Phosphatidylinositol phosphate kinases

Phosphatidylinositol 4,5-bisphosphate is a multi-functional lipid, whose proposed functions now number more than twenty, covering many aspects of cell biology in several different subcellular compartments. The enzymes primarily responsible for synthesizing this lipid, the Type I phosphatidylinositol 4-phosphate 5-kinases, are therefore a tightly regulated and diverse family. Here we review our current knowledge about these enzymes and how they may be regulated.

Tissue distribution of GAP1(IP4BP) and GAP1(m): two inositol 1,3,4,5-tetrakisphosphate-binding proteins

Two members of the GAP1 family, GAP1(IP4BP) and GAP1(m), have been shown to bind the putative second messenger Ins(1,3,4,5)P4 with high affinity and specificity, though other aspects of their behaviour suggest that in vivo, whereas GAP1(IP4BP) may function as an Ins(1,3,4,5)P4 receptor, GAP1(m) may be a receptor for the lipid second messenger PtdIns(3,4,5)P3. As a step towards clarifying their cellular roles, we describe here how we have raised and characterised antisera that are specific for the two proteins, and used these to undertake a comprehensive study of their tissue distribution. Both proteins are widely expressed, but there are several clear differences between them in the tissues that show the highest levels of expression.

Inositol 1,4,5-trisphosphate 3-kinase A associates with F-actin and dendritic spines via its N terminus

The consequences of the rapid 3-phosphorylation of inositol 1,4,5-trisphosphate (IP(3)) to produce inositol 1,3,4,5-tetrakisphosphate (IP(4)) via the action of IP(3) 3-kinases involve the control of calcium signals. Using green fluorescent protein constructs of full-length and truncated IP(3) 3-kinase isoform A expressed in HeLa cells, COS-7 cells, and primary neuronal cultures, we have defined a novel N-terminal 66-amino acid F-actin-binding region that localizes the kinase to dendritic spines. The region is necessary and sufficient for binding F-actin and consists of a proline-rich stretch followed by a predicted alpha-helix. We also localized endogenous IP(3) 3-kinase A to the dendritic spines of pyramidal neurons in primary hippocampal cultures, where it is co-localized postsynaptically with calcium/calmodulin-dependent protein kinase II. Our experiments suggest a link between inositol phosphate metabolism, calcium signaling, and the actin cytoskeleton in dendritic spines. The phosphorylation of IP(3) in dendritic spines to produce IP(4) is likely to be important for modulating the compartmentalization of calcium at synapses.

Inositol lipids are regulated during cell cycle progression in the nuclei of murine erythroleukaemia cells

Previous data suggest the existence of discrete pools of inositol lipids, which are components of a nuclear phosphoinositide (PI) cycle. However, it is not known whether the contents of these pools are regulated during cell proliferation. In the present study we demonstrate that the mass levels of three important constituents of the nuclear PI cycle are regulated during the cell cycle. Radioactive label incorporation into PtdIns(4,5)P(2) was seen to increase dramatically as synchronized cells entered S-phase. This did not coincide with any significant changes in the nuclear mass levels of this lipid, suggesting that the rate of turnover of this molecule was increased. Levels of PtdIns4P, the major substrate for PtdIns(4,5)P(2) production by Type I PtdInsP kinases (PIPkins), were regulated during the cell cycle and indicated a complex relationship between these two lipids. An alternative substrate for PtdIns(4,5)P(2), PtdIns5P, phosphorylated by Type II PIPkins, was present in nuclei at much smaller amounts than the PtdIns4P, and thus is unlikely to contribute significantly to PtdIns(4,5)P(2) turnover. However, a large increase in nuclear PtdIns5P mass was observed when murine erythroleukaemia cells are in G(1), and this could represent a potential pool of nuclear inositol lipid that has a specific signalling role. Analysis of extracted lipid fractions indicated the absence of any PtdIns3P in these nuclei.

Back in the water: the return of the inositol phosphates

Following the discovery of inositol-1,4,5-trisphosphate as a second messenger, many other inositol phosphates were discovered in quick succession, with some understanding of their synthesis pathways and a few guesses at their possible functions. But then it all seemed to go comparatively quiet, with an explosion of interest in the inositol lipids. Now the water-soluble phase is once again becoming a focus of interest. Old and new data point to a new vista of inositol phosphates, with functions in many diverse aspects of cell biology, such as ion-channel physiology, membrane dynamics and nuclear signalling.

Thrombin stimulation of platelets causes an increase in phosphatidylinositol 5-phosphate revealed by mass assay

Phosphatidylinositol 5-phosphate (PtdIns5P), a novel inositol lipid, has been shown to be the major substrate for the type II PtdInsP kinases (PIPkins) ¿Rameh et al. (1997) Nature 390, 192-196. A PtdInsP fraction was prepared from cell extracts by neomycin chromatography, using a protocol devised to eliminate the interaction of acidic solvents with plasticware, since this was found to inhibit the enzyme. The PtdIns5P in this fraction was measured by incubating with ¿gamma-(32)PATP and recombinant PIPkin IIalpha, and quantifying the radiolabelled PtdInsP(2) formed. This assay was used on platelets to show that during 10 min stimulation with thrombin, the mass level of PtdIns5P increases, implying the existence of an agonist-stimulated synthetic mechanism.

Type I phosphatidylinositol 4-phosphate 5-kinase directly interacts with ADP-ribosylation factor 1 and is responsible for phosphatidylinositol 4,5-bisphosphate synthesis in the golgi compartment

Phosphatidylinositol (PtdIns) 4,5-bisphosphate is involved in many aspects of membrane traffic, but the regulation of its synthesis is only partially understood. Golgi membranes contain PI 4-kinase activity and a pool of phosphatidylinositol phosphate (PIP), which is further increased by ADP-ribosylation factor 1 (ARF1). COS7 cells were transfected with alpha and beta forms of PI 4-kinase, and only membranes from COS7 cells transfected with PI 4-kinase beta increased their content of PIP when incubated with ARF1. PtdIns(4, 5)P(2) content in Golgi membranes was nonexistent but could be increased to a small extent upon adding either cytosol or Type I or Type II PIP kinases. However, when ARF1 was present, PtdIns(4,5)P(2) levels increased dramatically when membranes were incubated in the presence of cytosol or Type I, but not Type II, PIP kinase. To examine whether ARF1 could directly activate Type I PIP 5-kinase, we used an in vitro assay consisting of phosphatidycholine-containing liposomes, ARF1, and PIP 5-kinase. ARF1 increased Type I PIP 5-kinase activity in a guanine nucleotide-dependent manner, identifying this enzyme as a direct effector for ARF1.

