|Home | About | Journals | Submit | Contact Us | Français|
Although Sgf29p has been biochemically implicated as a component of SAGA (Spt-Ada-Gcn5 acetyltransferase), its precise mechanism of action in transcription is not clearly understood in vivo. Here, using a formaldehyde-based in vivo crosslinking and chromatin immunoprecipitation (ChIP) assay in conjunction with transcriptional and mutational analyses, we show that Sgf29p along with other SAGA components is recruited to the upstream activating sequence (UAS) of a SAGA-regulated gene, GAL1, in an activation-domain dependent manner. However, Sgf29p does not alter the recruitment of Spt20p that maintains the overall structural and functional integrity of SAGA. The recruitment of other SAGA components such as TAF10p, TAF12p, and Ubp8p to the GAL1 UAS is also not altered in the absence of Sgf29p. Interestingly, we find that the recruitment of TBP (TATA-box binding protein that nucleates the assembly of general transcription factors to form the pre-initiation complex for transcriptional initiation) to the core promoter of GAL1 is decreased in Δsgf29. Likewise, Sgf29p also enhances the recruitment of TBP to other SAGA-regulated promoters. Such reduction of TBP recruitment to these promoters subsequently decreases transcription. Taken together, these results support that SAGA-associated Sgf29p facilitates the recruitment of TBP (and hence transcription) without altering the global structural integrity of SAGA in vivo.
SAGA is a large multiprotein complex with histone acetyltransferase (HAT) and deubiquitinase activities (1, 2). In yeast, SAGA has at least fourteen non-essential and six essential components, and is required for transcription of approximately 10% genes (3). The non-essential components include Gcn5p-HAT, Ubp8p-histone deubiquitinase, Ada1p, Ada2p, Ada3p, Spt3p, Spt7p, Spt8p, Spt20p, Sgf11p, Sgf29p, Sgf73p, Sus1p and Chd1p. The essential components of SAGA include the ATM/PI-3-kinase-related protein Tra1p and a set of TBP associated factors (TAFs). SAGA’s global structural integrity is maintained by Ada1p, Spt7p and Spt20p (1, 2, 4, 5), and it is targeted to the promoter via interaction of its Tra1p component with transcriptional activators (6–9). For several yeast SAGA subunits, conserved homologues are found in mammals. For example, TRRAP (transformation/transcription domain-associated protein) has been identified as a mammalian homologue of yeast Tra1p. Like yeast Tra1p, TRRAP has been reported to be the target of several activators (10). Moreover, there is a TRRAP-containing mammalian complex, STAGA (SPT3-TAFII31-GCN5L acetylase), which is a homologue of yeast SAGA complex (11). Although many conserved homologues of yeast SAGA components are found in humans, homologues of yeast Spt8p and Chd1p are not yet identified in human SAGA or STAGA (12). Like in yeast and humans, the SAGA complex is also present in Drosophila (12). However, the homologues of yeast Spt8p, Spt20p, TAF6p, Sgf73p, and Chd1p have not yet been identified in Drosophila (12). Similar to the yeast and human SAGA complexes, Drosophila SAGA is also targeted by acidic activators (12, 13). Additionally, histone covalent modifications are also involved in recruiting SAGA to the active gene. For example, SAGA interacts with acetylated-histone H3, and such an interaction is mediated by the bromodomain of its Gcn5p component (1, 14–21). Similarly, the chromodomain of the Chd1p component of yeast SAGA interacts with di- and trimethylated-K4 (lysine 4) of histone H3 (22–24), and plays an important role in recruitment of SAGA onto chromatin. Collectively, these studies have demonstrated that activators and histone covalent modifications play crucial roles in recruiting and stabilizing SAGA onto promoter. Interestingly, a very recent structural and biochemical study (25) has revealed that yeast SAGA interacts with methylated-K4 of histone H3 via the tudor domain of its Sgf29p component. Consistently, this study further demonstrated that the absence of histone H3K4 methyltransferase lowers the recruitment of SAGA (25). Likewise, another recent study in human cell lines has also revealed the interaction of Sgf29p with methylated-K4 of histone H3 (26). Thus, the recognition of histone H3 K4 methylation by SAGA via its Sgf29p component appears to be evolutionary conserved among eukaryotes.
