KAT2A (GCN5) and KAT2B (PCAF) are histone acetyltransferases (HATs) that act as part of large multimember complexes to facilitate transcription by acetylating histones H3 and H4. In eukaryotes the SPT-ADA-GCN5 acetyltransferase (SAGA) complex has 19 subunits including TRRAP, ENY2, USP22 and subunits belonging to the ADA, SPT, TAF, and SGF group of proteins (Nagy et al. 2009). The ADA2A-containing (ATAC) complex shares with SAGA a core composed of KAT2A-TADA3 (ADA3)-CCDC101 (STAF36, SGF29) and either TADA2A (ADA2a) in ATAC, or TADA2B (ADA2b) in SAGA. ATAC complexes contain a second putative HAT, called CSRP2BP (ATAC2), and five other subunits; YEATS2, ZZZ3, MBIP, WDR5, and DR1 (NC2-Beta) (Guelman et al. 2009). CSRP2BP has weak HAT activity in vitro but it's main function is to maintain the structural integrity of ATAC (Guelman et al. 2009). At present, the biological function of the ATAC complex is not well understood. In vitro GCN5 acetylates mainly histone H3K14 (lysine-15 in the UniProt peptide which retains the initiating methionine), but when incorporated into the SAGA complex GCN5 becomes more efficient as an H3K14 acetylase and can also acetylate H3K9 and H3K18 (Brand et al. 1999, Grant et al. 1999), H3K23, and H3K27 (Kuo et al. 1996, Kuo & Andrews 2013). Drosophila ATAC mainly acetylates histone H4 (Ciurciu et al. 2006, Suganuma et al. 2008), suggested to be due to the presence of CSRP2BP in the complex (Suganuma et al. 2008) but different human ATAC preparations have exhibited a range of specificities with no clear difference between SAGA and ATAC (Guelman et al. 2009, Wang et al. 2008, Nagy et al. 2010). SAGA and ATAC complexes from mouse and human contain either GCN5 or PCAF in a mutually exclusive manner (Nagy et al. 2010, Krebs et al. 2010, Spedale et al. 2012).
The SAGA complex consists of KAT2A (hGCN5), TADA1 (STAF42), TADA2B (ADA2b), TADA3 (STAF54, ADA3), SUPT3H (SPT3), SUPT7L (STAF65G), TAF5L (PAF65B), TAF6L (PAF65A), TAF9 (TAFII31), TAF12 (TAFII20), TAF10 (TAFII31), TRRAP, SAP130 (Martinez et al. 2001), CCDC101, ATXN7, a factor termed STAF55 that cannot be identified, two further factors described as probable members that cannot be identified STAF46 and STAF60 (Nagy & Tora 2007) plus ATXN7L3, USP22, ENY2 (Zhao et al. 2008) and SUPT20H (Nagy et al. 2009).
N.B. Coordinates of post-translational modifications described here follow UniProt standard practice whereby coordinates refer to the translated protein before any further processing. Histone literature typically refers to coordinates of the protein after the initiating methionine has been removed. Therefore the coordinates of post-translated residues in the Reactome database and described here are frequently +1 when compared with the literature.