Gene activation by dissociation of an inhibitor from a transcriptional activation domain.

Direct masking of the activation domain (AD) of a transcriptional activator by an inhibitory protein and relief of such masking in response to signals is typical for several eukaryotic gene regulatory systems. Such is the case for the transcriptional inhibitors RB, MDM4/MDMX, ZFM1, Opi1, and Gal80, all of which target DNA-binding transcription activators to exert their inhibitory effect (5, 6, 16, 33). In the case of RB, it binds to a site embedded in the transactivation domain of the E2F protein. Phosphorylation of RB lowers its binding affinity to E2F and results in gene activation (6). For the Opi1 protein in Saccharomyces cerevisiae, signal-responsive tethering of Opi1 to the ER membrane physically separates Opi1 and its target, Ino2, the transcriptional activator protein involved in the inositol pathway (12). For the yeast Gal80 protein, its ability to mask the AD of the transcriptional activator Gal4 is somehow overcome through interaction with the galactose-activated form of the Gal3 protein. For each of these systems, the molecular activities that cause the unmasking event constitute the overall signal-responsive gene activation mechanism. In our efforts to elucidate the mechanism of GAL gene activation in yeast, we have focused on how galactose relieves Gal80 masking of the Gal4 AD. Gal4 is a prototypical acidic transcriptional activator that binds to a 17-bp upstream activation sequence within the GAL gene promoters (UASGAL) (3, 7). In the absence of galactose, Gal80 masks the Gal4 AD, preventing recruitment of the transcriptional machinery (13-15). In the presence of galactose and ATP, Gal3 physically interacts with Gal80, and such interaction is required for relieving the Gal80 inhibition of Gal4, resulting in induced GAL gene expression (2, 26). However, how the Gal3-Gal80 interaction leads to Gal4 activation has been unresolved. There are two contrasting bodies of evidence concerning how the Gal80-Gal3 interaction relieves Gal80 inhibition of Gal4. Historically, two independent studies led to the view that Gal80 remains associated with Gal4 at the promoter in galactose-induced cells (11, 21). The initial evidence came from experiments that used a Gal80-VP16 hybrid protein in which the transcriptional AD of VP16 was fused to Gal80. Gal80-VP16 was found to stimulate transcription of a GAL reporter gene in the presence of galactose, indicating that Gal80 stayed bound to Gal4 at the GAL gene promoter after induction (11). Later, others using a constitutive mutant of Gal3 protein at a 30× excess relative to Gal80 detected a complex of Gal3-Gal80-Gal4 associated with a UASGAL-containing DNA fragment in an electrophoresis mobility shift assay (21). Taken together, these data supported a nondissociation model proposing that galactose-activated Gal3 binds to Gal4-associated Gal80 at the GAL gene promoter in the nucleus and causes the Gal80-Gal4 complex to adopt a conformation that exposes the Gal4 AD (21). Challenging the nondissociation model is a more recent body of evidence that points to dissociation of Gal80 from Gal4 after induction. Gal3 was detectable by two different methods only in the cytoplasm, and cells in which Gal3 was tethered to membranes outside the nucleus exhibited a magnitude of induction similar to that exhibited by wild-type cells (18). In addition, chromatin immunoprecipitation experiments revealed reduced binding of Gal80 to Gal4 after galactose induction (19). Accordingly, a dissociation model was proposed, in which the binding of Gal80 to cytoplasm-localized Gal3 results in a decrease of Gal80 content in the nucleus, leading to its dissociation from Gal4 (19). However, Gal80 dissociation from Gal4 has in turn been called into question by the report of a fluorescence resonance energy transfer (FRET) between Gal80-enhanced cyan fluorescent protein (ECFP) and Gal4-enhanced yellow fluorescent protein (EYFP) in galactose-grown yeast (1). Here we present the results of new experiments aimed at resolving the conflicting models for how galactose triggers relief of Gal80 inhibition of Gal4. We demonstrate with the use of a GAL1 promoter-controlled reporter gene array and live-cell imaging that Gal80 rapidly dissociates from Gal4 in response to galactose. Our results further show that such dissociation depends on interaction between Gal3 and Gal80 and is temporally correlated with GAL reporter gene expression. We also find that Gal80 is able to reassociate with Gal4 when galactose is depleted and protein synthesis is blocked, suggesting that reversible binding of Gal80 by Gal3 contributes to the galactose-triggered Gal80-Gal4 dissociation event. We also detect a modest redistribution of Gal80 from the nucleus to the cytoplasm by 15 to 25 min following galactose addition. Finally, we provide here the first evidence that Gal3 is detectable within the nucleus before and after galactose addition. Based on these results, we conclude that the dissociation of Gal80 from Gal4 by Gal80 interaction with Gal3 is the event that initiates the active state of Gal4.