Abstract

The phosphoinositide (PIPn) family of signalling phospholipids are central regulators in membrane cell biology. Their varied functions are based on the phosphorylation pattern of their inositol ring, which can be recognized by selective binding domains in their effector proteins and be modified by a series of specific PIPn kinases and phosphatases, which control their interconversion in a spatial and temporal manner. Yet, a unique feature of PIPns remains largely unexplored: their unusually uniform acyl chain composition. Indeed, while most phospholipids present a range of molecular species comprising acyl chains of diverse length and saturation, PIPns in several organisms and tissues show the predominance of a single hydrophobic backbone, which in mammals is composed of arachidonoyl and stearoyl chains. Despite evolution having favoured this specific PIPn configuration, little is known regarding the mechanisms and functions behind it. In this review, we explore the metabolic pathways that could control the acyl chain composition of PIPns as well as the potential roles of this selective enrichment. While our understanding of this phenomenon has been constrained largely by the technical limitations in the methods traditionally employed in the PIPn field, we believe that the latest developments in PIPn analysis should shed light onto this old question.

Introduction

Phosphoinositides (PIPn) are a family of eight phospholipids (PI, PI3P, PI4P, PI5P, PI(4,5)P2, PI(3,4)P2, PI(3,5)P2, PI(3,4,5)P3), whose members define membrane identity, drive compartment-specific accumulation of other lipids and transduce extracellular signals [1]. They are interconverted by a series of highly regulated PIPn kinases and phosphatases and a key principle by which they act in cells is the specific recognition of their phosphorylated head-groups by enzymes and effector proteins [2] (Figure 1). Where known, this recognition is mediated primarily through electrostatic interactions between inositol ring phosphates and amino acid side chains in the conserved pockets of target proteins with less clear specific interactions with the hydrophobic ‘backbone’. Nonetheless, an under-appreciated characteristic of PIPn in mammalian cells is that they often display a relatively molecularly homogeneous acyl chain composition, the mechanistic significance of which is unclear.

Phosphoinositide structures, metabolism and interactions with effectors.

Figure 1.
Phosphoinositide structures, metabolism and interactions with effectors.

(A) A typical phosphoinositide (PIPn) structure is represented by C38:4 PI3P (sn-1-stearoyl-2-arachidonoyl phosphatidylinositol 3-phosphate). The fatty acids acylated in the sn-1 and sn-2 positions of the glycerol backbone (green) are labelled with both their common and systematic name, and with the lipid nomenclature based on the number of carbons and double bonds in brackets. The three positions in the inositol ring which can be phosphorylated or dephosphorylated by specific kinases and phosphatases to give rise to the different members of the PIPn family are indicated by red numbers. The pink cartoon engulfing the phosphorylated ring represents a specific interaction between the phosphorylated inositol ring and the binding domain of a PIPn effector protein. (B) Pathways of PIPn metabolism indicating their dynamic interconversion by kinase and phosphatase reactions acting on different positions on their inositol rings. Disputed activities are indicated with dashed arrows. (C) List of PIPn molecular species frequently identified by MS analysis. When their acyl chain composition is not resolved by further fragmentation, they are expressed by the total number of carbons and double bonds in both chains. (D) MS profile of the PIP2 molecular species found in mouse brain and mouse embryonic fibroblasts (MEFs) (adapted from [123]). The three PIP2 isomers are not resolved by this method, though PI(4,5)P2 generally accounts for most of the signal. While brain PIP2 shows the predominance of C38:4 species characteristic of mammalian PI, this enrichment can be lost in cultured cells such as MEFs, which present an heterogeneous distribution of molecular species in all PIPn.

Figure 1.
Phosphoinositide structures, metabolism and interactions with effectors.

(A) A typical phosphoinositide (PIPn) structure is represented by C38:4 PI3P (sn-1-stearoyl-2-arachidonoyl phosphatidylinositol 3-phosphate). The fatty acids acylated in the sn-1 and sn-2 positions of the glycerol backbone (green) are labelled with both their common and systematic name, and with the lipid nomenclature based on the number of carbons and double bonds in brackets. The three positions in the inositol ring which can be phosphorylated or dephosphorylated by specific kinases and phosphatases to give rise to the different members of the PIPn family are indicated by red numbers. The pink cartoon engulfing the phosphorylated ring represents a specific interaction between the phosphorylated inositol ring and the binding domain of a PIPn effector protein. (B) Pathways of PIPn metabolism indicating their dynamic interconversion by kinase and phosphatase reactions acting on different positions on their inositol rings. Disputed activities are indicated with dashed arrows. (C) List of PIPn molecular species frequently identified by MS analysis. When their acyl chain composition is not resolved by further fragmentation, they are expressed by the total number of carbons and double bonds in both chains. (D) MS profile of the PIP2 molecular species found in mouse brain and mouse embryonic fibroblasts (MEFs) (adapted from [123]). The three PIP2 isomers are not resolved by this method, though PI(4,5)P2 generally accounts for most of the signal. While brain PIP2 shows the predominance of C38:4 species characteristic of mammalian PI, this enrichment can be lost in cultured cells such as MEFs, which present an heterogeneous distribution of molecular species in all PIPn.

PIPn in mammalian tissues is enriched in the C38:4 molecular species

Glycerophospholipid classes are defined by their polar head-groups, which are bound to a diacylglycerol (DG) backbone through a diester phosphate. Each of these classes comprise a range of molecular species, which share the same head-group, but present different pairs of aliphatic chains (Figure 1). These chains can be linked via ester or ether bonds and derived from fatty acids of different length and degree of saturation [3]. Most phospholipid classes present a wide variety of alkyl/acyl compositions and only occasionally is the function of a particular molecular species known e.g. ‘rigid’ C32:0-PC dominates lung surfactant [4]. However, the acyl chain pairings of mammalian PI are unusually enriched (≥70%) in a single molecular species, composed of a stearoyl chain in sn-1 (18 carbons and no double bonds; C18:0) and an arachidonoyl chain in sn-2 (20 carbons and 4 double bonds; C20:4) (Figure 1) [57]. As both chains are simultaneously measured in most routine, mass spectrometry (MS)-based analyses, this molecular species is usually referred to as ‘C38:4’ PI. Early analyses focussed only on PI, but recent MS studies have shown that the acyl compositions of PI, PIP, PIP2 and PIP3 are very similar (current MS methods do not discriminate between regio-isomers), consistent with rapid interconversion between the various phosphorylated forms of these lipids and parental PI (Figure 1) [8,9]. These observations suggest evolution has favoured the selective enrichment of this species of PIPn in mammals, but we know very little about the advantages provided by this acyl chain species and the molecular mechanisms responsible for creating it.

How is the acyl composition of mammalian PIPn determined?

In principle, the C38:4 species of PI could be selectively synthesized de novo and/or enriched by subsequent acyl chain ‘remodelling’. The pathways for de novo synthesis of PI and acyl chain remodelling by the ‘Lands Cycle’ (Figure 2) have been discussed recently in an excellent review [10]; these reactions are envisaged to occur in the ER or membrane domains closely associated with it. Attempts to understand the major points of C38:4 enrichment have employed measurements of the acyl chain compositions of key intermediates, radioactive tracer studies and interrogation of the acyl chain selectivity of relevant enzymes in vitro. However, due to both technical and interpretational difficulties, the picture is still far from clear.

Synthesis of phosphatidylinositol and acyl chain remodelling.

Figure 2.
Synthesis of phosphatidylinositol and acyl chain remodelling.

De novo synthesis of PI starts with the formation of the simple phospholipid PA in the ER. At this point, glycerophospholipid synthesis diverges into two branches: the CDP-DG pathway for the synthesis of PI and the DG pathway for the other ER lipids. CDP-DG is also the precursor of CL but its synthesis takes place in the mitochondria using a different CDP-DG synthase (TAMM41). A selective channelling of C38:4 PA molecules towards the CDP-DG branch might contribute to the enrichment of this species in PI. Once synthesized, PI molecules can be remodelled into C38:4 PI by the Lands cycle, which involves the release of one of their acyl chains and the re-acylation of the resulting LPI via acyltransferases selectively incorporating C18:0 and C20:4 chains in sn-1 and sn-2, respectively. Abbreviations: ER, endoplasmic reticulum; G-3-P, glycerol-3-phosphate; LPA, lysophosphatidic acid; PA, phosphatidic acid; DG, diacylglycerol; PC, phosphatidylcholine; PE, phosphatidylethanolamine; PS, phosphatidylserine; TG, triacylglycerol; PGP, phosphatidylglycerol phosphate; CL, cardiolipin; CDP-DG, CDP-diacylglycerol; PI, phosphatidylinositol; LPI, lysophosphoinositol; FFA, free fatty acid; CTP, cytidine triphosphate; PPi, pyrophosphate; GPAT, glycerol-3-phosphate acyltransferase; AGPAT, acylglycerol-3-phosphate acyltransferase; PAP, phosphatidate phosphatase; DGK, diacylglycerol kinase; CDS, CDP-diacylglycerol synthase; TAMM41, TAM41 mitochondrial translocator assembly and maintenance homologue; PIS, phosphatidylinositol synthase; PLA1/2, phospholipases A1/A2; AGPAT8, 2-acylglycerol-3-phosphate acyltransferase 8; MBOAT7, membrane bound O-acyltransferase 7.

Figure 2.
Synthesis of phosphatidylinositol and acyl chain remodelling.

