Phospholipase C (PLC) is a receptor-regulated enzyme that hydrolyses phosphatidylinositol 4,5-bisphosphate (PI(4,5)P2) at the plasma membrane (PM) triggering three biochemical consequences, the generation of soluble inositol 1,4,5-trisphosphate (IP3), membrane-associated diacylglycerol (DG) and the consumption of PM PI(4,5)P2. Each of these three signals triggers multiple molecular processes impacting key cellular properties. The activation of PLC also triggers a sequence of biochemical reactions, collectively referred to as the PI(4,5)P2 cycle that culminates in the resynthesis of this lipid. The biochemical intermediates of this cycle and the enzymes that mediate these reactions are topologically distributed across two membrane compartments, the PM and the endoplasmic reticulum (ER). At the PM, the DG formed during PLC activation is rapidly converted into phosphatidic acid (PA) that needs to be transported to the ER where the machinery for its conversion into PI is localised. Conversely, PI from the ER needs to be rapidly transferred to the PM where it can be phosphorylated by lipid kinases to regenerate PI(4,5)P2. Thus, two lipid transport steps between membrane compartments through the cytosol are required for the replenishment of PI(4,5)P2 at the PM. Here, we review the topological constraints in the PI(4,5)P2 cycle and current understanding how these constraints are overcome during PLC signalling. In particular, we discuss the role of lipid transfer proteins in this process. Recent findings on the biochemical properties of a membrane-associated lipid transfer protein of the PITP family, PITPNM proteins (alternative name RdgBα/Nir proteins) that localise to membrane contact sites are discussed. Studies in both Drosophila and mammalian cells converge to provide a resolution to the conundrum of reciprocal transfer of PA and PI during PLC signalling.

Introduction

Phosphatidylinositol 4,5-bisphosphate (PI(4,5)P2) is the most versatile inositol-containing phospholipid enriched at the inner leaflet of the plasma membrane (PM). PI(4,5)P2 levels govern many cellular functions including endocytosis, ion channels and transporter activity and cytoskeleton–PM interactions [1]. Additionally, PI(4,5)P2 is a substrate for two signalling pathways, phospholipase C (PLC) and phosphoinositide 3-kinase (PI3K). Both are ubiquitous cell signalling pathways that can be activated by agonists acting on appropriate cell-surface receptors. Both G-protein-coupled receptors (GPCRs) and receptor tyrosine kinases can regulate PLC activity. In addition, both pathways can also be activated by small GTPases and some PLCs by a rise in cytosolic Ca2+. Hydrolysis of PI(4,5)P2 by PLC generates two second messengers, diacylglycerol (DG) and inositol 1,4,5-trisphosphate (IP3). A secondary consequence of PLC activation is a transient reduction in PI(4,5)P2 levels and this can have a profound impact on cellular functions that depend on the intact lipid. Thus, mechanisms have to be in place to replenish PI(4,5)P2 levels during cell signalling. This review focuses on these potential mechanisms.

PLC-mediated signalling has a long history and the first papers were published in the 1950s when it was noted that activation of cells with appropriate agonists caused an increase in the labelling of phosphatidylinositol (PI) and phosphatidic acid (PA) when cells had been pre-incubated with 32Pi to label the ATP pool. This stimulated increase in PA and PI labelling was a ubiquitous phenomenon observed in a wide range of cell types (see Figure 1). The time course of PA and PI labelling indicated a product–precursor relationship whereby PA was converted into PI and, by 1964, the metabolic interconversion between PA and PI was established and became known as the ‘PI cycle’ (Figure 1A) [2]. In 1975, it was proposed that the ‘PI cycle’ stimulated by agonist receptor interactions was responsible for causing the increase in cytosol Ca2+ [3]. Although much of the data fitted into this paradigm, there were clearly situations where Ca2+ was able to activate the ‘PI cycle’ [4]. The issue about Ca2+ and PLC activation was ultimately resolved when it became clear that there are multiple PLC isoenzymes with different characteristics. The ‘PI cycle’ was later modified to include PI(4,5)P2 as the cellular substrate for the PLC rather than PI and is referred to here as the ‘PI(4,5)P2 cycle’. PI became an intermediate in the pathway and was converted into PI(4,5)P2 by two lipid kinases, phosphoinositide 4-kinase (PI4K) and phosphatidylinositol 4-phosphate 5-kinase (PIP5K; Figure 1B). The discovery that IP3 was capable of mobilising Ca2+ from intracellular stores secured the link between the PI(4,5)P2 cycle and changes in cytosol Ca2+ [5]. The subsequent discovery of the IP3 receptor at the endoplasmic reticulum (ER) finally established the ‘PI(4,5)P2 cycle’ as an important signal transduction system. In addition to generating IP3, the other product of PLC activation is DG. This lipid was also identified as a second messenger; it activates several proteins including protein kinase C (PKC) by interactions with C1 domains. Thus, PLC activation became established as a signal transduction system at par with the adenylate cyclase–cAMP second messenger system. It has subsequently become clear that the receptor-activated PI(4,5)P2 cycle is a conserved feature of most metazoan cells.

The ‘PI(4,5)P2 cycle’: metabolic interconversions involved in resynthesis of phosphoinositides after hydrolysis by phospholipase C.

Figure 1.
The ‘PI(4,5)P2 cycle’: metabolic interconversions involved in resynthesis of phosphoinositides after hydrolysis by phospholipase C.

(A) The ‘PI cycle’ as proposed by Mabel Hokin and Lowell Hokin in 1964 [2]. (B) Topological organisation of the ‘PI(4,5)P2 cycle’: the ‘PI(4,5)P2 cycle’ operates between two membrane compartments, the ER and the PM requiring the transfer of PA and PI in opposite directions. The ‘PI(4,5)P2 cycle’ was routinely observed by monitoring the incorporation of 32Pi into PA and PI (phosphate labelled in blue).

Figure 1.
The ‘PI(4,5)P2 cycle’: metabolic interconversions involved in resynthesis of phosphoinositides after hydrolysis by phospholipase C.

(A) The ‘PI cycle’ as proposed by Mabel Hokin and Lowell Hokin in 1964 [2]. (B) Topological organisation of the ‘PI(4,5)P2 cycle’: the ‘PI(4,5)P2 cycle’ operates between two membrane compartments, the ER and the PM requiring the transfer of PA and PI in opposite directions. The ‘PI(4,5)P2 cycle’ was routinely observed by monitoring the incorporation of 32Pi into PA and PI (phosphate labelled in blue).

An outstanding question that was highlighted in a review in 1973 and again in 1975 was the issue of lipid transfer between the PM and the ER. PLC-mediated PI(4,5)P2 hydrolysis was confined to the PM, but the machinery to resynthesise PI was present at the ER (Figure 1B) [3,6]. In 1974, a cytosolic activity had been identified that was able to transport PI between membrane compartments and the concept arose that lipids could be moved by ‘lipid transfer proteins’ (LTPs) [7]. The identification of soluble PITPα as an exchange protein that could bind and transfer PI and phosphatidylcholine (PC) and subsequently shown to support IP3 production in permeabilised cells established the principle that PITPs participated in lipid transfer during PLC signalling [8,9]. However, the outstanding issue for PA transport from the PM (where it is produced by the sequential action of PLC and DG kinase, DGK) to the ER (where it is channelled into PI resynthesis) remained unclear until recently. In 2012, we identified that some members of the PITP family, namely the RdgB family of proteins (see Table 1) proteins, were able to bind and transfer PA in addition to PI [10]. This was the first example of a PA transfer protein. Previous experiments were conducted in permeabilised cells where the PI resynthesis cycle was reconstituted, indicated that PA transfer did not require cytosolic proteins or vesicular transport and it was suggested that PA transfer occurs at membrane contact sites (MCS) [11]. Here, we highlight the journey that has led to the identification of a class of PITPs (PITPNM/RdgB proteins) that can exchange PA for PI between the ER and the PM to complete the ‘PI(4,5)P2 cycle’. This journey has taken 40 years.

To a large extent, studies of the PI(4,5)P2 cycle have been performed in cell culture models. Although studies conducted in cell lines where the individual components can be overexpressed or knocked out have provided fundamental information about PLC signalling, there are limitations. Studies in cultured cell lines have relied on overexpression of multiple proteins simultaneously including receptors to enhance PLC signalling, PITPNM proteins and probes targeted to specific membranes to monitor changes in lipids. In these studies, the PLC signalling pathway has been hyperstimulated and, moreover, do not take into consideration the expression levels of endogenous PITP proteins or receptors. Nonetheless, these studies provide a valuable insight into the PLC signalling pathway. The human body contains ∼200 different cell types that are specialised to perform specific functions. More importantly, the protein composition and protein abundance varies depending on the cell type. Thus, it is critical to examine PLC signalling in the context of a specific cell function. The study by Kruse et al. [12] highlights the importance of studying PLC signalling under physiological conditions.

The Drosophila phototransduction cascade remains one of the best characterised systems where the application of genetics coupled with measurements of the physiological output, i.e. light response provides context to PLC signalling [1316]. In these cells, photon absorption by rhodopsin is transduced into opening of two classes of Ca2+-permeable ion channels, TRP and TRPL [13]. This transduction pathway requires Gq-mediated PLC activation leading to PI(4,5)P2 hydrolysis. Several generic elements of the ‘PI(4,5)P2 cycle’ including the GPCR (rhodopsin encoded by ninaE), heterotrimeric G-protein (Gq), phospholipase Cβ4 (PLCβ4 encoded by norpA), DGK (encoded by rdgA), cytidine diphosphate (CDP)-DG synthase (CDS; encoded by cds), PI synthase (encoded by pis) and PITPNM (encoded by rdgB) and PIP5K (reviewed in [13]) have been identified and characterised in fly photoreceptors. The ability to combine genetics with measurements of the physiological outcome has made Drosophila photoreceptors a valuable tool to dissect the questions related to the PI(4,5)P2 cycle. Thus, we will draw upon this system as an exemplar in our review and, in many instances, provide a comparative analysis between studies in mammalian cell lines and Drosophila photoreceptors.

Biochemistry of the PI cycle

PI constitutes ∼5–7% of the total cellular phospholipids in mammalian cells. De novo synthesis of PI occurs at the ER from PA (see Figure 1B). PA is synthesised by the acylation of glycerol 3-phosphate to lysophosphatidic acid and a second acylation reaction to generate PA. The PA is converted into cytidine diphosphate DG (CDP-DG) by the enzyme CDS in the presence of CTP. The final step is the condensation of CDP-DG and myo-inositol into PI by phosphatidylinositol synthase (PIS). An intriguing feature of PI in mammalian cells is its distinctive fatty acid composition; it is enriched in stearic acid (C18:0) at the sn-1 position and arachidonic acid (C20:4) at the sn-2 position of its glycerol backbone [17]. In mammalian cells, PI acquires its characteristic fatty acid composition by fatty acid remodelling after de novo synthesis of PI through sequential deacylation and reacylation reactions [18,19]. The mammalian enzyme responsible for the enrichment of arachidonic acid into PI is lyso-PI acyltransferase (also known as MBOAT7) [17,20,21]. The importance of this fatty acid configuration is underscored by the observation that disruption of this gene in mice leads to severe defects in brain development including a substantial reduction in the levels of PI [17,21].

PI can be sequentially phosphorylated by PI4K to phosphatidylinositol 4-phosphate (PI4P) and finally to PI(4,5)P2 by PIP5K. However, of the total cellular pool of PI, only 10–12% is present in the phosphorylated forms with PI4P and PI(4,5)P2 being the most abundant. The family of PI3K can also phosphorylate PI, PI4P and PI(4,5)P2 to make PI3P, PI(3,4)P2 and PI(3,4,5)P3, respectively, but these species are present at much lower abundance [22]. PI4P and PI(4,5)P2 are both enriched at the PM, although PI4P is also found in other compartments including the Golgi, endosomes and lysosomes [23,24]. The seven phosphorylated derivatives of PI are enriched in specific membrane compartments due to the presence of unique lipid kinases and phosphatases that, respectively, add and remove phosphates from unique positions on the inositol ring. In many internal compartments, these phosphoinositides are interconverted by lipid kinases and phosphatases, but in this case the basic backbone of PI is not destroyed. It is only the PLC signalling system that destroys the lipid requiring the resynthesis of PI.

Biochemical constraints in the PI(4,5)P2 cycle

One of the central assumptions of the ‘PI(4,5)P2 cycle’ is that the PA produced from PLC activity does not ‘mix’ with other cellular pools of PA and is maintained separately for recycling back to PI [3,25,26]. This assumption is partly based on observations that increased labelling of PI with [32P]Pi is not normally accompanied with enhanced labelling with radio-labelled glycerol or fatty acids. Additionally, measurements of the fatty acid composition of newly formed ‘stimulated PA’ reflect the fatty acid composition of PI, i.e. PA formed is enriched in stearate and arachidonate. In principle, there are at least two other major pathways that can generate PA which can then enter the PI(4,5)P2 cycle. One of these is the de novo synthesis from glycerol and fatty acids. In the second pathway, PA can also be produced by the hydrolysis of PC by phospholipase D (PLD). In mammalian cells, there are two PLDs, PLD1 and PLD2, that are also activated when cell-surface receptors bind their agonist [27]. It is not uncommon for both PLC and PLD to be activated by the same receptors. Since the fatty acid composition of PA derived from PLD activity will reflect the composition of PC, which mainly contains mono-saturated fatty acids, it is possible to distinguish the source of the PA [28], although previous studies with PLD2 knockout mice have failed to verify this model [29]. There is some evidence to suggest that the backbone of PI(4,5)P2 is recycled back into PI and the PA from the PLD pathway is kept separate [11,26]. This question needs to be revisited in intact cells where selective inhibition of the PLD pathway could be used to address this question.

Although the PI(4,5)P2 cycle is ubiquitous in distribution across cell types, it acquires complexity because the mammalian genome contains multiple genes that produce variants of a specific enzyme activity that can catalyse a specific step of the cycle (Table 2). For example, there are 13 PLCs and ten DGKs (excluding splice variants) in mammalian cells. As a further complexity, other lipid metabolising enzymes that are not core steps of the PI(4,5)P2 cycle can also generate lipid intermediates of the cycle. For example, PA can be produced by PLD (PLD1 and PLD2 and NAPE-PLD) [30,31] and DG can exit the cycle by being metabolised to 2-arachidonylglycerol, an activator of the endocannabinoid receptor, by DG lipase α [32]. This is particularly important in the brain as postsynaptic synthesis and release of endocannabinoid lipids activate presynaptic cannabinoid 1 receptor to mediate short- and long-term depression in many brain regions. In platelets, DG can exit the cycle by being metabolised by DG lipase to form arachidonic acid for thromboxane synthesis [33]. Thus, the assumption that the intermediates in the ‘PI(4,5)P2 cycle’ might be conserved [3] may not be valid. In platelets, mass measurements of phosphoinositides, PA and DG following stimulation with thrombin revealed that the loss of phosphoinositide was not completely recovered as either DG or PA [33]. What is clear is that hormone-stimulated PI resynthesis always occurs when PI(4,5)P2 hydrolysis by PLC takes place to replace the pool of degraded PI(4,5)P2 [3,34]. Below we describe the key enzymes that participate in the cycle and their subcellular locations.

Phospholipase C

Thirteen PLC enzymes have been identified and grouped into six classes, PLCβ1–4, PLCγ1,2, PLCδ1,3,4, PLCε, PLCζ and PLCη1,2, on the basis of domain structure and regulatory mechanisms (Table 2). The presence of a catalytic X and Y domain, a PH domain, a C2 domain and four EF hands are the common features shared by these PLCs. The presence of additional subtype-specific domains contributes to specific regulatory mechanisms. The enzymes of the PLCβ family can be stimulated by GPCRs and the Gq family of G-proteins is responsible for their activation. PLCε is regulated by Ras, Gβγ and by G12. Receptor tyrosine kinases, on the other hand, activate PLCγ1 and 2, while PLCδ family, PLCζ and PLCη are regulated by a rise in cytosol Ca2+ from the nanomolar to the micromolar range. In vitro, PLCs can hydrolyse PI, PI4P and PI(4,5)P2, but it is thought that, in cells, PI(4,5)P2 is their main substrate. However, there may be exceptions; for example, PLCε is present at the vicinity of the nuclear envelope where it hydrolyses Golgi PI4P [35]. The majority of the PLCs have been deleted in the mouse individually and the animals exhibit a range of phenotypes, including epilepsy (PLCβ1), embryonic lethality (PLCγ1) and B-cell development defects (PLCγ2) [36].

