Introduction

Yellow mosaic disease is a major constraint in improving productivity of grain legumes. The disease caused by whitefly transmitted geminiviruses is endemic to Indian sub-continent and south east Asian countries. There are two virus species [1], Mungbean yellow mosaic India virus (MYMIV) and Mungbean yellow mosaic virus (MYMV) that infect legume species causing an annual yield loss of about $300 million [2]. The typical symptoms of the disease are yellow mosaic or golden mosaic and stunting. MYMIV and MYMV belong to the genus Begomovirus in the family Geminiviridae comprising viruses having typical geminate particle morphology encapsidating a circular single stranded DNA genome of 2.7 kb length. The family [3] consists of four genera, Mastrevirus, Topocuvirus, Curtovirus and Begomovirus, distinguished based on their host range, genome and vector transmission.

The genus Begomovirus includes whitefly transmitted viruses that infect dicotyledonous plants and have either monopartite (DNA A) or bipartite (DNA A and DNA B) genomes. The DNA A component encodes for coat protein (CP, V1/AV1) in viral strand and replication initiation protein (Rep, C1/AC1), transcription activator protein (TrAP, C2/AC2), replication enhancer protein (REn, C3/AC3) and PTGS suppressor (C4/AC4) on complementary strand.

The DNA A component of Old World begomoviruses differs from those of New World in having an additional virion sense open reading frame upstream of coat protein gene. This ORF designated as ORF AV2 or V2 is also referred to as pre-coat protein [4]. An ORF is present in analogous position in leaf hopper transmitted, mastre (ORF V1) and curtovirus (ORF V2). Mutational analysis and localization studies have shown that in Maize streak virus, V1 protein is a symptom determinant [5] does not bind to DNA and is localized to cell periphery and facilitates movement of coat protein-DNA complex [6]. However, in Beet curly top virus, V2 does not affect cell to cell movement [7, 8]; modulates single stranded (ss) and double stranded (ds) DNA level and V2 mutants cause attenuation of symptoms. Attenuation of symptoms and reduction in viral DNA level were also observed in plants inoculated with AV2 mutants of a bipartite begomovirus Tomato leaf curl New Delhi virus [4]. In a monopartite begomovirus, Tomato yellow leaf curl virus-Israel, AV2 protein enhanced CP mediated nuclear export of viral DNA and exhibited limited capacity for cell to cell trafficking [6]. Ability of AV2 protein to suppress RNA silencing was demonstrated for Tomato yellow leaf curl virus-Israel [9] and East African cassava mosaic Cameroon virus [10]. From perusal of literature available, it is clear that the function of AV2/V2 is yet to be understood. In this communication, we present results on the expression of MYMIV-AV2 in bacterial system and show that recombinant protein influences cleaving and ATPase activity of Rep.

Materials and methods

Virus source

The functions of AV2 was investigated for blackgram (Bg) and cowpea (Cp) isolates of Mungbean yellow mosaic India virus (MYMIV-[IN:ND:Bg3:91] GenBank accession No. AF126406 and MYMIV-[IN:ND:Cp7:98] GenBank accession No. AF481865. These two isolates were chosen because in one of the isolates, the predicted AV2 protein is truncated due to C’ terminal deletion. The predicted protein for Bg isolate has 74 residues and in Cp isolate, the predicted protein is a full length protein comprising 113 amino acid residues.