How is the recruitment of SAGA correlated with the association of TBP [which nucleates the assembly of general transcription factors for pre-initiation complex (PIC) formation to initiate transcription] with promoter? Previous biochemical and genetic studies (5, 27, 28) have implicated the interaction of SAGA with TBP via its Spt3p and Spt8p components. Consistent with these studies, the ChIP experiments have demonstrated the requirement of Spt3p and Spt8p for recruitment of TBP at several SAGA-regulated promoters such as ADH1, PHO84, BDF2 and VTC3 (4). However, Spt3p and Spt8p have also been implicated in inhibiting TBP recruitment at several other promoters such as HIS3, TRP3 and HO (29, 30). Intriguingly, the ChIP experiments further revealed that Spt3p, but not Spt8p, is required for TBP recruitment at the promoter of a well-characterized SAGA-dependent gene, GAL1 (4). Together, these results support that Spt3p and Spt8p perform distinct functions in recruiting TBP. Further, Qiu et al (31) have demonstrated that SAGA can promote the recruitment of RNA polymerase II independently of TBP. In addition to its role in recruitment of TBP and RNA polymerase II, SAGA also regulates transcriptional initiation via its HAT (1, 4, 32–36) and histone H2B deubiquitinase (37–42) activities. Together, these studies reveal a complex regulation of transcriptional initiation by SAGA. Further, recent studies have revealed SAGA’s participation in transcriptional repression at the telomere, transcriptional elongation, mRNA export, active gene translocation, and genome repair (1, 2, 43–49), thus implicating it as an important regulator of gene expression.
Previous biochemical studies (50) have identified Sgf29 as a component of SAGA that is conserved from yeast to humans. A recent study in yeast has demonstrated the role of Sgf29p in transcriptional stimulation (25). Further, another study in rats has implicated Sgf29p in regulation of the expression of genes involved in tumourogenesis (51). While the function of Sgf29p in transcription has been documented in yeast and mammalian system, its precise mechanism of action in transcriptional regulation is not clearly understood in living cells. Recently, Bian et al (25) have implicated Sgf29p in recruitment of SAGA via its interaction with methylated-K4 on histone H3 in yeast. Similarly, another recent study in human cell lines (26) has also shown the requirement of Sgf29p in recruitment of SAGA via the interaction of Sgf29p with methylated-K4 of histone H3. Although these studies (25, 26) have demonstrated an important role of Sgf29p in recruitment of SAGA (and hence SAGA’s function in transcription) via histone H3 K4 methylation, it is not clearly known whether Sgf29p can also play additional function(s) in transcription by regulating SAGA’s structural integrity, TBP recruitment or histone H3 acetylation in vivo. A very recent MudPIT (multidimensional protein identification technology)-based biochemical study (52) has revealed that the absence of Sgf29p does not alter the global structural integrity of SAGA. Similarly, Bian et al. (25) have also demonstrated biochemically the dispensability of Sgf29p in maintaining SAGA’s overall structural integrity. However, in striking contrast to their in vitro results, Bian et al (25) have implicated that Sgf29p is required for SAGA’s structural integrity as well as histone H3 acetylation in vivo. Here, using the ChIP assay in conjunction with transcriptional and mutational analyses, we show that Sgf29p is not required for recruitment of SAGA components such as Spt20p, TAF10p, TAF12p and Ubp8p, supporting the dispensability of Sgf29p in maintaining SAGA’s global structural integrity in vivo. Further, our results reveal that Sgf29p facilitates the recruitment of TBP (and hence transcription) in vivo. Such role of Sgf29p does not appear to be mediated via histone H3 acetylation at the SAGA-dependent, but Gcn5p-independent gene.
