_Archives
Vi rology
Arch Virol (1992) 127:185-194
© Springer-Verlag 1992 Printed in Austria
Production, characterization and use of monoclonal antibodies to grapevine virus A D, Boscia 1, E. Aslouj 2, V. Elicio 1, V. Savino 1, M. A. Castellano a, and G. P. Martelli 1 i Dipartimento di Protezione delle Piante, Universitfi degli Studi and Centro di Studio del C. N. R. sui Virus e le Virosi delle Colture Mediterranee, Bari and 2 Istituto Agronomico Mediterraneo, Valenzano, Italy
Accepted April 9, 1992
Summary. Four stable hybridoma cell lines secreting monoclonal antibodies to grapevine closterovirus A (GVA) were obtained by fusing spleen cells of immunized BALB/c mice with mouse myeloma cell line Sp 2/0-Ag 14. In ELISA all MAbs reacted with virus in leaf extracts from Nicotiana benthamiana, glasshouse-, field-, or in vitro-grown grapevines, or with cortical shavings from mature grape canes. In IEM tests, only one of the MAbs (PA 3.F 5) decorated virus particles on the entire surface. This MAb was likely induced by a surface antigenic determinant, whereas the other three MAbs (PA 3.D 11, PA 3.C 6, and PA 3.B 9) were originated by cryptotopes. Coupling polyclonal antibodies for coating protein A-sensitized plates, and monoclonal antibody conjugates for antigen detection, gave highly efficient and reproducible results for identification of GVA in field-grown grapevines.
Introduction Grapevine virus A (GVA) is the only grapevine closterovirus whose biological [4, 15], serological [7], physicochemical [1, 2], and epidemiological [5, 17] properties have been extensively investigated. Nevertheless, the low concentration and erratic distribution of virus particles in host tissues [14] and the low titre of polyclonal antisera due to the poor immunogenic power of the virus, make GVA detection in naturally infected vines difficult. In our experience, also commercial ELISA kits do not yield reproducible results. As reported in the present paper, to overcome these difficulties hybridoma technology was applied to the production of monoclonal antibodies to GVA for development of a reliable ELISA protocol.
186
D. Boscia et al.
Materials and methods Virus sources Eight different virus isolates were used in this study. Cultures were established in Nicotiana benthamiana plants grown in a temperature-controlled glasshouse. Six isolates (TO 1, PA 2, PA 3, SA 36, SA 646, FR 1) were obtained by mechanical inoculation from grapevines, and two (PAl, FG1) by mealybug transmission. Isolates F G t , PA2 and PA3 were from infected vines grown at Bari, whereas the remaining isolates originated from other Italian regions (TO 1, Dr. M. Conti; PA 1, Dr. B. Rosciglione), France (FR 1, Dr. B. Walter), and Canada (SA 36 and SA 646, Dr. P. L. Monette).
Virus purification Isolate PA 3 obtained from Vitis vinifera cv. Perricone was used for production of monoclonal antibodies. Purified virus was obtained by differential centrifugation, sucrose and cesium sulfate density gradient centrifugation [15] and dialysis against 20 vol. of phosphate buffered saline (PBS).
Production and characterization of hybridomas secreting monoclonal antibodies to GVA Six week-old BALB/c female mice (Stefano Morini, S. Polo d'Enza) were immunized by injecting 500 ~tl (250 ~tl intraperitoneal and 250 gl subcutaneous) of an emulsion of 300 ~tg purified virus in an equal volume of Freund's complete adjuvant. A second injection of 300 gg virus in Freund's incomplete adjuvant was given intraperitoneally 15 days later, and a booster injection of 400 ~tg virus without adjuvant 75 days after first immunization. Fusion was performed three days afterwards by mixing 2 x 108 splenocytes with 2.4 x 107 SP 2/0 Ag 14 myeloma cells (American Type Culture Collection, Rockville) in presence of 50% polyethylene glycol MW 1300-1600. Fused cells in HAT medium (1 x I 0 - 4 M hypoxantine, 4 x 10- 7M aminopterin, and 1.6 x 10- 5M thymidine) were distributed in 100 gl aliquots in 96 well culture plates (Sterilin, Hounslow). Two different culture media were tested, (a) Roswell Park Memorial Institute (RPMI-1640) medium containing 2 mM L-glutamine and 20% fetal bovine serum (FBS), (b) HB basal medium (Irvine Scientific, Santa Ana) containing 1 mM Na-pyruvate, 2 mM L-glutamine and 1% HB-101 supplement. One day before fusion, culture plates were seeded with two types of feeder cells (103/well), i.e., peritoneal macrophages and splenocytes from mice primed 80 days previously with 500 ~tl pristane. Fifteen to 30 days after fusion, supernatant culture fluids were screened for the presence of GVA antibodies by indirect ELISA. Hybridoma cells secreting virus-specific antibodies were cloned and subcloned by the limiting dilution method on a feeder layer of peritoneal macrophages. Subctoning was repeated twice, and selected hybridoma lines were cultured in HB 101 medium with or without 5% FBS.
Identification of GVA-specific antibodies Supernatant culture fluids were screened with indirect DAS-ELISA. Polystyrene plates were pre-coated with purified IgG (0.1 gl/wetl) from a polyclonal antiserum to GVA in 0.05 M carbonate buffer pH 9.6, incubated at 37 °C for 2 h, then washed three times (3 rain each) with PBS-T (PBS 1 x plus 0.02% Tween 20). GVA PA 3-infected N. benthamiana sap diluted 1 : 20 in extraction buffer (PBS 1 x plus 2% polyvinylpyrrolidone and 0.02% Tween 20) was added and incubated at 37 °C for 2 h, or overnight at 4 °C. Plates were washed and loaded with 100 ~tl/welI of hybridoma culture fluids diluted 1:4 in PBS. Alkaline phosphatase-conjugated rabbit antimouse IgG were added, and incubated at 37 °C for 2 h prior to addition of substrate.