Although the PLC signalling pathway was first characterised as a means to generate the second messengers IP3 and DG, it is now clear that a decrease in PI(4,5)P2 also has significant consequences. The intact lipid PI(4,5)P2 is a key regulator of many ion channels [37,38]. The best example is the regulation of the M-current carried by KCNQ2/3 (Kv7.2/7.3) potassium channels [39,40]. The M-current is activated by muscarinic receptors and depletion of PI(4,5)P2 causes the closure of these channels; IP3 and DG do not play a significant role. In a previous study, detailed analysis of two GPCRs present in the same cell concluded that the number of receptors activated determined how much PI(4,5)P2 depletion occurred [41]. Interestingly, this concept has a longer history. Early studies had indicated that the ‘PI cycle’ tracked receptor occupancy but not changes in cytosol Ca2+ [42]. The recent studies used the M-current as a surrogate for PI(4,5)P2 hydrolysis, and substantial PI(4,5)P2 hydrolysis only occurred when receptors were present at a high density. Thus, only activation of the high-abundance muscarinic receptor led to closure of the KCNQ2/3 potassium channels but not by the activation of the low-abundance P2Y2 receptors for ATP [12,38,41]. Nonetheless, activation of either receptor by saturating concentrations of agonist provoked a full-amplitude Ca2+ response. This study confirms earlier work that only a small amount of PI(4,5)P2 needs to be hydrolysed to obtain a maximal rise in cytosol Ca2+. Secondly, a high receptor occupancy is required to obtain a substantial decrease of PI(4,5)P2. This is an important conclusion as it clarifies how different GPCRs in the same cell can produce different functional outputs. Furthermore, it implies that receptor density with maximal receptor occupancy will determine how much PI(4,5)P2 will get depleted. Under these circumstances, the PI resynthesis pathway must be tuned to cope with the extra demand when PI(4,5)P2 depletion occurs following high rates of receptor-triggered PLC activity. In this setting, another important factor that needs consideration is the speed of PI(4,5)P2 replenishment. Is the rate of PI(4,5)P2 synthesis variable between different cell types? Comparing the dynamics of PI(4,5)P2 resynthesis following PLC signalling in sympathetic neurons and in a non-excitable cell line reveals that synthesis of PI4P is restored much faster in sympathetic neurons than in a non-excitable cell line [12]. In neurons, PI(4,5)P2 has many essential functions, including exocytosis, endocytosis, actin assembly and ion channel regulation, and may be more vulnerable to PI(4,5)P2 depletion. This may require additional mechanisms to cope with the PI resynthesis pathway. The resynthesis of PI(4,5)P2 at the PM requires a supply of PI and it is here that lipid transporters may play an important role.

In the Drosophila genome, there are only three PLCs: two are highly related to the PLCβ family, NORPA and Plcβ21C, and a single PLCγ (Table 2) [4347]. Unlike in mammals, loss of PLCγ does not lead to lethality, but causes small wings [45]. Plcβ21C is implicated in olfaction transduction while NORPA is required for phototransduction. NORPA expression is enriched in photoreceptors and, in these cells, there are high levels of basal PLC activity [48]. Following illumination, PLC activity increases dramatically and, under bright light condition, ∼1 × 106 PLC molecules are active [49]. Thus, Drosophila photoreceptors are an example of cells in which very high rates of PLC activity are achieved. Despite these high levels of PLC activation, PI(4,5)P2 levels show only a 50% reduction under daylight equivalent illumination in vivo [50]. This was, however, not the case when recordings were done in the absence of extracellular Ca2+ or in the trp mutant, which reduced Ca2+ influx into photoreceptors [50,51]. Thus, the end point of PLC activation, namely Ca2+ influx, may offer a mechanism by which PLC activity can be inhibited and tuned to prevent PI(4,5)P2 depletion. Interestingly, a biochemical study of PLC activation in Drosophila heads has reported a bell-shaped dependence of PLC activity on Ca2+ concentration [52].

A further question that arises is the mechanism by which specificity of PI(4,5)P2 signalling arises during PLC activation. PI(4,5)P2 has multiple functions at the PM; therefore, during PLC signalling, it is essential to have mechanisms that prevent the non-PLC-dependent functions of PI(4,5)P2 being affected. There are multiple mechanisms by which the specificity of PI(4,5)P2 functions can be preserved; these have recently been reviewed [53]. In many neuronal and also some non-neuronal cells, many of the phospholipases form macro-molecular complexes with other proteins during cell activation, which means that PLC activation at the PM is likely to be restricted to microdomains. If PI(4,5)P2 hydrolysis at these domains is tightly coupled to the local delivery of PI for replenishment, this would lessen the impact of PI(4,5)P2 depletion on other PI(4,5)P2-dependent systems.

In Drosophila photoreceptors, NORPA is localised at the PM along with the GPCR rhodopsin and Gq. In this location, it is found as part of a macromolecular protein complex assembled on the multi-PDZ domain protein INAD. In addition to NORPA, this complex includes the TRP channel, PKCα and NINAC, an unconventional myosin (reviewed in [16,54]). The 90-amino-acid PDZ domain was originally identified as a conserved element in three structurally unrelated proteins, PSD-95, DLG and ZO-1. The PDZ domain binds to short C-terminal peptide sequences of target proteins. Multiple PDZ domains are often present in a single protein promoting the scaffolding of several target proteins [55]. All PLCβ subtypes have PDZ-binding motifs and the PDZ-containing Na+/H+ exchanger regulatory factor family, and SHANK proteins form signalling complexes bringing together the PLCβ isoenzymes with the different receptors [56]. Despite considerable interest in the idea that the presence of NORPA in the INAD complex of Drosophila photoreceptors provides speed and specificity to PLC signalling [57], many studies have failed to support the idea that the presence of an intact INAD complex is required for normal activation through PLC signalling [58,59]. However, the notion that confining PLC activation to signalling platforms would avoid effects on the cytoskeleton and vesicular transport and allow channelling of substrate to these regions remains attractive. As discussed below, these signalling platforms could reside at MCS allowing for efficient and localised signalling to occur.

DG kinases

The product of PLC activity is DG that is rapidly phosphorylated to PA by DGK. The first DGK was cloned in 1990, and now there are ten DGKs in the mammalian genome and five in the Drosophila genome (Table 2). DGK enzymes are subdivided into five subtypes based on the domain structures [60]. The expression of DGKs is cell-type-specific [60]; for example, in T-cells, the predominant isoforms expressed are DGKα and DGKζ. Since PI(4,5)P2 in mammalian cells is enriched in C18:0/C20:4, the DG formed will also have the same fatty acid composition. DGKε has been shown to specifically utilise stearoyl arachidonyl glycerol [61]. However, this specificity is not absolute as all the other DGKs can utilise C18:0/C20:4 species.

PA made following the activation of PLC activity is found at the PM as demonstrated using subcellular fractionation [62]. However, DGK may not be constitutively localised at the PM. For example, in T-cells, the predominant isoforms expressed, DGKα and DGKζ, are both recruited to the PM following T-cell receptor activation where they contribute to the termination of DG-mediated signalling downstream following PLCγ activation by T-cell receptor–ligand complexes [63,64]. An important consequence of this observation is that PA, the product of DGK activity being a lipid, will be spatially restricted to the PM whereas the next enzyme in the PI(4,5)P2 cycle, CDS that utilises PA as a substrate, is localised to the ER. Thus, a mechanism is required to transport PA from the PM to the ER.

In Drosophila photoreceptors, the DGK that is utilised in the removal of DG after PLC activation is encoded by rdgA [65]. The mammalian equivalent of the RdgA protein is the type IV family that includes DGKζ and DGKι. Loss of rdgA results in enhanced electrical responses to light, a defect in response termination [66,67] and severe retinal degeneration [68,69]. All of these phenotypes are dependent on ongoing PLC activity (reviewed in [13]) and underscore the importance of DGK in regulating light-induced PI(4,5)P2 turnover. The biochemical basis of phenotypes resulting from reduced rdgA function remains to be resolved. Reduction in rdgA function results in lower levels of PA; surprisingly, elevated DG levels have not been demonstrated [70,71]. It has been proposed that the DG may be metabolised via liberation of polyunsaturated fatty acids that may then activate TRP channels (reviewed in [72]). An interesting feature of RdgA is that in photoreceptors, the protein is localised to the sub-microvillar cisternae (SMC) [73], a specialisation of the ER that forms a MCS with the microvillar PM (reviewed in [74]) This localisation of RdgA to an ER sub-compartment resembles, conceptually, the recruitment of DGKα and DGKζ from the ER to the PM of mammalian cells following receptor-mediated PLC activation. In the case of Drosophila photoreceptors, RdgA could act in trans across the small 10 nm gap at this MCS to access DG generated at the microvillar PM by NORPA.

Type II PA phosphatase

Biochemical experiments have identified two PA phosphatase activities in mammalian cells. One of these, Type I PA phosphatase (also called Lipins), a Mg2+-dependent and N-ethylmaleimide-sensitive activity is associated with the cytosol and microsomal fraction and is involved in phospholipid biosynthesis. The other, a Type II PA phosphatase (also called lipid phosphate phosphohydrolase), is Mg2+-independent and N-ethylmaleimide-insensitive and is associated with the PM (reviewed in [75]). This latter class of PA phosphatases has been suggested to be involved in regulating signalling reactions and may in principle regulate the PI(4,5)P2 cycle where PA is generated as an intermediate by DGK activity. However, although a role for Type II PA phosphatase in regulating PA levels in Drosophila is shown, its role in regulating PI(4,5)P2 levels in mammalian cells remains to be established.

A Drosophila orthologue of Type II PA phosphatase has been implicated in the regulation of the PI(4,5)P2 cycle in Drosophila photoreceptors [71,76]. The Drosophila genome encodes seven Type II PA phosphatase-like genes. One of these, lazaro (laza), is enriched in the eye. Loss-of-function mutants of laza shows increased accumulation of PA during illumination that triggers G-protein-coupled hydrolysis of PI(4,5)P2. rdgA (DG kinase) mutants show reduced levels of PA and these are further reduced by overexpression of laza. Importantly, in addition to PA, the levels of PI are also reduced in rdgA mutants and these are further reduced by overexpression of laza. Thus, a Type II PA phosphatase activity regulates the levels of PA and PI, two key intermediates during G-protein-coupled PLC activation in photoreceptors. The retinal degeneration and electrophysiological defect in rdgA can be enhanced by overexpression of laza and suppressed in rdgA;laza double mutants. Taken together, these findings suggest that Type II PA phosphatase regulates PA and PI turnover during PLC activation. The subcellular localisation of endogenous LAZA remains to be established.

Cytidine diphosphate DG synthase

The next enzyme in the PI(4,5)P2 cycle is CDS that catalyses the conversion of PA into CDP-DG, an essential intermediate step in the biosynthesis of PI. In mammals, two homologous genes of CDS (CDS1 and CDS2) have been cloned that are 73% identical and 92% similar [7779]. Genes homologous with mammalian CDS are found in Escherichia coli, Saccharomyces cerevisiae, Drosophila, plants and cyanobacteria [7981]. In S. cerevisiae, the single CDS1 is an essential gene and appears to account for the majority of the synthase activity [82] (but see later). In Drosophila, complete loss of CDS results in lethality. However, an eye-enriched isoform of CDS is expressed and mutant alleles that result in its down-regulation result in defects in phototransduction and retinal degeneration [83]. In Drosophila photoreceptors, CDS is distributed throughout the cell body; it is likely to be localised to the ER as has been reported for mammalian cells [84,85].

The two isoforms of mammalian CDS show different substrate specificities. Mammalian CDS2 appears to be selective for the acyl chains of PA, the most preferred species being 1-stearoyl-2-arachidonoyl-sn-PA [84]. The fatty acid composition of PI and its phosphorylated derivative, PI(4,5)P2, is mainly 1-stearoyl-2-arachidonyl, and during PLC signalling, the PA generated would retain this composition. This is also supported by studies in zebrafish where CDS2 controls the availability of PI(4,5)P2 for VEGF signalling, and when mutated, there is a failure of angiogenesis [86]. CDS1, on the other hand, shows no particular substrate specificity [84]. Expression studies indicate that CDS1 and CDS2 exhibit quite different tissue specificity. In the mouse, CDS2 appears to be ubiquitously expressed while CDS1 has a restricted pattern of expression.

Early biochemical studies in yeast and mammalian cells showed that CDS activity was present in both microsomal (mainly ER) and mitochondrial fractions [8789], hence CDS1 was thought to be present in both ER and mitochondria. However, recent studies in yeast have identified a new enzyme, Tam41, which possesses CDP-DG synthase activity. Tam41 is a peripheral membrane protein localised in the inner mitochondrial membrane where it is involved in cardiolipin synthesis [90]. Tam41 and the yeast CDS1 have different evolutionary origins and are not related by sequence although both possess CDS activity [90]. It is of interest that the substrate, PA, is transported from the outer mitochondrial membrane to the inner mitochondrial membrane-localised Tam41 by the yeast PA transporter, Ups1/Mdm35 complex (TRIAP1/PRELI complex in mammals) [9195]. Interestingly, this complex has structural homology with mammalian PITPα [96,97].

PI synthase

PIS is involved in the final step of PI synthesis, converting CDP-DG and myo-inositol into PI. In addition to de novo biosynthesis, PIS has exchange activity where the inositol of the PI can be replaced for free inositol in the presence of Mn2+ and in the absence of CDP-DG [98]. PIS activity localises mainly to the ER and molecular identification was first achieved in yeast and ultimately made it possible to clone the mammalian enzyme [79,99]. There is a single gene for PIS in yeast and in mammalian cells, which shares sequence identity throughout their primary sequence. The human PIS gene is predicted to encode an integral protein of 213 amino acids (23 kDa) and studies in yeast indicate that the active site faces the cytosol [100]. In Drosophila, there is a single PIS gene (dPIS); loss-of-function mutants of dPIS result in both organismal and cell lethality underscoring the central role of PI signalling in eukaryotic cell biology. Hypomorphic alleles of dPIS show defects in phototransduction, a process that requires PI(4,5)P2 regeneration after its hydrolysis by PLC [101].

The ER is a heterogeneous compartment and studies conducted in COS-7 cells indicate that overexpressed PIS can co-localise with Rab10 to an ER-derived sub-compartment [85,102]. This sub-compartment makes contacts with other membranes, implying that PI can be delivered to other membranes presumably by a lipid transfer mechanism. These dynamic structures required the GTPase Sar1, indicating that they were generated from the ER. However, CDS1 or CDS2 did not co-localise in these dynamic structures [85]. The dependence on Sar1 implies that the vesicles are initially made as COPII-coated vesicles. One possibility is that PIS, normally an ER-resident enzyme, when overexpressed at high levels, results in being captured into COPII vesicles. The fundamental issue would be access to CDP-DG for the synthesis into PI in these mobile structures. A further study on these dynamic compartments is required, preferably in other cell types.

Phosphoinositide 4-kinase

Of the four PI4Ks encoded in the mammalian genome [PI4KIIIα (Stt4 in yeast), PI4KIIIβ (Pik1 in yeast) and PI4KIIα and PI4KIIβ (Lsb6 in yeast)], PI4KIIIα has recently been shown to be the isoform localised at the PM [103105]. The role of PI4KIIIα in regulating PI4P synthesis during PLC signalling was first studied using pharmacological inhibitors and siRNA [103]. In this setting, the role of PI4KIIIα in regulating PI4P synthesis during PLC signalling has been studied in the context of overexpressed GPCRs; the role of this enzyme in the context of an endogenous G-protein-coupled PLC pathway remains to be demonstrated. Also, the identity of the PI4K that regulates PI4P synthesis during Drosophila phototransduction remains to be established.

