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).
Deletions of host histone acetyltransferases and histone deacetylases affect Agrobacterium-mediated
transformation
Abstract
Agrobacterium tumefaciens-mediated transformation (AMT) is widely in use to genetically transform a broad range of plant and non-plant organisms including the yeast Saccharomyces cerevisiae. By using this yeast as a model host we have shown that disruption of a large number of host genes affects the efficiency of AMT (Chapter 3). One of the major effects was found in deletion mutants lacking genes encoding components of chromatin modifying complexes, i.e. histone acetyltransferase and histone deacetylase complexes. Here, we have further investigated the effect of diploid deletion of genes encoding histone acetyltransferases and histone deacetylases on AMT. Our experiments indicate that deletion of the ELP3, GCN5, HAT1 and HPA2 genes encoding histone acetyltransferases increases AMT. The largest effect was found for the deletion of GCN5. Reversely, deletion of the histone deacetylase genes HDA2, HDA3, HST3, and HST4 decreases AMT efficiency. The largest effect was found for the deletion of HDA2. Increasing the copy number of the histone deacetylases genes HDA2, HDA3 and HST4 by cloning on a multicopy plasmid slightly increases the AMT efficiency. Our results suggest that histone acetylation is inhibitory to AMT.
Introduction
Agrobacterium-mediated transformation (AMT) is widely in use to genetically transform a broad range of plant and non-plant organisms including the yeast Saccharomyces cerevisiae (Soltani et al., 2008). The agrobacterial genes involved in AMT are relatively well known. However, the effects that host genes have in the transformation process are only known to a limited extent. Expression profiling and forward genetic screening for Arabidopsis mutants which are resistant to Agrobacterium-mediated transformation have revealed a large number of plant effectors of AMT (Crane and Gelvin, 2007; Ditt et al., 2001, 2005, 2006; Veena et al., 2003; Zhu et al., 2003). Recently, we performed a systematic screening for mutants which increase or decrease Agrobacterium-mediated transformation (Chapter 3). By using the complete collection of viable S. cerevisiae deletion strains we identified 249 genes of which deletion resulted in an at least two-fold increased or decreased AMT
efficiency (Chapter 3). Several key effectors of AMT turned out to be chromatin- related genes. Deletion of genes encoding subunits of histone acetyltransferase (HAT) complexes, i.e. SAGA, SLIK, ADA and NuA4 complexes highly increased AMT, while deletion of genes encoding subunits of histone deacetylase complexes, i.e.
HDA2-HDA3, HST3-HST4 and SIN3-RPD3 complexes decreased AMT. These effects are specific for Agrobacterium-mediated transformation as the efficiency of chemical (lithium acetate) transformation was not altered in these mutants (Chapter 3). Here, we have further investigated the effect of deletion of genes for histone acetyltransferases and deacetylases in the yeast BY4743 background on AMT.
Deletion of GCN5 led to the largest increase in AMT, while deletion of HDA2 reduced AMT the most.
Materials and methods Yeast strains and media
The Saccharomyces cerevisiae homozygous diploid deletion strains of HAT genes (ELP3, GCN5, HAT1, HPA2, RTT109) and of HDAC genes (HDA2, HDA3, HST1, HST2, HST3, HST4), and heterozygous diploid deletion strain of the HAT gene ESA1 in the BY4743 background (MATa/MATα his3Δ1/his3Δ1 leu2Δ0/leu2Δ0 lys2Δ0/LYS2 MET15/met15Δ0 ura3Δ0/ura3Δ0)(Giaever et al., 2002) as well as the isogenic BY4743 parental strain were obtained from Invitrogen (Groningen, the Netherlands).
The genes are described in Table 1. Yeast cells were 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
Two derivatives of Agrobacterium tumefaciens strain LBA1100, one containing the binary vector pRAL7100 and the other one containing the binary vector pRAL7101 (Bundock et al., 1995) were used in Agrobacterium-mediated transformation (Chapter 2, figure 1; Table 2). Agrobacterium strains were grown and maintained as described (Hooykaas et al., 2006). AMT of S. cerevisiae according to the standard protocol (Bundock et al., 1995) was performed as described by Hooykaas et al., 2006. All materials and media for AMT were prepared as described (Hooykaas et al., 2006).
