Citation
Soltani, J. (2009, January 14). Host genes involved in Agrobacterium-mediated transformation. Retrieved from https://hdl.handle.net/1887/13400
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/1887/13400
Note: To cite this publication please use the final published version (if applicable).
General Introduction
Soltani, J., van Heusden, G.P.H. and Hooykaas, P.J.J. (2008) Agrobacterium-mediated transformation of non-plant organisms. In: Agrobacterium: from biology to biotechnology. pp 649-675, Eds Tzfira, T. and Citovsky, V. Springer press, New York, USA (with minor modifications).
Agrobacterium-mediated transformation of non-plant organisms
Agrobacterium tumefaciens is a plant pathogen, which causes crown gall by genetic transformation in more than 600 dicotyledonous plant species belonging to 90 families (De Cleene and De Ley, 1976). Although tumors are not formed on monocots, Agrobacterium can transform such plant species including the cereals as well (Hooykaas-Van Slogteren et al., 1984; Ishida et al., 1996). The capability of Agrobacterium to transform plants is widely used in plant biotechnology and plant research. This is based on the presence in Agrobacterium of a large Ti plasmid, which contains a set of genes (the virulence genes) that can mobilize a segment of the Ti plasmid called the T-DNA, which is surrounded by a direct repeat (border repeat) of 24 bp, to plant cells. The vir-genes are activated in the presence of plant-specific phenolic compounds such as acetosyringone. As shown in Figure 1, reception of the signal by VirA chemoreceptor leads to the activation of the transcription regulator VirG. Then VirG mediates transcription of the other vir-genes. Besides virA and virG the virB operon (with 11 genes) and the virD operon are essential for transformation.
The virB operon encodes the Type IV Secretion System (T4SS) which delivers the T- DNA and a number of virulence proteins into the plant cells. The virD operon encodes proteins involved in the production of the single stranded DNA copy of the T-DNA that is transferred to the plant cell. VirD2, in a complex with VirD1, VirC1 and VirC2, excises the T-strand from the Ti plasmid. VirC1/VirC2 function to stimulate relaxosome assembly at T-DNA border sequences (Atmakuri et al., 2007).
Upon excision, VirD2 remains covalently bound to the liberated 5 phosphate of the T-strand (Durrenberger et al. 1989). VirD2 then as a pilot protein mediates the translocation of the T-strand to the recipient cell via the T4SS that is formed by VirB1-11/VirD4 proteins. Inside the host cell the T- strand is covered by VirE2 protein molecules, which protect the T-strand against nucleases. In parallel, several virulence proteins like VirE2, VirE3, VirD5 and VirF translocate to the host cell (Vergunst et al., 2000; Schrammeijer et al., 2001, 2003). By a cooperation of virulence proteins and host factors the T-complex then moves to the nucleus where the T-DNA integrates into the genome. For a review of the transformation process and the functions of the virulence proteins see Citovsky et al., 2006; Gelvin, 2003;
Hooykaas and Schilperoort, 1992; McCullen and Binns, 2006; Tzfira et al., 2004; and Zhu et al., 2000.
Inside the host cell, VirD2 has critical functions for AMT. It has nuclear localization sequences and has interaction with plant importin/karyopherin proteins, implicating a role in the nuclear entry of the T-complex (Ballas and citovsky, 1997; Tzfira et al., 2001, 2004; Li et al., 2005). VirD2 interacts with Arabidopsis karyopherin via itsC- terminal bipartite NLS (Ballas and citovsky, 1997). It also interacts with several members of the plant cyclophilin family of peptidyl-prolyl cis-trans isomerases including Roc1, Roc2, Roc3, Roc4 and Roc5/CypA, (Deng et al., 1998). Cytoplasmic interaction of VirD2 with a plant protein phosphatase 2C (tomato Dig3, Arabidopsis Abi1) inhibits its nuclear import (Tao et al., 2004). Inside the host nucleus, VirD2 may play a role in the integration of T-DNA into the genome, although the T-DNA integration is mainly mediated by host factors (van Attikum et al., 2001; 2003).
Indeed, interaction of VirD2 with a conserved cyclin-dependent kinase-activating kinase (Cak2M) and the TATA-binding protein (TBP) is supposed to play a role in T- DNA integration via interaction with RNA polymerase II holoenzyme (Bako et al., 2003). The functional VirD2 protein is very important for the transformation of both plant and non-plant organisms (See section 3.2).
During the last decade it became clear that the ability of Agrobacterium to transform host organisms is not restricted to plants, but that numerous other eukaryotic and even prokaryotic organisms are transformable by Agrobacterium under laboratory conditions. Since the pioneering work on the yeast Saccharomyces cerevisiae (Bundock et al., 1995),
Agrobacterium-mediated transformation of other yeasts and many fungi has been shown (reviewed in Michielse et al., 2005; Hooykaas et al., 2006; see Table 1 for a complete list). In addition, Agrobacterium-mediated transformation of algae (Cheney et al., 2001; Kumar et al., 2004), oomycetes (Vijn and Govers, 2003), sea urchin (Bulgakov et al., 2006), mammalian cells (Kunik et al., 2001) and the gram positive bacterium Streptomyces lividans (Kelly and Kado, 2002) has been reported. At the moment genomes from many organisms have been sequenced or are being sequenced.