An interesting feature of PI4KIIIα that has emerged from recent studies is that, in multiple systems, the enzyme exists as part of a protein complex. In yeast, the PI4KIIIα orthologue Stt4 exists as part of a complex consisting of three adaptor proteins, the peripheral membrane protein Efr3, the scaffolding protein Ypp1 (TTC7 in mammals) and an integral membrane protein Sfk1 (TMEM150A in mammals); this has also been observed in mammalian cells [105,106]. EFR3 is palmitoylated and is targeted to the PM due to a basic surface patch and by interactions with TMEM150A, and recruits the PI4KIIIα-TTC7 complex by binding to the C-terminus of Efr3 [107]. Collectively, these studies also strongly suggest that PI4KIIIα is localised at the PM and the synthesis of PI4P occurs at this location. More recently, the leukodystrophy protein FAM126A (also known as hyccin) has also been shown to regulate PI4P synthesis and binds to TTC7 and possibly PI4KIIIα as well [108]. These findings suggest that the generation of PI4P by PI4KIIIα is a highly regulated process controlled by multiple proteins; this remains to be understood. Recent studies also indicate that in mammalian cells, PKC can stimulate PI4K activity and so increase PI4P levels [109,110].

PIP 5-kinase

Phosphorylation of PI4P to PI(4,5)P2 is catalysed by PIP5K. In the mammalian genome, there are three genes (PIP5Kα, -β and -γ) encoding PIP5K enzymes that are recruited at the PM as homodimers or heterodimers [111]. One of the striking features of the PIP5Ks is that, in vitro, their activity can be enhanced by PA [112]. This positive feedback by PA may in part ensure that PI(4,5)P2 levels are restored promptly since PI(4,5)P2 is important for other functions, including maintenance of the actin cytoskeleton. PIP5Kγ has many splice variants and one specific splice variant, PIP5Kγ87, was found to be responsible for the generation of PI(4,5)P2 utilised to make IP3 in histamine-stimulated HeLa cells [113]. In a separate study, PIP5Kα was found to be responsible for thrombin-stimulated IP3 generation in platelets [114]. Mice deleted of PIP5Kα or -β are born and develop normally, indicating that PIP5Kγ suffices to maintain PI(4,5)P2 levels. Mice deleted of PIP5Kγ, on the other hand, die shortly after birth due to defects in neuronal function. In the Drosophila genome, there are two PIP5K genes, sktl and dPIP5K. dPIP5K appears to be equivalent to mammalian PIP5Kγ; it is enriched in the adult head and appears to regulate PI(4,5)P2 resynthesis during phototransduction [115]. In contrast, loss of sktl is lethal during development and is also cell-lethal in multiple tissues where this has been tested ([116] and P. Raghu, unpublished work). Although some isoforms of PIP5K have been reported at intracellular membranes, the vast majority of isoforms across multiple species are localised to the PM. Thus, the synthesis of PI(4,5)P2 from PI4P occurs primarily at the PM.

Topological constraints in the PI(4,5)P2 cycle

Activation of both GPCRs and receptor tyrosine kinases stimulates the PLC signalling pathway to hydrolyse PI(4,5)P2 and generate the two ubiquitous second messengers IP3 and DG in almost all metazoans. The activity of PLC is auto-inhibited and external cues lead to its activation [117,118]. The substrate PI(4,5)P2 is mainly localised at the PM and, when cleaved by PLC, generates a series of lipid-soluble intermediates that are spatially restricted to and cannot diffuse away from the membrane at which they are generated. The immediate lipid product of PLC activity is DG that is rapidly converted into PA by DGK; these two reactions are thought to occur at the PM in mammalian cells (Figure 1B). PA is converted into PI by the action of two sequentially acting enzymes, CDS and PIS, which are localised to the ER. This separation of PA generation at the PM from the enzyme that metabolises it, and CDS at the ER membrane constitutes a topological problem necessitating the need to transfer PA from the PM to the ER (Figure 1B).

The final steps in the PI(4,5)P2 cycle are the sequential phosphorylation of PI at the fourth and fifth positions to regenerate PI(4,5)P2 (Figure 1B). The kinases that mediate the phosphorylation of PI to PI4P are not present in the ER but are thought to reside at the PM, once again generating a spatial separation of non-diffusible substrate from the enzymes. To maintain PI(4,5)P2 homeostasis following PLC signalling, the first question to be addressed is the subcellular distribution of PI and how changes in PI levels reflect changes in PI(4,5)P2 levels. PI distribution in cells has been monitored by subcellular fractionation and the majority of PI is found intracellularly, notably at the ER [119,120]. Unlike probes available for detecting the spatial distribution of phosphorylated PIs, no probes for monitoring the spatial distribution of PI have been identified so far. When PIS, an ER-localised enzyme was knocked down, a 40% loss of total PI was noted and this was accompanied by a corresponding decrease in both PI4P and PI(4,5)P2, indicating that PI, PI4P and PI(4,5)P2 were in equilibrium and changes were reflected in PI manifested itself in the total pool of phosphoinositides [85]. Thus, in PIS knockdown cells, changes in PI synthesis at the ER are reflected in changes in PM PI(4,5)P2 reflecting the dependence of PM PI(4,5)P2 synthesis on the availability of PI from the ER. Since it is generally the case that hydrolysis of PI(4,5)P2 manifests itself as a reduction in the PI pool [34,121123], this would imply that the PM depends on PI being made available from the ER. Indeed, in permeabilised cells when PLC activation is intensely driven, the amount of IP3 produced far exceeds the resting levels of PI(4,5)P2 and more importantly, a reduction in PI and not in PI(4,5)P2 is mainly observed [121]. Essentially, hydrolysis of PI(4,5)P2 is dependent on on-going conversion of PI(4,5)P2 from PI.

To demonstrate that PI at the ER contributes to PI(4,5)P2 synthesis at the PM in intact cells, a bacterial PLC was targeted to the ER to deplete PI in this compartment and noted that recovery of PI(4,5)P2 at the PM was delayed following PLC activation [85,124]. Studies by Dickson et al. [125] have recently suggested that PI4P at the Golgi participates in the maintenance of the PM PI(4,5)P2. Targeting of a PI4P phosphatase, Sac1, to the Golgi or the PM led to a decrease in PM PI(4,5)P2. They suggested that 30% of the PM PI(4,5)P2 is derived from a Golgi PI(4)P pool under steady state. In these studies, the KCNQ2/3 current was monitored as an indirect measure of PI(4,5)P2 content. They suggest that Golgi-derived vesicles are responsible for this transport. However, a recent study examined this question and concluded otherwise. The Golgi was disrupted using brefeldin A and, under these conditions, PI(4,5)P2 synthesis at the PM (as monitored by the translocation of the PH domain of PLCδ) was enhanced [124] and not reduced as expected. Thus, from these data, it is clear that PI originating from the ER is the precursor lipid for PI(4,5)P2 synthesis at the PM. Since the on-going production of IP3 in permeabilised cells was dependent on the presence of a PI transfer protein, this would suggest a supply of PI from the ER to the PM. Under these experimental conditions, any protein that has PI transfer activity was capable of supporting PLC signalling. Thus, PITPα, PITPβ and even yeast Sec14p, which has no structural or sequence identity with mammalian PITPs, were capable of supplying PI under these artificial conditions [126,127]. What these experiments establish is that the ER is the main source of PI and, moreover, it has to be transported to the PM.

Bridging the topological constraints in the PI(4,5)P2 cycle

Conceptually, one of the ways of bridging the topological gap in the PI(4,5)P2 cycle would be the use of LTPs that would move lipid molecules between the ER and PM across the aqueous cytosol. This process could also be further optimised if one were to use sites where the ER and PM come in close contact with each other, namely the ER–PM MCS. A brief introduction to the current state of ER–PM contact sites is given below and we then discuss LTPs in the context of the PI(4,5)P2 cycle.

ER–PM contact sites

The components of the PI(4,5)P2–PLC cycle are distributed in two membrane compartments, the ER and PM (Figure 1B). Recent studies indicate that these two compartments can come into close proximity allowing lipid exchange to occur. The ER exhibits a complex network of tubules and cisternae continuous with the nuclear membranes and is involved in vesicular transport, protein synthesis, Ca2+ storage and lipid metabolism [128130]. It is also the largest organelle of the cell representing nearly 50% of the total cellular membranes [131]. The ER can be closely opposed to other organelle membranes, including endosomes, peroxisomes, mitochondria and also the PM. These ER–PM MCS have been by observed by electron microscopy in both mammalian cells, e.g. in muscle cells and neurons, and in fly photoreceptors [74,132,133].

To maintain regions of close membrane contact, tethering proteins that bring the membranes together are necessary. The distance between the ER and the PM at ER–PM contact sites is estimated to be in the range 10–20 nm, sufficiently close to allow proteins in one compartment to directly interact with proteins or even lipids in the other compartment. Several tethering proteins have been identified in the formation of ER–PM contact sites. These include the junctophilins, Ist2 in yeast (related to TMEM16 proteins in mammals), vesicle-associated membrane protein-associated proteins (VAPs; Scs2 and 22 in yeast) and most recently the extended synaptotagmins (E-Syts; Tricalbins in yeast). These proteins are all anchored in the ER and many have domains allowing them to bind lipids on the PM in trans [130,134136]. The junctophilins are necessary for sarcoplasmic reticulum–PM junctions in muscle cells [137], whereas VAPs are involved in numerous interactions at ER–PM contact sites, notably with oxysterol-binding (OSBP)-related protein (ORP) and PITPNM family members. VAPs bind to an FFAT motif on LTPs such as ORPs and PITPNMs. E-Syts are transmembrane proteins with multiple C2 domains and are tethered at the ER, whereas the C2 domains interact with PM PI(4,5)P2. In yeast, ∼40% of the PM contacts the ER while in mammalian cells, ER–PM contacts are less prominent; in many non-excitable cells, <1% of the PM makes contact with the ER. In yeast, combined deletion of the six potential yeast ER–PM tethering proteins, Ist2, the VAP homologues Scs2/22 and all three Tricalbins was necessary to eliminate ER–PM junctions [138].

In mammalian cells, MCS can be induced during Ca2+ signalling, e.g. as it occurs with Stim and Orai [139,140]. Depletion of ER Ca2+ stores by IP3 binding to the IP3 receptor on the ER leads to the process whereby store-operated calcium entry can occur. In brief, the Ca2+ sensors, STIM1 and STIM2, detect the decrease in Ca2+ in the ER stores and this causes a conformational change in STIM proteins such that they expose the SOAR domain, resulting in STIM redistribution to ER–PM contact sites where they assemble with the PM-localised Orai1, which is the store-operated Ca2+ channel [141]. This transition is aided by an ancillary protein STIMATE (TMEM110), which stabilises the ER–PM contact sites [142,143].

In mammalian cells, ER-localised E-Syts have recently been identified as tether proteins that bind to the PM upon elevation of Ca2+ [144]. E-Syts are a family of ER integral membrane proteins with multiple C2 domains that are conserved from yeast to mammals. E-Syts contain an N-terminal ER-membrane anchor, an SMP (synaptotagmin-like mitochondrial-lipid-binding protein) domain and multiple C2 domains. The E-Syt family has three members; E-Syt1 has five C2 domains, and E-Syt2 and E-Syt3 each have three C2 domains (Table 2). C2 domains are independently folded domains that can mediate Ca2+/phospholipid interactions. The SMP domains belong to the TULIP superfamily of lipid/hydrophobic ligand-binding domains [145,146], and the SMP domain of E-Syt2 has a β-barrel structure and dimerises to form an ∼90 Å long cylinder traversed by a channel lined entirely with hydrophobic residues. The channel can accommodate lipids with the lipid fatty acid tails embedded in the hydrophobic channel, whereas the polar headgroup protrudes in the aqueous phase. The SMP channel lacks specificity and binds any glycerolipids without any discrimination, but does not bind cholesterol, sphingomyelin or ceramide [147,148]. The C2 domains of E-Syt2/3 interact with PM PI(4,5)P2, thus tethering the ER to the PM. Tethering of E-Syt1 to the PM is more complex; it occurs when Ca2+ is elevated and is dependent on the C2C domain of E-Syt1 [149]. In the presence of Ca2+, this domain binds to PI(4,5)P2. Since hydrolysis of PI(4,5)P2 by PLC causes a rise in Ca2+, E-Syt1 localises to the PM, but only transiently due to hydrolysis of PI(4,5)P2 [144,150].

E-Syt1 can apparently use their SMP domain to transfer lipids between two membrane compartments in vitro, provided that the C2C domain is intact. Mutations in the Ca2+-binding sites abrogate tethering as well as lipid transfer. Thus, E-Syt1 is a Ca2+-dependent LTP that functions at MCS [150,151]. Unlike other LTPs described later, SMP domains do not bind a single lipid like LTPs but form a channel and can transfer any glycerolipid.

PI transfer proteins

Given the distribution of lipid intermediates of the PI(4,5)P2 cycle between the PM and the ER, it is essential for cells to have in place a mechanism that allows the exchange of lipid intermediates between these two compartments (Figure 1B). Given the high rates that PLC signalling can achieve in vivo, transfer of lipid intermediates via vesicular transport is likely to be a slow and unwieldy solution. The use of LTPs is a potential solution for the exchange of such lipid intermediates. Multiple families of LTPs have been identified in eukaryotic cells that can engage in non-vesicular lipid transport. LTPs fall into two categories: they can be either single-domain cytosolic proteins or multidomain proteins with an LTP domain. To date, intracellular families of LTPs include START, OSBP, CRAL-TRIO, GLTP and PITP. The hallmark of all LTPs is their ability to accommodate a lipid within a hydrophobic cavity and shield it from the aqueous environment. Structures are available for many of the LTPs showing the lipid–protein interactions that take place to hold the lipid in position during transport between membranes [152]. These structures reveal that the entrance to the cavity is blocked by a flexible ‘lid’. In mammalian cells, the PITP family has been identified as the main LTP that can bind and transfer PI and therefore are the pertinent LTPs in ensuring a supply of PI for PI(4,5)P2 synthesis during PLC signalling [153,154].

PITP overview

Since the original discovery of PITPs as a biochemical activity that can mediate PI transfer between membranes in vitro [7] and support PLC signalling in vivo, a large number of genes encoding PITPs have been discovered in metazoa through both molecular cloning and genome sequencing efforts. In the mammalian genome, there are five genes that encode proteins with a PITP domain; these are subdivided into two classes based on sequence homology as well as their lipid-binding characteristics [155,156] (see Table 1 and Figure 2). Class I PITPs, PITPα and PITPβ are single-domain PITPs and can bind and transfer PI or PC, whereas Class II PITPs are proteins that can bind and transfer PI and PA, both of which are lipid intermediates that need to be exchanged during the PI(4,5)P2–PLC cycle. Class II PITPs comprise three members such as PITPNC1, PITPNM1 and PITPNM2. The founding member of the Class II PITPs is RdgB, a multidomain protein with an N-terminal PITP domain [157]. The N-terminal PITP domain of RdgB is followed by a region of low complexity, an FFAT motif, a DDHD domain and an LNS2 domain (Figure 2). The FFAT motif can bind to the ER-resident integral protein VAP. A third member of the PITPNM family is PITPNM3 that lacks the PITP domain. PITPNM3 also contains the FFAT motif and could potentially be a negative regulator by competing to bind with VAP. The structural arrangements of the different domains in PITPNM1 and -2 are not known, but one of the key questions that need to be addressed is whether the PI transfer activity of the PITP domain is kept inhibited by protein interactions and only made accessible after PLC activation. An acidic region has been shown to bind Ca2+, and therefore, changes in cytosol Ca2+ could potentially influence the structure [157,158]. In addition to PITPNM, Class II also includes the short Class IIB PITP, PITPNC1.

Restoration of PI(4,5)P2 levels following phospholipase C signalling is accomplished by PITPNM proteins that can reciprocally transfer PA and PI at MCS.

