Derepression of INO1 transcription requires cooperation between the Ino2p-Ino4p heterodimer and Cbf1p and recruitment of the ISW2 chromatin-remodeling complex

Abstract

The Saccharomyces cerevisiae INO1 gene encodes the structural enzyme inositol-3-phosphate synthase for the synthesis de novo of inositol and inositol-containing phospholipids. The transcription of INO1 is completely derepressed in the absence of inositol and choline (I(-) C(-)). Derepression requires the binding of the Ino2p-Ino4p basic helix-loop-helix (bHLH) heterodimer to the UAS(INO) promoter element. We report here the requirement of a third bHLH protein, centromere-binding factor 1 (Cbf1p), for the complete derepression of INO1 transcription. We found that Cbf1p regulates INO1 transcription by binding to sites distal to the INO1 promoter and encompassing the upstream SNA3 open reading frame (ORF) and promoter. The binding of Cbf1p requires Ino2p-Ino4p binding to the UAS(INO) sites in the INO1 promoter and vice versa, suggesting a cooperative mechanism. Furthermore, Cbf1p binding to the upstream sites was required for the binding of the ISW2 chromatin-remodeling complex to the Ino2p-Ino4p-binding sites on the INO1 promoter. Consistent with this, ISW2 was also required for the complete derepression of INO1 transcription.

Fig. 1.

Model for regulation of INO1 transcription. Shown is a schematic of the INO1 promoter containing two UAS_INO elements (shown in green) and a repressor site (URS1) (shown in red). Under derepressing conditions (I⁻ C⁻), the Ino2p-Ino4p heterodimer binds to two UAS_INO elements and recruits INO80 and Snf1p. INO80 is a chromatin-remodeling complex that recruits another remodeling complex, SWI/SNF. Snf1p is a kinase that phosphorylates serine 10 (S10) on histone H3, which in turn recruits SAGA, which acetylates lysine 14 of histone H3. Histone 3 S10-P promotes the interaction of the TATA-binding protein (TBP) with the INO1 TATA sequence. Under derepressing conditions, the Opi1p repressor is complexed with phosphatidic acid (PA) and retained in the endoplasmic reticulum (ER) (shown in yellow) by association with Scs2p. Under repressing conditions (I⁺ C⁺), PA levels drop, and Opi1p is released from the ER, translocates to the nucleus, and associates with Ino2p, repressing transcription. URS1 is a binding site for the general repressor Ume6p, which recruits the Sin3p corepressor and the Rpd3p histone deacetylase (HDAC) complex.

Fig. 2.

CBF1 regulates INO1-lacZ expression. (A) WT and isogenic bHLH knockout strains were transformed with an INO1-lacZ plasmid (pJH330). Transformants were grown in four different media: I⁻ C⁻ high P_i, I⁻ C⁻ low P_i, I⁺ C⁺ high P_i, and I⁺ C⁺ low P_i. Green and red bars indicate derepressing and repressing conditions, whereas dark and light indicate high and low P_i, respectively. In the case of ino2Δ and ino4Δ, the I⁻ C⁻ media had 10 μM inositol to allow the growth of these auxotrophic strains. Cells were harvested in mid-log phase and assayed for β-galactosidase activity. (B) Complementation test of the cbf1Δ INO1-lacZ phenotype. The cbf1Δ strain was cotransformed with the INO1-lacZ plasmid and either a pRS315-CBF1 plasmid or pRS315, and an isogenic WT strain was cotransformed with INO1-lacZ and plasmid pRS315. These transformants were assayed for β-galactosidase activity. The data represent means and standard errors of the means from at least three different experiments.

Fig. 3.