For effective functional genomics of these organisms and for their application in biotechnology highly efficient and facile genetic transformation protocols are needed.
In this respect, Agrobacterium-mediated transformation is becoming a very effective tool. The more traditional transformation methods such as particle bombardment and
the use of polyethylene glycol treated cells or protoplasts, have several drawbacks including the low transformation efficiency, the difficulty to control the copy number, the loss of molecular integrity of the DNA, the need to generate protoplasts and the limits on the size of the DNA (van den Eede et al., 2004). On the other hand, for Agrobacterium-mediated transformation it is possible to use different kinds of intact host cells such as conidia,
Figure 1. Overview of the main events in Agrobacterium-mediated transformation. In the vicinity of wounded plant cells Agrobacterium perceives the phytochemicals (1) by the VirA-VirG regulatory system (3) and attaches to the plant cell (2). This activates the expression of Vir proteins (4) and production of T-strand linked to VirD2 (5) and the movement of T-strand-VirD2 and a sera of Vir proteins to the host cell via a type four secretion system (T4SS) (6). Inside the host cytoplasm the T- complex is formed (7) and by using host motors and proteins moves to the nucleus (8). VirD2 has several interacting partners which mediate some events in the host cell (9). VirF is involved in protein degradation (10). Before the integration of T-DNA (13) it might be uncoated (11) and be converted to a ds-DNA (12) most likely by host machineries.
mycelia, sexual spores and fruiting body tissues from fungi without the need to make protoplasts (Michielse et al., 2005). Furthermore, it is possible to transfer relatively large segments of DNA, up to 150 kb with no or only little rearrangements (Hamilton et al., 1996) and the transformation frequencies are higher than with traditional transformation methods (de Groot et al., 1998; Bundock et al., 1999; Amey et al.,
2002; Campoy et al., 2003; Fitzgerald et al., 2003; Meyer et al., 2003; Idnurm et al., 2004; Michielse et al., 2004a; Rodriguez-Tovar et al., 2005). Also some species such as Agaricus bisporus and Calonecteria morganii that could not be transformed by traditional methods turned out to be transformable by Agrobacterium (de Groot et al., 1998; Malonek and Meinhardt, 2001). The potential of Agrobacterium to generate transformants having only a single integrated copy of the transgene in the genome, has been shown not only for plants but also for yeasts, fungi and mammalian cells (de Groot et al., 1998; Bundock et al., 1999; Kunik et al., 2001; Bundock et al., 2002).
This property makes Agrobacterium potentially a very powerful tool for insertial mutagenesis, gene tagging and gene targeting (Michielse et al., 2005). Indeed, Agrobacterium has been used as a tool for insertional mutagenesis in plants, such as Arabidopsis thaliana, Nicotiana species, and Oryza sativa (Koncz et al., 1989; Koncz et al., 1992; Krysan et al., 1999; Jeon et al., 2000) and more recently in different genera of fungi such as the model eukaryote S. cerevisiae (Bundock et al., 2002), the symbiotic fungus Hebeloma cylindrosporum (Combier et al., 2003), the biocontrol agents Beauveria bassiana (Leclerque et al., 2003), Coniothyrium minitans (Rogers et al., 2004), and Trichoderma reeisi (Zhong et al., 2007), the phytopathogens Colletotrichum. acutatum (Talhinhas et al., 2008) and Magnaporthe grisea (Betts et al., 2007; Choi et al., 2007; Li et al., 2003), and the human pathogens Aspegillus fumigatus (Sugui et al., 2005), Cryptococcus neoformans (Idnurm et al., 2004) and Penicillium marneffei (Zhang et al., 2008). The ability of Agrobacterium to genetically modify the yeast S. cerevisiae offers the possibility to use the many experimental tools available for this organism to study the transformation process in detail. A major issue in the transformation of eukaryotic cells is the integration of the foreign DNA at random positions in the genome rather than at specific locations. In contrast to most eukaryotic organisms, S. cerevisiae very efficiently integrates the foreign DNA by homologous recombination, allowing targeted integration at specific chromosomal locations. A comparison of the integration processes occurring in S.
cerevisae with those occurring in other eukaryotes may unravel the factors required for targeted integration of the foreign DNA. In this chapter we will review the Agrobacterium-mediated transformation of non-plant species. Several aspects of the Agrobacterium-mediated transformation of non-plant organisms especially fungi have been discussed in recent reviews (Hooykaas, 2005; Michielse et al., 2005; Lacroix et
al., 2006). Here, we focus on the transformation of both fungi and other non-plant organisms.
1. Non-plant organisms transformed by Agrobacterium
Since the observation that Agrobacterium is not only capable of transforming plant cells, but also cells from the yeast S. cerevisae, many other non-plant organisms have been successfully transformed by Agrobacterium (Table 1). Most of these organisms are fungi, but also algae and mammalian cells have been transformed. Agrobacterium- mediated transformation is not restricted to eukaryotes as Agrobacterium is also able to transform the gram positive bacterium Streptomyces lividans. Foreign DNA can be engineered to allow stable integration into nuclear as well as in extranuclear DNA, such as plastids and mitochondrial DNA. For instance, chloroplast DNA has successfully been engineered for resistance to herbicides, insects, disease and drought, and for the production of biopharmaceuticals (reviewed by Daniell et al., 2005). Most of the methods used for transformation of extranuclear DNA are based on polyethylene glycol treated protoplasts. More than a decade ago, two studies on Agrobacterium-mediated transformation of plastid DNA have been published (De Block et al., 1985; Venkateswarlu and Nazar, 1991), but these results have not been reproduced.
2. Experimental aspects of Agrobacterium-mediated transformation of non-plant organisms
Many different protocols for the transformation of non-plant organisms have been developed. In general, the procedures for the transformation of the different organisms are similar. For example, the binary system is standard for use in both plants and non-plant organisms. Most transformations of non-plant organisms are performed by co-cultivation of Agrobacterium and recipient cells on a solid support.
On the other hand, each organism requires its own optimal conditions to obtain maximal transformation frequencies. An optimized protocol for the Agrobacterium- mediated transformation of the yeast S. cerevisiae has recently been published (Hooykaas et al., 2006) which is the basic protocol for other fungi as well (Figure 2).
2.1 Agrobacterium strains
Various Agrobacterium strains have been used for the transformation of non-plant organisms, e.g. LBA4404, EHA105, and LBA1100. Systematic comparisons of different strains in relation to transformation frequencies have not been published, making it difficult to say which strain is best to use. The use of Agrobacterium strains derived from the supervirulent A281 strain which has a high level of vir gene expression, resulted in higher transformation frequencies in S. cerevisiae, Monascus purpureus, Phytophthora infestans and Cryphonectria parasitica (Piers et al., 1996;
Campoy et al., 2003; Vijn and Govers, 2003; Park and Kim, 2004). The introduction of a ternary plasmid carrying the virG mutant gene coding for the constitutively active Vir-GN54D protein into Agrobacterium strain LBA1100 resulted in a considerable improvement in the transformation efficiency of P. infestans (Vijn and Govers, 2003).
2.2 Requirement of acetosyringone
During plant transformation the T-DNA transfer machinery of Agrobacterium is induced by phenolic compounds such as acetosyringone originating from wounded plant cells. Also for transformation of non-plant organisms the virulence system has to be induced and in most transformation protocols the addition of acetosyringone to the induction medium is required. On the other hand, it has been reported that acetosyringone was not necessary for transformation of the alga Chlamydomonas reinhardtii (Kumar et al., 2004). Addition of acetosyringone not only to the induction medium, but also to the Agrobacterium pre-culture medium, improved transformation frequencies of the fungi Beauveria bassiana, Fusarium oxysporum, and Magnaporthe grisea (Mullins et al., 2001; Rho et al., 2001; Leclerque et al., 2003). Furthermore, omission of acetosyringone from the pre-culture medium delayed the formation of transformants. In contrast, addition of acetosyringone to the pre-culture medium did not affect transformation frequencies of the fungi Hebeloma cylindrosporum and Colletotrichum trifolii (Combier et al., 2003; Takahara et al., 2004). Thus, the requirement for the addition of acetosyringone to the pre-culture medium seems to be organism-dependent.
2.3 Effect of co-cultivation conditions
The conditions under which Agrobacterium is co-cultivated with the recipient organism have a major influence on the transformation frequency. Transformation efficiency is influenced by the ratio between Agrobacterium and recipient, the length of the co-cultivation period, temperature, pH, and the choice of filters. Increasing the amount of Agrobacterium cells relative to the recipient cells in the co-cultivation
Day 1
Day 2
Day 8
Day 11
Overnight culture in YPD
Yeast strains Agrobacterium strains
Overnight culture in LB+kan
In fresh YPD In Induction Medium + AS
6 hrs
Take 1 ml aliquot, pellet, wash cells, resuspend yeast cells in 1 ml of induction medium (without glucose).
Co-cultivate 100 l of yeast-Agrobacterium mix on cellulose nitrate filter on induction plate. Incubate 6-9 days at 22C.
- Transfer the filter containing co-cultivation mix into an eppendorf tube.
Vortex. Plate out 250 l on selection plate (without uracil).
- Make serial dilutions of co-cultivation mix to 10-5. Plate out 250 l on YPD plate containing cefotaxim.
- Incubate 3-5 days at 30C.
Count the number of yeast transformants, and the total number of yeast cells.
Calculate the relative frequency.
6 hrs
Take 1 ml aliquot, pellet, wash cells, resuspend cells in 1 ml of induction medium.
Mix 60 l aliquot of yeast with 60 l aliquot of Agrobacterium.
Figure 2. Schematic presentation of standard protocol for Agrobacterium-mediated transformation of yeast (Bundock et al., 1995). AS, acetosyringone; kan, kanamycin.
mixture may lead to an increase in the transformation frequency. However, addition of too many Agrobacterium cells can result in a decrease in transformation efficiency (Meyer et al., 2003). Several studies have shown that each organism has an optimal combination of co-cultivation period and temperature to obtain a maximum number of transformants (Mullins et al., 2001; Rho et al., 2001; Combier et al., 2003; Meyer et al., 2003; Rolland et al., 2003; Gardiner and Howlett, 2004; Michielse et al., 2004a).
In most transformation protocols optimal transformation is achieved at room temperature. An interesting aspect of Agrobacterium-mediated transformation of mammalian cells is that it occurred at 37°C after pre-growth of Agrobacterium at 28°C (Kunik et al., 2001). The effect of pH during co-cultivation on the transformation frequency was tested in Agrobacterium-mediated transformation of C.
trifolii and Colletotrichum lagenarium (Tsuji et al., 2003; Takahara et al., 2004). It was found that the optimal pH, leading to the highest transformation frequency, is between 5.0 and 5.3. Also for transformation of the yeast S. cerevisiae a small deviation from the optimal pH of 5.3 already resulted in a considerably lower transformation frequency (J. Soltani, unpublished observation). The optimal pH also depends on the Agrobacterium strain used, as the pH requirements for optimal vir- gene induction are slightly different for the different Agrobacterium strains (Turk et al., 1991). For efficient Agrobacterium-mediated transformation cells are co- cultivated on a solid support such as nitrocellulose filters, Hybond, filter paper, cellophane sheets, and polyvinylidene difluoride.
2.4 Markers used for Agrobacterium-mediated transformation
The vectors used for Agrobacterium-mediated transformation of non-plant organisms have similar requirements as the vectors used for plant transformations. The DNA sequences located inside the T-DNA borders will be transferred to the recipient cells and for selection of transformants a suitable selection marker is required. Frequently used markers in both plants and non-plant systems are different antibiotic resistance genes from bacterial plasmids. Also an herbicide resistance gene (bar) has been used as a selection marker in fungi (Fang et al., 2004). It is important that these markers are controlled by a promoter active in the host organism. For transformation of the yeast S. cerevisiae both auxotrophic markers such as URA3 and TRP1 as well as dominant resistance markers such as the G418 resistance marker KAN-MX are being used (Bundock et al., 1995; Piers et al., 1996; van Attikum, 2003). Uracil auxotrophy
markers have also been used during Agrobacterium-mediated transformation of filamentous fungi (Gouka et al., 1999; Sullivan et al., 2002; Michielse et al., 2005).
3. Role of virulence proteins in the Agrobacterium-mediated transformation of non-plant organisms
3.1 Chromosomally-encoded virulence proteins
Chromosomally-encoded virulence proteins (Chv proteins) are necessary for T-DNA transfer to plants. However, their role in transformation of non-plant organisms is not well-established. For Agrobacterium-mediated transformation of the yeast S.
cerevisiae the chromosomal virulence operons chvA, chvB, and exoC which are required for bacterial attachment to plant cells are not required (Piers et al., 1996). On the other hand, Agrobacterium-mediated transformation of mammalian cells depends on the presence of the chvA and chvB genes (Kunik et al., 2001). Reversely, it was reported that inactivation of one of the chromosomal genes involved in the biosynthesis of cellulose fibrils increases the frequency of transformation of Aspergillus awamori (Michielse et al., 2005).
3.2 Ti-plasmid encoded virulence proteins
The role of the Agrobacterium Ti-plasmid encoded virulence proteins (Vir proteins) in plant transformation is well studied. Some information is also available on the role of the Vir proteins in the transformation of non-plant organisms. The virA, virB, virD and virG genes are essential not only for plant transformation, but also for transformation of non-plant organism such as the yeast S. cerevisiae (Bundock et al., 1995) and the fungus A. awamori (Michielse et al., 2004b). An Agrobacterium strain with a deletion in the NLS and the omega sequences of virD2 was unable to transform yeast and was highly inefficient to transform A. awamori (Bundock et al., 1995;
Michielse et al., 2004b). The integrated T-DNAs of A. awamori transformants obtained with the virD2 mutant had truncated right ends (Michielse et al., 2004a).
Moreover, deletion of 5-end of virD2 which abolishes its exonuclease activity also abolished transformation of A. awamori by Agrobacterium (Michielse et al., 2004b).
Hence as in plants, these results indicate the significance of VirD2 in AMT of non plant organisms and that this protein plays a role in protecting the right T-DNA border. Although inactivation of virE2 almost eliminates the ability of Agrobacterium
to transform plants, transformation of S. cerevisiae by such mutant is only 10-fold reduced (Bundock et al., 1995) and transformation of A. awamori only less than 2- fold (Michielse et al., 2004a). Nevertheless, as in plants A. awamori transformants had left-border truncations (Michielse et al., 2004a), indicating that VirE2 in fungi as in plants helps to protect the T-strand against nucleases. The deletion of virC2, encoding a subunit of the T-DNA relaxosomal complex, reduced the transformation efficiency of A. awamori about 8-fold. Transformants in this case were characterized by the presence of complex T-DNA structures containing multicopy and truncated T- DNAs and vector backbone sequences (Michielse et al., 2004a). This suggests that VirC2 plays a role in correct T-DNA border processing and is required for single- copy T-DNA integration.
4. Targeted integration of T-DNA
Agrobacterium-mediated transformation of the yeast S. cerevisiae can result in random insertion of the T-DNA into the yeast genome by non-homologous end joining as is the common mechanism for T-DNA integration into the plant chromosome (Figure 3.A) (Bundock and Hooykaas, 1996). However, when DNA sequences homologous to those of the yeast genome are present, the DNA fragment will mostly integrate into the genome by homologous recombination (Figure 3.B).
This will result in integration of the T-DNA at a predetermined location of the chromosome. Integration of the T-DNA via homologous recombination occurs approximately 50-100-fold more efficient than via non-homologous recombination in the yeast (Bundock et al., 1995). When the T-DNA contains a yeast replicator such as an autonomously replicating sequence (ARS) or the replicator of the 2 plasmid, the T-DNA will be maintained in the yeast cell as a replicative plasmid (Bundock et al., 1995; Piers et al., 1996), after circularization of the T-DNA (Figure 3.C). In S.
cerevisiae the integration of T-DNA by homologous recombination is very efficient.
However, for most other organisms T-DNA insertion mainly occurs via non- homologous recombination, even when the T-DNA fragment has extensive sequence homology to the host chromosome. For the application of Agrobacterium-mediated transformation in functional genomics or biotechnology it is of great importance to improve the efficiency of integration via homologous recombination over non- homologous recombination. By using the yeast S. cerevisiae as a model, it was found that the proteins mediating T-DNA integration are the proteins involved in double
strand break (DSB) repair of the genomic DNA (van Attikum et al., 2001). Indeed, T- DNA integrates at preformed DSBs in both plant (Salomon and Puchta, 1998) and yeast chromosomes (van Attikum, 2003) (Reviewed in Tzfira et al., 2004). Several studies in S. cerevisiae have shed light on the mechanisms involved in DNA break repair, which may occur by homologous recombination or by non-homologous end joining (reviewed by Pâques and Haber, 1999; Jackson, 2002; Symington, 2002;
West, 2003; Dudasova et al., 2004; Krogh and Symington, 2004; Daley et al., 2005).
These mechanisms are summarized in Figure 4. Homologous recombination is initiated by a chromosomal double strand break (DSB) followed by the nucleolytic resection of the ends of the double stands break in the 5’ to 3’ direction. In yeast, and most likely also in mammals, this reaction relies on the Mre11 nuclease activity in a multiprotein complex consisting of Rad50, Mre11, and Xrs2 (Figure 4.A). The Rad52 protein is able to bind to the ends of double strand breaks. This has led to the proposal that this Rad52 binding channels to repair by homologous recombination instead of to non-homologous end joining. For T-DNA integration at double strand breaks by homologous recombination Rad52 is essential (van Attikum and Hooykaas, 2003).
The Rad51 nucleofilament interacts with undamaged DNA molecules and searches for a region with extensive homology. This process is influenced by other proteins such as replication protein A (Rpa), Rad52 and Rad54. When homology is found for example on the T-DNA, Rad51 catalyzes a strand invasion reaction in which the 3’
protruding end of the damaged molecule invades the undamaged DNA molecule. The 3’ end of the damaged DNA molecule is then extended by a DNA polymerase that copies information from the T-DNA molecule and the ends are ligated by a DNA ligase, resulting in the integration of the T-DNA into the chromosome.
Non-homologous end joining is initiated by a double strand break, followed by binding of the Ku70 and Ku80 proteins to the ends (Figure 4.B). With the help of other proteins the break is then sealed restoring the original sequence, with or without small deletions. Heterologous DNA sequences, including those of transposable elements and the Agrobacterium T-DNA can be captured during this process and be integrated at the break site. In S. cerevisiae at least six genes are required for efficient integration of T-DNA via non-homologous end joining: YKU70, LIG4, RAD50, MRE11, XRS2 and SIR4 (van Attikum et al., 2001). Interestingly, RAD50, MRE11 and XRS2 are also involved in (meiotic) homologous recombination, but not in T-DNA
integration by homologous recombination (van Attikum, 2003). Recently, it has been shown that also histone modifiers and ATP-dependent chromatin-remodeling complexes are recruited to sites of DNA damage (reviewed by Peterson and Côté, 2004; van Attikum and Gasser, 2005). Some plant homologs of the components of these complexes have already been found which show positive effects on or are differentially expressed during Agrobacterium-mediated transformation (Veena et al., 2003; Zhu et al., 2003; Loyter et al., 2005). For T-DNA integration at double strand breaks by non-homologous end joining Yku70 is essential (van Attikum, 2003). Thus, in S. cerevisiae the Rad52 and Yku70 proteins play a critical role in determining whether the T-DNA is integrated via homologous recombination or via non- homologous end joining (van Attikum, 2003). Microhomology between the T-DNA and the chromosomal DNA plays a role in the initial steps of the non-homologous end joining process (Tinland, 1996; Wurtele et al., 2003; Daley et al., 2005). As a result, truncations or deletions of the ends of the T-DNA may occur. This has been observed not only in plants but also in the yeast S. cerevisiae, in filamentous fungi and in mammalian cells (Bundock and Hooykaas, 1996; de Groot et al., 1998; Kunik et al., 2001; Mullins et al., 2001; van Attikum, 2003; Leclerque et al., 2003). It is expected that in the absence of the non-homologous end joining proteins Ku70 or Ku80 integration by homologous recombination will become relatively more frequent.
Indeed, after mutation of the YKU70 or YKU80 genes no integration by non- homologous end joining was seen in S. cerevisiae (van Attikum and Hooykaas, 2003).
It was also shown that in the yeast Kluyveromyces lactis disruption of YKU80 led to a large increase in integration of T-DNA by homologous recombination up to 97% of transformants analyzed (Kooistra et al., 2004). Also, in the filamentous fungus Neurospora crassa disruption of YKU70 or YKU80 vastly increased the integration of exogenous DNA into the genome by homologous recombination at a frequency of 100% (Ninomiya et al., 2004). These results support the great potential of manipulating the recombination machinery in optimizing targeted integration of T- DNA. Furthermore, as in plants in S. cerevisiae T-DNA can also be integrated via gap repair (Risseeuw et al., 1996). This integration event supports the model for the integration of T-DNA as a single strand in cooperation with VirD2 protein (reviewed in Tzfira et al., 2004).
Figure 3. Mechanisms involved in targeted and non-targeted integration of T-DNA and circularization into plasmids. (A) Random insertion of the T-DNA into the yeast genome by non-homologous end joining. (B) Insertion of the T-DNA at a predetermined location of the yeast genome by homologous recombination. (C) Circularization and maintenance in the yeast cell of the T-DNA containing replicator of the 2 plasmid. LB, left border; RB, right border; 2 ori, origin of replication from the yeast 2 plasmid, kanMX, kanamycin resistance gene.
Figure 4. Overview of the host mechanisms involved in T-DNA integration via homologous recombination (A) and non-homologous end joining (B) in S. cerevisiae. (A). Homologous recombination is initiated by a DSB followed by the nucleolytic resection of the ends of DSB by the Rad50, Mre11, and Xrs2. Rad51 interacts with undamaged DNA molecules (or T-DNA) and searches for a region with extensive homology and catalyzes a strand invasion reaction. The 3’ end of the damaged DNA molecule is then extended by a DNA polymerase and the ends are ligated by a DNA ligase. (B) Non-homologous end joining is initiated by a DSB, followed by binding of the Ku70 and Ku80 proteins to the ends. With the help of other proteins the break is then sealed restoring the original sequence or with small deletions. In S. cerevisiae at least six genes are required for efficient integration of T-DNA via non-homologous end joining: YKU70, LIG4, RAD50, MRE11, XRS2 and SIR4 (van Attikum et al., 2001). In this model it is assumed that the T-DNA is double stranded before integration.
DSB, double strands break. Adapted from Jackson, 2002 (Jackson, 2002) and van Attikum, 2003 (van Attikum, 2003).
5. Protein transfer from Agrobacterium to non-plant hosts
During the transformation process Agrobacterium transfers not only T-DNA but also a number of its virulence proteins into the host cell (Vergunst et al., 2000). This protein transfer is not restricted to plant cells, but it has been shown that the VirE2, VirE3 and VirF proteins can be transferred into cells from the yeast Saccharomyces cerevisiae as well (Schrammeijer et al., 2003). To study protein transfer from Agrobacterium to yeast, the Cre recombinase reporter assay for translocation has been used. Protein fusions between Cre and Vir proteins were expressed in Agrobacterium.
Transfer of the Cre-Vir fusion proteins from Agrobacterium to yeast can be monitored by a selectable excision event resulting from site-specific recombination mediated by Cre on a lox-flanked transgene in yeast. This assay illustrates the potential of Agrobacterium to introduce genome modifying enzymes into eukaryotic cells. As the signal for transport by the type four secretion system (TFSS) lies in the 30 C-terminal amino acids of transferred proteins, coupling of this transport signal to the C-terminus of heterologous proteins may allow their mobilization from Agrobacterium to eukaryotic target cells (Vergunst et al., 2005). This property of Agrobacterium is promising for its application in protein therapy of both plant and non-plant cells.
6. Prospects
Agrobacterium-mediated transformation has become a widely used tool for transformation of different types of eukaryotic cells. Especially for the transformation of various fungi, it has great advantages over other transformation methods. The efficiencies are much higher and the transformation protocols are relatively facile. It has also been shown that multiple copies of the T-DNA can be integrated by homologous recombination at a predetermined position of the genome of A. awamori allowing a high level of expression of the introduced gene of interest (Gouka et al., 1999). T-DNA integration at random positions in the genome of most eukaryotic organisms makes Agrobacterium a very useful tool for random mutagenesis and random gene tagging. Agrobacterium can transfer not only DNA but also proteins to the host organisms through its type four secretion system (TFSS). This protein transfer has been shown to occur independently of DNA transfer to both plants and the yeast S. cerevisiae. Most likely, protein transfer occurs during the transformation of all host cells, irrespective of their origin. Because of this property, Agrobacterium has a great potential for use in protein therapies. A major issue in the transformation
of eukaryotic cells is the integration of the foreign DNA at random positions in the genome rather than at specific locations. In contrast to most other organisms, S.
cerevisiae has a very efficient system for targeted integration of DNA fragments via homologous recombination, but will integrate the DNA at random positions if homology with the genome is lacking. By the use of the well developed genetics of S.
cerevisiae it was possible to identify key factors that control DNA integration by homologous recombination and non-homologous end-joining, respectively. As these key proteins are strongly conserved in other eukaryotes (from fungi to plants and animals) the knowledge obtained from yeast may be directly applicable in these other organisms to improve the frequency of targeted integration. Indeed, by disruption of YKU70 or YKU80 in the yeast K. lactis (Kooistra et al., 2004) and in the filamentous fungus Neurospora crassa (Ninomiya et al., 2004) the relative efficiency of targeted integration increased. Recently, it has been shown that expression in the plant Arabidopsis thaliana of the S. cerevisae RAD54 gene which is involved in chromatin remodeling, resulted in an increase of the integration of T-DNA by homologous recombination by two orders of magnitude (Shaked et al., 2005). The ability of Agrobacterium to transfer T-DNA, to a wide variety of eukaryotic and some prokaryotic organisms may have important consequences for evolution. In the rhizosphere where vir inducers are readily available and numerous microorganisms are living in close proximity, it is very likely that Agrobacterium-mediated transformation of non-plant organisms is occurring. This process may contribute to horizontal gene transfer. In this respect it is of interest that Agrobacterium DNA was found in the genomes of tobacco plants (Aoki and Syono, 1999; Frundt et al., 1998;
Furner et al., 1986; Intrieri and Buiatti, 2001; Meyer et al., 1995; Suzuki et al., 2002;
Tanaka et al., 2004). Future research has to show whether such horizontal gene transfer to non-plant organisms has happened and contributed to evolution.
7. Outline of this thesis
The virulence system of Agrobacterium tumefaciens is well studied, and the genes and mechanisms that mediate gene transfer from Agrobacterium to eukaryotic cells are relatively well known. After crossing the cell barriers of the host cell the introduced T-strand is translocated to the sites of integration in the host nuclear genome with the help of bacterial virulence proteins. The mechanisms by which Agrobacterium escapes from the host defense systems and utilizes host factors to
accomplish eventually T-DNA integration into the genome remain largely elusive. In this thesis, we used the yeast S. cerevisiae to find the host genes that are involved in Agrobaterium-mediated transformation (AMT). For this we used a collection of around 4800 yeast strains, each with a deletion in an individual ORF, and Agrobacterium strains carrying either of two plasmids allowing either the integration or the replication of T-DNA in yeast. Furthermore, a key protein which mediates the excision of T-DNA from Ti-plasmid, translocation of T-DNA to the host cell, nuclear entry of the T-DNA and full-length T-DNA integration into the plant- and non-plant host genomes is the Agrobaterium VirD2 protein. We used the yeast two-hybrid system to identify the yeast proteins which interact with VirD2 in order to better understand the mechanisms of T-DNA nuclear localization and integration into the yeast genome.
In Chapter 2 the modification of our standard protocol of AMT of yeast is described.
We adjusted the standard protocol to make it suitable for transformation of a large number of yeast strains maintained in microtiter plates.
In Chapter 3 the microtiter-based AMT protocol was used to screen the collection of approximately 4800 S. cerevisiae homozygous deletion strains to identify deletions that affect AMT. This resulted in the identification of 141 genes of which deletion increased AMT and of 108 genes of which deletion decreased AMT.
Of special interest were our observations that certain genes encoding chromatin modifying proteins affected AMT. In Chapter 4 we studied the effect of deletion of genes encoding histone acetyltransferases (HAT) and deacetylases (HDAC) in more detail. We used our standard protocol to quantify the AMT efficiency of such deletion mutants. This further confirmed the data obtained from the microtiter based AMT assays, and resulted in identification of several other HAT and HDAC deletions that affect AMT. We also studied the effect of complementation and overexpression of those genes on AMT.
Chapter 5 describes the experiments we initiated to unravel the subcellular localization of VirD2 after expression in yeast and to identify yeast factors that
interact with VirD2. Those experiments show the nuclear localization of VirD2, and identify 12 potential interaction partners of VirD2 in yeast.
Table 1. Fungi (including yeasts) transformed by A. tumefaciens (modified from Michielse et al., 2005).
Species Reference(s) Zygomycetes
Blakeslea trispora (Michielse et al., 2005) Backusella lamprospora (Nyilasi et al., 2008) Mucor circinelloides (Nyilasi et al., 2003)
M. miehei (Monfort et al., 2003)
Rhizopus oryzae (Michielse et al., 2004c)
Ascomycetes
Acremonium implicatum (Abello et al., 2008)
Ascochyta rabiei (Mogensen et al., 2006; White and Chen, 2006) Aspergilus awamori (de Groot et al., 1998; Gouka et al., 1999;
Michielse et al., 2004b)
A. giganteus (Meyer et al., 2003)
A. niger (de Groot et al., 1998)
Beauveria bassiana (dos-Reis et al., 2004; Leclerque et al., 2003;
Fang et al., 2004)
Blastomyces dermatiditis (Brandhorst et al., 2002; Sullivan et al., 2002) Botrytis cinerea (Rolland et al., 2003)
Cadophora finlandia (Gorfer et al., 2007)
Calonectria morganii (Malonek and Meinhardt, 2001) Candida albicans (Michielse et al., 2005)
C. glabrata (Michielse et al., 2005)
C. glycerinogenes (Zhiming et al., 2008)
C. tropicalis (Michielse et al., 2005)
Ceratocystis resinifera (Loppnau et al., 2004) Chaetomium globosum (Gao and Yang, 2005) Claviceps pururea (Michielse et al., 2005) Coccidiodes immitis (Abuodeh et al., 2000)
C. posadasii (Kellner et al., 2005)
Colletotrichum acutatum (Maruthachalam et al., 2008; Talhinhas et al., 2008)
C. destructivum (O’Connell et al., 2004)
C. falcatum (Maruthachalam et al., 2008)
C. gloeosporioides (de Groot et al., 1998)
Species Reference(s)
C. graminicola (Flowers and Vaillancourt 2005)
C. lagenarium (Tsuji et al., 2003)
C. trifolii (Takahara et al., 2004)
Coniothyrium minitans (Li et al., 2005; Rogers et al., 2004) Cryphonectria parasitica (Park and Kim 2004)
Fusarium circinatum (Covert et al., 2001)
F. culmorum (Michielse et al., 2005)
F. oxysporum (Khang et al., 2005; Mullins et al., 2001; Takken et al., 2004)
F. venenatum (de Groot et al., 1998)
Glarea lozoyensis (Zhang et al., 2003) Helminthosporium turcicum (Degefu and Hanif, 2003) Histoplasma capsulatum (Sullivan et al., 2002)
Kluyveromyces lactis (Bundock et al., 1999; Kooistra et al., 2004) Leptosphaeria biglobosa (Eckert et al., 2005)
L. maculans (Eckert et al., 2005; Gardiner and Howlett, 2004;
Gardiner et al., 2005)
Magnaporthe grisea (Khang et al., 2005; Li et al., 2003; Rho et al., 2001)
Metarhizium anisopliae (Fang et al., 2006) Metarhizium anisopliae var. acridum (Duarte et al., 2007)
Monascus ruber (Yang and Lee, 2008)
Monascus purpureus (Campoy et al., 2003) Monilinia fructicola (Dai et al., 2003) Mycosphaerella fijiensis (Michielse et al., 2005) M. graminicola (Zwiers and de Waard, 2001) Neurospora crassa (de Groot et al., 1998) Oculimacula acuformis (Eckert et al., 2005)
O. yallundae (Eckert et al., 2005)
Ophiostoma floccosum (Michielse et al., 2005)
O. piceae (Tanguay and Breuil, 2003)
O. piliferum (Hoffman and Breuil, 2004)
Paecilomyces fumosoroseus (Lima et al., 2006) Paracoccidioides brasiliensis (Leal et al., 2004) Penicillium chrysogenum (Sun et al., 2002)
P. marneffei (Zhang et al., 2008)
Phialocephala fortinii (Gorfer et al., 2007)
Saccharomyces cerevisiae (Bundock et al., 1995; Bundock et al., 2002;
Piers et al., 1996; Risseeuw et al., 1996) Sclerotinia sclerotiorum (Weld et al., 2006)
Trichoderma asperellum (Michielse et al., 2005)
Species Reference(s)
T. atroviride (Zeilinger, 2004)
T. harzianum (Michielse et al., 2005)
T. longibrachiatum (Michielse et al., 2005)
T. reesei (de Groot et al., 1998; Zhong et al., 2007)
Tuber borchii (Grimaldi et al., 2005)
Venturia inaequalis (Fitzgerald et al., 2003; Fitzgerald et al., 2004) Verticillium dahliae (Dobinson et al., 2003)
V. fungicola (Amey et al., 2002; Amey et al., 2003)
Basidiomycetes
Agaricus bisporus (Chen et al., 2000; de Groot et al., 1998; Mikosch et al., 2001)
Cryptococcus gattii (McClelland et al., 2005)
C. neoformans (Idnurm et al., 2004; McClelland et al., 2005) Hebeloma cylindrosporum (Combier et al., 2003; Pardo et al., 2002) Hypholoma sublateritium (Godio et al., 2004)
Laccaria bicolor (Kemppainen et al., 2005) Omphalotus olearius (Michielse et al., 2005) Paxillus involutus (Pardo et al., 2002) Phaffia rhodozyma (Michielse et al., 2005) Phanerochaete chrysosporium (Li and Zhang, 2005)
Pisolithus tinctorius (Rodriguez-Tovar et al., 2005)
P. microcarpus (Pardo et al., 2005)
Pseudozyma antarctica (Marchand et al., 2007) Volvariella volvacea (Wang et al., 2008)
Suillus bovines (Hanif et al., 2002; Pardo et al., 2002)