Figure 2.
Restoration of PI(4,5)P2 levels following phospholipase C signalling is accomplished by PITPNM proteins that can reciprocally transfer PA and PI at MCS.

Model illustrating how PITPNM participates in PA and PI transfer at MCS for the restoration of PI(4,5)P2 levels. The ‘PITP domain’ is modelled on the structure of PITPα. The ‘FFAT’ motif that binds to VAP is indicated. Participation of E-Syt in tethering the two membranes is not included in the diagram for clarity. MSP, major sperm protein homology domain; FFAT motif sequence, EFFDAxE; TM, transmembrane domain.

Figure 2.
Restoration of PI(4,5)P2 levels following phospholipase C signalling is accomplished by PITPNM proteins that can reciprocally transfer PA and PI at MCS.

Model illustrating how PITPNM participates in PA and PI transfer at MCS for the restoration of PI(4,5)P2 levels. The ‘PITP domain’ is modelled on the structure of PITPα. The ‘FFAT’ motif that binds to VAP is indicated. Participation of E-Syt in tethering the two membranes is not included in the diagram for clarity. MSP, major sperm protein homology domain; FFAT motif sequence, EFFDAxE; TM, transmembrane domain.

Table 1
Nomenclature of PITP proteins in common use
Gene name Full name Protein name Synonyms 
PITPNM1 Phosphatidylinositol transfer protein membrane-associated 1 PITPNM1 Nir2, RdgBα1 
PITPNM2 Phosphatidylinositol transfer protein membrane-associated 2 PITPNM2 Nir3, RdgBα2 
PITPNM3 PITPNM family member 3 (lacks a PITP domain) PITPNM3 Nir1, RdgBα3,
pI-RdgB 
PITPNC1 Phosphatidylinositol transfer protein, cytoplasmic 1 PITPNC1 RdgBβ 
PITPNA Phosphatidylinositol transfer protein α PITPα PITPNA
VIB1A 
PITPNB Phosphatidylinositol transfer protein β PITPβ PITPNB
VIB1B 
Gene name Full name Protein name Synonyms 
PITPNM1 Phosphatidylinositol transfer protein membrane-associated 1 PITPNM1 Nir2, RdgBα1 
PITPNM2 Phosphatidylinositol transfer protein membrane-associated 2 PITPNM2 Nir3, RdgBα2 
PITPNM3 PITPNM family member 3 (lacks a PITP domain) PITPNM3 Nir1, RdgBα3,
pI-RdgB 
PITPNC1 Phosphatidylinositol transfer protein, cytoplasmic 1 PITPNC1 RdgBβ 
PITPNA Phosphatidylinositol transfer protein α PITPα PITPNA
VIB1A 
PITPNB Phosphatidylinositol transfer protein β PITPβ PITPNB
VIB1B 
Table 2
Diversity in the enzymes of the PI(4,5)P2 cycle required for PLC signalling: comparison between mammals and Drosophila
Enzymes of the PI(4,5)P2 cycle Mammals Drosophila melanogaster 
PLCs 13 3 
 PLCβ PLCβ1; β2; β3; β4 norpA (CG3620); PLC21C (CG4574) 
 PLCγ PLCγ1; PLCγ2 PLCγ (CG4200) 
 PLCδ PLCδ1; δ3; δ4  
 PLCε PLCε  
 PLCζ PLCζ  
 PLCη PLCη1; η2  
DG kinase 10 5 
 Type I DGKα, DGKβ, DGKγ Dgk (CG34361) 
 Type II DGKδ, DGKη, DGKκ CG31187 
 Type III DGKε DGKε (CG8657) 
 Type IV DGKζ, DGKι rdgA (CG42667) 
 Type V DGKθ CG31140 
CDP-DG synthase 3 2 
 Cds1; Cds2 Cds (CG7962) 
 Tam41 Tam41 (CG33331) 
PI synthase 1 1 
 PIS PIS (CG9245) 
Class I PITPs 2 1 
 PITPα; PITPβ PITP (vib) (CG5269) 
Class II PITPs (RdgB) 3 2 
 IIA PITPNM1; PITPNM2 RdgB (CG11111) 
 IIB PITPNC1 RdgBβ (CG17818) 
PI4K 4 3 
 Type II PI4K PI4KIIA, PI4KIIB PI4KIIA (CG2929) 
 Type III PI4K PI4KIIIA, PI4KIIIB PI4KIIIA (CG10260); fwd, (PI4K3B; CG7004) 
PIP5K 3 2 
 PIP5K1A; PIP5K1B, PIP5K1C Sktl (CG9985); dPIP5K (CG3682
E-Synaptotagmin 3 1 
 E-Syt1; E-Syt2; E-Syt3 E-Syt2 (CG6643) 
VAMP-associated protein (VAP) 2 3 
 VAPA, VAPB dVAP-A (CG5014) 
  dVAP-B (CG33523) 
  fan (CG7919) 
Enzymes of the PI(4,5)P2 cycle Mammals Drosophila melanogaster 
PLCs 13 3 
 PLCβ PLCβ1; β2; β3; β4 norpA (CG3620); PLC21C (CG4574) 
 PLCγ PLCγ1; PLCγ2 PLCγ (CG4200) 
 PLCδ PLCδ1; δ3; δ4  
 PLCε PLCε  
 PLCζ PLCζ  
 PLCη PLCη1; η2  
DG kinase 10 5 
 Type I DGKα, DGKβ, DGKγ Dgk (CG34361) 
 Type II DGKδ, DGKη, DGKκ CG31187 
 Type III DGKε DGKε (CG8657) 
 Type IV DGKζ, DGKι rdgA (CG42667) 
 Type V DGKθ CG31140 
CDP-DG synthase 3 2 
 Cds1; Cds2 Cds (CG7962) 
 Tam41 Tam41 (CG33331) 
PI synthase 1 1 
 PIS PIS (CG9245) 
Class I PITPs 2 1 
 PITPα; PITPβ PITP (vib) (CG5269) 
Class II PITPs (RdgB) 3 2 
 IIA PITPNM1; PITPNM2 RdgB (CG11111) 
 IIB PITPNC1 RdgBβ (CG17818) 
PI4K 4 3 
 Type II PI4K PI4KIIA, PI4KIIB PI4KIIA (CG2929) 
 Type III PI4K PI4KIIIA, PI4KIIIB PI4KIIIA (CG10260); fwd, (PI4K3B; CG7004) 
PIP5K 3 2 
 PIP5K1A; PIP5K1B, PIP5K1C Sktl (CG9985); dPIP5K (CG3682
E-Synaptotagmin 3 1 
 E-Syt1; E-Syt2; E-Syt3 E-Syt2 (CG6643) 
VAMP-associated protein (VAP) 2 3 
 VAPA, VAPB dVAP-A (CG5014) 
  dVAP-B (CG33523) 
  fan (CG7919) 

Table 1 provides a summary of the PITP proteins and the various nomenclature in use. We propose that the HUGO nomenclature (PITPNA, PITPNB, PITPNM and PITPNC1) for the gene names should be adopted for the corresponding proteins and will be used throughout this review. Recent studies in both mammalian cells and Drosophila converge to suggest that the PITPNM proteins are the long-sought-for proteins that can facilitate the transfer of PI and PA between the PM and the ER at MCS during the PLC–PI(4,5)P2 cycle.

Biochemical properties of the PITP domain

PI transfer function

The PITP family of proteins is well characterised with respect to their lipid-binding specificities. PITPs have dual specificity. The first PITP protein to be identified, PITPα, was shown to be present in two complexes when purified from brain; PITPα bound to either PI or to PC [7,8]. The PITPα–PI (or PC) complex is stoichiometric, involving a single lipid molecule [159,160]. The binding pocket can accommodate either a PI or a PC molecule; whereas the phosphate moiety and the two lipid tails occupy identical positions, the headgroup makes contact with different amino acids. The inositol headgroup makes contact with four amino acid residues (T59, K61, E86 and N90) and these residues are conserved in all PITPs. Mutation of any of these amino acids leads to loss of PI binding and transfer, but has little effect on PC binding and transfer. The PI- and PC-bound forms are readily convertible by incubating with excess lipid and can be separated based on charge; the PI–PITPα complex is more negative than the PC–PITPα complex [161]. The lipid-bound forms of PITPs have a ‘closed’ conformation while the apo-form has an ‘open’ conformation. The apo-form is only transiently observed during lipid exchange at the membrane [152]. A recent study has shown that these key determinants of PI binding and transfer activity are also conserved in PITPNM1 when tested both in vitro [127] and in vivo [127,149].

PA transfer function

In contrast with PITPα, the PITP domain of Drosophila PITPNM1 does not transfer PC; rather it binds and transfers PA [127]. The PITP domains of PITPNM proteins bind PI at much higher levels compared with Class I PITPs and can also bind PA but not to the same extent as PITPNC1. PITPNC1 is a cytosolic protein with a PITP domain followed by an unstructured 80-amino-acid C-terminal tail. PITPNC1 bound similar amounts of PI and PA despite the fact that PA constitutes <1% of the total lipids, suggesting a very high affinity for PA binding [10]. Recent studies have shown that the isolated PITP domain of human PITPNM1 or Drosophila PITPNM was able to transfer PA between membrane compartments in vitro very efficiently [156].

An independent study also analysed the ability of PITPNM1 to remove PA from the PM to the ER in human embryonic kidney (HEK)-293 cells expressing the angiotensin II (AngII) receptor [162]. Using a PA reporter, these authors followed the decrease in cytoplasmic fluorescence as the PA probe translocated to the PM upon AngII stimulation. In cells that co-express PITPNM1, the PA probe was removed from the PM faster than in the surrounding cells not expressing the probe. From these experiments, the authors concluded that overexpression of PITPNM1 facilitates the removal of PA from the PM. The PITP domain was required for PA removal. Mutation of T59 to alanine or glutamate of the PITP domain of PITPNM1 prevented PA clearance following stimulation. Unfortunately, no analysis of the PA transfer and binding activity of these mutants in vitro were provided. The structure of the PITP domain of PITPα coupled with mutational analysis has identified that T59 is in the binding pocket and is important for PI binding and transfer. In PITPα, T59 when mutated to glutamate is unable to transfer PI but when mutated to alanine retains partial transfer activity [159,161]. This is also the case for the PITP domain of Drosophila PITPNM1. Thus, it is surprising that the T59A mutant completely prevented PA clearance following stimulation. Analysis of PA transfer in vitro of the two mutants shows that it was unaffected [127]. This result is not surprising because T59 only makes contact with the inositol ring and does not contribute to the binding of the phosphate moiety in the binding pocket. A detailed analysis of the PITP domain of the human PITPNM protein is warranted to clarify these discrepancies. Identification of mutants that are defective in PA but not PI transfer would assist in this analysis. One of the key questions is how the PITP domain accommodates PA in the lipid-binding cavity. It is interesting to note that PITPNC1, which can bind PA extremely well but transfer PA less efficiently, was without effect in clearing PA from the PM [162]. In contrast, the PITP domain of PITPNM1 alone was still partly functional, indicating that MCS make the transfer more efficient.

Spatial organisation of PITPNM in relation to function

PITPNM proteins are membrane-associated proteins

In Drosophila PITPNM, the founding member of the PITPNM family, early work had identified multiple hydrophobic segments downstream of the PITP domain and it was speculated that they could act as transmembrane segments [157]. Subsequently, PITPNM proteins were also identified in mammals where two separate genes encoding PITPNM1 and PITPNM2 were identified [163,164]. A study by Lu et al. [164] investigated whether PITPNM2 was an integral membrane protein. They performed subcellular fractionation studies and found that, although PITPNM2 was associated with the membrane pellet, conditions that disrupt protein–protein interactions, such as 1 M NaCl, high pH, or strong denaturants such as urea or guanidine, solubilised PITPNM2 partially or completely. In contrast, the detergent Triton X-100 failed to solubilise any PITPNM2. Thus, PITPNM2 did not behave as an integral membrane protein. Similar studies conducted with Drosophila PITPNM confirmed that the protein is membrane-associated and can be solubilised under conditions that disrupt protein–protein interactions, confirming that endogenous PITPNM proteins are not integral or cytosolic proteins but are membrane-associated [127].

In 2000, a zebrafish PITPNM homologue (alternative names: PITP-less RdgB, pl-RdgB and PITPNM3) that lacks the N-terminal PITP domain was identified [165]. PITPNM3 is also present in mammals but not in Drosophila. The mouse PITPNM3 shares over 45% amino acid identity with the two mouse PITPNM proteins (PITPNM1 and -2) and was also found to be membrane-associated; it could be released when membranes were treated with the protein denaturant guanidine. In contrast, Triton X-100 that solubilises the membranes did not solubilise the protein, but was found in the pellet probably associated with the cytoskeleton. Similar experiments have also been conducted for PITPNM1, and it was concluded that PITPNM proteins are not integral membrane proteins, but may associate with the membrane through protein–protein interactions [166]. Thus, three independent studies on different PITPNM proteins concluded that they are membrane-associated proteins [164166] and not integral membrane proteins as originally suggested [157]. Thus, it comes as a surprise that PITPNM3 has been recently identified as a transmembrane receptor for the chemokine, CCL18 [167,168]. It was proposed that extracellular CCL18 binds to PITPNM3 and can activate signal transduction pathways including PI3K [169]. This finding has been recently challenged, however [170], and we suggest that PITPNM proteins are membrane-associated proteins rather than integral proteins and it is unlikely that PITPNM3 is a transmembrane receptor for the chemokine, CCL18. Determinants that mediate membrane association of PITPNM proteins include interactions between the FFAT motif with the integral VAPs as well as the LNS2 that can interact with PA as reported recently.

Localisation of PITPNM proteins to MCS

Recent studies indicate that the transfer of PI from the ER to the PM takes place at MCS by PITPNM1. PITPNM1 was found to translocate to ER–PM contact sites following PLC activation by histamine [149,162]. Secondly, knockdown of PITPNM1 resulted in decreased recovery of PI(4,5)P2 at the PM while overexpression enhanced recovery. The PI transfer activity was essential for PI(4,5)P2 replenishment as mutation of key residues that abolished PI transfer was not functional, although they still translocated to the ER–PM contact sites. It is interesting to note that knockdown of PITPNM1 did not inhibit the initial loss of PI(4,5)P2 following histamine stimulation, indicating that basal levels of PI(4,5)P2 were not affected by loss of PITPNM1. This is in contrast with what is observed in Drosophila photoreceptors, where deletion of PITPNM led to a pronounced decrease in resting PI(4,5)P2 levels [127].

In mammalian cells, there are two PITPNM proteins and it is suggested that these two proteins have subtle differences in PI(4,5)P2 replenishment at the PM. PITPNM2 is more sensitive and functions at low levels of PI(4,5)P2 hydrolysis, while PITPNM1 restores PI(4,5)P2 homeostasis at intense levels of stimulation [124]. This was demonstrated by the use of two inhibitors, U73122 (PLC inhibitor) and R59022 (DGK inhibitor). Both drugs inhibited the translocation of PITPNM2 to the PM. Although U73122 has been used in the literature as a PLC inhibitor, it is clear that it is not a direct inhibitor of PLC enzymes. U73122 is a cationic amphiphilic drug and very likely decreases the availability of the PI(4,5)P2 substrate [171]. This would affect PI(4,5)P2 interactions with the cytoskeleton as well. Likewise, R59022 only inhibits the Type I DGKα and -γ and moderately inhibits Type II DGKδ and -κ [172]. In mammalian cells, there are ten DGKs which were subdivided into five classes (see Table 2 and [60]), and Type III DGKε, which is specific for arachidonyl stearoyl DG, is not inhibited by this inhibitor. Identification of the DGK responsible for conversion of PLC-derived DG to PA will assist in clarifying the situation.

What drives the translocation of PITPNM1 to ER–PM contact sites? Translocation to ER–PM contact sites requires PITPNM protein to be associated with the ER and the PM simultaneously. Several studies have shown that PITPNM1 is localised to the ER and translocated to ER–PM contact sites following PLC activation [124,149,162]. Moreover, co-expression of VAP-B increased the ER association of PITPNM1 and, upon stimulation, both proteins clustered and co-localised at MCS. It is interesting to note that CDS2 was also found in the same clusters [162].