Expression of AV2 (pre coat protein) using pMAL-p2X

The pMAL-p2X vector (New England Biolab) was used for cloning of ORF AV2 which was inserted downstream from the MAL E gene of Escherichia coli. The ORF AV2 was amplified from the full length DNA A clone using the following primers. The common forward primer is MYMIV AV2FP 5′ CATGAATTCATGTGGGATCC 3′ and reverse primer for Bg isolate is MYMIV BgAV2RP 5′ TGAAAGCTTTCAATCTCCTCC 3′ and for Cp isolate, MYMIV CpAV2RP 5′ TAGAAGCTTCTATACAGTCGG 3′. The calculated length of the amplified fragment was 222 and 339 bp for Bg and Cp isolates, respectively, and they were cloned between HindIII and EcoRI sites in pMAL-p2X vector. E. coli Rosetta strain cells were transformed with the recombinant plasmid and induced to express fusion protein MBP-BgAV2 and MBP-CpAV2. The crude lysate of the bacteria contained 52 and 55 kDa induced AV2 protein. Majority of the protein was solubilized easily by subjecting to sonication. Following is the protocol developed for large scale preparation of MBP-AV2 fusion protein. 500 ml culture of the E. coli strain Rosetta containing MBP-AV2 construct was grown to 0.5–0.6 OD at 600 nm and induced with 0.4 mM IPTG. Cultures were allowed to grow for 6 h at 32°C for Bg and for 15 h at 22°C for Cp. The cells were harvested by centrifugation at 4,000×g for 20 min and the pellet was resuspended in 60 ml column buffer, 50 mM Tris-HCl, pH 7.5; 200 mM NaCl; 1 mM EDTA, 10 mM β mercaptoethanol, 10% glycerol and 1 mM PMSF. After two cycles of freezing and thawing, cells were sonicated for short pulses of 10 sec for 2 min.

The lysate was centrifuged at 9,000×g at 4°C for 30 min. The supernatant was collected and bound to amylose resin (equilibrated with column buffer) by gently mixing on shaker for 2 h at 4°C. The column was washed with 20 column volumes of column buffer with 0.1% Triton X-100 and finally eluted with four column volumes of column buffer containing 10 mM maltose. After checking on a 12% SDS-PAGE, fractions containing higher amount of protein were pooled and dialysed against 50 mM Tris HCl, pH 8.0, 100 mM NaCl, 40% glycerol. The protein samples were divided into aliquots and stored at −20°C until further use. For biochemical assay, the amylose resin purified proteins were further purified through Heparin Sepharose CL-6B (Amersham Biosciences) column equilibrated with 50 mM Tris HCl, pH 8.0, 100 mM NaCl. The protein was bound to the matrix with rocking for 2 h at 4°C. The protein bound Sepharose beads were washed with 10 column volumes of buffer containing 150 mM NaCl. The MBP-AV2 proteins were eluted with NaCl gradient from 150 mM to 1.5 M. The pure protein was eluted at 1.2 M NaCl, 50 mM Tris-HCl which was dialyzed against 50 mM Tris HCl pH 8.0, 100 mM NaCl, 30% glycerol and used for further study. For biochemical assay, MBP protein alone was expressed from the vector, purified and used as control.

Polyclonal antibody to ORF AV2 of Bg isolate was prepared by immunizing white rabbit using MBP-BgAV2 protein in 1× column buffer and IgG was purified by affinity chromatography. The full length Rep protein of Bg isolate was over expressed in E. coli BL-21 (DE3) strain and purified to near homogeneity under non denaturing conditions as described previously [11].

Nucleic acid binding assay using South-western analysis

MYMIV-Bg isolate coat protein gene (CP) DNA fragment (~700 bp) was labelled with [α-32P]-dATP using the nick translation method [12]. The labelled DNA was purified from the unincorporated radio-activity by Sephadex G-50 column chromatography. Half of the labelled DNA was boiled for 5 min to make single stranded DNA (ssDNA) probe, and immediately kept on ice for 10 min. The rest of DNA was used as dsDNA probe. Five micrograms of MBP-AV2 fusion proteins of Bg and Cp were electrophoresed in SDS-PAGE along with MBP tagged coat protein (MBP-CP, [13]) of MYMIV and MBP proteins as positive and negative controls, respectively. The protein was transferred to nitrocellulose membrane (NCM) using semi- blot transfer apparatus (BioRad). The membrane was air dried followed by incubation in pre-hybridization solution [1× binding buffer (10 mM Tris-HCl, pH 7.5, 50 mM KCl, 1 mM DDT), 3% BSA] for 2 h at room temperature. The ssDNA and dsDNA probes were added along with boiled and sheared Salmon sperm DNA (50 μg/ml) to hybridization bottle separately. The hybridization was carried out for 3 h at room temperature.

