Plant Cell Rep (2004) 23:81–90 DOI 10.1007/s00299-004-0792-0 GENETICS AND GENOMICS D. Tamilselvi · G. Anand · S. Swarup A geminivirus AYVV-derived shuttle vector for tobacco BY2 cells Received: 21 January 2004 / Revised: March 2004 / Accepted: March 2004 / Published online: April 2004  Springer-Verlag 2004 Abstract We have developed a plant-Escherichia coli pASV shuttle vector from the essential elements of the Ageratum yellow vein virus (AYVV). The geminivirus vector contains the AYVV genome with the coat-protein deletion, the E. coli vector backbone of pUC19, a unique cloning site and gene expression cassettes for plant selection and reporter gene activity. The replication of pASV vectors was compared in Nicotiana benthamiana and N. tabacum BY2 cells, and the latter were found to be suitable for long-term maintenance of the vectors in culture. The vector DNA was detected at regular intervals by PCR, b-glucuronidase expression analysis and plasmid rescue during a 4-month culture period. A novel methylation-based PCR assay was carried out to show de novo replication for pASV-derived vectors in 2-month-old tobacco BY2 cell lines. This is the first report of the extrachromosomal replication of monopartite begomovirus with stability and foreign gene expression in long-term cell cultures. Keywords Ageratum yellow vein virus · Geminivirus · Shuttle vector · Replication · Foreign gene expression Abbreviations ACMV: African cassava mosaic virus · AYVV: Ageratum yellow vein virus · CP: Coat protein · GUS: b-Glucuronidase · TGMV: Tomato golden mosaic virus · Tobacco BY2: Tobacco L. cv. Bright Yellow Communicated by P.P. Kumar D. Tamilselvi · G. Anand · S. Swarup ()) Department of Biological Sciences, National University of Singapore, Singapore, 117 543 e-mail: dbsss@nus.edu.sg Tel.: +65-6874-7933 Fax: +65-6779-2486 Present address: G. Anand, Temasek Capital (Private) Limited, Temasek Tower, Singapore, 068811 Introduction Whole plants and plant cells are emerging as viable and competitive expression systems for large-scale protein production as a means of obtaining biologically active and safe biopharmaceutical proteins at affordable prices. Consequently, there is renewed interest in developing novel vectors to express foreign proteins in these systems. While the most widely used method for producing foreign proteins is via stably transformed plants, plant cell cultures provide an alternative. The main advantages of the latter lie in the large-scale production of proteins in industrially sized bioreactors under sterile, defined, and controllable conditions, all of which makes the plant cell culture system amenable to standard biomanufacturing practices. As a result of recent commercial interest, there have been many advances in the areas of transgenic plants (Fischer and Emans 2000; Ma et al. 2003), production in suspension cultures (Fischer et al. 1999a) and the application of plant vectors for foreign protein expression (Fischer et al. 1999b). Vast choices of plant transformation vectors are currently available due to the early start researchers have made in this field. Compared to the stable transformation vectors based on T-DNA, however, there are only a few reports on extrachromosomal shuttle vectors for expressing foreign genes. In plants, single-stranded DNA geminiviruses have been used as potential sources of extrachromosomal vector replicons to enable multiplication in the nuclei of infected cells (reviewed by Davies and Stanley 1989; Mullineaux et al. 1992; Stanley 1993; Timmermans et al. 1994). Geminiviruses have small single-stranded circular DNA genomes of 2.5–3.0 kb. They are characterized by twinned (geminate) icosahedral capsids and can replicate using plant host machinery via double-stranded DNA intermediates. The geminiviridae family has three subgroups, which can be distinguished based on their genetic organization, plant host, and insect vector. The genera Mastrevitesus and Curtovirus have a single genomic component and infect monocot and dicot species, respectively. Begomovirus members, on the other hand, 82 exclusively infect dicot species and have a genome comprising two similar-sized DNA components (DNA A and DNA B) (reviewed by Gutierrez 2000). DNA A encodes a replication-associated protein (Rep), coat-protein (CP), and proteins that participate in the control of replication and gene expression. The DNA B component encodes proteins required for nuclear trafficking and cellto-cell movement of the viral DNA. Relatively few begomoviruses have been described to possess a monopartite genome, which resembles DNA A. Those include the tomato leaf curl virus (TLCV), Ageratum yellow vein virus (AYVV), and cotton leaf curl virus (CLCuV) (Dry et al. 1993; Tan et al. 1995; Briddon et al. 2000). AYVV can replicate and form infectious symptoms in Ageratum, tomato, French bean and Nicotiana benthamiana (Tan et al. 1995). Transmission of AYVV is by whiteflies (Tan et al. 1995), and its monopartite genome has two overlapping virion-sense open reading frames (ORFs), V1 and V2 that encode the CP and movement protein, respectively. ORF C1 encodes the replicationassociated protein, ORFs C2 and C3 regulate virion-sense gene expression and DNA replication, respectively, ORF C4 is a pathogenicity determinant that may affect the host cell, and there is an additional ORF C5 of unknown function. The intergenic region (IR) contains the initiation site of rolling circle DNA replication. Irrespective of the monopartite or bipartite nature of geminiviruses, only their