Inositol and phosphate regulate GIT1 transcription and glycerophosphoinositol incorporation in Saccharomyces cerevisiae

Abstract

Glycerophosphoinositol is produced through deacylation of the essential phospholipid phosphatidylinositol. In Saccharomyces cerevisiae, the glycerophosphoinositol produced is excreted from the cell but is recycled for phosphatidylinositol synthesis when inositol is limiting. To be recycled, glycerophosphoinositol enters the cell through the permease encoded by GIT1. The transport of exogenous glycerophosphoinositol through Git1p is sufficiently robust to support the growth of an inositol auxotroph (ino1Delta). We now report that S. cerevisiae also uses exogenous phosphatidylinositol as an inositol source. Evidence suggests that phosphatidylinositol is deacylated to glycerophosphoinositol extracellularly before being transported across the plasma membrane by Git1p. A genetic screen identified Pho86p, which is required for targeting of the major phosphate transporter (Pho84p) to the plasma membrane, as affecting the utilization of phosphatidylinositol and glycerophosphoinositol. Deletion of PHO86 in an ino1Delta strain resulted in faster growth when either phosphatidylinositol or glycerophosphoinositol was supplied as the sole inositol source. The incorporation of radiolabeled glycerophosphoinositol into an ino1Delta pho86Delta mutant was higher than that into wild-type, ino1Delta, and pho86Delta strains. All strains accumulated the most GIT1 transcript when incubated in media limited for inositol and phosphate in combination. However, the ino1Delta pho86Delta mutant accumulated approximately threefold more GIT1 transcript than did the other strains when incubated in inositol-free media containing either high or low concentrations of P(i). Deletion of PHO4 abolished GIT1 transcription in a wild-type strain. These results indicate that the transport of glycerophosphoinositol by Git1p is regulated by factors affecting both inositol and phosphate availabilities and suggest a regulatory connection between phosphate metabolism and phospholipid metabolism.

FIG. 1.

Deletion of PHO86 results in enhanced growth of an ino1Δ strain in GroPIns-containing and PI-containing media. Strains were grown on synthetic media containing I⁺ (A), no inositol source (I⁻) (B), 75 mM GroPIns (G⁺) (C), and 75 mM PI (PI⁺) (D). At the indicated times, the A₆₀₀ was measured. Data for wild-type (wt), git1Δ, pho86Δ, and pho86Δ git1Δ strains are shown only in panel B.

FIG. 2.

Growth of an ino1Δ mutant on PI requires one or more PLB gene products. Strains were grown on synthetic media containing 75 mM PI, a source of inositol. At the indicated times, the A₆₀₀ was measured.

FIG. 3.

GIT1 transcripts accumulate in a wild-type strain limited for phosphate and inositol. A wild-type strain (JPV91) was grown in media containing (I+) or lacking (I−) 75 mM inositol and containing 0.2 mM (low Pi) or 10 mM (high Pi) inorganic phosphate. Cells were harvested in the logarithmic phase, and Northern analysis was performed with digoxigenin-labeled probes for GIT1 and control (SNR17) transcripts as described in Materials and Methods.

FIG. 4.

An ino1Δ pho86Δ mutant overexpresses GIT1, as measured by chloramphenicol acetyltransferase (CAT) activity. Cultures transformed with plasmid pCA998 were grown in high-P_i, I⁺ medium lacking tryptophan and were used to inoculate each of the following media lacking tryptophan: low P_i, I⁻; low P_i, I⁺; high P_i, I⁻; and high P_i, I⁺. Following three to five generations of growth, lysates obtained by glass bead breakage were assayed for chloramphenicol acetyltransferase activity as described in Materials and Methods. Reactions were carried out for 2.5 h with 5 to 25 μg of protein per assay. Data represent at least three separate determinations and standard errors of the means. wt, wild type.

FIG. 5.

Pho4p is required for GIT1 expression but not INO1 expression. (A) Strains (JPV203, JPV296, and JPV339) transformed with plasmid pCA999 were grown in high-P_i, I⁺ medium lacking leucine and were used to inoculate high-P_i, I⁻ and high-P_i, I⁺ media lacking leucine. Following three to five generations of growth, lysates obtained by glass bead breakage were assayed for chloramphenicol acetyltransferase (CAT) activity as described in Materials and Methods. Reactions were carried out for 2.5 h with 5 to 25 μg of protein per assay. (B) Strains (see above) transformed with plasmid pJH359 (INO1-CYC1-lacZ) were grown in high-P_i, I⁺ medium lacking uracil and were used to inoculate high-P_i, I⁻ and high-P_i, I⁺ media lacking uracil. Following three to five generations of growth, cells were assayed for β-galactosidase activity by using the Pierce yeast β-galactosidase assay kit. Data represent at least three separate determinations and standard errors of the means. wt, wild type; OD₆₀₀, optical density at 600 nm.