Plasmid SGP4 (53) was generated by cloning a DNA fragment containing three Gal4p-binding sites into the low copy number plasmid pRS416. The plasmid pFA6a-13Myc-KanMX6 (54) was used for genomic myc epitope-tagging of the proteins of interest. The Plasmid pRS416 (55) was used in the PCR-based gene disruption.
Multiple myc-epitope tags were added at the original chromosomal locus of SGF29 in W303a to generate ASY2 (Sgf29p-myc, Kan). The plasmid SGP4, carrying three Gal4p-binding sites, was transformed into ASY2 to generate ASY5. The endogenous SGF29 gene (encompassing the whole protein-coding sequence) of W303a was disrupted using the PCR-based gene knock-out method (56) to generate ASY8 (Δsgf29::URA3). Multiple myc-epitope tags were added at the original chromosomal locus of SPT20 in W303a and ASY8 to generate ASY10 (Spt20p-myc, Kan) and ASY12 (Spt20p-myc, Kan; Δsgf29::URA3), respectively. The strains, ASY4 (Ubp8p-myc, Kan) and SLY18 (Ubp8p-myc, Kan; Δsgf29::URA3), were generated by inserting multiple myc epitope tags at the original chromosomal locus of UBP8 in W303a and ASY8 strains.
For the studies at the GAL1 promoter in Δsgf29 and its isogenic wild type equivalent, cells were first grown in YPR (yeast extract containing peptone plus 2% raffinose) to an optical density at 600 nm (OD600) of 0.9 and then transferred to YPG (yeast extract-peptone plus 2% galactose) for 90 min at 30°C prior to formaldehyde-based in vivo crosslinking. Minimal media containing either 2% galactose or raffinose were used for the strain (ASY5) with reporter plasmid, SGP4. The yeast strains were grown in YPD (yeast extract containing peptone plus 2% dextrose) to an OD600 of 1.0 at 30°C for the studies at the ADH1, PHO84 and RPS5 promoters.
The ChIP assay was performed as described previously (57, 58). Briefly, yeast cells were treated with 1% formaldehyde, collected and resuspended in lysis buffer. Following sonication, cell lysates (400 μl lysate from 50 ml of yeast culture) were precleared by centrifugation, and then 100 μl lysate was used for each immunoprecipitation. Immunoprecipitated protein–DNA complexes were treated with proteinase K, the crosslinks were reversed, and then DNA was purified. Immunoprecipitated DNA was dissolved in 20 μl TE 8.0 (10 mM Tris-HCl pH 8.0, and 1 mM EDTA), and 1 μl of immunoprecipitated DNA was analyzed by PCR (polymerase chain reaction). PCR reactions contained [α-32P]dATP (2.5 μCi for each 25-μl reaction), and the PCR products were detected by autoradiography after separation on a 6% polyacrylamide gel. As a control, “input” DNA was isolated from 5 μl lysate without going through the immunoprecipitation step and was suspended in 100 μl TE 8.0. To compare PCR signal arising from the immunoprecipitated DNA with the input DNA, 1 μl of input DNA was used in the PCR analysis. The ChIP assay for histone H3 was carried out following the protocol as described previously (37, 59). Primer-pairs used for PCR analysis were as follows:
Primers flanking Gal4p-binding sites in the plasmid SGP4 are 5′ GGTGGCGGCCGCTCTAGAACTAGT-3′ and 5′ TTGACCGTAATGGGATAGGTCACG-3′.
Autoradiograms were scanned and quantitated by the National Institutes of Health image 1.62 program. Immunoprecipitated DNAs were quantitated as the ratio of immunoprecipitate to input in the autoradiogram.
