang, B. M. Zhu, and E. Seemuller. 1989. Plant diseases associ-ated with mycoplasma-like organisms, p. 545–640. In R. F. Whitcomb andJ. G. Tully (ed.), The mycoplasmas, vol. 5. Spiroplasmas, acholeplasmas, andmycoplasmas of plants and arthropods. Academic Press, Inc., New York, N.Y.

21. Orlandi, R., D. H. Gussow, P. T. Jones, and G. Winter. 1989. Cloningimmunoglobulin variable domains for expression by the polymerase chainreaction. Proc. Natl. Acad. Sci. USA 86:3833–3837.

22. Rosso, M. N., A. Schouten, J. Roosien, T. Borst-Vrenssen, R. S. Hussey, F. J.Gommers, J. Bakker, A. Schots, and P. Abad. 1996. Expression and func-tional characterization of a single chain FV antibody directed against secre-tions involved in plant nematode infection process. Biochem. Biophys. Res.Commun. 220:255–263.

23. Sambrook, J., E. T. Fritsch, and T. Maniatis. 1989. Molecular cloning: alaboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, ColdSpring Harbor, N.Y.

24. Schots, A. 1992. “Plantibodies”: a flexible approach to design disease resis-

VOL. 64, 1998 PHYTOPLASMA-DIRECTED PLANTIBODIES 4571

on May 28, 2020 by guest

http://aem.asm

.org/D

ownloaded from

Page 7: Engineering of a Single-Chain Variable-Fragment (scFv ... · FABRICE LE GALL, JOSEPH-MARIE BOVE´, AND MONIQUE GARNIER* Laboratoire de Biologie Cellulaire et Mole´culaire, Institut

tance. Neth. J. Plant Pathol. 98(Suppl. 2):123.25. Seemuller, E., B. Schneider, R. Maurer, U. Ahrens, X. Daire, H. Kison, K. H.

Lorenz, and E. Stackebrandt. 1994. Phylogenetic classification of plant-pathogenic mycoplasmas by sequence analysis of 16S rDNA. Int. J. Syst.Bacteriol. 44:440–446.

26. Shi, Y., M. B. Wang, K. S. Powell, E. Van Damme, V. A. Hilder, A. M. R.Gatehouse, D. Boulter, and J. A. Gatehouse. 1994. Use of the rice sucrosesynthase-1 promoter to direct phloem-specific expression of b-glucuronidase

and snowdrop lectin genes in transgenic tobacco plants. J. Exp. Bot. 45:623–631.

27. Tavladoraki, P., E. Benvenuto, S. Trinca, D. De Martinis, A. Cattaneo, andP. Galeffi. 1993. Transgenic plants expressing a functional single-chain Fvantibody are specifically protected from virus attack. Nature 366:469–472.

28. Voss, A., M. Niersback, R. Hain, H. J. Hirsch, Y. C. Liao, F. Kreuzaler, andR. Fischer. 1995. Reduced virus infectivity in N.tabacum secreting a TMV-specific full size antibody. Mol. Breed. 1:39–50.

4572 LE GALL ET AL. APPL. ENVIRON. MICROBIOL.

on May 28, 2020 by guest

http://aem.asm

.org/D

ownloaded from