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).
Genome-wide identification of host genes involved in Agrobacterium-mediated transformation by exploiting a
Saccharomyces cerevisiae deletion mutant collection
Abstract
Agrobacterium tumefaciens is a phytopathogen capable of genetically transforming plant cells. Under laboratory conditions it can also transform cells from many non- plant organisms including the yeast Saccharomyces cerevisiae. While the Agrobacterium genes required for the transformation process are relatively well known, most of the host genes involved still have to be identified. Here, we use the collection of ~4800 homozygous diploid deletion strains of non-essential genes in the S. cerevisiae strain BY4743 to find host genes which are involved in Agrobacterium- mediated transformation (AMT). By transformation of this strain collection with Agrobacterium strains carrying homologous integrative and replicating binary vectors, we identified 141 deletion strains that show at least two fold higher AMT efficiency and 108 deletion strains that show at least two fold lower AMT efficiency compared to the wild type strain BY4743. Classification of the identified genes according to the MIPS functional catalogs revealed that genes involved in protein activity regulation (7.6%), cell cycle and DNA processing (6.6%), metabolism (5.9%), cellular communication/signal transduction (5.6%), transcription (5.3%) and protein fate (5%) are well represented. These genes may be related to the different steps of the T-complex’s odyssey through the host cell. Deletion of several genes encoding components of ADA-, SAGA-, and SLIK- histone acetyltransferase complexes, i.e. NGG1, GCN5, SGF29, SGF73, VID21 and EAF7, of VPS36, encoding a component of the ESCRT-II complex involved in protein sorting and of THI21, involved in thiamine biosynthesis, resulted in a more than six fold increased AMT efficiency. Deletion of YHR151C, an uncharacterized gene, resulted in the lowest AMT efficiency. Deletion of genes encoding subunits of histone deacetylase complexes, i.e. HDA3, HST4, and SIN3 resulted in reduced AMT efficiency, in line with a major role of histone deacetylation in the transformation process.
Introduction
Agrobacterium tumefaciens is a soilborn phytopathogen capable of genetically transforming both di- and monocotyledonous plants from different families causing crown gall disease (Chilton et al., 1977, De Cleene and De Ley, 1976, Hooykaas-van Slogteren et al., 1984, Ishida et al., 1996). Under laboratory conditions it can also
transform cells from many non-plant organisms of different kingdoms including the yeast Saccharomyces cerevisiae (Bundock et al., 1995, Cheney at al, 2001, de Groot et al., 1998, Gouka et al., 1999, Kelly and Kado, 2002, Kumar et al., 2004, Kunik et al, 2001, Piers et al., 1996, reviewed by Soltani et al., 2008). This ability of Agrobacterium is based on the presence of a large tumor inducing (Ti) plasmid, which contains a set of virulence (vir) genes that can mobilize a segment of the Ti plasmid, i.e. the T-DNA in a single stranded form (T-strand), to host cells. Meanwhile, Agrobacterium also transfers a number of its virulence proteins to the host cell through its type four secretion system (Schrammeijer 2003, Vergunst et al., 2000, 2005). Delivered virulence proteins protect the T-strand from host nucleases, target it to the nucleus and may cooperate with host proteins to integrate it into the host genome.
Although the Agrobacterium genes required for the transformation process are relatively well known, investigations on host genes involved in this process are under way. A number of large-scale studies have already unraveled some host factors exploited by Agrobacterium to achieve transformation (reviewed by Gelvin, 2000, Citovsky et al., 2006, Tzfira and Citovsky, 2002). In a forward genetic screening of insertion mutant libraries of Arabidopsis thaliana 126 plant mutants resistant to Agrobacterium transformation (rat phenotype) were identified with mutations in genes related to chromatin remodeling, nuclear targeting, cytoskeleton, cell wall, metabolism, gene expression and signal transduction (Mysore et al., 2000, Zhu et al., 2003a,b). Agrobacterium-mediated root transformation of RNAi lines of chromatin- related genes of A. thaliana revealed 24 genes of which silencing resulted in decreased transformation efficiency (Crane and Gelvin, 2007).
Host factors involved in the transformation process have also been identified by their ability to interact with virulence proteins. Several plant proteins interacting with VirB2 VirD2, VirE2, VirE3 and VirF have been identified by protein-protein interaction assays. In Arabidopsis, VirB2 interacts with three proteins of unknown function and with AtRab8 which is a membrane-associated GTPase (Hwang and Gelvin, 2004). The VirD2 protein has different plant interactors, including the cyclophilins Roc1, Roc2, Roc3, Roc4 and Roc5, (Bako et al., 2003, Deng et al., 1998;
A.Briancon-Marjollet and H. van Attikum, unpublished data), the cyclin-dependent
kinase-activating kinase Cak2M (Bako et al., 2003), a TATA-box binding protein (Bako et al., 2003), the type 2C serine/threonine protein phosphatase Dig3 (Tao et al., 2004), and four members of the Arabidopsis importin α family (Bako et al., 2003, Ballas and Citovsky, 1997). The VirE2 protein interacts with the Arabidopsis basic- zipper (bZIP) protein VIP1 and the transcriptional regulator VIP2 (Tzfira et al., 2001;
Anand et al., 2007), and VirE3 interacts with the Arabidopsis importin α, the Csn5 subunit of the COP9 signalosome and with Brp which is a transcription factor of TFIIB family (Garcia-Rodriguez et al., 2006). VirF was shown to interact with the Arabidopsis homologues of the yeast Skp1 protein of SCF (Skp1-Cullin-F-box) complexes to mediate targeted proteolysis (Schrammeijer et al., 2001, Tzfira et al., 2004).
A number of studies have shown that the expression of a large number of plant genes is affected by Agrobacterium infection. In one of these studies with the plant Ageratum conyzoides it was shown that after 48 h of co-cultivation with Agrobacterium the expression of 251 plant genes was affected. For 56 of these genes the effect was already found after 24 h of co-cultivation (Ditt et al., 2001, 2005).
Several of the genes identified were related to plant defense and many of them were also differentially expressed in response to general stresses. Analysis of the effect of Agrobacterium infection on gene expression in the plant Arabidopsis showed that 48 h after inoculation a number of defense genes were induced, while a number of cell proliferation genes was repressed (Ditt et al., 2006). Another study showed that the transfer of T-DNA and Vir proteins affected the expression of 421 tobacco genes (Veena et al., 2003).