intergenic and the complementary strand ORFs are necessary for replication (Lazarowitz et al. 1989; Kammann et al. 1991; Ugaki et al. 1991). Hence, a popular strategy used in developing geminiviral vector backbone is the replacement of the CP gene with reporter genes. Such vectors have been developed using monopartite viruses such as wheat dwarf virus (WDV), maize streak virus (MSV) and bean yellow dwarf virus (BeYDV) (Ward et al. 1988; Lazarowitz et al. 1989; Topfer et al. 1989; Mor et al. 2003) and DNA A genomes of bipartite virus TGMV (Kanevski et al. 1992). Various vectors have used both the native CP promoter and foreign promoters such as the cauliflower mosaic virus (CaMV) 35S promoter to express foreign genes or reporter genes (Ugaki et al. 1991; Mor et al. 2003) for transient expression. Stable maize and tobacco cell lines containing the replicating viral episome for MSV or TGMV, respectively, have also been reported (Kanevski et al. 1992; Palmer et al. 1999). WDV, a monopartite mastrevirus, replicon-based shuttle vectors with bacterial replicons, namely ColE1 and p15A, have been used (Kammann et al. 1991; Ugaki et al. 1991) for extrachromosomal replication study. However, no shuttle vectors have been developed and evaluated based on the monopartite begomoviruses. We report here the development and evaluation of a plant-Escherichia coli shuttle vector using the replicon of the monopartite begomovirus AYVV and a bacterial replicon derived from pUC19 to ensure a high copy number in E. coli. Two different methods, namely electroporation and biolistic bombardment, were evaluated for their efficiency to transform the plant cells. A novel method based on the unique DNA methylation-specificity of the plant and E. coli cells was used to study de novo replication of the shuttle vector in the plant cells. The successful rescue and maintenance of structural integrity of the shuttle vector from plant cells into E. coli and the expression of the reporter gene was demonstrated in 4month-old cultures of tobacco BY2 cell suspension cultures. Materials and methods Construction of plasmids Plasmid pASV82, a shuttle vector for Escherichia coli and tobacco cells, was constructed based on the AYVV geminiviral and E. coli pUC19 backbone. The genealogy of vector construction is detailed in Fig. 1. A 2.7-kb AYVV DNA fragment was released from pHN419 (Tan et al. 1995) following BamHI digestion. This fragment was self-ligated to obtain a circular AYVV DNA template, which was then used to amplify a 2.3-kb fragment using primers extending away from the CP gene, thereby creating its deletion. The CP sense primer (50 -AATTCGTACTCATGCCAG-TAATCCAGTGTATGC-30 ) and Mlu antisense primer (50 -AATTCATTACCACGCGTGACATCACTAACAC-30 ) were used with the proofreading Vent DNA polymerase. PCR cycling parameters were 95C for min, 50C for and 72C for min. The number of PCR cycles was kept to a maximum of 20 to further minimize proofreading errors. EcoRI-compatible ends were generated using T4 DNA polymerase and the fragment ligated to the unique EcoRI site of the vector pNKA210.2, a derivative of pIBT210.1 (Haq et al. 1995). pNKA210.2 has a pUC19 (2.68 kb) backbone and a plant gene expression cassette consisting of a 35S CaMV promoter, 50 UTR of the tobacco etch virus (TEV-50 UTR), a translational enhancer, a replaceable stuffer fragment, and a vspB terminator sequence. This shuttle vector was named pASV. A 1.68-kb neomycin phosphotransferase II (NPTII) expression cassette from pNGI (Klien et al. 1989) was ligated into the unique HindIII site of pASV to generate pASVNPT. To facilitate cloning of the foreign gene in this shuttle vector, we created a unique enzyme site. There were two HindIII sites in pASVNPT. A single HindIII site was generated in pASVNPT by destroying one of the two HindIII sites by partial digestion, followed by blunt-ending and self-ligation to yield pASV82. Foreign gene expression cassettes can be inserted in pASV82 (8.2 kb) at its unique HindIII site. The constructs were selected in E. coli using ampicillin (50 mg/ml) and in plant cells using kanamycin (50 mg/ml). Plasmid pASVGUS, a derivative of pASV82 (8.2 kb) containing the GUS expression cassette, was constructed by inserting the 3-kb HindIII fragment of pRTL2-GUS (Carrington et al. 1991) into the HindIII site of pASV82. A replication-defective control plasmid, pASVDIR, was constructed by partial digestion of pASVNPT with BamHI, followed by inverse PCR amplification using primers flanking the IR region with a mixture of KlenTaq (Fermentas, Hanover, Md.) and Pfu polymerase (Promega, Madison, Wis.). The primers IRD (50 -TACTCTCCTGATACGATTGGGC-30 ) and C1 (50 -AATTCCCAAAGTGCCATTCGG-30 ) were used for PCR with the following parameters: 25 cycles of 30 s at 94C, at 55C, and at 72C. The PCR product lacking the IR (pASVDIR) was first blunted and then self-ligated to circularize it. All plasmids were constructed and propagated in E. coli strain DH5a and key junctions of the fragments sequenced after each construction. Large-scale amplification and purification of plasmids was performed by the alkaline lysis method followed by CsCl ultracentrifugation (Sambrook et al. 1989) or purified with a Nucleobond DNA purification AX500 column (Clontech laboratories, Palo Alto, Calif.) and compared. 