Overdrive, a T-DNA transmission enhancer on the A. tumefaciens tumour-inducing plasmid

During crown gall tumorigenesis a specific segment of the Agrobacterium tumefaciens tumour-inducing (Ti) plasmid, the T-DNA, integrates into plant nuclear DNA. Similar 23-bp direct repeats at each end of the T region signal T-DNA borders, and T-DNA transmission (transfer and integration) requires the right-hand direct repeat. A chemically synthesized right border repeat in its wild-type orientation promotes T-DNA transmission at a low frequency; Ti plasmid sequences which normally flank the right repeat greatly stimulate the process. To identify flanking sequences required for full right border activity, we tested the activity of a border repeat surrounded by different amounts of normal flanking sequences. Efficient T-DNA transmission required a conserved sequence (5' TAAPuTPy-CTGTPuT-TGTTTGTTTG 3') which lies to the right of the two known right border repeats. In either orientation, a synthetic oligonucleotide containing this conserved sequence greatly stimulated the activity of a right border repeat, and a deletion removing 15 bp from the right end of this sequence destroyed it stimulatory effect. Thus, wild-type T-DNA transmission required both the 23-bp right border repeat and a conserved flanking sequence which we call overdrive.     

Virulence genes, borders, and overdrive generate single-stranded T-DNA molecules from the A6 Ti plasmid of Agrobacterium tumefaciens.

Agrobacterium tumefaciens transfers the T-DNA portion of its Ti plasmid to the nuclear genome of plant cells. Upon cocultivation of A. tumefaciens A348 with regenerating tobacco leaf protoplasts, six distinct single-stranded T-DNA molecules (T strands) were generated in addition to double-stranded T-DNA border cleavages which we have previously reported (K. Veluthambi, R.K. Jayaswal, and S.B. Gelvin, Proc. Natl. Acad. Sci. USA 84:1881-1885, 1987). The T region of an octopine-type Ti plasmid has four border repeats delimiting three T-DNA regions defined as T left (TL), T center (TC), and T right (TR). The six T strands generated upon induction corresponded to the TL, TC, TR, TL + TC, TC + TR, and TL + TC + TR regions, suggesting that the initiation and termination of T-strand synthesis can occur at each of the four borders. Most TL + TC + TR T-strand molecules corresponded to the top T-DNA strand, whereas the other five T strands corresponded to the bottom T-DNA strand. Generation of T strands required the virA, virG, and virD operons. Extra copies of vir genes, harbored on cosmids within derivatives of A. tumefaciens A348, enhanced production of T strands. The presence of right and left border repeats in their native orientation is important for the generation of full-length T strands. When a right border repeat was placed in the opposite orientation, single-stranded T-DNA molecules that corresponded to the top strand were generated. Deletion of overdrive, a sequence that flanks right border repeats and functions as a T-DNA transmission enhancer, reduced the level of T-strand generation. Induction of A. tumefaciens cells by regenerating tobacco protoplasts increased the copy number of the Ti plasmid relative to the bacterial chromosome.     

Bidirectional transfer from a 24 bp border repeat of Agrobacterium tumefaciens.

T-region transfer from wild-type Agrobacterium strains is thought to be an orientated process, starting at the right border repeat and terminating at the left border repeat of the T-region. Here we demonstrate that a right border repeat in the inverted orientation relative to the onc-genes can also mediate transfer of the T-region to the plant cell, although with lower efficiency as a border repeat in the native orientation. Transfer mediated by an inverted right border repeat is stimulated by the presence of the T-region transfer enhancer. Similar single stranded molecules, comprising the bottom strand of the T-DNA, were isolated from acetosyringone induced bacteria, irrespective of the orientation of the right border. These findings show that border repeats work bidirectionally to some extent.     

Functional domains of Agrobacterium tumefaciens single-stranded DNA-binding protein VirE2.

