Production of polyclonal antisera using recombinant coat proteins of Grapevine leafroll-associated virus 2 and Grapevine virus B

The objective of this work was to produce and characterize specifi c antisera against Brazilian isolates of Grapevine leafroll-associated virus 2 (GLRaV-2) and Grapevine virus B (GVB), developed from expressed coat proteins (CPs) in Escherichia coli, and to test their possible use for the detection of these two viruses in diseased grapevines. The coat protein (CP) genes were RT-PCR-amplifi ed, cloned and sequenced. The CP genes were subsequently subcloned, and the recombinant plasmids were used to transform E. coli cells and express the coat proteins. The recombinant coat proteins were purifi ed, and their identities were confi rmed by SDS-PAGE and Western blot and used for rabbit immunizations. Antisera raised against these proteins were able to recognize the corresponding recombinant proteins in Western blots and to detect GLRaV-2 and GVB in infected grapevine tissues, by indirect ELISA, discriminating healthy and infected grapevines with absorbances (A 405 ) of 0.08/1.15 and 0.12/1.30, respectively. Expressing CP genes can yield high amount of viral protein with high antigenicity, and GLRaV-2 and GVB antisera obtained in this study can allow reliable virus disease diagnosis.


Introduction
Among grapevine (Vitis spp.) viral diseases, in Brazil, grapevine leafroll and corky bark stand out due to their economic relevance (Fajardo et al., 2003).
Infection with grapevine leafroll causes, in Vitis vinifera L., downward curling and discoloration of leaves, in the end of the growing season, and results in reduction of yield and quality of grape berries.Additionally, GLRaV-2 is also involved in the aetiology of disorders known as graft-incompatibility, young vine decline and rootstock stem lesion (Abou-Ghanem-Sabanadzovic et al., 2000;Bertazzon & Angelini, 2004;Beuve et al., 2007).Elimination of GLRaV-2 had benefi cial impact with a marked increase in fresh fruit yield (22%), cumulative weight growth (21%), and sugar concentration of fruit juice (9%) (Komar et al., 2007).
Corky bark causes yield decreases, incomplete ripening of grapes and progressive decline and death of plants (Kuhn, 1992).Some cultivars and rootstocks are symptomless carriers or exhibit only mild symptoms, while others suffer rapid decline.Typical swelling and longitudinal bark cracks of young branches, and early drying of branches occur in V. labrusca cultivars.Elimination of GVB had also a benefi cial impact, with an increase in vigor (35%) and production (16%) (Komar et al., 2007).There is experimental proof of controled transmission of GVB by several species of pseudococcid mealybugs (Boscia et al., 1993).
Antisera and several variants of enzyme-linked immunosorbent assay (ELISA) (Clark, 1981;Koenig, 1981;Barbara & Clark, 1982) are widely used tools for the routine detection of grapevine viruses (Kuniyuki et al., 2002;Fajardo et al., 2003).However, the production of high-quality virus-specifi c antisera to GLRaV-2 and GVB, suitable for their large scale detection and based on virus purifi cation procedures, faces substantial drawbacks, including complex virus infections, low yields of virus particles, presence of inhibitory compoundssuch as polyphenols, tannins and polysaccharides -, narrow range and poor mechanical transmission to herbaceous hosts, contamination of antigens with host proteins (Uyemoto et al., 1997;Xu et al., 2006;Beuve et al., 2007;Fajardo et al., 2007;Ling et al., 2007), and, especially in the case of GVB, low immunogenicity (Boscia et al., 1993).
The production of virus-specifi c antisera using recombinant proteins from cloned virus genes expressed in Escherichia coli, which can overcome the mentioned diffi culties, has been applied to produce a number of polyclonal antisera against coat protein (CP) of several grapevine viruses (Ling et al., 2000(Ling et al., , 2007;;Minafra et al., 2000;Xu et al., 2006;Fajardo et al., 2007) and nonstructural proteins (Saldarelli & Minafra, 2000).Expressed CP genes can yield high amounts of viral protein with adequate antigenicity, thereby surpassing the already mentioned drawbacks of viral antigen purifi cation from infected tissue, as well as contamination of antigens with plant proteins (Uyemoto et al., 1997;Ling et al., 2007).