De novo synthesis of PI starts with the formation of the simple phospholipid PA in the ER. At this point, glycerophospholipid synthesis diverges into two branches: the CDP-DG pathway for the synthesis of PI and the DG pathway for the other ER lipids. CDP-DG is also the precursor of CL but its synthesis takes place in the mitochondria using a different CDP-DG synthase (TAMM41). A selective channelling of C38:4 PA molecules towards the CDP-DG branch might contribute to the enrichment of this species in PI. Once synthesized, PI molecules can be remodelled into C38:4 PI by the Lands cycle, which involves the release of one of their acyl chains and the re-acylation of the resulting LPI via acyltransferases selectively incorporating C18:0 and C20:4 chains in sn-1 and sn-2, respectively. Abbreviations: ER, endoplasmic reticulum; G-3-P, glycerol-3-phosphate; LPA, lysophosphatidic acid; PA, phosphatidic acid; DG, diacylglycerol; PC, phosphatidylcholine; PE, phosphatidylethanolamine; PS, phosphatidylserine; TG, triacylglycerol; PGP, phosphatidylglycerol phosphate; CL, cardiolipin; CDP-DG, CDP-diacylglycerol; PI, phosphatidylinositol; LPI, lysophosphoinositol; FFA, free fatty acid; CTP, cytidine triphosphate; PPi, pyrophosphate; GPAT, glycerol-3-phosphate acyltransferase; AGPAT, acylglycerol-3-phosphate acyltransferase; PAP, phosphatidate phosphatase; DGK, diacylglycerol kinase; CDS, CDP-diacylglycerol synthase; TAMM41, TAM41 mitochondrial translocator assembly and maintenance homologue; PIS, phosphatidylinositol synthase; PLA1/2, phospholipases A1/A2; AGPAT8, 2-acylglycerol-3-phosphate acyltransferase 8; MBOAT7, membrane bound O-acyltransferase 7.

C38:4 enrichment by de novo synthesis

Phospholipid synthesis branches in the ER from the common precursor PA by the action of PA phosphatases, generating DG for PC/PE/PS/TG synthesis or, the action of CDS1/2 generating CDP-DG for PI synthesis [10] (Figure 2). PGP/PG/CL synthesis in mitochondria also uses a CDP-DG intermediate but this seems to be mostly synthesized in situ by TAMM41, a CDP-DG synthase that evolved independently of the ER enzymes, CDS1/2 [11,12]. As most CDP-DG in the ER will be committed to PI synthesis, a possible enrichment mechanism could be the selective channelling of C38:4 PA towards the CDP-DG branch and non-C38:4 PAs towards the DG branch (Figure 2). Indeed, the activation of PA phosphatases in ER domains proximal to the nuclear envelope represses PI synthesis in this region [13], but it is not known if this process is selective for certain molecular species. Most studies have reported the presence of multiple molecular species of PA in cells [14], but there are very few measurements of the acyl chain composition of CDP-DG (mainly due to its chemical and biological instability) [15]. Furthermore, given CDP-DG is used for both PI and PGP synthesis, ER and mitochondrial pools would need to be distinguished to obtain unambiguous evidence of C38:4 enrichment in the CDP-DG destined for PI synthesis. Some evidence suggesting selective channelling of PA species towards CDP-DG comes from investigations of the acyl chain selectivity of CDS1 and 2 in vitro. These studies suggest CDS2 has a significant preference for C20:4 in the sn-2 position of PA, but CDS1 is less selective [16], although it prefers unsaturated species of PA compared with fully saturated species [17]. CDS2 is ubiquitously expressed, while CDS1 is restricted to a subset of tissues [16,18] and might be induced when a high rate of de novo PI synthesis is required [19]. Furthermore, overexpressed CDS1 and CDS2 show slightly different patterns of localization in the ER [17]. These observations suggest that the formation of CDP-DG from PA is likely to be a regulated step and CDS1 and 2 have evolved to play differing roles, although definitive evidence for their involvement in establishing the acyl chain composition of CDP-DGs is still lacking.

There is only one PIS activity in mammalian cells [10]. This enzyme does not possess relevant acyl chain selectivity in vitro [20] and there are several examples where mammalian cells can synthesise a range of different PI species in culture (see below). This suggests PIS will use any CDP-DGs available to it and is very unlikely to select C38:4 CDP-DG for PI synthesis.

Early metabolic labelling studies comparing the incorporation rate of 14C-glycerol, 14C-glucose, and 32P-Pi showed that the initial stages of de novo PI synthesis produce multiple PI species, which are then followed by a progressive enrichment of the C38:4 species [2124]. This is the best evidence to date that the de novo synthesis pathway does not selectively produce C38:4 PI, but that this species is enriched by subsequent steps. It is therefore surprising that, apart from one preliminary report [25], these pioneering studies have not yet been confirmed and extended in any detail with modern cell culture and MS methods.

C38:4 enrichment by acyl chain remodelling

The foregoing discussion suggests at least some, if not all, of the C38:4 enrichment in PIs is achieved by acyl chain remodelling of multiple PIs synthesized de novo. The primary mechanism for engineering new acyl chain variants of pre-existing phospholipids is the Lands Cycle. In this cycle, a phospholipase A (PLA) hydrolyzes an acyl chain from the sn-1 or sn-2 position and the resulting lysophospholipid is reacylated with a new acyl chain by a lysophospholipid acyltransferase (Figure 2). In principle, the process of acyl chain enrichment could be governed by the acyl chain selectivity of phospholipases A1/2 or the acyl-CoA transferases, together with the availability of different acyl-CoAs. Thus far, the clearest examples of selectivity are found amongst the large family of acyl-CoA transferases, some of which have strong preferences for a particular combination of lyso-lipid and acyl-CoA [26].

MBOAT7 (Membrane-bound O-acyltransferase 7) can selectively reacylate LPI in sn-2 with polyunsaturated substrates, such as arachidonoyl-coA [27,28], while AGPAT8 (1-acylglycerol-3-phosphate acyltransferase 8), can reacylate LPI in sn-1 using stearoyl-coA [29]. Despite initial studies suggesting that AGPAT8 can acylate lysocardiolipin and other lysophospholipids in vitro [3032], the main effect of AGPAT8 deletion in cells and animals is a decrease in the stearoyl content of PIPn [7,29,33], suggesting that its main function is the sn-1 remodelling of PI. The extent to which MBOAT7 and AGPAT8 contribute to C38:4 enrichment in PI by selectively incorporating arachidonoyl and stearoyl chains in sn-2 and sn-1, respectively, is shown by the reduction in C38:4 PIPn species in both MBOAT7−/− [5,6] and AGPAT8−/− [7] mice. In both cases, the reduced frequency of C38:4 species was accompanied by a decrease in PI levels, possibly due to the impairment of LPI reacylation, though it could also indicate that the C38:4 backbone is involved in the maintenance of the total PIPn pool. Nevertheless, despite the relative increase in other species, C38:4 was still predominant in both KO models, indicating that other mechanisms must also contribute to its enrichment.

AGPAT8−/− mice appeared healthy, despite the decrease in PI relative to other lipids and a reduced proportion of C18:0 in PIPn [7]. However, its transient silencing in ARPE-19 cells, derived from human retinal pigment epithelium, which also reduced the C18:0 content in PIPn, selectively altered the levels and localization of PI3P and PI(4,5)P2, hence affecting membrane trafficking dependent on these lipids [33]. The mechanisms behind these alterations remain unknown, though it has been suggested that the substitution of the C18:0 chain in sn-1 for the shorter C16:0 could affect PIPn interaction with proteins sensitive to the orientation or distance of the polar head with respect to the membrane surface [34].

MBOAT7−/− mice displayed neonatal lethality and aberrant brain development, but it is not clear whether these anomalies are caused by the accumulation of LPI and a decrease in the total levels of PIPn rather than the modest changes in their distribution of molecular species [5,6]. The rs641738 C > T variant in the locus containing MBOAT7 has been associated with liver diseases such as alcohol-related cirrhosis [35], inflammation and fibrosis in chronic hepatitis B and C [36,37] and non-alcoholic fatty liver disease, though this later association has been disputed [38]. This variant is associated with lower protein levels of MBOAT7 in liver and changes in the levels of minor PI species in plasma [39] and liver [40]. Nevertheless, it seems likely that this pro-inflammatory phenotype may be related to defective processing of C20:4 by MBOAT7 rather than small changes in PI species.

The first committed step in PI remodelling is its hydrolysis into LPI by PLAs, and thus coupling this process to de novo PI synthesis might be envisaged to facilitate remodelling. Indeed, it has been suggested that the faster remodelling of atypical PE and PS species produced de novo involves their selective degradation by PLAs [41]. Similarly, PIs generated from exogenous C32:0 CDP-DG in liver microsomes were rapidly hydrolyzed into LPI, while little LPI was formed from pre-existing PIs, suggesting a selective de-acylation of newly synthesized PIs [42]. As yet, no PI-PLA activity selective for these intermediate species has been identified and the roles of the PLA superfamily of enzymes in phospholipid remodelling are not clearly defined [43]. The inhibition of Ca2+-independent phospholipase A2b (iPLA2b) reduces the incorporation of C20:4 into phospholipids [44,45] while its overexpression increases the levels of C38:4 PI [46]. The analysis of brains from iPLA2b−/− mice, however, suggest that it is not required for C38:4 enrichment [47]. DDHD1 seems to be the PLA1 involved in the generation of C20:4 LPI [48], an endogenous activating ligand of G protein-coupled receptor GPR55 [49,50]. Although loss of the homologue of DDHD1 in C. elegans, IPLA-1, results in a dramatic change in their predominant PI species, from C18:0/20:5 to 18:1/20:5 [29], in DDHD1−/− mice there were no changes in C38:4 PIPn whilst there was a relative increase in 18:1/C20:4 PIPn at the expense of shorter and more saturated species [51]. Regarding the products of PLA activity on PIs, C18:0 LPI shows the highest levels in mouse tissues, with C20:4 LPI also being abundant in the nervous system [52]. However, this provides little information about the kinetics of remodelling in sn-1 and sn-2 as these species can be formed as remodelling intermediates but also from C38:4 PI for signalling purposes. Thus, if it is not reacylated, LPI can be metabolized by lysophospholipases A, C or D to generate glycerophosphoinositol [53], monoacylglycerol [54] or LPA [55], which will all have different signalling roles [56].

