Glycogen synthase kinase-3 is required for optimal de novo synthesis of inositol

Studies have shown that the inositol biosynthetic pathway and the enzyme glycogen synthase kinase-3 (GSK-3) are targets of the mood-stabilizing drugs lithium and valproate. However, a relationship between these targets has not been previously described. We hypothesized that GSK-3 may play a role in inositol synthesis, and that loss of GSK-3 may lead to inositol depletion, thus providing a mechanistic link between the two drug targets. Utilizing a yeast Saccharomyces cerevisiae gsk-3Delta quadruple-null mutant, in which all four genes encoding homologues of mammalian GSK-3 are disrupted, we tested the hypothesis that GSK-3 is required for de novo inositol biosynthesis. The gsk-3Delta mutant exhibited multiple features of inositol depletion, including defective growth in inositol-lacking medium, decreased intracellular inositol, increased INO1 and ITR1 expression, and decreased levels of phosphatidylinositol. Treatment of wild-type cells with a highly specific GSK-3 inhibitor led to a significant increase in INO1 expression. Supplementation with inositol alleviated the temperature sensitivity of gsk-3Delta. Activity of myo-inositol-3 phosphate synthase, the rate-limiting enzyme in inositol de novo biosynthesis, was decreased in gsk-3Delta. These results demonstrate for the first time that GSK-3 is required for optimal myo-inositol-3 phosphate synthase activity and de novo inositol biosynthesis, and that loss of GSK-3 activity causes inositol depletion.     

Lithium and valproate decrease the membrane phosphatidylinositol/phosphatidylcholine ratio

Lithium and valproate, two structurally different anti-bipolar drugs, cause decreased intracellular inositol in the yeast Saccharomyces cerevisiae and an in-crease in expression of a structural (INO1) and a regulatory (INO2) gene for phospholipid synthesis that responds to inositol depletion (Vaden, D., Ding, D., Peterson, B., and Greenberg, M.L., 2001, J Biol Chem 276: 15466-15471). We report here that both drugs decrease the relative rate of membrane phosphatidylinositol synthesis and, to a lesser but still significant degree, the steady state relative phosphatidylinositol composition. In addition, both drugs increase the rate of phosphatidylcholine (PC) synthesis. Finally, valproate, but not lithium, increases expression of phosphatidylcholine pathway genes CHO1 and OPI3. The overall effect on membrane phospholipid composition is a reduction in the phosphatidylinositol/phosphatidylcholine ratio by both drugs. Because maintenance of the appropriate phosphatidylinositol/phosphatidylcholine ratio is required for secretory vesicle formation, a decrease in this ratio may have far-reaching implications for understanding the therapeutic mechanisms of action of these drugs.     

Expression of yeast INM1 encoding inositol monophosphatase is regulated by inositol, carbon source and growth stage and is decreased by lithium and valproate

Inositol monophosphatase plays a vital role in the de novo biosynthesis of inositol and in the phosphoinositide second messenger signalling pathway. We cloned the Saccharomyces cerevisiae open reading frame (ORF) YHR046c (termed INM1), which encodes inositol monophosphatase, characterized the protein Inm1p and analysed expression of the INM1 gene. INM1 was expressed in bacteria under the control of the lacZ promoter. The purified protein has inositol monophosphatase activity that is inhibited by the antibipolar drug lithium, but not valproate. In the inm1Delta:URA3 null mutant, inositol monophosphatase activity was reduced but not eliminated. The disruption had little effect on growth in the presence of lithium or valproate and no effect on growth in the absence of inositol. To characterize the regulation of INM1, we examined the effects of inositol, carbon source, growth phase, and the antibipolar drugs lithium and valproate on INM1 expression using an INM1-lacZ reporter gene. Unlike all other phospholipid biosynthetic enzyme-encoding genes studied, which contain the UASINO regulatory element, INM1 expression is increased in the presence of inositol. In addition, INM1 expression was repressed during growth in glycerol and derepressed as glucose-grown cells entered stationary. Both lithium and valproate, which cause a decrease in intracellular inositol, effect a decrease in INM1 expression. A model is presented to account for regulation of INM1 expression.     

Regulation of the yeast INO1 gene. The products of the INO2, INO4 and OPI1 regulatory genes are not required for repression in response to inositol

The ino2Delta, ino4Delta, opi1Delta, and sin3Delta mutations all affect expression of INO1, a structural gene for inositol-1-phosphate synthase. These same mutations affect other genes of phospholipid biosynthesis that, like INO1, contain the repeated element UAS(INO) (consensus 5' CATGTGAAAT 3'). In this study, we evaluated the effects of these four mutations, singly and in all possible combinations, on growth and expression of INO1. All strains carrying an ino2Delta or ino4Delta mutation, or both, failed to grow in medium lacking inositol. However, when grown in liquid culture in medium containing limiting amounts of inositol, the opi1Delta ino4Delta strain exhibited a level of INO1 expression comparable to, or higher than, the wild-type strain growing under the same conditions. Furthermore, INO1 expression in the opi1Delta ino4Delta strain was repressed in cells grown in medium fully supplemented with both inositol and choline. Similar results were obtained using the opi1Delta ino2Delta ino4Delta strain. Regulation of INO1 was also observed in the absence of the SIN3 gene product. Therefore, while Opi1p, Sin3p, and the Ino2p/Ino4p complex all affect the overall level of INO1 expression in an antagonistic manner, they do not appear to be responsible for transmitting the signal that leads to repression of INO1 in response to inositol. Various models for Opi1p function were tested and no evidence for binding of Opi1p to UAS(INO), or to Ino2p or Ino4p, was obtained.     