The effect of inositol 1,3,4,5-tetrakisphosphate on inositol trisphosphate-induced Ca2+ mobilization in freshly isolated and cultured mouse lacrimal acinar cells

Earlier reports have shown a remarkable synergism between InsP(4) and InsP(3) [either Ins(1,4,5)P(3) or Ins(2,4,5)P(3)] in activating Ca(2+)-dependent K(+) and Cl(-) currents in mouse lacrimal cells [Changya, Gallacher, Irvine, Potter and Petersen (1989) J. Membr. Biol. 109, 85-93; Smith (1992) Biochem. J. 283, 27-30]. However, Bird, Rossier, Hughes, Shears, Armstrong and Putney [(1991) Nature (London) 352, 162-165] reported that they could see no such synergism in the same cell type. A major experimental difference between the two laboratories lies in whether or not the cells were maintained in primary culture before use. Here we have compared directly the responses to inositol polyphosphates in freshly isolated cells versus cells cultured for 6-72 h. In the cultured cells, Ins(2,4,5)P(3) at 100 microM produced a robust stimulation of K(+) and Cl(-) currents, as much as an order of magnitude greater than that observed in the freshly isolated cells. However, the freshly isolated cells could be restored to a sensitivity similar to cultured cells by the addition of InsP(4) at a concentration two orders of magnitude lower than that of Ins(2,4,5)P(3). We discuss the implications of this with respect to the actions of InsP(4), including the possibility that disruption of the cellular structure during the isolation of the cells exposes an extreme manifestation of a possible physiological role for InsP(4) in controlling calcium-store integrity.

Nuclear targeting of the beta isoform of type II phosphatidylinositol phosphate kinase (phosphatidylinositol 5-phosphate 4-kinase) by its alpha-helix 7

Type II phosphatidylinositol phosphate kinases (PIPkins) have recently been found to be primarily phosphatidylinositol 5-phosphate 4-kinases, and their physiological role remains unclear. We have previously shown that a Type II PIPkin [isoform(s) unknown], is localized partly in the nucleus [Divecha, Rhee, Letcher and Irvine (1993) Biochem. J. 289, 617-620], and here we show, by transfection of HeLa cells with green-fluorescent-protein-tagged Type II PIPkins, that this is likely to be the Type IIbeta isoform. Type IIbeta PIPkin has no obvious nuclear localization sequence, and a detailed analysis of the localization of chimaeras and mutants of the alpha (cytosolic) and beta PIPkins shows that the nuclear localization requires the presence of a 17-amino-acid length of alpha-helix (alpha-helix 7) that is specific to the beta isoform, and that this helix must be present in its entirety, with a precise orientation. This resembles the nuclear targeting of the HIV protein Vpr, and Type IIbeta PIPkin is apparently therefore the first example of a eukaryotic protein that uses the same mechanism.

PiUS (Pi uptake stimulator) is an inositol hexakisphosphate kinase

A cDNA cloned from its ability to stimulate inorganic phosphate uptake in Xenopus oocytes (phosphate uptake stimulator (PiUS)) shows significant similarity with inositol 1,4,5-trisphosphate 3-kinase. However, the expressed PiUS protein showed no detectable activity against inositol 1,4,5-trisphosphate, nor the 1,3,4,5- or 3,4,5, 6-isomers of inositol tetrakisphosphate, whereas it was very active in converting inositol hexakisphosphate (InsP(6)) to inositol heptakisphosphate (InsP(7)). PiUS is a member of a family of enzymes found in many eukaryotes and we discuss the implications of this for the functions of InsP(7) and for the evolution of inositol phosphate kinases.

Regulation of type IIalpha phosphatidylinositol phosphate kinase localisation by the protein kinase CK2

Inositol lipid synthesis is regulated by several distinct families of enzymes [1]. Members of one of these families, the type II phosphatidylinositol phosphate kinases (PIP kinases), are 4-kinases and are thought to catalyse a minor route of synthesis of the multifunctional phosphatidylinositol 4,5-bisphosphate (PI(4,5)P(2)) from the inositide PI(5)P [2]. Here, we demonstrate the partial purification of a protein kinase that phosphorylates the type IIalpha PIP kinase at a single site unique to that isoform - Ser304. This kinase was identified as protein kinase CK2 (formerly casein kinase 2). Mutation of Ser304 to aspartate to mimic its phosphorylation had no effect on PIP kinase activity, but promoted both redistribution of the green fluorescent protein (GFP)-tagged enzyme in HeLa cells from the cytosol to the plasma membrane, and membrane ruffling. This effect was mimicked by mutation of Ser304 to alanine, although not to threonine, suggesting a mechanism involving the unmasking of a latent membrane localisation sequence in response to phosphorylation.

Nuclei contain two differentially regulated pools of diacylglycerol

A number of recent studies have highlighted the presence of a nuclear pool of inositol lipids [1] [2] that is regulated during progression through the cell cycle [1] [3], differentiation [1] [2] and after DNA damage [2], suggesting that a number of different regulatory pathways impinge upon this pool of lipids. It has been suggested that the downstream consequence of the activation of one of these nuclear phosphoinositide (PI) regulatory pathways is the generation of nuclear diacylglycerol (DAG) [1] [3] [4], which is important in the activation of nuclear protein kinase C (PKC) [5] [6] [7]. Activation of PKC in turn appears to regulate the progression of cells through G1 and into S phase [4] and through G2 to mitosis [3] [8] [9] [10] [11]. Although the evidence is enticing, there is as yet no direct demonstration that nuclear PIs can be hydrolysed to generate nuclear DAG. Previous data in murine erythroleukemia (MEL) cells have suggested that nuclear phosphoinositidase Cbeta1 (PIC-beta1) activity is important in the generation of nuclear DAG. Here, we demonstrate that the molecular species of nuclear DAG bears little resemblance to the PI pool and is unlikely to be generated directly by hydrolysis of these inositol lipids. Further, we show that there are in fact two distinct subnuclear pools of DAG; one that is highly disaturated and mono-unsaturated (representing more than 90% of the total nuclear DAG) and one that is highly polyunsaturated and is likely to be derived from the hydrolysis of PI. Analysis of these pools, either after differentiation or during cell-cycle progression, suggests that the pools are independently regulated, possibly by the regulation of two different nuclear phospholipase Cs (PLCs).