The large number of nowadays available nucleotide sequences of GLRaV-2 (Meng et al., 2005) and GVB (Saldarelli et al., 1996;Nickel et al., 2002), associated with the last development of molecular biology techniques, have turned expressing viral genes into an extremely powerful strategy for the production of antibodies against plant viruses.Despite the development of nucleotide sequence-based detection methods, ELISA remains highly relevant for large scale identifi cation of grapevine virus (Fajardo et al., 2003).Thus, the availability of GLRaV-2 and GVB antisera, with adequate sensitivity and specifi city, will allow reliable virus disease diagnosis in grapevines from vineyards and indexing materials from virus-free clone production programs.
The objective of this study was to produce and characterize specifi c antisera against Brazilian isolates of GLRaV-2 and GVB, developed from CP expressed in E. coli, and to test the possible use of the raised antisera for the detection of these two viruses in diseased grapevines.
One recombinant clone, containing the CP (rCP) gene of each GLRaV-2 (GLRaV-2cp, 597 bp) and GVB (GVBcp, 594 bp), was cultured for 8 hours at 37°C, in LB medium with 100 µg mL -1 ampicilin (LB/ ampicilin), and the plasmid DNA was purifi ed using the Flexi Prep kit (Amersham Biosciences), according to the manufacturer's instructions.The GLRaV-2cp and GVBcp genes were removed from the cloning plasmids, by digestion with EcoRI, and ligated to the expression vectors pRSET-A and pRSET-C (Invitrogen), respectively, previously digested with the same enzyme.
In order to confi rm the correct orientation of the inserts (CP) into pRSET-A and -C, the obtained recombinant plasmids were sequenced, PCR was performed using T7 foward primer placed in pRSET-C and complementary sense primer (GVB7038) or plasmids were digested with the restriction enzyme (PstI), with a unique restriction site present both in vector pRSET-A and in the CP gene of GLRaV-2.
For in vitro expression, the constructs pRSET-A/ GLRaV-2cp and pRSET-C/GVBcp were transferred to E. coli strain BL21:DE3 by heat shock, and one colony of each clone was cultivated at 37°C in 200 mL LB/ ampicilin, until an OD 600 of approximately 0.5, when expression was induced by addition of isopropylbeta-D-thiogalactopyranoside (IPTG) to a fi nal concentration of 2 mM.Six hours post-induction, the bacterial cells were collected by centrifugation (5,000 g/10 min) and stored at -80°C.Total protein extracts were obtained by resuspension in lysis buffer (50 mM Tris-HCl, 100 mM NaCl, 2 mM EDTA, pH 8), lysozyme treatment and sonication, as described by Noueiry et al. (1994).The CP extract, resuspended in 1 mL of 100 mM NaHCO 3 , pH 9, plus 0.5% SDS (w/v), was purifi ed by affi nity cromatography in Ni-NTA columns (Qiagen), according to the manufacturer's instructions.GLRaV-2cp and GVBcp expressions were evaluated by 4%/12% discontinuous polyacrylamide gel electrophoresis (SDS-PAGE) and Western blot, using commercial antisera against the CPs of GLRaV-2 (Sanofi -Pasteur) or GVB (Agritest), respectively, and antibodies against the six histidine tag (anti-his C-terminal/AP Ab) (Invitrogen).The protein quantifi cation was done by comparing one aliquot of GLRaV-2 or GVB expressed CP with serial dilution aliquots of BSA (0.25, 0.5, 1, 1.4, 2 and 2.5 mg mL -1 BSA) in SDS-PAGE, since Bradford reagent was incompatible with the SDS concentration present in recombinant protein solutions.