Total mRNA was prepared from yeast cell culture following the standard protocol. Briefly, 10 ml yeast culture of a total OD600 of 1.0 was harvested, and then was suspended in 100 μl RNA preparation buffer (500 mM NaCl, 200 mM Tris–HCl, 100 mM Na2 EDTA and 1% SDS) along with 100 μl phenol/chloroform/isoamyl alcohol and 100 μl volume-equivalent of glass beads (acid washed; Sigma). Subsequently, yeast cell suspension was vortexed with a maximum speed (10 in VWR mini-vortexer; Cat. No. 58816-121) for five times (30 seconds each). Cell suspension was put in ice for 30 seconds between pulses. After vortexing, 150 μl RNA preparation buffer and 150 μl phenol/chloroform/isoamyl alcohol were added to yeast cell suspension followed by vortexing for 15 seconds with a maximum speed on VWR mini-vortexer. phase was collected following 5 min centrifugation at a maximum speed in microcentrifuge machine. The total mRNA was isolated from aqueous phase by ethanol precipitation.
Primer extension analysis was performed as described previously (7). The primers used for analysis of GAL1, ADH1, PHO84 and RPS5 mRNAs were as follows:
Sanders et al (50) have implicated biochemically Sgf29p as a component of SAGA. To determine whether Sgf29p exists in the same form, as that defined by its biochemical co-purification with SAGA, at the promoter in vivo, we analyzed the recruitment of Sgf29p to the UAS of a well-characterized SAGA-dependent gene, GAL1, using a ChIP assay. If Sgf29p is an integral component of SAGA in vivo, it will be recruited to the GAL1 UAS but not to the core promoter, since we have previously shown that SAGA components including TAFs are predominantly recruited to the GAL1 UAS but not core promoter (53). Figures 1A and 1B show that the TAF12p subunit of SAGA as well as Sgf29p were present at the GAL1 UAS but not the core promoter following 7 The aqueous cycles (each cycle with 10 seconds at a power output of 8 in Misonix sonicator) of sonication in the ChIP assay. Similarly, Gal4p was recruited to the GAL1 UAS (Figure 1B). These observations suggest that Sgf29p is a component of SAGA in vivo. However, it remained possible that other transcription components might have played a crucial role in recruitment of Sgf29p to the GAL1 UAS, since the above experiments were performed on the intact GAL1. To address this issue, we next analyzed whether Sgf29p is recruited by Gal4p to a plasmid bearing only Gal4p-binding sites but not other promoter elements. Figure 1C shows that Sgf29p was recruited to the Gal4p-binding sites in the plasmid when Gal4p activation domain was active in galactose-containing growth medium. On the other hand, Sgf29p was not associated with the Gal4p-binding sites when Gal4p activation domain was inactive in raffinose-containing growth medium. However, the levels of Gal4p recruitment to the Gal4p binding sites in both galactose and raffinose-containing growth media were the same (Figure 1C). Similarly, the Ubp8p and Sgf73p components of SAGA were also recruited to the Gal4p-binding site by Gal4p activation domain (Figure 1D; 37, 58). Further, our previous studies (53) have demonstrated that other SAGA components such as Spt3p and Spt20p were also recruited to the minimal Gal4p-binding sites in the plasmid when Gal4 activation domain was active in galactose-containing growth medium. Significantly, the general transcription factors were not recruited to the minimal Gal4p binding sites (53). Together, these results support that Sgf29p exists in the same form at the promoter in vivo, as observed biochemically (50), and thus, it is recruited to the GAL1 UAS or Gal4p activation domain along with other SAGA components.
To determine the role of Sgf29p in transcriptional regulation, we next analyzed its contribution in SAGA recruitment or integrity. If Sgf29p maintains SAGA’s integrity, the SAGA components will not be recruited to the GAL1 UAS in the Δsgf29 strain. Figure 2A shows that the deletion of SGF29 did not alter the recruitment of the SAGA components such as Spt20p, Ubp8p, TAF10p and TAF12p. Similarly, the recruitment of the activator Gal4p to the GAL1 UAS was not altered in the Δsgf29 strain (Figure 2B). Likewise, our previous studies have demonstrated that the recruitment of the SAGA components was not changed in the deletion mutant of GCN5 that is dispensable for the integrity of SAGA (4, 5, 53). On the other hand, recruitment of SAGA was almost lost in the SPT20 (that integrates SAGA (5)) deletion mutant strain (4, 53). Since Spt20p maintains the global structural integrity of SAGA and its recruitment (along with other SAGA components such as TAF10p, TAF12p and Ubp8p) is not altered in the absence of Sgf29p, thus, like Gcn5p, Sgf29p does not appear to regulate the global structural integrity of SAGA in vivo.