Inositol 1,3,4,5-tetrakisphosphate as a second messenger--a special role in neurones?

There has been much controversy over the possibility that inositol 1,3,4,5-tetrakisphosphate (InsP4) may have a second messenger function. A possible resolution to this controversy may stem from the recent cloning of two putative receptors for InsP4, GAP1IP4BP and GAP1m. Both these proteins are expressed at high levels in neurones, as is inositol 1,4,5-trisphosphate 3-kinase, the enzyme that makes InsP4. In this review we discuss the possible relevance of these high expression levels to the complex way in which neurones control Ca2+ and use it as a second messenger.

Tissue-specific expression and endogenous subcellular distribution of the inositol 1,3,4,5-tetrakisphosphate-binding proteins GAP1(IP4BP) and GAP1(m)

GAP1(IP4BP) and GAP1(m) belong to the GAP1 family of Ras GTPase-activating proteins that are candidate InsP4 receptors. Here we show they are ubiquitously expressed in human tissues and are likely to have tissue-specific splice variants. Analysis by subcellular fractionation of RBL-2H3 rat basophilic leukemia cells confirms that endogenous GAP1(IP4BP) is primarily localised to the plasma membrane, whereas GAP1(m) appears localised to the cytoplasm (cytosol and internal membranes) but not the plasma membrane. Subcellular fractionation did not indicate a specific co-localisation between membrane-bound GAP1(m) and several Ca2+ store markers, consistent with the lack of co-localisation between GAP1(m) and SERCA1 upon co-expression in COS-7 cells. This difference suggests that GAP1(m) does not reside at a site where it could regulate the ability of InsP4 to release intracellular Ca2+. As GAP1(m) is primarily localised to the cytosol of unstimulated cells it may be spatially regulated in order to interact with Ras at the plasma membrane.

PIPkins1, their substrates and their products: new functions for old enzymes

The phosphatidylinositolphosphate kinases (PIPkins) are a unique family of enzymes that catalyse the production of phosphorylated inositol lipids. Recent advances have revealed that, due to their ability to utilise a number of different lipid substrates (at least in vitro), this family is potentially able to generate several distinct, physiologically important inositol lipids. Despite their importance, however, our understanding of the regulation of the PIPkins and of their physiological role in cellular signalling and regulation is still poor. Here we describe in turn the diverse physiological functions of the known substrates and major products of the PIPkins. We then examine what is known about the members of the PIPkin family themselves, and their characteristics and regulation.

Regulation of adenylyl cyclase by membrane potential

Mammalian adenylyl cyclases possess 12 transmembrane-spanning domains and bear a superficial resemblance to certain classes of ion channels. Some evidence suggests that bacterial and sea urchin sperm adenylyl cyclases can be regulated by membrane depolarization. In the present study, we explored the effect of altering membrane potential on the adenylyl cyclase activity of cerebellar granule cells with acute potassium depolarization. A biphasic stimulatory and then inhibitory response is evoked by progressive increases in the extracellular [K]:[Na] ratio in the absence of extracellular Ca2+. This effect does not mimic the linear increase in membrane potential elicited under the same conditions. Instead it appears as though membrane depolarization opens L-type (nimodipine-sensitive) Ca2+ channels, allowing the entry of Na+, which directly stimulates adenylyl cyclase activity. Gramicidin, which generates pores that are permeable to monovalent cations, and concurrently eliminates the membrane potential, permits a similar stimulation by extracellularly applied Na+. Although the results indicate no direct sensitivity of cerebellar granule cell adenylyl cyclase to membrane potential, they do demonstrate that, as a result of membrane depolarization, the influx of Na+, as well as Ca2+, will elevate cAMP levels.

Manganese-stimulated phosphatidylinositol headgroup exchange in rat liver microsomes

Manganese-dependent, CMP-independent incorporation of myo-[3H]inositol into phospholipids of rat liver microsomes was studied in an attempt to clarify the physiological significance of this headgroup-exchange reaction. The enzyme responsible worked best with Mn2+ as a co-factor, but Mg2+ at physiological concentrations supported a significant rate of incorporation. The K(m) for myo-inositol was around 11 microM, yet incorporation of myo-[3H]inositol was unaffected by as much as 5 mM choline, ethanolamine, glycerol or serine; as this is a reversible reaction, these data imply that phosphatidylinositol is the most likely lipid substrate. Similarly, other inositols showed an apparent affinity at least two orders of magnitude lower than myo-inositol. Glucosamine alpha 1-6 myo-inositol also had a low affinity for the enzyme, making it unlikely that this headgroup-exchange activity is part of a metabolic pathway for glycosyl phosphatidylinositols. The phosphatidylinositol radiolabelled by headgroup exchange was deacylated and deglycerated, and the resulting inositol phosphate headgroup cochromatographed on anion exchange HPLC with myo-inositol l-phosphate. The simplest interpretation of all the data is the apparent paradox that this enzyme functions at a slow rate under physiological conditions to remove the myo-inositol headgroup from phosphatidylinositol, only to replace it with another myo-inositol.