Monoclonal antibodies to grapevine virus A
187
Controls consisted of healthy N. benthamiana extracts diluted 1 : 20 in extraction buffer. Reactions with absorbance values threefold those of controls or higher were regarded as positive. All wells that gave negative readings were retested with antimouse IgA and IgM conjugates.
Determination of monoclonal antibody isotypes Isotyping was done on MAbs produced in serum-free medium. Determination was indirect by eluting IgGs from protein A-sepharose columns (Sigma Chemical Co., St. Louis) with 0.1 M citrate buffer at pH 6 for IgG 1, pH 5 for IgG 2 a, pH 4 for IgG 2 b and IgG 3, and confirmed in ELISA using subclass-specific rabbit antimouse antisera (Sigma).
Production of ascitic fluids Ascitic fluids were produced by peritoneal injection of 1 to 5 x 106 hybrid cells into BALB/ c mice which had been given 500 ~tl of Freund's incomplete adjuvant on the day before.
Purification of monoclonal antibodies MAbs were purified from ascitic fluids or supernatant culture fluids by affinity chromatography on protein A-sepharose columns.
ELISA protocols for G VA identification Five different ELISA protocols [31 were tested (Table 1) using extracts from leaves of GVA-infected N. benthamiana, petioles of basal 2-month-old leaves of glasshouse- or fieldgrown grapevines, cortical shavings from mature canes or in vitro grown grapevine explants. Protein A (Pharmacia, Uppsala), and monoclonal and polyclonal antibodies were both used at a concentration of 1 gg/ml. Polyclonal antisera were A 110 (titre 1 : 16) obtained from M. Conti, and BA-453 (titre 1 : 16) produced locally in the course of previous studies. Absorbance values (Aa0s) were read with a Titertek Multiscan photometer (Flow Laboratories, Mc Lean). All grapevine accessions examined for presence of GVA were affected by different diseases with a prevalence of leafroll and rugose wood.
Table 1. ELISA protocols 1-5 used for GVA detection in host tissues 1 P A b - A g - MAb - GaM - S 2 PAb - A g - P A b E - S 3 M A b - Ag - MAbE - S 4 PAb- Ag- MAbE- S 5 Prot A - P A b - A g - M A b E - S
Ag Antigen GaM Goat anti-mouse antibody MAb Monoclonal antibodies (mouse) PAb Polyclonal antibodies (rabbit) Prot A Protein A E Enzyme (alkaline phosphatase) S Substrate (P-nitrophenyt-phosphate)
188
D. Boscia et al.
Immunoelectron microscopy Immunoelectron microscopy (decoration) was according to Milne and Luisoni [13] and gold immunolabelling as described by Hu etal. [10]. A goat-antimouse conjugate with gold particles 10 nm in diameter (Sigma) was used in these latter tests after staining with 2% aqueous uranyl acetate. Grids were examined with a Philips 210 C electron microscope.
Results
Production and characterization of monoclonal antibodies Fusion products were subdivided in four aliquots, two for each condition of culture medium and feeder cells. Each aliquot was plated in 177 culture wells. The highest number of colonies was obtained with RPMI medium containing 20% FBS, in the presence of peritoneal macrophages. Of the 112 cell cultures obtained, 13 (11.6%) reacted positively in indirect DAS-ELISA with GVA but not with healthy plant extracts. Four of these cell lines denoted PA 3.F 5, PA 3.C 6, PA 3.B 9, and PA3.D 11 were subcloned twice and propagated for characterization. All four hybridoma lines were stable. They continued to secrete large quantities of GVA-specific antibodies (direct DAS-ELISA titres between 1 : 8,000 and 1 : 32,000) after three successive cycles of freezing and thawing in liquid nitrogen (Table 2). All antibodies belonged to IgG 1 isotype. Serum-free HB medium proved suitable for antibody production in artificial culture for cell lines PA 3.F 5, PA 3.D 11 and PA 3.C 6, whereas for line PA 3.B 9 it required the addition of 5% FBS. All four hybrid cell lines induced in mice the formation of peritoneal tumors secreting antibodies.
Epitope identification Identification of reaction sites of GVA particles to the four MAbs was attempted by decoration and gold immunolabelling tests. GVA particles were uniformly decorated with or without colloidal gold along their whole length only by MAb Table 2. Properties of monoclonal antibodies specific to PA 3 isolate of GVA Monoclonal antibody
MAbPA 3.F 5 MAbPA3.Dll MAbPA3.B9 MAbPA 3.C 6
Hybridoma cell lines
PA3 F 4 F 5 C l l F 10 PA3F4DllF12F10 PA3F4B9E10G9 PA3F4C6B7F10
Isotype Dilution end points in ELISA
IgG 1 IgG1 IgG1 IgG1
indirect
direct
t : 2.048,000 1:8.192,000 1:2.048,000 1:1.024,000
1 : 32,000 1:64,000 1:8,000 1:8,000
Decoration in IEM
+ -
Ascitic fluids were used in all tests. Culture serum-free media were used for isotyping monoclonal antibodies. Indirect ELISA was protocol 1, and direct ELISA was protocol 3 (see Table 1)
Monoclonal antibodies to grapevine virus A
189
PA 3.F 5 (Fig. 1 B and C), also when used at a concentration of 1 gg/ml. Notwithstanding the clear-cut reactivity in ELISA, no decoration was visible on virus particles exposed to the other three MAbs, even at concentrations of 3 mg/ ml (Table 2). Likewise, no precipitate was apparently formed in tube precipitin tests, and no antibodies were detected on stained particles observed under the electron microscope (not shown). ELISA reactivity of all MAbs was completely destroyed by the harsh degradation treatment used for separation of protein coat subunits for polyacrylamide gel electrophoresis [11]. However, when particles were exposed to a milder dissociation procedure, using calcium chloride [8], ELISA reaction with MAbs PA 3, D 11, PA 3.B 9, PA 3.C 6, and polyclonal antiserum was retained, and was higher than with intact virus (Fig. 2). Conversely, the reactivity of MAb PA 3. F 5 was reduced by about 60%. These results could be explained by assuming that MAb PA 3.F 5 is specific for an antigenic determinant found on the surface of intact virus particles and less active in dissociated subunits.