Table 1. Description of S. cerevisiae genes of which homozygous diploid deletion mutants were used in this study.
Standard name Systematic name Description taken from SGDa WT BY4743 Parental Wild type
Histone acetyltransferases
ELP3 YPL086C Subunit of Elongator complex; exhibits histone acetyltransferase activity that is directed to histones H3 and H4
ESA1b YOR244W Catalytic subunit of NuA4 complex that acetylates four conserved internal lysines of histone H4 N-terminal tail
GCN5 YGR252W catalytic subunit of the ADA and SAGA histone acetyltransferase complexes; acetylates N-terminal lysines on histones H2B and H3 HAT1 YPL001W Catalytic subunit of the Hat1p-Hat2p histone acetyltransferase
complex that acetylates free nuclear and cytoplasmic histone H4 HPA2 YPR193C Tetrameric histone acetyltransferase; acetylates histones H3 and
H4 in vitro
RTT109 YLL002W acetylates histone H3 at K56 Histone deacetylases
HDA2 YDR295C Subunit of a possibly tetrameric trichostatin A-sensitive class II histone deacetylase complex containing an Hda1p homodimer and an Hda2p-Hda3p heterodimer
HDA3 YPR179C Subunit of a possibly tetrameric trichostatin A-sensitive class II histone deacetylase complex
HST1 YOL068C NAD(+)-dependent class III histone deacetylase;
HST2 YPL015C Cytoplasmic member of the Sir2 family of NAD(+)-dependent class III protein deacetylases
HST3 YOR025W Member of the Sir2 family of NAD(+)-dependent class III protein deacetylases
HST4 YDR191W Member of the Sir2 family of NAD(+)-dependent class III protein deacetylases
a) Saccharomyces Genome Database (http://www.yeastgenome.org).
b) ESA1 strain is a heterozygous diploid deletion mutant.
Construction of plasmids
Escherichia coli strain XL-1 Blue (supE44 hsdR17 recA1 endA1 gyrA46 thi relA1 lac− F′ [proAB+ lacIq lacZΔM15 Tn10] Tcr) was grown at 37°C in Luria-Bertani (LB) broth and was used for all cloning experiments. When required, antibiotics were added at the following concentrations: 40 µg/ml of kanamycin or 60 µg/ml of ampicilin.E. coli XL-1 blue was transformedby the heat shock protocol (Takahashi et al., 1992). Plasmids were isolated using a QIAprep Spinminiprep kit (QIAGEN) as recommended by the supplier. For complementation and overexpression experiments primers were designed to amplify the GCN5 (2.16 kb), HDA2 (2.8 kb), HDA3 (2.3 kb) and HST4 (1.7 kb) genes, including their promoter and terminator regions, and to add additional restriction sites for cloning. Primers are listed in Table 3. Chromosomal DNA from wild type BY4743 was isolated as described in Chapter 3. PCR amplifications werecarried out by using proofreading Vent polymerase (New England BioLabs)with thefollowing conditions: denaturation at 94°C for 3 min followed by
30 cycles of annealing for 1 min at 52°C (HDA2, HST4 and GCN5) or 60°C (HDA3), extensionfor 1 min per 1 kb DNA at 72°C, and denaturation for 1 min at 94°C. PCR products were gel-purified usinga QIAquick gel extraction kit (QIAGEN) according to the manufacturer'sprotocol. Purified PCR fragments were inserted into the pCR Blunt II TOPO vector (Invitrogen) as recommended by the manufacturer. DNA fragments containing GCN5, HDA2, HDA3 or HST4 genes were obtained by digestion with XbaI and SacI restriction endonucleases and were cloned into the single copy CEN plasmid pRS315 and into the multicopy 2µ plasmid p425TEF, digested with the same enzymes (Table 2). p425TEF is a TEF-promoter carrying derivative of pRS425 (Christianson et al., 1992; Mumberg et al., 1995). XbaI and SacI digestion of p425TEF removes the TEF promoter. All constructed plasmids were analyzed by restriction analyses using: EcoRI, SacI, XbaI, SacI/XbaI, BamHI, SpeI, BamHI/SpeI, and HindIII restriction endonucleases and by sequence analyses (BaseClear, The Netherlands). Using the lithium acetate transformation protocol (Gietz et al., 1995) the resulting pRS315-derived plasmids were transferred to the corresponding yeast deletion strains and the resulting pRS425-derived plasmids were transferred to wild type strain BY4743. The presence of the introduced plasmids was confirmed by plasmid isolation followed by restriction analyses. Allligations, restriction digestions and gel electrophoreses were performed using the standard techniques (Sambrook et al., 1989). Restriction enzymes were supplied by New England BioLabs (Beverly, MA).