Modulation of Ins(2,4,5)P3-stimulated Ca2+ mobilization by ins(1,3,4, 5)P4: enhancement by activated G-proteins, and evidence for the involvement of a GAP1 protein, a putative Ins(1,3,4,5)P4 receptor

We have previously shown that addition of Ins(1,3,4,5)P4 to permeabilized L1210 cells increases the amount of Ca2+ mobilized by a submaximal concentration of Ins(2,4,5)P3, and we suggested that, in doing this, Ins(1,3,4,5)P4 is not working via an InsP3 receptor but indirectly via an InsP4 receptor [Loomis-Husselbee, Cullen, Dreikhausen, Irvine and Dawson (1996) Biochem. J. 314, 811-816]. Here we have investigated whether this effect might be mediated by GAP1(IP4BP), recently identified as a putative receptor for Ins(1,3, 4,5)P4. GAP1(IP4BP) is a protein that interacts with one or more monomeric G-proteins, so we sought evidence for involvement of monomeric G-proteins in the effects of Ins(1,3,4,5)P4 in permeabilized L1210 cells. Guanosine 5'-[gamma-thio]triphosphate (GTP[S]) enhanced the effect of Ins(1,3,4,5)P4 on Ins(2,4, 5)P3-stimulated Ca2+ mobilization, but had no effect on the action of Ins(2,4,5)P3 alone. A specific enhancement of only the action of Ins(1,3,4,5)P4 was also seen with GTP[S]-loaded R-Ras or Rap1a (two G-proteins known to interact with GAP1(IP4BP)), whereas H-Ras was inactive at similar concentrations. Guanosine 5'-[beta-thio]diphosphate (GDP[S]) did not alter the action of either Ins(2,4,5)P3 or Ins(1,3,4,5)P4. Finally, the addition of exogenous GAP1(IP4BP), purified from platelets, markedly enhanced the effect of Ins(1,3,4,5)P4, and again, the amount of Ca2+ mobilized by Ins(2,4,5)P3 alone was unaltered. We conclude that the increase in Ins(2,4,5)P3-stimulated Ca2+ mobilization by Ins(1,3,4, 5)P4 may be mediated by GAP1(IP4BP) or a closely related protein (such as GAP1(m)), and if so, the action of the GAP1 is not solely to regulate GTP loading of a G-protein, but rather it acts with a G-protein to cause its effect.

Inositide signalling in Chlamydomonas: characterization of a phosphatidylinositol 3-kinase gene

Phosphoinositide (PI) 3-kinases, which phosphorylate the D-3 position of the inositol ring, function in several different signalling pathways. The phosphatidylinositol (PtdIns)-specific PI 3-kinase of yeast (Vps34p) is part of a receptor signalling protein complex associated with the trans-Golgi membranes, whereas PI 3-kinases that phosphorylate polyphosphoinositides in animal cells form a major receptor-controlled signalling pathway in the plasma membrane. Recent studies have indicated the presence of active PLC, PLD, and PI 3-kinase-dependent signalling systems in the unicellular green alga Chlamydomonas, and PtdIns-3P in Chlamydomonas shows a particularly high rate of turnover. Here we report the cloning of the Chlamydomonas Vps34p, and some characterisation of its properties, regulation and localisation. A single-copy 12 kb gene was present. The corresponding protein of 122 kDa had full-length homology with Vps34ps from other species, but it contained a novel spacer-like insert region of 148 amino acid residues between homology region 2 (HR2) and the C-terminal catalytic core domain, and three other shorter putative inserts. Available cDNAs were used to assemble a pBluescript clone expressing a recombinant protein which had PtdIns-specific 3-kinase activity. However, an unexpected observation was that recombinant proteins containing the complete catalytic core, but lacking HR2, had no lipid kinase activity, pointing to a previously unsuspected role for this domain, possibly in substrate binding. VPS34 mRNA and protein levels, as determined by RNAse protection assays and by immunological methods respectively, were low in all cell stages that were examined. Western blotting of subcellular fractions revealed that most of Vps34p in cell lysates of cw-15 (a cell wall-deficient mutant) could be recovered in a NP-40-resistant 100000 x g pellet, suggesting that the enzyme may have a location different from that found in higher plants.

Regulation of PtdIns4P 5-kinase C by thrombin-stimulated changes in its phosphorylation state in human platelets

PtdIns(4,5)P2 production by the enzyme PtdIns4P 5-kinase C (PIPkin C) was examined in thrombin-stimulated human platelets. Thrombin caused a rapid, transient 2-3-fold increase in PIPkin activity and a transient net dephosphorylation of the enzyme. PIPkin C was phosphorylated on serine and threonine residues in unstimulated platelets; no evidence for tyrosine phosphorylation was found. The phosphatase inhibitor okadaic acid promoted PIPkin C hyperphosphorylation and a concomitant marked inhibition of its activity in immunoprecipitates. Activity was restored by treatment with alkaline phosphatase, suggesting the existence of an inhibitory phosphorylation site. In support of this idea, alkaline phosphatase treatment of PIPkin C immunoprecipitated from unstimulated platelets caused a modest (1.6-fold) but significant activation of the enzyme. However, alkaline phosphatase treatment of PIPkin C immunoprecipitated from thrombin-stimulated platelets caused a decrease in activity to approximately the same levels, suggesting that the phosphorylation of PIPkin C also contributes to the observed stimulation. Two-dimensional phosphopeptide mapping of immunoprecipitated PIPkin C revealed that the enzyme is multiply phosphorylated and that, whereas some phosphopeptides are indeed lost on stimulation, consistent with the net dephosphorylation of the enzyme, at least two novel sites become phosphorylated. This suggests that thrombin causes complex changes in the phosphorylation state of PIPkin C, one consequence of which is its activation.