In addition to the acyl-coA acyltransferases, PI acyl chains might also be remodelled by poorly characterized transacylation reactions, in which an acyl chain is transferred directly from a phospholipid donor to a lysophospholipid [57]. This may be an important reaction for CL remodelling, as evidenced by the mitochondrial dysfunction caused by defects in the CL transacylase Tafazzin [58], but the potential for this type of reaction in PI remodelling remains unknown.

Selective remodelling of de novo PI species could also be facilitated by a spatial confinement of the enzymes involved in PI synthesis and remodelling. Regarding this possibility, a growing body of evidence based on the localization of overexpressed PIS indicates that PI synthesis occurs in discrete locations in the ER [5962]. PI synthesis and remodelling could be coupled in these domains, as indicated by the co-localization AGPAT8 with PIS [33].

In summary, we still do not fully understand the process of acyl chain enrichment for the C38:4 species of PI in mammalian cells. It remains possible that some selectivity occurs through the actions of CDS2 to supply C38:4-enriched CDP-DG to PIS, but most evidence points to a significant role for acyl chain remodelling of PIs synthesized de novo by the LPI acyltransferases MBOAT7 and AGPAT8, and possibly the PLA1 DDHD1. However, the deletion of these acyl transferases in mice does not affect the relative levels of C38:4 in PIPn to the same extent as it does in C. elegans, suggesting redundant mechanisms are co-ordinated in mammals to ensure a robust C38:4 enrichment. Further studies are clearly needed to establish the mechanisms controlling this efficient remodelling of PIs, including its compartimentalization and its coordination with PI synthesis.

What are the selective properties of the acyl chain composition of PIP(n)?

The important properties conveyed by C38:4 enrichment in mammalian cells are still unclear. Plausible arguments can be made in favour of a creating a distinct molecular label to distinguish the PIPn backbone from other lipids, a biophysical advantage to the mechanisms by which PIPn interact with their protein effectors or, a regulated supply of arachidonic acid for the synthesis of bioactive lipids.

Does C38:4 enrichment create an efficient ‘PI’ cycle?

The field of PIPn signalling originated from the discovery of a hormone-stimulated increase in the metabolism of PI relative to other phospholipids. Some of these early studies described a selective increase in the incorporation of 32P-Pi or 3H-inositol compared with 14C-glycerol, which was interpreted to mean a relatively ‘closed’ PI cycle may exist in which a stimulated PLC produces DG and PA that are selectively recycled back into PI (Figure 3; [63]). The activation of a PI(4,5)P2-selective PLC by receptors is now accepted to be a central signal transduction mechanism in multicellular organisms and liberates the ‘second messengers’ IP3 and DG, which stimulate intracellular Ca2+-release and the activation of PKC, respectively [64,65]. Whilst PIPn can be interconverted by kinases and phosphomonoesterases, the action of a PLC produces DG, which is a common biosynthetic intermediate in lipid metabolism. This creates a potential problem for cells in segregating PLC-derived-DG from other sources of DG, enabling it to both act as a selective signal for the activation of PKC and also as a source of PA for the PI-cycle. Furthermore, PLC activation is often observed to occur in parallel with activation of a PLD which usually appears to be directed predominantly against PC, adding a further source of complexity in segregating pools of PA (and DG via PA phosphatases) destined for different purposes [66].

The signal PLC transduction pathway is coupled to PI re-synthesis to maintain the PIPn pool.

Figure 3.
The signal PLC transduction pathway is coupled to PI re-synthesis to maintain the PIPn pool.

The activation of PLC (phospholipase C) in response to GPCR agonists (G protein-coupled receptors) induces the cleavage of PI(4,5)P2, generating DG (diacylglycerol) and IP3 (inositol 1,4,5-trisphosphate), which leads to the release of Ca2+ from intracellular stores and the activation of PKC (protein kinase C), respectively. The drop in the total PIPn pool caused by PI(4,5)P2 hydrolysis is compensated by the activation of PI synthesis. It has been proposed that this PI re-synthesis involves a closed ‘PI cycle’ (red arrows), where the DG generated by the hydrolysis of PI(4,5)P2 is selectively recycled back into PI to maintain the total PIPn pool. This process involves the transport of the generated DG and/or PA from the PM (plasma membrane) to the ER (endoplasmic reticulum), where it can be utilized by PI synthesis enzymes CDS and PIS. This cycle could be spatially confined to PM-ER contact sites, where LTPs (lipid transfer proteins) transport the lipid intermediates between membranes (dotted lines). The C38:4 backbone could facilitate the PI cycle by providing molecular identity with the DG and PA molecules, preventing their consumption by other biosynthetic pathways as well as the utilization of intermediates from other origins, such as PA generated from PC via PLD (phospholipase D).

Figure 3.
The signal PLC transduction pathway is coupled to PI re-synthesis to maintain the PIPn pool.

The activation of PLC (phospholipase C) in response to GPCR agonists (G protein-coupled receptors) induces the cleavage of PI(4,5)P2, generating DG (diacylglycerol) and IP3 (inositol 1,4,5-trisphosphate), which leads to the release of Ca2+ from intracellular stores and the activation of PKC (protein kinase C), respectively. The drop in the total PIPn pool caused by PI(4,5)P2 hydrolysis is compensated by the activation of PI synthesis. It has been proposed that this PI re-synthesis involves a closed ‘PI cycle’ (red arrows), where the DG generated by the hydrolysis of PI(4,5)P2 is selectively recycled back into PI to maintain the total PIPn pool. This process involves the transport of the generated DG and/or PA from the PM (plasma membrane) to the ER (endoplasmic reticulum), where it can be utilized by PI synthesis enzymes CDS and PIS. This cycle could be spatially confined to PM-ER contact sites, where LTPs (lipid transfer proteins) transport the lipid intermediates between membranes (dotted lines). The C38:4 backbone could facilitate the PI cycle by providing molecular identity with the DG and PA molecules, preventing their consumption by other biosynthetic pathways as well as the utilization of intermediates from other origins, such as PA generated from PC via PLD (phospholipase D).

Recent studies have suggested an important element in the organisation of a ‘PI cycle’ that may allow segregation from other biosynthetic pathways is the close and discrete apposition of contact sites between the ER and PM (Figure 3) [6769]. These ER-PM contact sites offer an opportunity for lipid transfer proteins (LTPs) to locally exchange PLC-derived DG and PA produced in the PM, with newly synthesized PI produced in the ER, offering the potential for close coupling of PI-resynthesis and -consumption and the possibility of a diffusion barrier with bulk lipid synthesis in more distant regions of the ER [7073]. There are still many details missing however, as to the precise composition of different types of ER-PM contact sites and how effectively they are isolated from other regions of the ER.

It seems conceivable that the spatial segregation offered by ER-PM contact sites might be supplemented in mammals by molecular recognition of the C38:4 backbone. Thus, the common C38:4 signature in intermediates generated from PI(4,5)P2 hydrolysis could influence their selective recycling back into PI. A preference for C38:4 substrates has been observed with CDS2 [17] and other PI cycle enzymes, such as DGKε [74] and PI4P5Ks [75], and an interesting idea has been proposed that the PI cycle could act as a ‘molecular sieve’, progressively enriching the C38:4 species through multiple passages around the cycle [76]. However, the relevance of this selectivity for C38:4 substrates is still unclear. DGKε−/− mice do not present major anomalies, but they do display decreased levels of C20:4-containing PIP2 in the brain and an apparent attenuation of PLC signalling induced by electroconvulsive shock [77]. PLC signalling was also affected by the deletion of MBOAT7 in the RAW264.7 macrophage cell line, which caused a significant reduction in the % of C38:4 in PIPn that was accompanied by prolonged Ca2+ oscillations upon stimulation, though it is not clear if the apparent reduction in Ca2+ efflux could be linked to a defective PI cycle [78].

Does C38:4 enrichment affect interactions with effectors?

Many of the well-established interactions between PIPn and their effector proteins take place through the recognition of phosphorylated inositol head-groups by conserved domains in soluble, or extrinsic membrane proteins [2]. These interactions can induce a conformational change in the target protein and/or promote a significant change in its cellular location, causing an increase in effective concentration at the membrane where the particular PIPn resides [2]. Coupled with tight control of the synthesis and degradation of individual PIPn, these types of interactions allow PIPns to regulate compartment-specific events, for example the synthesis of PI(3,4,5)P3 at the plasma membrane regulates signal transduction through PH-domain containing effectors in response to growth factors, or synthesis of PI3P in early endosomes regulates protein trafficking through PX and FYVE domain-containing effectors. Several structures are now available for these complexes, which suggest the key interactions are electrostatic in nature and are confined to the inositol ring (an example of a PX domain bound to PI3P is shown in Figure 4A). However, many of these structures were obtained from crystals containing only a soluble inositol phosphate head-group or short-chain versions of PIPn, precluding the discovery of important interactions with the acyl chains. Furthermore, when specifically interrogated in biochemical assays, some of these interactions have been shown to be more efficient with C38:4 PIPn e.g. in the PDK1-mediated phosphorylation of PKB stimulated by PIP3 [79].

Phosphoinositide–protein interaction modes.

Figure 4.
Phosphoinositide–protein interaction modes.