Growth studies of yeast strains containing the new constructs
The yeast strains were grown overnight in YPD medium. The next day, cells were harvested and resuspended in fresh liquid MY medium supplemented with histidine (20 μg/ml), and uracil (20 μg/ml), [and leucine (30 μg/ml) for wild type strain] to OD620 of 0.05, and grown for 8 hours at 30°C (300 rpm). The cultures were sampled at 7 time points, and OD620 was measured. Moreover, after overnight growth the cells were also diluted to an OD620 of 0.1 and grown in microtiter plates for 4 hours at 30°C (900 rpm). Then, serial dilutions of each strain (0, 10-1, 10-2, 10-3, 10-4, 10-5, and 10-6) were plated on MY medium supplemented with the required nutrients as mentioned above. The plates were incubated at 30°C for 2-3 days.
Table 2. Plasmids used in this study.
Plasmid name Features Reference
pRAL7100 Binary vector (15.4 kb) derived from pBIN19 containing the S. cerevisiae URA3 gene flanked by PDA1 sequences between the left border and right border repeats; allows integration of URA3 at the PDA1 locus via homologous recombination
Bundock et al., 1995
pRAL7101 Binary vector (16.5 kb) derived from pBIN19 containing the S. cerevisiae URA3 gene and the 2µ origin of replication between the left border and right border repeats; allows replication after circularization of the plasmid
Bundock et al., 1995
pCR-Blunt-II TOPO A plasmid vector for cloning of blunt-end PCR products
Invitrogen
pRS315 A pBluescript-based centromere S.cerevisiae/E.coli phagemid vector, containing the LEU2 yeast selectable marker gene and CEN6 - ARSH4
Sikorsky & Hieter, 1989
pRS315.GCN5 pRS315 carrying GCN5 This study, pRUL1136 pRS315.HDA2 pRS315 carrying HDA2 This study, pRUL1137 pRS315.HDA3 pRS315 carrying HDA3 This study, pRUL1138 pRS315.HST4 pRS315 carrying HST4 This study, pRUL1139 p425TEF A pBluescript-based 2µ S.cerevisiae/E.coli phagemid
vector, containing the LEU2 yeast selectable marker gene, 2µ replicator and TEF promoter
Mumberg et al., 1995
pRS425.GCN5 pRS425 carrying GCN5 at XbaI/SacI restriction site This study, pRUL1142 pRS425.HDA2 pRS425 carrying HDA2 at XbaI/SacI restriction site This study, pRUL1143 pRS425.HDA3 pRS425carrying HDA3 at XbaI/SacI restriction site This study, pRUL1144 pRS425.HST4 pRS425 carrying HST4 at XbaI/SacI restriction site This study, pRUL1145
Table 3. Primers used to amplify the corresponding genes.