Metabolism and possible compartmentalization of inositol lipids in isolated rat-liver nuclei

(1) The removal of the nuclear envelope from isolated rat-liver nuclei by washing with Triton X-100 (TX-100) was assessed by electron microscopy. All the envelope was removed by 0.04% (w/v) TX-100. (2) After this removal, phosphorylation of inositol lipids and diacylglycerol (DAG) from [gamma-32P]ATP still occurs, despite the near complete absence of detectable (by mass assay) DAG and PtdIns. This suggests that the majority of these two lipids in nuclei are present in the nuclear membrane, but the small amounts remaining after extraction, defined as intranuclear, are available for phosphorylation by lipid kinases (36% for DAG and 24% for PtdIns respectively, when expressed as a percentage of incorporation of intact nuclei). (3) PtdIns(4,5)P2 did not follow the same pattern as PtdIns and DAG; after removal of the nuclear membrane, 40% of the mass of this lipid was left in the nucleus. Moreover, a similar amount of PtdIns(4,5)P2 was also resistant to extraction with even higher concentrations of detergent, suggesting that PtdIns(4,5)P2 has a discrete intranuclear location, probably bound to nuclear proteins. (4) Addition of exogenous substrates, PtdIns, PtdIns(4)P and DAG, to membrane-depleted nuclei resulted in reconstitution of the majority of lipid phosphorylations from [gamma-32P]ATP (70%, 90% and 94% of intact nuclei respectively), suggesting a predominantly intranuclear location for the respective kinases. (5) Nuclei also showed phosphomonoesterase and phosphatidic acid hydrolase activity; dephosphorylation of pre-radiolabelled PtdIns(4)P, PtdIns(4,5)P2 and phosphatidic acid was observed when [gamma-32P]ATP was removed. However, some of the radioactivity was apparently resistant to these enzymes, suggesting the existence of multiple pools of these lipids. (6) Addition of excess non-radiolabelled ATP to nuclei pre-labelled with [gamma-32P]ATP resulted in an initial increase in the label in PtdIns(4,5)P2, implying a precursor-product relationship between the radiolabelled pools of PtdIns(4)P and PtdIns(4,5)P2. This was confirmed by analysis of the incorporation of 32P into the 4'-phosphate group of PtdIns(4)P and the individual 4'- and 5'-phosphate groups of PtdIns(4,5)P2. The data from these experiments also indicated that PtdIns(4,5)P2 can be produced from a pre-existing pool of PtdIns(4)P, as well as de novo from PtdIns. (7) Taken together our data suggest that isolated rat-liver nuclei have an intranuclear inositol lipid metabolism mechanism utilizing enzymes and substrates equivalent to those found in cytosol and plasma membrane, and that there may be some, but not complete, compartmentalization of the components of the nuclear inositol cycle.

Rapid kinetic measurements of 45Ca2+ mobilization reveal that Ins(2,4,5)P3 is a partial agonist at hepatic InsP3 receptors

Ins(2,4,5)P3, a metabolically stable analogue of Ins(1,4,5)P3, is widely used in analyses of Ca2+ signalling pathways, but its utility depends upon it faithfully mimicking the effects of the natural messenger, Ins(1,4,5)P3, at InsP3 receptors. To compare the kinetics of InsP3-evoked 45Ca2+ mobilization, Ins(1,4,5)P3- and Ins(2,4,5)P3-stimulated 45Ca2+ release from the intracellular stores of permeabilized rat hepatocytes was measured using rapid superfusion. Both Ins(1,4,5)P3 and Ins(2,4,5)P3 caused concentration-dependent increases in the rate of 45Ca2+ efflux, which accelerated towards a peak and then abruptly switched to a bi-exponentially decaying release rate. However, the peak rate of 45Ca2+ mobilization evoked by maximal concentrations of Ins(2,4,5)P3 was only 65+/-3% (n = 3) of that evoked by Ins(1,4,5)P3. Furthermore, Ins(2,4,5)P3 inhibited the peak rate of 45Ca2+ efflux evoked by Ins(1,4,5)P3. These results indicate that Ins(2,4,5)P3 is a partial agonist at hepatic Ins(1,4,5)P3 receptors. Additionally, responses to Ins(2,4,5)P3 were less positively cooperative [Hill coefficient (h) = 1.9+/-0.3] than were those to Ins(1,4,5)P3 (h = 3.0+/-0.2) and the kinetics of termination of 45Ca2+ mobilization were slower. The lesser efficacy of Ins(2,4,5)P3 may account for the lower cooperativity in the responses it evokes, the slower inactivation of InsP3 receptors and the characteristic patterns of Ca2+ spiking it evokes in intact cells.

Aggregation-dependent, integrin-mediated increases in cytoskeletally associated PtdInsP2 (4,5) levels in human platelets are controlled by translocation of PtdIns 4-P 5-kinase C to the cytoskeleton

Thrombin-stimulated aggregation of human platelets promotes an increase in the phosphatidylinositol 4-phosphate (PtdIns 4-P) 5-kinase (PIPkin) activity in the cytoskeleton. This phenomenon is associated with translocation of PIPkin isoform C to the cytoskeleton and with an increase in the amount of phosphatidylinositol bisphosphate (PtdInsP2) bound to the cytoskeletal pellet. All three of these effects are prevented if the platelets are not stirred or if RGD-containing peptides are present, demonstrating that they require integrin activation. All three are also abolished by pretreatment with okadaic acid, which also prevents the aggregation-dependent translocation of pp60(c-src) to the cytoskeleton. The results point to the existence of a cytoskeletally associated PtdInsP2 pool under the control of integrin-mediated signals that act via PIPkin C and suggest that a common, okadaic acid-sensitive mechanism may underlie the aggregation-dependent translocation of certain signalling molecules to the platelet cytoskeleton.

Identification of diacylglycerol pyrophosphate as a novel metabolic product of phosphatidic acid during G-protein activation in plants

