Research Paper On Ti Plasmid And Topo

Citation: Zhang Y, Lee C-W, Wehner N, Imdahl F, Svetlana V, Weiste C, et al. (2015) Regulation of Oncogene Expression in T-DNA-Transformed Host Plant Cells. PLoS Pathog 11(1): e1004620.

Editor: Darrell Desveaux, University of Toronto, CANADA

Received: July 31, 2014; Accepted: December 10, 2014; Published: January 23, 2015

Copyright: © 2015 Zhang et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited

Data Availability: All relevant data are within the paper and its Supporting Information files.

Funding: This study was financially supported by the Deutsche Forschungsgemeinschaft, GRK1342 (TP A7), SFB 567 (TP B5) and the China Scholarship Council (CSC) [to YZ]. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.


Agrobacterium tumefaciens is a pathogenic bacterium that infects several plant species. A region in the tumor inducing (Ti) plasmid, the transfer DNA (T-DNA), is integrated into the plant genome causing crown gall disease [1]. There are essentially two groups of genes encoded on the T-DNA of virulent A. tumefaciens strains [2]. The first is responsible for producing opines, so providing a carbon and nitrogen source for A. tumefaciens, with the second group expressing the oncogenes required for crown gall development. These oncogenes include IaaH, IaaM, Ipt, gene 6b and gene 5. It is assumed that although gene 6b and gene 5 are expendable, IaaH, IaaM and Ipt are crucial for crown gall development [3–5]. IaaH and IaaM code for enzymes that catalyze biosynthesis of auxin and Ipt mediates cytokinin biosynthesis [5,6]. IaaM encodes a tryptophan monooxygenase that converts tryptophan (Trp) into indole-3-acetamide (IAM), and IaaH an indole-3-acetamide hydrolase, converts IAM into indole-3-acetic acid (IAA) [7–9]. Ipt (isopentenyl transferase) catalyzes the rate-limiting step in cytokinin biosynthesis [2,5,10]. Cytokinins can also be synthesized in A. tumefaciens cells by the chromosomal encoded miaA enzyme [11,12] and the trans-zeatin synthesizing (tzs) enzyme encoded on the nopaline-type pTi-plasmid [13–15]. A. tumefaciens secretes auxin and cytokinin from the cells to initiate crown gall development [16] and pretreatment of plant tissues with auxin and cytokinin promotes A. tumefaciens-mediated transformation efficiency [14,17,18]. Very recently it was shown that cytokinins secreted by A. tumefaciens repress a Myb transcription factor in host plant cells, resulting in an enhanced transformation efficiency [18].

The increased production of auxin and cytokinin in T-DNA transformed plant cells expressing the IaaH, IaaM and Ipt oncogenes induces cell proliferation and differentiation [19,20]. Therefore, a T-DNA harboring plant cell needs to initiate transcription of the three oncogenes in order to express their function. In eukaryotic cells, the RNA polymerase II complex mediates transcription of mRNAs from protein-coding genes. This complex recognizes the TATA box and the transcription start site (TSS) [21] within upstream promoter regions that drive the expression of the downstream coding sequence (CDS). These two sequence features build the core promoter and this is sufficient to transcribe a gene [21]. TATA boxes were predicted to be present 5’ upstream of the CDS of the IaaH, IaaM and Ipt oncogenes [22–24]. In addition to initiation of transcription by the RNA polymerase II complex, expression of eukaryotic genes is usually regulated by transcription factors. These bind to regulatory sequence elements localized in the promoter regions of many eukaryotic genes and are oriented in a sense or anti-sense direction distant from the TSS [21]. For the Ipt gene of the octopine Ti plasmid pTiAch5, a 184 bp fragment upstream of the CDS is sufficient for transcription in plant cells [25]. In particular, the region between −185 and −139 bp from the translational start codon are essential [26]. Within that region, the 30 bp sequence cyt-1 binds an as yet unknown protein from tobacco nuclear protein extracts, designated CBF (cyt-1 binding factor) [27]. This suggests that expression of the agrobacterial oncogenes can be regulated by host transcription factors that await discovery.

A well-known response of plants to microbial pathogens is the microbe associated molecular pattern (MAMP)-induced innate immunity response, which includes expression of several WRKY transcription factors [28]. The expression profiles of 72 WRKY genes in Arabidopsis revealed that 49 genes are responsive to salicylic acid (SA) and pathogen treatment [29]. The WRKY transcription factor binding elements, the W-boxes (TGAC), are present in many defense related gene promoters [28]. In addition to the induction of pathogen defense responses, crown gall development requires cell proliferation and differentiation, such as vascularization [30]. These developmental programs are synergistically controlled by auxin and cytokinin signaling pathways that lead to changes in the regulation of gene expression. The expression of some auxin responsive factor (ARF) genes is induced by auxin, particularly in developing embryos and vascular tissues [31]. ARFs are known to induce the transcription of genes in an auxin-dependent manner by binding to auxin response elements (AuxREs) in auxin responsive promoters [31,32]. The regulation of ARF function involves auxin/indole acetic acid (Aux/IAA) proteins and TIR1 (transport inhibitor response 1) [33,34]. Aux/IAA proteins interact and repress the transcriptional activity of ARFs [35,36]. The F-box auxin receptor TIR1 is part of the SCFTIR ubiquitin ligase complex [37,38]. At increasing auxin concentrations, Aux/IAA proteins are recognized and ubiquitinylated by the SCFTIR complex and subsequently degraded by the 26S proteasome [39,40]. The de-repressed ARF proteins can activate target promoters.

This study focuses on the transcriptional regulation of the A. tumefaciens genes IaaH, IaaM and Ipt in the host plant. The intergenic regions between the CDSs of IaaH, IaaM and Ipt of the virulent T-DNA of A. tumefaciens strain C58 (pTiC58, AE007871) showed promoter activity in Arabidopsis cells. The IaaH and IaaM genes involved in auxin biosynthesis in T-DNA transformed cells, were ubiquitously expressed at low levels. In contrast, the Ipt promoter was activated by the transcription factor WRKY40 (AT1G80840), a transcription factor that responded rapidly to A. tumefaciens infection. WRKY40 together with ARF5 (AT1G19850), which is part of an auxin-dependent signaling pathway, boosted Ipt promoter activity in an auxin dependent manner. This enhanced activity correlated with cis-regulatory elements such as W-boxes and AuxREs in the Ipt promoter and the protein interaction of WRKY40 with ARF5. Our findings suggest that A. tumefaciens recruits the WRKY-dependent pathogen defense pathway to activate Ipt gene expression. This can be substantially increased when the auxin-dependent developmental process mediated by ARF5 is switched on.


The intergenic regions between the oncogenes function as promoters in plant cells

To discover how the expression of the agrobacterial oncogenes IaaH, IaaM and Ipt is regulated in plant cells, we analyzed the structure of the T-DNA region of the nopaline-type Ti plasmid pTiC58. The CDS of the three oncogenes are sequentially arranged and interrupted by two non-coding intergenic regions (IGR1 and IGR2; Fig. 1A). The IaaM and Ipt genes are transcribed from the sense strand and the IaaH gene is encoded on the opposite strand. If IGR1 functions as a promoter for both the IaaH and IaaM oncogenes, it must be a bidirectional promoter: one direction being 5’ upstream of the TSS of the IaaH CDS (IGR1a) and the other, 5’ upstream of IaaM (IGR1b).

Figure 1. IGR1 and IGR2 function as promoters in Arabidopsis cells.

(A) Arrangement of the coding sequences of the IaaH, IaaM and Ipt oncogenes and the intergenic regions (IGRs) in the T-DNA region of the Ti plasmid of A. tumefaciens strain C58, pTiC58. (B) Arabidopsis crown gall callus cells expressing the green fluorescing protein (GFP) under the control of IGR1a (IGR1a::GFP), IGR1b (IGR1b::GFP) and IGR2 (IGR2::GFP). IGR1 was used in two orientations; one is upstream of IaaH CDS (IGR1a) and the other upstream of IaaM (IGR1b). The universal cauliflower mosaic virus promoter was used as a positive control (2× CaMV35S::GFP) and the GFP CDS without promoter, as the negative control (GFP). Images show crown gall callus cells in the transmission microscopy (top row) and the UV light mode (bottom row, excitation: 490 nm, emission: 510 nm). The UV-light intensity used for excitation is the same for both pictures. Bars, 50 μm.

To prove whether the IGRs function as promoters in plant cells, the complete IGR sequences were fused with the CDS of the green fluorescent protein (GFP) in a binary vector. The IGR1a and IGR1b sequences included the 5’ untranslated regions (5’ UTR) of both the IaaH and IaaM genes, whereas IRG2 contained the 3’ UTR of IaaM and 5’ UTR of the Ipt gene. We generated stable transformed Arabidopsis crown gall tumor cell lines by infecting Arabidopsis root segments with the virulent A. tumefaciens strain C58, which, in addition to their pTiC58, harbor a binary vector with the IGR::GFP constructs. Detection of GFP fluorescence in the IGR1a::GFP, IGR1b::GFP and IGR2::GFP crown gall cell lines demonstrated that the IGRs drive GFP expression, so function as promoters in plant cells (Fig. 1B). Furthermore, as the IGR1 sequence is a bidirectional promoter, it can drive transcription of both the IaaH and IaaM genes.

Since IGR1a, IGR1b and IGR2 all function as promoters in eukaryotic cells, their sequences should contain the core promoter elements, such as the initiator (Inr) sequence and TATA box. To localize these in the promoters, we determined the TSSs of the IaaH, IaaM and Ipt genes using the 5’ rapid amplification of cDNA ends (5’ RACE) assay, finding that the translational start codon of the IaaH, IaaM and Ipt CDSs are at positions +12 bp, +26 bp and +44 bp in respect to the TSS (Table 1). Upstream of the TSSs, the typical eukaryotic Inr box (YYANWYY, TSS is underlined, Y = C/T, W = A/T, N = A/G/C/T) was present in the three promoter sequences. This is in agreement with the plant specific “YR Rule” (YR, TSS is underlined, Y = C/T, R = A/G [41,42]). The TATA boxes, the binding sites for the general transcription factor complex, are found in the promoter regions −25 bp to −35 bp and another feature of many eukaryotic promoters, the CAAT boxes, are localized approximately −70 bp upstream of the TSSs within the oncogene promoter regions (Table 1).

To ascertain whether the regulatory promoter elements of pTiC58 are conserved, we performed a sequence alignment with the promoter and 5’ untranslated regions (5’ UTRs) of the three oncogenes from different Ti plasmids. We compared the upstream sequences of the three oncogene CDSs of the Ti plasmids from two nopaline-types (pTiC58, pTiSAKURA), three octopine-types (pTiA6NC, pTiAch5, pTi15955) and one agropine-type (pTiBo542). The alignment shows that the TSSs (arrows), TATA boxes and CAAT boxes of the promoters for IaaH (S1 Fig.), IaaM (S2 Fig.) and Ipt (S3 Fig.) are conserved between the pTi plasmids of the different A. tumefaciens strains. In contrast, two TATA boxes are present 5’ upstream of the CDS in the Ipt genes from the octopine Ti plasmids (S3 Fig.). In the Ipt promoter of pTiC58, two CAAT boxes were predicted (S3 Fig.), one of which (GGTAAAGCC, from −72 to −64 bp) is conserved and also found in other nopaline type and in the octopine type pTi plasmids, but not in the agropine type Ipt promoter where no CAAT box was predicted. The second CAAT box (AAGGAATCT, −49 to −41 bp) is specific for the Ipt promoters of the nopaline type Ti-plasmids (S3 Fig.). Cis-regulatory binding elements for transcription factors were also determined in the IaaH, IaaM and Ipt promoters on the Watson and Crick strand using PLACE ( [43–45]. Several binding elements for different transcription factor families including MYB, DOF, WRKY, bHLH, ARR1 and ARF, were localized within the Ipt promoter (Table 1). In the IaaH and IaaM promoters, the binding element for the ARR1 (AT3G16857) transcription factor was dominant and there were eight ARR1 elements altogether.

To identify potential transcription factors that may be involved in enhancing the expression of the oncogenes, we analyzed existing microarray data of Arabidopsis crown galls [20,46], based on the Arabidopsis transcription factors listed in the Plant Transcription Factor Database v3.0 [47] ( A total of 151 transcription factor genes were found to be differentially transcribed in inflorescence stems inoculated with the virulent A. tumefaciens strain C58 compared to non-inoculated stems (S1 Table; fold change ≥ 2 or ≤ 0.5, p value < 0.01). As early as three hours post inoculation (hpi), three of these genes were up-regulated: WRKY53 (AT4G23810, 2.47 fold), WRKY40 (2.22 fold), and NAC102 (AT5G63790, 2.18 fold). WRKY53 was also up-regulated by the disarmed A. tumefaciens strain GV3101 (2.37 fold) 3 hpi. Six days post inoculation (dpi), the expression of six transcription factor genes was up- or down-regulated (S1 Table). In Arabidopsis crown gall material of A. tumefaciens strain C58, 141 transcription factor genes were transcriptionally changed compared to reference tissue 35 days post wounding (dpw). Amongst these, 74 genes were up-regulated, with 67 down-regulated (S1 Table) and all belong to various families including WRKYs, MYBs, DOFs, and NACs. The DNA binding elements and the microarray data both suggest that the MYB, DOF, WRKY, bHLH, ARR1 and ARF transcription factors are potential candidates for involvement in the regulation of Ipt expression, while ARR1 could regulate transcription of the IaaH or IaaM genes. The core promoter sequence elements could contribute to the basal expression of the three oncogenes in plant cells, whereas the binding sites for transcription factors might function in enhancing their transcription.

WRKY18, WRKY40, WRKY60 and ARF5 activate the Ipt oncogene promoter

To begin to study the regulation of onocgene expression, we first used quantitative real-time PCR (qRT-PCR). We assessed the relative transcript numbers of IaaH, IaaM and Ipt genes in 25-day-old Arabidopsis thaliana crown galls induced by the virulent A. tumefaciens strain C58, finding that the transcript levels of IaaH and IaaM were much lower compared to those of the Ipt gene in the crown galls (Fig. 2A). The high-throughput protoplast transactivation (PTA) system was then used [48] to identify transcription factors that could activate the three oncogene promoters in plant cells. To do so, the complete promoters of IaaH (IGR1a, 337 bp), IaaM (IGR1b, 337 bp) and Ipt (IGR2, 697 bp) of the pTiC58-encoded oncogenes (Fig. 1A) were fused with the CDS of the firefly luciferase (LUC) reporter gene. The plasmids containing the oncogene promoter-LUC constructs were transfected into Arabidopsis mesophyll protoplasts, either alone, or together with a second plasmid containing the CDS of a transcription factor fused to the constitutive cauliflower mosaic virus (CaMV35S) promoter. The relative luminescence, a measure for the oncogene promoter activity since it drives luciferase gene expression, was then determined. Mesophyll protoplasts transfected only with the oncogene promoter-LUC constructs showed the same pattern of promoter activity as that determined for the relative transcript numbers in crown galls (Fig. 2B). The Ipt promoter induced a higher relative luminescence than the IaaH and IaaM promoters.

Figure 2. Transcripts of oncogenes in crown galls and activity of oncogene promoters in protoplasts.

(A) Relative abundance of IaaH, IaaM and Ipt transcripts in crown gall tumors 25 days after inoculation of A. tumefaciens strain C58 into Arabidopsis inflorescence stems. Relative transcript numbers were quantified by qRT-PCR and normalized to 10,000 molecules of ACTIN2/8. Bars show mean values (±SD) of three independent samples. (B) Relative luciferase activity (firefly LUC/renilla LUC) driven by oncogene promoters (IaaH pro, IaaM pro and Ipt pro). Relative luciferase activity is calculated by firefly luminescence/renilla luminescence. Bars show mean values (±SD) of three independent experiments.

Next, a library containing the CDS of more than 400 transcription factors was screened. Among the included family members, WRKY, AP2/ERF, bHLH, bZIP, DOF, MYB and NAC, only WRKY18 (AT4G31800), WRKY40, WRKY60 (AT2G25000) and ARF5 were found to specifically activate the Ipt promoter in protoplasts (Fig. 3A). Protoplasts co-transfected with the WRKY or ARF effector and the Ipt-promoter-LUC reporter constructs exhibited a significantly higher promoter activity (reflected by luciferase activity) compared to the control samples that only harbored the reporter. Despite several attempts, no transcription factor was found to activate the IaaH and IaaM promoters. Comparison of the three WRKYs alone and in combination both showed that WRKY40 exerts the strongest impact on Ipt promoter-driven luciferase expression (S4 Fig.). Even all three WRKYs together did not increase the relative luminescence more than WRKY40 alone. This observation points towards a dominant role for WRKY40 in Ipt promoter regulation.

Figure 3. Activation of the Ipt promoter and gene expression of WRKY18, WRKY40, WRKY60 and ARF5.

(A) Fold induction of Ipt promoter-driven luminescence (Ipt pro) by WRKY18, WRKY40, WRKY60 and ARF5 in Arabidopsis mesophyll protoplasts transfected with two plasmid types. One harbors the Ipt promoter upstream of the firefly luciferase coding sequence (CDS) and the other, the universal cauliflower mosaic virus promoter (CaMV35S) upstream of a transcription factor CDS. The relative luminescence induced by the Ipt promoter in the absence of a transcription factor expression plasmid was set to 1. Bars show mean values (±SD) of three independent experiments. (B) Relative transcript numbers of WRKY18, WRKY40, WRKY60 and ARF5 genes in crown galls 25 days after inoculation with the virulent A. tumefaciens strain C58 (C58 Crown gall) and the disarmed strain (GV3101 Stems). (C) Time-dependent expression of the WRKY18, WRKY40, WRKY60 and ARF5 genes upon infiltration of five-week-old Arabidopsis leaves with suspension (OD600 1.0) of strain C58 and GV3101 as well as an Agromix buffer as control. Relative transcript numbers were quantified by qRT-PCR and normalized to 10,000 molecules of ACTIN2/8. Bars show mean values (±SD) of three independent samples. ** P<0.01 *** P<0.001 (Student’s t-test).

The transcript numbers of WRKY18, WRKY40, WRKY60 and ARF5 genes in crown gall tissues of A. tumefaciens strain C58 were determined using qRT-PCR. In agreement with the published microarray data [20,46], the transcript levels were clearly elevated in crown gall tumors compared to inflorescence stems inoculated with the disarmed A. tumefaciens strain GV3101 (Fig. 3B). It is already known that WRKY18, WRKY40 and WRKY60 are induced early after bacterial and fungal pathogen infection [49,50]. To analyze the impact of A. tumefaciens on gene induction, we analyzed the time-dependent expression of the three WRKY genes in Arabidopsis thaliana (Col-0) leaf tissues infiltrated with either the virulent A. tumefaciens strain C58, the disarmed strain GV3101 or buffer as a control. The qRT-PCR results demonstrated that the three WRKY genes responded to a certain degree to the infiltrated buffer solution at all analyzed time points (2 hpi to 72 hpi), indicating that they respond to wounding (Fig. 3C). The transcript levels of WRKY18 began to increase significantly at 8 hpi after infiltration by strain GV3101. The WRKY40 and WRKY60 genes were significantly induced by both A. tumefaciens strains as early as 2 and 4 hpi, respectively (Fig. 3C). In contrast, transcription of the ARF5 gene was still very low after 72 hpi, suggesting that this gene is not responsive to A. tumefaciens or wounding at the time points analyzed (Fig. 3C). The gene expression patterns imply that at the very beginning of A. tumefaciens infection (2 to 4 hpi), WRKY40 and WRKY60 genes are already expressed.

WRKY and ARF transcription factors bind respectively to specific DNA sequences, W-box (TGAC) and AuxRE (TGTCNC or TGTCTN). Sequence analysis of the two IGRs of pTiC58 revealed that seven W-boxes (one W-box is localized in the 5’ UTR of the Ipt gene) and five AuxREs are located in IGR2 (Table 1, 2), which are equally distributed along the promoter sequence (S5 Fig.). IGR1 drives expression of IaaH and IaaM and contains only one W-box and AuxRE sequence motif, and this is more closely localized upstream of the IaaM than that of the IaaH TATA box. Sequence comparisons of IGR1 and IGR2 regions illustrate that W-boxes and AuxREs are also conserved in the T-DNA regions of several A. tumefaciens strains (Table 2). Similar to the pTiC58, the majority of these elements are enriched in the Ipt promoters whereas only one or two of them are located in the IaaH and IaaM promoter sequences. From this in silico result, it can be concluded that the Ipt oncogenes, rather than IaaH and IaaM of the different A. tumefaciens strains are regulated by WRKY and ARF transcription factors in planta.

Table 2. Number of WRKY-boxes (W-boxes) and auxin response elements (AuxREs) within the intergenic regions (IGRs) of the tumor inducing (Ti) plasmids from different A. tumefaciens strains.

WRKY18, WRKY40 and WRKY60 mutants display an impaired crown gall development

To unravel the role of WRKY18, WRKY40 and WRKY60 in A. tumefaciens-mediated crown gall development, we performed a crown gall growth assay with mutant plants of the three WRKY genes inoculated with the tumorigenic A. tumefaciens strain C58, determining the crown gall weights 25 days later. All mutant genotypes developed smaller crown galls than the wild-type Col-0 (Fig. 4A, B), with the double mutant wrky18/wrky40 and the triple mutant wrky18/40/60 developing the smallest crown galls. The triple mutant was most resistant to crown gall development; about 30% of the mutant plants did not development any crown gall material at all after 25 days. Unfortunately, the role of the ARF5-mediated auxin signaling pathway on crown gall development could not be analyzed due to the strong developmental phenotypes of arf5 mutant plants [51,52].

Figure 4. Arabidopsis wrky mutants develop smaller crown galls.

(A) Crown gall weights of wrky18, wrky40, wrky60 mutants and the wild type Col-0 25 days after inoculation of Arabidopsis inflorescence stems with the tumorigenic A. tumefaciens strain C58. Bars show mean values of crown gall weight (±SE) separated from the stems of at least 40 plants from each genotype. (B) Representative pictures of the stems of the different genotypes 25 days after inoculation of A. tumefaciens. (C) Relative transcript numbers of the Ipt oncogene in stems of the wild-type plant Col-0 and the wrky18/40/60 triple mutant 6 days post inoculation (6 dpi) of A. tumefaciens strain C58. Relative transcript numbers were quantified by qRT-PCR and normalized to 10,000 molecules of ACTIN2/8. Bars show mean values (±SD) of three independent samples. * P<0.05; ** P<0.01; *** P<0.001 (Student’s t-test).

If WRKY18, WRKY40 and WRKY60 activate the Ipt promoter, it would be expected that Ipt oncogene expression would be altered in the WRKY mutant plants. To investigate this, we used quantitative RT-PCR to measure the relative transcript numbers of the Ipt oncogene in Arabidopsis crown gall material of the wrky mutants inoculated with A. tumefaciens strain C58. Compared to crown galls from the wild-type (Col-0) plants, the Ipt transcript levels were similar in crown galls from the wrky18, wrky40 and wrky60 mutants (S6A Fig.). Due to this similarity, i.e., no obvious impact of WRKY on long term Ipt gene expression in crown galls, earlier time points of C58 Arabidopsis stem inoculations were analyzed. At 2 dpi, the Ipt transcript levels were far too low to reliably quantify differences (S6B Fig.). Only at 6 dpi did Ipt transcription reach a measureable level (S6B Fig.) and showed in the triple mutant (wrky18/40/60) a moderate reduction compared to the wild-type (Fig. 4C). The moderate reduction of Ipt transcription may be due to the function of ARF5, which is still expressed in the wrky triple mutant. This assumption is supported by the observation that in crown galls of the wrky single mutants gene expression of ARF5 was elevated and that of IAA12, an inhibitor of ARF5 function, was reduced (S6C Fig.).

WRKY40 and ARF5 proteins interact and synergistically potentiate Ipt promoter activity

The PTA data revealed that the Ipt promoter can be activated by WRKY18, WRKY40, WRKY60 and ARF5. To test whether these transcription factors cooperatively regulate the Ipt promoter, we co-expressed the WRKY40 protein with ARF5 in the presence of the Ipt promoter-LUC construct in Arabidopsis mesophyll protoplasts. The Ipt promoter-driven luciferase activity was clearly higher, particularly in the presence of ARF5 and WRKY40 compared to ARF5 or WRKY40 alone (Fig. 5A). In contrast, expression of ARF5 together with WRKY18 or WRKY60 did not further enhance the Ipt promoter activity. This also indicates that WRKY40 is more important than WRKY18 and WRKY60 for activating the Ipt promoter.

Figure 5. WRKY40 and ARF5 protein interaction potentiates Ipt promoter activity.

(A) Fold induction of Ipt promoter-driven luminescence in the presence of WRKY18, WRKY40, WRKY60 and ARF5 transcription factor expression plasmids in the protoplast transactivation system. The relative luminescence induced by the Ipt promoter in protoplasts without transfection of any of the transcription factor expression plasmids was set to 1. Bars show mean values (±SD) of three independent experiments *** P<0.01 (Student’s t-test). NS: not significant. (B) Bimolecular fluorescence (BiFC) assay with ARF5-cYFP and ARF5-nYFP, WRKY18-nYFP, WRKY40-nYFP, WRKY60-nYFP, (C) with WRKY40-cYFP, (D) with cYFP, (E) with a C-terminal deletion of ARF5 (ARF5Δ722-cYFP), (F) with ARF3-cYFP and (G) with WRKY53-cYFP in Arabidopsis mesophyll protoplasts. YFP, image in fluorescence mode of reconstituted yellow fluorescent proteins; BF, images in bright filed mode; merge, overlay of YFP with the corresponding BF image. Bars, 10 μm.

These results imply that the WRKY40 and ARF5 proteins interact to synergistically activate Ipt gene expression. This was tested using the Bimolecular Fluorescence Complementation (BiFC) assay to study protein interactions between the WRKYs and ARF5. The C-terminal half of the yellow fluorescent protein (cYFP) was fused to the C-terminus of the ARF5 and WRKY40 proteins to express ARF5- and WRKY40-cYFP fusion proteins, respectively. The N-terminal half of YFP (nYFP) was fused to the C-terminus of the three WRKY proteins as well as to ARF5 to generate WRKY18-, WRKY40-, WRKY60-nYFP and ARF5-nYFP. Observation of YFP-mediated fluorescence demonstrates that both WRKY40 and ARF5 interacted with themselves and with all the other expressed genes, when transiently co-expressed in Arabidopsis mesophyll protoplasts (Fig. 5B, C). The fluorescence signal was always restricted to the nucleus. The free cYFP construct was used as negative control, and showed no YFP fluorescence when co-expressed with the WRKY-nYFPs and ARF5-nYFP in protoplasts (Fig. 5D).

It has been reported that the domain III and IV at the C-terminus of the ARF5 protein is important for dimerization and protein-protein-interaction [53–55]. To prove whether these domains are required for the interaction with the WRKY proteins, we fused a C-terminal deletion of ARF5 (1–722 aa) to cYFP (ARF5Δ722-cYFP) and co-expressed them with either ARF5-nYFP or the three WRKY-nYFPs. Although stable [53], the truncated ARF5Δ722 protein was unable to interact with the intact ARF5 protein or with WRKY18, WRKY40 and WRKY60 (Fig. 5E). This indicates that the domains III and IV are not only required for self-interaction, but also for interaction with the three WRKYs. The specificity of the interactions between ARF5 and the three WRKYs was confirmed by co-expressing ARF3 (AT2G33860)-cYFP, which naturally lacks domain III and IV, and WRKY53-cYFP, expressed early after infection with A. tumefaciens strain C58 (3 hpi; S2 Table) [20,53]. Neither ARF3 nor WRKY53 interacted with ARF5, WRKY18, WRKY40, and WRKY60, thus verifying that the interactions between the WRKYs and ARF5 are specific (Fig. 5F, G).

The PTA assays indicate that WRKY40 has a stronger potential to activate the Ipt promoter than WRKY18 and WRKY60 (Fig. 3A and 5A). This implies that WRKY40 regulates the Ipt promoter directly. We therefore analyzed binding of WRKY40 to the Ipt promoter using the electrophoretic mobility shift assay (EMSA). The recombinant WRKY40 protein fused to six histidine amino acids at the N-terminus (6×His-WRKY40) was expressed and purified from E. coli and a 50 bp fragment (−184 bp to −135 bp) of the Ipt promoter, which contains three of the six W-boxes located in the promoter region, was radioactively labeled and served as a probe for EMSA (Fig. 6A). Only a weak band of the shifted Ipt promoter fragment (Fig. 6B, WRKY40-Ipt complex) was observed in the presence of 150 ng purified recombinant 6×His-WRKY40 protein, but a doubled amount of the His-tagged WRKY40 protein (300 ng) exhibited a much stronger band. Addition of unlabeled Ipt promoter fragments as competitor to the reaction mixture significantly reduced the binding of WRKY40 to the labeled Ipt promoter probe. Thus, the WRKY40 protein binds to the Ipt probe in vitro, suggesting that the Ipt promoter is a direct target of the WRKY40 transcription factor in plant cells.

Figure 6. WRKY40 binds to the Ipt promoter in vitro.

(A) Positions of W-boxes (TGAC, grey bars) in the sense (above the line) and anti-sense strand (below the line) of the Ipt promoter. The line below the Ipt promoter (−184 bp to −135 bp) indicates the fragment used as Ipt probe for electrophoretic mobility shift assay (EMSA). (B) EMSA with the labeled Ipt promoter probe in the absence (−) and in the presence of 150 ng (+) or 300 ng (++) of purified recombinant histidine-tagged WRKY40 protein. Competitor indicates without (−) and with (+) addition of unlabeled Ipt promoter probe. The WRKY40-Ipt complex indicates binding of WRKY40 protein to the labeled Ipt probe and the free Ipt probe no protein binding.

The ARF5-mediated auxin-signaling pathway induces Ipt, but not IaaH and IaaM gene expression

That ARF5 enhances the WRKY40-mediated activation of the Ipt promoter suggests that the auxin signaling pathway is involved in regulating Ipt expression. Previous studies have shown that the levels of unconjugated IAA in infected Arabidopsis stems are more than two-fold higher six days after inoculation with A. tumefaciens strain C58 compared to non-inoculated plant stems [20]. We found that crown galls accumulate four times more unconjugated IAA than control tissues and the total level of cytokinins in Arabidopsis crown gall tissues infected with A. tumefaciens strain C58 are 10 times higher than in crown gall-free stem tissues (8414 vs. 849 ng/g dry weight). The dominant cytokinin forms in Arabidopsis crown gall tissues were zeatin conjugates, including zeatin nucleotide (3657 vs. 308 ng/g dry weight) and zeatin riboside (2294 vs. 76 ng/g dry weight). The content of free zeatin was also higher in crown gall tissues than in mock-inoculated stems (544 vs. 34 ng/g dry weight).

Based on these results, we used the PTA system to analyze the impact of auxin and cytokinin on IaaH, IaaM and Ipt promoter activity. The Ipt promoter was highly activated by the bioactive auxin type 1-naphthaleneacetic acid (1-NAA) and the cytokinin type trans-zeatin (Fig. 7A), with the latter much less effective. Increasing concentrations of auxin and cytokinin had no strong enhancing effect on the activity of the three oncogene promoters (Fig. 7A). The Ipt promoter sequence contains five auxin response elements (AuxREs, TGTCNC or TGTCTN) for binding of ARF transcription factors, whereas only one AuxRE is present in the bidirectional IaaH and IaaM promoter sequence (Table 1, S5 Fig.) and ARF transcription factors usually regulate their target genes in an auxin-dependent manner [33,34]. Thus, we analyzed the regulatory effect of ARF5 on the Ipt promoter in the presence of auxin in the PTA system. ARF5 activated the Ipt promoter, an activation that was even stronger when the mesophyll protoplasts were treated with auxin (1-NAA, Fig. 7B). Mutations in the AuxREs (sense TGTCNC or TGTCTN, anti-sense GNGACA or NAGACA) in the Ipt promoter abolished the auxin induction and the enhancing effect of ARF5 (Fig. 7C). It is known that auxin/indole-3-acetic acid (Aux/IAA) proteins can inhibit ARF mediated promoter activation and the repressor of ARF5 is IAA12 (also known as BODENLOS, BDL, AT1G04550) [54]. When we co-transfected Arabidopsis mesophyll protoplasts with the ARF5 and IAA12 plasmid constructs, a significant reduction in the Ipt promoter-driven luciferase activity was found compared to protoplasts transfected with only ARF5 (Fig. 7B). Nonetheless, the level of the Ipt promoter activity was not as low as it was in the absence of any transcription factor, indicating that not all ARF5 proteins are inhibited by IAA12.

Figure 7. ARF5 activates the Ipt promoter in an auxin-dependent manner.

(A) Fold induction of oncogene promoter-driven luminescence in Arabidopsis mesophyll protoplasts treated with auxin (1-NAA) or cytokinin (trans-zeatin) of different concentrations. Protoplasts were transfected with the IaaH, IaaM, and Ipt promoter-luciferase reporter constructs, and then incubated with different concentrations of 1-NAA or trans-zeatin overnight. (B) Fold induction of Ipt promoter-driven luminescence in Arabidopsis mesophyll protoplasts (Ipt pro) in the presence of the transcription factor expressing plasmids ARF5 (Ipt pro ARF5) or ARF5 plus IAA12 (Ipt pro ARF5 IAA12) with (+ 1-NAA) and without auxin (− 1-NAA) addition. (C) Mutations in the five auxin responsive elements (AuxREm, TGTCNC to TGGCNC and TGTCTN to TGGCTN) of the Ipt promoter abolish the ARF5- and auxin-dependent luminescence induction. The relative luminescence of intact or mutated Ipt promoters in the absence of any transcription factor expression plasmids and auxin treatment was set to 1. Bars show mean values (±SD) of three independent experiments. * P<0.05; ** P<0.01; *** P<0.001 (Student’s t-test).

In addition to the W-boxes and AuxREs, the IaaH, IaaM and Ipt promoters also contain ARR1 binding elements (GATT; Table 1), suggesting that the three oncogenes are regulated by type-B ARR transcription factors to mediate cytokinin signaling. The ARR1 gene is expressed at low levels in crown gall tissue of the virulent A. tumefaciens strain C58 and in stems infected with the disarmed strain GV3101 according to real time PCR measurements (S7A Fig.). ARR4 (AT1G10470), a type A transcription factor gene, was strongly expressed in crown gall tumors (S7A Fig.). The ability of both the ARR1 and ARR4 transcription factors to activate the IaaH, IaaM and Ipt promoters was tested in the PTA system. Neither ARR1 nor ARR4 significantly increased luciferase activity driven by the three oncogene promoters, even in the presence of trans-zeatin (S7B Fig.). Hence, the ARF5-mediated auxin signaling pathway, but not that of cytokinin, regulates Ipt expression, whereas the expression of IaaH and IaaM is not affected by either of the two signaling pathways.


The plant pathogen, Agrobacterium tumefaciens takes advantage of the host transcriptional machinery to express its own T-DNA encoded oncogenes IaaH, IaaM and Ipt in plant cells. Expression of the oncogenes results in increased production of the phytohormones auxin and cytokinin, which induce uncontrolled cell proliferation and crown gall development. The T-DNA transformation process and the roles of the encoded oncogene enzymes is far better understood [10,56,57] than the regulation of oncogene expression in plant cells. We therefore examined this regulation, asking whether the expression of the IaaH, IaaM and Ipt oncogenes is regulated by host transcription factors and how oncogene expression is coordinated to obtain tumor-inducing auxin/cytokinin levels in a T-DNA transformed cell.

Agrobacterium tumefaciens utilizes a transcription factor of the pathogen defense pathway to induce Ipt oncogene expression

Expression of a gene in a eukaryotic cell requires general sequence features (e.g. TATA, CAAT) and potentially cis-regulatory elements for the binding of transcription factors. For the Ipt promoter of the octopine Ti plasmid pTiAch5, previous studies have shown that it binds CBF, a protein of unknown function from tobacco nuclear protein extracts [25–27]. This implies that at the least, expression of the Ipt oncogene is regulated by plant derived transcription factors. Nonetheless, using the PTA screening system we found that no transcription factor activated the IaaH and IaaM promoters of pTiC58. This may be because the transcription factor collection used for screening did not cover all the encoded Arabidopsis transcription factors; candidates for binding to the IaaH and IaaM promoters may have been missed. However, the very few cis-regulatory elements in the relatively short promoter sequence and the low level of transcription in crown galls, in addition to the low promoter activity in protoplasts, suggest that the IaaH and IaaM genes are not strongly activated by transcription factors, but instead are constitutively expressed at low basal levels. In contrast, the Ipt oncogene promoter of pTiC58 contains several W-boxes and is activated by the WRKY18, WRKY40 and WRKY60 proteins. The impaired crown gall growth on the wrky18, wrky40 and wrky60 mutant plants indicates that these WRKY transcription factors have a positive effect on crown gall development. The three WRKYs are paralogous transcription factors that cooperatively regulate biotic and abiotic stress responses in Arabidopsis [49,58–63] and the respective wrky mutants are known to be more resistant to biotrophic pathogens such as Pseudomonas syringae and powdery mildew Golovinomyces orontii [50]. Hence, the smaller crown galls on these wrky mutants may result from both fewer transformation events due to the stronger resistance response towards biotrophic pathogens and/or from reduced Ipt expression due to the loss of wrky function. Unfortunately, these two processes are not easy to separate in infection-based assays.

It is known that transcription of WRKY40 and WRKY60 is induced by fungal and bacterial pathogens [49,50]. Likewise, A. tumefaciens inoculation induced their transcription within two hours, indicating that they are expressed quite early in response to this pathogen. Thus, it is conceivable that the WRKYs are needed to trigger Ipt oncogene expression from the very beginning in a T-DNA transformed cell, so these pathogen responsive genes are already expressed when the T-DNA enters the host cell. Consequently, a reduction in Ipt promoter activity can be observed early on in the wrky triple mutant, vanishing at later infection stages. The relatively moderate difference in Ipt gene expression between the wrky triple mutant and wild-type most likely results from the increased expression of ARF5 and reduced expression of its inhibitor IAA12 in the mutant background. Thus, A. tumefaciens hijacks a host transcription factor, which is part of the plant pathogen defense machinery, to initiate expression of its own oncogene in the host cell.

Auxin, but not cytokinin signaling is important for inducing Ipt oncogene expression

A. tumefaciens and T-DNA transformed plant cells produce auxin and cytokinin [13,20]. Cytokinin affects cell division, essential for cell proliferation and initiation of crown gall development. Only the activity of the Ipt promoter, not that of the IaaH and IaaM genes, increased upon application of trans-zeatin, the dominant cytokinin in Arabidopsis crown galls. Eight binding elements for the ARR1 transcription factor are located in the bidirectional promoter of IaaH/IaaM and seven in the Ipt promoter. ARR1 is a type-B ARR transcription factor that activates transcription of cytokinin responsive genes [64,65]. Nonetheless, the activity of all three oncogene promoters was not influenced either by ARR1 or ARR4, even in the presence of trans-zeatin. This indicates that cytokinin signaling does not have a dominant role in oncogene expression. The auxin type 1-NAA was much more effective than trans-zeatin in activating the Ipt promoter, but again, not for the promoters of IaaH or IaaM. Elevated levels of free IAA are detectable in infected tissues six days after inoculation with A. tumefaciens strain C58 [20] and at the same infection stage, expression of the ARF5 gene begins to increase, as shown in the microarray data (1.49 fold, P value = 0.006) [20,46]. The Ipt promoter contains five AuxREs and is activated by 1-NAA and by the auxin response factor ARF5 upon release from inhibition by IAA12 in an auxin-dependent manner. Expression of the ARF5 gene is induced by auxin [66] and the elevated auxin levels in plant tissues infected and T-DNA transformed by A. tumefaciens most likely induce ARF5 gene expression and de-repress the ARF5 protein by proteolysis of IAA12. The release of ARF5 inhibition in the presence of auxin leads finally to activation of the Ipt promoter in the T-DNA transformed plant cell and may contribute to the moderate differences of Ipt transcript numbers in the wrky mutants and wild-type. Taken together, the results indicate that auxin is an important factor in regulating Ipt oncogene expression, which exerts its function through the auxin-sensitive transcription factor ARF5.

WRKY40 and ARF5 synergistically boost Ipt gene expression, thereby integrating host pathogen defense and auxin signaling

Our study shows that WRKY40 binds directly to the Ipt promoter in vitro and has the strongest effect on Ipt promoter activation in plant cells, an activation that increases even further in the presence of the ARF5 transcription factor. It has been shown that WRKY transcription factors specifically interact with different kinds of proteins [67] and WRKY18, WRKY40 and WRKY60 interact with each other and themselves [49], a result confirmed in this study. Moreover, WRKY18, WRKY40, and WRKY60 interact with ARF5. Most ARFs contain four important domains, except for ARF3, ARF13 and ARF17, which lack domain III and IV and ARF23, which has only domain I [31]. Domain III and IV are localized at the C-terminus of ARF proteins and are important for dimerization and interaction with Aux/IAA proteins [53]. According to our study, the domain III and IV of ARF5 seem to be required for the interaction with the three WRKY transcription factors. The interaction of ARF5 with WRKY40, but not that with WRKY18 and WRKY60, greatly enhances the activation of the Ipt promoter, so emphasizing the role of WRKY40 as the most important transcriptional activator of Ipt gene expression. Moreover, the WRKY40-ARF5 interaction links two signaling pathways for the regulation of Ipt gene expression: the ARF5-dependent auxin and WRKY-mediated pathogen defense pathway. Both pathways are activated in the host plant upon infection with A. tumefaciens and synergistically boost expression of the Ipt gene in T-DNA transformed cells.


This study suggests a bifactorial regulation of oncogene expression in T-DNA transformed plant host cells (Fig. 8). Just after A. tumefaciens infection, auxin and cytokinin levels are as low as in an untransformed plant cell. The WRKY40 gene is soon expressed in response to infection, and the protein binds to W-boxes in the Ipt promoter to induce gene expression (Fig. 8A). Under low auxin conditions, ARF5 interacts with IAA12, so is inactivated. Over time, the auxin concentration increases in the T-DNA transformed cell, the result of the ubiquitous expression of IaaH and IaaM, driven by binding the RNA polymerase II complex to the promoter and additional auxin that can be secreted from the A. tumefaciens cells into the apoplast. Under high auxin concentrations, the ARF5 inhibitor IAA12 is poly-ubiquitinylated and degraded, thus releasing the transcription factor ARF5. The de-repressed ARF interacts via domain III and IV with WRKY40, resulting in strong expression of the Ipt oncogene. Taken together, this transcription factor interaction integrates two signaling pathways: the WRKY-based pathogen defense pathway for initial induction of Ipt gene expression and later, the auxin signaling pathway to boost Ipt expression. Moreover, the alterations in Ipt expression levels may be a mechanism to fine-tune the cytokinin to auxin ratios in a transformed plant cell. The appropriate auxin/cytokinin balance is an important mechanism to control whether a crown gall will proliferate or grow and differentiate.

Construction of Disarmed Ti Plasmids Transferable between Escherichia coli and Agrobacterium Species▿†

  1. Kazuya Kiyokawa,
  2. Shinji Yamamoto,
  3. Kei Sakuma,
  4. Katsuyuki Tanaka,
  5. Kazuki Moriguchi and
  6. Katsunori Suzuki*
  1. Department of Biological Science, Graduate School of Science, Hiroshima University, Higashi-Hiroshima, Hiroshima 739-8526, Japan


Agrobacterium-mediated plant transformation has been used widely, but there are plants that are recalcitrant to this type of transformation. This transformation method uses bacterial strains harboring a modified tumor-inducing (Ti) plasmid that lacks the transfer DNA (T-DNA) region (disarmed Ti plasmid). It is desirable to develop strains that can broaden the host range. A large number of Agrobacterium strains have not been tested yet to determine whether they can be used in transformation. In order to improve the disarming method and to obtain strains disarmed and ready for the plant transformation test, we developed a simple scheme to make certain Ti plasmids disarmed and simultaneously maintainable in Escherichia coli and mobilizable between E. coli and Agrobacterium. To establish the scheme in nopaline-type Ti plasmids, a neighboring segment to the left of the left border sequence, a neighboring segment to the right of the right border sequence of pTi-SAKURA, a cassette harboring the pSC101 replication gene between these two segments, the broad-host-range IncP-type oriT, and the gentamicin resistance gene were inserted into a suicide-type sacB-containing vector. Replacement of T-DNA with the cassette in pTiC58 and pTi-SAKURA occurred at a high frequency and with high accuracy when the tool plasmid was used. We confirmed that there was stable maintenance of the modified Ti plasmids in E. coli strain S17-1λpir and conjugal transfer from E. coli to Ti-less Agrobacterium strains and that the reconstituted Agrobacterium strains were competent to transfer DNA into plant cells. As the modified plasmid delivery system was simple and efficient, conversion of strains to the disarmed type was easy and should be applicable in studies to screen for useful strains.

Agrobacterium-mediated transformation has been considered the most efficient and reliable method for plant biology and biotechnology. This methodology has been established for many plants, but not for others. One of the major factors affecting its applicability is the limited number of donor Agrobacterium strains, because the method depends exclusively on the host ranges of the strains.

Wild-type Agrobacterium strains harboring a tumor-inducing (Ti) plasmid are the causative agent of crown gall tumor disease in dicotyledonous plants (35). The transfer DNA (T-DNA) and virulence gene (vir) regions in the Ti plasmid are essential for tumorigenesis. The vir gene products nick the T-DNA region at its left border (LB) and right border (RB) and then transfer T-DNA into plant cells. T-DNA contains phytohormone synthesis genes, whose expression causes infected plants to suffer from unregulated growth (5, 26). The hairy-root-inducing (Ri) plasmid has a similar T-DNA system.

The binary vector system (11) is widely used for Agrobacterium-mediated transformation. Binary vectors are small plasmids with a cloning site and a selectable marker gene between LB and RB (2). To ensure transformation without tumorigenicity, Agrobacterium strains used for transformation contain a modified Ti plasmid, which lacks T-DNA (disarmed) but retains the entire vir region. Unfortunately, only a small number of Ti plasmids have been disarmed.

Most pathogenic Agrobacterium strains are classified into three species: Agrobacterium tumefaciens (biovar 1, Rhizobium radiobacter), Agrobacterium rhizogenes (biovar 2, Rhizobium rhizogenes), and Agrobacterium vitis (biovar 3, Rhizobium vitis) (33). The genomic organizations of the Agrobacterium species are diverse (25, 27, 29). Pathogenic strains of each species are variable (1), and some of them might be potentially more effective for transformation than the strains used previously. For instance, Agrobacterium strain KAT23 causes tumors in legume plants, including common bean and soybean, very effectively (34). Disarmed Ti or Ri plasmids are either chosen from mutants or created by homologous recombination with a plasmid designed for this purpose (12, 16, 17). Both methods require either extensive screening or knowledge of structural and functional information for the plasmids. However, the large size of Ti and Ri plasmids, approximately 200 kbp, makes structural analysis and modification difficult. Complete nucleotide sequences of several Ti and Ri plasmids (for example, pTi-SAKURA, pTiC58, and pRi1724) have been reported (9, 14, 24, 26, 31). Accumulation of such nucleotide sequence information makes targeted replacement easier than it was previously. However, the large size of T-DNA obstructs the double crossover in the removal process during engineering. In addition to Ti plasmids, chromosomal virulence genes are necessary for plant transformation. It has been pointed out that combining a Ti plasmid with certain chromosomal backgrounds can markedly influence virulence (8). Thus, transfer of large plasmids to various Agrobacterium strains is another important engineering step, which is still not easy for researchers who are not familiar with Agrobacterium biology.

In this study, we describe a simple method and tool plasmids for constructing versatile disarmed nopaline-type Ti plasmids mobilizable from Escherichia coli to Agrobacterium strains, conversion of nopaline-type Agrobacterium strains to disarmed strains using the tool plasmids and simple selection media, and conversion of Ti-less Agrobacterium strains to disarmed strains using the modified Ti plasmids.

Previous SectionNext Section


Bacterial strains and culture conditions.Bacterial strains used in this study are listed in Table 1. E. coli strains were grown at 37°C in LB medium (1% Bacto tryptone, 0.5% NaCl, 0.5% yeast extract). A. tumefaciens strains were cultured at 28°C in LB medium or IFO medium (1% polypeptone, 0.2% yeast extract, 0.1% MgSO4). A. rhizogenes strains were cultured at 28°C in IFO medium. Antibiotics were added at the following final concentrations: 50 μg/ml gentamicin, 50 μg/ml kanamycin, 30 μg/ml nalidixic acid, 50 μg/ml rifampin, 50 μg/ml ampicillin, and 50 μg/ml neomycin.


Bacterial strains and plasmids used in this study

Plant materials used for transformation.Nicotiana tabacum SR-1 and Kalanchoe sp. were used as host plants for infection and DNA transfer experiments. N. tabacum SR-1 was cultured azenically on MS medium solidified with 0.8% agar at 28°C with continuous illumination. Kalanchoe sp. was cultured in a greenhouse. Leaves were surface sterilized with 1% sodium hypochlorite for 15 min and rinsed for 2 min with sterile distilled water four times before azenic experiments were performed.

Plasmid construction.Construction of tool plasmids pLRS-GmsacB and pLRS-Gms2 is described in the supplemental material. The 1.4-kbp left fragment (LL) just outside the left border of T-DNA and the 1.0-kbp right fragment (RR) just outside the right border of T-DNA were derived from pTi-SAKURA (24). The gentamicin resistance (Gmr) gene was obtained from pUCGm2, the sacB gene and the Kmr gene were obtained from pK18mobsacB (21), IncP type (RK2) oriT was obtained from pJP5603 (18), and low-copy-number pSC101 oriV was obtained from pMW119 (Nippon Gene, Tokyo, Japan).

The binary plasmid pBIN-GI was prepared as follows. A 2.6-kbp HindIII-EcoRI fragment containing the β-glucuronidase (GUS) gene with an intron was obtained from pIG221 (15) and inserted into pBIN19 (2).

DNA preparation and analysis.Plasmid DNA was extracted from bacterial cells by the alkaline sodium dodecyl sulfate method (3). Manipulation of plasmid DNA was performed using standard methods.

Bacterial transformation.Modified shuttle Ti plasmids were extracted from A, tumefaciens strains by the modified alkaline sodium dodecyl sulfate method and purified by ethidium bromide-CsCl gradient ultracentrifugation. Purified shuttle Ti plasmids were introduced into E. coli strains by electroporation as described previously (20, 32).

Plasmids were transferred from E. coli to Agrobacterium strains by conjugal transfer as described elsewhere (28), with some modifications. The E. coli-Agrobacterium cell mixture was spotted onto LB agar for conjugation of A. tumefaciens and onto IFO agar for conjugation of A. rhizogenes. After overnight incubation at 28°C, cells were resuspended and spread onto appropriate selective agar media.

Plant transformation.Transformation of tobacco leaf disks was carried out as described by Clemente (6), with some modifications. Agrobacterium strains transformed with the binary vector pBIN-GI were grown overnight in liquid media supplemented with the appropriate antibiotics. Tobacco leaf disks (diameter, 1 cm) were immersed in the Agrobacterium suspension (optical density at 660 nm, 0.8) for 5 min and cocultivated for 2 days at 22°C with continuous fluorescent light illumination. After cocultivation, the leaf disks were cultivated on MS selective agar with 200 μg/ml claforan and 100 μg/ml kanamycin at 28°C with fluorescent light illumination. Kalanchoe leaf disks were subjected to the same transformation procedure but with different phytohormone and antibiotic concentrations (0.5 mg/liter benzyladenine, 2.0 mg/liter naphthylacetic acid, and 50 μg/ml kanamycin).

Quantitative and histochemical analyses of GUS activity were carried out as described by Jefferson et al. (13).

Previous SectionNext Section


Construction of disarmed shuttle Ti plasmids.We designed a simple engineering scheme that can make pathogenic Ti plasmids disarmed, stably maintainable in E. coli, and mobilizable between E. coli and Agrobacterium species. As an example, we used the scheme with nopaline-type plasmids. We first constructed pLRS-GmsacB and pLRS-Gms2 (see Fig. S1 in the supplemental material) as tool plasmids to modify nopaline-type Ti plasmids. These tool plasmids are pK18mobsacB containing two fragments, LL and RR, which neighbor to the left of LB and to the right of RB of T-DNA in pTi-SAKURA, respectively, and a cassette containing a gentamicin resistance gene, the low-copy-number type replication origin (oriV) derived from pSC101, and the IncP-type transfer origin (oriT) sandwiched between LL and RR. The pSC101 replication ori should allow the chimeric plasmids to replicate at a very low copy number in E. coli.

Two nopaline-type Ti plasmids, pTiC58 and pTi-SAKURA, were modified using pLRS-GmsacB, as shown in Fig. 1. First, the pLRS-GmsacB plasmid in E. coli was introduced by conjugation into two pathogenic nopaline-type strains belonging to A. tumefaciens (biovar 1). C58rif is a pathogenic strain harboring pTiC58. C58C1 is a Ti-less strain. C58C1 harboring pTi-SAKURA is another pathogenic strain. Because pLRS-GmsacB cannot replicate in Agrobacterium cells, the tool plasmid should integrate into the Ti plasmids by homologous recombination at either LL or RR in the transformants (Fig. 1A). The Agrobacterium transconjugants were resistant to gentamicin and kanamycin and sensitive to sucrose due to the Gmr, Kmr, and sacB genes on the fusion plasmids.

FIG. 1.

Conversion of pathogenic Ti plasmids so that they are disarmed and transferable between E. coli and Agrobacterium. The modification of pTiC58 and pTi-SAKURA consists of two steps. (A) pLRS-GmsacB was inserted in vivo into pTiC58 and pTi-SAKURA by homologous recombination at either RR or LL. (B) Cells harboring the fused plasmid DNA were cultivated on LB agar containing sucrose and gentamicin in order to select for the subsequent crossover products. Only the recombinant that did not include the T-DNA portion was selected by cultivation on the medium.

Next, the transconjugants harboring the resulting fusion plasmid were cultured on LB agar supplemented with gentamicin and sucrose. Cultivation in a sucrose-containing medium selects for cells that do not have the sacB gene. Loss of the fusion plasmid can occur at a high frequency. Loss of this plasmid converts cells to Gms, Kms, sucrose-resistant cells. Deletion of the sacB gene from the plasmid can take place at a high frequency through homologous recombination in two ways: recombination between two RR segments, resulting in removal of the pLRS-GmsacB portion, or, alternatively, recombination between two LL segments, resulting in loss of the T-DNA region (Fig. 1B). The former recombination converts cells to Gms, whereas the latter maintains Gmr genes. Thus, colonies on the selective agar plate were expected to have a disarmed type of pTi. To confirm the lack of T-DNA in the derivatives of pTiC58 and pTi-SAKURA, for each Ti plasmid four colonies were randomly chosen from the selective agar culture and analyzed by PCR. T-DNA products were not detected in any of the colonies examined, whereas the virB gene was detected in every colony examined in another PCR experiment (data not shown). These results suggest that there was accurate and frequent removal of the long T-DNA region by replacement using pLRS-GmsacB and the simple selection media. The resultant Ti plasmids were designated pTiC58-S and pTi-SAKURA-S.

Introduction of modified Ti into Agrobacterium species via E. coli.Modified Ti plasmids pTiC58-S and pTi-SAKURA-S were extracted from the Agrobacterium strains. The plasmid DNAs were introduced into two E. coli strains, S17-1λpir and SURE. In order to check the structural integrity of the modified Ti plasmids during maintenance in E. coli, the plasmid DNAs were extracted from the E. coli transformants. The EcoRI fragment ladder profiles suggest that pTi-SAKURA-S was maintained stably in S17-1λpir (Fig. 2A) and that pTiC58-S was also maintained stably in the same E. coli strain (data not shown). Structural alteration was not detectable even after three serial repetitions of the E. coli culture (Fig. 2B). In contrast to the plasmids in S17-1λpir, pTi-SAKURA-S suffered from significant deletions in the other E. coli strain, strain SURE (Fig. 2A).

FIG. 2.

Stability of the modified Ti plasmids. pTiC58-S and pTi-SAKURA-S were extracted from Agrobacterium cells and then introduced into two E. coli strains, S17-1λpir and SURE. Plasmid DNA was extracted from each E. coli transformant culture and then digested with EcoRI before electrophoretic separation in a 0.8% agarose gel. (A) pTi-SAKURA-S transformant colonies of S17-1λpir and of SURE were cultivated in liquid medium. (B) Cultivation of one S17-1λpir transformant was repeated serially three times. The presence (+Gm) or absence (−Gm) of gentamicin in the medium is indicated.

Because S17-1λpir possesses the IncP-type tra genes in its chromosome, it was expected that the S17-1λpir transformants could mobilize the modified Ti plasmids to various bacteria by conjugation. The Ti plasmid-less Agrobacterium strain C58C1 was cocultivated with the S17-1λpir transformants harboring the modified Ti plasmids. The resulting Rifr Gmr transconjugant frequencies were 5 × 10−5 for pTiC58-S and 4 × 10−5 for pTi-SAKURA-S. Similarly, the modified Ti plasmids were also introduced successfully by conjugation into another Ti plasmid-less A. tumefaciens strain, strain MNS-1, and into an Ri plasmid-less A. rhizogenes strain, strain A4RL.

Evaluation of reconstructed Agrobacterium strains.We performed plant transformation experiments to confirm the ability of the Agrobacterium transconjugants constructed as described above. For this experiment, the Agrobacterium transconjugants were transformed with an intron-containing GUS reporter plasmid pBIN-GI. The activity of the reconstructed Agrobacterium strains for transformation of tobacco leaf disks was as high as that of the original Agrobacterium strains in which the Ti plasmids were modified (see Fig. S2 in the supplemental material). This result indicates that the modified Ti plasmids have T-DNA transfer ability even after transmission from E. coli to Agrobacterium.

As shown above, pTiC58-S and pTi-SAKURA-S in S17-1λpir were mobilizable into Agrobacterium strains, and this enabled us to easily convert Agrobacterium strains to a disarmed type. We also tried to evaluate the disarmed Ti plasmids, as well as the Ti- and Ri-free strains. As mentioned above, we introduced each of the two disarmed Ti plasmids into two A. tumefaciens strains, C58C1 and MNS-1, and one A. rhizogenes strain, A4RL. The disarmed-plasmid-containing strains were transformed with the GUS reporter binary plasmid pBIN-GI. Then transformation of tobacco and Kalanchoe leaf disks was carried out with these reconstructed Agrobacterium strains. Two weeks after cocultivation with the donor Agrobacterium strains, kanamycin-resistant (Kmr) calluses were observed on the tobacco leaf disks. pTi-SAKURA-S was as effective as pTiC58-S in all strains tested (data not shown). Kmr calluses were induced in tobacco frequently by C58C1 strains containing this plasmid and less frequently by A4RL strains containing the same disarmed plasmid. However, Kmr calluses were rarely induced by MNS-1 strains containing this plasmid. The data for GUS activity in the tobacco leaf disks (Fig. 3A) was comparable to the data for formation of Kmr calluses. Regenerated recombinant tobacco plants were obtained from the Kmr calluses and showed GUS activity in their leaves and roots (see Fig. S3 in the supplemental material). When we treated Kalanchoe leaf disks, however, A4RL strains containing the disarmed plasmid induced higher GUS activity than C58C1 strains containing the same plasmid, as shown in Fig. 3B. The preference for A4RL of Kalanchoe sp. was in contrast to the preference for C58C1 rather than A4RL of tobacco.

FIG. 3.

Evaluation of the plant transformation efficiencies of reconstructed Agrobacterium strains with different genome backbones. (A) Expression of GUS activity in tobacco leaf disks cocultivated with reconstructed Agrobacterium strains harboring pBIN-GI. (B) Expression of GUS activity in Kalanchoe leaf disks cocultivated with reconstructed Agrobacterium strains harboring pBIN-GI. Cell extracts of the leaf disks were prepared. The filled bars indicate the relative GUS activity of leaf disks transformed with C58C1 harboring pTiC58-S and pBIN-GI. The open bars indicate specific GUS activity. The data averages and with standard deviations of three independent experiments (five leaf disks each). 4MU, 4-methylumbelliferone.

Previous SectionNext Section


In this study, we described a new disarming scheme and construction of versatile disarmed nopaline-type Ti plasmids mobilizable from E. coli to Agrobacterium strains and then conversion of Ti-less Agrobacterium strains to disarmed strains taking advantage of the modified Ti plasmids. Stable maintenance of Ti plasmids both in E. coli and during the transfer step is a prerequisite for delivering the disarmed plasmids to many strains of Agrobacterium and related genera and subsequent examination of their plant transformation abilities. Several research groups have tried to maintain Ti plasmids in E. coli. Native Ti plasmids cannot replicate in E. coli and therefore require additional replication genes that are functional in E. coli. Sprinzl and Geider (23) added the phage fd ori to a nopaline-type Ti plasmid. However, the modified Ti plasmid was inserted into chromosomal DNA of E. coli. Velikov and Buryanov (30) added ColE1 ori to a nopaline-type Ti plasmid, but the modified Ti plasmid was either inserted into chromosomal DNA or maintained as a much smaller plasmid resulting from large deletions.

In this study, we replaced T-DNA with a cassette containing oriT derived from RK2 and oriV derived from pSC101. This replacement was efficient using the tool plasmid constructed in this study. Two modified Ti plasmids were stably maintained in E. coli strain S17-1λpir. Substitution of low-copy-number oriV for high copy-number-number oriV is likely to be effective for stable maintenance in E. coli. On the other hand, the modified Ti plasmids were damaged in another E. coli strain, strain SURE, due to large deletions, even though SURE was developed using a scheme to increase plasmid structural stability by mutating genes related to DNA recombination and repair pathways (10). In any case, it is clear that the E. coli strain used is very important for Ti plasmid maintenance.

It was easy to transfer the modified Ti plasmids from S17-1λpir to Agrobacterium strains. Moreover, reconstructed A. tumefaciens and A. rhizogenes strains harboring the modified Ti plasmids successfully transformed plant cells. Therefore, using E. coli strain S17-1λpir harboring the shuttle Ti plasmids, various Ti- and Ri-less Agrobacterium strains could be easily converted to disarmed strains useful for plant transformation tests. Plasmid delivery by IncP-type system conjugation does not require addition of any special inducer molecules and enables transfer to wide range of bacteria, while conjugation with the tra regulon on Ti plasmids requires a special inducer, such as agrocinopine (7, 19), which is not available commercially.

Broothaerts et al. (4) mobilized pTiEHA101 derivatives that contain IncP-type oriT using transferable helper plasmid RP4-4 into Sinorhizobium meliloti, Mesorhizobium loti, and a Rhizobium species. They detected T-DNA transfer ability in the transconjugant bacteria. It was necessary to remove the helper plasmid from the transconjugants, because the transconjugants received not only Ti but also the helper plasmid and the latter suppressed the T-DNA transfer ability. E. coli donor strain S17-1λpir employed in this study was easy to select against and moreover is convenient in that it does not deliver the helper IncP plasmid to recipient cells.

The C58C1 strains having modified Ti transformed tobacco leaf disks more efficiently than the A4RL strains harboring the same modified Ti did. On the other hand, the latter strains were more effective at transforming Kalanchoe leaf disks. These results suggest that the various genomic backgrounds of the Agrobacterium strains differentially influence the fitness for each plant. There might be strains among pathogenic Agrobacterium strains that are more efficacious than the commonly used Agrobacterium strains. The disarmed Ti plasmids constructed in this study would help screening for such strains.

Complete nucleotide sequences are available for several different types of Ti and Ri plasmids (26). The difference in the auxiliary vir region affects the host range in part. It is worth replacing the LL and RR segments in the tool plasmids with the corresponding segments of various types of plasmids in order to develop disarmed strains of a type other than the nopaline type.

In addition to pLRS-GmsacB, we constructed pLRS-Gms2 (see Fig. S2 in the supplemental material). The latter tool plasmid can also be used to disarm nopaline-type plasmids and is superior to pLRS-GmsacB since it lacks the Apr gene in the cassette and therefore does not increase the resistance to β-lactam antibiotics in the disarmed strains. Using a simple and efficient Ti-curing method which we reported previously (32) and the shuttle Ti plasmids constructed in this study, it would be easy to convert many pathogenic Agrobacterium strains to disarmed strains, even for researchers who are not familiar with Agrobacterium biology.

Previous SectionNext Section


This research was supported in part by the Ministry of Education, Science, Sports and Culture (grant-in-aid for scientific research 20570221) and by the Japan Science and Technology Agency.

Previous SectionNext Section


    • Received 11 August 2008.
    • Accepted 22 January 2009.
  • ↵*Corresponding author. Mailing address: Department of Biological Science, Graduate School of Science, Hiroshima University, Higashi Hiroshima 739-8526, Japan. Phone: 81-82-424-7455. Fax: 81-82-424-0734. E-mail: ksuzuki{at}
  • ↵▿ Published ahead of print on 30 January 2009.

  • ↵† Supplemental material for this article may be found at


  1. Bautista-Zapanta, J., H. Arafat, K. Tanaka, H. Sawada, and K. Suzuki. 9 November 2007, posting date. Variation of 16S-23S internally transcribed spacer sequence and intervening sequence in rDNA among the three major Agrobacterium species. Microbiol. Res.doi:10.1016/j.micres.2007.08.003.

  2. Bevan, M.1984. Binary Agrobacterium vectors for plant transformation. Nucleic Acids Res.12:8711-8721.

  3. Birnboim, H. C., and J. Doly.1979. A rapid alkaline extraction procedure for screening recombinant plasmid DNA. Nucleic Acids Res.7:1513-1523.

  4. Broothaerts, W., H. J. Mitchell, B. Weir, S. Kaines, L. M. Smith, W. Yang, J. E. Mayer, C. Roa-Rodríguez, and R. A. Jefferson.2005. Gene transfer to plants by diverse species of bacteria. Nature433:629-633.

  5. Citovsky, V., V. Kozlovsky, B. Lacroix, A. Zaltsman, M. Dafny-Yelin, S. Vyas, A. Tovkach, and T. Tzfira.2007. Biological systems of the host cell involved in Agrobacterium infection. Cell. Microbiol.9:9-20.

  6. Clemente, T.2006. Nicotiana (Nicotiana tobaccum, Nicotiana benthamiana). Methods Mol. Biol.343:143-154.

  7. Ellis, J. G., A. Kerr, A. Petit, and J. Tempe.1982. Conjugal transfer of nopaline and agropine Ti-plasmids—the role of agrocinopines. Mol. Gen. Genet.186:269-274.

  8. Gelvin, S. B., and C. Liu.1994. Genetic manipulation of Agrobacterium tumefaciens strains to improve transformation of recalcitrant plant species, p. B4/1-B4/13 In S. B. Gelvin and R. A. Schilperoort (ed.), Plant molecular biology manual, 2nd ed. Kluwer Academic Publishers, Dordrecht, The Netherlands.

  9. Goodner, B., G. Hinkle, S. Gattung, N. Miller, M. Blanchard, et al.2001. Genome sequence of the plant pathogen and biotechnology agent Agrobacterium tumefaciens C58. Science294:2323-2328.

  10. Greener, A. L. September 1996. Cloning host organisms. U.S. patent 5552314.

  11. Hoekema, A., P. R. Hirsch, P. J. J. Hooykaas, and R. A. Schiperoort.1983. A binary plant vector strategy based on separation of vir- and T-region of the Agrobacterium tumefaciens Ti-plasmid. Nature303:179-180.

  12. Hood, E. E., G. L. Helmer, R. T. Fraley, and M. D. Chilton.1986. The hypervirulence of Agrobacterium tumefaciens A281 is encoded in a region of pTiBo542 outside of T-DNA. J. Bacteriol.168:1291-1301.

  13. Jefferson, R. A., T. A. Kavanagh, and M. W. Bevan.1987. GUS fusions: β-glucuronidase as a sensitive and versatile gene fusion marker in higher plants. EMBO J.6:3901-3907.

  14. Moriguchi, K., Y. Maeda, M. Satou, N. S. N. Hardayani, M. Kataoka, N. Tanaka, and K. Yoshida.2001. The complete nucleotide sequence of a plant root-inducing (Ri) plasmid indicates its chimeric structure and evolutionary relationship between tumor-inducing (Ti) and symbiotic (Sym) plasmids in Rhizobiaceae. J. Mol. Biol.307:771-784.

  15. Ohta, S., S. Mita, T. Hattori, and K. Nakamura.1990. Construction and expression in tobacco of a β-glucuronidase (GUS) reporter gene containing an intron within the coding sequence plant. Plant Cell Physiol.31:805-813.

  16. Ooms, G., P. J. Hooykaas, R. J. Van Veen, P. Van Beelen, T. J. Regensburg-Tuïnk, and R. A. Schilperoort.1982. Octopine Ti-plasmid deletion mutants of Agrobacterium tumefaciens with emphasis on the right side of the T-region. Plasmid7:15-29.

  17. Palanichelvam, K., P. Oger, S. J. Clough, C. Cha, A. F. Bent, and S. K. Farrand.2000. A second T-region of the soybean-supervirulent chrysopine-type Ti plasmid pTiChry5, and construction of a fully disarmed vir helper plasmid. Mol. Plant-Microbe Interact.13:1081-1091.

  18. Penfold, R. J., and J. M. Pemberton.1992. An improved suicide vector for construction of chromosomal insertion mutations in bacteria. Gene118:145-146.

  19. Piper, K. M., and S. K. Farrand.2000. Quorum sensing but not autoinduction of Ti plasmid conjugal transfer requires control by the opine regulon and the antiactivator TraM. J. Bacteriol.182:1080-1088.

  20. Sawahel, W., G. Sastry, C. Knight, and D. Cove.1993. Development of an electro-transformation system for Escherichia coli DH10B. Biotechnol. Tech.7:261-266.

  21. Schäfer, A., A. Tauch, W. Jäger, J. Kalinowski, G. Thierbach, and A. Pühler.1994. Small mobilizable multi-purpose cloning vector derived from the Escherichia coli plasmids pK18 and pK19: selection of defined deletions in the chromosome of Corynebacterium glutamicum. Gene145:69-73.

  22. Simon, R., U. Priefer, and A. Pühler.1983. A broad host range mobilization system for in vivo genetic engineering: transposon mutagenesis in Gram-negative bacteria. Bio/Technology1:784-794.

  23. Sprinzl, M., and K. Geider.1988. Transfer of the Ti plasmid from Agrobacterium tumefaciens into Escherichia coli cells. J. Gen. Microbiol.134:413-424.

  24. Suzuki, K., Y. Hattori, M. Uraji, N. Ohta, K. Iwata, K. Murata, A. Kato, and K. Yoshida.2000. Complete nucleotide sequence of a plant tumor-inducing Ti plasmid. Gene242:331-336.

  25. Suzuki, K., K. Iwata, and K. Yoshida.2001. Genome analysis of Agrobacterium tumefaciens: construction of physical maps for linear and circular chromosomal DNAs, determination of copy number ratio and mapping of chromosomal virulence genes. DNA Res.8:141-147.

  26. Suzuki, K., K. Tanaka, S. Yamamoto, K. Kiyokawa, K. Moriguchi, and K. Yoshida.2009. Ti and Ri plasmids, p. 133-150. In E. Schwartz (ed.), Microbial megaplasmids. Microbiology monographs, vol. 11. Springer-Verlag, Heidelberg, Germany.

  27. Tanaka, K., H. Urbanczyk, H. Matsui, H. Sawada, and K. Suzuki.2006. Construction of physical map and mapping of chromosomal virulence genes of the biovar 3 Agrobacterium (Rhizobium vitis) strain K-Ag-1. Genes Genet. Syst.81:373-380.

  28. Uraji, M., K. Suzuki, and K. Yoshida.2002. A novel plasmid curing method using incompatibility of plant pathogenic Ti plasmids in Agrobacterium tumefaciens. Genes Genet. Syst.77:1-9.

  29. Urbanczyk, H., K. Suzuki, K. Yoshida, and K. Kondo.2003. Physical and gene maps of Agrobacterium biovar 2 strains and their relationship to biovar 1 chromosomes. Microbiology149:3035-3042.

  30. Velikov, V. A., and Y. I. Buryanov.1998. Formation of deletional derivatives of the Ti-plasmid pGV3850 in a conjugative transfer from Agrobacterium tumefaciens to Escherichia coli. Genetica34:1056-1062.

  31. Wood, D. W., J. C. Setubal, R. Kaul, D. E. Monks, J. P. Kitajima, et al.2001. The genome of the natural genetic engineer Agrobacterium tumefaciens C58. Science294:2317-2323.

  32. Yamamoto, S., M. Uraji, K. Tanaka, K. Moriguchi, and K. Suzuki.2007. Identification of pTi-SAKURA DNA region conferring enhancement of plasmid incompatibility and stability. Genes Genet. Syst.82:197-206.

  33. Young, J. M., L. D. Kuykendall, E. Martínez-Romero, A. Kerr, and H. Sawada.2001. A revision of Rhizobium Frank 1889, with an emended description of the genus, and the inclusion of all species of Agrobacterium Conn 1942 and Allorhizobium undicola de Lajjudie et al. 1998 as new combinations: Rhizobium radiobacter, R. rhizogenes, R. rubi, R. undicola and R. vitis. Int. J. Syst. Evol. Microbiol.51:89-103.

  34. Yukawa, K., H. Kaku, H. Tanaka, Y. Koga-Ban, and M. Fukuda.2007. Characterization and host range determination of soybean super virulent Agrobacterium tumefaciens KAT23. Biosci. Biotechnol. Biochem.71:1676-1682.

  35. Zambryski, P., J. Tempe, and J. Schell.1989. Transfer and function of T-DNA genes from Agrobacterium Ti and Ri plasmids in plants. Cell56:193-201.

  • American Society for Microbiology

0 thoughts on “Research Paper On Ti Plasmid And Topo

Leave a Reply

Your email address will not be published. Required fields are marked *