Gene name Primer Sequences 5’-3’
Fw ATTCTAGAATCTTAAACACTTATGGGCAGC GCN5
rev TAGAGCTCTCCAGAAGAAGCGGATGTTG Fw ATTCTAGAACCTTCATGCTTTTGCCCTG HDA2
rev TAGAGCTCAACATGAAATAATGCACCAGAG Fw ATTCTAGAGTTACTGATTTAATCCACTCAG HDA3
rev TAGAGCTCAAGTTTAGCAGGCCACACTG Fw ATTCTAGATGACTTCACTGATAGCGACC HST4
rev TAGAGCTCTGCATTGGTTGCATAGTAAGG
Results
Effect of deletion of genes encoding histone acetyltransferases and histone decetylases on AMT efficiency
By screening the collection of viable S. cerevisiae deletion strains we found that deletion of several genes encoding components of histone acetyltransferase complexes resulted in an increased sensitivity towards AMT, whereas deletion of several genes encoding histone deacetylase subunits resulted in a decreased sensitivity (Chapter 3). In order to investigate the role of histone (de) acetylation in more detail, we reanalyzed the histone acetyltransferase (HAT) and deacetylase (HDAC) deletion strains using the standard AMT protocol. To this end, homozygous diploid deletion strains of the HAT genes ELP3, GCN5, HAT1 and HPA2, of the HDAC genes HDA2, HDA3, HST1, HST2, HST3 and HST4, and the parental BY4743 strain were transformed using two derivatives of A. tumefaciens LBA1100. These Agrobacterium strains contain either the binary vector pRAL7100 of which the T-DNA integrates into the genome by homologous recombination or pRAL7101 of which the T-DNA can replicate as a plasmid in yeast. As ESA1, encoding a HAT, is essential, the heterozygous ESA1 deletion strain was used in this case. Data from AMT experiments show that there is a large variation in the transformation frequencies between the different experiments for both HAT (Table 4) and HDAC (Table 5) deletion strains.
However, it is clear that deletion of GCN5, which acetylates lysine residues on histone H3 and H2B, and has a global role in transcription activation, increases AMT by both the integrative (Table 4A) and replicating (Table 4B) T-DNA by about 10 fold. Deletion of HAT1, encoding a protein that acetylates histone H4, and deletion of HPA2 encoding a protein that acetylates lysine residues on histone H3 and H4, also resulted in an increased AMT. Deletion of ELP3, encoding a subunit of the elongator complex that acetylates histones H3 and H4 seems to increase AMT frequency as well, although the results are fairly variable. Interestingly, also deletion of one copy of the ESA1 gene, encoding the acetyltransferase subunit of the NuA4 complex which acetylates lysine residues on histone H4 and H2A, resulted in an increased AMT frequency. Recently, RTT109 was shown to encode a HAT acetylating specifically the histone H3 lysine 56 residue (Driscoll et al., 2007; Han et al., 2007; Schneider et al., 2006). Therefore, we also investigated the effect of deletion of this gene on AMT efficiency. In two independent experiments we did not find a significant effect of this deletion on AMT efficiency (average frequency of 1.7×10-5 versus 1.5×10-5 (Wt) for
AMT with pRAL7100, and average frequency of 1.9×10-5 versus 2.3×10-5 (Wt) for AMT with pRAL7101).
Table 4. Agrobacterium-mediated transformation of yeast histone acetyltransferase deletion mutants. Yeast strains were transformed by A. tumefaciens LBA1100 containing either pRAL7100 (A) or pRAL7101 (B) binary vectors.
A pRAL7100 Yeast strain AMT frequency (×10-5)a
Experiment 1 Experiment 2 BY4743 Wt 0.5 (100) 0.8 (100)
ΔELP3 0.2 (40) 2.5 (312)
ΔESA1(heterozygous) 1.0 (200) 2.5 (312)
ΔGCN5 7.0 (1400) 6.4 (800)
ΔHAT1 1.3 (260) 2.0 (250)
ΔHPA2 - 2.3 (287)
B pRAL7101 Yeast strain AMT frequency (×10-5)a
Experiment 1 Experiment 2 Experiment 3 BY4743 Wt 0.2 (100) 1.7 (100) 0.5 (100)
ΔELP3 0.8 (400) 3.2 (188) 1.1 (220) ΔESA1(heterozygous) 2.4 (1200) 1.7 (100) 0.9 (180) ΔGCN5 11.8 (5900) 7.1 (418) 0.8 (160) ΔHAT1 4.2 (2100) 3.0 (176) 1.5 (300) ΔHPA2 4.4 (2200) 2.4 (141) -
a, AMT frequencies are depicted as the number of Ura+ colonies divided by the output number of each yeast strain after transformation; Relative frequency which is frequency of the strain / frequency of wild type BY4743 ×100 is given in parenthesis.