The influence of the acyl chains in PIPn–protein interactions is often missed in structural studies by the utilization of soluble, short-chain versions of PIPn, the conformational dynamics of longer chains and the lack of a membrane context. (AE) Examples of protein–PIPn interactions adapted from the Protein Data Bank, where the protein backbone is shown interacting with PIPn of diverse acyl chain composition (in spacefill representation: carbon, grey; phosphorus, orange; oxygen, red). The original lipid ligand is indicated, although shorter versions are represented in the structures of PITPα (B) and GIRK2 (C) where the PIPn acyl chains were not fully solved. Although the interaction with PIPn-binding domains is often restricted to the phosphorylated inositol ring (A), its presentation relative to the membrane surface could be affected by the acyl chain length/saturation. The PIPn acyl chains can directly interact with hydrophobic residues inserted in the membrane in transmembrane proteins such as GIRK2 (C). Interestingly, the activation of certain members of this family of ion channels by PIPn was largely dependent of their acyl chains [80]. The influence of the acyl chains could be more evident in PIPn–protein interactions that involve the extraction of the whole lipid from the membrane (B,D,E). In these proteins, the polar head can be buried in the protein (B), with the acyl chains occupying two hydrophobic channels, but can also be exposed to the solvent (E), which could be indicative of a major binding contribution for the acyl chains. (A) PX domain from p40phox bound to di-C4-PI3P (PDB 1H6H) [124]. (B) PITPα complexed to PI from bovine liver (PDB 1UW5) [125]. (C) G protein-gated inward rectifier K+ channel GIRK2 in complex with sodium and di-C8-PI(4,5)P2 (PDB 3SYA) [126]. (D) Osh6p in complex with brain PI4P (4PH7) [127]. (E) nuclear receptor SF-1 bound to di-C16:0-PIP3 (PDB 4QJR) [92].

Figure 4.
Phosphoinositide–protein interaction modes.

The influence of the acyl chains in PIPn–protein interactions is often missed in structural studies by the utilization of soluble, short-chain versions of PIPn, the conformational dynamics of longer chains and the lack of a membrane context. (AE) Examples of protein–PIPn interactions adapted from the Protein Data Bank, where the protein backbone is shown interacting with PIPn of diverse acyl chain composition (in spacefill representation: carbon, grey; phosphorus, orange; oxygen, red). The original lipid ligand is indicated, although shorter versions are represented in the structures of PITPα (B) and GIRK2 (C) where the PIPn acyl chains were not fully solved. Although the interaction with PIPn-binding domains is often restricted to the phosphorylated inositol ring (A), its presentation relative to the membrane surface could be affected by the acyl chain length/saturation. The PIPn acyl chains can directly interact with hydrophobic residues inserted in the membrane in transmembrane proteins such as GIRK2 (C). Interestingly, the activation of certain members of this family of ion channels by PIPn was largely dependent of their acyl chains [80]. The influence of the acyl chains could be more evident in PIPn–protein interactions that involve the extraction of the whole lipid from the membrane (B,D,E). In these proteins, the polar head can be buried in the protein (B), with the acyl chains occupying two hydrophobic channels, but can also be exposed to the solvent (E), which could be indicative of a major binding contribution for the acyl chains. (A) PX domain from p40phox bound to di-C4-PI3P (PDB 1H6H) [124]. (B) PITPα complexed to PI from bovine liver (PDB 1UW5) [125]. (C) G protein-gated inward rectifier K+ channel GIRK2 in complex with sodium and di-C8-PI(4,5)P2 (PDB 3SYA) [126]. (D) Osh6p in complex with brain PI4P (4PH7) [127]. (E) nuclear receptor SF-1 bound to di-C16:0-PIP3 (PDB 4QJR) [92].

PIPn have also been found to interact specifically with a small number of intrinsic membrane proteins, such as members of the family of inward-rectifier potassium channels (IRKs) [1]; an example of PI(4,5)P2 bound to GIRK2 is shown in Figure 4C. Although the positions of the acyl chains are undefined in these structures (presumably because of their mobility), there is other evidence that the acyl chain composition is indeed relevant to these types of interactions. Thus, while constitutively active IRK1 selectively interacted with PI(4,5)P2 irrespective of its acyl chain composition, the G-protein activated GIRK1 and GIRK4 were promiscuously stimulated by PI(4,5)P2, PI(3,4)P2, PI(3,5)P2 or PIP3, but showed a strong preference for C38:4 over C32:0 species [80].

In principle, protein interactions involving the whole PIPn molecule, rather than just the head-group, might be more sensitive to acyl chain composition. This would be the case for LTPs, some of which can selectively extract individual phospholipid molecules from a donor membrane and transfer them to an acceptor membrane in vitro [68,81]. It is now clear that many LTPs are involved in non-vesicular transport of lipids between cellular membranes, though precisely how they achieve this is still under some debate [82]. An intriguing recent model suggests some members of the ORP/Osh family facilitate counter transport of distinct lipids, the direction of net transport of one lipid being created by the concentration gradient of a ‘driver’ lipid, PI4P, which is, in turn, determined by the separate locations of PI 4-kinases (on the acceptor membrane) and PI4P-4-phosphatases (on the ‘donor’ membrane), e.g. in PI4P-driven transport of PS between the ER and PM or, PI4P-driven transport of sterol between the ER and Golgi [83]. The crystal structures of several LTPs are known and they present various topologies for lipid binding, but often possess a hydrophobic groove or channel which can accommodate the acyl chains of phospholipids and separate them from the aqueous environment [81]; an example of PITPα bound to PI is shown in Figure 4B and Osh6p bound to PI4P is shown in Figure 4D. However, there is still very little data generated on the acyl chain selectivity of these LTPs for their cargo and the precise positions of the acyl chains in the available structures are frustratingly inexact, through a combination of acyl chain mobility and use of heterogenous lipid ligands (see Figure 4). The LTPs ORP8 and E-SYT2 expressed in mammalian cell lines co-purified with multiple phospholipid species [84,85] and other work has indicated that ORP8 can selectively interact with different acyl chain species of PI4P (and PI(4,5)P2) and suggests it might preferentially extract saturated species of PI4P from the plasma membrane [86]. Recent work has also suggested that the ORP homologues in yeast which regulate the transport of PS and cholesterol from the ER to the PM in exchange for PI4P, allow the creation of PM lipid nanodomains where sterol/PS aggregate with PI4P molecules remaining in the PM and facilitate their phosphorylation by PI4P5K. Remarkably, these sterol/PS/PI4P nanodomains were more efficiently established with unsaturated versions of PS and PI4P, and promoted the activation of PI4P5K by engaging with its amphipathic specificity loop [87]. Furthermore, whilst an analysis of PIs bound to PITPNC1 suggested a preference for C38:4 PI species, while shorter and more saturated species were bound to PITPα [88], there has been no systematic analysis of the acyl chain specificities of LTPs thought to be relevant to the counter exchange of DG, PA and PI across ER-PM contact sites, such as Nir2/3 and TMEM24. This would represent an obvious point at which selectivity for the C38:4 backbone might contribute to an effective PI cycle (see above).

It is also interesting to consider PIPn–protein interactions in the nucleus. Although PIPn are involved in the regulation of intra-nuclear processes, such as transcription or DNA-repair, their biophysical state in such a hydrophilic environment remains unclear, and little is known about their origin and acyl chain composition [89,90]. Small structures enriched in PI(4,5)P2 named Nuclear Lipid Islets have been immunoprecipitated and subjected to lipid analysis [91], but the distribution of PIPn molecular species has not been described. Crystal structures of PI(4,5)P2 and PIP3 bound to nuclear receptor SF-1 (Steroidogenic Factor-1) revealed that this interaction takes place through their hydrophobic tails (see Figure 4E), leaving the phosphorylated head exposed to the surface, where it could modulate co-activator binding [92] and be modified by PIPn kinases/phosphatases [93]. It is not known if this binding is affected by the acyl chain composition of the PIPn, though C32:0 species were used to derive the SF-1 crystals (Figure 4E). It has also been described that nuclear p53 can bind PI(4,5)P2, which results in its stabilization [94]. This interaction was resistant to denaturation, and was observed with natural PI(4,5)P2 but not with dioctanoyl-PI(4,5)P2.

Lessons from evolution

The synthesis of PI from CDP-DG and free inositol via PIS is conserved in all eukaryotes (reviewed in [95]), which also present most of their phosphorylated forms [96]. While these PIPn participate in homologous functions in different organisms, the PI enrichment in C38:4 backbones has only been observed thus far in vertebrates. Therefore, tracing the emergence of C38:4 enrichment during the evolution of PIPn signalling could provide some insights into the evolutionary pressures favouring a distinct and homogeneous acyl chain composition in this key family of signalling lipids.

While C38:4 enrichment might be restricted to vertebrates, a common feature of PIPn across many different organisms is the concentration of saturated acyl chains in sn-1 compared with other phospholipids [29,97,98]. In yeast, the enrichment in C16:0 or C18:0 in sn-1 appears to take place by distinct mechanisms; the PI synthase Pis1p seems to select CDP-DG molecules containing C16:0 [99], while C18:0 chains are incorporated via remodelling by the LPI acyltransferase Psi1p, their closest homologue to mammalian AGPAT8 [100]. Psi1Δ yeast strains present an abnormal localization of PI4P and PI(4,5)P2 and alterations in vesicle trafficking and cell polarity, which has been attributed to non-redundant functions of the PIPn pool containing C18:0 [101]. The main lysophospholipid sn-2 acyltransferase in yeast microsomes is Ale1p, an MBOAT family enzyme with a strong preference for mono- and poly-unsaturated acyl-CoA chains, though the longer chain C20:4-CoA is a poor substrate [102]. Studies investigating the substrate preferences of the mammalian MBOAT family suggest some overlapping selectivity for the lyso-phospholipid head group and varying selectivity for acyl-CoA chain length and degree of unsaturation [26], but MBOAT7 appears remarkably selective for both LPI and C20:4-CoA [28]. Similar results are observed in vitro with homologues in nematode [27] and fly [103], though C20:4 is absent from flies [104]. Therefore, during the evolution of PIPn signalling, two different families of acyltransferases (AGPAT and MBOAT) have evolved to generate a pool of PIPn with a stearoyl chain in sn-1 and a polyunsaturated fatty acid in sn-2, which in vertebrates becomes C20:4. Remarkably, the social amoeba Dictyostelium discoideum, which branched before the divergence of fungi and metazoa, also presents a unique hydrocarbon backbone in PIPn, composed of an ether-linked C16:0 chain in sn-1 and a ester-linked C18:1 chain in sn-2 (C34:1e) [105]. Taken together, a theme emerges that suggests a core functional advantage to the selected PI backbone, such as availability within the bilayer and/or IPn head-group presentation to effectors. This advantage may be fine-tuned in individual organisms to the precise temperature and lipid composition of the membranes in which they must act. Moreover, in some organisms, such as vertebrates and Dictyostelium, there appears to have been further enrichment to more precise molecular species, which may imply additional advantages, such as metabolic identity and optimal PIPn–protein interactions.