Washing was performed thrice at room temperature using wash buffer (1× binding buffer, 3% BSA/5% nonfat dry milk, and 0.5 mM DTT). The NCM was air-dried and autoradiographed. The autoradiograph was analyzed using phosphorimager (Amersham Biosciences, Typhoon 9210).

ATPase assay

To investigate the ability of the AV2 protein of Bg and Cp isolates to hydrolyse ATP, ATPase assay was carried out essentially as described by Bagewadi et al. [14]. The buffer used for the ATPase reaction contained 20 mM Tris-HCl (pH 8.0), 1 mM MgCl2, 100 mM KCl, 8 mM DTT and 80 μg/ml BSA. In brief, 1 μl of [γ-32P]-ATP (10 μCi, 6000 Ci/mmol) was diluted 50-fold with 5 mM ATP. 1 μl of dilute radiolabelled ATP was mixed with the desired amounts of protein in ATPase reaction buffer and incubated at 37°C for 30 min. On completion of the reaction, 1 μl of the reaction mix was spotted onto a thin layer chromatography (TLC) plate, (polyethyleneimine PEI, Sigma-Aldrich, USA), air-dried and chromatography was performed using 0.5 M LiCl and 1 M HCOOH as the running solvent at room temperature for ~25 min. Following completion of chromatography, the TLC plate was air-dried and autoradiographed using phosphorimager (Amersham Biosciences, Typhoon, 9210).

Monitoring nicking activity

Nicking activity of MYMIV-Rep protein in the presence of AV2 protein was examined. A 26-mer oligonucleotide containing sequence from hairpin region of MYMIV DNA A, 5′-CGACTCAGCTATAATATTACCTGAGT-3′ was used for nicking assay. The 26-mer was 5′-end labelled as described in Sambrook and Russel, 2001 using T4 polynucleotide kinase. As Rep nicks specifically at TAATATT↓AC sequence, the nicked product that could be visualized was a 5′-labelled 18-mer. About 1–2 ng of radiolabelled 26-mer (specific activity ~10,000 cpm/μl) was incubated with 50 ng of Rep protein in absence or presence of MBP-AV2 fusion protein ranging from 1–4 ng, in 40 μl reaction mixture in a buffer (25 mM Tris-HCl, pH 8.0, 75 mM NaCl, 2.5 mM EDTA, 10 mM MgCl2, 5 mM DTT) at 37°C for 1 h. The reaction was terminated by adding 5 μl of stop buffer (1% SDS, 25 mM EDTA, 10% glycerol) and 12 μl loading dye. The reaction mix was boiled at 100°C for 5 min and the products were resolved on 15% native PAGE and autoradiographed using phosphorimager.

Results

The crude lysates of bacterial clones having MBP-BgAV2 and MBP-CpAV2 inserts were electrophoresed in SDS-PAGE. Coomassie blue staining revealed a band of high intensity in the lysate from induced cells compared to lysate from uninduced cells. The molecular weight of fusion protein was ~52 kDa for Bg isolate. In the case of Cp isolate there were two bands one at 55 kDa and one at 42 kDa, due to proteolytic cleavage. Concentration of IPTG, 0.1–0.5 mM was evaluated for induction and 0.4 mM was found to be suitable. Growth conditions required to express the protein were evaluated and found that MBP-BgAV2 expressed well when grown at 32°C for 6 h post induction. However, for Cp isolate the temperature had to be reduced to 22°C for 16 h to produce MBP-CpAV2 fusion protein. Sonication for 2 min, 10 sec pulse each time, with gap to cool gave good yield of protein. The MBP-AV2 protein after sonication were purified through amylose resin and eluted with 10 mM maltose second and third fractions that showed higher concentration (Fig. 1) were pooled and used for further study.

Fig. 1
figure 1

SDS-PAGE electrophoresis of purified MBP-CpAV2 protein. Recombinant proteins were purified by amylose resin column. The purified proteins were eluted from amylose column in different fractions in 10 mM maltose. M MW marker, lanes 2–7 aliquots of 10 mM eluted fractions

Full size image

The amount of expressed AV2 in the soluble fraction of E. coli was reasonably high for Bg isolate (1 mg/100 ml culture) compared to Cp isolate (0.3 mg/100 ml culture). Both the proteins reacted to MBP antibody and to the antibody raised against MBP-BgAV2 in western blot. In western blot analysis of crude protein extract of MYMIV inoculated plants, very faint bands were detected with MBP-BgAV2 antibody (data not shown).