Data from AMT experiments with the HDAC mutants indicate that in the absence of HDA2 and HDA3 AMT with both integrative (Table 5A) and replicating (Table 5B) T-DNAs is strongly decreased. Results obtained with the HST3 and HST4 deletion mutants were variable. In most experiments a decrease in the AMT efficiency was seen, but in some experiments there was no effect, or even a positive effect was seen (Table 5).
Table 5. Agrobacterium-mediated transformation of yeast histone deacetylase deletion mutants. All yeast strains were transformed by A. tumefaciens LBA1100 containing either pRAL7100 (A) or pRAL7101 (B) binary vectors.
A pRAL7100 Yeast strain Frequency (×10-5)a
Experiment 1 Experiment 2 Experiment 3 BY4743 Wt 0.70 (100) 0.30 (100) 0.80 (100)
ΔHDA2 0.04 (5) 0.02 (6) 0.01 (1) ΔHDA3 0.08 (11) 0.16 (55) 0.16 (20)
ΔHST1 1.08 (154) 0.56 (186) 0.35 (44) ΔHST2 0.21 (30) 0.79 (263) 0.75 (94) ΔHST3 0.05 (7) 0.36 (120) 0.16 (20) ΔHST4 0.24 (34) 0.05 (17) 0.13 (16)
B pRAL7101 Yeast strain Frequency (×10-5)a
Experiment 1 Experiment 2 Experiment 3 BY4743 Wt 0.40 (100) 0.40 (100) 0.60 (100)
ΔHDA2 - 0.05 (12) 0.04 (7)
ΔHDA3 0.09 (22) 0.01 (2) 0.35 (58) ΔHST1 0.48 (120) 0.40 (100) 0.83 (138) ΔHST2 2.48 (620) 0.32 (80) 0.40 (67) ΔHST3 0.51 (127) 0.47 (117) - ΔHST4 0.60 (150) 0.12 (30) 0.45 (75)
a, AMT frequencies are depicted as the number of Ura+ colonies divided by the output number of each yeast strain after transformation; Relative frequency which is frequency of the strain / frequency of wild type BY4743 ×100 is given in parenthesis.
Complementation experiments
In order to find out whether the alterations in the frequencies of AMT seen with the HAT and HDAC mutants were indeed due to the described mutations, we performed complementation experiments. To this end, we introduced the complementing gene on the pRS315 plasmid into the mutants. As a control, also the vector alone was introduced. Subsequently, the standard AMT protocol (Bundock et al., 1995) was used to transform these strains with two derivatives of A. tumefaciens strain LBA1100 containing either homologous integrative T-DNA (pRAL7100) or replicating T-DNA (pRAL7101) plasmids. The results of these AMT experiments are shown in Table 6.
Table 6. Effect of complementation of selected genes deletions on frequency of Agrobacterium-mediated transformation. All yeast strains were transformed by A.
tumefaciens LBA1100 containing either pRAL7100 (A) or pRAL7101 (B) binary vectors.
A pRAL7100
Yeast strain Frequency (×10-5)a Frequency relative to Wt Experiment 1 Experiment 2 Average
BY4743. pRS315 5.3 6 5.6 100
ΔGCN5. pRS315 51.8 (100) 28.4 (100) 40.1 (100) 712 ΔGCN5. pRS315.GCN5 7.4 (14) 9.7 (34) 8.6 (21) 152 ΔHDA2. pRS315 1.0 (100) 0.8 (100) 0.9 (100) 16 ΔHDA2. pRS315.HDA2 4.6 (446) 4.2 (525) 4.4 (493) 79 ΔHDA3. pRS315 3.1 (100) 1.9 (100) 2.5 (100) 44 ΔHDA3. pRS315.HDA3 6.8 (219) 8.2 (431) 7.5 (301) 133 ΔHST4. pRS315 5.6 (100) 5.4 (100) 5.5 (100) 98 ΔHST4. pRS315.HST4 4.1 (73) 4.2 (78) 4.2 (75) 74
B pRAL7101
Yeast strain Frequency (×10-5)a Frequency relative to Wt Experiment 1 Experiment 2 Average
BY4743. pRS315 2.8 2.3 2.5 100
ΔGCN5. pRS315 35.8 (100) 24.4 (100) 30.1 (100) 1205 ΔGCN5. pRS315.GCN5 2.8 (8) 14.2 (58) 8.5 (28) 339 ΔHDA2. pRS315 0.4 (100) 1.0 (100) 0.7 (100) 28 ΔHDA2. pRS315.HDA2 1.8 (486) 2.8 (269) 2.3 (330) 92 ΔHDA3. pRS315 0.3 (100) 0.9 (100) 0.6 (100) 24 ΔHDA3. pRS315.HDA3 0.9 (300) 1.4 (155) 1.2 (201) 48 ΔHST4. pRS315 3.6 (100) 4.9 (100) 4.0 (100) 161 ΔHST4. pRS315.HST4 3.2 (89) 4.0 (82) 3.4 (81) 136
a, frequencies are depicted as the number of Ura+ colonies divided by the output number of yeast cells;
Relative frequency i.e. frequency of the complemented strain / frequency of its empty vector control is given in parenthesis.