After dialysis in 10 mM phosphate buffer, pH 7.4, plus 0.425% NaCl (w/v), 180, 250, 400, 500 and 670 µg of the in vitro-expressed denatured protein were injected intramuscularly into the hind legs of a white, New Zealand rabbit, approximately 35 days old, at weekly intervals.One rabbit was used for each expressed CP.The fi rst injection was performed with complete Freund's adjuvant (1:1 v/v), and the four remaining injections with incomplete Freund's adjuvant (1:1 v/v).Beginning one week after the last injection, seven weekly bleedings were carried out (25-30 mL bleeding by animal).Blood samples were allowed to coagulate for 1 hour at 37 o C, and 30 min at 4 o C, and then centrifuged at 3,000 g/10 min.The antisera were aliquoted and stored at -20 o C.
Western blot tests were performed to confi rm the effective production and specifi city of antibodies against GLRaV-2 and GVB expressed CP. Aliquots (2 µL) of these proteins were eletrophoresed in 12% SDS-PAGE gels and transferred onto a nitrocellulose membrane using a wet blotter (BioRad).The membrane was blocked (2% w/v powdered milk), incubated with crude GLRaV-2 and GVB antisera (1:250 v/v), washed, incubated with goat-anti-rabbit antibodies-AP conjugate (Sigma), and revealed with NBT/BCIP as substrate.
Purifi cation of the globuline fraction (IgG) from the antisera was carried out by ion exchange chromatography, in a DEAE-sephacel resin (Sigma) equilibrated with 25 mM sodium acetate buffer, pH 5.2, and the antisera was diluted 1:10 (v/v) in distilled water.An equal volume of saturated ammonium sulphate was used to precipitate proteins in antisera.The suspension was centrifuged at 3,000 g/10 min, the pellet was resuspended in 2 mL of half-strength PBS and dialysed for 12 hours at 4 o C in the same buffer.Globuline fractions were quantifi ed at 280 nm and stored at -20 o C.
Plate-trapped antigen-ELISA (PTA-ELISA), a type of indirect ELISA (Clark, 1981), was performed by using purifi ed IgG produced against rCP of GLRaV-2 and GVB (at 4 µg µL -1 ) and general conjugate (at 1:1,000 v/v) (goat-anti-rabbit antibodies-alkaline phosphatase conjugate, Sigma).This allowed evaluating the specifi city and sensitivity of the obtained antisera against GLRaV-2 and GVB, in order to detect these viruses in grapevines previously tested positive for them.The samples, consisting of leaf veins and petioles of six GLRaV-2-infected grapevines and six GVB-infected grapevines, were powdered in the presence of liquid nitrogen and diluted 1:3 (w/v) in coating buffer (50 mM sodium carbonate, pH 9.6).Samples were considered infected, when absorbance at 405 nm was at least twice the average value of the healthy controls.

Results and Discussion
Expressions of GLRaV-2cp and GVBcp were induced four times each in 200 mL cultures, resulting in 0.750, 3.125, 0.750 and 0.875 mg CP of GLRaV-2 (a total of 5.5 mg) and 0.750, 1.625, 0.875 and 1.250 mg CP of GVB (a total of 4.5 mg).Compared to other results (Ling et al., 2000(Ling et al., , 2007;;Fajardo et al., 2007) on fusion protein expressions, the average protein yields obtained in the present study (6.875 and 5.625 µg mL -1 culture medium of GLRaV-2cp and GVBcp, respectively) are in the range of similar studies and adequate for the required schedule of rabbit immunizations.
In SDS-PAGE gels, the presence of bands with a molecular mass corresponding to approximately 25 and 26 kDa (expected values for GLRaV-2 and GVB fusion proteins, respectively) was observed (Figure 1).Of these values, ca.22 and 23 kDa corresponded to the CPs of GLRaV-2 and GVB, respectively, increased by approximately 3 kDa of six histidine fused to the N-terminal of these proteins (6 His-tag).The band corresponding to the fusion protein was absent in the sample transformed with the empty vector.
The identities of the expressed proteins were confi rmed by Western blot, using commercial antisera against the CP of GLRaV-2 or GVB, and antibodies against the 6 His-tag (Figure 2).
Testing the produced antisera for the detection of expressed GLRaV-2 and GVB CPs by Western blot revealed that they reacted strongly and specifi cally to the homologous immunogens (Figure 3), which demonstrates the effectiveness of the immunization procedure, as well as the immunogenicity of the fusion protein.
