Yg. Gangloff et al., The human TFIID components TAF(II)135 and TAF(II)20 and the yeast SAGA components ADA1 and TAF(II)68 heterodimerize to form histone-like pairs, MOL CELL B, 20(1), 2000, pp. 340-351
It has been previously proposed that the transcription complexes TFIID and
SAGA comprise a histone octamer-like substructure formed from a heterotetra
mer of H4-like human hTAF(II)80 (or its Drosophila melanogaster dTAF(II)60
and yeast [Saccharomyces cerevisiae] yTAF(II)60 homologues) and H3-like hTA
F(II)31 (dTAF(II)40 and yTAF(II)17) along with two homodimers of H2B-like h
TAF(II)20 (dTAF(II)30 alpha and yTAF(II)61/68). However, it has not been fo
rmally shown that hTAF(II)20 heterodimerizes via its histone fold. By two-h
ybrid analysis with yeast and biochemical characterization of complexes for
med by coexpression in Escherichia coli, we showed that hTAF(II)20 does not
homodimerize but heterodimerizes with hTAF(II)135. Heterodimerization requ
ires the alpha 2 and alpha 3 helices of the hTAF(II)20 histone fold and is
abolished by mutations in the hydrophobic face of the hTAF(II)20 alpha 2 he
lix. Interaction with hTAF(II)20 requires a domain of hTAF(II)135 which sho
ws sequence homology to H2A. This domain also shows homology to the yeast S
AGA component ADA1, and we show that yADA1 heterodimerizes with the histone
fold region of yTAF(II)61/68, the yeast hTAF(II)20 homologue. These result
s are indicative of a histone fold type of interaction between hTAF(II)20-h
TAF(II)135 and yTAF(II)68-yADA1, which therefore constitute novel histone-l
ike pairs in the TFIID and SAGA complexes.