A dominant mutation that alters the regulation of INO1 expression in Saccharomyces cerevisiae.

A dominant, single nuclear gene mutation, CSE1, caused inositol auxotrophy in yeast cells. The inositol requirement was marked when choline was present in the medium. Inositol-1-phosphate synthase, the regulatory enzyme of inositol synthesis, is repressed by inositol, or more profoundly by a combination of inositol and choline in the wild type. In CSE1, the level of inositol-1-phosphate synthase was low and was greatly repressed on the addition of choline alone. In accordance with this, INO1 mRNA encoding the enzyme was low even under the depressed conditions and was profoundly decreased by choline in CSE1. But in the wild type, the addition of choline alone had little effect. An INO1-lacZ fusion was constructed and the control of the INO1 promoter in CSE1 was studied. lacZ expression was repressed not only by inositol, but also by choline in CSE1, whereas it was repressed by inositol, but only slightly by choline in the wild type. CSE1 was unlinked to the INO1 structural gene. Thus CSE1 was thought to be a regulatory mutation. Furthermore, when the CDP-choline pathway was mutationally blocked, choline did not affect INO1 expression, indicating that the metabolism of choline via the CDP-choline pathway is required for INO1 repression.     

Valproate disrupts regulation of inositol responsive genes and alters regulation of phospholipid biosynthesis

Valproate (VPA) is one of the two drugs approved by the Food and Drug Administration (FDA) for the treatment of bipolar disorder. The therapeutic mechanism of VPA has not been established. We have shown previously that growth of the yeast Saccharomyces cerevisiae in the presence of VPA causes a decrease in intracellular inositol and inositol-1-P, and a dramatic increase in expression of INO1, which encodes the rate limiting enzyme for de novo inositol biosynthesis. To understand the underlying mechanism of action of VPA, INO1, CHO1 and INO2 expression, intracellular inositol and phospholipid biosynthesis were studied as a function of acute and chronic exposure of growing cells to the drug. A decrease in intracellular inositol was apparent immediately after addition of VPA. Surprisingly, expression of genes that are usually derepressed during inositol depletion, including INO1, CHO1 and INO2 (that contain inositol-responsive UASINO sequences) decreased several fold during the first hour, after which expression began to increase. Incorporation of 32Pi into total phospholipids was significantly decreased. Pulse labelling of CDP-DG and PG, shown previously to increase during inositol depletion, increased within 30 min. However, pulse labelling of PS, which normally increases during inositol depletion, was decreased within 30 min. PS synthase activity in cell extracts decreased with time, although VPA did not directly inhibit PS synthase enzyme activity. Thus, in contrast to the effect of chronic VPA treatment, short-term exposure to VPA abrogated the normal response to inositol depletion of inositol responsive genes and led to aberrant synthesis of phospholipids.     

Neurodevelopment and mood stabilizers

Mood disorders and schizophrenia share a number of common properties, including: genetic susceptibility; differences in brain structure and drug based therapy. Some genetic loci may even confer susceptibility for bipolar mood disorder and schizophrenia, and some atypical antipsychotic drugs are used as mood stabilizers. As schizophrenia is associated with aberrant neurodevelopment, could this also be true for mood disorders? Such changes could arise pre- or post-natal, however the recent interest in neurogenesis in the adult brain has suggested involvement of these later processes in the origins of mood disorders. Interestingly, the common mood stabilizing drugs, lithium, valproic acid (VPA) and carbamazepine, are teratogens, affecting a number of aspects of animal development. Recent work has shown that lithium and VPA interfere with normal cell development, and all three drugs affect neuronal morphology. The molecular basis for mood stabilizer action in the treatment of mood is unknown, however these studies have suggested both targets and potential mechanisms. Lithium directly inhibits two evolutionarily conserved signal transduction pathways: the protein kinase Glycogen Synthase Kinase-3 (GSK-3) and inositol signaling. VPA can up-regulate gene expression through inhibition of histone deacetylase (HDAC) and indirectly reduce GSK-3 activity. VPA effects are not conserved between cell types, and carbamazepine has no effect on the GSK-3 pathway. All three mood stabilizers suppress inositol signaling, results further supported by studies on the enzyme prolyl oligopeptidase (PO) and the sodium myo-inositol transporter (SMIT). Despite these intriguing observations, it remains unclear whether GSK-3, inositol signaling or both underlie the origins of bipolar disorder.     