The transferred DNA (T-DNA) portion of the Agrobacterium tumefaciens tumor-inducing (Ti) plasmid enters infected plant cells and integrates into plant nuclear DNA. Direct repeats define the T-DNA ends; transfer begins when the VirD2 endonuclease produces a site-specific nick in the right-hand border repeat and attaches to the 5' end of the nicked strand. Subsequent events liberate the lower strand of the T-DNA from the Ti plasmid, producing single-stranded DNA molecules (T strands) that are covalently linked to VirD2 at their 5' ends. A. tumefaciens appears to transfer T-DNA into plant cells as a T-strand-VirD2 complex. The bacterium also transports VirE2, a cooperative single-stranded DNA-binding protein, into plant cells during infection. Both VirD2 and VirE2 contain nuclear localization signals that may direct these proteins, and bound T strands, into plant nuclei. Here we report the locations of functional regions of VirE2 identified by eight insertions of XhoI linker oligonucleotides, and one deletion mutation, throughout virE2. We examined the effects of these mutations on virulence, single-stranded DNA (ssDNA) binding, and accumulation of VirE2 in A. tumefaciens. Two of the mutations in the C-terminal half of VirE2 eliminated ssDNA binding, whereas two insertions in the N-terminal half altered cooperativity. Four of the mutations, distributed throughout virE2, decreased the stability of VirE2 in A. tumefaciens. In addition, we isolated a mutation in the central region of VirE2 that decreased tumorigenicity but did not affect ssDNA binding or VirE2 accumulation. This mutation may affect export of VirE2 into plant cells or nuclear localization of VirE2, or it may affect an uncharacterized activity of VirE2.     

The DNA sequences of T-DNA junctions suggest that complex T-DNA loci are formed by a recombination process resembling T-DNA integration

After Agrobacterium-mediated plant transformation, multiple T-DNAs frequently integrate at the same position in the plant genome, resulting in the formation of inverted and direct repeats. Because these inverted repeats cannot be amplified and analyzed by PCR, Arabidopsis root cells were co-transformed with two different T-DNAs with distinct sequences adjacent to the T-DNA borders. Nine direct or inverted T-DNA border junctions were analyzed at the sequence level. Precise end-to-end fusions were found between two right border ends, whereas imprecise fusions and filler DNA were present in T-DNA linkages containing a left border end. The results suggest that end-to-end ligation of double-stranded T-DNAs occurs especially between right T-DNA ends and that illegitimate recombination on the basis of microhomology, deletions, repair activities and insertions of filler DNA is involved in the formation of left border T-DNA junctions. Therefore, a similar illegitimate recombination mechanism is proposed that is involved in the formation of complex T-DNA inserts as well as in the integration of the T-DNA in the plant genome.     

Right 25 bp terminus sequence of the nopaline T-DNA is essential for and determines direction of DNA transfer from agrobacterium to the plant genome

We have determined which sequences at the right border of the T-DNA region of the nopaline C58 Ti plasmid are required for transfer and/or integration of the T-DNA into the plant cell genome. The results indicate that the 25 bp T-DNA terminus repeat sequence, TGACAGGATATATTGGCGGGTAAAC, is directly responsible for T-DNA transfer; furthermore, this sequence is directional in its mode of action. A transfer-negative nononcogenic Ti plasmid derivative, pGV3852, was constructed, in which 3 kb covering the right T-DNA border region was substituted for by pBR322 sequences. The pBR322 sequences in pGV3852 provide a site for homologous recombination with pBR-derived plasmids containing sequences to assay for transfer activity. First, a 3.3 kb restriction fragment overlapping the deleted region in pGV3852 was shown to restore transfer activity. Second, a sequence of only 25 bp, the T-DNA terminus sequence, was shown to be sufficient to restore normal transfer activity. The transfer-promoting sequences are most active when reinserted in one orientation, that normally found in the Ti plasmid.     

Genetic analysis of integration mediated by single T-DNA borders.