As can be seen in Table 6A and 6B, introduction of the GCN5 gene in the gcn5 deletion mutant lowered the frequency of AMT to about the levels seen with the wild type. Introduction of the vector alone did not have this effect. This shows that the increased AMT seen with the gcn5 mutant is indeed due to the absence of GCN5 gene. It can be seen in Table 6 that also introduction of HDA2 and HDA3 into the respective mutants restores the frequency of AMT to the wild type level. For reasons unknown, the hst4 deletion mutant carrying the pRS315 empty vector show similar (with pRAL7100) or even higher (with pRAL7101) transformation frequencies than the wild type strain (Table 6), while the deletion strains lacking plasmids show lower transformation frequencies in most experiments (Table 5). Therefore, the effect of complementation could not be determined.
We noticed that the gcn5 mutant did not grow as well as the wild type and therefore we checked the growth of several mutants. As can be seen in Figure 1A, whereas the gcn5 mutant grows more slowly than the wild type, the other mutants grow similarly as the wild type. Comparable effects were seen in liquid cultures (data not shown).
Complementation of the gcn5 mutant with the wild type GCN5 gene leads to a restoration of the growth rate both on solid media (Figure 1B) and in liquid culture (data not shown).
A B C
Figure 1. Growth on minimal media of yeast deletion strains. A, Growth of yeast deletion strains. B, the Δgcn5 containing pRS315 alone and pRS315 carrying GCN5 gene. C, Growth on minimal media of yeast wild type strain BY4743 containing p425TEF and pRS315 alone, and pRS425 carrying selected genes
Overexpression experiments
In order to find out whether overexpression of HAT (GCN5) and HDAC (HDA2, HDA3, HDA3) genes would have an opposite effect on AMT as that of the mutation, we introduced those genes on the multicopy vector pRS425 in the wild type strain.
We first checked the effect of this overexpression on the growth rate. As can be seen
in Figure 1C, growth of the overexpressors was normal on solid media. Similar effects were seen in liquid cultures (data not shown). Subsequently, these strains were transformed with two derivatives of A. tumefaciens strain LBA1100 containing either homologous integrative T-DNA (pRAL7100) or replicating T-DNA (pRAL7101) plasmids. As shown in Table 7, the frequency of transformation with Agrobacterium carrying pRAL7100 was slightly higher for strains carrying HDA2, HDA3 and HST4 on a multicopy vector than for the strain carrying an empty vector. A minor decrease in AMT frequency using strains with pRAL7101 was observed for strains carrying GCN5 on the multicopy vector.
Table 7. Effect of overexpression of selected yeast genes on frequency of Agrobacterium- mediated transformation. All yeast strains were transformed by A. tumefaciens LBA1100 containing either pRAL7100 or pRAL7101 binary vectors.