Biosynthesis of inositol in yeast. Primary structure of myo-inositol-1-phosphate synthase (EC 5.5.1.4) and functional analysis of its structural gene, the INO1 locus

A biochemical, molecular, and genetic analysis of the Saccharomyces cerevisiae INO1 gene and its product, L-myo-inositol-1-phosphate synthase (EC 5.5.1.4) has been carried out. The sequence of the entire INO1 gene and surrounding regions has been determined. Computer analysis of the DNA sequence revealed four potential peptides. The largest open reading frame of 553 amino acids predicted a peptide with a molecular weight of 62,842. The amino acid composition and amino terminus of purified L-myo-inositol-1-phosphate synthase were chemically determined and compared to the amino acid composition and amino terminus of the protein predicted from the DNA sequence of the large open reading frame. This analysis established that the large open reading frame encodes L-myo-inositol-1-phosphate synthase. The largest of several small open reading frames adjacent to INO1 predicted a protein of 133 amino acids with a molecular weight of 15,182 and features which suggested that the encoded protein may be membrane-associated. A gene disruption was constructed at INO1 by eliminating a portion of the coding sequence and replacing it with another sequence. Strains carrying the gene disruption failed to express any protein cross-reactive to antibody directed against L-myo-inositol-1-phosphate synthase. Although auxotrophic for inositol, strains carrying the gene disruption were completely viable when supplemented with inositol. In a similar fashion, a gene disruption was constructed in the chromosomal locus of the 133-amino acid open reading frame. This mutation did not affect viability but did cause inositol to be excreted from the cell.     

Identification of the INO1 gene of Mycobacterium tuberculosis H37Rv reveals a novel class of inositol-1-phosphate synthase enzyme.

1L-myo-inositol (inositol) is vital for the biogenesis of mycothiol, phosphatidylinositol and glycosylphosphatidylinositol anchors linked to complex carbohydrates in Mycobacterium tuberculosis. All these cellular components are thought to play important roles in host-pathogen interactions and in the survival of the pathogen within the host. However, the inositol biosynthetic pathway in M. tuberculosis is not known. To delineate the pathways for inositol formation, we employed a unique combination of tertiary structure prediction and yeast-based functional assays. Here, we describe the identification of the gene for mycobacterial INO1 that encodes inositol-1-phosphate synthase distinct in many respects from the eukaryotic analogues.     

The INO2 and INO4 loci of Saccharomyces cerevisiae are pleiotropic regulatory genes.

We isolated a mutant of Saccharomyces cerevisiae defective in the formation of phosphatidylcholine via methylation of phosphatidylethanolamine. The mutant synthesized phosphatidylcholine at a reduced rate and accumulated increased amounts of methylated phospholipid intermediates. It was also found to be auxotrophic for inositol and allelic to an existing series of ino4 mutants. The ino2 and ino4 mutants, originally isolated on the basis of an inositol requirement, are unable to derepress the cytoplasmic enzyme inositol-1-phosphate synthase (myo-inositol-1-phosphate synthase; EC 5.5.1.4). The INO4 and INO2 genes were, thus, previously identified as regulatory genes whose wild-type product is required for expression of the INO1 gene product inositol-1-phosphate synthase (T. Donahue and S. Henry, J. Biol. Chem. 256:7077-7085, 1981). In addition to the identification of a new ino4-allele, further characterization of the existing series of ino4 and ino2 mutants, reported here, demonstrated that they all have a reduced capacity to convert phosphatidylethanolamine to phosphatidylcholine. The pleiotropic phenotype of the ino2 and ino4 mutants described in this paper suggests that the INO2 and INO4 loci are involved in the regulation of phospholipid methylation in the membrane as well as inositol biosynthesis in the cytoplasm.     

Inositol regulates phosphatidylglycerolphosphate synthase expression in Saccharomyces cerevisiae.

The enzyme phosphatidylglycerolphosphate synthase (PGPS; CDPdiacylglycerol-glycerol-3-phosphate 3-phosphatidyltransferase; EC 2.7.8.5) catalyzes the committed step in the synthesis of cardiolipin, a phospholipid found predominantly in the mitochondrial inner membrane. To determine whether PGPS is regulated by cross-pathway control, we analyzed PGPS expression under conditions in which the regulation of general phospholipid synthesis could be examined. The addition of inositol resulted in a three- to fivefold reduction in PGPS expression in wild-type cells in the presence or absence of exogenous choline. The reduction in enzyme activity in response to inositol was seen in minutes, suggesting that inactivation or degradation of the enzyme plays an important role in inositol-mediated repression of PGPS. In cho2 and opi3 mutants, which are blocked in phosphatidylcholine synthesis, inositol-mediated repression of PGPS did not occur unless choline was added to the media. Three previously identified genes that regulate general phospholipid synthesis, INO2, INO4, and OP11, did not affect PGPS expression. Thus, ino2 and ino4 mutants, which are unable to derepress biosynthetic enzymes involved in general phospholipid synthesis, expressed wild-type levels of PGPS activity under derepressing conditions. PGPS expression in the opi1 mutant, which exhibits constitutive synthesis of general phospholipid biosynthetic enzymes, was fully repressed in the presence of inositol and partially repressed even in the absence of inositol. These results demonstrate for the first time that an enzymatic step in cardiolipin synthesis is coordinately controlled with general phospholipid synthesis but that this control is not mediated by the same genetic regulatory circuit.