83 Fig. The Ageratum yellow vein virus (AYVV) genome and cloning strategy of the AYYVderived plant-Escherichia coli shuttle vector. In the AYVV genome, ORF V1 encodes coatprotein, V2 encodes movement protein, C1 encodes replicationassociated protein, C2 regulates virion-sense gene expression, C3 regulates DNA replication, C4 encodes pathogenicity determinant, C5 encodes an unknown function. Exp cas1 Expression cassette 1, IR intergenic region. The recombinant plasmids described in the Results are indicated in boxes. The unique HindIII in pASV82 can be used to clone expression cassettes with the gene of interest Optimization of electroporation conditions for N. benthamiana mesophyll-derived protoplasts Protoplast isolation, electroporation medium and culture conditions for stable transformation were as described by Sala et al. (1989). To determine the optimum electroporation conditions for introducing plasmid DNA into protoplasts, we used two electroporation devices, namely T820, a square wave pulse generator (BTX, San Diego, Calif.) and a Gene Pulser II exponential wave pulse generator (Bio-Rad, Hercules, Calif.). The optimization methods were based on Trypan blue uptake and fluorescein diacetate staining to determine viability according to Saunders et al. (1995) with minor modifications. A suspension of 1”106 protoplasts in 600 ml of HeNa/F buffer (10 mM HEPES, pH 7.1, mM CaCl2, 150 mM NaCl, 0.2 M mannitol) was used for electroporation in 4-mm gap cuvettes obtained from the manufacturers of the two electroporators. The electroporation of protoplasts with the BTX T820 was optimized with a single pulse of 80 ms and with varied levels of field strength under low and high voltage modes. The Bio-Rad Gene Pulser II was operated with the capacitance set at 1,000 mF, the resistance set at 100–200 W and the time constant at approximately 18–26 ms with varied field strength. Trypan blue uptake was observed under bright field microscopy, and the FDA staining was observed under UV fluorescence in a dark field with the inverted fluorescent Leitz Fluovert FU microscope (Leitz, Wetzlar, Germany) equipped with a UV lamp. 84 Transformation of N. benthamiana mesophyll-derived protoplasts with plasmid DNA and confirmation using PCR For transformation using our optimized conditions, we generally electroporated 50 mg of purified closed circular plasmid DNA into 1”106 protoplasts using the BTX electroporator. To confirm the transformation, we extracted the DNA according to Townsend et al. (1986) at various intervals during protoplast culture. Linear PCR for the NPTII gene and NAD5 gene was carried out for 15 cycles of at 96C, at 55C, and at 72C with primers NPTfor (50 -GAAGGCGATAGAAGGCGA-30 ) and NPTrev (50 GGGTGGAGAGGCTATTCGGC-30 ). To amplify the NAD5 gene, we used the PCR primers NAD5for (50 -TAGCCCGACCGTAGTGATGTTAA-30 ) and NAD5rev (50 -ATCACCGAACCTGCACTCAGGAA-30 ) with the following parameters: 15 cycles of 30 s at 96C, at 55C, at 72C. Transformation of tobacco BY2 by particle bombardment and PCR detection of transformed lines Cells of N. tabacum L. cv. BY2 were maintained in MS medium (Murashige and Skoog 1962) supplemented with 0.18 mg/l K2HPO4, 100 mg/l myoinositol, mg/l thiamine HCl, 0.5 mg/l MES, 30 g/l sucrose in the dark at 120 rpm and 25C. A biolistic particle gun (model PDS-1000/He; Bio-Rad) was used for biolistic bombardment of the tobacco BY2 cells as described by Kikkert (1993). Conditions were optimized for various rupture disks (900, 1,100, 1,300 psi), and 1-mm gold particles as microcarriers coated with pRTL2-GUS vector DNA were used. The BY2 cells were bombarded and assayed for transient GUS expression days after bombardment. Under the optimized conditions the biolistic bombardment was carried out for the tobacco BY2 cells using the shuttle plasmids, and the transformed cell lines were screened on selection medium containing 50 mg/ml kanamycin. Total DNA was extracted according to Dellaporta et al. (1983). Transformed tobacco cells were confirmed by PCR amplification of the AYVV region with the primers C1 and CP sense. PCR cycling parameters were 25 cycles of at 96C, at 57C, and at 72C. The same DNA extract used for the AYVV primers was also used for a tobacco chromosomal gene, SAMDC (S-adenosyl methionine decarboxylase). The SAMDC gene was amplified using the primers samdcfor (50 -CGGCTGCTCACATGACTGTTAGTTCTGGC-30 ) and samdcrev (50 -AACATGCAAGCACCTTCTCAACCAG-30 ) with the following PCR cycling parameters: 25 cycles of at 95C, 30 s at 50C, 30 s at 72C. Rescue of the pASVNPT and pASVGUS shuttle vectors from tobacco BY2 cells in E. coli Total DNA was isolated from tobacco BY2 transformed cells as described by Dellaporta et al. (1983). Five micrograms of total plant DNA was transformed into E. coli strain DH5a to rescue the shuttle vector. Transformed cells were selected on LB plates containing ampicillin (50 mg/ml). Plasmid DNA was isolated for further analysis using the alkaline lysis method. Analysis of replicating DNA by coupled restriction enzyme digestion-random amplification PCR (CREDRA-PCR) and Southern blot analysis CREDRA-PCR has been used previously to identify DNA methylation in plants (Cai et al. 1996; Prakash and Kumar 1997). We modified CREDRA-PCR to study plasmid replication in plant cells as follows. Total DNA (5 mg) from DH5a, DNA from the tobacco transformed cell line, and rescued DNA from DH5a were restricted with m6A methylation-sensitive (DpnI) and methylation-resistant (BclI) enzymes for h in a 20-ml reaction volume. The digested DNA was re-purified and amplified using AYVV primers (C1 and CP sense) under the conditions described earlier. Five-microliter aliquots of the PCR products were fractionated by agarose gel electrophoresis and analyzed by Southern blot