Association with the ER is dependent on the FFAT motif that binds to ER-localised VAPs. This is supported by studies in HeLa cells where PITPNM1 mutated at the FFAT-binding motif failed to get recruited at ER–PM contacts [149,162]. Recruitment of PITPNM1 was dependent on prior recruitment of E-Syt1 that decreased the distance between the ER–PM contacts [149]. Thus, these results suggested that the increase in cytosol Ca2+ following PLC activation results in E-Syt1 recruitment, which reduces the distance between the two membranes allowing PITPNM1 to bridge the gap with the FFAT motif facilitating ER association. In this scenario, it is suggested that PITPNM1 can promote the transfer of PI for PI(4,5)P2 replenishment after receptor-induced hydrolysis via its PITP domain [149]. The main evidence to support this can be summarised [149]: (1) down-regulation of E-Syt1 results in a delay in the kinetics of PI(4,5)P2 resynthesis measured using fluorescent reporters for this lipid; (2) knockdown of PITPNM1 leads to a decrease in PI(4,5)P2 replenishment as monitored by measuring the dynamics of PI(4,5)P2 reporters; and (3) overexpression of PITPNM1 enhances PI(4,5)P2 replenishment via PI binding to the PITP domain. These findings suggest that ER–PM MCS may support the PI(4,5)P2 cycle. The effect of reducing MCS function on clearing PA from the PM is not known.

Saheki et al. [150] have examined the recruitment of endogenous E-Syt1 to ER–PM contacts and report that while ionomycin or thapsigargin rapidly translocated E-Syt1, the translocation observed with oxotremorine as a stimulus was transient. Since E-Syt1 binds to PI(4,5)P2 at the membrane in a Ca2+-dependent manner, the pronounced decrease in PI(4,5)P2 by oxotremorine is likely to be responsible for the transient nature of E-Syt1 recruitment [150].

In one study, a synthetic construct, MAPPER (membrane-attached peripheral ER), was used that selectively monitors ER–PM contact sites. The MAPPER construct comprises a signal peptide, GFP, transmembrane (TM) segment, a linker, FRB (FKBP12 rapamycin-binding) domain and a linker followed by a phosphoinositide-binding (PB) domain. Essentially the transmembrane segment resides in the ER while the PB domain can bind to PI(4,5)P2 at the PM. It was found that the increase in cytosol Ca2+ increased ER–PM junctions and this was dependent on the translocation of E-Syt to the junctions [149]. Ca2+-dependent translocation of E-Syt1 to ER–PM junctions has also been demonstrated by another study [144].

To examine how PITPNM proteins sense PI(4,5)P2 hydrolysis, and translocate to the ER–PM contact sites, it was shown that the LNS2 domain of PITPNM1 is recruited by increases in PA [173]. During cell stimulation of cell-surface receptors, many agonists stimulate both PLC and PLD. Both phospholipases produce PA; PLD produces PA directly while PLC produces DG that is rapidly phosphorylated to PA by DGK. A commonly used PLC inhibitor, U73122, was found to inhibit translocation of PITPNM1 to the ER–PM contact sites as was a DGK inhibitor, R59022, suggesting that PLC was the main source of PA [162].

However, in HeLa cells, endogenous PITPNM1 was found to be localised to the Golgi in interphase cells and translocated from the Golgi to the PM upon stimulation with epidermal growth factor (EGF) [173]. Localisation to the Golgi required the PITP domain while translocation to the PM required the C-terminal LNS2 domain (amino acid 911–1244). The LNS2 domain has been identified in lipins (PA phosphatase) and is the catalytic domain; however, in PITPNM proteins, a critical residue is replaced by a serine, meaning that PITPNM would lack intrinsic PA phosphatase activity but could bind PA. Both exogenous addition of PA and stimulation of PLD activity led to the translocation of PITPNM1 to the PM. The isolated LNS2 domain was found to bind liposomes when PA was present. Overexpression of PITPNM1 was found to enhance EGF-stimulated Akt phosphorylation while knockdown reduced EGF-stimulated Akt phosphorylation [173]. The effects of PITPNM1 on Akt required both the PITP domain and the LNS2 domain. Additional studies showed that PITPNM1 regulated both PI(4,5)P2 and PI(3,4,5)P3 levels, and this required both the PITP domain intact for PI transfer and also the LNS2 domain that could bind PA. The main conclusion from this study was that PITPNM1 could translocate to the PM upon increases in PA levels, and that the PITP domain transferred PI to the PM [173].

In contrast with mammalian tissue culture cells, in Drosophila photoreceptors, the single PITPNM is constitutively localised at the SMC, an MCS between the ER and the PM [74]. There is no detectable change in this localisation on receptor stimulation. However, it is possible that there are subtle changes or rearrangements in 3D conformation between the various domains of PITPNM, thus regulating lipid transfer function. These remain to be discovered.

Concluding remarks

In this review, we put the PITPNM proteins as major players in moving PI and PA during the resynthesis of PI(4,5)P2 after PLC signalling (Figure 2). We answer the major outstanding question that was highlighted in the review by Michell [3]. A neat solution is found by identifying a single protein that can accomplish reciprocal transfer of PA for PI to complete the ‘PI cycle’. An important feature that emerges is that lipid transfer takes place at privileged sites where the ER and the PM come into close proximity such that lipid exchange occurs over very short distances. One outstanding question that remains is the structures of the different domains of the PITPNM proteins. This will shed light on whether the PITP domain is held in an inactive state and only becomes operational during PLC signalling. A second question is to identify key residues in the PITP domain of the PITPNM proteins that are important for PA binding as these mutants will facilitate in further testing of the model.

Abbreviations

     
  • AngII

    angiotensin II

  •  
  • CDP

    cytidine diphosphate

  •  
  • CDS

    CDP-DG synthase

  •  
  • COPII

    coat protein II

  •  
  • DG

    diacylglycerol

  •  
  • DGK

    DG kinase

  •  
  • EGF

    epidermal growth factor

  •  
  • ER

    endoplasmic reticulum

  •  
  • E-Syts

    extended synaptotagmins

  •  
  • FRB

    FKBP12 rapamycin-binding

  •  
  • GPCR

    G-protein-coupled receptor

  •  
  • INAD

    inactivation –no-after-potential D

  •  
  • IP3

    inositol 1,4,5-trisphosphate

  •  
  • laza

    lazaro

  •  
  • LTPs

    lipid transfer proteins

  •  
  • MCS

    membrane contact site(s)

  •  
  • MAPPER

    membrane-attached peripheral ER

  •  
  • NAPE-PLD

    N-acyl phosphatidylamine phospholipase D

  •  
  • NORPA

    no receptor Potentail A

  •  
  • ORP

    oxysterol-binding (OSBP)-related protein

  •  
  • OSBP

    oxysterol-binding

  •  
  • PA

    phosphatidic acid

  •  
  • PB

    phosphoinositide-binding

  •  
  • PC

    phosphatidylcholine

  •  
  • PI(4,5)P2

    phosphatidylinositol 4,5-bisphosphate

  •  
  • PI

    phosphatidylinositol

  •  
  • PI3K

    phosphoinositide 3-kinase

  •  
  • PI4K

    phosphoinositide 4-kinase

  •  
  • PI4P

    phosphatidylinositol 4-phosphate

  •  
  • PIP5K

    phosphatidylinositol 4-phosphate 5-kinase

  •  
  • PIS

    PI synthase

  •  
  • PITP

    phosphatidylinositol transfer protein

  •  
  • PITPNM

    PITP membrane-associated

  •  
  • PKC

    protein kinase C

  •  
  • PLC

    phospholipase C

  •  
  • PLCβ4

    phospholipase Cβ4

  •  
  • PLD

    phospholipase D

  •  
  • PM

    plasma membrane

  •  
  • RdgA

    retinal degeneration A

  •  
  • RdgB

    Drosophila retinal degeneration B

  •  
  • SMC

    sub-microvillar cisternae

  •  
  • SMP

    synaptotagmin-like mitochondrial lipid-binding protein

  •  
  • Tm

    transmembrane

  •  
  • TRP

    transient receptor potential

  •  
  • TRPL

    transient receptor potential like

  •  
  • VAMP

    vesicle-associated membrane protein

  •  
  • VAPs

    vesicle-associated membrane protein-associated proteins

  •  
  • VEGF

    vascular endothelial growth factor.

Funding

Work in the Cockcroft Laboratory is funded by grants from Biotechnology and Biological Sciences Research Council [grant ref. BB/J005606/1] and the British Heart Foundation [grant number 044/25411]. Work in the Raghu Laboratory is funded by the National Centre for Biological Sciences — TIFR and a Wellcome Trust — DBT India Alliance Senior Fellowship to P.R. [grant number 1A/S/14/2/501540].

Acknowledgments

We thank Shweta Yadav for her comments and Nicolas Blunsom for his help in creating the diagrams. We also thank Bob Michell for his comments.

Competing Interests

The Authors declare that there are no competing interests associated with the manuscript.