Transformation of plant cells by the T-DNA of the Ti plasmid of Agrobacterium tumefaciens depends in part upon a sequence adjacent to the right T-DNA end. When this sequence is absent, the T-DNA is almost avirulent; when it is present, DNA between it and the left T-DNA border region becomes integrated in plants. To investigate further this process of DNA transfer and integration, we introduced the right border region and the nopaline synthase (nos) gene of plasmid pTiC58 into a variety of new positions around Ti plasmids. The border region functioned when separated from the remainder of the T-DNA by almost 50 kilobases. It also worked when placed outside of the T-DNA region where there were no known left-border sequences with which to interact. Indeed, the nos gene could be transferred to plants even when no other Ti plasmid sequences were present on the same plasmid. These results may indicate that the sequence requirements for the left borders are not as stringent as those for the right borders. In addition, mutants with an extra copy of the right border region within their T-DNA were found to transfer or integrate only parts of the bacterial T-DNA region. It is possible that abnormally placed T-DNA borders interfere with the normal process of DNA transfer, integration, or both.     

Direct repeats of T-DNA integrated in tobacco chromosome: characterization of junction regions

Plant transformation via Agrobacterium frequently results in formation of multiple copy T-DNA arrays at one target site of the chromosome. The T-DNA copies are arranged in repeats, direct or inverted around one of the T-DNA borders. A Ti plasmid-derived transformation vector has been constructed enabling direct selection of transformants carrying at least two linked copies of T-DNA in the same orientation. The selection is based on expression of a promoterless neomycin phosphotransferase gene on one T-DNA copy from a promoter located on the other T-DNA copy. After co-cultivation of tobacco protoplasts with Agrobacterium, as many as 30% of regenerated transformed plants carried directly repeated T-DNA copies. The junction regions between two T-DNAs were amplified and 13 amplified fragments were cloned and sequenced. The involvement of T-DNA left and right border sequences in direct repeat junctions was determined. In some junctions, additional filler DNA was detected. The length of filler DNA varied from a few up to almost 300 bp. The longer filler DNAs from two clones were found to be T-DNA fragments in direct or reverse orientation. We discuss the recently suggested models for T-DNA integration and propose that the formation of direct repeats in genomes does not necessarily result from ligation of intermediates (i.e. T-strands), but more likely from the co-integration of several intermediates into one target site.     

vir-induced recombination in Agrobacterium. Physical characterization of precise and imprecise T-circle formation

Induction of Ti plasmid virulence (vir) gene expression during the early stages of plant cell transformation by Agrobacterium tumefaciens initiates the generation of several T-DNA-associated molecular events: (1) site-specific nicks at T-DNA border sequences (border nicks); (2) free, unipolar, linear, single-stranded T-DNA copies (T-strands); and (3) double-stranded, circular T-DNA molecules (T-circles). The first two T-DNA products have been detected in A. tumefaciens, while T-circles have only been detected following Escherichia coli transformation or transduction. The relationship between the three events has not been evaluated since the genesis of T-circles in A. tumefaciens has not been clarified. Evidence is presented here that T-circles are not an artefact of E. coli transformation, but are present as free, double-stranded molecules in A. tumefaciens resulting from site-specific reciprocal recombination between the left and right 25-base-pair border sequences that flank the T-DNA. Furthermore, the frequency of T-circle formation correlates with the frequency of formation of its reciprocal product, the Ti plasmid deleted in the T-DNA region. Several types of recombinant T-DNA circles arise after activation of vir gene expression, a major class representing precise site-specific recombination between both T-DNA borders, and a minor class representing recombination events either utilizing only one T-DNA border sequence and other Ti plasmid sequences, or utilizing only Ti plasmid sequences (i.e. no T-DNA borders). Nucleotide sequence analyses show that when one (nicked) border recombines with other Ti plasmid sequences, a small stretch (16 to 17 base-pairs) of local homology suffices to allow crossing over.     

Trans-kingdom T-DNA transfer from Agrobacterium tumefaciens to Saccharomyces cerevisiae.