In South-western assay, AV2 protein was tested for its DNA binding activity, using labelled Bg isolate coat protein gene fragment. Autoradiogram did not show any hybridization signal indicating absence of binding between the protein and the probe in both Bg and Cp isolates. It is inferred that ORF AV2 does not bind ss or ds DNA.

In the experiment on ATPase activity, liberation of radioactive orthophosphate from [γ-32P]-ATP by AV2 protein was not seen. It is concluded that AV2 does not have ATPase activity (Fig. 2 lanes. 7, 2).

Fig. 2
figure 2

Modulation of ATPase activity of Rep by AV2. An autoradiogram showing the release of labelled phosphate from [α-32P]-ATP using 100 ng 6×His-Rep (lanes 3–6, 8–11) and in the presence of increasing amounts of MBP-BgAV2 (lanes 3–6) and Cp (lanes 8–11); lane 1 shows the untreated labelled ATP, lane 2 Rep protein alone, lane 7 and 12 MBP-BgAV2 and MBP-CpAV2 alone, lane 13 MBP + Rep, lane 14 ATP incubated with 100 ng of MBP alone

Full size image

ATPase activity of Rep in the presence of AV2

The release of radioactive phosphate (Pi) from (γ−32)-ATP by Rep protein of Bg isolate in the presence of AV2 protein was examined. The autoradiogram of TLC plate showed increase in the hydrolysis of ATP by Rep in the presence of AV2. Figure 2, lane 2 shows the ATPase activity by Rep protein alone. Compared to that, in the presence of either BgAV2 or CpAV2 (Fig. 2, lanes 3–6 and 8–11) increase in hydrolysis of ATP by Rep was seen. There was at least two folds increase in the presence of AV2. This increase was seen with even as low as 1 ng of AV2 and continued till 2–4 ng of AV2, after which the increment reaches a plateau.

Nicking activity of Rep in the presence of AV2 protein

Rep protein of geminiviruses which is highly conserved in all members is a sequence specific DNA binding protein with site specific nicking and closing activity. The Rep protein is the only virus encoded protein required for initiation and termination of rolling circle replication. The Rep protein cleaves the phosphodiester bond between the seventh and eighth nucleotide of the nonanucleotide motif (TAATATT↓AC). As nicking activity of the Rep protein is its key function, the modulation of such activity in the presence of AV2 was studied.

6×His-Rep was incubated along with AV2 protein at different concentration in a reaction mixture having 5′ end labelled 26-mer, containing the nonanucleotides TAATATT↓AC, sequence (nicking site is indicated by the arrow); the 26-mer was expected to get cleaved into 18-mer and 8-mer. As 26-mer is 5′ end labelled, the 18-mer released can be visualized after autoradiography. MBP-AV2 protein alone and MBP with Rep served as controls. Figure 3, lane 2 shows the nicking activity of 6×His-Rep, which is considerably low; compared to the activity shown by Rep alone (lane 2), in the reaction having Rep along with MBP-BgAV2 or MBP-CpAV2 protein, there was tremendous (10 folds) increase in the nicking activity as is evident from Fig. 3 lanes 3–5 for MBP-BgAV2, lanes 7–9 for MBP-CpAV2. The increase in nicking activity was seen when 60 ng of Rep was mixed with 1–4 ng of AV2 protein. However, AV2 concentration higher than 20 ng had not increased the cleavage further (Fig. 3, lanes 3–5 for MBP-BgAV2 and lanes 7–9 for MBP-CpAV2). In the reaction of Rep protein with MBP protein, there was no increase in the release of 18-mer; activity was comparable with activity seen with Rep protein alone (Fig. 3, lanes 11 and 12). MBP protein alone did not show any nicking activity (Fig. 3, lane 13), confirming that in the MBP-AV2 fusion protein it is AV2 which is causing the increase in Rep nicking activity.

Fig. 3
figure 3

Modulation of nicking activity of Rep by AV2. Autoradiogram of a 15% PAGE/urea gel showing the γ-32P labelled 26-mer substrate (lane 1) and 18-mer cleaved products (lanes 2–5 and 7–9). The substrate was treated with 60 ng of 6×His-Rep alone (lane 2) or in the presence of 20, 60 and 100 ng of MBP-BgAV2 or MBP-CpAV2 proteins (lanes 3, 4, 5 and 7, 8, 9) or with only MBP-BgAV2/CpAV2 (lanes 6 & 10)

Full size image

Discussion

MYMIV is a member of the genus Begomovirus in the Geminiviridae family comprising plant viruses with single stranded circular genome of 2.7 kb length. Yellow mosaic disease caused by MYMIV and MYMV is a major constraint in improving the productivity of grain legumes. Management of the disease through host resistance is the only strategy to mitigate yield loss. In the absence of reliable source of resistance to YMVs, viral derived sequences are looked upon to confer resistance to viruses. In this context, it is necessary to understand the biological role of viral gene products before they are introduced as transgene.

Investigation on begomoviruses and monopartite mastreviruses have suggested that, V2/AV2 upstream of coat protein gene is a symptom determinant, facilitates movement of CP-DNA complex, and interferes with viral DNA accumulation. In all these studies inoculation with wild type and AV2 mutants, or transient assays have been employed to ascertain the role of AV2 protein.

In the present study, AV2 protein has been expressed in bacterial system and the expressed recombinant protein was used to analyze its function. In our earlier studies [15], we found that inoculation of legume hosts with AV2 mutants led to reduced viral DNA accumulation similar to results of Padidam et al. [4], with Tomato leaf curl New Delhi virus. The mechanism by which AV2 protein contributes to regulation of replication is not understood. Since results obtained in in planta system indicated AV2 protein role in regulation of replication, its role in replication was studied.

AV2 is a small protein of 10 kDa/13 kDa size and 17.5 and 13.2% of the protein is composed of basic amino acid in Bg and Cp isolate, respectively. The estimated PI of the protein is 6.47 for Bg isolate and 9.61 for Cp isolate. Except for one motif for phosphorylation site in the Cp isolate no other characteristic motif could be deciphered. In the present study, we did not see any binding between the expressed protein and single or double stranded DNA. AV2/V2 protein in other viruses like ToLCNDV, TYLCV-Is and the analogous protein V1 of MSV too do not show any binding activity. Even in cases where V1 protein is attributed with movement function, it interacts through coat protein and facilitates docking of CP-DNA complex to periphery [6, 16].

The silencing suppression ability of AV2/V2 has been demonstrated at least for two begomoviruses, EACMV and TYLCV-Is [9, 10]. In both the viruses, the suppressing activity did not involve interference with siRNA accumulation. It is suggested that AV2 may intervene downstream of siRNA production, unlike TrAP or AC4. Karjee et al. [17] hypothesized that viral suppressors which suppress RNA silencing should enhance replication and demonstrated it for AC2 suppressor of MYMIV. The link between suppressors mediated enhanced replication and RNAi mediated host defence may be one key event in viral pathogenicity which needs to be investigated.

Modulation of nicking and ATPase activity of Rep by AV2 protein was interesting. AV2-Rep interaction at very low concentration of 1–4 ng enhanced the nicking activity at least by 20 folds. With the increased concentration of AV2 protein, further enhancement was not seen indicating how site involved in nicking could be saturated even at low concentration. ATPase activity is only two folds enhanced in the presence of AV2 protein and not a dramatic increase as in the case of nicking activity. In the present study AV2 protein of both Bg and Cp isolates showed similar enhancement activity though Bg isolate lacks 39 amino acids at C-terminal region. It is postulated that N-terminal region up to 74 alone is perhaps enough to contribute to enhancement. The truncated and complete proteins differed only in their conditions required for optimum production. No such information is available on any protein that modulates the activities of Rep in other geminiviruses; their activity is readily comparable with any accessory protein. AV2 protein may function as a molecular chaperone to bring Rep and DNA together in a controlled fashion. The very low concentration at which it is efficient, also may explain its lack of detection inside nucleus in planta system.