hybridization according to manufacturer’s protocol (Boehringer Mannheim, Germany). The 1.7-kb AYVV gene probe was generated by PCR amplification and was labeled with digoxygenin-labeled dUTP by random priming. Histological GUS assays Putative transformed cell lines and control BY2 cells were tested for histochemical localization of GUS (Gallagher 1992). The cells were incubated overnight at 37C in an assay buffer consisting of 100 mM NaPO4 (pH 7.0), mM X-Gluc (5-bromo-4-chloro-3-indolyl glucuronide cyclohexylammonium salt), examined for fluorescence, and photographed under a bright field using an inverted Lietz Fluovert FU microscope (Leitz, Wetzlar, Germany). Results and discussion Construction of pASV plant-E. coli shuttle vectors We chose a monopartite AYVV clone from pHN419 to construct a pASV series of plant-E. coli shuttle vectors pASVNPT, pASV82, pASVGUS and pASVDIR (Fig. 1). The vector pASV has all of the essential elements of AYVV except the CP (V1) to allow AYVV viral replicon for extrachromosomal replication in plant cells. The pUC19 backbone enables the plasmids to replicate in E. coli under ampicillin selection for easy manipulation and recovery of the clones from plants in E. coli. The NPTII expression cassette [consisting of the CaMV 35S promoter and nopaline synthase (NOS) gene terminator] in pASVNPT allows kanamycin selection in plant cells. The pASVDIR plasmid, with the deletion of the intergenic region in pASVNPT, serves as a replication-defective control. An intermediate cloning vector, pASV82, was derived from pASVNPT and contains a unique HindIII site for subsequent cloning of the expression cassette of genes of interest. The GUS expression cassette was cloned at this HindIII site to yield pASVGUS for its expression in plant cells. The expression cassette has a dual CaMV 35S promoter, 50 UTR translation enhancer of TEV and a CaMV 35S terminator. An alternate expression cassette1 with the CaMV 35S promoter, 50 UTR of TEV, a stuffer fragment, and vspB terminator is also available in this vector. This stuffer fragment can be replaced with the gene of interest between NcoI and KpnI sites. However, the unique HindIII site is recommended for the insertion of foreign gene expression cassette since NcoI and KpnI are not unique sites in the vector. Optimization of electroporation condition Our objective was to study the replication of pASV-derived vectors in N. benthamiana and N. tabacum protoplasts. We first standardized the electroporation conditions for introducing the vectors to the isolated mesophyll protoplasts of N. benthamiana using two electroporators. A comparison of the effect of the two different pulse 85 determined that the optimum field strength should be 0.3 kV/cm for the exponential wave pulse generator and 0.5–0.6 kV/cm for the square wave pulse generator. The protoplasts showed decreased viability under a low voltage mode with BTX T820 (data not shown). In subsequent experiments we used the BTX electroporator with a pulse field strength of 0.55 kV/cm and a single pulse of 80 ms. Transformation of mesophyll-derived protoplasts of N. benthamiana with pASVNPT Fig. Electroporation efficiency of mesophyll protoplasts. Trypan-blue uptake and viability as determined by fluorescein diacetate staining of mesophyll protoplasts electroporated using: a BTX model T820 (square wave pulse generator), b Bio-Rad Gene pulser II (exponential pulse generator). Arrows indicate the optimum conditions used in subsequent experiments types revealed a general trend. Freshly isolated protoplasts had 50–60% viability, and in both types of pulse generators the viability dropped below 10% with higher pulse strength (Fig. 2). The number of Trypan bluestained protoplasts increased and viability rapidly decreased with increasing field strength with both electroporators. To maintain a protoplast viability of 50%, we Fig. 3a–c PCR detection of pASVNPT DNA in Nicotiana benthamiana protoplasts. a pASVNPT construct. IR Intergenic region. b Amplification products of the vector-borne NPTII gene. L Low-mass DNA ladder. c Amplification products of the chromosomal NAD5 gene. L 1-kb ladder (Fermentas). Lanes 1–6 PCR for days 1– in duplicate, C negative control Using the optimized electroporation conditions with highquality (circular form) input DNA (derived from Nucleobond column-purified kit), we electroporated mesophyll protoplasts with pASVNPT and followed the replication of vector DNA over a 6-day period. Microscopic observation showed no active protoplast division during this period. The NPTII PCR products of pASVNPT were detectable throughout the period tested, but their levels declined with time in the cells (Fig. 3b). As expected, there was no increase in the level of PCR products for the chromosomal gene NAD5 (Fig. 3c). The steady-state levels of the pASVNPT vector decreased slightly during the 6-day period; this could be due to either decreased replication or to the same level of replication ability but an increased turnover of vector DNA. In practical terms, either situation would lead to a lower copy number of vector molecules. These results show that the pASVNPT vector did not replicate efficiently in the N. benthamiana cells. Previous reports of transient replication experiments with protoplasts derived from mesophyll cells have shown that host cell division is a prerequisite for the replication of some types of begomovirus but not for all. For example, replication of the ACMV requires host cell division (Townsend et al.1986), while TGMV does not (Brough et al. 1992). In other geminiviruses also, such dependency on host cell division varies. In the Mastrevirus group, WDV replication is dependent on cell divi- 86 Fig. 4a–f Selection of transgenic tobacco BY2 calluses 30 days following bombardment with pASV vectors. a pASVNPT-transformed calluses in selection medium. b pASVGUS-transformed calluses in selection medium. c, d Control tobacco calluses in the absence of kanamycin. e Control tobacco calluses in selection medium. f pASVNPTDIR-transformed replication-defective control on selection medium sion of the host Triticum monococcum (Matzeit et al. 1991), while in another study with maize it was shown to proceed in the absence of cell division (Timmermans et al. 1992). Hence, the dependency of geminiviral replication on host cell division may be affected by both viral and host factors. Because there was a lack of replication of pASVNPT in the N. benthamiana system in our studies, we conclude that AYVV replication is dependent on host cell division and that a slowly or non-dividing host system such as N. benthamiana would not be suitable for pASV-derived vectors. Consequently, we selected the related N. tabacum BY2 cells, which are known for sustaining rapid division in long-term cultures, for further testing. Transformation of pASV-derived vectors in N. tabacum BY2 cells by particle bombardment In order to efficiently transform tobacco BY2 cells we optimized the particle bombardment conditions using pRTL2-GUS. In transient GUS expression studies with pRTL2-GUS, the highest number of uniformly blue-colored cells (4,423€225 cells) with 1,300 psi were observed days post-bombardment. This is in contrast to the low number of stained cells with 900 and 1,100 psi (271€11, 1,019€45 cells), respectively, and no stained cells observed in the control mock bombarded cells. Subsequent experiments were conducted under these optimized conditions. With the objective to study the replication of pASVderived vectors in tobacco BY2 cells, the pASVNPT, pASVGUS and a replication-defective pASVDIR DNA were bombarded into BY2 tobacco cells for clonal selection of transformed lines on antibiotic-containing medium. Two hundred putative transformed independent lines from each vector were further cultured on selection plates for an extended period of months. The calluses transformed with pASVNPT and pASVGUS appeared as tiny white clumps between 20 days and 30 days post- bombardment, thereby allowing clonal selection of transformed cell lines. Both non-transformed tobacco calluses and cells transformed with pASVDIR showed no callus proliferation on selection plates (Fig. 4e,f), while the growth of control non-transformed calluses on nonselection plates was normal. The absence of colonies with the negative control pASVDIR-transformed cells proved that the intergenic region is necessary for viral replication and that the selection process of transformed cells was efficient. This observation is in accordance with deletion analyses of other geminiviruses such as WDV (Ugaki et al. 1991) and MSV (Shen and Hohn 1994), as this region contains the initiation site for rolling circle DNA replication. The transformed calluses maintained for months kept their ability to stably replicate the pASV vectors. In other analyses cell suspension cultures of TGMV and MSV transformed tobacco and maize lines were maintained for months and year, respectively (Kanevski et al. 1992; Palmer et al. 1999). Because calluses require less frequent transfers than cell suspensions, the methods described here would allow easier maintenance in prolonged periods. Replication studies of pASV-derived shuttle vectors As the maintenance of long-term cultures provides only indirect evidence for the stable replication of vectors, replication was studied using four methods to obtain independent validation: (1) detecting the presence of the vector DNA by PCR in long-term cultures of transformed calluses, (2) rescuing pASVGUS from plant cells into E. coli, (3) CREDRA-PCR assays based on methylation differences of DNA replicated in plant and E. coli cells, and (4) assaying for foreign reporter genes in transformed tobacco BY2 cells. 87 Fig. 5a, b PCR detection of pASVGUS in tobacco BY2 transformed calluses. a pASVGUS construct. b Lanes: L 1-kb ladder (Fermentas), negative control lacking total DNA from calluses, PCR products of AYVV gene from pASVNPT, pASVGUStransformed callus DNA, 4, PCR products of SAMDC gene— internal positive control for chromosomal DNA from calluses PCR detection of pASVNPT and pASVGUS vectors in transformed calluses tobacco suspension culture, while DNA prepared from the untransformed control BY2 cells did not yield any E. coli transformants, as was expected. This further confirmed that pASVGUS DNA was propagated along with the chromosomal DNA in transformed plant cells. The structural integrity of the AYVV vector was studied by PCR and the sizing of the constructs was determined by restriction profiling. Plasmid DNA was prepared from the rescued colonies and screened for the presence of the AYVV and GUS regions, respectively (Fig. 6a,b). All of the rescued clones that were screened showed the 1.7-kb AYVV fragment and the 0.7-kb GUS gene fragment upon PCR-based amplification. Uncut plasmids were compared to detect any size differences, but no major differences in the sizes of the plasmids were found. Additional restriction analyses of the rescued clones were performed to further investigate the possibilities of any vector DNA rearrangements: no