Agrobacterium tumefaciens transfers part of its tumour-inducing (Ti) plasmid, the transferred or T-DNA, to plants during tumourigenesis. This represents the only example of naturally occurring trans-kingdom transfer of genetic material. Here we report that A.tumefaciens can transfer its T-DNA not only to plant cells, but also to another eukaryote, namely the yeast Saccharomyces cerevisiae. The Ti plasmid virulence (vir) genes that mediate T-DNA transfer to plants were found to be essential for transfer to yeast as well. Transgenic S.cerevisiae strains were analysed for their T-DNA content. Results showed that T-DNA circles were formed in yeast with precise fusions between the left and right borders. Such T-DNA circles were stably maintained by the yeast if the replicator from the yeast 2 mu plasmid was present in the T-DNA. Integration of T-DNA in the S.cerevisiae genome was found to occur via homologous recombination. This contrasts with integration in the plant genome, where T-DNA integrates preferentially via illegitimate recombination. Our results thus suggest that the process of T-DNA integration is predominantly determined by host factors.     

Genetic studies on the role of octopine T-DNA border regions in crown gall tumor formation

Summary Crown gall tumors result from transfer and integration of the T-DNA from the Ti plasmid of Agrobacterium tumefaciens into plant nuclear DNA. In the present study, recombinant plasmids containing deletion and rearrangement deriviatives of the T-DNA region of the octopine Ti plasmid pTiA6 were tested in a binary tumorigenesis system (Hoekema et al. 1983) to determine the requirements for T-DNA border regions in tumor formation. Since two defined segments of the T-DNA region of octopine Ti plasmids can be detected in tumor DNA (the left (T_L-) and right (T_R-) DNA), four border regions exist in this Ti plasmid. Agrobacteria harboring plasmid constructs which contain a T-DNA gene capable of inciting tumors (gene 4, the tmr gene, which is involved in cytokinin biosynthesis) and various T-DNA border regions were tested for ability to cause tumors on Nicotiana glauca and other host plants. Such tmr constructs containing as their only border region the right border of either the T_L-DNA or the T_R-DNA are fully tumorigenic. Analogous tmr constructs containing only the T_L-DNa left border region are not tumorigenic. These results do not depend on the orientation or position of the single border with respect to the tmr gene; furthermore, the T_R-DNA right border can confer tumor-forming ability despite the presence of an intervening copy of the T_L-DNA left border.These results for relatively small plasmids are contrasted with previously determined requirements for border regions in tumorigenesis by intact Ti plasmids. A model previously proposed by Wang et al. (1984) for the role of border regions in DNA transfer to plant cells is extended in order to explain the tumor-forming ability of plasmid constructs containing a single border region. The results of this study interpreted according to the model suggest that the octopine T_L-DNA left border is defective in this DNA-transfer process.     

Function of heterologous and pseudo border repeats in T region transfer via the octopine virulence system of Agrobacterium tumefaciens

The successful transfer of the Ti plasmid T region to the plant cell is mediated by its 24 bp border repeats. Processing of the T-region prior to transfer to the plant cell is started at the right border repeat and is stimulated by a transfer enhancer sequence called overdrive. Left and right border repeats differ somewhat in nucleotide sequence; moreover, the repeats of different Ti and Ri plasmids are slightly different. Our data indicate that these differences do not have a significant influence on border activity. However, the overdrive sequence is essential for the efficient transfer of a T region via an octopine transfer system. Our data suggest that an overdrive sequence must also be present next to the right border repeats of the nopaline Ti plasmid and the agropine of octopine and nopaline Ti plasmids express some differences in T-DNA processing activities. of cotopine and nopaline Ti plasmids express some differences in T-DNA processing activities.Furthermore, we demonstrate that certain pseudo border repeats, sequences that resemble the native 24 bp border repeat and naturally occur within the octopine Ti plasmid T-region, are able to mediate T region transfer to the plant cell, albeit with much reduced efficiency as compared to wild-type border repeats.     