References

References
1
Balla
,
T.
(
2013
)
Phosphoinositides: tiny lipids with giant impact on cell regulation
.
Physiol. Rev.
93
,
1019
1137
doi:
2
Hokin
,
M.R.
and
Hokin
,
L.E.
(
1964
) Interconversions of phosphatidylinositol and phosphatidic acid involved in the response to acetylcholine in the salt gland. In
Metabolism and Physiological Significance of Lipids.
(
Dawson
,
R.M.C.
and
Rhodes
,
D.N.
, eds), pp.
423
434
.
John Wiley
,
New York
3
Michell
,
R.H.
(
1975
)
Inositol phospholipids and cell surface receptor function
.
Biochim. Biophys. Acta Rev. Biomembranes
415
,
81
147
doi:
4
Jafferji
,
S.S.
and
Michell
,
R.H.
(
1976
)
Investigation of the relationship between cell-surface calcium-ion gating and phosphatidylinositol turnover by comparison of the effects of elevated extracellular potassium ion concentration on ileum smooth muscle and pancreas
.
Biochem. J.
160
,
397
399
doi:
5
Berridge
,
M.J.
(
1993
)
Inositol trisphosphate and calcium signalling
.
Nature
361
,
315
325
doi:
6
Lapetina
,
E.G.
and
Michell
,
R.H.
(
1973
)
Phosphatidylinositol metabolism in cells receiving extracellular stimulation
.
FEBS Lett.
31
,
1
10
doi:
7
Helmkamp
, Jr,
G.M.
,
Harvey
,
M.S.
,
Wirtz
,
K.W.
and
van Deenen
,
L.L.
(
1974
)
Phospholipid exchange between membranes. Purification of bovine brain proteins that preferentially catalyze the transfer of phosphatidylinositol
.
J. Biol. Chem.
249
,
6382
6389
PMID:
[PubMed]
8
Thomas
,
G.M.H.
,
Cunningham
,
E.
,
Fensome
,
A.
,
Ball
,
A.
,
Totty
,
N.F.
,
Truong
,
O.
et al. 
(
1993
)
An essential role for phosphatidylinositol transfer protein in phospholipase C-mediated inositol lipid signaling
.
Cell
74
,
919
928
doi:
9
Kauffmann-Zeh
,
A.
,
Thomas
,
G.M.H.
,
Ball
,
A.
,
Prosser
,
S.
,
Cunningham
,
E.
,
Cockcroft
,
S.
et al. 
(
1995
)
Requirement for phosphatidylinositol transfer protein in epidermal growth factor signaling
.
Science
268
,
1188
1190
doi:
10
Garner
,
K.
,
Hunt
,
A.N.
,
Koster
,
G.
,
Somerharju
,
P.
,
Grover
,
E.
,
Li
,
M.
et al. 
(
2012
)
Phosphatidylinositol transfer protein, Cytoplasmic 1 (PITPNC1) binds and transfers phosphatidic acid
.
J. Biol. Chem.
287
,
32263
32276
doi:
11
Whatmore
,
J.
,
Wiedemann
,
C.
,
Somerharju
,
P.
,
Swigart
,
P.
and
Cockcroft
,
S.
(
1999
)
Resynthesis of phosphatidylinositol in permeabilized neutrophils following phospholipase Cβ activation: transport of the intermediate, phosphatidic acid, from the plasma membrane to the endoplasmic reticulum for phosphatidylinositol resynthesis is not dependent on soluble lipid carriers or vesicular transport
.
Biochem. J.
341
,
435
444
doi:
12
Kruse
,
M.
,
Vivas
,
O.
,
Traynor-Kaplan
,
A.
and
Hille
,
B.
(
2016
)
Dynamics of phosphoinositide-dependent signaling in sympathetic neurons
.
J. Neurosci.
36
,
1386
1400
doi:
13
Raghu
,
P.
,
Yadav
,
S.
and
Mallampati
,
N.B.
(
2012
)
Lipid signaling in Drosophila photoreceptors
.
Biochim. Biophys. Acta Mol. Cell Biol. Lipids
1821
,
1154
1165
doi:
14
Katz
,
B.
and
Minke
,
B.
(
2009
)
Drosophila photoreceptors and signaling mechanisms
.
Front. Cell. Neurosci.
3
,
2
doi:
15
Hardie
,
R.C.
and
Juusola
,
M.
(
2015
)
Phototransduction in Drosophila
.
Curr. Opin. Neurobiol.
34
,
37
45
doi:
16
Wang
,
T.
and
Montell
,
C.
(
2007
)
Phototransduction and retinal degeneration in Drosophila
.
Pflugers Arch. Eur. J. Phys.
454
,
821
847
doi:
17
Anderson
,
K.E.
,
Kielkowska
,
A.
,
Durrant
,
T.N.
,
Juvin
,
V.
,
Clark
,
J.
,
Stephens
,
L.R.
et al. 
(
2013
)
Lysophosphatidylinositol-acyltransferase-1 (LPIAT1) is required to maintain physiological levels of PtdIns and PtdInsP2 in the mouse
.
PLoS ONE
8
,
e58425
doi:
18
Holub
,
B.J.
and
Kuksis
,
A.
(
1971
)
Differential distribution of orthophosphate-32P and glycerol-14C among molecular species of phosphatidylinositols of rat liver in vivo
.
J. Lipid Res.
12
,
699
705
19
Luthra
,
M.G.
and
Sheltawy
,
A.
(
1976
)
The metabolic turnover of molecular species of phosphatidylinositol and its precursor phosphatidic acid in Guinea-pig cerebral hemispheres
.
J. Neurochem.
27
,
1503
1511
doi:
20
Imae
,
R.
,
Inoue
,
T.
,
Nakasaki
,
Y.
,
Uchida
,
Y.
,
Ohba
,
Y.
,
Kono
,
N.
et al. 
(
2012
)
LYCAT, a homologue of C. elegans acl-8, acl-9, and acl-10, determines the fatty acid composition of phosphatidylinositol in mice
.
J. Lipid Res.
53
,
335
347
doi:
21
Lee
,
H.-C.
,
Inoue
,
T.
,
Sasaki
,
J.
,
Kubo
,
T.
,
Matsuda
,
S.
,
Nakasaki
,
Y.
et al. 
(
2012
)
LPIAT1 regulates arachidonic acid content in phosphatidylinositol and is required for cortical lamination in mice
.
Mol. Biol. Cell
23
,
4689
4700
doi:
22
Stephens
,
L.
,
Jackson
,
T.R.
and
Hawkins
,
P.T.
(
1993
)
Agonist-stimulated synthesis of phosphatidylinositol(3,4,5)-trisphosphate: a new intracellular signalling system?
Biochim. Biophys. Acta Mol. Cell Res.
1179
,
27
75
doi:
23
Cockcroft
,
S.
,
Taylor
,
J.A.
and
Judah
,
J.D.
(
1985
)
Subcellular localisation of inositol lipid kinases in rat liver
.
Biochim. Biophys. Acta Mol. Cell Res.
845
,
163
170
doi:
24
Hammond
,
G.R.V.
,
Machner
,
M.P.
and
Balla
,
T.
(
2014
)
A novel probe for phosphatidylinositol 4-phosphate reveals multiple pools beyond the Golgi
.
J. Cell Biol.
205
,
113
126
doi:
25
Geison
,
R.L.
,
Banschbach
,
M.W.
,
Sadeghian
,
K.
and
Hokin-Neaverson
,
M.
(
1976
)
Acetylcholine stimulation of selective increases in stearic and arachidonic acids in phosphatidic acid in mouse pancreas
.
Biochem. Biophys. Res. Commun.
68
,
343
349
doi:
26
Broekman
,
M.J.
,
Ward
,
J.W.
and
Marcus
,
A.J.
(
1981
)
Fatty acid composition of phosphatidylinositol and phosphatidic acid in stimulated platelets. Persistence of arachidonyl-stearyl structure
.
J. Biol. Chem.
256
,
8271
8274
PMID:
[PubMed]
27
Liscovitch
,
M.
(
1991
)
Signal-dependent activation of phosphatidylcholine hydrolysis: role of phospholipase D
.
Biochem. Soc. Trans.
19
,
402
407
doi:
28
Pettitt
,
T.R.
,
Martin
,
A.
,
Horton
,
T.
,
Liossis
,
C.
,
Lord
,
J.M.
and
Wakelam
,
M.J.O.
(
1997
)
Diacylglycerol and phosphatidate generated by phospholipases C and D, respectively, have distinct fatty acid compositions and functions. Phospholipase D-derived diacylglycerol does not activate protein kinase C in porcine aortic endothelial cells
.
J. Biol. Chem.
272
,
17354
17359
doi:
29
Oliveira
,
T.G.
,
Chan
,
R.B.
,
Tian
,
H.
,
Laredo
,
M.
,
Shui
,
G.
,
Staniszewski
,
A.
et al. 
(
2010
)
Phospholipase D2 ablation ameliorates Alzheimer's disease-linked synaptic dysfunction and cognitive deficits
.
J. Neurosci.
30
,
16419
16428
doi:
30
Cockcroft
,
S.
(
2001
)
Signalling roles of mammalian phospholipase D1 and D2
.
Cell. Mol. Life Sci.
58
,
1674
1687
doi:
31
Ueda
,
N.
,
Okamoto
,
Y.
and
Morishita
,
J.
(
2005
)
N-acylphosphatidylethanolamine-hydrolyzing phospholipase D: a novel enzyme of the β-lactamase fold family releasing anandamide and other N-acylethanolamines
.
Life Sci.
77
,
1750
1758
doi:
32
García del Caño
,
G.
,
Aretxabala
,
X.
,
González-Burguera
,
I.
,
Montaña
,
M.
,
López de Jesús
,
M.
,
Barrondo
,
S.
et al. 
(
2015
)
Nuclear diacylglycerol lipase-α in rat brain cortical neurons: evidence of 2-arachidonoylglycerol production in concert with phospholipase C-β activity
.
J. Neurochem.
132
,
489
503
doi:
33
Mauco
,
G.
,
Dangelmaier
,
C.A.
and
Smith
,
J.B.
(
1984
)
Inositol lipids, phosphatidate and diacylglycerol share stearoylarachidonoylglycerol as a common backbone in thrombin-stimulated human platelets
.
Biochem. J.
224
,
933
940
doi:
34
Michell
,
R.H.
,
Kirk
,
C.J.
,
Maccallum
,
S.H.
and
Hunt
,
P.A.
(
1988
)
Inositol lipids: receptor-stimulated hydrolysis and cellular lipid pools
.
Philos. Trans. R. Soc. Lond. B Biol. Sci.
320
,
239
246
doi:
35
Zhang
,
L.
,
Malik
,
S.
,
Pang
,
J.
,
Wang
,
H.
,
Park
,
K.M.
,
Yule
,
D.I.
et al. 
(
2013
)
Phospholipase Cε hydrolyzes perinuclear phosphatidylinositol 4-phosphate to regulate cardiac hypertrophy
.
Cell
153
,
216
227
doi:
36
Fukami
,
K.
,
Inanobe
,
S.
,
Kanemaru
,
K.
and
Nakamura
,
Y.
(
2010
)
Phospholipase C is a key enzyme regulating intracellular calcium and modulating the phosphoinositide balance
.
Prog. Lipid Res.
49
,
429
437
doi:
37
Hilgemann
,
D.W.
and
Ball
,
R.
(
1996
)
Regulation of cardiac Na+, Ca2+ exchange and KATP potassium channels by PIP2
.
Science
273
,
956
959
doi:
38
Hille
,
B.
,
Dickson
,
E.J.
,
Kruse
,
M.
,
Vivas
,
O.
and
Suh
,
B.-C.
(
2015
)
Phosphoinositides regulate ion channels
.
Biochim. Biophys. Acta Mol. Cell Biol. Lipids
1851
,
844
856
doi:
39
Hughes
,
S.
,
Marsh
,
S.J.
,
Tinker
,
A.
and
Brown
,
D.A.
(
2007
)
PIP2-dependent inhibition of M-type (Kv7.2/7.3) potassium channels: direct on-line assessment of PIP2 depletion by Gq-coupled receptors in single living neurons
.
Pflugers Arch. Eur. J. Physiol.
455
,
115
124
doi:
40
Brown
,
D.A.
,
Hughes
,
S.A.
,
Marsh
,
S.J.
and
Tinker
,
A.
(
2007
)
Regulation of M(Kv7.2/7.3) channels in neurons by PIP2 and products of PIP2 hydrolysis: significance for receptor-mediated inhibition
.
J. Physiol.
582
,
917
925
doi:
41
Dickson
,
E.J.
,
Falkenburger
,
B.H.
and
Hille
,
B.
(
2013
)
Quantitative properties and receptor reserve of the IP3 and calcium branch of Gq-coupled receptor signaling
.
J. Gen. Physiol.
141
,
521
535
doi:
42
Michell
,
R.H.
,
Jafferji
,
S.S.
and
Jones
,
L.M.
(
1976
)
Receptor occupancy dose-response curve suggests that phosphatidylinositol breakdown may be intrinsic to the mechanism of the muscarinic cholinergic receptor
.
FEBS Lett.
69
,
1
5
doi:
43
Shortridge
,
R.D.
,
Yoon
,
J.
,
Lending
,
C.R.
,
Bloomquist
,
B.T.
,
Perdew
,
M.H.
and
Pak
,
W.L.
(
1991
)
A Drosophila phospholipase C gene that is expressed in the central nervous system
.
J. Biol. Chem.
266
,
12474
12480
PMID:
[PubMed]
44
Bloomquist
,
B.T.
,
Shortridge
,
R.D.
,
Schneuwly
,
S.
,
Perdew
,
M.
,
Montell
,
C.
,
Steller
,
H.
et al. 
(
1988
)
Isolation of a putative phospholipase C gene of Drosophila, norpA and its role in phototransduction
.
Cell
54
,
723
733
doi:
45
Thackeray
,
J.R.
,
Gaines
,
P.C.
,
Ebert
,
P.
and
Carlson
,
J.R.
(
1998
)
Small wing encodes a phospholipase C-(gamma) that acts as a negative regulator of R7 development in Drosophila
.
Development
125
,
5033
5042
PMID:
[PubMed]
46
Balakrishnan
,
S.S.
,
Basu
,
U.
and
Raghu
,
P.
(
2015
)
Phosphoinositide signalling in Drosophila
.
Biochim. Biophys. Acta Mol. Cell Biol. Lipids
1851
,
770
784
doi:
47
Kain
,
P.
,
Chakraborty
,
T.S.
,
Sundaram
,
S.
,
Siddiqi
,
O.
,
Rodrigues
,
V.
and
Hasan
,
G.
(
2008
)
Reduced odor responses from antennal neurons of Gqα, phospholipase Cβ, and rdgA mutants in Drosophila support a role for a phospholipid intermediate in insect olfactory transduction
.
J. Neurosci.
28
,
4745
4755
doi:
48
Hardie
,
R.C.
,
Gu
,
Y.
,
Martin
,
F.
,
Sweeney
,
S.T.
and
Raghu
,
P.
(
2004
)
In vivo light-induced and basal phospholipase C activity in Drosophila photoreceptors measured with genetically targeted phosphatidylinositol 4,5-bisphosphate-sensitive ion channels (Kir2.1)
.
J. Biol. Chem.
279
,
47773
47782
doi:
49
Hardie
,
R.C.
and
Raghu
,
P.
(
2001
)
Visual transduction in Drosophila
.
Nature
413
,
186
193
doi:
50
Hardie
,
R.C.
,
Liu
,
C.-H.
,
Randall
,
A.S.
and
Sengupta
,
S.
(
2015
)
In vivo tracking of phosphoinositides in Drosophila photoreceptors
.
J. Cell Sci.
128
,
4328
4340
doi:
51
Hardie
,
R.C.
,
Raghu
,
P.
,
Moore
,
S.
,
Juusola
,
M.
,
Baines
,
R.A.
and
Sweeney
,
S.T.
(
2001
)
Calcium influx via TRP channels is required to maintain PIP2 levels in Drosophila photoreceptors
.
Neuron
30
,
149
159
doi:
52
Running Deer
,
J.L.
,
Hurley
,
J.B.
and
Yarfitz
,
S.L.
(
1995
)
G protein control of Drosophila photoreceptor phospholipase C
.
J. Biol. Chem.
270
,
12623
12628
doi:
53
Kolay
,
S.
,
Basu
,
U.
and
Raghu
,
P.
(
2016
)
Control of diverse subcellular processes by a single multi-functional lipid phosphatidylinositol 4,5-bisphosphate [PI(4,5)P2]
.
Biochem. J.
473
,
1681
1692
doi:
54
Raghu
,
P.
(
2006
)
Regulation of Drosophila TRPC channels by protein and lipid interactions
.
Semin. Cell Dev. Biol.
17
,
646
653
doi:
55
Tsunoda
,
S.
,
Sierralta
,
J.
,
Sun
,
Y.
,
Bodner
,
R.
,
Suzuki
,
E.
,
Becker
,
A.
et al. 
(
1997
)
A multivalent PDZ-domian protein assembles signalling complexes in a G-protein-coupled cascade
.
Nature
388
,
243
249
doi:
56
Kim
,
K.-H.
,
Lim
,
S.
,
Kim
,
J.
,
Kim
,
S.
,
Kim
,
J.H.
,
Ryu
,
S.H.
et al. 
(
2011
)
Subtype-specific roles of phospholipase C-β via differential interactions with PDZ domain proteins
.
Adv. Enzyme Regul.
51
,
138
151
doi:
57
Scott
,
K.
and
Zuker
,
C.S.
(
1998
)
Assembly of the Drosophila phototransduction cascade into a signalling complex shapes elementary responses
.
Nature
395
,
805
808
doi:
58
Henderson
,
S.R.
,
Reuss
,
H.
and
Hardie
,
R.C.
(
2000
)
Single photon responses in Drosophila photoreceptors and their regulation by Ca2+
.