Quantification of INO1 transcript levels in WT and cbf1Δ strains. (A) Isogenic WT and cbf1Δ strains were grown to mid-log phase under derepressing (I⁻ C⁻) and repressing (I⁺ C⁺) conditions, and INO1 transcript levels were quantified by QRT-PCR. (B) Similarly, isogenic WT, isw2Δ, and itc1Δ strains were grown as described above. INO1 transcript levels were normalized to TCM1 transcript levels. The data represent means and standard errors of the means from at least three different experiments.

Fig. 4.

ChIP analysis of Ino2p-Ino4p binding to the INO1 promoter and upstream regions. (A) Schematic showing primer positions (A to E) and E boxes (E1 to E5) relative to the INO1 and SNA3 ORFs. (B and C) Ino2p-TAP and Ino4p-TAP bind to the INO1 promoter. ChIP analysis was performed by using TAP-tagged strains grown under derepressing (I⁻ C⁻) and repressing (I⁺ C⁺) conditions. Enrichment on the INO1 promoter and upstream regions was quantified by using QRT-PCR. ChIP/immunoprecipitation (IP) ratios were normalized by using TCM1. The INO1 ORF primers cover a region within the INO1 coding sequence and serve as a negative control. The data represent means and standard errors of the means from at least three different experiments.

Fig. 5.

ChIP analysis of Cbf1p binding to the INO1 promoter and upstream regions. Cbf1p-TAP binds to regions upstream of the INO1 promoter within and upstream of the SNA3 ORF. ChIP analysis was performed by using a CBF1-TAP-tagged strain grown under derepressing (I⁻ C⁻) and repressing (I⁺ C⁺) conditions. Enrichment on the INO1 promoter and upstream regions was quantified by using QRT-PCR. ChIP/IP ratios were normalized by using TCM1. The data represent means and standard errors of the means from at least three different experiments.

Fig. 6.

Cbf1p regulation of INO1 expression depends on Ino2p-Ino4p-binding sites in the INO1 promoter. (A) WT and isogenic cbf1Δ knockout strains were transformed with an INO1-1200-lacZ plasmid (wild-type INO1 promoter) or an INO1-100-lacZ plasmid (UAS_INO-deleted INO1 promoter). (B) WT and isogenic cbf1Δ knockout strains were transformed with pJH330 containing mutations of the E1 box, the E2 box, or both the E1 and E2 boxes. These transformants were assayed for β-galactosidase activity as described above. The data represent means and standard errors of the means from at least three different experiments.

Fig. 7.

Cbf1p and Ino2p-Ino4p bindings to the INO1 promoter are interdependent. (A) Cbf1p binding requires the Ino2p-Ino4p heterodimer. ChIP analysis was performed by using CBF1-TAP-tagged ino2Δ and ino4Δ strains under derepressing (I⁻ C⁻) and repressing (I⁺ C⁺) conditions. (B) Ino2p-Ino4p binding requires Cbf1p. ChIP analysis was performed by using INO2- and INO4-TAP-tagged cbf1Δ strains as described above. Enrichment on the INO1 promoter and upstream regions was quantified by using QRT-PCR and the primer pairs described in the legend of Fig. 4A. ChIP/IP ratios were normalized by using TCM1. The data represent means and standard errors of the means from at least three different experiments.

Fig. 8.

Isw2p recruitment to the INO1 promoter requires Cbf1p. ChIP analysis was performed by using ISW2-TAP (A) and ISW2-TAP cbf1Δ strains (B) under derepressing (I⁻ C⁻) and repressing (I⁺ C⁺) conditions. Enrichment on the INO1 promoter was quantified by using QRT-PCR and the primer pairs described in the legend of Fig. 4A. ChIP/IP ratios were normalized by using TCM1. The data represent means and standard errors of the means from at least three different experiments.

Fig. 9.

Model for regulation of INO1 transcription by Ino2p-Ino4p, Cbf1p, and ISW2. Black arrows indicate the positions of genes, and green bars indicate the positions of UAS_INO elements and other potential E boxes. Numbered arrows indicate the sequence of events. Refer to Discussion for a complete description of the model.