While the hydrolysis of PI(4,5)P2 by PLC seems to be ubiquitous in eukaryotes, the emergence of PLC signalling in metazoans might have increased the pressure for PIPn homeostatic mechanisms. In fly photoreceptor cells, which present heterogeneous PIPn species and require the rapid replenishment of PI(4,5)P2 to sustain their light-activated PLC signalling, an efficient PI cycle is organised by a specialized ER-PM contact site, the Submicrovillar Cisternae, where PI is resynthesized using PA molecules transferred from the PM by RdgB proteins (a homologue of mammalian Nir2) [68]. In mammals, the efficient operation of a PI cycle in different cell types might be facilitated by an additional layer of selectivity provided by the C38:4 backbone (see discussion above). This may be reflected in the divergence of C38:4-selective CDS2 and non-selective CDS1 in vertebrates [106] and it would be informative to investigate whether other proteins involved in the mammalian PI cycle, such as Nir2/3, have evolved selectivity for C38:4 substrates.

The presence of alternative PIPn species

Whilst the forgoing discussion has focussed on the enrichment of the C38:4 species of PIPn in primary mammalian tissue, several examples have been reported where this enrichment does not occur, or at least not to the same extent [107]. For instance, PI species in mouse testis and platelets present a relatively high proportion of palmitoyl (C16:0) chains [7,108]. High levels of a rare CL species composed by C16:0 chains have also been reported in testis, suggesting that these saturated phospholipids could have specific roles during spermatogenesis [109]. Regarding platelets, C32:0 is moderately abundant in PI but not in PIP, PIP2 or PIP3, suggesting that these denucleated cells may contain a PI pool which does not cycle through its phosphorylated forms in the same manner as C38:4 PIPn. In addition, the stimulation of the PI3K and PLC pathways in platelets indicated that some minor PIPn species present different dynamics to the C38:4 PIPn pool [108].

Alternative PIPn species might also be enriched in certain cell compartments. The exosomes released by prostate cancer cell line PC-3, showed a higher proportion of C34:1 and C36:1 species in PI compared with the parental cells [110]. As PI levels were very low in exosomes, their unusual composition could indicate an exclusion of polyunsaturated PI species during exosome formation. This lack of C38:4 enrichment in PIs has also been observed in extracellular vesicles derived from other prostate cell lines [111] or differentiated 3T3-L1 adipocytes [112], although their levels were not compared with those of the parental cells. Similarly, PI in lung surfactant showed a predominance of saturated/monounsaturated species, which could be related to its biophysical functions [25,113].

A higher saturation degree in alternative PIPn species could render functional advantages, such as facilitating their packing [3]. PIPn can aggregate in microdomains, as seen with clusters of synthasin-1A/PI(4,5)P2 controlling neuronal exocytosis [114]. Although these clusters could be spontaneously recreated in artificial membranes with both di-C18:1-PI(4,5)P2 and pig brain PI(4,5)P2 the influence of their acyl chain composition was not explored. Alternatively, the presence of more saturated, shorter chain species may simply reflect the lack of a requirement to remodel towards C38:4, because their main function might be distinct from their better established signalling roles e.g simply to provide more membrane.

While the enrichment in C38:4 species is generally preserved in primary cultures of mammalian cells, it is often lost in immortalized cell lines [115] (Figure 1D). The distribution of species in PIPn in these cells can be highly influenced by the fatty acid composition of the culture media, with C38:4 being dependent on the availability of C20:4 [9]. In fact, the distribution of PI species in vivo might be influenced by diet, as shown in obese rats fed with an insect-based chow, presenting a decreased proportion of C38:4 PI in the liver, though the total PUFA-PI was sustained by an increase in C38:3 PI [116]. Yet, different cell lines cultured in the same conditions can present extremely different proportions of C38:4 in PIPn. High levels of short and less saturated PI species have been observed in certain cancer cell lines [117], a phenotype that has also been observed in human tumours [118,119]. Mutations in p53 could be one of the factors favouring this unusual composition in PIs [120], possibly by altering the expression of genes involved in fatty acid synthesis [121]. Therefore, the systematic comparison of gene expression and mutation patterns in cell lines and tumours in relation to their % of C38:4 in PIPn could provide new insights into the mechanisms and functions behind this enrichment.

Future perspectives

There are still some major gaps in our knowledge of how the acyl chain composition of PIPn are created in different organisms and, in particular, how enrichment for the C38:4 species occurs in most mammalian tissues. To some extent, this reflects the technical challenges in measuring these different species, particularly the less abundant, more highly phosphorylated PIPn. Hopefully, recent developments in MS methods will enable more routine measurements and, importantly, allow independent regio-isomers of PIPn to be interrogated (ie PI3P vs PI4P vs PI5P; PI(3,4)P2 vs PI(4,5)P2 vs PI(3,5)P2 [122]. Without this knowledge, it is very difficult to envisage selective interventions to directly test hypotheses regarding the function of acyl chain composition. However, it is also the case that the acyl chains of PIPn have been largely ignored in considering their interaction with effectors, and this should be more carefully addressed in future structural and biochemical studies, particularly where the acyl chains are directly engaged e.g. in LTPs. There has clearly been evolutionary pressure to engineer the acyl chain composition of PIPn to different degrees, in different tissues and organisms, we just do not at present have a satisfactory explanation for it.

Perspectives

  • PIPns present an unusually uniform distribution of molecular species, with a marked predominance of the stearoyl/arachidonoyl hydrophobic backbone.

  • This enrichment is achieved by the acyl chain remodelling of PI molecules involving selective acyltransferases and might provide molecular identity with the PIPn pool.

  • New insights should arise from the study of the metabolic dynamics of PIPn molecular species and the inclusion of the acyl chain variable in functional and structural assays.

Abbreviations

     
  • AGPAT

    acylglycerol-3-phosphate acyltransferase

  •  
  • CDP-DG

    CDP-diacylglycerol

  •  
  • CDS

    CDP-diacylglycerol synthase

  •  
  • CL

    cardiolipin

  •  
  • DG

    diacylglycerol

  •  
  • DGK

    diacylglycerol kinase

  •  
  • ER

    endoplasmic reticulum

  •  
  • IRKs

    inward-rectifier potassium channels

  •  
  • LPA

    lysophosphatidic acid

  •  
  • LTPs

    lipid transfer proteins

  •  
  • MEFs

    mouse embryonic fibroblasts

  •  
  • MS

    mass spectrometry

  •  
  • PIPn

    Phosphoinositides

Acknowledgements

We are grateful to Bob Michell for critical reading of the manuscript. Len Stephens and Phillip Hawkins acknowledge financial support from the BBSRC (BB/P013384/1). David Barneda is an employee of AstraZeneca, Sabina Cosulich is an employee and stakeholder of AstraZeneca.