Fig. 1. A Purified GVA preparation used for immunizing mice. Bar: 200 nm. B Two GVA particles heavily decorated along the entire surface by monoclonal antibody PA 3.F 5. Bar: 200nm. C A GVA particle submitted to double decoration by MAb PA 3.F 5 followed by goat anti-mouse colloidal gold conjugate. Labelling is uniformly distributed on the particle. Bar: 100nm
190
D. Boscia etal, 13.00
DISSOCIATED
INTACT
HEALTHY
Fig. 2. Effect of mild dissociation on the reactivity of GVA preparations to different monoclonal antibodies. Whereas MAb PA 3.F 5 reacts better with intact than with dissociated virus, the reverse is true for the other three MAbs which give exceedingly low readings with intact virus. Note that polyclonal antiserum behaves as MAbs PA 3.D 11, PA 3.B 9, PA 3.C 6, suggesting that most of the antibodies derive from cryptotopes
The other three MAbs may be cryptotope-specific, thus unable to decorate virus particles, but reacting very efficiently with dissociated virus under conditions that do not damage their determinants. Further evidence of the different origin of MAbs was obtained by competition binding assay [9]. Alakaline phosphatase-unlabelled PA 3.B 9 totally inhibited binding of PA 3.D 11 at a concentration of 10 gg/ml, whereas no competition at any concentration was observed w h e n unlabelled PA 3.F 5 used as first antibody, was challenged with labelled PA 3.D 11 (data not shown).
Serological comparison of GVA isolates Seven GVA isolates of different geographical origin were compared serologicatly among each other and with isolate PA 3 used for raising MAbs. The comparison was made with all four MAbs in E L I S A and IEM. In E L I S A (Table 3) no significant differences were observed in the reactivity of the three cryptotope-specific MAbs (D 11, B 9, and C 6) with any of the isolates. With PA 3.F 5, the French isolate F R 1 gave a lower reaction, which was consistently observed in tests where monoclonals were compared in pairs. However, in I E M tests, PA 3.F 5 decorated equally well all virus isolates, including F R 1.
GVA detection in infected hosts According to the source material, two or more of the ELISA protocols outlined in Table 1 were used in these tests.
Monoclonal antibodies to grapevine virus A
191
Table 3. Serologicaldetection of GVA isolates by indirect ELISA protocol 1 Virus isolates PA 3 PA 2 PA 1 FG 1 FR 1 TO 1 SA 36 SA 646 Healthy
Monoclonal antibodies PA 3.F 5
PA 3.D 11
PA 3.B 9
PA 3.C 6
1.04 t.40 0.66 1.84 0.19 1.69 1.19 1.07 0.04
1.37 0.52 0.64 0.78 1.28 0.55 0.51 0.63 0.05
1.35 0.58 0.77 0.95 0.80 0.86 0.64 0.79 0.06
1.12 0.46 0.74 0.76 0.94 0.70 0.93 0.85 0.04
Polyclonal All0 0.63 0.33 0.34 0.45 0.41 0.37 0.38 0.47 0.15
Figures are mean A405 values of eight replicates
N. benthamiana
ELISA procedures no. 1 (indirect), and nos. 2 and 3 (direct) gave highly reproducible results with all four MAbs when applied to GVA-infected N. benthamiana leaf extract. Thus, ELISA 3 was routinely employed for GVA detection in herbaceous hosts. Glasshouse and in vitro-grown vines These tests utilized ELISA nos. 3 and 4 with MAb PA 3.F 5. Both protocols gave satisfactory results. However, A405 readings were consistently higher with in vitro explants (data not shown), thus confirming previous observations reporting a higher concentration of GVA in tissues of in vitro-grown plants [16]. Field-grown vines All five ELISA procedures were evaluated by testing leaf samples or cortical shavings from mature canes of 63 field-grown vines, 10 of which were known to contain GVA. A preliminary screening in which only MAb PA 3.F 5 was used as coating antibody for antigen trapping (ELISA 3), showed that cotour reactions were weak and, sometimes, barely above those of healthy controls. The results, however, were not improved when other monoclonals were employed, alone or in various combinations, including a cocktail of all four (data not shown). Therefore, all tests were performed with MAb PA 3.F 5. The results of these tests (Table 4) confirmed that in the case of this plant material, MAbs did not perform satisfactorily as trapping antibodies (ELISA 3) but improved virus detection when used as second (revealing) antibodies (ELISA 1). The background given by this antibody combination was much lower than that observed with ELISA 2. Highly reproducible results were obtained with a protocol in which polyclonal antibodies (antiserum A 110 or BA-453) were used
192
D. Boscia et al.
Table 4. Comparison of five ELISA wocedures (1-5) for GVA detection in grapevine tissues
Vines
1
2
3
4
5
GVA-infecteda ELISA-negativeb
0.48 0.03
0.33 0.25
0.09 0.01
0.53 0.01
0.96 0.02
aMean values of three determinations made on 10 different grapevine accessions b Mean values of three determinations made on 53 different grapevine accessions
for trapping, and monoclonal PA 3.F 5 as second antibody (ELISA 4). A405 readings were improved by pre-coating plates with protein A (ELISA 5).