We provide evidence that phosphatidic acid (PtdOH) formed during signaling in plants is metabolized by a novel pathway. In much of this study, 32Pi-labeled Chlamydomonas cells were used, and signaling was activated by adding the G-protein activator mastoparan. Within seconds of activation, large amounts of [32P]PtdOH were formed, with peak production at about 4 min, when the level was 5-25-fold higher than the control. As the level of [32P]PtdOH subsequently decreased, an unknown phospholipid (PLX) increased in radiolabeling; before activation it was barely detectable. The chromatographic properties of PLX resembled those of lyso-PtdOH and CMP.PtdOH but on close inspection were found to be different. PLX was shown to be diacylglycerol pyrophosphate (DGPP), the product of a newly discovered enzyme, phosphatidate kinase, whose in vitro activity was described recently (Wissing, J. B., and Behrbohm, H. (1993) Plant Physiol. 102, 1243-1249). The identity of DGPP was established by co-chromatrography with a standard and by degradation analysis as follows: [32P]DGPP was deacylated, and the product (glycerolpyrophosphate, GroPP) was hydrolyzed by mild acid treatment or pyrophosphatase to produce GroP and Pi as the only radioactive products. Since DGPP is the pyrophosphate derivative of PtdOH and is formed as the concentration of PtdOH decreases, we assumed that PtdOH was converted in vivo to DGPP. This was confirmed by showing that during a short labeling protocol while the specific radioactivity of DGPP was increasing, the specific radioactivity of the 32Pi derived from DGPP as above was higher than that of [32P]GroP. DGPP was also formed in suspension cultures of tomato and potato cells, and its synthesis was activated by mastoparan. Moreover, it was also found in intact tissues of a number of higher plants, for example, carnation flower petals, vetch roots, leaves of fig-leaved goosefoot, and common persicaria and microspores of rape seed. Our results suggest that DGPP is a common but minor plant lipid that increases in concentration when signaling is activated. Possible functions of DGPP in phospholpase C and D signaling cascades are discussed.

Synergistic effects of inositol 1,3,4,5-tetrakisphosphate on inositol 2,4,5-triphosphate-stimulated Ca2+ release do not involve direct interaction of inositol 1,3,4,5-tetrakisphosphate with inositol triphosphate-binding sites

We have previously found that for permeabilized L1210 cells, low micromolar concentrations of Ins(1,3,4,5)P4 added prior to Ins(2,4,5)P3 enhance the effects of suboptimal concentrations of Ins(2,4,5)P3 in causing Ca2+ release from InsP3-sensitive Ca2+ stores [Cullen, Irvine and Dawson (1990) Biochem J. 271, 549-553]. If this was due either to some conversion of added Ins(1,3,4,5)P4 into Ins(1,4,5)P3 by the 3-phosphatase, or to Ins(1,3,4,5)P4 acting as a weak (or partial) agonist on the InsP3 receptor it would be expected that,in the presence of thimerosal to sensitize the InsP3 receptor, the dose-response curve to Ins(1,3,4,5)P4 would be left-shifted by the same extent as that of Ins(1,4,5)P3. This was found not to be the case; the dose-response curve to Ins(1,3,4,5)P4 was not shifted at all by thimerosal. Furthermore, L-Ins(1,3,4,5)P4, which can displace radiolabelled D-Ins(1,3,4,5)P4 but not D-Ins(1,4,5)P3 from their respective high-affinity binding sites, mimicked the effects of D-Ins(1,3,4,5)P4 in enhancing the slow phase of Ins(2,4,5)P3-stimulated Ca2+ release. Ins(1,3,4,5)P4 caused an increase in magnitude of the slow phase of InsP3-stimulated Ca2+ release leaving the magnitude of the fast phase unaltered, in contrast to increasing Ins(2,4,5)P3 concentrations which increased the size of both phases. In addition, Ins(1,3,4,5)P4 decreased the rate constant for the slow phase of Ca2+ release. These findings point strongly to the conclusion that InsP4 is not working directly via the InsP3 receptor but indirectly via an InsP4 receptor.

Changes in the components of a nuclear inositide cycle during differentiation in murine erythroleukaemia cells

Differentiation of murine erythroleukaemia cells with the chemical agent DMSO leads to a cessation of proliferation and the production of a number of erythrocyte markers such as haemoglobin. We have previously demonstrated that activation of proliferation leads to an increase in the production of nuclear diacylglycerol (DAG). Here we demonstrate that differentiation leads to a decrease in the levels of nuclear DAG and the activity of the nuclear-associated phosphoinositidase C (PIC). The change in activity appears to be due to a decrease in the mass levels of the beta 1 isoform, as demonstrated by the use of isoform-specific antibodies. Moreover, the changes correlate with the cessation of proliferation and an increase in the number of cells in G1 phase of the cell cycle, rather than with the number of cells which have differentiated. Indeed, although treatment of the cells with phorbol 12-myristate 13-acetate (PMA) inhibits the differentiation programme as assessed by haemoglobin staining, it does not inhibit the number of cells blocking in G1 of the cell cycle or the changes in nuclear DAG or PIC activity. The possible involvement of this nuclear inositide cycle during progression through the cell cycle is discussed.

Identification of a specific Ins(1,3,4,5)P4-binding protein as a member of the GAP1 family

Inositol 1,3,4,5-tetrakisphosphate (Ins(1,3,4,5)P4) is produced rapidly from inositol 1,4,5-trisphosphate (Ins(1,4,5)P3) in stimulated cells. Despite extensive experimentation, no clearly defined cellular function has yet been described for this inositol phosphate. Binding sites specific for Ins(1,3,4,5)P4 have been identified in several tissues, and we have purified one such protein to homogeneity. Its high affinity for Ins(1,3,4,5)P4, and its exquisite specificity for this isomeric configuration, suggest it may be an Ins(1,3,4,5)P4 receptor. Here we report the cloning and characterization of this protein as a GTPase-activating protein, specifically a member of the GAP1 family. In vitro it shows GAP activity against both Rap and Ras, but only the Ras GAP activity is inhibited by phospholipids and is specifically stimulated by Ins(1,3,4,5)P4.

The cloning and sequence of the C isoform of PtdIns4P 5-kinase

In this study we describe the purification and sequencing of the C isoform of platelet PtdIns4P 5-kinase. Subsequently a cDNA was isolated from a human circulating-leucocyte library, which when sequenced was shown to contain all of the peptides identified in the purified protein. In addition, expression of this cDNA in bacteria led to the production of a protein which was recognized by specific monoclonal antibodies raised to the bovine brain enzyme [Brooksbank, Hutchings, Butcher, Irvine and Divecha (1993) Biochem. J. 291, 77-82] and also led to the appearance of PtdIns4P 5-kinase activity in the bacterial lysates. Interestingly, the cDNA showed no similarity to any of the previously cloned inositide kinases. A search of the DNA databases showed that two proteins from Saccharomyces cerevisiae shared close similarity to this enzyme, one of which, the mss4 gene product, has been implicated in the yeast inositol lipid pathway. These data suggest that the PtdIns4P 5-kinases are a new family of inositide kinases unrelated to the previously cloned phosphoinositide 3/4-kinases.