So far, we find that the absence of Sgf29p did not alter the global structural integrity of SAGA in vivo. Similarly, previous studies revealed that the Spt3p component of SAGA is also not required for SAGA’s global structural integrity (5, 53, 60, 61). Rather, it promotes transcription by facilitating the recruitment of TBP at the promoter (53, 60, 61). Thus, Spt3p is essential for SAGA’s functional integrity, even though it is dispensable for maintaining the overall structural integrity of SAGA. Based on these studies, we hypothesize that like Spt3p, Sgf29p might be regulating transcription by modulating the recruitment of TBP at the promoter for nucleation of the PIC assembly to initiate transcription. To test this hypothesis, we next analyzed Sgf29p’s contribution in recruitment of TBP to the GAL1 core promoter. Figure 3A shows that the recruitment of TBP to the GAL1 core promoter was decreased in the SGF29 deletion mutant strain. Such a decrease in TBP recruitment to the GAL1 core promoter in Δsgf29 would reduce transcription. Indeed, we find that the transcription of GAL1 was also decreased in Δsgf29 (Figure 3B). As a control, we show that transcription of a SAGA-independent gene, RPS5 (53), was not altered in Δsgf29 (Figure 3B). Consistent with the RPS5 transcription data, we further show that the recruitment of TBP to the RPS5 core promoter was not altered in the absence of Sgf29p (Figure 3C). Together, these results demonstrate that the Sgf29p component of SAGA promotes GAL1 transcription by facilitating the recruitment of TBP to the core promoter.
To determine whether Sgf29p also facilitates TBP recruitment (and hence transcription) at other genes, we analyzed the association of TBP with the core promoters of two other SAGA-dependent genes, namely ADH1 and PHO84 (4) in the SGF29 deletion mutant and its isogenic wild type equivalent. Figures 4A and 4B show that the recruitment of TBP to the ADH1 and PHO84 core promoters was decreased in Δsgf29. Consistently, transcription of these genes was also decreased in the SGF29 deletion mutant strain (Figure 4C). Thus, SAGA-associated Sgf29p enhances transcription by facilitating the recruitment of TBP in vivo.
We next asked whether Gcn5p-HAT activity of SAGA is regulated by Sgf29p in vivo. Previous biochemical studies have demonstrated that Gcn5p-HAT activity on nucleosomal histone H3 is dependent on the Ada2p and Ada3p components of SAGA (4, 58, 62–65). Further, a recent MudPIT-based study (52) implicated that Sgf29p along with Ada2p, Ada3p and Gcn5p form a HAT module within SAGA. Thus, like Ada2p and Ada3p (4, 58, 64, 65), Sgf29p may regulate the HAT activity and hence histone H3 acetylation at the SAGA-regulated promoters. To test this, we analyzed the levels of histone H3 acetylation and histone H3 at the core promoters of the SAGA-dependent ADH1 and PHO84 genes in the Δsgf29 mutant and wild type strains. Interestingly, our ChIP data revealed that the levels of histone H3 at the ADH1 and PHO84 core promoters were increased in Δsgf29 when compared with the wild type equivalent (Figure 5). This is as expected, since previous studies (58, 66) have implicated an inverse correlation between TBP recruitment (or PIC formation/transcription) and nucleosomal disassembly. As TBP recruitment at the ADH1 and PHO84 core promoters was decreased in Δsgf29 (Figures 4A and 4B), an increased level of histone H3 was observed at these core promoters in the absence of Sgf29p (Figure 5). Similar increase in the level of histone H3 acetylation would be observed in the ChIP assay in Δsgf29, if Sgf29p did not regulate histone H3 acetylation at these promoters. Intriguingly, we find that the levels of histone H3 acetylation at these promoters did not increase in the same proportions as of histone H3 in Δsgf29 (Figure 5). These results support that histone H3 acetylation is decreased in the absence of Sgf29p in vivo.