J. Physiol.
524
,
179
194
doi:
59
Li
,
H.-S.
and
Montell
,
C.
(
2000
)
TRP and the PDZ protein, INAD, form the core complex required for retention of the signalplex in Drosophila photoreceptor cells
.
J. Cell Biol.
150
,
1411
1422
doi:
60
Mérida
,
I.
,
Ávila-Flores
,
A.
and
Merino
,
E.
(
2008
)
Diacylglycerol kinases: at the hub of cell signalling
.
Biochem. J.
409
,
1
18
doi:
61
Lung
,
M.
,
Shulga
,
Y.V.
,
Ivanova
,
P.T.
,
Myers
,
D.S.
,
Milne
,
S.B.
,
Brown
,
H.A.
et al. 
(
2009
)
Diacylglycerol kinase ε is selective for both acyl chains of phosphatidic acid or diacylglycerol
.
J. Biol. Chem.
284
,
31062
31073
doi:
62
Bennett
,
J.P.
,
Cockcroft
,
S.
,
Caswell
,
A.H.
and
Gomperts
,
B.D.
(
1982
)
Plasma-membrane location of phosphatidylinositol hydrolysis in rabbit neutrophils stimulated with formylmethionyl-leucylphenylalanine
.
Biochem. J.
208
,
801
808
doi:
63
Flores
,
I.
,
Casaseca
,
T.
,
Martinez
,
A.
,
Kanoh
,
H.
and
Merida
,
I.
(
1996
)
Phosphatidic acid generation through interleukin 2 (IL-2)-induced alpha-diacylglycerol kinase activation is an essential step in IL-2-mediated lymphocyte proliferation
.
J. Biol. Chem.
271
,
10334
10340
doi:
64
Merida
,
I.
,
Andrada
,
E.
,
Gharbi
,
S.I.
and
Avila-Flores
,
A.
(
2015
)
Redundant and specialized roles for diacylglycerol kinases α and ζ in the control of T cell functions
.
Sci. Signal.
8
,
re6
doi:
65
Masai
,
I.
,
Okazaki
,
A.
,
Hosoya
,
T.
and
Hotta
,
Y.
(
1993
)
Drosophila retinal degeneration A gene encodes an eye-specific diacylglycerol kinase with cysteine-rich zinc-finger motifs and ankyrin repeats
.
Proc. Natl Acad. Sci. USA
90
,
11157
11161
doi:
66
Raghu
,
P.
,
Usher
,
K.
,
Jonas
,
S.
,
Chyb
,
S.
,
Polyanovsky
,
A.
and
Hardie
,
R.C.
(
2000
)
Constitutive activity of the light-sensitive channels TRP and TRPL in the Drosophila diacylglycerol kinase mutant, rdgA
.
Neuron
26
,
169
179
doi:
67
Hardie
,
R.C.
,
Martin
,
F.
,
Cochrane
,
G.W.
,
Juusola
,
M.
,
Georgiev
,
P.
and
Raghu
,
P.
(
2002
)
Molecular basis of amplification in Drosophila phototransduction: roles for G protein, phospholipase C, and diacylglycerol kinase
.
Neuron
36
,
689
701
doi:
68
Harris
,
W.A.
and
Stark
,
W.S.
(
1977
)
Hereditary retinal degeneration in Drosophila melanogaster. A mutant defect associated with the phototransduction process
.
J. Gen. Physiol.
69
,
261
291
doi:
69
Georgiev
,
P.
,
Garcia-Murillas
,
I.
,
Ulahannan
,
D.
,
Hardie
,
R.C.
and
Raghu
,
P.
(
2005
)
Functional INAD complexes are required to mediate degeneration in photoreceptors of the Drosophila rdgA mutant
.
J. Cell Sci.
118
,
1373
1384
doi:
70
Inoue
,
H.
,
Yoshioka
,
T.
and
Hotta
,
Y.
(
1989
)
Diacylglycerol kinase defect in a Drosophila retinal degeneration mutant rdgA
.
J. Biol. Chem.
264
,
5996
6000
PMID:
[PubMed]
71
Garcia-Murillas
,
I.
,
Pettitt
,
T.
,
Macdonald
,
E.
,
Okkenhaug
,
H.
,
Georgiev
,
P.
,
Trivedi
,
D.
et al. 
(
2006
)
lazaro Encodes a lipid phosphate phosphohydrolase that regulates phosphatidylinositol turnover during Drosophila phototransduction
.
Neuron
49
,
533
546
doi:
72
Raghu
,
P.
and
Hardie
,
R.C.
(
2009
)
Regulation of Drosophila TRPC channels by lipid messengers
.
Cell Calcium
45
,
566
573
doi:
73
Masai
,
I.
,
Suzuki
,
E.
,
Yoon
,
C.-S.
,
Kohyama
,
A.
and
Hotta
,
Y.
(
1997
)
Immunolocalization of Drosophila eye-specific diacylgylcerol kinase, rdgA, which is essential for the maintenance of the photoreceptor
.
J. Neurobiol.
32
,
695
706
doi:
74
Yadav
,
S.
,
Cockcroft
,
S.
and
Raghu
,
P.
(
2016
)
The Drosophila photoreceptor as a model system for studying signalling at membrane contact sites
.
Biochem. Soc. Trans.
44
,
447
451
doi:
75
Brindley
,
D.N.
and
Waggoner
,
D.W.
(
1998
)
Mammalian lipid phosphate phosphohydrolases
.
J. Biol. Chem.
273
,
24281
24284
doi:
76
Kwon
,
Y.
and
Montell
,
C.
(
2006
)
Dependence on the lazaro phosphatidic acid phosphatase for the maximum light response
.
Curr. Biol.
16
,
723
729
doi:
77
Inglis-Broadgate
,
S.L.
,
Ocaka
,
L.
,
Banerjee
,
R.
,
Gaasenbeek
,
M.
,
Chapple
,
J.P.
,
Cheetham
,
M.E.
et al. 
(
2005
)
Isolation and characterization of murine Cds (CDP-diacylglycerol synthase) 1 and 2
.
Gene
356
,
19
31
doi:
78
Halford
,
S.
,
Dulai
,
K.S.
,
Daw
,
S.C.
,
Fitzgibbon
,
J.
and
Hunt
,
D.M.
(
1998
)
Isolation and chromosomal localization of two human CDP-diacylglycerol synthase (CDS) genes
.
Genomics
54
,
140
144
doi:
79
Lykidis
,
A.
,
Jackson
,
P.D.
,
Rock
,
C.O.
and
Jackowski
,
S.
(
1997
)
The role of CDP-diacylglycerol synthetase and phosphatidylinositol synthase activity levels in the regulation of cellular phosphatidylinositol content
.
J. Biol. Chem.
272
,
33402
33409
doi:
80
Zhou
,
Y.
,
Peisker
,
H.
,
Weth
,
A.
,
Baumgartner
,
W.
,
Dörmann
,
P.
and
Frentzen
,
M.
(
2013
)
Extraplastidial cytidinediphosphate diacylglycerol synthase activity is required for vegetative development in Arabidopsis thaliana
.
Plant J.
75
,
867
879
doi:
81
Haselier
,
A.
,
Akbari
,
H.
,
Weth
,
A.
,
Baumgartner
,
W.
and
Frentzen
,
M.
(
2010
)
Two closely related genes of Arabidopsis encode plastidial cytidinediphosphate diacylglycerol synthases essential for photoautotrophic growth
.
Plant Physiol.
153
,
1372
1384
doi:
82
Shen
,
H.
,
Heacock
,
P.N.
,
Clancey
,
C.J.
and
Dowhan
,
W.
(
1996
)
The CDS1 gene encoding CDP-diacylglycerol synthase in Saccharomyces cerevisiae is essential for cell growth
.
J. Biol. Chem.
271
,
789
795
doi:
83
Wu
,
L.
,
Niemeyer
,
B.
,
Colley
,
N.
,
Socolich
,
M.
and
Zuker
,
C.S.
(
1995
)
Regulation of PLC-mediated signalling in vivo by CDP-diacylglycerol synthase
.
Nature
373
,
216
222
doi:
84
D'Souza
,
K.
,
Kim
,
Y.J.
,
Balla
,
T.
and
Epand
,
R.M.
(
2014
)
Distinct properties of the two isoforms of CDP-diacylglycerol synthase
.
Biochemistry
53
,
7358
7367
doi:
85
Kim
,
Y.J.
,
Guzman-Hernandez
,
M.L.
and
Balla
,
T.
(
2011
)
A highly dynamic ER-derived phosphatidylinositol-synthesizing organelle supplies phosphoinositides to cellular membranes
.
Dev. Cell
21
,
813
824
doi:
86
Pan
,
W.
,
Pham
,
V.N.
,
Stratman
,
A.N.
,
Castranova
,
D.
,
Kamei
,
M.
,
Kidd
,
K.R.
et al. 
(
2012
)
CDP-diacylglycerol synthetase-controlled phosphoinositide availability limits VEGFA signaling and vascular morphogenesis
.
Blood
120
,
489
498
doi:
87
Mok
,
A.
,
Wong
,
T.
,
Filgueiras
,
O.
,
Casola
,
P.G.
,
Nicholson
,
D.W.
,
McMurray
,
W.C.
et al. 
(
1988
)
Characterization of CDPdiacylglycerol hydrolase in mitochondrial and microsomal fractions from rat lung
.
Biochem. Cell Biol.
66
,
425
435
doi:
88
Mok
,
A.Y.
,
McDougall
,
G.E.
and
McMurray
,
W.C.
(
1992
)
CDP-diacylglycerol synthesis in rat liver mitochondria
.
FEBS Lett.
312
,
236
240
doi:
89
Kuchler
,
K.
,
Daum
,
G.
and
Paltauf
,
F.
(
1986
)
Subcellular and submitochondrial localization of phospholipid-synthesizing enzymes in Saccharomyces cerevisiae
.
J. Bacteriol.
165
,
901
910
PMID:
[PubMed]
90
Tamura
,
Y.
,
Harada
,
Y.
,
Nishikawa
,
S.-i.
,
Yamano
,
K.
,
Kamiya
,
M.
,
Shiota
,
T.
et al. 
(
2013
)
Tam41 is a CDP-diacylglycerol synthase required for cardiolipin biosynthesis in mitochondria
.
Cell Metab.
17
,
709
718
doi:
91
Connerth
,
M.
,
Tatsuta
,
T.
,
Haag
,
M.
,
Klecker
,
T.
,
Westermann
,
B.
and
Langer
,
T.
(
2012
)
Intramitochondrial transport of phosphatidic acid in yeast by a lipid transfer protein
.
Science
338
,
815
818
doi:
92
Potting
,
C.
,
Tatsuta
,
T.
,
König
,
T.
,
Haag
,
M.
,
Wai
,
T.
,
Aaltonen
,
M.J.
et al. 
(
2013
)
TRIAP1/PRELI complexes prevent apoptosis by mediating intramitochondrial transport of phosphatidic acid
.
Cell Metab.
18
,
287
295
doi:
93
Potting
,
C.
,
Wilmes
,
C.
,
Engmann
,
T.
,
Osman
,
C.
and
Langer
,
T.
(
2010
)
Regulation of mitochondrial phospholipids by Ups1/PRELI-like proteins depends on proteolysis and Mdm35
.
EMBO J.
29
,
2888
2898
doi:
94
Scharwey
,
M.
,
Tatsuta
,
T.
and
Langer
,
T.
(
2013
)
Mitochondrial lipid transport at a glance
.
J. Cell Sci.
126
,
5317
5323
doi:
95
Tatsuta
,
T.
,
Scharwey
,
M.
and
Langer
,
T.
(
2014
)
Mitochondrial lipid trafficking
.
Trends Cell Biol.
24
,
44
52
doi:
96
Watanabe
,
Y.
,
Tamura
,
Y.
,
Kawano
,
S.
and
Endo
,
T.
(
2015
)
Structural and mechanistic insights into phospholipid transfer by Ups1-Mdm35 in mitochondria
.
Nat. Commun.
6
,
7922
doi:
97
Yu
,
F.
,
He
,
F.
,
Yao
,
H.
,
Wang
,
C.
,
Wang
,
J.
,
Li
,
J.
et al. 
(
2015
)
Structural basis of intramitochondrial phosphatidic acid transport mediated by Ups1-Mdm35 complex
.
EMBO Rep.
16
,
813
823
doi:
98
Holub
,
B.J.
(
1974
)
The Mn2+-activated incorporation of inositol into molecular species of phosphatidylinositol in rat liver microsomes
.
Biochim. Biophys. Acta Lipid Lipid Metab.
369
,
111
122
doi:
99
Tanaka
,
S.
,
Nikawa
,
J.-i.
,
Imai
,
H.
,
Yamashita
,
S.
and
Hosaka
,
K.
(
1996
)
Molecular cloning of rat phosphatidylinositol synthase cDNA by functional complementation of the yeast Saccharomyces cerevisiae pis mutation
.
FEBS Lett.
393
,
89
92
doi:
100
Bochud
,
A.
and
Conzelmann
,
A.
(
2015
)
The active site of yeast phosphatidylinositol synthase Pis1 is facing the cytosol
.
Biochim. Biophys. Acta Mol. Cell Biol. Lipid
1851
,
629
640
doi:
101
Wang
,
T.
and
Montell
,
C.
(
2006
)
A phosphoinositide synthase required for a sustained light response
.
J. Neurosci.
26
,
12816
12825
doi:
102
English
,
A.R.
and
Voeltz
,
G.K.
(
2013
)
Rab10 GTPase regulates ER dynamics and morphology
.
Nat. Cell Biol.
15
,
169
178
doi:
103
Balla
,
A.
,
Kim
,
Y.J.
,
Varnai
,
P.
,
Szentpetery
,
Z.
,
Knight
,
Z.
,
Shokat
,
K.M.
et al. 
(
2008
)
Maintenance of hormone-sensitive phosphoinositide pools in the plasma membrane requires phosphatidylinositol 4-Kinase IIIα
.
Mol. Biol. Cell
19
,
711
721
doi:
104
Bojjireddy
,
N.
,
Botyanszki
,
J.
,
Hammond
,
G.
,
Creech
,
D.
,
Peterson
,
R.
,
Kemp
,
D.C.
et al. 
(
2014
)
Pharmacological and genetic targeting of the PI4KA enzyme reveals its important role in maintaining plasma membrane phosphatidylinositol 4-phosphate and phosphatidylinositol 4,5-bisphosphate levels
.
J. Biol. Chem.
289
,
6120
6132
doi:
105
Nakatsu
,
F.
,
Baskin
,
J.M.
,
Chung
,
J.
,
Tanner
,
L.B.
,
Shui
,
G.
,
Lee
,
S.Y.
et al. 
(
2012
)
Ptdins4p synthesis by PI4KIIIα at the plasma membrane and its impact on plasma membrane identity
.
J. Cell Biol.
199
,
1003
1016
doi:
106
Baird
,
D.
,
Stefan
,
C.
,
Audhya
,
A.
,
Weys
,
S.
and
Emr
,
S.D.
(
2008
)
Assembly of the PtdIns 4-kinase Stt4 complex at the plasma membrane requires Ypp1 and Efr3
.
J. Cell Biol.
183
,
1061
1074
doi:
107
Wu
,
X.
,
Chi
,
R.J.
,
Baskin
,
J.M.
,
Lucast
,
L.
,
Burd
,
C.G.
,
De Camilli
,
P.
et al. 
(
2014
)
Structural insights into assembly and regulation of the plasma membrane phosphatidylinositol 4-kinase complex
.
Dev. Cell
28
,
19
29
doi:
108
Baskin
,
J.M.
,
Wu
,
X.
,
Christiano
,
R.
,
Oh
,
M.S.
,
Schauder
,
C.M.
,
Gazzerro
,
E.
et al. 
(
2016
)
The leukodystrophy protein FAM126A (hyccin) regulates PtdIns(4)P synthesis at the plasma membrane
.
Nat. Cell Biol.
18
,
132
138
doi:
109
Wuttke
,
A.
,
Sagetorp
,
J.
and
Tengholm
,
A.
(
2010
)
Distinct plasma-membrane PtdIns(4)P and PtdIns(4,5)P2 dynamics in secretagogue-stimulated β-cells
.
J. Cell Sci.
123
,
1492
1502
doi:
110
Tóth
,
J.T.
,
Gulyás
,
G.
,
Tóth
,
D.J.
,
Balla
,
A.
,
Hammond
,
G.R.
,
Hunyady
,
L.
et al. 
(
2016
)
BRET-monitoring of the dynamic changes of inositol lipid pools in living cells reveals a PKC-dependent PtdIns4P increase upon EGF and M3 receptor activation
.
Biochim. Biophys. Acta Mol. Cell Biol. Lipids
1861
,
177
187
doi:
111
Lacalle
,
R.A.
,
de Karam
,
J.C.
,
Martinez-Munoz
,
L.
,
Artetxe
,
I.
,
Peregil
,
R.M.
,
Sot
,
J.
et al. 
(
2015
)
Type I phosphatidylinositol 4-phosphate 5-kinase homo- and heterodimerization determines its membrane localization and activity
.
FASEB J.
29
,
2371
2385
doi:
112
Cockcroft
,
S.
(
2009
)
Phosphatidic acid regulation of phosphatidylinositol 4-phosphate 5-kinases
.
Biochim. Biophys. Acta Mol. Cell Biol. Lipids
1791
,
905
912
doi:
113
Wang
,
Y.J.
,
Li
,
W.H.
,
Wang
,
J.
,
Xu
,
K.
,
Dong
,
P.
,
Luo
,
X.
et al. 
(
2004
)
Critical role of PIP5KIγ87 in InsP3-mediated Ca2+ signaling
.
J. Cell Biol.
167
,
1005
1010
doi:
114
Wang
,
Y.
,
Chen
,
X.
,
Lian
,
L.
,
Tang
,
T.
,
Stalker
,
T.J.
,
Sasaki
,
T.
et al. 
(
2008
)
Loss of PIP5KIβ demonstrates that PIP5KI isoform-specific PIP2 synthesis is required for IP3 formation
.
Proc. Natl Acad. Sci. USA
105
,
14064
14069
doi:
115
Chakrabarti
,
P.
,
Kolay
,
S.
,
Yadav
,
S.
,
Kumari
,
K.
,
Nair
,
A.
,
Trivedi
,
D.
et al. 
(
2015
)
A dPIP5K dependent pool of phosphatidylinositol 4,5 bisphosphate (PIP2) is required for G-protein coupled signal transduction in Drosophila photoreceptors
.
PLoS Genet.
11
,
e1004948
doi:
116
Hassan
,
B.A.
,
Prokopenko
,
S.N.
,
Breuer
,
S.
,
Zhang
,
B.
,
Paululat
,
A.
and
Bellen
,
H.J.
(
1998
)