GVA distribution in naturally infected vines ELISA protocol 5 was used for a large scale survey, the preliminary steps of which involved testing a field-grown collection of varieties of different geographical origin. Out of a total of 1100 vine s examined, 138 (11.5 %) were visibly affected by rugose wood, and many of the others (percentage undetermined) by leafroll. GVA was detected in 20% of the samples, with great variation of incidence between different geographic areas. Albania and Malta had the highest frequency of GVA infection (61 and 58 %, respectively) whereas the lowest was recorded in the Maghreb (12.5 %). Italian grapevine accessions averaged a GVA infection of about 30%. Vines affected by rugose wood had the lowest infection rate (7%). Discusssion
The results of the present investigation demonstrate that low titre, a frequent shortcoming of polyclonal GVA antisera raised in rabbits, can be overcome by the production of large amounts of high-titre monoclonal antibodies in mouse ascitic fluids. The four MAbs obtained were GVA-specific and could all be used for virus detection in host tissues regardless of their type (leaf veins, petioles, bark scrapings), and conditions under which source plants were grown (field, glasshouse, in vitro culture). ELISA reactions were equally satisfactory irrespective of the origin of the MAbs, i.e., whether their formation was elicited by surface determinants (PA 3.F 5) or cryptotopes (PA 3.D 11, PA 3.B 9 and PA 3.C 6). However, only MAb PA 3.F 5 was able to decorate virus particles, lending support to the hypothesis that its determinant is an epitope of the intact virus. MAbs did not yield satisfactory results in ELISA when used for coating, possibly because absorption to polystyrene reduced their activity [12]. Thus, in the present case, the full advantages of MAb reagents could not be exploited. However, an efficient ELISA protocol was developed, in which polyclonal antibodies were employed for antigen trapping in protein A-precoated plates,
Monoclonal antibodies to grapevine virus A
193
and were followed by enzyme-conjugated MAbs for antigen detection (ELISA 5). By this procedure two main results were obtained: (a) one of the GVA isolates could be differentiated from the others (although only in ELISA), confirming that there are serological differences in GVA populations [7], (b) large scale surveys of field-grown vines are now possible, with a level of confidence higher than that afforded by previous tests in which serology or molecular probes [14] were employed. The results of a preliminary survey in which a relatively large number of samples were examined did not yield conclusive results concerning the association of GVA with a specific disease. Gugerli et al. [7] suggested that GVA may be implicated in the etiology of Rupestris stem pitting, one of' the diseases of the rugose wood complex [18]. In the course of the present study only 7% of the vines affected by rugose wood contained GVA, a very low figure considering that American data indicate Rupestris stem pitting to o ~ u r in a high proportion of grapevine accessions from Western Europe [6].
Acknowledgements Research supported by the National Research Council of Italy, Special Project RAISA, Subproject No. 2 Paper No. 358. The authors wish to thank Drs. M. Conti, the late B. Rosciglione, B. Walter and P. L. Monette for supplying virus isolates. Dr. R. Lupo for in vitro cultures and Mr. G. Netti for valuable technical assistance.
References 1. Boccardo G, D'Aquilio M (1981) The protein and nucleic acid ofa closterovirus isolated from grapevine with stem pitting symptoms. J Gen Virol 53:17%182 2. Castrovilli S, Gallitelli D (1985) A comparison of two isolates of Grapevine virus A. Phytopathol Mediterr 24:219-220 3. Clark MF, Adams AN (1977) Characteristics of the microplate method of enzymelinked immunosorbent assay for detection of plant viruses. J Gen Virol 34:475483 4. Conti M, Milne RG, Luisoni E, Boccardo G (1980) A closterovirus from stem pitting diseased grapevine. Phytopathology 70:394-399 5. Engelbrecht D J, Kasdorf GGF (1985) Association of a closterovirus with grapevines indexing positive for grapevine teafrolt disease and evidence for its natural spread in grapevines. Phytopathol Mediterr 24:101-105 6. Goheen HC (1988) Ruperstris stem pitting. In: Pearson RP, Goheen AC (eds) Compendium of grape diseases. American Phytopathological Society Press, St. Paul, p 53 7. Gugerli P, Rosciglione B, Brugger JJ, Bonnard S, Ramel ME, Tremea F (1991) Further characterization of grapevine leafroll disease. In: Proceedings of the 10 th Meeting ICVG, Volos 1990, pp 5%60 8. Hajimorad MR, Francki RIB (1989) Preparation of soluble, biologically active alfalfa mosaic virus coat protein and its CaC12-induced degradation. J Virol Methods 25: 4961 9. Hill JH (1990) Competitive binding between monoclonal antibodies to define epitope relationships or characterize immunoglobulin idiotypes. In: Hampton R, Ball E, De Boer S (eds) Serological methods for detection and identification of viral and bacterial plant pathogens. American Phytopathological Society Press, St. Paul, pp 215-221
194
D. Boscia etal.: Monoclonal antibodies to grapevine virus A
10. Hu JS, Gonsalves D, Boscia D, Namba S (1991) Use of monoclonal antibodies to characterize grapevine leafroll associated closteroviruses. Phytopathology 80:920-925 11. Laemmli UK (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T 4. Nature 227:680-685 12. Martin RR, Stace-Smith R (1984) Production and characterization of monoclonal antibodies specific to potato leaf roll virus. Can J Plant Pathol 6:206-210 13. Milne RG, Luisoni E (1977) Rapid immune electron microscopy of virus preparations. Methods Virol 6:265-281 14. Minafra A, Russo M, Martelli GP (1992) Further studies on the use of molecular probes to grapevine closterovirus A. Vitis 31:26--31 15. Monette PL, James D (1990) Detection of two strains of grapevine virus A. Plant Dis 74:898-900 16. Monette PL, James D (1990) Use of in vitro cultures of Nicotiana benthamiana for the purification of grapevine virus A. Plant Cell Tissue Org Cult 23:131-134 17. Rosciglione B, Castellano MA, Martelli GP, Savino V, Cannizzaro G (1983) Mealybugs transmission of grapevine virus A. Vitis 22:331-347 18. Savino V, Boscia D, Martelli GP (1989) Rugose wood complex of grapevine: can grafting to Vitis indicator discriminate between diseases? In: Proceedings of the 9th Meeting ICVG, Kiryat Anavim 1987, pp91-94 Authors' address: Dr. D. Boscia, Dipartimento di Protezione delte Piante, Via Amendola 165/A, 1-70126 Bari, Italy. Received April 4, 1992
Vi rology
Arch Virol (1992) 127:185-194
© Springer-Verlag 1992 Printed in Austria
Production, characterization and use of monoclonal antibodies to grapevine virus A D, Boscia 1, E. Aslouj 2, V. Elicio 1, V. Savino 1, M. A. Castellano a, and G. P. Martelli 1 i Dipartimento di Protezione delle Piante, Universitfi degli Studi and Centro di Studio del C. N. R. sui Virus e le Virosi delle Colture Mediterranee, Bari and 2 Istituto Agronomico Mediterraneo, Valenzano, Italy
Accepted April 9, 1992
Summary. Four stable hybridoma cell lines secreting monoclonal antibodies to grapevine closterovirus A (GVA) were obtained by fusing spleen cells of immunized BALB/c mice with mouse myeloma cell line Sp 2/0-Ag 14. In ELISA all MAbs reacted with virus in leaf extracts from Nicotiana benthamiana, glasshouse-, field-, or in vitro-grown grapevines, or with cortical shavings from mature grape canes. In IEM tests, only one of the MAbs (PA 3.F 5) decorated virus particles on the entire surface. This MAb was likely induced by a surface antigenic determinant, whereas the other three MAbs (PA 3.D 11, PA 3.C 6, and PA 3.B 9) were originated by cryptotopes. Coupling polyclonal antibodies for coating protein A-sensitized plates, and monoclonal antibody conjugates for antigen detection, gave highly efficient and reproducible results for identification of GVA in field-grown grapevines.
Introduction Grapevine virus A (GVA) is the only grapevine closterovirus whose biological [4, 15], serological [7], physicochemical [1, 2], and epidemiological [5, 17] properties have been extensively investigated. Nevertheless, the low concentration and erratic distribution of virus particles in host tissues [14] and the low titre of polyclonal antisera due to the poor immunogenic power of the virus, make GVA detection in naturally infected vines difficult. In our experience, also commercial ELISA kits do not yield reproducible results. As reported in the present paper, to overcome these difficulties hybridoma technology was applied to the production of monoclonal antibodies to GVA for development of a reliable ELISA protocol.
186
D. Boscia et al.
Materials and methods Virus sources Eight different virus isolates were used in this study. Cultures were established in Nicotiana benthamiana plants grown in a temperature-controlled glasshouse. Six isolates (TO 1, PA 2, PA 3, SA 36, SA 646, FR 1) were obtained by mechanical inoculation from grapevines, and two (PAl, FG1) by mealybug transmission. Isolates F G t , PA2 and PA3 were from infected vines grown at Bari, whereas the remaining isolates originated from other Italian regions (TO 1, Dr. M. Conti; PA 1, Dr. B. Rosciglione), France (FR 1, Dr. B. Walter), and Canada (SA 36 and SA 646, Dr. P. L. Monette).
Virus purification Isolate PA 3 obtained from Vitis vinifera cv. Perricone was used for production of monoclonal antibodies. Purified virus was obtained by differential centrifugation, sucrose and cesium sulfate density gradient centrifugation [15] and dialysis against 20 vol. of phosphate buffered saline (PBS).
Production and characterization of hybridomas secreting monoclonal antibodies to GVA Six week-old BALB/c female mice (Stefano Morini, S. Polo d'Enza) were immunized by injecting 500 ~tl (250 ~tl intraperitoneal and 250 gl subcutaneous) of an emulsion of 300 ~tg purified virus in an equal volume of Freund's complete adjuvant. A second injection of 300 gg virus in Freund's incomplete adjuvant was given intraperitoneally 15 days later, and a booster injection of 400 ~tg virus without adjuvant 75 days after first immunization. Fusion was performed three days afterwards by mixing 2 x 108 splenocytes with 2.4 x 107 SP 2/0 Ag 14 myeloma cells (American Type Culture Collection, Rockville) in presence of 50% polyethylene glycol MW 1300-1600. Fused cells in HAT medium (1 x I 0 - 4 M hypoxantine, 4 x 10- 7M aminopterin, and 1.6 x 10- 5M thymidine) were distributed in 100 gl aliquots in 96 well culture plates (Sterilin, Hounslow). Two different culture media were tested, (a) Roswell Park Memorial Institute (RPMI-1640) medium containing 2 mM L-glutamine and 20% fetal bovine serum (FBS), (b) HB basal medium (Irvine Scientific, Santa Ana) containing 1 mM Na-pyruvate, 2 mM L-glutamine and 1% HB-101 supplement. One day before fusion, culture plates were seeded with two types of feeder cells (103/well), i.e., peritoneal macrophages and splenocytes from mice primed 80 days previously with 500 ~tl pristane. Fifteen to 30 days after fusion, supernatant culture fluids were screened for the presence of GVA antibodies by indirect ELISA. Hybridoma cells secreting virus-specific antibodies were cloned and subcloned by the limiting dilution method on a feeder layer of peritoneal macrophages. Subctoning was repeated twice, and selected hybridoma lines were cultured in HB 101 medium with or without 5% FBS.
Identification of GVA-specific antibodies Supernatant culture fluids were screened with indirect DAS-ELISA. Polystyrene plates were pre-coated with purified IgG (0.1 gl/wetl) from a polyclonal antiserum to GVA in 0.05 M carbonate buffer pH 9.6, incubated at 37 °C for 2 h, then washed three times (3 rain each) with PBS-T (PBS 1 x plus 0.02% Tween 20). GVA PA 3-infected N. benthamiana sap diluted 1 : 20 in extraction buffer (PBS 1 x plus 2% polyvinylpyrrolidone and 0.02% Tween 20) was added and incubated at 37 °C for 2 h, or overnight at 4 °C. Plates were washed and loaded with 100 ~tl/welI of hybridoma culture fluids diluted 1:4 in PBS. Alkaline phosphatase-conjugated rabbit antimouse IgG were added, and incubated at 37 °C for 2 h prior to addition of substrate.
Monoclonal antibodies to grapevine virus A
187
Controls consisted of healthy N. benthamiana extracts diluted 1 : 20 in extraction buffer. Reactions with absorbance values threefold those of controls or higher were regarded as positive. All wells that gave negative readings were retested with antimouse IgA and IgM conjugates.
Determination of monoclonal antibody isotypes Isotyping was done on MAbs produced in serum-free medium. Determination was indirect by eluting IgGs from protein A-sepharose columns (Sigma Chemical Co., St. Louis) with 0.1 M citrate buffer at pH 6 for IgG 1, pH 5 for IgG 2 a, pH 4 for IgG 2 b and IgG 3, and confirmed in ELISA using subclass-specific rabbit antimouse antisera (Sigma).
Production of ascitic fluids Ascitic fluids were produced by peritoneal injection of 1 to 5 x 106 hybrid cells into BALB/ c mice which had been given 500 ~tl of Freund's incomplete adjuvant on the day before.
Purification of monoclonal antibodies MAbs were purified from ascitic fluids or supernatant culture fluids by affinity chromatography on protein A-sepharose columns.
ELISA protocols for G VA identification Five different ELISA protocols [31 were tested (Table 1) using extracts from leaves of GVA-infected N. benthamiana, petioles of basal 2-month-old leaves of glasshouse- or fieldgrown grapevines, cortical shavings from mature canes or in vitro grown grapevine explants. Protein A (Pharmacia, Uppsala), and monoclonal and polyclonal antibodies were both used at a concentration of 1 gg/ml. Polyclonal antisera were A 110 (titre 1 : 16) obtained from M. Conti, and BA-453 (titre 1 : 16) produced locally in the course of previous studies. Absorbance values (Aa0s) were read with a Titertek Multiscan photometer (Flow Laboratories, Mc Lean). All grapevine accessions examined for presence of GVA were affected by different diseases with a prevalence of leafroll and rugose wood.
Table 1. ELISA protocols 1-5 used for GVA detection in host tissues 1 P A b - A g - MAb - GaM - S 2 PAb - A g - P A b E - S 3 M A b - Ag - MAbE - S 4 PAb- Ag- MAbE- S 5 Prot A - P A b - A g - M A b E - S
Ag Antigen GaM Goat anti-mouse antibody MAb Monoclonal antibodies (mouse) PAb Polyclonal antibodies (rabbit) Prot A Protein A E Enzyme (alkaline phosphatase) S Substrate (P-nitrophenyt-phosphate)
188
D. Boscia et al.
Immunoelectron microscopy Immunoelectron microscopy (decoration) was according to Milne and Luisoni [13] and gold immunolabelling as described by Hu etal. [10]. A goat-antimouse conjugate with gold particles 10 nm in diameter (Sigma) was used in these latter tests after staining with 2% aqueous uranyl acetate. Grids were examined with a Philips 210 C electron microscope.
Results
Production and characterization of monoclonal antibodies Fusion products were subdivided in four aliquots, two for each condition of culture medium and feeder cells. Each aliquot was plated in 177 culture wells. The highest number of colonies was obtained with RPMI medium containing 20% FBS, in the presence of peritoneal macrophages. Of the 112 cell cultures obtained, 13 (11.6%) reacted positively in indirect DAS-ELISA with GVA but not with healthy plant extracts. Four of these cell lines denoted PA 3.F 5, PA 3.C 6, PA 3.B 9, and PA3.D 11 were subcloned twice and propagated for characterization. All four hybridoma lines were stable. They continued to secrete large quantities of GVA-specific antibodies (direct DAS-ELISA titres between 1 : 8,000 and 1 : 32,000) after three successive cycles of freezing and thawing in liquid nitrogen (Table 2). All antibodies belonged to IgG 1 isotype. Serum-free HB medium proved suitable for antibody production in artificial culture for cell lines PA 3.F 5, PA 3.D 11 and PA 3.C 6, whereas for line PA 3.B 9 it required the addition of 5% FBS. All four hybrid cell lines induced in mice the formation of peritoneal tumors secreting antibodies.
Epitope identification Identification of reaction sites of GVA particles to the four MAbs was attempted by decoration and gold immunolabelling tests. GVA particles were uniformly decorated with or without colloidal gold along their whole length only by MAb Table 2. Properties of monoclonal antibodies specific to PA 3 isolate of GVA Monoclonal antibody
MAbPA 3.F 5 MAbPA3.Dll MAbPA3.B9 MAbPA 3.C 6
Hybridoma cell lines
PA3 F 4 F 5 C l l F 10 PA3F4DllF12F10 PA3F4B9E10G9 PA3F4C6B7F10
Isotype Dilution end points in ELISA
IgG 1 IgG1 IgG1 IgG1
indirect
direct
t : 2.048,000 1:8.192,000 1:2.048,000 1:1.024,000
1 : 32,000 1:64,000 1:8,000 1:8,000
Decoration in IEM
+ -
Ascitic fluids were used in all tests. Culture serum-free media were used for isotyping monoclonal antibodies. Indirect ELISA was protocol 1, and direct ELISA was protocol 3 (see Table 1)
Monoclonal antibodies to grapevine virus A
189
PA 3.F 5 (Fig. 1 B and C), also when used at a concentration of 1 gg/ml. Notwithstanding the clear-cut reactivity in ELISA, no decoration was visible on virus particles exposed to the other three MAbs, even at concentrations of 3 mg/ ml (Table 2). Likewise, no precipitate was apparently formed in tube precipitin tests, and no antibodies were detected on stained particles observed under the electron microscope (not shown). ELISA reactivity of all MAbs was completely destroyed by the harsh degradation treatment used for separation of protein coat subunits for polyacrylamide gel electrophoresis [11]. However, when particles were exposed to a milder dissociation procedure, using calcium chloride [8], ELISA reaction with MAbs PA 3, D 11, PA 3.B 9, PA 3.C 6, and polyclonal antiserum was retained, and was higher than with intact virus (Fig. 2). Conversely, the reactivity of MAb PA 3. F 5 was reduced by about 60%. These results could be explained by assuming that MAb PA 3.F 5 is specific for an antigenic determinant found on the surface of intact virus particles and less active in dissociated subunits.
Fig. 1. A Purified GVA preparation used for immunizing mice. Bar: 200 nm. B Two GVA particles heavily decorated along the entire surface by monoclonal antibody PA 3.F 5. Bar: 200nm. C A GVA particle submitted to double decoration by MAb PA 3.F 5 followed by goat anti-mouse colloidal gold conjugate. Labelling is uniformly distributed on the particle. Bar: 100nm
190
D. Boscia etal, 13.00
DISSOCIATED
INTACT
HEALTHY
Fig. 2. Effect of mild dissociation on the reactivity of GVA preparations to different monoclonal antibodies. Whereas MAb PA 3.F 5 reacts better with intact than with dissociated virus, the reverse is true for the other three MAbs which give exceedingly low readings with intact virus. Note that polyclonal antiserum behaves as MAbs PA 3.D 11, PA 3.B 9, PA 3.C 6, suggesting that most of the antibodies derive from cryptotopes
The other three MAbs may be cryptotope-specific, thus unable to decorate virus particles, but reacting very efficiently with dissociated virus under conditions that do not damage their determinants. Further evidence of the different origin of MAbs was obtained by competition binding assay [9]. Alakaline phosphatase-unlabelled PA 3.B 9 totally inhibited binding of PA 3.D 11 at a concentration of 10 gg/ml, whereas no competition at any concentration was observed w h e n unlabelled PA 3.F 5 used as first antibody, was challenged with labelled PA 3.D 11 (data not shown).
Serological comparison of GVA isolates Seven GVA isolates of different geographical origin were compared serologicatly among each other and with isolate PA 3 used for raising MAbs. The comparison was made with all four MAbs in E L I S A and IEM. In E L I S A (Table 3) no significant differences were observed in the reactivity of the three cryptotope-specific MAbs (D 11, B 9, and C 6) with any of the isolates. With PA 3.F 5, the French isolate F R 1 gave a lower reaction, which was consistently observed in tests where monoclonals were compared in pairs. However, in I E M tests, PA 3.F 5 decorated equally well all virus isolates, including F R 1.
GVA detection in infected hosts According to the source material, two or more of the ELISA protocols outlined in Table 1 were used in these tests.
Monoclonal antibodies to grapevine virus A
191
Table 3. Serologicaldetection of GVA isolates by indirect ELISA protocol 1 Virus isolates PA 3 PA 2 PA 1 FG 1 FR 1 TO 1 SA 36 SA 646 Healthy
Monoclonal antibodies PA 3.F 5
PA 3.D 11
PA 3.B 9
PA 3.C 6
1.04 t.40 0.66 1.84 0.19 1.69 1.19 1.07 0.04
1.37 0.52 0.64 0.78 1.28 0.55 0.51 0.63 0.05
1.35 0.58 0.77 0.95 0.80 0.86 0.64 0.79 0.06
1.12 0.46 0.74 0.76 0.94 0.70 0.93 0.85 0.04
Polyclonal All0 0.63 0.33 0.34 0.45 0.41 0.37 0.38 0.47 0.15
Figures are mean A405 values of eight replicates
N. benthamiana
ELISA procedures no. 1 (indirect), and nos. 2 and 3 (direct) gave highly reproducible results with all four MAbs when applied to GVA-infected N. benthamiana leaf extract. Thus, ELISA 3 was routinely employed for GVA detection in herbaceous hosts. Glasshouse and in vitro-grown vines These tests utilized ELISA nos. 3 and 4 with MAb PA 3.F 5. Both protocols gave satisfactory results. However, A405 readings were consistently higher with in vitro explants (data not shown), thus confirming previous observations reporting a higher concentration of GVA in tissues of in vitro-grown plants [16]. Field-grown vines All five ELISA procedures were evaluated by testing leaf samples or cortical shavings from mature canes of 63 field-grown vines, 10 of which were known to contain GVA. A preliminary screening in which only MAb PA 3.F 5 was used as coating antibody for antigen trapping (ELISA 3), showed that cotour reactions were weak and, sometimes, barely above those of healthy controls. The results, however, were not improved when other monoclonals were employed, alone or in various combinations, including a cocktail of all four (data not shown). Therefore, all tests were performed with MAb PA 3.F 5. The results of these tests (Table 4) confirmed that in the case of this plant material, MAbs did not perform satisfactorily as trapping antibodies (ELISA 3) but improved virus detection when used as second (revealing) antibodies (ELISA 1). The background given by this antibody combination was much lower than that observed with ELISA 2. Highly reproducible results were obtained with a protocol in which polyclonal antibodies (antiserum A 110 or BA-453) were used
192
D. Boscia et al.
Table 4. Comparison of five ELISA wocedures (1-5) for GVA detection in grapevine tissues
Vines
1
2
3
4
5
GVA-infecteda ELISA-negativeb
0.48 0.03
0.33 0.25
0.09 0.01
0.53 0.01
0.96 0.02
aMean values of three determinations made on 10 different grapevine accessions b Mean values of three determinations made on 53 different grapevine accessions
for trapping, and monoclonal PA 3.F 5 as second antibody (ELISA 4). A405 readings were improved by pre-coating plates with protein A (ELISA 5).
GVA distribution in naturally infected vines ELISA protocol 5 was used for a large scale survey, the preliminary steps of which involved testing a field-grown collection of varieties of different geographical origin. Out of a total of 1100 vine s examined, 138 (11.5 %) were visibly affected by rugose wood, and many of the others (percentage undetermined) by leafroll. GVA was detected in 20% of the samples, with great variation of incidence between different geographic areas. Albania and Malta had the highest frequency of GVA infection (61 and 58 %, respectively) whereas the lowest was recorded in the Maghreb (12.5 %). Italian grapevine accessions averaged a GVA infection of about 30%. Vines affected by rugose wood had the lowest infection rate (7%). Discusssion
The results of the present investigation demonstrate that low titre, a frequent shortcoming of polyclonal GVA antisera raised in rabbits, can be overcome by the production of large amounts of high-titre monoclonal antibodies in mouse ascitic fluids. The four MAbs obtained were GVA-specific and could all be used for virus detection in host tissues regardless of their type (leaf veins, petioles, bark scrapings), and conditions under which source plants were grown (field, glasshouse, in vitro culture). ELISA reactions were equally satisfactory irrespective of the origin of the MAbs, i.e., whether their formation was elicited by surface determinants (PA 3.F 5) or cryptotopes (PA 3.D 11, PA 3.B 9 and PA 3.C 6). However, only MAb PA 3.F 5 was able to decorate virus particles, lending support to the hypothesis that its determinant is an epitope of the intact virus. MAbs did not yield satisfactory results in ELISA when used for coating, possibly because absorption to polystyrene reduced their activity [12]. Thus, in the present case, the full advantages of MAb reagents could not be exploited. However, an efficient ELISA protocol was developed, in which polyclonal antibodies were employed for antigen trapping in protein A-precoated plates,
Monoclonal antibodies to grapevine virus A
193
and were followed by enzyme-conjugated MAbs for antigen detection (ELISA 5). By this procedure two main results were obtained: (a) one of the GVA isolates could be differentiated from the others (although only in ELISA), confirming that there are serological differences in GVA populations [7], (b) large scale surveys of field-grown vines are now possible, with a level of confidence higher than that afforded by previous tests in which serology or molecular probes [14] were employed. The results of a preliminary survey in which a relatively large number of samples were examined did not yield conclusive results concerning the association of GVA with a specific disease. Gugerli et al. [7] suggested that GVA may be implicated in the etiology of Rupestris stem pitting, one of' the diseases of the rugose wood complex [18]. In the course of the present study only 7% of the vines affected by rugose wood contained GVA, a very low figure considering that American data indicate Rupestris stem pitting to o ~ u r in a high proportion of grapevine accessions from Western Europe [6].
Acknowledgements Research supported by the National Research Council of Italy, Special Project RAISA, Subproject No. 2 Paper No. 358. The authors wish to thank Drs. M. Conti, the late B. Rosciglione, B. Walter and P. L. Monette for supplying virus isolates. Dr. R. Lupo for in vitro cultures and Mr. G. Netti for valuable technical assistance.
References 1. Boccardo G, D'Aquilio M (1981) The protein and nucleic acid ofa closterovirus isolated from grapevine with stem pitting symptoms. J Gen Virol 53:17%182 2. Castrovilli S, Gallitelli D (1985) A comparison of two isolates of Grapevine virus A. Phytopathol Mediterr 24:219-220 3. Clark MF, Adams AN (1977) Characteristics of the microplate method of enzymelinked immunosorbent assay for detection of plant viruses. J Gen Virol 34:475483 4. Conti M, Milne RG, Luisoni E, Boccardo G (1980) A closterovirus from stem pitting diseased grapevine. Phytopathology 70:394-399 5. Engelbrecht D J, Kasdorf GGF (1985) Association of a closterovirus with grapevines indexing positive for grapevine teafrolt disease and evidence for its natural spread in grapevines. Phytopathol Mediterr 24:101-105 6. Goheen HC (1988) Ruperstris stem pitting. In: Pearson RP, Goheen AC (eds) Compendium of grape diseases. American Phytopathological Society Press, St. Paul, p 53 7. Gugerli P, Rosciglione B, Brugger JJ, Bonnard S, Ramel ME, Tremea F (1991) Further characterization of grapevine leafroll disease. In: Proceedings of the 10 th Meeting ICVG, Volos 1990, pp 5%60 8. Hajimorad MR, Francki RIB (1989) Preparation of soluble, biologically active alfalfa mosaic virus coat protein and its CaC12-induced degradation. J Virol Methods 25: 4961 9. Hill JH (1990) Competitive binding between monoclonal antibodies to define epitope relationships or characterize immunoglobulin idiotypes. In: Hampton R, Ball E, De Boer S (eds) Serological methods for detection and identification of viral and bacterial plant pathogens. American Phytopathological Society Press, St. Paul, pp 215-221
194
D. Boscia etal.: Monoclonal antibodies to grapevine virus A
10. Hu JS, Gonsalves D, Boscia D, Namba S (1991) Use of monoclonal antibodies to characterize grapevine leafroll associated closteroviruses. Phytopathology 80:920-925 11. Laemmli UK (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T 4. Nature 227:680-685 12. Martin RR, Stace-Smith R (1984) Production and characterization of monoclonal antibodies specific to potato leaf roll virus. Can J Plant Pathol 6:206-210 13. Milne RG, Luisoni E (1977) Rapid immune electron microscopy of virus preparations. Methods Virol 6:265-281 14. Minafra A, Russo M, Martelli GP (1992) Further studies on the use of molecular probes to grapevine closterovirus A. Vitis 31:26--31 15. Monette PL, James D (1990) Detection of two strains of grapevine virus A. Plant Dis 74:898-900 16. Monette PL, James D (1990) Use of in vitro cultures of Nicotiana benthamiana for the purification of grapevine virus A. Plant Cell Tissue Org Cult 23:131-134 17. Rosciglione B, Castellano MA, Martelli GP, Savino V, Cannizzaro G (1983) Mealybugs transmission of grapevine virus A. Vitis 22:331-347 18. Savino V, Boscia D, Martelli GP (1989) Rugose wood complex of grapevine: can grafting to Vitis indicator discriminate between diseases? In: Proceedings of the 9th Meeting ICVG, Kiryat Anavim 1987, pp91-94 Authors' address: Dr. D. Boscia, Dipartimento di Protezione delte Piante, Via Amendola 165/A, 1-70126 Bari, Italy. Received April 4, 1992