Although in recent years a substantial number of host genes involved in AMT has been identified, probably a large number of other genes relevant for AMT still have to be disclosed. Even though collections of Arabidopsis T-DNA insertion mutants are quite valuable for functional genomics and for the identification of host genes involved in AMT, a recent study has shown that these collections are incomplete as there are still a fairly high number of annotated genes without T-DNA insertions (Li et al., 2006). Moreover, due to the complex nature of Agrobacterium–host interactions, genome-wide screens for host genes affecting AMT are difficult to perform and many relevant genes may have been missed. Furthermore, except for
adenine mutants in yeast (Roberts et al., 2003), in none of the forward genetic screens hypersensitive host mutants have been found. During recent years the budding yeast S. cerevisiae has been developed as an excellent model host to study AMT. Studies in yeast have identified key genes involved in T-DNA integration via homologous recombination and non-homologous end joining (van Attikum et al., 2001, van Attikum and Hooykaas, 2003). Involvement of these NHEJ proteins in T-DNA integration in plants was less clear (van Attikum et al., 2003). Yeast has a small genome-size ( 6200 annotated open reading frames) and a relatively low level of gene redundancy. Yeast null deletion mutant collections, both haploid and diploid (homozygous and heterozygous), have been constructed (Giaever et al., 2002). These collections have been used for several genome-wide screens including screens for host genes involved in replication of plant viruses (Ishikawa et al., 1997, Panavas and Nagy, 2003, Pantaleo et al., 2003, Price et al., 2002, 2005) and screens for genes involved in translocation of the Ty transposon (Griffith et al., 2003, Irwin et al., 2005, Scholes et al., 2001). In this study we have used the collection of ~4800 non-essential diploid homozygous deletion mutants in a systematic approach to identify genes involved in AMT. By transformation of this strain collection with Agrobacterium strains carrying homologous integrative and replicating binary vectors, we identified 141 deletion strains that show at least two fold higher AMT efficiency and 108 deletion strains that show at least two fold lower AMT efficiency compared to the parental wild type strain BY4743. Genes identified in this study belong to almost all functional categories and may be related to the different steps of the T-complex’s odyssey through the host cell.
Materials and Methods
Bacterial strains, plasmids and media
Two derivatives of Agrobacterium tumefaciens strain LBA1100, containing either the homologous integrative plasmid pRAL7100 or the replicating plasmid pRAL7101 (Bundock et al., 1995), were used in Agrobacterium-mediated transformation screens (See Figure 1, Chapter 2). Agrobacterium was grown and maintained as described (Hooykaas et al., 2006). E. coli XL1-Blue (supE44 hsdR17 recA1 endA1 gyrA46 thi relA1 lac− F′ [proAB+ lacIq lacZΔM15 Tn10] Tcr) from Stratagene was used for
plasmid amplification. It was grown at 37°C in Luria-Bertani medium containing 60 µg/ml ampicillin.
Yeast strains and media
S. cerevisiae BY4743 (MATa/MATα his3Δ1/his3Δ1 leu2Δ0/leu2Δ0 lys2Δ0/LYS2 MET15/met15Δ0 ura3Δ0/ura3Δ0) and the complete collection (~4800 mutants) of homozygous diploid kanMX4 deletion strains of BY4743 (Giaever et al., 2002) were obtained from Euroscarf (Frankfurt, Germany) and InVitrogen (Groningen, the Netherlands), respectively. BY4743 was grown at 30°C in liquid YPD medium (Sherman et al., 1986) and deletion mutant strains were grown at 30°C in liquid YPD medium containing G418 (150 μg /ml).
Agrobacterium-mediated transformation
For AMT of S. cerevisiae strain collections cultivated in microtiter plates, we modified the standard protocol introduced by Bundock et al. (1995). Development of the new protocol is described in Chapter 2. Briefly, the wells of a 96-well microtiter plate containing 190µl of YPD with 150 μg/ml G418 were inoculated with deletion strains from the collection using 10 μl aliquots of each yeast mutant, followed by 48 hours growth at 30˚C under continuous shaking at 900 rpm. On each plate two wells were used for the wild type strain BY4743, cultivated in YPD medium lacking G418.
Aliquots of 20 μl of the overnight cultures were added to 180 μl of fresh YPD medium in new microtiter plates, followed by incubation for 6 hr at 30˚C under continuous shaking at 900 rpm. Then, yeast cells were collected by centrifugation for 2 min at 3000 rpm, and were washed once with 160 μl of induction medium.
Agrobacterium cultures were obtained by adding 1.2 ml of an overnight culture of the Agrobacterium strains to 20 ml of induction medium containing acetosyringone (20 mM) followed by incubation for 6 hrs at 29-30 ˚C under continuous shaking at 250- 300 rpm to an OD620 of approximately 0.6. The yeast cells were resuspended in 180 μl of Agrobacterium culture and the mixtures were shaken for 3 min. Then, 8 μl of the mixture from each well was pipetted onto rectangular nitrate cellulose filters (Schleicher and Schuell, Dassel, Germany) of the same size as the microtiter plate.
The filters were laid onto rectangular induction medium (IM) plates, supplemented with histidine (20 μg/ml), leucine (30 μg/ml) and uracil (20 μg/ml) and when required with adenine (20 μg/ml), arginine (20 μg/ml), methionine (20 μg/ml), serine (360 μg/ml)
or tryptophan (20 μg/ml). After 6-7 days of co-cultivation at 22˚C, emerged colonies were streaked onto rectangular selection plates using a swab, 24 colonies on each rectangular plate. Selection plates consisted of MY medium (Zonneveld, 1986) supplemented with histidine (20 μg/ml), leucine (30 μg/ml) and cefotaxim (200 μg/ml).
After 3 days incubation at 30˚C the transformants were counted.
The original AMT protocol introduced by Bundock et al (1995) was used for re- testing of selected yeast mutants.
Isolation of yeast chromosomal DNA
Yeast transformants were colony purified by streaking them on a selection plate followed by incubation for 2 days to obtain single cell colonies. These colonies were picked, cultivated overnight in selective medium and collected by centrifugation. The yeast pellet was washed with 1 ml sterile water, resuspended in freshly made lysis buffer (1ml of 0.9 M sorbitol / 50 mM EDTA, pH 8) containing Lyticase (1mg/ml) and incubated for 60 min at 30˚C. After centrifugation for 1 min at 5000 rpm, 0.8 ml of 10 mM Tris-HCl / 20 mM EDTA / 1% SDS was added and incubated at 65˚C for 10 min. Then, 500 μl phenol (saturated with 10 mM Tris-HCl (pH 8.0 / 1 mM EDTA) was added and the mix vortexed for 1 min before centrifugation for 5 min at 13000 rpm. The upper phase was extracted 3 times with phenol and one time with chloroform until the interphase disappeared. DNA was precipitated by addition of 0.1 volume of 3 M sodium acetate (pH 5) and 2 volumes of absolute ethanol (-20˚C), followed by an incubation for 30 min at -20˚C. Chromosomal DNA was collected by centrifugation for 10 min at 13000 rpm, washed with 200 μl 70% ethanol (4˚C), dried under vacuum, dissolved in 100 μl 1x TE (10 mM Tris/1 mM EDTA, pH 8) with 1 μl RNase (containing 1.5 mg/ml RNase A (Roche, Mannheim, Germany), 300 U/ml RNase T1 (Boehringer Mannheim) in 10 mM Tris, 15 mM NaCl, pH.7.5), incubated 30 min at 65˚C and stored at -20˚C.