significant restriction fragment length polymorphisms were seen between the rescued and control vector DNA (Fig. 6c). Similar results were obtained from 1-, 2-, and 3-month-old cultures. Taken together, these results confirm that the pASV backbone vector can replicate in both plant and E. coli without undergoing any detectable size alterations or rearrangements. Previous studies on the rescue of geminiviral shuttle vectors were carried out on 6-day- and 7-day-old cultures with the graminaceous host containing mastrevirus, WDV (Ugaki et al. 1991; Kammann et al. 1991). We report here longterm maintenance of the structural integrity of begomovirus-based pASV vectors. Although, these results show no major rearrangements, minor ones would not have been detected using the methods described here. Also, the rearranged vectors could not have been rescued if the rearrangement had affected their ability to replicate in E. coli. The versatility of the vectors to replicate in other host plants needs to be tested with respect to broader applications. Randomly picked healthy calluses at the 6-week stage growing on selection plates were screened for the presence of the vector. A 1.7-kb AYVV DNA fragment from the pASVNPT and pASVGUS constructs was amplified from the DNA of the transformed calluses. A 0.3-kb SAMDC DNA fragment from the chromosomal gene was amplified as a control (Fig. 5b). Twenty transformed calluses that showed the presence of pASVGUS DNA were transferred to MS medium with kanamycin for vector selection in suspension culture. Shuttling ability and rescue of pASVGUS from plant cells to E. coli Suspensions of transformed cell lines, as described above, were used for studying shuttle replication of the constructs in plant and E. coli cells. The suspension cell cultures were maintained by subculturing in selection medium once a week. The morphology and growth rates were similar in all cell suspension cultures up to 2– months in comparison to the control BY2 cells without kanamycin. The control BY2 cells on kanamycin did not multiply in suspension cells. Subculturing was continued for a period of months. We hypothesized that presence of the 1.7-kb AYVV PCR fragment would not directly confirm the shuttling ability of the vector. Hence, we attempted to rescue the vector from plant cells into E. coli at monthly intervals. This also allowed us to study any major rearrangements in the vector as a consequence of replication in both the plant and E. coli cells. The rescue of pASVGUS from plant cells into E. coli was carried out on 1-, 2- and 3month-old suspension cultures maintained on selection medium. Between 10 and 20 E. coli colonies were obtained with mg of total DNA derived from cells of the 88 Fig. 6a–c Rescue of pASVGUS shuttle vectors from transformed BY2 plant cells in E. coli. PCR detection of pASVGUS plasmid DNA with: a AYVV primers, b GUS primers. Lanes: L 1-kb ladder (Fermentas), C negative control with non-transformed tobacco cell, PCR with control pASVGUS plasmid DNA, 3–8 PCR with rescued pASVGUS clones (RC 3–8) from E. coli clones. c Restriction profiles of pASVGUS rescued clones RC (3*) and RC (4*) from panels a and b. Lanes: 1, 4, Non-digested DNA controls, 2, 5, BamHI-digested, 3, 6, HindIII-digested. The experiment was repeated with 1-, 2-, and 3-month-old suspension cultures. Results from the 2-month-old suspension cultures are shown in this figure Fig. 7a–c Verification of pASVGUS shuttle vector replication in tobacco cells using CREDRA-PCR. a Experimental setup for studying extrachromosomal replication. b CREDRA-PCR products using AYVV primers for pASVGUS detection. Lanes: 1–3 Control input pASVGUS m6A methylated DNA, 4–6 total genomic DNA from pASVGUS-transformed tobacco cells, 7–9 rescued plasmid DNA in DH5a (U un-digested, BclI BclI-digested, DpnI DpnI-digested). c Southern blot hybridization of the CREDRA-PCR products from the gel shown in panel b with DIG-labeled 1.7-kb AYVV fragment Replication studies based on DNA methylation differences We reasoned that direct evidence for vector replication in plant cells would require either distinguishing the replicative intermediates of the vector or showing the presence of DNA replication features specific to plants. We chose the latter line of proof and based our experiments on the m5C methylation that occurs during the replication of DNA in plants but which is absent in E. coli. The converse argument that m6A methylation is found in E. coli but is absent in eukaryotic DNA was also used in these studies. Other researchers have proposed that the extrachromosomal replicons may not be acces- sible for the host methylases, although the barrier preventing access of the methylases was not defined (Brough et al. 1992; Doerfler 1993). Based on these considerations, we studied the differences in the methylation status of the input DNA obtained from E. coli and that of the DNA obtained from newly transformed plant cell lines using a combination of methylation-sensitive restriction enzymes followed by PCR. To study the replication of extrachromosomal DNA of the pASVGUS shuttle vector we used a modified methylation-based PCR method termed CREDRA-PCR (Fig. 7a). This method was carried out on total genomic DNA isolated on the third day of subculture from 89 2-month-old cell suspension cultures. To discriminate between the input and the newly replicated DNA, both the DNA from suspension tobacco cells and the DNA rescued from E. coli were digested with either dam-methylationdependent restriction enzyme DpnI