Opi1 mediates repression of phospholipid biosynthesis by phosphate limitation in the yeast Saccharomyces cerevisiae.

Structural genes of phospholipid biosynthesis in the yeast Saccharomyces cerevisiae are transcribed when precursor molecules inositol and choline (IC) are limiting. Gene expression is stimulated by the heterodimeric activator Ino2/Ino4, which binds to ICRE (inositol/choline-responsive element) promoter sequences. Activation is prevented by repressor Opi1, counteracting Ino2 when high concentrations of IC are available. Here we show that ICRE-dependent gene activation is repressed not only by an excess of IC but also under conditions of phosphate starvation. While PHO5 is activated by phosphate limitation, INO1 expression is repressed about 10-fold. Repression of ICRE-dependent genes by low phosphate is no longer observed in an opi1 mutant while repression is still effective in mutants of the PHO regulon (pho4, pho80, pho81 and pho85). In contrast, gene expression with high phosphate is reduced in the absence of pleiotropic sensor protein kinase Pho85. We could demonstrate that Pho85 binds to Opi1 in vitro and in vivo and that this interaction is increased in the presence of high concentrations of phosphate. Interestingly, Pho85 binds to two separate domains of Opi1 which have been previously shown to recruit pleiotropic corepressor Sin3 and activator Ino2, respectively. We postulate that Pho85 positively influences ICRE-dependent gene expression by phosphorylation-dependent weakening of Opi1 repressor, affecting its functional domains required for promoter recruitment and corepressor interaction. Copyright © 2016 John Wiley & Sons, Ltd.

Dimerization of yeast transcription factors Ino2 and Ino4 is regulated by precursors of phospholipid biosynthesis mediated by Opi1 repressor.

Structural genes of phospholipid biosynthesis in the yeast S. cerevisiae are activated by the heterodimeric transcription factor Ino2 + Ino4, binding to ICRE (inositol/choline-responsive element) promoter motifs. In the presence of phospholipid precursors inositol and choline, Ino2-dependent activation is inhibited by the Opi1 repressor which interacts with Ino2. In this work, we systematically investigated the importance of regulatory mechanisms possibly affecting ICRE-dependent gene expression. Autoregulatory expression of INO2, INO4 and OPI1 was abolished by promoter exchange experiments, showing that autoregulation of regulators contributes to the degree of differential gene expression but is not responsible for it. Using GFP fusion proteins, Ino2 and Ino4 were found to localize to the nucleus under conditions of repression and derepression. Interestingly, nuclear localization of Ino2 required a functional INO4 gene. Targeting of a lexA-Ino2 fusion to a heterologous promoter containing lexA operator motifs revealed a constitutive gene activation which was not influenced by phospholipid precursors. We could show that Ino2-dependent activation of a lexA-Ino4 fusion is affected by inositol and choline. Since gene activation required interaction of Ino2 and Ino4 mediated by their helix-loop-helix domains, formation/dissociation of the heterodimer must be considered as an important step of target gene regulation.

Constitutive expression of yeast phospholipid biosynthetic genes by variants of Ino2 activator defective for interaction with Opi1 repressor.

Regulated expression of structural genes involved in yeast phospholipid biosynthesis is mediated by inositol/choline-responsive element (ICRE) upstream motifs, bound by the heterodimeric activator complex Ino2 + Ino4. Gene repression occurs in the presence of sufficient inositol and choline, requiring an intact Opi1 repressor which binds to Ino2. For a better understanding of interactions among regulators, we mapped an 18 aa repressor interaction domain (RID, aa 118-135) within Ino2 necessary and sufficient for binding by Opi1. By alanine scanning mutagenesis of the entire RID we were able to identify nine residues critical for Opi1-dependent repression of Ino2 function. Consequently, the corresponding dominant Ino2 variants conferred constitutive expression of an ICRE-dependent reporter gene and were no longer inhibited even by overproduction of Opi1. Interestingly, Ino2 RID partially overlaps with transcriptional activation domain TAD2. As certain mutations exclusively affect repression while others affect both repression and activation, both functions of Ino2 can be functionally uncoupled. Correspondingly, we mapped the RID-binding activator interaction domain (AID, aa 321-380) at the C-terminus of Opi1 and introduced missense mutations at selected positions. An Opi1 variant simultaneously mutated at three highly conserved positions showed complete loss of repressor function, confirming RID-AID interaction as the crucial step of regulated expression of ICRE-dependent genes.

Transcriptional activators Cat8 and Sip4 discriminate between sequence variants of the carbon source-responsive promoter element in the yeast Saccharomyces cerevisiae.

The structural genes for gluconeogenesis in the yeast Saccharomyces cerevisiae are activated by the carbon source-responsive element (CSRE) found in the respective upstream regions. Regulatory genes CAT8 and SIP4 both encode zinc-cluster proteins which can bind to CSRE motifs and activate target genes under conditions of glucose deprivation. In this work, we describe a functional analysis of sequence variants containing single mutations within the strongly activating CSRE(ICL1) motif. While the sequence CCNNNNNNCCG was required as the minimal UAS for gene activation by both Cat8 and Sip4, the activators responded differently to sequence variations in the central part of the CSRE. Our results allowed us to derive a consensus sequence for efficient gene activation by Cat8 (YCCNYTNRKCCG), while a more specific motif is required for activation by Sip4 (TCCATTSRTCCGR). Although their zinc cluster domains are clearly related, Cat8 and Sip4 are not isofunctional. This conclusion is further supported by the finding that biosynthetic derepression of Cat8 in the presence of a nonfermentable carbon source precedes that of Sip4 by about 90 min.