Analysis of T-DNA integration at the PDA1 locus
Integration of T-DNA by homologous recombination at the PDA1 locus was checked by PCR and Southern blotting on chromosomal DNA. For PCR the following oligonucleotide primers were used: PDA-fw (5´
AAGAAAAGGAGACCCCCTATGG 3´), PDA-rev (5´
TCAGGGCCTGAGGATTCTATTG 3´), URA3-fw (5´
GATGAGTAGCAGCACGTTCC 3´) and URA3-rev (5´
GCAGGCTGGGAAGCATATTTG AG 3´). Reactions were performed in a total volume of 50 μl using Goldstar DNA polymerase (Eurogentec, Seraing, Belgium).
Amplification conditions were: 3 min at 94˚C, followed by 30 cycles of 1 min at 94˚C, 1 min at 54˚C and 3 min at 72˚C, ending with 10 min at 72˚C. PCR products were analyzed on 0.7% agarose gels in a 1x TAE (40 mM Tris/20 mM Acetate/1 mM EDTA) buffer. Expected bands for the combination of PDA-fw/URA3-fw which amplifies N-terminal parts and areas upstream of inserted T-DNA was ~1.5 kb, and for the combination of PDA-rev/URA3-rev which amplifies C-terminal parts and areas downstream of inserted T-DNA was ~2.5 kb. For Southern blot analysis, yeast chromosomal DNA was digested overnight with either SpeI or EcoRV restriction enzymes, run on a 0.7% agarose gel and transferred onto positively charged nylon membranes (Roche) using a vacuum pump (Pharmacia). Cross-linking of DNA, followed by hybridization with a 0.9 kb XhoI-BglII fragment obtained from pUC4α10 vector which has been radiolabeled with 32P. The bands were detected by autoradiography.
Yeast transformation using the lithium acetate (LiAc) transformation protocol
Yeast was transformed using the lithium acetate transformation protocol as described (Gietz and Schiestl, 1995). pUC4α10::HindIII, a pUC4 derived plasmid containing the S. cerevisiae URA3 gene flanked by PDA1 sequences, was used as a linear plasmid which allows integration into the PDA1 locus via homologous recombination (Steensma et al., 1990), and YCplac33 centromeric plasmid containing URA3 and CEN4/ARS1 (5.60 kb) (Gietz and Sugino, 1988) was used as a replicating plasmid.
Results and discussion
Screening yeast deletion strains for AMT
In order to identify host genes involved in AMT of the yeast S. cerevisiae we screened the collection of homozygous diploid non-essential deletion strains for mutants with an altered sensitivity towards AMT. For this screen we made use of Agrobacterium strains carrying either a homologous integrative or a replicating binary vector
(pRAL7100 and pRAL7101, respectively, Figure 1, Chapter 2). Following AMT the T-DNA part of the first vector will mainly integrate into the PDA1 locus by homologous recombination, whereas the T-DNA part of the latter plasmid will be maintained as an extrachromosomal plasmid after circularization (Bundock et al., 1995). Due to the low efficiency of microtiter-based AMT with the Agrobacterium strain carrying the binary vector pRAL7102 allowing random integration (Chapter 2), this Agrobacterium strain was not used in this screen. Using our novel microtitre plate based transformation protocol (Chapter 2) each of the ~4800 deletion mutants was independently co-cultivated with each of the two Agrobacterium strains and the transformation efficiencies were determined. Transformation efficiencies were compared to that of the parental wild type strain BY4743 present in duplicate on the same microtitre plate. In the first screen, deletion strains that gave at least 1.5-fold more (424 strains) or at least 1.5-fold less (475 strains) transformants than the wild type strain after co-cultivation with one of the two Agrobacterium strains were selected. These strains were screened twice more for AMT using our microtitre-based protocol. A reduced number of transformants may not only be the result of a lower transformation efficiency but also of a reduced viability of the deletion strain. For example, deletion strains may be viable on rich medium (YPD) but may be unable to grow on the selection plates containing minimal medium. Therefore, the deletion strains with apparent reduced transformation efficiency were tested for viability.
Thirteen deletion strains with decreased AMT that gave poor growth (an at least 100- fold lower number of colonies on the YPD plates) were removed from further investigating. These deletion strains include ARD1 (YHR013C), CDC26 (YFR036W), CST6 (YIL036W), HFM1 (YGL251C), MTO1 (YGL236C), OPI3 (YJR073C), TOS1 (YBR162C), TPS1 (YBR126C), UBP6 (YFR010W), YCR193C, YDR008C, YER091C-A and YPT6 (YLR262C). Finally, 249 mutants were identified that repeatedly showed an at least two fold increased (141 mutants) or an at least two fold decreased (108 mutants) AMT efficiency with one or both of the Agrobacterium strains. The identified genes - grouped into 10 functional categories partly based on the description in the Saccharomyces Genome Database (SGD) - are shown in Tables 1 and 2, and are discussed in this chapter. They have also been grouped into 18 functional categories according to the Munich Information Center for Protein Sequences (MIPS) Comprehensive Yeast Genome Database (http://mips.gsf.de/proj/yeast/CYGD/db) (Table 3). Deletion of several genes
encoding components of ADA-, SAGA-, and SLIK- histone acetyltransferase complexes, i.e. NGG1, GCN5, SGF29, SGF73, VID21 and EAF7, deletion of VPS36, encoding a component of the ESCRT-II complex involved in protein sorting, and deletion of THI21, involved in thiamine biosynthesis, resulted in the highest (more than 6.0-fold increased) AMT efficiencies. Deletion of HDA3, HST4, and SIN3 encoding subunits of histone deacetylase complexes resulted in reduced AMT efficiency, in line with a major role of histone deacetylation in the transformation process. However, the lowest AMT efficiencies are seen in the deletion mutants of YHR151C, which encodes an uncharacterized protein. Beside these, deletion in HAP4, encoding a subunit of CCAAT-binding complex, CLN3, encoding a G1 cyclin, GTR2, encoding a GTP-binding protein, SLT2, encoding a MAP kinase, ACB1, encoding Acyl-CoA-binding protein, ERG28, encoding an endoplasmic reticulum membrane protein, GND1, encoding 6-phosphogluconate dehydrogenase, SHM2, encoding serine hydroxymethyltransferase, ILM1, which maybe involved in mitochondrial DNA maintenance and MGR2 that is required for growth of cells lacking the mitochondrial genome resulted in strong decrease in AMT.