Sequence context of the T-DNA border repeat element determines its relative activity during T-DNA transfer to plant cells

Summary We present a detailed analysis of the function of the right and left T-DNA border regions of the nopaline Ti plasmid of Agrobacterium tumefaciens. An avirulent deletion of the right border of the nopaline Ti plasmid (pGV3852) was used as an acceptor for 14 different T-DNA border constructs. The functional activities of these constructs were assayed by their ability to restore virulence, i.e. transformation on inoculated plants. Tumorigenicities were measured in several independent experiments over a 2 year period and the statistical significance of their relative levels was evaluated. The data indicate: (i) the entire sequence of the 25 bp direct repeat of the T-DNA is required to provide an efficient substrate for mediating T-DNA transfer events; deletion derivatives of either the conserved or the vaiable domain of the repeat are defective in T-DNA transfer; (ii) while the 25 bp direct repeat alone can promote the T-DNA transfer, the flanking sequences of the repeats enhance (on the right) or attenuate (on the left) their activity; and (iii) tumorigenicity measurements vary depending on the plant assay system: potato discs are more sensitive than wounded tobacco leaves in detecting differences in T-DNA border activity.     

The interaction of Agrobacterium Ti-plasmid DNA and plant cells

The tumour-inducing plasmids of Agrobacterium tumefaciens (Ti-plasmids) reveal several interesting properties. They are catabolic plasmids, which, instead of rendering Agrobacterium strains capable of catabolizing compounds found in Nature, force a plant to synthesize these catabolites (denoted 'opines'). This situation is obtained by insertion of a segment of the Ti-plasmid (the T-DNA) into the plant nucleus, where T-DNA genes become expressed and intervene in the biosynthesis of these opines. Cells containing the T-DNA behave as neoplasms (crown gall cells). Southern blotting shows that the insertion process responsible for T-DNA transfer probably recognizes special sequences on the T-DNA since the length of the T-DNA segment observed in different, independently isolated tumour lines was found to be similar. For the nopaline Ti-plasmids both left-hand and right-hand borders were found to be constant. For the octopine plasmid the left border was constant and at least two classes of right-hand borders were found. Upon redifferentiation of the transformed plant cells, the T-DNA was found to be conserved in all somatic cells examined. However, small deletions at the border fragments of the T-DNA have been observed. The exact arrangement and copy number of the T-DNA in a nucleus is still under study, but genomic cloning has already revealed that an interspersed tandem arrangement is dominant in nopaline tumours. Clones containing both the right border of one T-DNA and the left border of the neighbouring tandem T-DNA were isolated. In order to identify the different T-plasmid encoded functions an extensive use was made of transposon insertion mutagenesis. When an antibiotic resistance transposon was inserted into the non-essential regions of the T-DNA, a linked transfer to the plant DNA of the transposon together with the T-DNA was observed. This indicates that Ti-plasmids are possible vectors for genetic engineering in plants. A strategy is described for insertion of any cloned DNA segment into the T-DNA.     

Transgene repeats in aspen: molecular characterisation suggests simultaneous integration of independent T-DNAs into receptive hotspots in the host genome

Rearrangements of T-DNAs during genetic transformation of plants can result in the insertion of transgenes in the form of repeats into the host genome and frequently lead to loss of transgene expression. To obtain insight into the mechanism of repeat formation we screened 45 transgenic lines of aspen and hybrid aspen transformed with six different gene constructs. The frequency of T-DNA repeat formation among randomly screened transgenic lines was found to be about 21%. In ten transgenic lines direct repeats were detected. An inverted repeat was found in one other transgenic line. Sequencing of the junctions between the T-DNA inserts revealed identical residual right-border repeat sequences at the repeat junctions in all ten transgenic lines that had direct repeats. Formation of "precise" junctions based on short regions of sequence similarity between recombining strands was observed in three transgenic lines transformed with the same plasmid. Additional DNA sequences termed filler DNAs were found to be inserted between the T-DNA repeats at eight junctions where there was no similarity between recombining ends. The length of the filler DNAs varied from 4 to almost 300 bp. Small filler DNAs--a few base pairs long--were in most cases copied from T-DNA near the break points. The large filler sequences of about 300 bp in two transgenic lines were found to be of host plant origin, suggesting that transgene repeat formation occurred as a result of the simultaneous invasion of a receptive site in the host genome by two independent T-DNA strands. On the basis of the results obtained, and in the light of previous reports on T-DNA/plant DNA junctions in aspen and other crop plants, a mechanistic model for transgene rearrangement and filler formation is suggested.     