A pRAL7100
Strain Frequency (×10-5)a
Experiment 1
Experiment 2
Experiment 3
Experiment 4
BY4743 Wt - 3.0 2.3 3.1
BY4743. p425TEF 2.8 (100)c 2.1 (100) 2 (100) 3.4 (100) BY4743. pRS425. GCN5 2.1 (75) 2.6 (124) 4.7 (172) 2.3 (68) BY4743. pRS425. HDA2 3.2 (114) 4.3 (223) 4.0 (163) 4.7 (138) BY4743. pRS425. HDA3 5.6 (200) 3.2 (191) 7.3 (238) 5.4 (159) BY4743. pRS425. HST4 3.3 (118) 3.6 (346) 4.5 (269) 4.8 (141)
B pRAL7101
Strain Frequency (×10-5)a
Experiment
1 Experiment
2 Experiment
3 Experiment 4
BY4743 Wt 1.4 2.7 2.0 1.4
BY4743. p425TEF 3.0 (100)c 2.1 (100) 2.0 (100) 3.0 (100) BY4743. pRS425. GCN5 2.9 (97) 1.1 (52) 1.2 (58) 1.5 (51) BY4743. pRS425. HDA2 2.3 (77) 3.1 (148) 4.1 (207) 2.0 (68) BY4743. pRS425. HDA3 3.8 (127) 2.4 (114) - 2.8 (93) BY4743. pRS425. HST4 - 2.8 (133) 2.6 (128) 1.5 (50)
a, frequencies are depicted as the number of Ura+ colonies divided by the output number of yeast cells;
Relative frequency, i.e. frequency of the strain / frequency of the empty vector control, is given in parenthesis.
Discussion
By forward genetic screening of the yeast S. cerevisiae homozygous diploid deletion collection in the BY4743 background we found 249 yeast genes whose deletion affects AMT (Chapter 2 and Chapter 3). Major effects were foundin deletion mutants lacking components of chromatin modifying complexes, i.e. histone acetyltransferase
(HAT) SAGA, SLIK, ADA and NuA4 complexes and histone deacetylase complexes (HDAC). Hence, we were interested in further analyzing the effect of deletion of yeast HAT and HDAC subunits on AMT. Here, we reanalyzed the viable yeast histone acetyltransferase and histone deacetylase deletion mutants in the BY4743 background. To quantify the relative frequency of AMT, we made use of our standard AMT protocol (Bundock et al., 1995). Homozygous deletion mutants of five HAT genes (ELP3, GCN5, HAT1, HPA2 and RTT109), of two class II HDAC genes (HDA2 and HDA3) and of four class III HDAC gens (HST1, HST2, HST3 and HST4) were selected for further analyses (Table 1). Because ESA1 codes for the histone acetyltransferase activity of the NuA4 complex and is an essential gene we also performed AMT on its heterozygous diploid deletion strain. Deletion mutants of the HDA1, RPD3 and SIR2 HDAC genes were not present in the yeast strain collections.
Using the microtiter-based AMT protocol the gcn5 deletion strain was found to have an increased AMT efficiency with both integrative and replicating plasmids (Chapter 3). Using the standard AMT protocol we now have confirmed that deletion of GCN5 results in a strongly increased AMT efficiency (Table 4). Complementation of the gcn5 deletion strain by the wild type GCN5 gene lowers AMT frequency, indicating that the effect was indeed caused by deletion of the GCN5 gene. Gcn5 is the HAT subunit of at least three large complexes, i.e. SAGA, ADA and SLIK, which are involved in RNA polymerase II transcription (for review: Nagy and Tora, 2007). The SAGA complex is required for optimal transcription elongation, and besides for mRNA export and probably for nucleotide excision repair (for review: Baker and Grant, 2007). Deletion of other SAGA subunits also increases AMT (Chapter 3), although to a lesser extent. In vitro, Gcn5 mainly acetylates histones H2B and H3, whereas in vivo GCN5 is required for acetylation of histones H3 and H4 (Zhang et al, 1998). The molecular mechanisms leading to the increased AMT in gcn5 deletion mutants remain unclear. Deletion of GCN5 may affect transcription of genes relevant for AMT and/or of genes involved in defense mechanisms. In this respect it is of interest that it was found that of the 4912 genes analyzed 185 transcripts were reduced by 2-fold or more and 83 increased by 2-fold or more in a gcn5 deletion mutant (Holstege et al., 1998). These genes belong to different functional categories. The most strongly affected genes were not found in our screen for deletion mutants with
SRB8. The YAK1 gene encodes a serine-threonine protein kinase and its expression is 3.2-fold lower in the gcn5 deletion strain than in the wild type, whereas deletion of YAK1 resulted in a 3.8- and 3.3-fold higher AMT efficiency using Agrobacterium strains carrying pRAL7100 and pRAL7101, respectively (Chapter 3, Table 1).