From the collection of hyper- and hypo-sensitive deletion strains 21 and 32 strains, respectively, were re-tested for AMT efficiency using the standard yeast AMT protocol. In the latter test the number of transformants was corrected for the number of viable yeast cells after co-cultivation and the corrected transformation efficiencies relative to that of the wild type strain BY4743 are determined. Results are shown in parenthesis in Tables 1 and 2. This retest showed that all tested deletion strains with increased sensitivity towards AMT in the microtiter plate based protocol also showed increased sensitivity in the standard protocol. However, for some strains the transformation efficiency determined by the two protocols differs to some extent. As can be seen in Table 1, in general the relative transformation efficiencies were higher using the standard transformation protocol than in the mictotiter plate based protocol.
LiAc transformation
To investigate whether the effects of deletion of the identified genes are specific to AMT or whether also transformation by other methods is affected, a number of deletion strains and the wild type strain BY4743 were transformed using the widely used lithium acetate transformation protocol. For this purpose these strains were
transformed with a replicating plasmid (YCplac33) with URA3 marker and with a linear DNA fragment containing the URA3 gene flanked by PDA1 sequences allowing integration at the PDA1 locus. As shown in figure 1 the altered transformation frequencies of these deletion strains observed for AMT (Tables 1 and 2) were not found for transformation using the lithium acetate protocol. For example, deletion of GCN5 and NGG1 resulted in 6-8 fold increased transformation efficiency for AMT using either integrative or replicating plasmid (Table 1), whereas only an about 2-fold increased efficiency was found for the lithium acetate transformation using integrative plasmid. For YAF9 and EAF7 deletions, a 5-7 fold increased transformation efficiency was found for AMT using either integrative or replicating plasmid (Table 1), whereas an unaffected or even a decreased efficiency was found for the lithium acetate transformation. Deletion of HDA3, HST4 and SIN3 greatly reduced AMT efficiency (Table 2), but had hardly any negative effect on the lithium acetate transformation efficiency (Figure 1). This indicates that the effects found for these deletion strains are specific to AMT.
0 1 2 3 4
WT ΔGCN5 ΔNGG1 ΔYAF9 ΔEAF7 ΔHST4 ΔHDA3 ΔSIN3
Yeast strains
R elativ e freq ue nc y
0 1 2 3 4 5 6 7 8 9 10
WT ΔGCN5 ΔNGG1 ΔYAF9 ΔEAF7 ΔHST4 ΔHDA3 ΔSIN3
Yeast strains
Relative frequency
Figure 1. Transformation efficiencies of selected S. cerevisiae deletion strains using the lithium acetate (LiAc) method (A) with either pUC4α10 (pda1::URA3) ( ) or YCplac33 (URA3/CEN4-ARS1) ( ), compared with the microtiter-based AMT (B) of the same strains using either pRAL7100 (pda1::URA3) ( ) or pRAL7101 (URA3/2µ Ori) ( ). Relative frequency of Ura+ colonies is frequency in the mutant strain / frequency of wild type strain. Data are averages of 3 independent experiments (Except for hst4, hda3 and sin3 deletions for AMT which are from 2 experiments). Error bars: standard deviation.
A
B
Effect of gene deletions on T-DNA integration by homologous recombination
For a number of deletion strains with an altered sensitivity towards AMT we verified whether integration by homologous recombination still occurred at the correct chromosomal location. The integrative binary vector pRAL7100 has homology with the yeast PDA1 locus and after AMT the T-DNA insert will integrate into this locus (Bundock et al. 1995). Integration of the T-DNA at the PDA1 locus was analyzed by PCR using primers specific for upstream and downstream sequences of PDA1 and for URA3. Transformants of the parental strain BY4743 and of strains with deletions of ADA2, ARG4, ARP6, EAF7, GPB2, HDA3, HST4, NGG1, SIN3, SPT8, SNF7, and YAF9, obtained after transformation with Agrobacterium carrying pRAL7100 were analyzed. The correct PCR fragments (1.5 and 2.5 kb) were detected in five of the six transformants of BY4743, in all five transformants of APR6Δ, in four of the five transformants of YAF9Δ and GPB2Δ, in three of the five transformants of ARG4Δ, HDA3Δ, NGG1Δ, SPT8Δ and EAF7Δ, in two of the five transformants of ADA2Δ, HST4Δ, and SNF7Δ, and only one of the five transformants of SIN3 strains.
Integration at the PDA1 locus was further analyzed by Southern blotting. To this end chromosomal DNA was isolated from several of the above mentioned transformants that scored negatively in the PCR assay, digested by SpeI and hybridized to a probe containing sequences just upstream of the PDA1 locus. PDA1 is located on a 5.3 kb SpeI fragment and after integration of the T-DNA at this locus the fragment becomes slightly smaller (4.8 kb). As shown in figure 2, the PDA1 probe hybridizes to the expected fragments of 5.3 and 4.8 kb of SpeI-digested chromosomal DNA from four transformants of the sin3 deletion strain, to three out of five transformants of HDA3, and to two out of four transformants of HST4. This indicates that the T-DNA is integrated into one of the PDA1 loci of the diploid strain. For several transformants of the other deletion strainsthe expected hybridizations were not found (Fig 2). Instead, the probe hybridized strongly to a fragment of approximately 8 kb. This was found for all three transformants analyzed of the ada2 deletion strain, two of hda3 and hst4, and one of spt8 and snf7 transformants. The T-DNA construct located between right and left borders in pRAL7100 is around 8 kb and contains one SpeI restriction site.
Therefore, we hypothesized that the 8 kb band observed on the Southern blot corresponds to circularized T-DNA present at multiple copies. To test this hypothesis we analyzed twelve additional ada2 transformants for growth on plates containing 5-
fluoro-orotic acid. These twelve transformants were all able to grow, indicating that the URA3 marker has not been stably integrated into the genome (data not shown). In addition, by PCR we were able to show the presence of the left border – right border junction in these transformants (data not show).
Figure 2. Southern blot analysis of transformants that didn’t give the signature for HR integration in PCR analyses from hda3, hst4, sin3, spt8, snf7, ada2, and ngg1 mutants. Extracted chromosomal DNA from each transformant was digested with Spe1 restriction enzyme. t1, transformant number 1, etcetera.
Functional grouping of the identified host genes
In this screen we identified 249 genes of which deletion resulted in at least two fold increased or decreased transformation efficiency (Tables 1 and 2). Grouping of these genes into 18 functional categories according to the MIPS revealed that genes involved in protein activity regulation (7.6%), cell cycle and DNA processing (6.6%), metabolism (5.9%), cellular communication/signal transduction (5.6%), transcription (5.3%) and protein fate (folding, modification, destination) (5%) are well represented in our screening (Table 3).