Here, we show that, like other SAGA components, Sgf29p is recruited to the UAS, but not core promoter of a well-characterized SAGA-dependent gene, GAL1 (Figures 1A and 1B). Further, we demonstrate that Gal4p activation domain targets the recruitment of Sgf29p to the minimal Gal4p-binding site as a SAGA component in vivo (Figure 1C). These results implicate that Sgf29p exists in vivo in the same form as that defined by its biochemical co-purification with SAGA. Further, we find that the deletion of SGF29 does not impair the recruitment of SAGA components such as TAF10p, TAF12p, Spt20p and Ubp8p (Figure 2A). These observations support that Sgf29p is not essential to maintain the global structural integrity of SAGA in vivo.
Consistent with our in vivo results, a recent study (25) has also demonstrated that Sgf29p is not required for SAGA’s overall integrity in vitro. Further, another MudPIT-based recent biochemical study (52) has revealed the dispensability of Sgf29p in maintaining SAGA’s global structural integrity. However, in striking contrast to these biochemical data (25, 52) and our in vivo results, Bian et al (25) have implicated the requirement of Sgf29p in maintaining SAGA’s structural integrity in vivo. Using the ChIP assay, they demonstrated the decrease in the recruitment of the Ada2p component of SAGA in Δsgf29. Ada2p is not essential for maintaining SAGA’s global structural integrity. It is also an integral component of the ADA complex. Bian et al. (25) used Ada2p as a representative component of SAGA to evaluate the effect of Sgf29p in recruitment of SAGA, and found that the absence of Sgf29p decreased the recruitment of Ada2p by ~2-fold at the GAL1 UAS. A recent study (52) implicated Sgf29p in the HAT module (containing Sgf29p, Ada2p, Ada3p and Gcn5p) within SAGA, and thus, the absence of Sgf29 may alter the association of Ada2p with SAGA in vivo. Indeed, a decrease in the recruitment of Ada2p in Δsgf29 was observed in the previous studies (25). Further, Lee et al (52) have demonstrated Sgf29p as a common component of the SAGA and ADA complexes. Thus, the decrease in the recruitment of Ada2p in Δsgf29 could also occur via the ADA complex if it associates with GAL1 UAS. Therefore, it is important to look at several components of the SAGA complex to evaluate the role of Sgf29p in maintaining SAGA’s structural integrity in vivo. In this study, we analyzed the recruitment of several SAGA components (TAF10p, TAF12p, Spt20p and Ubp8p) to the GAL1 UAS. These SAGA components represent different structural modules of the SAGA complex. Recently, MudPIT-based biochemical studies (52) have implicated four structural modules (TAF, SPT, DUB and HAT) within SAGA. Our ChIP data demonstrate that the recruitment of TAF10p (TAF module), TAF12p (TAF module), Spt20p (SPT module), and Ubp8p (DUB module) is not altered in the absence of Sgf29p (Figure 2A). These results support that SAGA’s overall integrity is not altered in the absence of Sgf29p in vivo. Further, our results suggest that Sgf29p could be a peripheral component of SAGA. Indeed, a recent biochemical study (52) has implicated Sgf29p as a peripheral SAGA component. This is further substantiated by the fact that Sgf29p is accessible to interact with methylated-K4 of histone H3 to enhance the recruitment of SAGA (25, 26).
