Analysis of Spt7 Function in the Saccharomyces cerevisiae SAGA Coactivator Complex

The Saccharomyces cerevisiae SAGA (Spt-Ada-Gcn5 acetyltransferase) complex is a multisubunit coactivator complex that is important for transcription in vivo (27, 32). Whole-genome mRNA analysis of SAGA mutants has shown that the expression of approximately 10% of S. cerevisiae genes are affected by the loss of the SAGA complex (23). Both in vivo and in vitro experiments have shown that SAGA activates transcription after its recruitment by transcriptional activators (5, 22, 42, 43). Other results have suggested that SAGA also represses transcription at particular promoters (4, 23). In addition, several studies have shown that SAGA often acts coordinately with other coactivator complexes at a promoter to achieve normal levels of transcription (27). The SAGA complex is conserved between yeast and humans, strongly suggesting that this type of coactivator is also important in mammalian transcription (6, 25, 26, 28, 46). The subunits in SAGA can be grouped functionally based on a large body of in vivo and in vitro experiments (27). Three classes of SAGA proteins are involved in distinct aspects of transcriptional control. First, Gcn5 contains the catalytic activity for the histone acetyltransferase (HAT) activity of SAGA (8), and Gcn5's HAT activity is modulated by the Ada2 and Ada3 proteins (2, 16, 39). Second, Spt3 and Spt8 of SAGA have been shown to control the TATA box binding protein (TBP)-TATA interaction at particular promoters (4, 5, 9, 11, 22). Third, Tra1 has been shown to interact with several transcriptional activators in vitro, suggesting that SAGA is recruited to promoters via this subunit and subsequently activates or represses transcription through its different activities (7). SAGA also contains two additional classes of proteins that are not known to participate directly in regulation and therefore may serve structural roles. In the first class are three proteins, Spt7, Spt20, and Ada1, that function as SAGA core components based on their requirement for integrity of the complex. SAGA is absent from spt7Δ, spt20Δ, and ada1Δ mutants, as assayed by Western analysis and HAT assays (16, 38). Moreover, these mutations cause a broad variety of severe phenotypes, consistent with a complete loss of SAGA function (21, 30, 38). The second class contains a subset of the Tafs (Tafs 5, 6, 9, 10, and 12), proteins initially identified as components of the TFIID complex (17). In referring to the Tafs, we are using the recently revised uniform Taf nomenclature (41). At least one of the Tafs, Taf12, appears to be important for SAGA structure, as its removal via a taf12 temperature sensitivity mutation affects both SAGA integrity as well as its nucleosomal HAT activity (17). While the control of transcription by SAGA has been extensively studied, less is known about the control of its assembly and the protein-protein associations within the complex. Previous studies have shown that both Spt7 and Ada1 have histone fold motifs that interact with Taf components of SAGA: Ada1 with Taf12 and Spt7 with Taf10 (13, 14). However, it is not known how the core subunits contribute toward the structural integrity of the complex. To learn more about the roles of one of the SAGA core components, we have chosen Spt7 as the focus of our studies. Spt7 is a 1,332-amino-acid, highly negatively charged protein whose sequence contains two motifs of note. First, as mentioned previously, Spt7 contains a histone fold (amino acids 979 to 1045) required for interaction with the Taf10 subunit of SAGA (13). Second, Spt7 contains a bromodomain (amino acids 463 to 523), a motif found in many transcription factors (20; reviewed in reference 44). A deletion that removes the Spt7 bromodomain does not cause any detectable phenotypes, suggesting that this domain is either redundant or not crucial for Spt7 function (15, 38). Previous studies suggest that Spt7 may act dynamically to regulate SAGA function. These studies demonstrated that SAGA exists in at least one alternate form that has been named SLIK (SAGA-like), SAGAalt, or SALSA (4, 18, 34; D. Sterner and S. Berger, personal communication). This complex is referred to hereafter as SLIK/SALSA. The SLIK/SALSA complex has two identified differences from SAGA: a smaller form of Spt7 and the absence of Spt8 (4; Patrick Grant, personal communication). Although little is known about the formation of SLIK/SALSA, evidence suggests involvement of the general amino acid control pathway, because SLIK/SALSA levels increase in cells grown in amino acid starvation conditions (4). However, the requirements for SLIK/SALSA formation, as well as its functions in vivo, remain unknown. In this study, we examine several aspects of Spt7 protein structure and function. We demonstrate that Spt7 is involved in regulating the levels of Spt20 and Ada1, the two other SAGA core components, suggesting that Spt7 levels control the amount of SAGA present in vivo. Additional experiments reveal that partial Spt7-containing complexes form in the absence of the other core components. To determine regions of function and interaction with other subunits, we generated and analyzed a set of spt7 partial deletion mutants. This analysis allowed us to delineate a region of minimal function in Spt7 as well as a domain for Spt8 interaction. Moreover, we have defined a region of Spt7 that is required for its processing and, hence, for the formation of SLIK/SALSA. Analysis of the spt7 mutant that impairs processing suggests that SLIK/SALSA is not strongly required for transcriptional activation.