Specificity of the purified inositol (1,3,4,5) tetrakisphosphate-binding protein from porcine platelets

The specificity of the inositol 1,3,4,5-tetrakisphosphate binding protein purified from porcine platelets [Cullen et al. (1995) Biochem. J. 305, 139-143] was examined using all the isomers of myo-inositol tetrakisphosphate. From the relative potencies of these compounds it appears that phosphorylation of the 1, 3 and 5 positions is essential for high affinity binding, that there is some tolerance of phosphorylation of the 6-hydroxyl, but none of a phosphate in the 2-position, and that phosphorylation of the 4-hydroxyl has very little influence. The binding of Ins(1,3,4,5)P4 was not appreciably altered by physiological Mg2+ concentrations, and the pH dependence of binding under physiological conditions showed a decline from pH 5.5 to pH 9.0.

Purification and characterization of an Ins(1,3,4,5)P4 binding protein from pig platelets: possible identification of a novel non-neuronal Ins(1,3,4,5)P4 receptor

A novel Ins(1,3,4,5)P4-binding protein has been purified to apparent homogeneity from solubilized membranes derived from pig platelets. It has a high affinity for Ins(1,3,4,5)P4 (Kd 6.3 +/- 0.4 nM), a Bmax of 2.5-6.0 nmol/mg of protein, and a high specificity for Ins(1,3,4,5)P4 [Kd values for Ins(1,3,4,5,6)P5, InsP6, GroPtdIns(3,4,5)P3, Ins(1,4,5)P3, Ins(3,4,5,6)P4 and L-Ins(1,3,4,5)P4 of 85.0 +/- 4.1 nM, 800.0 +/- 20.2 nM, 65.6 +/- 2.6 nM, > 10 microM, 793.3 +/- 55.6 nM and 81.0 +/- 5.9 nM respectively]. The protein has an apparent molecular mass of 104 kDa, suggesting that this peripheral tissue protein may be different from Ins(1,3,4,5)P4 binding proteins previously isolated from neuronal tissues.

The nuclear phosphoinositide cycle--does it play a role in nuclear Ca2+ homoeostasis?

The probable answer to this question is no. Much of the current evidence summarised elsewhere in this issue points to nuclear Ca2+ changes changing in response to cytosolic Ca2+, with little evidence for an independently controlled nuclear Ca2+ homeostasis. There are InsP3 receptors in the nuclear membrane, and it is possible that during nuclear membrane assembly the InsP3 acting on these (Sullivan and Wilson, this issue) is formed by an inositide cycle located on the assembling nuclear skeleton. But our current experimental data suggest that when the nucleus is intact, InsP3 generated by this cycle would have to exit through the nuclear pores to act on any known InsP3 receptors. Thus the nuclear inositide cycle appears more likely to serve to generate diacylglycerol to activate protein kinase C, and/or to generate inositol phosphates such as InsP2, which may have distinct intranuclear functions.

Specific binding sites for inositol 1,3,4,5-tetrakisphosphate are located predominantly in the plasma membranes of human platelets

In the present study we describe the characterization and localization of Ins(1,3,4,5)P4-binding sites in human platelet membranes. Specific binding sites for Ins(1,3,4,5)P4 have been identified on mixed, plasma and intracellular membranes from neuraminidase-treated platelets using highly purified carrier-free [32P]Ins(1,3,4,5)P4. The displacement of Ins(1,3,4,5)P4 from these sites by Ins(1,4,5)P3 and InsP6 occurs at greater than two orders of magnitude higher concentrations and with Ins(1,3,4,5,6)P5 at about 40-fold higher concentrations than with Ins(1,3,4,5)P4. The membranes were further separated by free-flow electrophoresis into plasma and intracellular membranes. The Ins(1,3,4,5)P4-binding sites separated with plasma membranes, and showed similar affinities and specificities as mixed membranes, whereas Ins(1,4,5)P3-binding sites were predominantly in the intracellular membranes. These results suggest a predominantly plasma membrane location for putative Ins(1,3,4,5)P4 receptors in human platelets.

Rapid turnover of phosphatidylinositol 3-phosphate in the green alga Chlamydomonas eugametos: signs of a phosphatidylinositide 3-kinase signalling pathway in lower plants?

When Chlamydomonas eugametos gametes were incubated in carrier-free [32P]P1, the label was rapidly incorporated into PtdInsP and PtdInsP2 and, after reaching a maximum within minutes, was chased out by recirculating unlabelled P1 in the cell. This pulse-chase labelling pattern reflects their rapid turnover. In contrast, 32P incorporation into the structural lipids was slow and continued for hours. Of the radioactivity in the PtdInsP spot, 15% was in PtdIns3P and the rest in PtdIns4P, and of that in the PtdInsP2 spot, 1% was in PtdIns(3,4)P2 and the rest in PtdIns(4,5)P2, confirming the findings by Irvine, Letcher, Stephens and Musgrave [(1992) Biochem. J. 281, 269-266]. When cells were labelled with carrier-free [32P]P1, both PtdInsP isomers incorporated label in a pulse-chase-type pattern, demonstrating for the first time in a plant or animal system that D-3 poly-phosphoinositides turn over rapidly in non-stimulated cells, with kinetics similar to those shown by the D-4 isomers. In animal systems such lipids are already established as signalling molecules, and the data suggest that a similar role must be sought for them in lower plants such as Chlamydomonas.