or BclI followed by PCR amplification and Southern analyses. DpnI is dependent on the methylation of adenine in the recognition sequence GA6/TC and, consequently, it cleaved the both the input and rescued DNA that was methylated in E. coli. BclI did not cleave the m6A methylated sequence (Fig. 7b,c). In contrast, the replicative DNA in tobacco cells was resistant to digestion with DpnI since they lacked the m6A site, consequently giving a PCR product with the AYVV primers. BclI did cleave the de novo synthesized DNA that lacked m6A methylation in the tobacco cells. Hence, vector DNA was not amplified and there was no corresponding signal in either the agarose gel or its Southern blot probed with vector DNA (Fig. 7c) that showed a higher sensitivity of detection. Based on these observations, we conclude that de novo replication of the pASVGUS vector occurred in plant cells in 2month-old suspension cells. The mastrevirus replicating shuttle vector in protoplasts of Triticum monococcum (Kammann et al. 1991), maize endosperm-derived protoplasts for WDV (Timmermans et al. 1992), and tobacco NT1 cells for BeYDV (Mor et al. 2003) have also been analyzed by methylation sensitivity. In these reports, methylation-based studies of replication were carried out during the first week following electroporation. In order to study de novo replication in long-term plant cell cultures, we adopted a methylation-based PCR assay—the CREDRA-PCR method—which is more sensitive, requires little starting material, and is simpler to perform than the previous methods. Expression of foreign reporter gene Once the replication ability of pASVGUS was established, we further tested its application as an extrachromosomal expression vector using the GUS reporter gene. Transgenic cell lines and control BY2 cells were tested for histochemical localization of GUS over an extended period of time spanning months. GUS assays were performed daily for days beginning with the first day of subculturing on 2-week- and 1-, 2-, 3,- and 4-month-old suspension cultures. In all cases GUS expression was detected in 0.001–0.006% of the cells screened. GUS expression was maximal on third day, and expression was not found when the cells had reached the seventh day, a time when the tobacco BY2 cells are likely to be in the stationary stage (Nagata et al.1992). The reason for maximal expression on third day is not yet clear. The highest percentage of GUS-expressing cells occurred in the first month, with the percentage declining gradually later in the fourth month. However, we speculate that this could be due to changes in the copy number of the vector during the different growth phases of the culture. This variation was also noticed with TGMV (Kanevski et al. 1992) and BeYDV in tobacco cells (Mor et al. 2003). The difference in copy number could arise from interference of geminiviral replication with plant cell-cycle machinery or other host cell pathways (Gutierrez 2000). Conclusion We have shown here the utility of pASV-derived shuttle vectors for long-term stable maintenance of constructs and expression of foreign genes in cultured plant cells. Using methylation-based PCR assays we have also shown de novo replication of pASV vectors in long-term cultures. On the basis of the system we have described here pASV vectors will be suitable for use when researchers are looking to express foreign proteins in a closed sterile system rather than in whole transgenic plants. They will be useful in long-term cell cultures by improving the replication and expression levels through extensively characterizing the host factors in synchronized cell cultures and promoters highly active in early cell division of BY2 cells. Acknowledgements The authors would like to thank Dr. Wong Sek Man for providing pHN419 containing the full-length AYVV coding sequence, Prof. Charles Arntzen and Dr. Hugh S. Mason for their gift of the pIBT210.1 plasmid, and Dr. Jaideep Mathur for providing the tobacco BY2 suspension cells. TS and GA have been supported by NUS research scholarships. References Briddon RW, Mansoor S, Bedford ID, Pinner MS, Markham PG (2000) Clones of cotton leaf curl geminivirus induce symptoms atypical of cotton leaf curl disease. Virus Genes 20:19–26 Brough CL, Gardiner WE, Inamdar NM, Zhang XY, Ehrlich M, Bisaro DM (1992) DNA methylation inhibits propagation of tomato golden mosaic virus DNA in transfected protoplasts. Plant Mol Biol 18:703–712 Cai Q, Guy CL, Moore GA (1996) Detection of cytosine methylation and mapping of a gene influencing cytosine methylation in the genome of citrus. Genome 39:235–242 Carrington JC, Freed DD, Leinicke AJ (1991) Bipartite signal sequence mediates nuclear translocation of the plant potyviral NIa protein. Plant Cell 3:953–962 Davies JW, Stanley J (1989) Geminivirus genes and vectors. Trends Genet 5:77–81 Dellaporta SL, Wood J, Hicks JB (1983) Maize DNA miniprep. In: Malmberg R, Messing J, Sussex I (eds) Molecular biology of plants: a laboratory course manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor Doerfler W (1993) Pattern of de novo methylation and promoter inhibition: studies on the adenovirus and the human genome. In: Jost JP, Saluz HP (eds) DNA methylation: molecular biology and biological significance. Birkhauser Vedlag, Basel, pp 262–299 Dry IB, Rigden JE, Krake LR, Mullineaux PM, Rezaian MA (1993) Nucleotide sequence and genome organization of tomato leaf curl geminivirus. J Gen Virol 74:147–151 Fischer R, Emans N (2000) Molecular farming of pharmaceutical