Competing Interests

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

References

References
1
Dickson
,
E.J.
and
Hille
,
B.
(
2019
)
Understanding phosphoinositides: rare, dynamic, and essential membrane phospholipids
.
Biochem. J.
476
,
1
23
2
Hammond
,
G.R.
and
Balla
,
T.
(
2015
)
Polyphosphoinositide binding domains: key to inositol lipid biology
.
Biochim. Biophys. Acta
1851
,
746
758
3
Harayama
,
T.
and
Riezman
,
H.
(
2018
)
Understanding the diversity of membrane lipid composition
.
Nat. Rev. Mol. Cell Biol.
19
,
281
4
Harayama
,
T.
,
Eto
,
M.
,
Shindou
,
H.
,
Kita
,
Y.
,
Otsubo
,
E.
,
Hishikawa
,
D.
et al.  (
2014
)
Lysophospholipid acyltransferases mediate phosphatidylcholine diversification to achieve the physical properties required in vivo
.
Cell Metab.
20
,
295
305
5
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 PtdInsP(2) in the mouse
.
PLoS ONE
8
,
e58425
6
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
7
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
8
Gaudette
,
D.C.
,
Aukema
,
H.M.
,
Jolly
,
C.A.
,
Chapkin
,
R.S.
and
Holub
,
B.J.
(
1993
)
Mass and fatty acid composition of the 3-phosphorylated phosphatidylinositol bisphosphate isomer in stimulated human platelets
.
J. Biol. Chem.
268
,
13773
13776
PMID:
[PubMed]
9
Anderson
,
K.E.
,
Juvin
,
V.
,
Clark
,
J.
,
Stephens
,
L.R.
and
Hawkins
,
P.T.
(
2016
)
Investigating the effect of arachidonate supplementation on the phosphoinositide content of MCF10a breast epithelial cells
.
Adv. Biol. Regul.
62
,
18
24
10
Blunsom
,
N.J.
and
Cockcroft
,
S.
(
2019
)
Phosphatidylinositol synthesis at the endoplasmic reticulum
.
Biochim. Biophys. Acta Mol. Cell Biol. Lipids
11
Tamura
,
Y.
,
Harada
,
Y.
,
Nishikawa
,
S.-.
,
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
12
Blunsom
,
N.J.
,
Gomez-Espinosa
,
E.
,
Ashlin
,
T.G.
and
Cockcroft
,
S.
(
2018
)
Mitochondrial CDP-diacylglycerol synthase activity is due to the peripheral protein, TAMM41 and not due to the integral membrane protein, CDP-diacylglycerol synthase 1
.
Biochim. Biophys. Acta Mol. Cell Biol. Lipids
1863
,
284
298
13
Bahmanyar
,
S.
,
Biggs
,
R.
,
Schuh
,
A.L.
,
Desai
,
A.
,
Muller-Reichert
,
T.
,
Audhya
,
A.
et al.  (
2014
)
Spatial control of phospholipid flux restricts endoplasmic reticulum sheet formation to allow nuclear envelope breakdown
.
Genes Dev.
28
,
121
126
14
Triebl
,
A.
,
Trötzmüller
,
M.
,
Eberl
,
A.
,
Hanel
,
P.
,
Hartler
,
J.
and
Köfeler
,
H.C.
(
2014
)
Quantitation of phosphatidic acid and lysophosphatidic acid molecular species using hydrophilic interaction liquid chromatography coupled to electrospray ionization high resolution mass spectrometry
.
J. Chromatogr. A
1347
,
104
110
15
Thompson
,
W.
and
Macdonald
,
G.
(
1976
)
Cytidine diphosphate diglyceride of bovine brain: positional distribution of fatty acids and analysis of major molecular species
.
Eur. J. Biochem.
65
,
107
111
16
Saito
,
S.
,
Goto
,
K.
,
Tonosaki
,
A.
and
Kondo
,
H.
(
1997
)
Gene cloning and characterization of CDP-diacylglycerol synthase from rat brain
.
J. Biol. Chem.
272
,
9503
9509
17
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
18
Papatheodorou
,
I.
,
Fonseca
,
N.A.
,
Keays
,
M.
,
Tang
,
Y.A.
,
Barrera
,
E.
,
Bazant
,
W.
et al.  (
2017
)
Expression atlas: gene and protein expression across multiple studies and organisms
.
Nucleic Acids Res.
46
,
D246
D251
19
Blunsom
,
N.J.
,
Gomez-Espinosa
,
E.
,
Ashlin
,
T.G.
and
Cockcroft
,
S.
(
2019
)
Sustained phospholipase C stimulation of H9c2 cardiomyoblasts by vasopressin induces an increase in CDP-diacylglycerol synthase 1 (CDS1) through protein kinase C and cFos
.
Biochim. Biophys. Acta Mol. Cell Biol. Lipids
1864
,
1072
1082
20
D'Souza
,
K.
and
Epand
,
R.M.
(
2015
)
The phosphatidylinositol synthase-catalyzed formation of phosphatidylinositol does not exhibit acyl chain specificity
.
Biochemistry
54
,
1151
1153
21
Holub
,
B.
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
PMID:
[PubMed]
22
Nakagawa
,
Y.
,
Rüstow
,
B.
,
Rabe
,
H.
,
Kunze
,
D.
and
Waku
,
K.
(
1989
)
The de novo synthesis of molecular species of phosphatidylinositol from endogenously labeled CDP diacylglycerol in alveolar macrophage microsomes
.
Arch. Biochem. Biophys.
268
,
559
566
23
Luthra
,
M.
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
24
Holub
,
B.
and
Piekarski
,
J.
(
1976
)
Biosynthesis of molecular species of CDP-diglyceride from endogenously-labeled phosphatidate in rat liver microsomes
.
Lipids
11
,
251
257
25
Postle
,
A.D.
,
Heeley
,
E.L.
and
Wilton
,
D.C.
(
2001
)
A comparison of the molecular species compositions of mammalian lung surfactant phospholipids
.
Comp. Biochem. Physiol. A Mol. Integr. Physiol.
129
,
65
73
26
Hishikawa
,
D.
,
Hashidate
,
T.
,
Shimizu
,
T.
and
Shindou
,
H.
(
2014
)
Diversity and function of membrane glycerophospholipids generated by the remodeling pathway in mammalian cells
.
J. Lipid Res.
55
,
799
807
27
Lee
,
H.-C.
,
Inoue
,
T.
,
Imae
,
R.
,
Kono
,
N.
,
Shirae
,
S.
,
Matsuda
,
S.
et al.  (
2008
)
Caenorhabditis elegans mboa-7, a member of the MBOAT family, is required for selective incorporation of polyunsaturated fatty acids into phosphatidylinositol
.
Mol. Biol. Cell
19
,
1174
1184
28
Gijon
,
M.A.
,
Riekhof
,
W.R.
,
Zarini
,
S.
,
Murphy
,
R.C.
and
Voelker
,
D.R.
(
2008
)
Lysophospholipid acyltransferases and arachidonate recycling in human neutrophils
.
J. Biol. Chem.
283
,
30235
30245
29
Imae
,
R.
,
Inoue
,
T.
,
Kimura
,
M.
,
Kanamori
,
T.
,
Tomioka
,
N.H.
,
Kage-Nakadai
,
E.
et al.  (
2010
)
Intracellular phospholipase A1 and acyltransferase, which are involved in Caenorhabditis elegans stem cell divisions, determine the sn-1 fatty acyl chain of phosphatidylinositol
.
Mol. Biol. Cell
21
,
3114
3124
30
Cao
,
J.
,
Liu
,
Y.
,
Lockwood
,
J.
,
Burn
,
P.
and
Shi
,
Y.
(
2004
)
A novel cardiolipin-remodeling pathway revealed by a gene encoding an endoplasmic reticulum-associated acyl-CoA: lysocardiolipin acyltransferase (ALCAT1) in mouse
.
J. Biol. Chem.
279
,
31727
31734
31
Agarwal
,
A.K.
,
Barnes
,
R.I.
and
Garg
,
A.
(
2006
)
Functional characterization of human 1-acylglycerol-3-phosphate acyltransferase isoform 8: cloning, tissue distribution, gene structure, and enzymatic activity
.
Arch. Biochem. Biophys.
449
,
64
76
32
Zhao
,
Y.
,
Chen
,
Y.-Q.
,
Li
,
S.
,
Konrad
,
R.J.
and
Cao
,
G.
(
2009
)
The microsomal cardiolipin remodeling enzyme acyl-CoA lysocardiolipin acyltransferase is an acyltransferase of multiple anionic lysophospholipids
.
J. Lipid Res.
50
,
945
956
33
Bone
,
L.N.
,
Dayam
,
R.M.
,
Lee
,
M.
,
Kono
,
N.
,
Fairn
,
G.D.
,
Arai
,
H.
et al.  (
2017
)
The acyltransferase LYCAT controls specific phosphoinositides and related membrane traffic
.
Mol. Biol. Cell
28
,
161
172
34
Choy
,
C.H.
,
Han
,
B.-K.
and
Botelho
,
R.J.
(
2017
)
Phosphoinositide diversity, distribution, and effector function: stepping out of the box
.
Bioessays
39
,
1700121
35
Buch
,
S.
,
Stickel
,
F.
,
Trépo
,
E.
,
Way
,
M.
,
Herrmann
,
A.
,
Nischalke
,
H.D.
et al.  (
2015
)
A genome-wide association study confirms PNPLA3 and identifies TM6SF2 and MBOAT7 as risk loci for alcohol-related cirrhosis
.
Nat. Genet.
47
,
1443
36
Thabet
,
K.
,
Asimakopoulos
,
A.
,
Shojaei
,
M.
,
Romero-Gomez
,
M.
,
Mangia
,
A.
,
Irving
,
W.L.
et al.  (
2016
)
MBOAT7 rs641738 increases risk of liver inflammation and transition to fibrosis in chronic hepatitis C