Skittles, a Drosophila phosphatidylinositol 4-phosphate 5-kinase, is required for cell viability, germline development and bristle morphology, but not for neurotransmitter release
.
Genetics
150
,
1527
1537
117
Bunney
,
T.D.
and
Katan
,
M.
(
2011
)
PLC regulation: emerging pictures for molecular mechanisms
.
Trends Biochem. Sci.
36
,
88
96
doi:
118
Hicks
,
S.N.
,
Jezyk
,
M.R.
,
Gershburg
,
S.
,
Seifert
,
J.P.
,
Harden
,
T.K.
and
Sondek
,
J.
(
2008
)
General and versatile autoinhibition of PLC isozymes
.
Mol. Cell
31
,
383
394
doi:
119
Kirk
,
C.J.
,
Michell
,
R.H.
and
Hems
,
D.A.
(
1981
)
Phosphatidylinositol metabolism in rat hepatocytes stimulated by vasopressin
.
Biochem. J.
194
,
155
165
doi:
120
Whatmore
,
J.
,
Morgan
,
C.P.
,
Cunningham
,
E.
,
Collison
,
K.S.
,
Willison
,
K.R.
and
Cockcroft
,
S.
(
1996
)
ADP-ribosylation factor 1-regulated phospholipase D activity is localized at the plasma membrane and intracellular organelles in HL60 cells
.
Biochem. J.
320
,
785
794
doi:
121
Cunningham
,
E.
,
Thomas
,
G.M.H.
,
Ball
,
A.
,
Hiles
,
I.
and
Cockcroft
,
S.
(
1995
)
Phosphatidylinositol transfer protein dictates the rate of inositol trisphosphate production by promoting the synthesis of PIP2
.
Curr. Biol.
5
,
775
783
doi:
122
Jones
,
L.M.
,
Cockcroft
,
S.
and
Michell
,
R.H.
(
1979
)
Stimulation of phosphatidylinositol turnover in various tissues by cholinergic and adrenergic agonists, by histamine and by caerulein
.
Biochem. J.
182
,
669
676
doi:
123
Jones
,
L.M.
and
Michell
,
R.H.
(
1974
)
Breakdown of phosphatidylinositol provoked by muscarinic cholinergic stimulation of rat parotid-gland fragments
.
Biochem. J.
142
,
583
590
doi:
124
Chang
,
C.-L.
and
Liou
,
J.
(
2015
)
Phosphatidylinositol 4,5-bisphosphate homeostasis regulated by Nir2 and Nir3 proteins at endoplasmic reticulum-plasma membrane junctions
.
J. Biol. Chem.
290
,
14289
14301
doi:
125
Dickson
,
E.J.
,
Jensen
,
J.B.
and
Hille
,
B.
(
2014
)
Golgi and plasma membrane pools of PI(4)P contribute to plasma membrane PI(4,5)P2 and maintenance of KCNQ2/3 ion channel current
.
Proc. Natl Acad. Sci. USA
111
,
E2281
E2290
doi:
126
Cunningham
,
E.
,
Tan
,
S.W.
,
Swigart
,
P.
,
Hsuan
,
J.
,
Bankaitis
,
V.
and
Cockcroft
,
S.
(
1996
)
The yeast and mammalian isoforms of phosphatidylinositol transfer protein can all restore phospholipase C-mediated inositol lipid signaling in cytosol-depleted RBL-2H3 and HL-60 cells
.
Proc. Natl Acad. Sci. USA
93
,
6589
6593
doi:
127
Yadav
,
S.
,
Garner
,
K.
,
Georgiev
,
P.
,
Li
,
M.
,
Gomez-Espinosa
,
E.
,
Panda
,
A.
et al. 
(
2015
)
RDGBα, a ptdIns-PtdOH transfer protein, regulates G-protein-coupled PtdIns(4,5)P2 signalling during Drosophila phototransduction
.
J. Cell Sci.
128
,
3330
3344
doi:
128
English
,
A.R.
and
Voeltz
,
G.K.
(
2013
)
Endoplasmic reticulum structure and interconnections with other organelles
.
Cold Spring Harb. Perspect. Biol.
5
,
a013227
doi:
129
Lahiri
,
S.
,
Toulmay
,
A.
and
Prinz
,
W.A.
(
2015
)
Membrane contact sites, gateways for lipid homeostasis
.
Curr. Opin. Cell Biol.
33
,
82
87
doi:
130
Henne
,
W.M.
,
Liou
,
J.
and
Emr
,
S.D.
(
2015
)
Molecular mechanisms of inter-organelle ER-PM contact sites
.
Curr. Opin. Cell Biol.
35
,
123
130
doi:
131
Jelsema
,
C.L.
and
Morre
,
D.J.
(
1978
)
Distribution of phospholipid biosynthetic enzymes among cell components of rat liver
.
J. Biol. Chem.
253
,
7960
7971
PMID:
[PubMed]
132
Henkart
,
M.
,
Landis
,
D.M.
and
Reese
,
T.S.
(
1976
)
Similarity of junctions between plasma membranes and endoplasmic reticulum in muscle and neurons
.
J. Cell Biol.
70
,
338
347
doi:
133
Suzuki
,
E.
and
Hirosawa
,
K.
(
1994
)
Immunolocalization of a Drosophila phosphatidylinositol transfer protein (rdgB) in normal and rdgA mutant photoreceptor cells with special reference to the subrhabdomeric cisternae
.
J. Electron Microsc. (Tokyo)
43
,
183
189
PMID:
[PubMed]
134
Stefan
,
C.J.
,
Manford
,
A.G.
and
Emr
,
S.D.
(
2013
)
ER-PM connections: sites of information transfer and inter-organelle communication
.
Curr. Opin. Cell Biol.
25
,
434
442
doi:
135
Kunzelmann
,
K.
,
Cabrita
,
I.
,
Wanitchakool
,
P.
,
Ousingsawat
,
J.
,
Sirianant
,
L.
,
Benedetto
,
R.
et al. 
(
2016
)
Modulating Ca2+ signals: a common theme for TMEM16, Ist2, and TMC
.
Pflugers Arch. Eur. J. Physiol.
468
,
475
490
doi:
136
Kunzelmann
,
K.
(
2015
)
TMEM16, LRRC8A, bestrophin: chloride channels controlled by Ca2+ and cell volume
.
Trends Biochem. Sci.
40
,
535
543
doi:
137
Takeshima
,
H.
,
Hoshijima
,
M.
and
Song
,
L.S.
(
2015
)
Ca2+ microdomains organized by junctophilins
.
Cell Calcium
58
,
349
356
doi:
138
Manford
,
A.G.
,
Stefan
,
C.J.
,
Yuan
,
H.L.
,
Macgurn
,
J.A.
and
Emr
,
S.D.
(
2012
)
ER-to-plasma membrane tethering proteins regulate cell signaling and ER morphology
.
Dev. Cell
23
,
1129
1140
doi:
139
Prakriya
,
M.
and
Lewis
,
R.S.
(
2015
)
Store-operated calcium channels
.
Physiol. Rev.
95
,
1383
1436
doi:
140
Chang
,
C.-L.
and
Liou
,
J.
(
2016
)
Homeostatic regulation of the PI(4,5)P2–Ca2+ signaling system at ER–PM junctions
.
Biochim. Biophys. Acta Mol. Cell Biol. Lipids
1861
,
862
873
doi:
141
Kim
,
J.Y.
and
Muallem
,
S.
(
2011
)
Unlocking SOAR releases STIM
.
EMBO J.
30
,
1673
1675
doi:
142
Quintana
,
A.
,
Rajanikanth
,
V.
,
Farber-Katz
,
S.
,
Gudlur
,
A.
,
Zhang
,
C.
,
Jing
,
J.
et al. 
(
2015
)
TMEM110 regulates the maintenance and remodeling of mammalian ER-plasma membrane junctions competent for STIM-ORAI signaling
.
Proc. Natl Acad. Sci. USA
112
,
E7083
E7092
doi:
143
Jing
,
J.
,
He
,
L.
,
Sun
,
A.
,
Quintana
,
A.
,
Ding
,
Y.
,
Ma
,
G.
et al. 
(
2015
)
Proteomic mapping of ER-PM junctions identifies STIMATE as a regulator of Ca2+ influx
.
Nat. Cell Biol.
17
,
1339
1347
doi:
144
Giordano
,
F.
,
Saheki
,
Y.
,
Idevall-Hagren
,
O.
,
Colombo
,
S.F.
,
Pirruccello
,
M.
,
Milosevic
,
I.
et al. 
(
2013
)
PI(4,5)P2-dependent and Ca2+-regulated ER-PM interactions mediated by the extended synaptotagmins
.
Cell
153
,
1494
1509
doi:
145
Kopec
,
K.O.
,
Alva
,
V.
and
Lupas
,
A.N.
(
2010
)
Homology of SMP domains to the TULIP superfamily of lipid-binding proteins provides a structural basis for lipid exchange between ER and mitochondria
.
Bioinformatics
26
,
1927
1931
doi:
146
Alva
,
V.
and
Lupas
,
A.N.
(
2016
)
The TULIP superfamily of eukaryotic lipid-binding proteins as a mediator of lipid sensing and transport
.
Biochim. Biophys. Acta Mol. Cell Biol. Lipids
1861
,
913
923
doi:
147
Schauder
,
C.M.
,
Wu
,
X.
,
Saheki
,
Y.
,
Narayanaswamy
,
P.
,
Torta
,
F.
,
Wenk
,
M.R.
et al. 
(
2014
)
Structure of a lipid-bound extended synaptotagmin indicates a role in lipid transfer
.
Nature
510
,
552
555
doi:
148
Reinisch
,
K.M.
and
De Camilli
,
P.
(
2016
)
SMP-domain proteins at membrane contact sites: structure and function
.
Biochim. Biophys. Acta Mol. Cell Biol. Lipids
1861
,
924
927
doi:
149
Chang
,
C.-L.
,
Hsieh
,
T.-S.
,
Yang
,
T.T.
,
Rothberg
,
K.G.
,
Azizoglu
,
D.B.
,
Volk
,
E.
et al. 
(
2013
)
Feedback regulation of receptor-induced Ca2+ signaling mediated by E-Syt1 and Nir2 at endoplasmic reticulum-plasma membrane junctions
.
Cell Rep.
5
,
813
825
doi:
150
Saheki
,
Y.
,
Bian
,
X.
,
Schauder
,
C.M.
,
Sawaki
,
Y.
,
Surma
,
M.A.
,
Klose
,
C.
et al. 
(
2016
)
Control of plasma membrane lipid homeostasis by the extended synaptotagmins
.
Nat. Cell Biol.
18
,
504
515
doi:
151
Yu
,
H.
,
Liu
,
Y.
,
Gulbranson
,
D.R.
,
Paine
,
A.
,
Rathore
,
S.S.
and
Shen
,
J.
(
2016
)
Extended synaptotagmins are Ca2+-dependent lipid transfer proteins at membrane contact sites
.
Proc. Natl Acad. Sci. USA
113
,
4362
4367
doi:
152
Shadan
,
S.
,
Holic
,
R.
,
Carvou
,
N.
,
Ee
,
P.
,
Li
,
M.
,
Murray-Rust
,
J.
et al. 
(
2008
)
Dynamics of lipid transfer by phosphatidylinositol transfer proteins in cells
.
Traffic
9
,
1743
1756
doi:
153
Cockcroft
,
S.
and
Garner
,
K.
(
2013
)
Potential role for phosphatidylinositol transfer protein (PITP) family in lipid transfer during phospholipase C signalling
.
Adv. Biol. Regul.
53
,
280
291
doi:
154
Cockcroft
,
S.
(
2012
)
The diverse functions of phosphatidylinositol transfer proteins
.
Curr. Top. Microbiol. Immunol.
362
,
185
208
doi:
155
Allen-Baume
,
V.
,
Ségui
,
B.
and
Cockcroft
,
S.
(
2002
)
Current thoughts on the phosphatidylinositol transfer protein family
.
FEBS Lett.
531
,
74
80
doi:
156
Cockcroft
,
S.
,
Garner
,
K.
,
Yadav
,
S.
,
Gomez-Espinoza
,
E.
and
Raghu
,
P.
(
2016
)
Rdgbα reciprocally transfers PA and PI at ER–PM contact sites to maintain PI(4,5)P2 homoeostasis during phospholipase C signalling in Drosophila photoreceptors
.
Biochem. Soc. Trans.
44
,
286
292
doi:
157
Vihtelic
,
T.S.
,
Goebl
,
M.
,
Milligan
,
S.
,
O'Tousa
,
S.E.
and
Hyde
,
D.R.
(
1993
)
Localization of Drosophila retinal degeneration B, a membrane-associated phosphatidylinositol transfer protein
.
J. Cell Biol.
122
,
1013
1022
doi:
158
Lev
,
S.
,
Hernandez
,
J.
,
Martinez
,
R.
,
Chen
,
A.
,
Plowman
,
G.
and
Schlessinger
,
J.
(
1999
)
Identification of a novel family of targets of PYK2 related to Drosophila retinal degeneration B (rdgB) protein
.
Mol. Cell. Biol.
19
,
2278
2288
doi:
159
Tilley
,
S.J.
,
Skippen
,
A.
,
Murray-Rust
,
J.
,
Swigart
,
P.
,
Stewart
,
A.
,
Morgan
,
C.P.
et al. 
(
2004
)
Structure-function analysis of phosphatidylinositol transfer protein alpha bound to human phosphatidylinositol
.
Structure
12
,
317
326
doi:
160
Yoder
,
M.D.
,
Thomas
,
L.M.
,
Tremblay
,
J.M.
,
Oliver
,
R.L.
,
Yarbrough
,
L.R.
and
Helmkamp
, Jr,
G.M.
(
2001
)
Structure of a multifunctional protein. Mammalian phosphatidylinositol transfer protein complexed with phosphatidylcholine
.
J. Biol. Chem.
276
,
9246
9252
doi:
161
Morgan
,
C.P.
,
Skippen
,
A.
,
Segui
,
B.
,
Ball
,
A.
,
Allen-Baume
,
V.
,
Larijani
,
B.
et al. 
(
2004
)
Phosphorylation of a distinct structural form of phosphatidylinositol transfer protein α at Ser166 by protein kinase C disrupts receptor-mediated phospholipase C signaling by inhibiting delivery of phosphatidylinositol to membranes
.
J. Biol. Chem.
279
,
47159
47171
doi:
162
Kim
,
Y.J.
,
Guzman-Hernandez
,
M.L.
,
Wisniewski
,
E.
and
Balla
,
T.
(
2015
)
Phosphatidylinositol-phosphatidic acid exchange by Nir2 at ER-PM contact sites maintains phosphoinositide signaling competence
.
Dev. Cell
33
,
549
561
doi:
163
Chang
,
J.T.
,
Milligan
,
S.
,
Li
,
Y.
,
Chew
,
C.E.
,
Wiggs
,
J.
,
Copeland
,
N.G.
et al. 
(
1997
)
Mammalian homolog of Drososphila retinal degeneration B rescues the mutant fly phenotype
.
J. Neurosci.
17
,
5881
5890
PMID:
[PubMed]
164
Lu
,
C.
,
Vihtelic
,
T.S.
,
Hyde
,
D.R.
and
Li
,
T.
(
1999
)
A neuronal-specific mammalian homolog of the Drosophila retinal degeneration B gene with expression restricted to the retina and dentate gyrus
.
J. Neurosci.
19
,
7317
7325
PMID:
[PubMed]
165
Elagin
,
V.A.
,
Elagina
,
R.B.
,
Doro
,
C.J.
,
Vihtelic
,
T.S.
and
Hyde
,
D.R.
(
2000
)
Cloning and tissue localization of a novel zebrafish RdgB homolog that lacks a phospholipid transfer domain
.
Vis. Neurosci.
17
,
303
311
doi:
166
Litvak
,
V.
,
Shaul
,
Y.D.
,
Shulewitz
,
M.
,
Amarilio
,
R.
,
Carmon
,
S.
and
Lev
,
S.
(
2002
)
Targeting of Nir2 to lipid droplets is regulated by a specific threonine residue within its PI-transfer domain
.
Curr. Biol.
12
,
1513
1518
doi:
167
Chen
,
J.
,
Yao
,
Y.
,
Gong
,
C.
,
Yu
,
F.
,
Su
,
S.
,
Chen
,
J.
et al. 
(
2011
)
CCL18 from tumor-associated macrophages promotes breast cancer metastasis via PITPNM3
.
Cancer Cell
19
,
541
555
doi:
168
Bonecchi
,
R.
,
Locati
,
M.
and
Mantovani
,
A.
(
2011
)
Chemokines and cancer: a fatal attraction
.
Cancer Cell
19
,
434
435
doi:
169
Zhang
,
B.
,
Yin
,
C.
,
Li
,
H.
,
Shi
,
L.
,
Liu
,
N.
,
Sun
,
Y.
et al. 
(
2013
)
Nir1 promotes invasion of breast cancer cells by binding to chemokine (C–C motif) ligand 18 through the PI3K/Akt/GSK3β/Snail signalling pathway
.
Eur. J. Cancer
49
,
3900
3913
doi:
170
Krohn
,
S.
,
Garin
,
A.
,
Gabay
,
C.
and
Proudfoot
,
A.E.
(
2013
)
The activity of CCL18 is principally mediated through interaction with glycosaminoglycans
.
Front. Immunol.
4
,
193
doi:
171
Osman
,
S.
,
Taylor
,
K.A.
,
Allcock
,
N.
,
Rainbow
,
R.D.
and
Mahaut-Smith
,
M.P.
(
2016
)
Detachment of surface membrane invagination systems by cationic amphiphilic drugs
.
Sci. Rep.
6
,
18536
doi:
172
Sato
,
M.
,
Liu
,
K.
,
Sasaki
,
S.
,
Kunii
,
N.
,
Sakai
,
H.
,
Mizuno
,
H.
et al. 
(
2013
)
Evaluations of the selectivities of the diacylglycerol kinase inhibitors R59022 and R59949 among diacylglycerol kinase isozymes using a new non-radioactive assay method
.
Pharmacology
92
,
99
107
doi:
173
Kim
,
S.
,
Kedan
,
A.
,
Marom
,
M.
,
Gavert
,
N.
,
Keinan
,
O.
,
Selitrennik
,
M.
et al. 
(
2013
)
The phosphatidylinositol-transfer protein Nir2 binds phosphatidic acid and positively regulates phosphoinositide signalling
.
EMBO Rep.
14
,
891
899
doi: