The effect of bacterial isochorismate synthase on the Brassica rapa metabolome
Simoh, S.
Citation
Simoh, S. (2008, June 11). The effect of bacterial isochorismate synthase on the Brassica rapa metabolome. Retrieved from https://hdl.handle.net/1887/12944
Version: Corrected Publisher’s Version
License: Licence agreement concerning inclusion of doctoral thesis in the Institutional Repository of the University of Leiden
Downloaded from: https://hdl.handle.net/1887/12944
Note: To cite this publication please use the final published version (if applicable).
Chapter 2
Host-bacterium interactions in Agrobacterium tumefaciens-mediated plant transformation:
Mechanism of action and Agrobacterium/plant factors involved
Published in Current Topics in Plant Biology (2007) 8: 1-20 Sanimah Simoh1,2, Huub JM Linthorst3, Robert Verpoorte1
1 Section Metabolomics, Institute of Biology, Leiden University, Leiden, The Netherlands
2 Biotechnology Research Centre, Malaysian Agricultural Research & Development Institute (MARDI), Kuala Lumpur, Malaysia
3 Section Plant Cell Physiology,Institute of Biology, Leiden University, Leiden, The Netherlands
Abstract
Agrobacterium–mediated plant transformation has been widely used to transfer genes of interest in many varieties of crops to obtain specific desired traits. Most of the molecular mechanisms that underlie the transformation steps have been well elucidated over the years. However, a few steps, such as nuclear targeting, T-DNA integration, and Agrobacterium-plant proteins involved remain largely obscure and are still under extensive studies. This review describes the major steps involved in the molecular mechanism of Agrobacterium-mediated transformation and provides insight in the recent developments in studies on the Agrobacterium/plant transformation system. Some factors affecting the transformation efficiency are also briefly discussed
Keywords: Agrobacterium tumefaciens, molecular mechanisms, T-DNA integration, plant factors, transformation
2.1 Introduction
Agrobacterium tumefaciens and A. rhizogenes are bacterial pathogens that possess the ability to transfer part of their genetic material to plants by invading the host plant and stably integrating part of their DNA into the plant genome. Thus, the genus of Agrobacterium has been used as a major tool for plant genetic transformation. Both bacteria share many similarities, including the mechanism by which they transfer the genetic material. One distinct feature of the infection relies on the presence of the Ti plasmid in A. tumefaciens, which induces formation of tumors on plant stems, and the Ri plasmid in A. rhizogenes, which is responsible for hairy root formation. Due to its broad host range and ability to infect many plant tissues, A. tumefaciens has become a primary option for plant transformation experiments.
Agrobacterium tumefaciens is a soil pathogen, a gram negative bacterium which infects many species of plants causing a disease known as “crown gall”. The existence of the disease in plants was first discovered by Smith and Townsend in 1907. To date, the Agrobacterium-plant cell interaction is the only known natural occurrence of inter- kingdom DNA transfer (Veluthambi et al., 2003). However, recent discovery by Hotopp et al. (2007) showed the evidence of gene transfer from Wolbachia bacterial DNA into host insect and nematode cells. The oncogenic activity of Agrobacterium is due to the presence of a large (200-kb) tumor inducing plasmid (pTi) and tumor formation results from its infection into the host plant. Upon infection of the wounded plant, the bacterium transfers a small segment of DNA, the so called T-DNA from the Ti plasmid into the plant cell. The T-DNA is then translocated into the plant nucleus and stably integrated into the chromosome (Zupan et al., 2000; Tzfira and Citovsky, 2002; Kohli, 2003). This mechanism is regulated by several virulence genes (vir) that are present in the Ti plasmid (Erkhard, 2004) and leads to the expression of the introduced gene in the host cell (Gelvin, 1998). The wild type T-DNA encodes a specific set of oncogenic (onc) genes that when it is expressed in the host cell, leads to the formation of a tumor. The tumor contains enzymes for synthesizing amino acids and sugar conjugates called opines which serve as a source of carbon and nitrogen for Agrobacterium. Thus, the Agrobacterium strains are usually classified according to the type of opine gene present on their T- DNA : nopaline, octopine, agropine or succinamopine types (Dessaux et al.,
1992). Nopaline and octopine strains are the most common (Hooykaas and Beijersbergen, 1994). For the purpose of plant transformation, the tumorous growth activity caused by Agrobacterium is prevented by deleting the oncogenes or making the genes nonfunctional by interrupting their sequence in such a way that the transformation is still effective without developing the disease (Gartland, 1995). This process is called disarming.
Replacing the tumor encoding region by DNA of interest does not affect the transfer of T-DNA into the host plant. This opened the way for Agrobacterium as the tool to produce transgenic plants that can express specific genes of interest. Figure 2.1 illustrates some of the structure of opines.
Nowadays, the use of Agrobacterium’s Ti plasmid as a plant vector for introduction of foreign genes is a well established technology. Because of Ti plasmid’s large size (about 200 kbp), direct genetic manipulation in vitro is difficult. To overcome this problem, scientists developed alternative strategies by designing several types of vectors. The most successful is the development of the binary vector system (Hoekema et al., 1983), which has been commonly used for plant genetic engineering. This strategy was based on the fact that two main components for successful Agrobacterium gene transfer, the T-DNA region and the vir genes, can reside on separate plasmids (Hoekema et al., 1983; Hellen and Mullineaux, 2000). In this system, the vir gene functions are provided by a
‘disarmed’ Ti plasmid in which the T-DNA with all the onc genes and its left (LB) and right (RB) border has been deleted, leaving the vir region and other plasmid parts intact.
A second plasmid, which is able to replicate and be easily manipulated in both Escherichia coli and Agrobacterium, is smaller and carries a modified T-DNA between LB and RB flanking regions containing the genes to be transferred. Typically, this plasmid has a broad host range origin of replication (ori), a bacterial gene for antibiotic resistance which serves for the selection of transformed plant cells, a marker for selection and maintenance in both E. coli and Agrobacterium, and multiple restriction endonuclease sites for insertion of additional genes in the T-DNA region. The plasmid harboring the vir gene function is termed vir helper while the plasmid harboring the T- region is termed binary vector. When these two vectors, which remain as separate entities, are present within the same Agrobacterium cell, the vir genes that reside on the helper vector can function in trans on the T-region of the binary vector to promote T-
DNA processing and transfer to the recipient plant cell (Hoekema et al., 1983).
The
Figure 2.1 Some of the opines structures
Figure 2.2 Binary Ti Plasmid: The T-DNA region from the plasmid carrying the virulence region is removed and the left and right border of the T-DNA region is placed into another plasmid.
HOOC CH CH2 CH2 CH2 NH C
NH2
NH NH
CH
HOOC CH2 CH2 COOH
HOOC CH CH2 CH2 CH2 NH C
NH2
NH NH
CH
H3C COOH
HOOC CH CH2 C NH
CH
HOOC CH2
NH2 O
CH2 COOH
Right Border
ORI of E.coli
ORI of Agrobacterium
Bacterial Selectable Marker Left Border
Plant Selectabel Marker
Gene of Interest
Binary Vector
239 bp
Agrobacterium chromosomal DNA
ORI of Agrobacterium
ORI of E.coli Plant Selectable
Marker
Octopine
Succinamopine Nopaline
development of this basic binary vector system has prompted other researchers to construct T-DNA containing plasmid vectors with additional features to enhance transformation efficiency. The use of these vectors for plant transformation excluded the requirement for homologous recombination in plasmid T-DNA, thus simplifying the task for introducing and analyzing the recombinant T-DNA in Agrobacterium (McBride and Summerfelt, 1990). Figure 2.2 schematically illustrates the binary vector system.
The unique ability of the bacterium to transfer the T-DNA from the Ti plasmid into the plant nuclear genome has made Agrobacterium-mediated transformation widely adopted for generating transgenic plants. The use of Agrobacterium for plant transformation was first reported by Hoekema et al. (1983) although the molecular mechanisms of the process were unknown at that moment. Since then, Agrobacterium transformation technology has been greatly improved for transferring foreign genes encoding the desired traits to a wide variety of commercially important crops. “Flavr Savr” tomato with prolonged shelf life was the first commercial application of this technology introduced in the US market in 1994.
Agrobacterium tumefaciens is a broad host range plant pathogenic bacterium for dicotyledonous plants (Escobar and Dandekar, 2003). The bacterium has been found to infect around 600 species in 90 families of dicotyledonous plants making it the main method of choice for plant transformation for dicots. Despite this success, many plant species or genotypes, especially monocots, remain highly recalcitrant to Agrobacterium- mediated transformation (Chateau, 2000). For example, it was only in 1994 that scientists finally achieved the transformation of rice by Agrobacterium-mediated gene transfer (Hiei et al., 1994). The mechanism of non-host resistance in certain recalcitrant species and monocots to A. tumefaciens is still poorly understood. Hooykaas-Van Slogteren et al.
(1984) reported for the first time the occurrence of tumor formation and the evidence of T-DNA transfer and its expression in two monocot species, from the Liliaceae and Amaryllidaceae after infection with octopine and nopaline strains. As this T-DNA transfer occurred from wild type Ti plasmid but not from a plasmid that is defective in one of its vir genes, the authors suggested that the introduction of T-DNA by A.
tumefaciens into monocots might follow the similar mechanism as in dicots. In recent years, a remarkable progress of transformation using A. tumefaciens in recalcitrant and
monocot species has been observed (Sparrow, 2004; Bliss et al., 1999). It was also discovered that Agrobacterium has the ability to transform non plant or organisms.
Research by several groups has demonstrated the successful T-DNA transfer and its integration into filamentous fungi, Aspergillus awamori (de Groot et al., 1998) and yeast, Saccharomyces cerevisiae (Bundock et al., 1995). Development of new bacterial strains, vectors and Ti plasmids to enhance the transformation efficiency and also the increasing knowledge of transformation mechanisms, may lead to a new era of transformation technology, especially for non-plant host range and species that are recalcitrant to transformation.
Agrobacterium-based transformation relies on the susceptibility of the host plant to infection, so the T-DNA can be successfully integrated into the host genome. If this mechanism is possible, the selected genotype can be transformed. The differences in susceptibility to infection may be attributed to environmental and physiological factors as well as to the type of tissues (Gelvin, 2000). A second bottleneck for obtaining transformed plants is the possibility for the initially transformed cells to regenerate into viable plants. Agrobacterium-mediated plant transformation has been most successful when the plants are amenable to tissue culture and regeneration. However, an adequate number of regenerable cells accessible to gene transfer treatment, that can retain the regeneration capacity throughout the whole process of transformation until the stable transgenic plant is obtained is a crucial factor that must first be achieved (Birch, 1997). In addition, an efficient plant regeneration system from explants is dependent on the genotype used together with the specific combination and concentration of basal medium and plant growth regulators in the culture medium (Kumar and Rajam, 2005). Several factors such as the type of explants, the concentration of A. tumefaciens, the co- cultivation period, the selectable marker gene and the type of vector used are also prerequisites for an efficient plant transformation (Hiei et al., 1997).
Undoubtedly, the ability of A. tumefaciens to transfer a segment of its DNA to the plant nuclear genome provides a powerful tool for plant biotechnology and therefore A.
tumefaciens mediated gene transfer is currently the most used method for plant transformation. Agrobacterium-mediated transformation technology has unique advantages, such as the straightforwardness of the gene transfer system, the resulting low
copy number of transgenes integrated into the host genome (Gelvin, 1998), high frequency of stable transformations and low incidence of gene silencing (Veluthambi et al., 2003), as opposed to any other method of transformation. In addition, large pieces of DNA are able to be transferred with minimal chances of rearrangements (Roy et al., 2000) and the introduced genes are usually transmitted to next generations in a Mendelian manner (Rhodora and Thomas, 1996). Recent developments in the knowledge of the molecular biology of A. tumefaciens have made that most of the research has been focused on the improvement of factors within the bacterium itself. This includes for example, the introduction of multiple copies of different vir genes, modification of transformation vectors and optimization of the tissue culture conditions (Tzfira and Citovsky, 2002). Despite the extensive use of Agrobacterium as a transformation tool, the molecular mechanism of the gene transfer from Agrobacterium to the plant cell beginning with its attachment to the plant cell to the integration of the T-DNA into the plant genome is still largely unknown and its elucidation remains a challenge. The success of Agrobacterium-mediated plant transformation also depends on the presence of several proteins encoded by genes of the host plant. Thus, detailed knowledge of every step involved in this complex mechanism, as well as the key role of the Agrobacterium/plant factors involved at the molecular level will likely facilitate the further improvement of this technology for future plant genetic engineering.
2.2 Mechanism of infection at molecular level
Three genetic elements of Agrobacterium are needed for plant transformation (Sheng and Citovski, 1996). The first element is the T-DNA flanked by two 25-bp imperfect direct repeats, also known as left border (LB) and right border (RB). The LB and RB are the only necessary sequences in T-DNA that play an important role for gene transfer to the plant cell. The sequences between the two borders can be replaced by specific genes of interest. The fact that any DNA inserted between the T-DNA borders will be transferred to the plant genome (Sheng and Citovsky, 1996) indicates that the T-DNA molecule itself does not confer the protein transport machinery from the bacterial cell to the host plant (Baron and Zambryski, 1996), but this role is fulfilled by the second essential element for transformation i.e. virulence (vir) genes, as well as their plant cellular partners.
The second element is the 35-kb virulence (vir) region which is located in the Ti plasmid but separated from the T-DNA (Stachel and Nester, 1986). This region consists of at least six major operons i.e. VirA, VirB, VirC, VirD, VirE, and VirG, which encode the virulence proteins of DNA transport. Twenty one genes in this region are essential for wild type tumorigenesis whereas other vir operons including VirD5, VirE3, VirF, VirH, VirJ, VirK, VirL, VirM, VirP and VirR are not essential for tumorigenesis in all host plants or play a role in a specific host plant (Zhu et al., 2000). The third element necessary for transformation are the chromosomal virulence (chv) genes which are located in the bacterial chromosome and are involved in recognition and early attachment of Agrobacterium to the plant cell (Zambry et al., 1989). It is those three elements that work in a synergistic fashion and enable the gene transfer by Agrobacterium. Intensive studies of several decades ago have revealed relatively detailed knowledge of the mechanism and function of chv and vir genes that mediate the gene transfer from the Agrobacterium to the host plant. However, the molecular elements participating during the transfer of the T-DNA complex and T-DNA integration into the host nuclear genome are still largely unknown (Jiang et al., 2003). In addition, little is known about the plant cell factors involved during the process of Agrobacterium-mediated transformation.
There are a few major steps that play vital roles for successful gene transfer from Agrobacterium into plant cells. Figure 2.3 illustrates the process of transformation from the beginning of virulence gene induction to the integration of T-DNA into the plant genome.
2.2.1 Chemotaxis and attachment of Agrobacterium to the host cell
Recognition of the Agrobacterium and its subsequent attachment to the targeted host plant is an initial step and a pre-requisite for the infection process. Before this step can be established, it requires wounding of the host plant which hypothetically could allow the infection process by activating specific receptor sites on the plant cell (Escudero and Hohn, 1997). The target plant’s response to the injury is to release phenolic compounds and/or sugars that attract the Agrobacterium to the wounded tissues. The movement of the bacterium in response to the chemical stimuli produced by wounded tissue is called chemotaxis which is a crucial step in the early events of Agrobacterium–plant cell
Att, chvA chvE, chvB, psc
T-DNA strand
Vir region (A,B,C,D,E,F,G)
Vir G
Vir A
phenolic inducer, chvG, chvI
Agrobacterium cell
3’
T-DNA
Ti plasmid
Plant cell
Figure 2.3 The passage of transformation process by Agrobacterium to the plant cell.
T-DNA complex
NUCLEUS VirE2/ D2 receptor
Integrated T-DNA Vir E2
D2
Wounding
Vir B channel
Nuclear Targeting CYTOPLASM
Agrobacterium chromosome
V
Vir D4
5’
Vir D2
Vir E2, Vir E1 Vir B
VirG-P
sugar, vitronectin-like protein, RAT1, RAT3, acidity chvE
D2
Vir B channel
chvD
interaction (Winans, 1992). Research conducted by several researchers showed that Agrobacterium contains a chemotaxis operon which resembles the chemotaxis operon commonly found in the α-subgroup of proteobacteria (Wright et al., 1998). Mutation of orf1 and cheA in this operon resulted in impaired chemotaxis of the Agrobacterium.
However, the function of the protein encoded in the operon and the mechanism of interaction with virulence genes leading to the chemotactic pathway has yet to be determined.
The second early step of infection involves the attachment of the Agrobacterium to the target plant cell. As the binding of Agrobacterium to the plant cell is saturable and other genera of bacteria cannot compete with these binding sites, it is thought that specific receptors may exist on Agrobacterium and the plant cell surface (Winans, 1992).
The Agrobacterium chromosomal genes i.e. chvA, chvB, pscA (exoC) and att (Matthysse, 1987; Tzfira and Citovsky, 2000) play a key role in the recognition process of these specific plant cell surface receptors. Agrobacterium mutants of these loci are unable to bind with plant cells resulting in a loss of virulence in many plant species (Ziemienowicz, 2001). In Agrobacterium, chvA is thought to encode a transporter of β-1,2-glucan while chvB encodes its synthesis enzyme (Peng et al., 2001). It is believed that the plant vitronectin-like protein is one of the receptors that may be involved in this attachment process (Wagner and Matthysse, 1992). In animals, vitronectin acts as a specific receptor for different pathogenic bacteria (Tzfira and Citovsky, 2002). This function is also supported by the fact that mutants of chvA, chvB, pscA (exoC) and att are unable to bind to plant cells and also show low binding activity to plant vitronectin (Wagner and Matthysse, 1992). Other plant proteins such as pectin acceptors (Ziemienowicz, 2001) and the rhicadesin binding protein (Swart et al., 1994) are also thought to play a role in the attachment process.
The attachment process involves two steps (Gelvin, 2000). The early step involves the synthesis of an acidic polysaccharide which requires the attR locus. Agrobacterium mutants on this locus were avirulent and lacked attachment activity (Reuhs et al., 1997).
This step is reversible, which means that the bacterium can be easily dislodged from the plant cell by sheer forces (Gelvin, 2000). The second step involves the formation of
cellulose (β-1,4-glucan) fibrils once the Agrobacterium is attached to the receptor on the plant cell surface (Sheng and Citovski, 1996). The formation of cellulose fibrils is thought to play a role in initial binding of Agrobacterium to the plant cell. The cellulose filaments also help in entrapping further Agrobacterium cells to form a cluster of bacterial aggregates that tightly bound to the cell wall and form a stable bacteria-plant cell connection (Hernandaz et al., 1999). It is thought that the cellulose fibril is also essential for T-DNA transfer, but its role in this mechanism remains unproven.
Agrobacterium mutants that cannot synthesize cellulose fibrils and cannot form the attachment still show virulence (Matthysse, 1987). Several recent studies showed that the plant cell surface protein genes, RAT1, encoding for arabinogalactine protein (AGP) and RAT3, encoding a small protein secreted to the apoplast may also be involved in the attachment process (Gelvin, 2000). Mutants in these loci showed an impaired attachment of Agrobacterium to plant cells.
2.2.2 Activation of Vir genes
The activation of vir genes is dependent on a two component signal transduction system of Agrobacterium i.e. consisting of 1) a membrane sensor protein, VirA and 2) a cytoplasmic transducer protein, VirG (Tzfira and Citovsky, 2000). The constitutively expressed VirA acts as a ‘sensor’ for plant-released phenolic compounds upon wounding of the plant host cell, and in turn activates the VirG product. The signal transduction pathways of further vir induction begins when VirA interacts with the phenolic signal in combination with a chromosomally encoded sugar-binding protein (ChvE), (Shimoda et al., 1993) and extracellular acidity resulting in VirA autophosphorylation at a specific histidine residue (His-474) of the protein. Since the phosphorylation at this residue is unstable due to the formation of a high energy phosphate bond (Pollard and Cooper, 1986), VirA then transfers the phosphate group to aspartate residue 52 in VirG where it is in an active and stable form. The transphosphorylated form of VirG then binds with a conserved 12-bp sequence of a vir box enhancer element located in the promoter region of vir genes (Sheng and Citovsky, 1996). This interaction leads to transcriptional activation of the VirB, VirC, VirD, and VirE gene in the vir operon (Escobar et al., 2003).
The phosphorylation process is an essential step in inducing vir gene activity as it is
thought to enhance the ability of VirG to recruit another protein involved in transcriptional activation (Tzfira and Chitovsky, 2000). Another chromosomally encoded gene, ChvD which is a homologue to ATP-binding cassette transporters is also thought to be involved in inducing the virulence through an effect on VirG expression (Liu et al., 2001).
As phenolic compounds secreted by plants are part of the defense mechanism after pathogen attack, plant-inducible loci in the vir region, i.e. virH and pinF are believed to be involved in detoxifying these compounds when they are present at high concentrations (Sheng and Citovsky, 1996). The virH operon consists of virH1 and virH2 genes, which encode cytochrome P450-type enzymes that could be responsible for O-demethylation activities (Brencic et al., 2004). Kalogeraki et al. (1999) reported that VirH2 protein acted as O-demethylase that demethylated the methoxy group of ferulic acid, one of the strong vir inducers, to form caffeic acid which lacks inducer activity. Brencic et al.
(2004) proposed three possible roles of VirH2, i) to limit the duration of vir gene induction. Because virH2 is tightly regulated by VirA and VirG, the virulence gene will only be inactivated later after the phenolic compounds are able to completely induce the vir regulon to express all the vir genes. ii) To detoxify the phenolic compounds as suggested by Kalogeraki et al. (1999). The authors concluded that the demethylation of phenolic compounds decreases their toxicity and during infection and colonization, this would be an added advantage to the Agrobacterium. iii) For catabolism of phenolics as carbon and energy sources. The authors also reported that three phenolic compounds tested, i.e. vanillyl alcohol, vanillin and vanillate, could serve as sole sources of carbon for Agrobacterium.
It is widely accepted that the induction of vir genes is essential for the effective gene transfer. The essential structures required for the phenolic compounds involved in inducing the virulence activity are a benzene ring with a hydroxyl group at position 4 and the methoxy group at position 3 whereas the presence of another methoxy group at position 5 increases the virulence inducer activity (Brencic and Winans, 2005). These findings supported the previous research carried out by Melchers et al. (1989) who showed that molecules involved in inducing the vir genes must have at least one methoxy group as well as para-hydroxyl groups. To date, more than 80 compounds are known to
have the ability to induce the vir gene expression. Acetosyringone and α- hydroxyacetosyringone are the first two phenolic compounds isolated from tobacco root cultures that have been identified as specific virulence inducers (Stachel et al. 1985;
Brencic and Winans, 2005). The other families of phenolic compounds that have been reported to have a positive effect on vir gene induction are hydroxycinnamides (Berthelot et al., 1998), benzalacetone and chalcones (Joubert et al., 2002) as well as lignin precursors in dicots such as sinapyl alcohol and coniferyl alcohol (Melchers et al., 1989).
Table 2.1 illustrates the chemical structures of some of the phenolic derivatives that show inducer activity on Agrobacterium vir genes.
Table 2.1 Chemical structures of some of the vir gene inducers from a family of phenolic-related compounds.
Basic structure Compounds R1 R2 R3
Acetosyringone OMe OMe COCH3
Hydroxyacetosyringone OMe OMe COCH2OH
Syringic acid OMe OMe COOH
Coniferyl alcohol -H OMe CH=CHCH2OH Sinapyl alcohol OMe OMe CH=CHCH2OH
Sinapic acid OMe OMe CH=CHCOOH
Syringaldehyde OMe OMe CHO
Acetovanillone -H OMe COCH3
Vanillacetone -H OMe CH=CHCOCH3
Vanillin -H OMe CHO
Ferulic acid -H OMe CH=CHCOOH
R2 R1
OH R3
Ethyl ferulate -H OMe CH=CHCOOCH2CH3
Environmental stimuli such as acidic pH, temperature and monosaccharides may also be involved synergistically with phenolic compounds to enhance the vir gene expression (Winans, 1992; Cangelosi et al., 1989; Jin et al., 1993). Induction of vir gene expression upon low pH was suggested by two mechanisms (Liu et al., 2001), one involves VirA.
Acidic pH may stimulate the activity of the transmembrane sensory kinase of VirA. The other mechanism is through the activation of the P2 promoter of VirG that acts independently from VirA (Chang et al., 1996). Another VirG promoter, P1 is suggested to be responsible for induction by phenolic compounds and phosphate starvation (Winans, 1990). Li et al. (2002) hypothesized that the ChvG protein, which is part of the two component regulatory system ChvG/ChvI that is important for virulence, is responsible for regulating two acid inducible genes; aopB and katA in Agrobacterium.
The authors suggested that ChvG functions as a global pH sensor protein that directly or indirectly senses the extracellular acidity, which plays an important role in causing the Agrobacterium tumorigenesis. The induction of vir genes by phenolic compounds is also greatly influenced by monosaccharides (Brencic and Winans, 2005). Ankenbauer and Nester (1990) suggested that arabinose, galactose, galacturonic acid, glucose, glucuronic acid, mannose and cellobiose are among the specific monosaccharides that can enhance the vir gene activity. Galacturonic and glucuronic acid are considered as the most potent inducers, which are effective even at a very low concentration (100 µM).
The molecular mechanism underlying the interaction of phenolic compounds and VirA is presumably due to VirA acting as a phenolic binding protein. The presence of a double bond in phenolic compounds is essential for the induction (Joubert et al., 2002).
VirA comprises of a periplasmic and a cytoplasmic domain in which the latter, consists of a linker, kinase and receiver domain. The periplasmic domain is likely to be involved in sensing monosaccharides involved in vir gene induction (Peng et al., 1998) whilst the sensor domain for phenolics is probably located adjacent to the TM2 membrane-spanning region and consists of the linker domain and the adjoining part of the cytoplasmic domain. The proposed mechanism begins when an acidic residue of VirA interacts with the phenolic compound by protonating the carbonyl group (Joubert et al., 2002). The electron resonance delocalization effect that occurs during the process enables the transfer of an electron from the phenol to the VirA acidic residue. The activated phenol
activation
B
H3CO
HO
H3CO
O
CH3
HO O
HB
H3CO
O
H3CO
OH
CH3
O O
Figure 2.4 Proposed mechanism of phenolic induction of vir gene expression. Electron resonance delocalization occurred during this process. An acidic residue at the binding site protonates the carbonyl group and activates the phenol. The phenol then protonates the basic residue (B) at the surface receptor that leads to the conformational change of VirA. This process induces the phosporylation cascade of VirG. (Hess et al., 1991, Joubert et al., 2002)
Figure 2.5 Generation of T-strand. VirD2 by the help of VirD1 nicks the bottom strand of the RB and LB sequence of the T-DNA to form a single stranded endonucleolytic cleavage within the both sites. This process creates free 3’OH in which the new DNA synthesis is initiated by using the top strand of T-DNA as a template. The DNA synthesis begins at the RB cleavage sites, proceeds across the T-region and terminates at the LB cleavage sites. This synthesis displaces bottom strand of T-DNA and eventually this strand is released from the T-region to form a free single stranded T-DNA molecule called T-strand (Stachel et al., 1986)
5’ 3’ 5’ 3’
3’
LB RB
3’ 5’
5’
T-strand +
then protonates a VirA basic residue leading to a conformational change in the protein and subsequent initiation of the phosporylation cascade leading to VirG activation (Hess et al., 1991) (Figure 2.4). The chromosomally encoded sugar binding protein, chvE, which interacts with the periplasmic domain of VirA is thought to be involved in this process by enhancing the sensitivity of VirA to phenolic compounds (Campbell et al., 2000). Previous research carried out by several workers however, showed that instead of VirA, two other low molecular weight binding proteins encoded by the bacterial chromosome, p10 and p21 may directly participate in binding to the phenolic compounds, suggesting that these proteins may mediate the signal transduction system prior to VirA autophosporylation (Lee et al., 1992).
2.2.3 Generation of T-DNA complex
Induction of virulence genes leads to the generation of a linear single stranded T-DNA, called T-strand. VirD2 and VirD1 proteins encoded by the Agrobacterium VirD locus are responsible for generating the T-strand from the double stranded T-DNA region (Herrera- Estrella et al., 1990). VirD2, a site specific endonuclease, with the help of VirD1 (Zupan and Zambryski, 1995) recognizes and nicks the 25-bp border sequences to create a single stranded endonucleolytic cleavage in the lower strand between the 3rd and 4th base pair of each border of the T-DNA. This process produces, by the help of DNA replication enzyme and cellular repair machinery, a single stranded copy of the T-DNA representing the bottom strand of T-DNA. Upon nicking of the right border, VirD2 forms a phospodiester bond and remains covalently bound at the 5’end of the T-strand (Tzfira et al., 2000) via tyrosine residue 29 (Ziemiennowicz, 2001) to form the ssT-DNA/VirD2 complex as the final outcome before it is transferred into the plant cell. A model for the generation of single stranded T-DNA as proposed by Stachel et al. (1986) is illustrated in Figure 2.5.
Other additional vir genes, VirC1 and VirC2, are also suggested to participate in generation of the T-strand. VirC1 was proposed to specifically bind with the ‘overdrive’
sequence (Toro et al. 1989), which is thought to play an essential role in enhancing the level of single stranded DNA formation (Toro et al., 1989). Previous research by van
Haaren et al. (1987) showed that Agrobacterium strains that had both the right border repeat and overdrive sequence produced a large amount of T-strands upon induction with acetosyringone whereas the strain that had only the right border produced lesser amounts of T-strand. This cis-active 24-bp overdrive sequence, located on the right of the T-DNA left border, is proposed to provide an active binding site for VirC1 through direct interaction with VirD endonuclease, as well as for optimal transfer of wild type T-DNA (Toro et al., 1989; Shurvinton and Ream, 1991; Zhu et al., 2000).
2.2.4 Transfer of T-DNA into plant cell
Agrobacterium needs to undergo a few major obstacles in order to achieve a successful gene transfer. The bacterium must transfer the T-DNA across the plant cell wall and plasma membrane into the cytoplasm to allow it to reach the nucleus. Eventually, the T- DNA must be stably integrated into the plant genome. Agrobacterium tumefaciens exploits the bacterial type IV secretion system (T4SS) which requires eleven virB genes and virD4 to transfer the T-DNA complex and virulence protein into the plant cell cytoplasm. T4SS consists of two major components: 1) a T-pilus, a filamentous appendage composed of VirB2 as the main component with VirB5 and VirB7 as associated components, and 2) a membrane associated transporter complex which is able to span the bacterial double membrane (Eckardt, 2004). The T-pilus is generated in the surface presumably at one end of the Agrobacterium cell upon induction of virulence genes by plant chemical signals. It is thought to play a key role in establishing bacterial- plant cell contact and also acts as a transenvelope mating channel (Atmakuri et al., 2004) in assisting the transport of ssT-DNA/VirD2 complex from Agrobacterium into the plant cytoplasm. The mating channel of VirB2 probably interacts with the T-DNA complex with the help of VirD4 (Ziemienowicz, 2001) to form the VirB/VirD4 transport apparatus prior to transporting the T-DNA complex and virulence genes into the plant cytoplasm.
Research conducted by Hwang et al. (2004), demonstrated that three plant proteins, the VirB2-interacting proteins (BTI1-BTI3) and a membrane-associated GTPase, AtRAB8 in Arabidopsis may interact with the T-pilus during the early interaction of Agrobacterium and plant cell. Although it is apparent that the T-pilus is an essential part of the process responsible for transporting the T-DNA complex and virulence proteins from
Agrobacterium to the plant cell, the precise mechanism on how this transfer process occurs is still poorly understood.
The transfer of ssT-DNA/VirD2 complex from the bacterial cytoplasm into the plant nucleus requires the presence of two other vir proteins i.e. VirE1 and VirE2. VirE1 acts as a specific molecular chaperone to prevent VirE2 from self-aggregation as well as to stabilize and maintain VirE2 in an export-competent state in Agrobacterium (Deng et al.