.
Nat. Commun.
7
,
12757
37
Thabet
,
K.
,
Chan
,
H.L.Y.
,
Petta
,
S.
,
Mangia
,
A.
,
Berg
,
T.
,
Boonstra
,
A.
et al.  (
2017
)
The membrane-bound O-acyltransferase domain-containing 7 variant rs641738 increases inflammation and fibrosis in chronic hepatitis B
.
Hepatology
65
,
1840
1850
38
Xia
,
Y.
,
Huang
,
C.X.
,
Li
,
G.Y.
,
Chen
,
K.H.
,
Han
,
L.
,
Tang
,
L.
et al.  (
2019
)
Meta-analysis of the association between MBOAT7 rs641738, TM6SF2 rs58542926 and nonalcoholic fatty liver disease susceptibility
.
Clin. Res. Hepatol. Gastroenterol.
43
,
533
541
39
Mancina
,
R.M.
,
Dongiovanni
,
P.
,
Petta
,
S.
,
Pingitore
,
P.
,
Meroni
,
M.
,
Rametta
,
R.
et al.  (
2016
)
The MBOAT7-TMC4 variant rs641738 increases risk of nonalcoholic fatty liver disease in individuals of european descent
.
Gastroenterology
150
,
1219
1230.e6
40
Luukkonen
,
P.K.
,
Zhou
,
Y.
,
Hyötyläinen
,
T.
,
Leivonen
,
M.
,
Arola
,
J.
,
Orho-Melander
,
M.
et al.  (
2016
)
The MBOAT7 variant rs641738 alters hepatic phosphatidylinositols and increases severity of non-alcoholic fatty liver disease in humans
.
J. Hepatol.
65
,
1263
1265
41
Kainu
,
V.
,
Hermansson
,
M.
and
Somerharju
,
P.
(
2008
)
Electrospray ionization mass spectrometry and exogenous heavy isotope-labeled lipid species provide detailed information on aminophospholipid acyl chain remodeling
.
J. Biol. Chem.
283
,
3676
3687
42
Darnell
,
J.C.
,
Osterman
,
D.G.
and
Saltiel
,
A.R.
(
1991
)
Synthesis of phosphatidylinositol in rat liver microsomes is accompanied by the rapid formation of lysophosphatidylinositol
.
Biochim. Biophys. Acta Lipids Lipid Metab.
1084
,
269
278
43
Schaloske
,
R.H.
and
Dennis
,
E.A.
(
2006
)
The phospholipase A2 superfamily and its group numbering system
.
Biochim. Biophys. Acta Mol. Cell Biol. Lipids
1761
,
1246
1259
44
Balsinde
,
J.
,
Ackermann
,
E.J.
,
Conde-Frieboes
,
K.
and
Dennis
,
E.A.
(
1995
)
Inhibition of calcium-independent phospholipase A2 prevents arachidonic acid incorporation and phospholipid remodeling in P388D1 macrophages
.
Proc. Natl Acad. Sci. U.S.A.
92
,
8527
8531
45
Balsinde
,
J.
,
Balboa
,
M.A.
and
Dennis
,
E.A.
(
1997
)
Antisense inhibition of group VI Ca2+-independent phospholipase A2 blocks phospholipid fatty acid remodeling in murine P388D1 macrophages
.
J. Biol. Chem.
272
,
29317
29321
46
Murakami
,
M.
,
Masuda
,
S.
,
Ueda-Semmyo
,
K.
,
Yoda
,
E.
,
Kuwata
,
H.
,
Takanezawa
,
Y.
et al.  (
2005
)
Group VIB Ca2+-independent phospholipase A2γ promotes cellular membrane hydrolysis and prostaglandin production in a manner distinct from other intracellular phospholipases A2
.
J. Biol. Chem.
280
,
14028
14041
47
Cheon
,
Y.
,
Kim
,
H.-W.
,
Igarashi
,
M.
,
Modi
,
H.R.
,
Chang
,
L.
,
Ma
,
K.
et al.  (
2012
)
Disturbed brain phospholipid and docosahexaenoic acid metabolism in calcium-independent phospholipase A2-VIA (iPLA2β)-knockout mice
.
Biochim. Biophys. Acta Mol. Cell Biol. Lipids
1821
,
1278
1286
48
Yamashita
,
A.
,
Kumazawa
,
T.
,
Koga
,
H.
,
Suzuki
,
N.
,
Oka
,
S.
and
Sugiura
,
T.
(
2010
)
Generation of lysophosphatidylinositol by DDHD domain containing 1 (DDHD1): possible involvement of phospholipase D/phosphatidic acid in the activation of DDHD1
.
Biochim. Biophys. Acta Mol. Cell Biol. Lipids
1801
,
711
720
49
Oka
,
S.
,
Nakajima
,
K.
,
Yamashita
,
A.
,
Kishimoto
,
S.
and
Sugiura
,
T.
(
2007
)
Identification of GPR55 as a lysophosphatidylinositol receptor
.
Biochem. Biophys. Res. Commun.
362
,
928
934
50
Alhouayek
,
M.
,
Masquelier
,
J.
and
Muccioli
,
G.G.
(
2018
)
Lysophosphatidylinositols, from cell membrane constituents to GPR55 ligands
.
Trends Pharmacol. Sci.
39
,
586
604
51
Inloes
,
J.M.
,
Jing
,
H.
and
Cravatt
,
B.F.
(
2018
)
The spastic paraplegia-associated phospholipase DDHD1 is a primary brain phosphatidylinositol lipase
.
Biochemistry
57
,
5759
5767
52
Masquelier
,
J.
and
Muccioli
,
G.G.
(
2016
)
Development and validation of a specific and sensitive HPLC-ESI-MS method for quantification of lysophosphatidylinositols and evaluation of their levels in mice tissues
.
J. Pharm. Biomed. Anal.
126
,
132
140
53
Mariggiò
,
S.
,
Sebastià
,
J.
,
Filippi
,
B.M.
,
Iurisci
,
C.
,
Volonté
,
C.
,
Amadio
,
S.
et al.  (
2006
)
A novel pathway of cell growth regulation mediated by a PLA2α-derived phosphoinositide metabolite
.
FASEB J.
20
,
2567
2569
54
Murase
,
S.
and
Okuyama
,
H.
(
1985
)
A membrane-bound phospholipase C with an apparent specificity for lysophosphatidylinositol in porcine platelets
.
J. Biol. Chem.
260
,
262
265
PMID:
[PubMed]
55
Aoki
,
J.
,
Taira
,
A.
,
Takanezawa
,
Y.
,
Kishi
,
Y.
,
Hama
,
K.
,
Kishimoto
,
T.
et al.  (
2002
)
Serum lysophosphatidic acid is produced through diverse phospholipase pathways
.
J. Biol. Chem.
277
,
48737
48744
56
Yamashita
,
A.
,
Oka
,
S.
,
Tanikawa
,
T.
,
Hayashi
,
Y.
,
Nemoto-Sasaki
,
Y.
and
Sugiura
,
T.
(
2013
)
The actions and metabolism of lysophosphatidylinositol, an endogenous agonist for GPR55
.
Prostaglandins Other Lipid Mediat.
107
,
103
116
57
Yamashita
,
A.
,
Hayashi
,
Y.
,
Matsumoto
,
N.
,
Nemoto-Sasaki
,
Y.
,
Koizumi
,
T.
,
Inagaki
,
Y.
et al.  (
2017
)
Coenzyme-A-independent transacylation system; possible involvement of phospholipase A2 in transacylation
.
Biology
6
,
23
58
Schlame
,
M.
(
2013
)
Cardiolipin remodeling and the function of tafazzin
.
Biochim. Biophys. Acta Mol. Cell Biol. Lipids
1831
,
582
588
59
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
60
English
,
A.R.
and
Voeltz
,
G.K.
(
2013
)
Rab10 GTPase regulates ER dynamics and morphology
.
Nat. Cell Biol.
15
,
169
61
Tábara
,
L.C.
,
Vicente
,
J.-J.
,
Biazik
,
J.
,
Eskelinen
,
E.-L.
,
Vincent
,
O.
and
Escalante
,
R.
(
2018
)
Vacuole membrane protein 1 marks endoplasmic reticulum subdomains enriched in phospholipid synthesizing enzymes and is required for phosphoinositide distribution
.
Traffic
19
,
624
638
62
Nishimura
,
T.
,
Tamura
,
N.
,
Kono
,
N.
,
Shimanaka
,
Y.
,
Arai
,
H.
,
Yamamoto
,
H.
et al.  (
2017
)
Autophagosome formation is initiated at phosphatidylinositol synthase-enriched ER subdomains
.
EMBO J.
36
,
1719
63
Michell
,
R.H.
(
1975
)
Inositol phospholipids and cell surface receptor function
.
Biochim. Biophys. Acta Rev. Biomembr.
415
,
81
147
64
Thillaiappan
,
N.B.
,
Chakraborty
,
P.
,
Hasan
,
G.
and
Taylor
,
C.W.
(
2018
)
IP3 receptors and Ca2+ entry
.
Biochim. Biophys. Acta Mol. Cell Res.
1866
,
1092
1100
65
Gallegos
,
L.L.
and
Newton
,
A.C.
(
2008
)
Spatiotemporal dynamics of lipid signaling: protein kinase C as a paradigm
.
IUBMB Life
60
,
782
789
66
Pettitt
,
T.R.
and
Wakelam
,
M.J.
(
1999
)
Diacylglycerol kinase ε, but not ζ, selectively removes polyunsaturated diacylglycerol, inducing altered protein kinase C distribution in vivo
.
J. Biol. Chem.
274
,
36181
36186
67
Chen
,
Y.-J.
,
Quintanilla
,
C.G.
and
Liou
,
J.
(
2019
)
Recent insights into mammalian ER–PM junctions
.
Curr. Opin. Cell Biol.
57
,
99
105
68
Cockcroft
,
S.
and
Raghu
,
P.
(
2018
)
Phospholipid transport protein function at organelle contact sites
.
Curr. Opin. Cell Biol.
53
,
52
60
69
Saheki
,
Y.
and
Camilli
,
P.D.
(
2017
)
Endoplasmic reticulum–plasma membrane contact sites
.
Annu. Rev. Biochem.
86
,
659
684
70
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
71
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
72
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