Categorization of the identified genes using FunSpec (http://funspec.med.utoronto.ca/) shows that in hypersensitive mutants, genes involved in regulation of C-compound and carbohydrate utilization, mRNA synthesis, transcriptional control and metabolism are overrepresented, and that in hyposensitive mutants, genes involved in nitrogen and sulfur metabolism, amino acid biosynthesis, DNA repair and metabolism are overrepresented (Tables 4 and 5). In a systematic study on the properties of the S. cerevisiae deletion strains it was shown that disruption of 1105 of the total 5916 S. cerevisiae genes (18.7%)
10 kb 8 kb 6 kb 4 kb 5 kb 3 kb
t2 t3 t4
Δada2 Δsnf
Δspt8 Δngg1
t3 t4 t5
Δsin3 Δhst4
Δhda3
t1 t2 t3 t4 t5 t1 t2 t3 t4 t5 t2 t1 t2 t3 t3
DNA ladder DNA
ladder
Table 3. Functional analysis of identified deletion mutants according to MIPS entries (http://mips.gsf.de/proj/yeast/CYGD/db).
a. Number of identified genes in the given category b. Percentage of identified genes
Table 4. Main categories overrepresented in mutants with increased AMT according to FunSpec analysis. (http://funspec.med.utoronto.ca/)
Category P-value k f
Regulation of C-compound and carbohydrate utilization 1.38e-05 12 120
mRNA synthesis 3.11e-05 23 406
Transcriptional control 0.00015 19 334
Metabolism 0.00094 39 1066
P-value: The probability that the intersection of a given list with any given functional category occurs by chance.
k: Number of genes
f: Total number of genes in the given category
Table 5. Main categories overrepresented in mutants with decreased AMT according to FunSpec analysis. (http://funspec.med.utoronto.ca/)
Category P-value k f
Nitrogen and sulfur metabolism 1.78e-06 9 67
Amino acid biosynthesis 4.89e-06 11 118
DNA repair 1.76e-05 9 88
Metabolism 5.99e-05 35 1066
P-value: The probability that the intersection of a given list with any given functional category occurs by chance.
k: Number of genes
f: Total number genes in the given category
MIPS category Total number
of genes in given category
Hyper- sensitive to
AMT
Hypo- sensitive to
AMT
No.a %b No.a %b
1 Biogenesis of cellular components 865 21 2.4 22 2.5 2 Cellular communication/Signal transduction
mechanism 234 7 3.0 6 2.6
3 Cell cycle and DNA processing 1010 39 3.8 28 2.8 4 Cell fate (Cell aging) 28 1 3.6 0.0 0.0 5 Cell fate (Cell growth) 239 3 1.2 0.0 0.0.
6 Cell rescue, defense and virulence 554 13 2.3 14 2.5 7 Cellular transport, transport facilitation and
transport routes
1043 29 2.8 14 1.3
8 Cell type differentiation 454 11 2.4 11 2.4
9 Development (Systemic) 69 0.0 0.0 2 2.9
10 Energy 396 8 4.4 8 2.2
11 Interaction with the cellular environment 463 9 1.9 14 3.0
12 Metabolism 1521 45 2.9 46 3.0
13 Protein activity regulation 249 12 4.8 7 2.8 14 Protein fate (folding, modification,
destination) 1156 41 3.5 17 1.5
15 Protein synthesis 480 10 2.1 3 0.0
16 Protein with binding function or cofactor
requirement 1049 19 1.8 20 1.9
17 Transcription 1078 36 3.2 23 2.1
18 Unclassified proteins 1995 16 0.8 16 0.8
results in a reduced fitness on YPD (Giaever et al., 2002). We found that 55 of the 141 (39%) deletion strains with an increased AMT efficiency (Table 1) and 34 of the 108 (31%) deletion strains with a reduced AMT efficiency (Table 2) have a reduced viability on YPD.
Identified genes which are involved in histone acetylation, transcription and DNA repair
As shown in Table 1, many of the strains with the highest transformation efficiencies have deletions in genes involved in histone acetylation, transcription, DNA repair and chromosome remodeling. Deletions of genes encoding subunits of ADA-, SAGA- and SLIK- transcriptional regulatory histone acetyltransferase (HAT) complexes (NGG1, GCN5, SGF29, SGF73) and of the NuA4 HAT complex (VID21, EAF7) highly increased AMT efficiency by both integrative (6.0-8.3 fold) and replicating T-DNA (3.0-6.7 fold), whereas deletion of other non-essential genes encoding components of these HAT complexes ADA2, EAF6, SGF11, SGF29, SGF73, SPT3, SPT8, SPT20 and YAF9) resulted in more than 2.0-fold increased transformation efficiency. On the other hand, deletion of genes encoding subunits of yeast histone deacetylase complexes (HDACs)(HDA3, HST4, and SIN3) resulted in decreased AMT. Thus, histone acetylation seems to have a negative effect on AMT.
One of the main functions of HAT complexes is the regulation of RNA polymerase II- dependent transcription. Therefore, it is very interesting that deletion of a number of genes related to RNA polymerase II (PAF1, SRB8, SSN8 and TAF14) resulted in a highly increased AMT efficiency by both integrative (4.1-4.8 fold) and replicating T- DNA (2.0-3.5 fold), while deletion of genes encoding components of the HAP subcomplex of the RNA polymerase II elongator (ELP4, HAP2, and HAP4) caused a decreased AMT. In line with this, a number of transcription related genes were identified whose deletion either increases AMT, i.e. GAL80, GCN4, MSN2, or decreases it, i.e. AZF1, PHD1, SPT10, SPT21. Thus, the effects of deletion of genes involved in histone acetylation or deacetylation on AMT may be explained by an altered expression of one or more genes required for AMT. Alternatively, the acetylation status of the PDA1 locus may be changed in the HAT and HDAC mutants, resulting in an altered transformation frequency. On the other hand, also the transformation by Agrobacterium strains carrying pRAL7101 is affected. Otherwise,
it has been shown that the Arabidopsis VirE2 Interacting Protein 1 (AtVIP1) binds to plant histones, and may help chromatin targeting of the T-complex (Djamei et al., 2007, Loyter et al., 2005). Moreover, the VirD2 protein interacts with the TATA box- binding protein (TBP) and a nuclear kinase (Cak2M) (Bako et al., 2003). Hence, it is possible that a reduced acetylation may allow the T-DNA to integrate into the host chromosome more efficiently.
Interestingly, the Eaf6 and Yaf9 subunits of the NuA4 HAT complex are shared with the SWR1 chromatin remodeling complex. Deletion of EAF6 and YAF9 resulted in a 4.0-5.2 fold increased AMT efficiency with both integrative and replicating T-DNA (Table 1). Also deletion of genes encoding other subunits of the SWR1 complex (SWC3, SWC5, SWR1, VPS71, and VPS72) caused an increased AMT. Moreover, deletion of RSC2 encoding a subunit of the RSC chromatin remodeling complex, and deletion of SNF11 encoding a subunit of the SWI/SNF chromatin remodeling complex, resulted in hypersensitivity towards AMT. Yeast RSC, SWI/SNF and Swr1 complexes also direct histone exchange (Mizuguchi et al., 2004, Bruno et al., 2003).