Since Sgf29p is dispensable for SAGA’s overall structural integrity, it may not have any effect on TBP recruitment and hence transcriptional initiation. Alternatively, it is quite possible that it can promote TBP recruitment (and hence transcription) without altering the recruitment of SAGA, similar to the functional role of Spt3p (5, 53, 60, 61). Indeed, our data support this possibility. We find that the recruitment of TBP to the promoters of several SAGA-dependent genes such as GAL1, ADH1 and PHO84 is decreased in the absence of Sgf29p (Figures 3A, 4A and 4B). Such decrease in TBP recruitment subsequently reduces the transcription of these genes (Figures 3B and and4C).4C). Consistent with our results, Lee et al (52) have demonstrated the growth defect of the Δsgf29 cells in galactose as well as ethanol/glycerol-containing growth media. Yeast cells require the expression of the GAL1 and ADH1 genes for growth in these media (52). Further, Bian et al (25) have also demonstrated the transcriptional defect of GAL1 in the absence of Sgf29p. Taken together, Sgf29p promotes transcription by facilitating the recruitment of TBP without altering SAGA’s global structural integrity.
Previous studies (58) have demonstrated histone H3 acetylation at the promoters of above three genes. Such an acetylation is significantly lost in the absence of Gcn5p (58). However, the recruitment of TBP to the core promoters of GAL1 and ADH1 is not altered in the deletion mutant of GCN5 (4, 58), implicating that histone H3 acetylation is dispensable for TBP recruitment at these promoters. Therefore, these studies support that the decrease in TBP recruitment to the GAL1 and ADH1 core promoters in the absence of Sgf29p is not mediated via histone H3 acetylation. On the other hand, Gcn5p or histone H3 acetylation promotes the recruitment of TBP to the PHO84 promoter (4, 58). Thus, it may be likely that the decrease in TBP recruitment to the PHO84 promoter in Δsgf29 is due to an impaired recruitment of Gcn5p (or associated HAT activity). We rule out this possibility as the deletion of SGF29 does not alter the overall recruitment of SAGA (Figure 2A). In addition, Lee et al (52) have recently demonstrated biochemically that Gcn5p remains associated with SAGA in the absence of Sgf29p. Similarly, Bian et al (25) have also shown that Sgf29p does not alter SAGA’s global structural integrity in vitro. If Sgf29p would have regulated the recruitment of TBP via the modulation of histone H3 acetylation, a decreased level of TBP recruitment to the core promoter of GAL1 or ADH1 in Δsgf29 would not have been observed (as the recruitment of TBP to the GAL1 and ADH1 promoters is not altered in the absence of Gcn5p or HAT activity; 4, 58). However, our data (Figures 3A and and4A)4A) have demonstrated a decreased recruitment of TBP to the ADH1 and GAL1 promoters in vivo. Thus, the role of Sgf29p in promoting the recruitment of TBP to these three promoters does not appear to be mediated via histone H3 acetylation. This is further substantiated by a recent observation that the deletion of SGF29 does not alter the HAT activity of SAGA in vitro (25).
As mentioned above, a recent MudPIT-based study implicated the presence of Sgf29p in SAGA as well as ADA HAT complex (52). Further, the deletion of SGF29 has been shown to reduce global acetylation of histone H3 in the western blot analysis (25). Such a decrease in global histone H3 acetylation may be mediated via ADA, SAGA or both. Since recent MudPIT-based study (52) has implicated the presence of Sgf29p in the HAT module of SAGA, Sgf29p may likely to regulate the HAT activity of SAGA (and hence histone H3 acetylation), similar to the roles of Ada2p and Ada3p (that are present in the HAT module) in regulation of SAGA’s HAT activity (4, 58, 64, 65). However, previous biochemical studies (25, 52) have demonstrated that the structural integrity of SAGA is not altered in the absence of Sgf29p, and such SAGA complex (without Sgf29p) does not alter histone H3 acetylation in vitro (25), and hence SAGA’s HAT activity. Thus, the global decrease in histone H3 acetylation in the absence of Sgf29p may occur through Sgf29p-associated ADA HAT complex.