However, decreased expression of YAK1 can at the best only partly explain the strongly increased AMT of gcn5 mutants. The expression of SRB8, encoding subunit of the RNA polymerase II mediator complex, was increased 4.9-fold in the gcn5 deletion mutant (Holstege et al., 1998), whereas AMT efficiency was 4.3- and 2.7- fold higher for the srb8 deletion mutant than for the wild type using Agrobacterium strains carrying pRAL7100 and pRAL7101, respectively (Chapter 3, Table 1). Thus, deletion of gcn5 may disturb a complicated gene expression network ultimately resulting in increased AMT efficiency. On the other hand, disturbing Gcn5-containing complexes may affect the DNA repair pathways, possibly influencing T-DNA integration. Furthermore, the absence of active Gcn5-containing complexes may lead to an altered chromatin structure favorable for T-DNA integration.
We found that deletion of two other HAT genes, HAT1 and HPA2, also leads to increased AMT efficiency, although the effect was less than for the gcn5 deletion (Table 4). Hat1, which acetylates histone H4, is linked to histone deposition and DNA double-strand break repair (Qin and Parthun, 2002, 2006). Hpa2 is a tetrameric histone acetyltransferase which acetylates histones H3 and H4 in vitro and exhibits autoacetylation activity (Angus-Hill et al., 1999). In most experiments, AMT efficiency was slightly increased for the ELP3 deletion mutant and for the heterozygous ESA1 deletion mutant. Elp3 is a subunit of the elongator complex which exhibits histone acetyltransferase activity on histones H3 and H4. The ESA1 gene is essential in yeast and encodes the catalytic subunit of the NuA4 HAT complex which is linked to global and targeted histone H4 acetylation, regulation of transcription, cell-cycle progression and to the DNA double-strand break repair (Doyon and Côté, 2004). On the other hand, deletion of RTT109 does not affect AMT. In contrast to the above mentioned HATs, which acetylate the N-terminal tails of histones, Rtt109 acetylates lysine 56 within the core domain of histone H3 in a cell cycle-specific manner (Driscoll et al., 2007; Han et al., 2007; Schneider et al., 2006; Xu et al., 2005). Hence, the acetylation status of lysine residues in the histone tails may be more important for AMT than the acetylation status of internal lysine residues.
Surprisingly, deletion of a putative HAT gene, SPT10, resulted in a decreased instead of increased AMT in microter-based AMT assays (Chapter 3). Although this observation still has to be confirmed by transformation using our standard transformation protocol, it indicates that Spt10 is involved in AMT differently from the other HATs.
Deletion of HDAC genes HDA2 and HDA3 significantly decreased the AMT efficiency with both integrative and replicating binary vectors (Table 5).
Complementation of the hda2 and hda3 disruption strains by the wild type alleles restored AMT efficiencies. In addition, increasing the copy number of HDA2 and HDA3 by cloning on a multicopy plasmid somewhat increased transformation by Agrobacterium carrying the integrative plasmid. HDA2 and HDA3 encode subunits of a class II histone deacetylase complex containing an Hda1p homodimer and an Hda2p-Hda3p heterodimer (Wu et al., 2001). Results of the experiments to study the effect of deletion of the class III HDAC genes, HST1, HST2, HST3 and HST4 are inconclusive, although deletion of HST4 seems to have a negative effect on AMT efficiency (Table 5).
The results of the experiments described in this Chapter together with the results of the large-scale screening of yeast deletion strains (Chapter 3) suggest that histone acetylation inhibits AMT and histone deacetylation supports it. The effects were found for transformation by Agrobacterium strains carrying both pRAL7100 allowing T-DNA integration by homologous recombination and pRAL7101 allowing the formation of a self-replicating plasmid after circularization of the T-DNA. This indicates that the effects of disturbing histone acetylation are not restricted to the DNA integration process. Future research should show the mechanisms by which AMT is affected by HATs and HDACs.