Transfer of the Ti plasmid from Agrobacterium tumefaciens into Escherichia coli cells

We have screened strains of Agrobacterium tumefaciens for spontaneous mutants showing constitutive transfer of the nopaline Ti plasmid pTiC58 during conjugation. The Ti plasmid derivatives obtained could be transferred not only to A. tumefaciens but also to E. coli cells. The Ti plasmid cannot survive as a freely replicating plasmid in E. coli, but it can occasionally integrate into the E. coli chromosome. However, insertion in tandem of plasmids carrying fd replication origins (pfd plasmids) into the T-DNA provides an indicator for all transfer events into E. coli cells, providing fd gene 2 protein is present in these cells. This viral protein causes the excision of one copy of the pfd plasmid and allows its propagation in the host cell. By using this specially designed Ti plasmid, which was also made constitutive in transfer functions, we found plasmid exchange among A. tumefaciens strains and between A. tumefaciens and E. coli cells to be equally efficient. A Ti plasmid with repressed transfer functions was transferred to E. coli with a rate similar to the low frequency at which it was transferred to A. tumefaciens. The expression of transfer functions of plasmid RP4 either in A. tumefaciens or in E. coli did not increase the transfer of the Ti plasmid into E. coli cells, nor did the addition of acetosyringone, an inducer of T-DNA transfer to plant cells. The results show that A. tumefaciens can transfer the Ti plasmid to E. coli with the same efficiency as within its own species. Conjugational transmission of extrachromosomal DNA like the narrow-host-range Ti plasmid may often not only occur among partners allowing propagation of the plasmid, but also on a 'try-all' basis including hosts which do not replicate the transferred DNA.     

Right-hand border regions of octopine T-DNA are recognized by RNA polymerase of Agrobacterium as well as by VirD1 and VirD2 proteins.

The T-DNA of octopine Ti plasmid of Agrobacterium tumefaciens contains TL- and TR-DNA regions each bounded by 25 base-pair-repeats (designated A, B, C and D from left to right). Short DNA segments containing the borders B, C and D were found to function as promoter when placed in the rightward orientation upstream of promoter-less lacZ. Promoter consensus sequence of Agrobacterium were found within these border repeats and in their adjacent regions. The expression of lacZ was low when the segments contained the overdrive, a sequence known to enhance T-DNA transfer. Simultaneous overproduction of VirD1 and D2 proteins, endonuclease acting on the border repeats, interfered with the promoter functions of the border segments. In spite of their activity under these conditions, the border regions do not seem to be involved in the gene expression, because they are not followed by appropriate open reading frames. We propose that RNA polymerase of Agrobacterium competes with VirD products for T-DNA borders and thereby affects the transfer of T-DNA.     

Activity of T-DNA borders in plant cell transformation by mini-T plasmids.

By using a binary vector system, we examined the requirements for border sequences in T-DNA transformation of plant genomes. Mini-T plasmids consisting of small replicons with different extents of pTiT37 T-DNA were tested for plant tumor-inducing ability in Agrobacterium tumefaciens strain LBA4404 containing helper plasmid pAL4404 (which encodes virulence genes needed for T-DNA transfer). Assays of these bacteria on carrot disks, Kalanchoë leaves, and SR1 Nicotiana tabacum plantlets showed that mini-T plasmid containing full length T-DNA including left and right borders was highly virulent, as were mini-T plasmids containing all onc (oncogenicity) genes and only the right border. In contrast, mini-T plasmids containing all onc genes and only the left border induced tumors only rarely, and a mini-T plasmid containing all onc genes but no T-DNA borders was completely avirulent. Southern hybridization analyses of tumor DNA showed that T-DNA border sequences delimited the extent of the two-border mini-T plasmid transferred and integrated into the plant genome. When only one T-DNA border was present, it formed one end of the transferred DNA, and the other end mapped in the vector sequences. The implications of these results for the mechanism of T-DNA transfer and integration are discussed.