proteins. Transgenic Res 9:279–299 Fischer R, Emans N, Schuster F, Hellwig S, Drossard J (1999a) Towards molecular farming in the future: using plant cellsuspension cultures as bioreactors. Biotechnol Appl Biochem 30:109–112 90 Fischer R, Vaquero CM, Sack M, Drossard J, Emans N, Commandeur U (1999b) Towards molecular farming in the future: transient protein expression in plants. Biotechnol Appl Biochem 30:113–116 Gallagher SR (1992) GUS Protocols: using the GUS gene as a reporter of gene expression. Academic, San Diego, Calif. Gutierrez C (2000) DNA replication and cell cycle in plants: learning from geminiviruses. EMBO J 19:792–799 Haq TA, Mason HS, Clements JD, Arntzen CJ (1995) Oral immunization with a recombinant bacterial antigen produced in transgenic plants. Science 268:714–716 Kammann M, Schalk HJ, Mazeit V, Schaefer S, Schell J, Gronenborn B (1991) DNA replication of wheat dwarf virus, a geminivirus, requires two cis-acting signals. Virology 184:786–790 Kanevski IF, Thakur S, Cosowsky L, Sunter G, Brough C, Bisaro D, Maliga P (1992) Tobacco lines with high copy number of replicating recombinant geminivirus vectors after biolistic DNA delivery. Plant J 2:457–463 Kikkert JR (1993) The biolistic PDS-1000/He device. Plant Cell Tissue Organ Cult 33:221–226 Klien TM, Kornstein L, Stanford JC, Fromm ME (1989) Genetic transformation of maize cells by particle bombardment. Plant Physiol 91:440–444 Lazarowitz SG, Pinder AJ, Damsteegt VD, Rogers SG (1989) Maize streak virus genes essential for systemic spread and symptom development. EMBO J 8:1023–1032 Ma JKC, Drake PMW, Christou P (2003) The production of recombinant pharmaceutical proteins in plants. Nat Rev Genet 4:794–805 Matzeit V, Schaefer S, Kammann M, Schalk HJ, Schell J, Gronenborn B (1991) Wheat dwarf virus vectors replicate and express foreign genes in cells of monocotyledonous plants. Plant Cell 3:247–258 Mor TS, Moon YS, Palmer KE, Mason HS (2003) Geminivirus vectors for high level expression of foreign proteins in plant cells. Biotechnol Bioeng 81:430–437 Mullineaux PM, Davies JW, Woolston CJ (1992) Geminiviruses as gene vectors. In: Michael T, Wilson A, Davies JW (eds) Genetic engineering with plant viruses. CRC, Boca Raton, pp 187–215 Murashige T, Skoog F (1962) A revised medium for rapid growth and bioassays with tobacco tissue cultures. Physiol Plant 15:473–497 Nagata T, Nemoto Y, Hasezawa S (1992) Tobacco BY-2 cell line as the ‘HeLa’ cells in the cell biology of higher plants. Int Rev Cytol 132:1–30 Palmer KE, Thomson A, Rybicki EP (1999) Generation of maize cell lines containing autonomously replicating maize streak virus-based gene vectors. Arch Virol 144:1345–1360 Prakash AP, Kumar PP (1997) Inhibition of shoot induction by 5-azacytidine and 5-aza-20 deoxycitidine in Petunia involves DNA hypomethylation. Plant Cell Rep 16:719–724 Sala F, Marchesi ML, Castiglione S, Paszkowski J, Saul M, Potrykus I, Negrutiu I (1989) Direct gene transfer in protoplasts of Nicotiana plumbaginifolia. In: Bajaj YPS (ed) Biotechnology in agriculture and forestry, vol 9. Plant protoplasts and genetic engineering II. Springer, Berlin Heidelberg New York, pp 217–227 Sambrook J, Fritsch EF, Maniatis T (1989) Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor Saunders JA, Lin CH, Hou BH, Cheng JP, Tsengwa N, Lin JJ, Smith CR, McIntosh MS, Wert SV (1995) Rapid optimization of electroporation conditions for plant cells, protoplasts and pollen. Mol Biotechnol 3:181–189 Shen WH, Hohn B (1994) Amplification of the b-glucuronidase gene in maize plants by vector based on maize streak virus. Plant J 5:227–236 Stanley J (1993) Geminiviruses: plant viral vectors. Curr Biol 3:91– 96 Tan HN Priscilla, Wong SM, Wu M, Bedford ID, Saunders K, Stanley J (1995) Genome organization of ageratum yellow vein virus, a monopartite whitefly-transmitted geminivirus isolated from a common weed. J Gen Virol 76:2915–2922 Timmermans MCP, PremDas O, Messing J (1992) Trans replication and high copy numbers of wheat dwarf virus vectors in maize cells. Nucleic Acids Res 20:4047–4045 Timmermans MCP, PremDas O, Messing J (1994) Geminiviruses and their uses as extrachromosomal replicons. Annu Rev Plant Physiol Plant Mol Biol 45:79–112 Topfer R, Gronenborn B, Schell J, Steinbiss HH (1989) Uptake and transient expression of chimeric genes in seed-derived embryos. Plant Cell 1:133–139 Townsend R, Watts J, Stanley J (1986) Synthesis of viral DNA forms in Nicotiana plumbaginifolia protoplasts inoculated with cassava latent virus (CLV); evidence for the independent replication of one component of the CLV genome. Nucleic Acids Res 14:1253–1265 Ugaki M, Ueda T, Timmermans MCP, Vieira J, Elliston KO, Messing J (1991) Replication of a geminivirus derived shuttle vector in maize endosperm cells. Nucleic Acids Res 19:371– 377 Ward A, Etessami P, Stanley J (1988) Expression of a bacterial gene in plants mediated by infectious geminivirus DNA. EMBO J 7:1583–1587 . obtaining biologically active and safe biopharmaceutical proteins at affordable prices. Consequently, there is renewed interest in developing novel vectors to express foreign proteins in these systems. While. producing foreign proteins is via stably transformed plants, plant cell cul- tures provide an alternative. The main advantages of the latter lie in the large-scale production of proteins in in- dustrially. and in plant cells using kanamycin (50 mg/ml). Plasmid pASVGUS, a derivative of pASV82 (8.2 kb) containing the GUS expression cassette, was constructed by inserting the 3-kb HindIII fragment of