The purifi ed IgG were obtained in concentrations of 1,396 and 2,010 µg mL -1 for GLRaV-2 and GVB, respectively.Their excellent recognition of GLRaV-2 and GVB in infected grapevines was confi rmed by PTA-ELISA (Table 1).Neither of the antisera showed relevant absorbance with healthy samples, which demonstrates high sensitivity and specifi city (Tables 1  and 2).Considering absorbance value range, obtained as a result of routine indexing tests performed in the lab, the performance of GLRaV-2 and GVB produced antisera was similar to that obtained with commercial antibodies.PTA-ELISA was chosen to check the antisera, because of its low cost, reliability and Pesq.agropec.bras., Brasília, v.43, n.10, p.1405Brasília, v.43, n.10, p. -1411Brasília, v.43, n.10, p. , out. 2008 practicability as a large scale virus diagnostic method (Ling et al., 2000).Also indirect ELISA variants showed to yield a broader specifi city in heterologous reactions and were reliably sensitive, thus are more suitable for large scale analyses (Koenig, 1981;Barbara & Clark, 1982).Differences observed in the nucleotide or amino acids sequences of the CP gene may have an effect on all protein functions, including the serological relationship.Therefore, the CP gene variability should be considered, since some epitopes of the coat protein are the basis for antibody recognition by ELISA.It was demonstrated that GLRaV-2 and GVB Brazilian isolates show 94-98% and 82% nucleotide identities, respectively, with foreign isolates from different geographic origins (Nickel et al., 2002).Thus, it is reasonable to assume that antisera produced in this study could not recognize all foreign isolates or strains, and vice-versa, highlighting the importance of antisera production against local isolates.GLRaV-2 occurs in nature as a number of biological and molecular variants, grouped into distinct strains, based on their CP gene sequences, such as PN (= Sem, 94/970), 93/955, H4, RG, Alfi e and BD (Bertazzon & Angelini, 2004;Meng et al., 2005).A similar observation was done with GVB (Shi et al., 2004).Despite this fact, the antisera against GLRaV-2, isolate L/I, produced in this study, reacted positively to several GLRaV-2 isolates, including fi ve GLRaV-2 isolates (Table 2), which showed among them 87-99% and 90-99% of nucleotide and deduced amino acids identities, respectively, in the coat protein gene, according to nucleotide sequences at GenBank (http://www.ncbi.nlm.nih.gov).Similar results were Table 1.Absorbance values at 405 nm, observed in a plate-trapped antigen-ELISA using grapevine (Vitis spp.) samples and antisera against GLRaV-2 and GVB fusion proteins, expressed in Escherichia coli cells.Conditions: IgG at 4 µg mL -1 and conjugate at 1:1000 (v/v).
obtained, using the produced antiserum against GVB, isolate BR1, to detect the virus in other GVB-infected grapevines, but, in this case, the variability of GVB isolates was not molecularly characterized as GLRaV-2.These results show that the produced antisera reliably detected several molecularly different isolates of GLRaV-2 (Table 2), as well as GVB, in six different infected grapevine accesses of unknown origin.
The variable viral titer, in GLRaV-2-infected grapevine plants, was demonstrated in two indirect ELISA tests, spaced one and a half month from each other.In the fi rst test, the absorbance value of GLRaV-2-infected grapevine cultivar Riesling was 1.15, and in the second one it was 0.42, considering precisely the same experimental conditions (Tables 1  and 2).The titer fl uctuation is well known of many grapevine viruses and determines when, how and where the sample to be analysed must be collected to perform serological tests (Fajardo et al., 2003).
The present results confi rm previous reports stating the suitability of antisera obtained against recombinant CP for detection of viruses by serological tests (Xu et al., 2006;Ling et al., 2007), including indirect ELISA.

Conclusions
1. Antigens of GLRaV-2 and GVB can be conveniently obtained by gene expression in bacteria.
2. Expressing CP genes can yield high amounts of viral protein with high antigenicity, and surpass the drawbacks of viral antigen purifi cation from infected tissue, as well as contamination of antigens with plant proteins.