Inhibition of iron-catalysed hydroxyl radical formation by inositol polyphosphates: a possible physiological function for myo-inositol hexakisphosphate

1. The ability of myo-inositol polyphosphates to inhibit iron-catalysed hydroxyl radical formation was studied in a hypoxanthine/xanthine oxidase system [Graf, Empson and Eaton (1987) J. Biol. Chem. 262, 11647-11650]. Fe3+ present in the assay reagents supported some radical formation, and a standard assay, with 5 microM Fe3+ added, was used to investigate the specificity of compounds which could inhibit radical generation. 2. InsP6 (phytic acid) was able to inhibit radical formation in this assay completely. In this respect it was similar to the effects of the high affinity Fe3+ chelator Desferral, and dissimilar to the effects of EDTA which, even at high concentrations, still allowed detectable radical formation to take place. 3. The six isomers of InsP5 were purified from an alkaline hydrolysate of InsP6 (four of them as two enantiomeric mixtures), and they were compared with InsP6 in this assay. Ins(1,2,3,4,6)P5 and D/L-Ins(1,2,3,4,5)P5 were similar to InsP6 in that they caused a complete inhibition of iron-catalysed radical formation at > 30 microM. Ins(1,3,4,5,6)P5 and D/L-Ins(1,2,4,5,6)P5, however, were markedly less potent than InsP6, and did not inhibit radical formation completely; even when Ins(1,3,4,5,6)P5 was added up to 600 microM, significant radical formation was still detected. Thus InsP5s lacking 2 or 1/3 phosphates are in this respect qualitatively different from InsP6 and the other InsP5s. 4. scyllo-Inositol hexakisphosphate was also tested, and although it caused a greater inhibition than Ins(1,3,4,5,6)P5, it too still allowed detectable free radical formation even at 600 microM. 5. We conclude that the 1,2,3 (equatorial-axial-equatorial) phosphate grouping in InsP6 has a conformation that uniquely provides a specific interaction with iron to inhibit totally its ability to catalyse hydroxyl radical formation; we suggest that a physiological function of InsP6 might be to act as a 'safe' binding site for iron during its transport through the cytosol or cellular organelles.

Evidence for receptor-specific regulation of the metabolism of the second messenger, inositol trisphosphate, in human endothelial cells

Primary cultures (1-2 weeks-old, no passage) of endothelial cells from the human umbilical vein were pre-labelled with [3H]inositol, stimulated for varying times (15 seconds to 5 minutes) in the presence of LiCl with either thrombin or histamine, and the extracted radiolabelled isomers of inositol phosphates were analysed by anion-exchange chromatography. The doses of each receptor-agonist were chosen to stimulate a large accumulation of inositol phosphates at approximately the same rate. At all times of stimulation the relative accumulation of inositol trisphosphate isomers and inositol tetrakisphosphate was different upon stimulation with these agonists, particularly at earlier times. These results argue for the receptor-specific regulation of the inositol trisphosphate 3-kinase. The significance of this observation is discussed with respect not only to the current ideas on the control of calcium homoeostasis by inositol phosphates, but also in consideration of their role as second messengers controlling different end-points of vascular function.

Monoclonal antibodies to phosphatidylinositol 4-phosphate 5-kinase: distribution and intracellular localization of the C isoform

We have raised a panel of monoclonal antibodies to PtdIns4P 5-kinase C purified from bovine brain [Divecha, Brooksbank and Irvine (1992) Biochem. J. 288, 637-642]. This panel includes antibodies which specifically recognize PtdIns4P 5-kinase C both in a native catalytically active condition, and/or when presented on Western blots. Some of the former antibodies will also inhibit PtdIns4P 5-kinase C activity. We have used the blotting antibodies to study the bovine tissue distribution of PtdIns4P 5-kinase C and its distribution in mammalian species. We have also studied its localization in Jurkat cells and found it to be predominantly bound to membranes, with only a minority localized to the cytoskeleton. Neither PtdIns4P 5-kinase activity nor PtdIns4P 5-kinase C, as detected by Western blotting, were increased in the cytoskeleton after stimulation of Jurkat cells with OKT3. These antibodies should prove to be extremely useful tools with which to study the regulation of PtdIns4P 5-kinase C.

Nuclear diacylglycerol is increased during cell proliferation in vivo

Highly purified nuclei were prepared from livers and kidneys of rats undergoing compensatory hepatic or renal growth, the former being predominantly by cellular proliferation, and the latter mostly by cellular enlargement. In liver, an increase in nuclear diacylglycerol (DAG) concentration occurred between 16 and 30 h, peaking at around 20 h. At the peak of nuclear DAG production a specific translocation of protein kinase C to the nucleus could be detected; no such changes occurred in kidney. There was no detectable change in whole-cell DAG levels in liver, and the increase in DAG was only measurable in nuclei freed of their nuclear membrane. Overall, these results suggest that there is a stimulation of intranuclear DAG production, possibly through the activation of an inositide cycle [Divecha, Banfić and Irvine (1991) EMBO J. 10, 3207-3214] during cell proliferation in vivo.

Phosphoinositide signalling enzymes in rat liver nuclei: phosphoinositidase C isoform beta 1 is specifically, but not predominantly, located in the nucleus

The presence of phosphoinositide-mobilizing enzymes has been investigated in purified rat liver nuclei by radiolabelling and by probing with antibodies. A Ca(2+)-activated phosphoinositidase C (PIC) is present and was shown immunologically to be the beta 1 isoform. No gamma- or delta-PIC was found. However, only 5% of the total beta 1-PIC isoform is nuclear, with the majority being cytosolic. G alpha q and G alpha 11, the suggested physiological activators of beta 1-PIC, were not present. A PtdIns4P 5-kinase is also present, which immunologically is shown to be the C isoform. All of these nuclear inositide enzymes still remained after the removal of the nuclear envelope with Triton X-100, consistent with the concept of an intranuclear inositide cycle [Divecha, Banfic and Irvine (1991) EMBO. J. 10, 3207-3214].

Purification and characterization of phosphatidylinositol 4-phosphate 5-kinases

We detail the purification and characterization of three distinct isoforms of PtdIns4P 5-kinase present in bovine brain. One of these, PtdIns4P 5-kinase C, was purified to apparent homogeneity, and SDS/PAGE analysis demonstrated a single polypeptide and molecular mass 53 KDa. These three isoforms were shown to differ in their kinetic properties, and immunological characterization with an antibody raised to PtdIns4P 5-kinase C demonstrated that this isoform was unrelated to the other two. Furthermore, PtdIns4P 5-kinase C was shown to be the bovine brain homologue of the Type II PtdIns4P 5-kinase previously purified from human erythrocytes [Bazenet, Ruano, Brockman & Anderson (1990) J. Biol. Chem. 265, 18012-18022].