1999) before transport into the plant cytoplasm. Hypothetically, VirE1 also can prevent binding of VirE2 to newly formed ssT-DNA in the bacterial cell as well as prevent the formation of VirE2-dependent channels in the bacterial membrane (Duckely and Hohn, 2003). In the plant cytoplasm, VirE2 coats the entire ssT-DNA/VirD2 complex before it is transferred into the plant nucleus. Thus it was proposed that the T-DNA complex comprises of a single stranded DNA with VirD2 capped at its 5’end and coated with VirE2 protein. It is thought that around 600 molecules of VirE2 cover the entire strand of T-DNA (Rossi et al., 1998). Initially it was proposed that VirE2 binds to ssT- DNA/VirD2 in the bacterial cytoplasm prior to export to the plant cytoplasm (Zupan and Zambryski, 1995), but more recent studies conducted by several researchers suggest that Agrobacterium exports ssT-DNA/VirD2 complex and VirE2 independently (Gelvin, 1998). Later, research by Vergunst et al. (2000) has proven that the transfer of VirE2 into plant cells is mediated by the Agrobacterium VirB/D4 transport system but independently from T-DNA transport. All these findings confirmed earlier work by Regensburg-Tuïnk et al. (1993) which indicated that some vir-proteins such as VirF could be transferred into plant cells by the vir-encoded transport system independently from T-DNA.
The proteins encoded by the VirD and VirE operons that form the T-DNA complex are not only essential to ensure the integration of intact T-DNA (Pelczar et al., 2004), but may also function to protect the T-DNA strand against degradation (Ward, 1988; Rossi et al., 1996), to recognize transmembrane channels from the bacterium to the host cell (Zambryski, 1988), as well perform as a nuclear targeting signal in the plant cell (Herrera-Estrella et al., 1990). Recent biophysical studies demonstrated that VirE2 is able to interact with lipids in the plant plasma membrane to form a transmembrane channel allowing the ssT-DNA/VirD2 complex to be transported into the plant cytoplasm (Dumas, 2001; Duckely and Hohn, 2003).
Both the VirD2 and VirE2 protein posses a short peptide sequence that act as a nuclear localization signal (NLS) mediating the import of T-DNA into the plant nucleus (Escobar et al., 2003). VirD2 protein contains a carboxyl-terminal NLS which was suggested to bind with the Arabidopsis nuclear import receptor, AtKAP-α (karyopherin α) to mediate the nuclear import of the T-DNA strand (Ballas and Chitovsky, 1997). AtKAP-α is a plant importin that is known to be responsible for nuclear translocation of proteins containing NLS sequences (Eckardt, 2004). An additional plant protein, cyclophilin (CyPs) in three isoforms, i.e. Roc1, Roc4 and CypA was also reported to specifically interact with VirD2 and probably play a role in T-DNA complex transfer both in the plant cytoplasm and nucleus (Deng et al., 1998). CyPs-VirD2 interaction is disrupted by cyclosporine A, a compound that binds cyclophilins and inhibits their peptidyl-prolyl cis- trans isomerase (PPIase) activity in vitro (Tzfira et al., 2000), leading to the inhibition of Agrobacterium-mediated transformation in Arabidopsis and tobacco. The authors hypothesized that due to its transfer from bacterium to plant cell, VirD2, along with the T-DNA complex may change its conformation to execute its function in the new environment. CyPs, known to have the PPIase activity, may aid in proper folding of VirD2 into a different conformation inside the plant cell. The authors also suggested that as CyPs may have chaperone activities, the CyPs-VirD2 binding may be required to maintain VirD2 in a functional and transfer-competent state condition.
Despite the presence of an NLS in VirE2, research showed that it is not directly involved in nuclear import of the T-complex into the plant nucleus as has been observed for VirD2 (Rossi et al., 1996). In fact, it has been proposed that VirE2 plays a role in unfolding the T-DNA strand to pass through the nuclear pore (Citovsky et al., 1992).
Tzfira et al. (2001) proposed that Arabidopsis VirE2-interacting factor (VIP1), a plant nuclear protein, may specifically interact with VirE2 during the nuclear import of the latter. The deduced amino acid sequence of VIP1 consists of structural features characteristic of a basic zipper (bZIP) protein known to be localized in the cell nucleus (Tzfira et al., 2001; van der Krol and Chua, 1991), suggesting a role of this protein in facilitating VirE2 nuclear import and the T-DNA transfer to the plant cell nucleus.
Research conducted by Citovsky et al. (2004) showed that VIP1 specifically interacted with a component of cellular export machinery, karyopherin α and VirE2 to form ternary
complexes. In these complexes, VIP1 may act as a molecular adapter between VirE2 and karyopherin α protein, allowing VirE2 to utilize the plant cellular export machinery. Li et al. (2005) found that the VIP1 was also functional beyond nuclear import. It was shown that the ability of the N-terminal portion of VIP1 to bind with VirE2 was sufficient to facilitate transient expression in Arabidopsis. Moreover, the C-terminal portion of VIP1 functions in homomultimerization and interaction with the plant H2A histone protein, supporting the idea that this VIP1 is also involved in promoting tumorigenesis and consequently in stable genetic transformation. Research by Lacroix et al. (2005) demonstrated the possible role of another virulence protein, VirE3 in the nuclear import.
In the host cell cytoplasm, VirE3 binds with the karyopherin α protein which subsequently mediates its transfer into the nucleus. In addition, VirE3 was also reported to interact with VirE2. Because the nuclear import of VirE2 is likely to be mediated by the VIP1 protein, it was proposed that VirE3 resembles the function of VIP1 and acts as an ‘adapter’ between VirE2 and karyopherin α, ‘piggy backing’ VirE2 into the host nucleus. The authors suggest that because VIP1 is found as non-abundant protein in the host cell and as one of the host limiting factors for plant transformation, Agrobacterium evolved its own virulence protein to compensate for the absence or low level of VIP1 in certain plant species.
It is also presumed that another virulence protein, VirF, may play a vital role in the transfer process (Gelvin, 2000). Bound VirD2, VirE2 and probably VIP1 protein have to be removed from the T-DNA complex before its integration into the plant genome.
Recently, Tzfira et al. (2004) suggested that VirF is one of the factors involved in uncoating the T-DNA complex from its cognate proteins VIP1 and VirE2. It was proposed that VirF destabilized VIP1 and VirE2 by specifically targeting both proteins for proteosomal degradation in which the Skp1-Cdc53-cullin-F-box (SCF) complex is involved. Schrammeijer et al. (2001) reported that VirF contains the F box by which it can directly interact with plant homologs of the yeast Skp1, the ASK (Arabidopsis Skp1- like) protein. The SCF complex, which consists of four subunits: cullin (Cdc53 in yeast), Skp1, Rbx1 and F-Box protein, is a family of ubiquitin-protein ligase (E3) that play a role in selective protein degradation by the ubiquitin-proteosome pathway. The Skp1 probably interacts and activates the F-box which functions as a receptor to attract the specific
target protein for proteolysis (Schrammeijer et al., 2001). Research by Tzfira et al. (2004) also demonstrated that VirF is not bound to VirD2, suggesting that VirD2 may not be affected or remains bound to the T-DNA complex during the integration (Eckardt, 2004).
2.2.5 Integration of T-DNA into plant nuclear genome and its expression
Whereas extensive studies have been carried out to reveal the molecular mechanisms that occur during the integration of T-DNA into the host genome, some of the steps remain largely obscure. The T-DNA enters the plant cell nucleus as a complex with VirD2 bound to the 5’end of the single stranded molecule. Once in the nucleus, VirE2 coats the entire T-DNA strand prior to its integration into the plant chromosome by illegitimate recombination or non-homologous recombination (Ohba, 1995), i.e. a type of recombination between two molecules that do not share extensive sequence homology (Gheysen et al., 1991; Mayerhofer et al., 1991). It was proposed that the integration process begins when the 3’end of the T-DNA strand finds microhomologies on the complementary target DNA strand and subsequently invades the plant DNA (Tinland and Hohn, 1995). An A-T rich region in plant DNA is likely the favorable entry site for integration (Brunaud et al., 2002). Upon generation of nicks in the upper strand of the host DNA, the complementary strand of the invading T-DNA is synthesized by DNA polymerase (Pelczar et al., 2004) until it reaches the 5’end of the right border which is covalently attached to the VirD2 molecule (Brunaud et al., 2002). This step is thought to be accompanied by removal of VirE2 molecules from the T-DNA strand (Pelczar et al., 2004). The right end of the T-strand is then ligated into the lower strand of the host DNA while the nicked upper strand is extended upon repair synthesis (Tinland and Hohn, 1995) and becomes ligated into the synthesized complementary T-DNA. This process results in newly synthesized double stranded T-DNA as a final outcome (Brunaud et al., 2002). This integration pattern frequently results in a short deletion of target DNA and also invading DNA. It is presumed that the integration of T-DNA into the plant chromosome is mediated by a double-strand break (DSB) repair mechanism in plant DNA in which this site of DNA damage appears to be the entry point for the site of integration (Salomon and Puchta, 1998). This mode of T-DNA insertion was proposed as the normal mode of T-DNA entry into the plant genome (Chilton and Que, 2003).
The Agrobacterium and plant proteins that may be involved in the T-DNA integration process are still largely unknown and under extensive study. Initially it is thought that the VirD2 protein is responsible for the process, as it functions as a ‘pilot’ protein to guide the T-DNA into the plant nucleus (Gelvin, 2000) and remains covalently attached at the 5’end of T-DNA up to the integration process (Ziemienowicz, 2001). Prior to this finding, research by Mysore et al. (1998) indicated that a conserved region in VirD2, called the ω domain is involved in T-DNA integration. The authors hypothesized that this domain may be involved in establishing the correct conformation of the T-DNA protein complex or may interact with plant factors that mediate the T-DNA integration into the plant genome. However, the precise role of VirD2 in the integration is still unclear. Based on the model of T-DNA integration by Tinland (1996), it was suggested that VirD2 acts as integrase-ligase in T-DNA integration. However, recent work conducted by several researchers revealed that VirD2 proteins play no role in T-DNA integration (Ziemienowicz et al., 2000; van Attikum et al., 2001). Ziemienowicz et al. (2000) reported that VirD2 is not involved in T-DNA ligation in vitro. Instead, the authors suggested that the ligation process is performed by plant enzymes, most probably by DNA ligase. Using the yeast S. cerevisiae as a model, van Attikum et al. (2001) demonstrated that T-DNA integration into the yeast genome by non homologous recombination (NHR) requires the non homologous end joining (NHEJ) host proteins YKU70, LIG4, RAD50, MRE11, XRS2 and SIR4 in which the genes that encode these proteins are also involved in repair of DNA double strand breaks for NHEJ. Later, the authors demonstrated that RAD52 and YKU70 are the key proteins required for T-DNA integration, channeling the T-DNA integration into either the NHR or HR pathway (van Attikum et al., 2003). As NHR is the preferable mechanism for T-DNA integration in plants, study of the plant orthologs of the genes that code for these proteins will further clarify the mechanisms of the T-DNA integration in plant.
Research conducted by Bako et al. (2003) suggested that two further proteins, a plant protein kinase, CAK2Ms in alfalfa and the TATA box binding protein (TBP) in Arabidopsis, may also be involved in the T-DNA integration into the plant genome. The CAK2Ms protein was found to be able to bind and phosphorylate the C-terminal regulatory domain of the RNA polymerase II largest subunit that serves as the TBP
binding platform. It was proposed that VirD2 interacts with and is phosphorylated by the CAK2Ms protein and also binds tightly to the TATA box binding protein, suggesting it is implicated in targeting of the T-DNA to the plant nucleus. Mysore et al. (2000) reported that an Arabidopsis RAT5 histone H2A protein also plays a role in T-DNA integration.
This was based on the finding that the Arabidopsis rat5 mutant with a disrupted histone H2A gene is unable to exhibit T-DNA integration in its genome, whereas overexpression of RAT5 increased the transformation efficiency. Later, research by Yi et al. (2002) demonstrated that there is a strong correlation of the expression of the histone H2A gene with the susceptibility to Agrobacterium transformation. They found that the elongation zone of Arabidopsis root, which is the most susceptible part for Agrobacterium-mediated transformation, highly expresses the histone H2A protein. Furthermore, the pattern of root susceptibility to phytohormone treatment is correlated with the histone H2A expression pattern. All these findings suggest that the histone H2A protein may be an important target for improving the transformation efficiency in recalcitrant species.
2.3 Factors influencing plant transformation efficiency
Rapid development of plant transformation methodologies has led to the situation that it is possible to transfer genes with desired traits to a range of agronomically important crops. As Agrobacterium-mediated transformation has played a major role in the advancement of plant genetic engineering, a lot of basic scientific and applied research nowadays is focused on the factors that can enhance the transformation efficiency. This goal is achieved by manipulating various factors that may contribute to the success of plant transformation, including the Agrobacterium strain or by optimizing the plant tissue culture condition, inoculation and cocultivation techniques (Nontaswatsri et al., 2004), the selectable marker gene, the plant genotype, the type of vector used (Hiei et al., 1997), as well as the improvement of the plant cellular factors involved in Agrobacterium- mediated transformation. However, the most critical task in achieving high transformation efficiency is to find the right combination of the factors mentioned above, since each acts differently depending on the behavior under the culture conditions of the selected explants to be transformed (Wei et al., 2000).
The success of plant transformation depends on the stable incorporation of foreign genes into the plant genome, followed by regeneration of the intact plant and subsequently the expression of the introduced gene in the transgenic plant (Walden and Wingender, 1995).
However, it is much more difficult to obtain a stable transformant than to obtain high transient gene expression in a plant (Lessard et al., 2002). Generally, high transient expression has not led to more readily obtainable stable transformants. Whereas a large number of dicots can easily be transformed by Agrobacterium, some of them are recalcitrant and remain unreceptive to Agrobacterium infection. This may be attributed to the lack of good regeneration capacity under tissue culture conditions (Wei et al., 2000), as well as to a problem in T-DNA integration itself. Thus it is a prerequisite to first establish an efficient regeneration protocol of the plant species to be transformed.
Generally, when using Agrobacterium-mediated transformation, one needs to select the transformed plant cells from the total tissue containing a majority of non-transformed cells and if complete transgenic plants are required, plant regeneration procedures through tissue culture techniques are necessary. Although high-frequency plant regeneration has been reported from many varieties of crops, the regeneration capacity is still depending on the explant sources, the species and even on the genotype of a given species (Pua and Li, 2004). Nontaswatsri et al. (2004) showed efficient genetic transformation of carnation by just cocultivating the explants on filter paper soaked with water and acetosyringone (AS) instead of using a phytohormone and nutrient-rich media during the cocultivation procedure.
Another important factor to be considered that can improve the transformation efficiency is the choice of Agrobacterium strain, whether it is an octopine, nopaline or another strain. This is due to the different virulence systems present on different Ti plasmids that may cause the host range differences. Melchers et al. (1990) demonstrated that octopine and nopaline strains had difference virulence levels on transformation of N.
glauca, and the differences were due to the absence of the virF gene in the nopaline strains. The use of a ‘super binary’ vector in which the binary Ti vector contains additional vir genes and ‘super virulent’ Agrobacterium strains in which the disarmed Ti plasmid contains different vir genes have improved plant transformation efficiencies considerably (Komari, 1990; Hiei et al., 1994; Zhao et al., 2000), especially for
recalcitrant species (Vain, 2004). Research by Srivatanakal et al. (2000) showed that additional virulence genes on the helper plasmid doubled or tripled the yield of transformed tobacco. Similar results have been reported by Khanna and Daggard (2003) who demonstrated that low efficiency of wheat transformation was overcome by using an Agrobacterium strain possessing a super binary vector carrying an extra set of vir genes.
The authors also reported that the use of the polyamine spermidine in the regeneration medium increased the transformation recovery, suggesting another exogenous element that may contribute to high frequency of transformation. The use of polyamines in Agrobacterium-mediated transformation has also been reported by other researchers (Kumar and Rajam, 2005). The results suggest that the polyamines enhance vir gene induction and subsequently the T-DNA transfer in tobacco cells. The function of polyamines in enhancing vir gene activity and T-DNA transfer is not known, but it could be due to the interaction of polyamines with the molecular machinery of vir gene induction and particularly with VirA-VirG signal transduction (Kumar and Rajam, 2005).
Another limiting factor in Agrobacterium-mediated transformation is the effect of tissue necrosis of the explants after infection by Agrobacterium. This is due to the hypersensitive response of the plant. Upon infection with a pathogen or after wounding, the plant activates its defense machinery, which induces an oxidative burst (Wojtaszek, 1997) causing the production of reactive oxygen species (ROS). The ROS intermediates are thought to mediate programmed cell death of plant cells, which generates a barrier of dead cells around the site of infection (Olhoft et al., 2001). The addition of an antioxidant in cocultivation and/or regeneration media is thought to reduce this hypersensitive response. Several researchers reported the efficient transformation of certain monocots after supplementing the medium or pre-treatment of the explants with antinecrotic compounds with known antioxidant activity such as the thiol L-cystein (Olhoft and Somers, 2001), ascorbic acid (Enriquez-Obregon et al., 1999), dithiothreitol and polyvinylpolypyrrolidone (Perl et al., 1996).
Wounding of explant tissue is also one of the critical steps leading to the successful Agrobacterium-mediated transformation as it allows Agrobacterium to infect the target tissue as well as to produce the phenolic chemical signals that induce the T-DNA transfer (Stachel et al., 1985). Some researchers reported that the wounding of the explant could
be enhanced by subjecting the plant tissue to sonication (Santare et al., 1998; Trick and Finer, 1997; Weber et al., 2003). This little explored technology or so-called sonication- assisted Agrobacterium-mediated transformation (SAAT) tremendously improved the transformation efficiency of several crops that were recalcitrant to Agrobacterium- mediated transformation. The enhanced transformation efficiency by this method was probably due to the microwounding that occurred on the surface and inside of the target tissue which allows the bacterial cells to reach the proper target tissue.
2.4 Conclusion and perspective
As Agrobacterium-mediated plant transformation has become the most used method to introduce foreign genes to obtain a desired phenotype in a variety of crops, the fundamental knowledge underlying the molecular mechanism of Agrobacterium- mediated plant transformation has been a hot topic since many years. The important events including the bacterial attachment to the plant cell, vir gene activation, T-DNA processing, nuclear targeting and T-DNA integration have been quite well studied, although the role of the host cellular proteins involved in the transformation process remains largely obscure and is still under extensive investigation. A better understanding of all molecular events in the process as well as the plant proteins involved could be exploited for the further improvement of Agrobacterium-mediated plant transformation.
In addition, the knowledge on the factors that influence the transformation efficiency is also crucial. The detailed knowledge on the factors limiting the transformation efficiency will broaden the range of the crop species that can be transformed by A. tumefaciens especially for the recalcitrant species.
2.5 Acknowledgements
The authors thank Prof. Dr. P.J.J. Hooykaas for his comments and critical reading of this review. We also thank Malaysian Agricultural Research and Development Institute (MARDI), Malaysia for the Ph.D grant to Sanimah Simoh.