To determine the role of Sgf29p in histone H3 acetylation at the promoter in vivo, we analyzed histone H3 acetylation at the promoters of the SAGA-regulated ADH1 and PHO84 genes in the presence and absence of Sgf29p. We find that Sgf29p moderately enhances histone H3 acetylation at these SAGA-regulated promoters (Figure 5). However, Bian et al (25) have demonstrated a dramatic decrease in histone H3 acetylation at the GAL1 promoter in Δsgf29, even though their in vitro results shows no effect of Sgf29p on histone H3 acetylation. One possible reason for this discrepancy could be the absence of histone H3 control in their ChIP experiments (25). Therefore, it is not clear how much histone H3 acetylation signal was altered in proportion to histone H3 in Δsgf29. Further, they have used K9 mono-acetylated histone H3 antibody in their study (25). On the other hand, K9/14-diacetylated histone H3 antibody was used in this study. Although our results demonstrate that Sgf29p moderately enhances histone H3 acetylation, such acetylation does not appear to enhance TBP recruitment at the GAL1 and ADH1 promoters as described above. However, such a moderate regulation of histone H3 acetylation by Sgf29p may have a positive effect on TBP recruitment at the Gcn5p-regulated genes, as previous studies (36, 58) have shown a correlation of histone H3 acetylation with TBP recruitment.
Previous studies have implicated the enhancement of SAGA recruitment via the interaction of the Sgf29p’s tudor domain with K4-methylated-histone H3 (25, 26). Therefore, the deletion of SET1 (that is involved in histone H3 K4 mono, di and trimethylation) would impair the recruitment of TBP (and hence transcription) at the SAGA-regulated genes. However, previous studies have demonstrated that the absence of Set1p does not alter TBP recruitment at the core promoters of several SAGA-dependent genes such as GAL1, ADH1 and PHO84 (37, 67). Consistently, the transcription of these genes is also not altered in the SET1 deletion mutant when compared with wild type equivalent (37, 67). Although previous studies (25, 26) have implicated the interaction of Sgf29p with methylated-K4 of histone H3, the loss of such interaction does not appear to have a dramatic effect on transcription. Similarly, Gcn5p has been implicated to enhance the recruitment of SAGA via the interaction of its bromodomain with acetylated-histone H3 (1, 14–21, 63, 68, 69). However, the deletion of GCN5 does not alter global recruitment of SAGA or its integrity (4, 5). Further, Gcn5p is not required for transcription of all SAGA-dependent genes (3, 58).
The role of Sgf29p in promoting the recruitment of TBP does not appear to be primarily mediated via histone H3 acetylation, neither is the decrease in recruitment of TBP due to disintegration of the SAGA complex in the absence of Sgf29 (Figure 2A; 25, 52). Although we find that the absence of Sgf29p lowers the recruitment of TBP, but its effect on TBP recruitment is not as dramatic as observed in other SAGA mutants such as Δspt20, Δada1, Δspt7 and Δsgf73 (4, 37, 58). Even though the effect of Sgf29p on TBP recruitment is relatively small, it may have important roles during differentiation and development in higher eukaryotes. Indeed, the absence of Sgf29p in rats has been implicated in oncogenic transformation via modest transcriptional alteration (51). Further, the loss of Sgf29p homologue in plant delays flowering (70), and has a moderate effect on transcription (70).
In summary, we show here that Sgf29p exists in vivo in the same form as that defined by its biochemical co-purification with SAGA. Sgf29p is not required for recruitment of the SAGA components such as Spt20p, TAF10p, TAF12p and Ubp8p (or its overall structural integrity), but facilitates the association of TBP (and hence transcription) in vivo. These results provide significant insight as to how yeast Sgf29p (and possibly its mammalian homologue) functions physiologically to regulate transcription.
We thank Michael R. Green for TBP, TAF10p and TAF12p antibodies; Priyasri Chaurasia and Shruti Bagla for technical assistance; Shivani Malik for critical reading of the manuscript, and Sarah Frankland-Searby for editorial assistance.
Fundings: This work was supported by a National Scientist Development Grant (0635008N) from American Heart Association, a Research Scholar Grant (06-52) from American Cancer Society, a National Institutes of Health grant (1R15GM088798-01), and internal grants of Southern Illinois University. A.S. was supported by a predoctoral fellowship (0710187Z) from American Heart Association.