In this report, we have analyzed the in vivo role of SAGA components for TBP recruitment to SAGA-dependent promoters. Our major findings are summarized in Table and are discussed below. All of the SAGA-dependent promoters examined required Ada1p, Spt3p, Spt7p, and Spt20p for TBP recruitment. Previous biochemical experiments have indicated that Ada1p, Spt7p, and Spt20p are required for the integrity of the SAGA complex (16
), which is consistent with the general requirement for these components that we observed. However, our results do not rule out that, in addition to complex integrity, Ada1p, Spt7p, and Spt20p also have a more direct role in TBP recruitment. In addition, genetic and biochemical experiments have suggested that Spt3p interacts with TBP (13
). Our finding that Spt3p is generally required for TBP recruitment is again consistent with this possibility.
Summary of the requirement for SAGA components for TBP recruitment to SAGA-dependent promotersa
In contrast to the general requirement of Ada1p, Spt3p, Spt7p, and Spt20p, we found that Spt8p, Ada2p, Ada3p, Gcn5p, and the TAF components of SAGA were differentially required for TBP recruitment to SAGA-dependent promoters. Significantly, we found that the promoter requirements for Gcn5p, Ada2p, and Ada3p were identical, which is consistent with previous work showing that Ada2p and Ada3p are required for Gcn5p's HAT activity on nucleosomal histones (3
). TAF12 was also required at the same subset of promoters as Gcn5p, Ada2p, and Ada3p, consistent with the finding that TAF12 is required for SAGA's in vitro HAT activity on nucleosomal substrates (17
). Finally, genetic evidence has suggested that Spt8p is required for the functional interaction between Spt3p and TBP (15
), which may be mediated through an Spt8p-TBP interaction (37
). Our results suggest that if this interaction occurs in vivo, it is required only at certain promoters.
The SAGA-dependent promoters displayed differential sensitivities to temperature-sensitive mutations in TAF6
, and TAF12
. The differential requirement for TAF6, TAF9, and TAF12 at SAGA-dependent promoters is consistent with the results of genome-wide expression analysis (26
). However, several studies have found that different temperature-sensitive mutant alleles within a single TAF can result in distinct phenotypes and gene expression profiles (for an example, see reference 23
). This finding is most likely explained by the fact that TAFs can have multiple domains, can be present in more than one complex, can mediate different functions, and can also be required for complex integrity (1
). Thus, although our results suggest that SAGA-dependent promoters will have differential TAF requirements, it will be important to verify this supposition by analysis of additional TAF mutants.
SAGA is a complex, multisubunit transcription factor that has at least two distinct activities: it can serve as the direct target (adaptor) for transcriptional activation domains, and it has a HAT activity that can modify chromatin structure. These two activities are carried out by distinct subunits of the SAGA complex. We speculate that the multiplicity of SAGA functions is related to our finding that various SAGA-dependent promoters differentially require specific SAGA subunits. For example, the differential requirement for Gcn5p's HAT activity suggests that the chromatin structure differs at various SAGA-dependent promoters. Likewise, the activators at some SAGA-dependent promoters, such as Gal4p, may require the SAGA adaptor function; however, at other SAGA-dependent promoters, SAGA's adaptor function may not be essential because, for example, SAGA is redundant with other targets. Recognizing and understanding the specific features of SAGA-dependent promoters that determine their requirement for specific SAGA subunits will require further research. It is intriguing, however, that, like SAGA, several other multisubunit transcription complexes, including TFIIA (12
), TFIIE (33
), and TAFs (26
), are also differentially required.
Previous studies have shown that transcription of the TAF-independent GAL1
promoter is dependent on SAGA (13
). On the basis of genome-wide transcription profiling, we predicted that other TAF-independent promoters would, like GAL1
, also require SAGA for transcription (26
). Consistent with this prediction, we found that the TAF-independent ADH1
promoter required SAGA for transcription as well as TBP recruitment. In contrast, SAGA was dispensable for transcription from the TAF-independent promoters SED1
(Fig. ) and PGK1
(data not shown) (28
). These results indicate that only a subset of TAF-independent promoters are SAGA dependent. The mechanism by which promoters that are both TAF and SAGA independent are transcriptionally activated remains to be determined. Our results raise the possibility that another complex in addition to TFIID and SAGA is involved in TBP recruitment and transcription activation.
In summary, we have shown here that SAGA components are differentially required for TBP binding to SAGA-dependent promoters in vivo. Our in vivo analysis is remarkably consistent with previous genetic, biochemical, and genome-wide expression data that SAGA components are differentially required for the gene activity (16
). However, the molecular basis of the distinct but selective effects of individual SAGA components on TBP recruitment and hence transcription remains to be elucidated. The present functional analysis of SAGA reinforces an important concept: the individual subunits of transcription complexes may have distinct and selective functions.