The NuA4 complex acetylates H2A.Z after the incorporation into the chromatin by the Swr1 complex (Keogh et al., 2006). Furthermore, the Swr1 and the NuA4 complexes prevent the spreading of silencing of telomeric heterochromatin (Boudreault et al., 2003, Meneghini et al., 2003). It has been shown that chromatin remodeling complexes have a key role in the double strand break repair process (van Attikum et al., 2005). Hence, a fourth possibility is that decreased histone acetylation results in an altered double strand break (DSB) repair. Current models of DNA repair in the context of chromatin by both homologous recombination and non-homologous end-joining (NHEJ) implicate the NuA4 HAT complex, and the Ino80, Swr1 and RSC chromatin remodeling complexes in early stages of DSB repair, while DNA polymerases, histone deacetylases and the Rad52 epistasis group proteins, are all involved in later processes of DSB repair (Costelloe et al., 2006, Peterson and cote, 2004, van Attikum and Gasser, 2005). In our screen, mutants of the RAD52 epistasis group i.e. RAD50, RAD51, RAD54, RAD57, and XRS2 showed a decreased AMT efficiency. We expected to retrieve the rad52 mutant as a mutant with a strongly decreased AMT by the T-DNA directed towards integration by HR at the PDA1 locus, but the mutant labeled rad52 did not have the RAD52 deletion. Furthermore, strains with deletions in MMS22, MSH3, SLX8 and DPB3, which are also involved in DNA
repair, have an increased AMT efficiency. Taken together, these observations are in line with a role of the host DNA repair machinery in AMT. Since AMT efficiency using a replicating T-DNA is also increased in most of those mutants, multiple mechanisms may affect AMT in a more general manner.
Identified genes involved in protein sorting
Deletion of VPS36, encoding a component of the ESCRT-II (Endosomal Sorting Complex Required for Transport) complex involved in the MVB (Multi Vesicular Body) protein sorting pathway, resulted in approximately 6.0- and 3.0-fold increased transformation using Agrobacterium carrying pRAL7100 or pRAL7101, respectively.
Deletion of genes encoding components of the ESCRT-I complex, i.e. STP22 and VPS28, and of genes encoding components of the ESCRT-III complex, i.e. SNF7 and VPS20 and of genes encoding components of other complexes which are associated with ESCRT, i.e. APM3, and CHC1, resulted in an increased AMT efficiency. One of the functions of the ESCRT complexes is sorting of ubiquitinated membrane proteins destined for degradation (Slagsvold et al., 2006). Deletion of parts of these complexes may prevent degradation of proteins related to AMT. Furthermore, deletion of a number of other genes involved in protein sorting, i.e. ATG15, DFM1, RIM20, VAC8, VAC14, VPL23 and VPS66, resulted in increased AMT efficiency (Table 1). On the other hand, deletion of genes which are related to the Golgi and secretory pathway, i.e. ARF1, ERP3, GYP1, KES1, LST7, STV1, and YPT6 and of genes encoding the membrane-associated retromer complex, i.e. PEP8 and VPS29 resulted in a decreased AMT. Thus, while the functional ESCRT and its associated complexes seem to inhibit AMT, a functional Golgi secretory pathway and membrane-associated retromer complex seem to favor AMT.
Identified genes involved in proteolysis
Deletion of the F-box protein component of the SCF ubiquitin-ligase complex required for Cln1 and Cln2 degradation, i.e. GRR1, resulted in a 4.2-fold increased AMT. Deletion of CLN1 and CLN3 resulted in a decreased AMT. Cln1, Cln2 and Cln3 are cyclins that promote the G1- to S-phase transition (Richardson, 1989). We also observed that deletion strains of cln1, cln2 and cln3 had slightly more unbudded cells than the wild type BY4743 (data not shown). This may indicate that the cell
cycle has influence on AMT, as previously reported for plant cells (Villemont et al., 1997). Deletion of four other genes involved in proteolysis, i.e. UBP3, UBX6, NAS6 and RIM13 resulted in around 3-fold increased AMT. This may indicate that host functional proteolysis system has an inhibitory effect on AMT. An involvement of protein degradation in the transformation process is further supported by the observation that the VirF protein of Agrobacterium interacts with plant homologs of the yeast Skp1 protein (Schrammeijer et al., 2001), which is a component of the SCF ubiquitin-ligase complex, to direct the targeted proteolysis of VIP1 and VirE2 (Tzfira et al., 2004).
Identified genes involved in Metabolism
As shown in Tables 1 and 2, deletion of 91 genes involved in metabolism affected the AMT efficiency more than 2-fold. Deletion of THI21 encoding hydroxymethylpyrimidine phosphate kinase involved in the last steps of thiamine biosynthesis gave an around 6-fold increased AMT. Deletion of the other gene in this pathway, i.e. THI22 also increased AMT around 2.5-fold. Hence, thiamine biosynthesis might protect host cells against AMT, maybe by strengthening cell defense. Furthermore, it has been shown that disruption of purine biosynthesis greatly affects AMT (Roberts et al., 2003). In our screen, the ade1, ade4 and ade12 deletion mutants showed hypersensitivity towards AMT (using both plasmids) in medium deprived of adenine, whereas in the presence of adenine only ade12 deletion mutant was sensitive (Table 1). Indeed ade12 was the most sensitive adenine deletion mutant to AMT. There are two distinct subpathways in the purine biosynthesis pathway producing adenine and guanine. Since deletion of genes involved in the guanine biosynthesis subpathway did not affect AMT efficiency, and since ade12 is involved in regulating the adenine biosynthesis subpathway, disruption of adenine biosynthesis might be the key of hypersensitivity of adenine mutants to AMT. On the other hand, deletion of several genes involved in arginine biosynthesis, i.e. ARG1, ARG2, ARG4, and ECM40 and in methionine metabolism, i.e. MET1, MET3, MET8, MET10, MET14, and MET16 decreased AMT efficiency. This suggests that Agrobacterium benefits from the host arginine biosynthesis and methionine metabolism pathways. A number of other genes involved in metabolism also affected AMT (Table 1 and 2)
Identified genes involved in translation
Deletions of two components of the large (60S) ribosomal subunits, i.e. RPL16B and RPL40A, showed 4.0-4.5 folds increased AMT. Furthermore, deletions of RPS6A, RPS14A, and RPS27B, components of the small (40S) ribosomal subunits, as well as deletion of translation initiation factor eIF-5A, i.e. ANB1 also increased AMT between 2.3-3.8 folds. In contrast, deletion of RPL12B showed a decreased AMT.
Hence, interfering with the function of host translation machinery apparently affects the AMT efficiency.
Identified genes with mitochondrial function
Deletion of the mitochondria-related genes, RML2, FIS1, YIA6, and POR12 increased AMT, whereas deletion of other mitochondria-related genes, i.e. PET122, COX23, ILM1, MRPL39 and UPS1 decreased AMT efficiency. Deletion of these genes might affect the respiratory growth of yeast cells, thereby affecting the sensitivity of the host to AMT.
Effect of host nuclear envelope interactors on AMT
Deletion of the NUP120 gene encoding a subunit of the Nup84p subcomplex of the nuclear pore complex, and deletion of the gene encoding the Kap123 karyopherin beta, which mediates nuclear import, increased AMT. On the other hand, deletion of SOY1 which encodes a protein that is associated with the nuclear pore complex decreases AMT. Perinuclear compartments are implicated in various nuclear activities including transcription and DNA repair (reviewed in Akhtar and Gasser, 2007). The nuclear pores are also considered as the main gateways to the nucleus for T-DNA and Vir proteins. Hence, deletion of host nuclear pore proteins may affect the nuclear entry of T-DNA directly or affect the AMT through transcription and DNA repair indirectly.
Effect of host signal transduction pathways on AMT
Deletions of a number of genes involved in signal transduction pathways showed great effects on AMT efficiency. For example, deletion of IRA2, encoding a GTPase- activating protein that negatively regulates RAS, and deletion of RAS2, encoding a GTP-binding protein, showed 4.7- and 5.0-fold, respectively, increased AMT using Agrobacterium carrying pRAL7100. Furthermore, deletion of genes encoding protein
kinases, i.e. ATG1, NPR1, RIM15, SAK1 and YAK1, as well as the PHO81 cyclin- dependent kinase (CDK) inhibitor, and the heterotrimeric G protein GPB2, increased the AMT efficiency 3.0-4.6 fold. Also the deletion mutants of HSL1 and YCK3 kinases, and SOK2, a regulatory protein in the cAMP-dependent protein kinase signal transduction pathway, showed an increased AMT in a range of 2.2-2.8 fold. Since signal transduction pathways play important roles in the response of cells to environmental changes and defense mechanisms of the cell, deletion of these genes might prevent stress responses in response to the presence of Agrobacterium. On the other hand, three deletion mutants related to the mitogen-activated protein kinase (MAPK) pathway, i.e. BCK1, SLT2 and RLM1, as well as SOK1, involved in the cAMP-mediated signaling pathway showed a decreased AMT efficiency. It was recently shown that Agrobacterium uses the host MAPK defense signaling pathway to deliver the VirE2 protein into the host nucleus (Djamei et al., 2007).
Effect of host cell envelope and cytoskeleton on AMT
The first interaction between Agrobacterium and the host cell takes place at the cell envelope. Once, the T-complex is assembled in the host cytoplasm the cytoskeleton may lead it to the nucleus and the sites of integration (Citovsky et al., 2006). Two permeases, i.e. AGP3 and SAM3, two transporters, i.e. YEA4 and YOR1, a transmembrane protein, i.e. DFG16, and a subunit of the GET complex, GET2, when deleted resulted in increased AMT (Table 1). On the other hand, deletion of the CAN1 and UGA4 permeases, SKN1, a type II membrane protein, SMI1, and ROT2 which are involved in cell wall synthesis, and a lectin-like protein, i.e. FLO10, led to reduced AMT (Table 2). Interestingly, deletions of actin related proteins i.e. APR1, a subunit of the dynactin complex, BBC1, which is involved in assembly of actin patches, and the SAC6 fimbrin resulted in a decreased AMT. This may be due to interfering with the movement of the T-complex by the host motors, since it was shown in a cell-free system that T-complexes actively transport along microtubules in a dynein (but not kinesin) dependent manner (Salman et al., 2005).
RNA-related proteins
Deletions of NAM7, encoding an RNA helicase, of PUF2 and REF2, encoding RNA- binding proteins, of RNH202, encoding the RNase H2 subunit, and of TSR2 with a potential role in pre-rRNA processing increased AMT efficiency in a range of 3.0-4.2
fold for integrative T-DNA and 1.5-2.8 fold for the replicating one (Table 1). This signifies the inhibitory effects of these RNA related factors on AMT.
Miscellaneous
Several genes with different functions which were not a member of the complexes mentioned in above groups were categorized as miscellaneous (Table 1 and 2). In this group, deletions of APJ1, BSD2, ROX1, CAT8, CLG1, RGP1, UGA2, and STP4 genes resulted in increased AMT. Deletion of MAK3, MDS3, MGR2, SRC1, STB5, UBA4, UME1, and YRF1-6 resulted in decreased AMT.
Uncharacterized proteins
Deletion of 33 genes of uncharacterized function affected AMT (Table 1 and 2) Among these, deletion of YDL025C resulted in a 5.3 fold increased AMT with integrative T-DNA and a 7.0 fold increased AMT with the replicating T-DNA.
Moreover, the AMT efficiency increased in a number of deletion mutants, i.e. in which YCR045C, YDL023C, YKL023W, YMR100W, and YLL007C were deleted, in a range of 4.1-5.3 fold for integrative T-DNA and 3.0-3.5 fold for the replicating T- DNA. Among the deletions which decreased the AMT efficiency, YHR151C had the highest effects.
Effect of reduced fitness of yeast cells on AMT
Deletion strains with a reduced fitness on YPD are overrepresented in the set of deletion strains with an altered sensitivity towards AMT. However, reduced fitness is not the most important factor as only in 89 non-essential deletion mutants of 1105 deletion strains that have been described to have a reduced viability on YPD an altered AMT efficiency is seen, indicating that fitness is not the only factor determining AMT efficiency. Another factor could be a prolongation of some stages of the cell cycle in some of these deletion mutants as it has been shown that the AMT efficiency is cell cycle dependent (Villemont et al., 1997). In this respect it is of interest that deletion of CLN1 and CLN3 promoting the G1 to S transition and of the checkpoint genes RAD17 and MEC3 resulted in a decreased AMT (Table 2).
Reversely, the deletion of FAR10, which can release G1 cell cycle arrest, increased AMT 4.2-5.8 folds (Table 1).
Yeast interactors of Ty- transposon
The collections of yeast deletion strains have also been used to identify genes involved in transposition of the Ty- transposon (Griffith et al. 2003). Thirty-one of the identified genes in our screen also affect yeast transposon mobilization and integration. Indeed, mutants of EAF7, DBP3, ACS1, VPS27, VPS28, VPS36, VPS20, SNF7, RIM20, RIM13 reduced the transposition, while mutants of ADA2, SGF73, RSC1, BDF1, NGG1, GAL80, CLN2, NUP120, POP2, VAC8, YEA4 increased transposition (Griffith et al., 2003; Irwin et al., 2005; Scholes et al., 2001).
Furthermore, SIN3, SPT10, SPT21, MMS22, RAD52, XRS2, DPB3, RPL16B, SRB8, ELP4 and IKI3 change the retrotransposition activity of Ty1 in yeast (Griffith et al., 2003). Moreover, while in our screen the functioning of the RRD1, RPS6A, RPS14A, SSN8, PEP8, KAP123, CLN1, CLN3, SAC6, ATG1 and ATG15 genes turned out to affect AMT, their very close relatives, i.e. RRD2, RPL6A, RPL14A, SSN2, PEP7, KAP122, CLN2, SAC3 and ATG2 were seen to affect the retrotransposition of Ty elements in yeast (Griffith et al., 2003; Irwin et al., 2005; Scholes et al., 2001). It has also been shown that mutation in ECM17, RIM13 and YPL150W decreases the Ty3 transposition, whereas mutation in HSL1 increases it (Aye et al., 2004). Except for RIM13, this is also the case for AMT. That study also has shown the negative effect of mutation in GTR1, SHM1 on the Ty3 transposition, while in our screen we identified the same effect for GTR2 and SHM2 on AMT. In total, this may indicate a conserved role of those genes in foreign DNA integration.
Conclusion
In this study, it was found that mutations in 249 out of ~4800 yeast non-essential genes resulted in an altered AMT. Functional grouping showed that they have diverse functions all over the cell. An overview of these processes is given in figure 3. By screening libraries of Arabidopsis T-DNA insertion mutants and of RNAi lines in a root-based transformation assay 150 mutants were identified that showed a decreased AMT efficiency (Zhu et al., 2003a; Crane and Gelvin, 2007). As in our study, proteins encoded by those genes are involved in chromatin modification and remodeling, nuclear targeting, cell wall structure and metabolism, cytoskeleton structure and function, and signal transduction. However, in addition to the mutants in which AMT efficiency is decreased, we further have identified host mutants in which
AMT efficiency is increased. In line with the report of Roberts et al., (2003), yeast adenine mutants are identified in our screen and cause an increased AMT. Besides these, we have identified ade12 as a key adenine mutant for AMT, and a large number of other yeast effectors of AMT which were not reported by Roberts et al., (2003).
This difference could mainly be caused by the different method, different yeast and Agrobacterium strains carrying different vectors, and different cocultivation conditions. Taken together, both in yeast and in plants genes involved in many cellular processes are related to AMT. In both organisms disruption of genes involved in chromatin structure have severe effects on AMT. Recently, the construction of haploid and diploid collections of ~900 (out of ~ 1100) hypomorphic alleles of yeast essential genes, and a haploid collection of temperature-sensitive mutants of 250 yeast essential genes, has been reported (Ben-Aroya et al., 2008; Breslow et al., 2008).
Screening of such collections for AMT would shed light on the role of essential yeast genes on AMT.
In conclusion, we have identified genes involved in the different steps of the T- complex’s odyssey through the host cell. Since most of the identified genes are conserved among eukaryotes, the corresponding genes in other hosts may be involved in AMT as well. These findings may help to improve AMT of other hosts, including different plants and fungi, especially those hosts that are recalcitrant towards AMT, and to improve plant resistance to crown gall disease.
Table 1. S. cerevisiae genes of which deletion results in increased AMT efficiency. Gene Descriptiona Increased transformation (fold)b Systematic name Standard name Integrative plasmid (pRAL7100) Replicating plasmid (pRAL7101) Transcription, DNA repair, chromosome remodelling YDR448W ADA2 Component of the ADA and SAGA complexes 3.1±1.0 1.8±0.8 YGR252W GCN5Component of the ADA and SAGA complexes 6.7±1.5 6.0±1.7 YDR176W NGG1Component of the ADA and SAGA complexes 8.3±1.5 6.7±1.5 YPL047W SGF11 Component of the SAGA complex 2.4±0.5 1.0±0 YCL010C SGF29 SAGA associated Factor 6.0±2.6 3.0±2 YGL066W SGF73 SAGA associated Factor 7.1±2.2 3.7±1.5 YDR392W SPT3 Component of the SAGA and SAGA-like complexes 3.0±0.1 1.7±0.4 YLR055C SPT8 Component of the SAGA complex 3.0±0.1 (7.4) 2.0±1.0 (5.3) YOL148C SPT20 Component of the SAGA complex 3.0±0.1 2.3±1.0 YBR279W PAF1 RNA polymerase II-associated protein 4.6±1.5 3.0±1.0 YCR081W SRB8 Subunit of the RNA polymerase II mediator complex 4.3±2.0 2.7±1.2 YNL025C SSN8 Cyclin-like component of the RNA polymerase II holoenzyme 4.8±0.7 2.0±1.4 YPL129WTAF14 Subunit of TFIID, TFIIF, and SWI/SNF complexes 4.1±1.8 3.5±0.7 YML051WGAL80 Transcriptional regulator 3.5±0.5 3.4±2.7 YEL009C GCN4Transcriptional activator 1.7±0.252.5±0.7 YMR037CMSN2 Transcriptional activator 2.0±0.0 2.0±1.4 YNL139C RLR1 Subunit of the THO complex 3.0±0.1 1.8±0.3 YDR359CVID21 Component of the NuA4 complex 6.2±1.5 5.0±0.0 YJR082C EAF6 Component of the NuA4 and the SWR1 complexes 5.0±0.0 4.0±1.0 YNL136W EAF7 Component of the NuA4 complex 7.0±0.0 5.0±3.5 YNL107WYAF9 Component of the NuA4 and the SWR1 complexes 5.2±0.3 5.0±1.0 YLR085C ARP6 Nuclear actin-related protein involved in chromatin remodeling 1.7±0.4 (3.5) 1.2±0.2 YPL116W HOS3Trichostatin A-insensitive homodimeric histone deacetylase 1.8±0.3 2.2±1.0 YLR399C BDF1 Protein involved in transcription initiation at TATA-containing promoters 4.5±0.5 1.8±0.4 YAL011WSWC3 Component of the Swr1p complex 2.6±0.7 (17) 1.8±0.4 (2.6) YBR231CSWC5 Component of the Swr1p complex 2.7±0.6 2.5±2.2 YDR334W SWR1 Swi2/Snf2-related ATPase, component of the SWR1 complex 1.7±0.3 (3.0) 1.7±1.2 YML041C VPS71 Component of the Swr1p complex 3.8±0.3 1.5±0.7 YDR485CVPS72 Component of the Swr1p complex 2.2±0.76 2.0±1.4