73
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
74
Lung
,
M.
,
Shulga
,
Y.V.
,
Ivanova
,
P.T.
,
Myers
,
D.S.
,
Milne
,
S.B.
,
Brown
,
H.A.
et al.  (
2009
)
Diacylglycerol kinase ɛ is selective for both acyl chains of phosphatidic acid or diacylglycerol
.
J. Biol. Chem.
284
,
31062
31073
75
Shulga
,
Y.V.
,
Anderson
,
R.A.
,
Topham
,
M.K.
and
Epand
,
R.M.
(
2012
)
Phosphatidylinositol-4-phosphate 5-kinase isoforms exhibit acyl chain selectivity for both substrate and lipid activator
.
J. Biol. Chem.
287
,
35953
35963
76
Epand
,
R.M.
(
2017
)
Features of the phosphatidylinositol cycle and its role in signal transduction
.
J. Membr. Biol.
250
,
353
366
77
de Turco
,
E.B.R.
,
Tang
,
W.
,
Topham
,
M.K.
,
Sakane
,
F.
,
Marcheselli
,
V.L.
,
Chen
,
C.
et al.  (
2001
)
Diacylglycerol kinase ɛ regulates seizure susceptibility and long-term potentiation through arachidonoyl–inositol lipid signaling
.
Proc. Natl Acad. Sci. U.S.A.
98
,
4740
4745
78
Takemasu
,
S.
,
Ito
,
M.
,
Morioka
,
S.
,
Nigorikawa
,
K.
,
Kofuji
,
S.
,
Takasuga
,
S.
et al.  (
2019
)
Lysophosphatidylinositol-acyltransferase-1 is involved in cytosolic Ca2+ oscillations in macrophages
.
Genes Cells
24
,
366
376
79
Stephens
,
L.
,
Anderson
,
K.
,
Stokoe
,
D.
,
Erdjument-Bromage
,
H.
,
Painter
,
G.F.
,
Holmes
,
A.B.
et al.  (
1998
)
Protein kinase B kinases that mediate phosphatidylinositol 3,4,5–trisphosphate-dependent activation of protein kinase B
.
Science
279
,
710
714
80
Rohács
,
T.
,
Chen
,
J.
,
Prestwich
,
G.D.
and
Logothetis
,
D.E.
(
1999
)
Distinct specificities of inwardly rectifying K+ channels for phosphoinositides
.
J. Biol. Chem.
274
,
36065
36072
81
Wong
,
L.H.
,
Gatta
,
A.T.
and
Levine
,
T.P.
(
2019
)
Lipid transfer proteins: the lipid commute via shuttles, bridges and tubes
.
Nat. Rev. Mol. Cell Biol.
20
,
85
101
82
Wang
,
Y.
,
Mousley
,
C.J.
,
Lete
,
M.G.
and
Bankaitis
,
V.A.
(
2019
)
An equal opportunity collaboration between lipid metabolism and proteins in the control of membrane trafficking in the trans-Golgi and endosomal systems
.
Curr. Opin. Cell Biol.
59
,
58
72
83
Antonny
,
B.
,
Bigay
,
J.
and
Mesmin
,
B.
(
2018
)
The oxysterol-binding protein cycle: burning off PI(4)P to transport cholesterol
.
Annu. Rev. Biochem.
87
,
809
837
84
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
85
Chung
,
J.
,
Torta
,
F.
,
Masai
,
K.
,
Lucast
,
L.
,
Czapla
,
H.
,
Tanner
,
L.B.
et al.  (
2015
)
PI4P/phosphatidylserine countertransport at ORP5-and ORP8-mediated ER–plasma membrane contacts
.
Science
349
,
428
432
86
Ghai
,
R.
,
Du
,
X.
,
Wang
,
H.
,
Dong
,
J.
,
Ferguson
,
C.
,
Brown
,
A.J.
et al.  (
2017
)
ORP5 and ORP8 bind phosphatidylinositol-4, 5–biphosphate (PtdIns (4, 5) P 2) and regulate its level at the plasma membrane
.
Nat. Commun.
8
,
757
87
Nishimura
,
T.
,
Gecht
,
M.
,
Covino
,
R.
,
Hummer
,
G.
,
Surma
,
M.A.
,
Klose
,
C.
et al.  (
2019
)
Osh proteins control nanoscale lipid organization necessary for PI(4,5)P2 synthesis
.
Mol. Cell
75
,
1043
1057.e8
88
Garner
,
K.
,
Hunt
,
A.N.
,
Koster
,
G.
,
Somerharju
,
P.
,
Groves
,
E.
,
Li
,
M.
et al.  (
2012
)
Phosphatidylinositol transfer protein, cytoplasmic 1 (PITPNC1) binds and transfers phosphatidic acid
.
J. Biol. Chem.
287
,
32263
32276
89
Jacobsen
,
R.G.
,
Mazloumi Gavgani
,
F.
,
Edson
,
A.J.
,
Goris
,
M.
,
Altankhuyag
,
A.
and
Lewis
,
A.E.
(
2019
)
Polyphosphoinositides in the nucleus: roadmap of their effectors and mechanisms of interaction
.
Adv. Biol. Regul.
72
,
7
21
90
Fiume
,
R.
,
Faenza
,
I.
,
Sheth
,
B.
,
Poli
,
A.
,
Vidalle
,
M.C.
,
Mazzetti
,
C.
et al.  (
2019
)
Nuclear phosphoinositides: their regulation and roles in nuclear functions
.
Int. J. Mol. Sci.
20
,
2991
91
Sobol
,
M.
,
Krausová
,
A.
,
Yildirim
,
S.
,
Kalasová
,
I.
,
Fáberová
,
V.
,
Vrkoslav
,
V.
et al.  (
2018
)
Nuclear phosphatidylinositol 4, 5–bisphosphate islets contribute to efficient RNA polymerase II-dependent transcription
.
J Cell Sci
131
,
jcs211094
92
Blind
,
R.D.
,
Sablin
,
E.P.
,
Kuchenbecker
,
K.M.
,
Chiu
,
H.-J.
,
Deacon
,
A.M.
,
Das
,
D.
et al.  (
2014
)
The signaling phospholipid PIP3 creates a new interaction surface on the nuclear receptor SF-1
.
Proc. Natl Acad. Sci. U.S.A.
111
,
15054
15059
93
Blind
,
R.D.
,
Suzawa
,
M.
and
Ingraham
,
H.A.
(
2012
)
Direct modification and activation of a nuclear receptor–PIP2 complex by the inositol lipid kinase IPMK
.
Sci. Signal.
5
,
ra44
94
Choi
,
S.
,
Chen
,
M.
,
Cryns
,
V.L.
and
Anderson
,
R.A.
(
2019
)
A nuclear phosphoinositide kinase complex regulates p53
.
Nat. Cell Biol.
21
,
462
475
95
Michell
,
R.H.
(
2008
)
Inositol derivatives: evolution and functions
.
Nat. Rev. Mol. Cell Biol.
9
,
151
96
Brown
,
J.R.
and
Auger
,
K.R.
(
2011
)
Phylogenomics of phosphoinositide lipid kinases: perspectives on the evolution of second messenger signaling and drug discovery
.
BMC Evol. Biol.
11
,
4
97
Schneiter
,
R.
,
Brügger
,
B.
,
Sandhoff
,
R.
,
Zellnig
,
G.
,
Leber
,
A.
,
Lampl
,
M.
et al.  (
1999
)
Electrospray ionization tandem mass spectrometry (ESI-MS/MS) analysis of the lipid molecular species composition of yeast subcellular membranes reveals acyl chain-based sorting/remodeling of distinct molecular species en route to the plasma membrane
.
J. Cell Biol.
146
,
741
754
98
McCourt
,
P.
and
Somerville
,
C.
(
1986
)
A mutant of Arabidopsis deficient in C18: 3 and C16: 3 leaf lipids
.
Plant Physiol.
81
,
859
864
99
Kaliszewski
,
P.
,
Ferreira
,
T.
,
Gajewska
,
B.
,
Szkopinska
,
A.
,
Berges
,
T.
and
Żołądek
,
T.
(
2006
)
Enhanced levels of Pis1p (phosphatidylinositol synthase) improve the growth of Saccharomyces cerevisiae cells deficient in Rsp5 ubiquitin ligase
.
Biochem. J.
395
,
173
181
100
Le Guédard
,
M.
,
Bessoule
,
J.-J.
,
Boyer
,
V.
,
Ayciriex
,
S.
,
Velours
,
G.
,
Kulik
,
W.
et al.  (
2009
)
PSI1 is responsible for the stearic acid enrichment that is characteristic of phosphatidylinositol in yeast
.
FEBS J.
276
,
6412
6424
101
Doignon
,
F.
,
Laquel
,
P.
,
Testet
,
E.
,
Tuphile
,
K.
,
Fouillen
,
L.
and
Bessoule
,
J.-J.
(
2016
)
Requirement of phosphoinositides containing stearic acid to control cell polarity
.
Mol. Cell. Biol.
36
,
765
780
102
Riekhof
,
W.R.
,
Wu
,
J.
,
Gijón
,
M.A.
,
Zarini
,
S.
,
Murphy
,
R.C.
and
Voelker
,
D.R.
(
2007
)
Lysophosphatidylcholine metabolism in Saccharomyces cerevisiae the role of P-type ATPases in transport and a broad specificity acyltransferase in acylation
.
J. Biol. Chem.
282
,
36853
36861
103
Steinhauer
,
J.
,
Gijón
,
M.A.
,
Riekhof
,
W.R.
,
Voelker
,
D.R.
,
Murphy
,
R.C.
,
Treisman
,
J.E.
et al.  (
2009
)
Drosophila lysophospholipid acyltransferases are specifically required for germ cell development
.
Mol. Biol. Cell
20
,
5224
5235
104
Yoshioka
,
T.
,
Inoue
,
H.
,
Kasama
,
T.
,
Seyama
,
Y.
,
Nakashima
,
S.
,
Nozawa
,
Y.
et al.  (
1985
)
Evidence that arachidonic acid is deficient in phosphatidylinositol of Drosophila heads
.
J. Biochem.
98
,
657
662
105
Clark
,
J.
,
Kay
,
R.R.
,
Kielkowska
,
A.
,
Niewczas
,
I.
,
Fets
,
L.
,
Oxley
,
D.
et al.  (
2014
)
Dictyostelium uses ether-linked inositol phospholipids for intracellular signalling
.
EMBO J.
33
,
2188
2200
106
Schreiber
,
F.
,
Patricio
,
M.
,
Muffato
,
M.
,
Pignatelli
,
M.
and
Bateman
,
A.
(
2014
)
Treefam v9: a new website, more species and orthology-on-the-fly
.
Nucleic Acids Res.
42
,
D922
D925
107
Hicks
,
A.M.
,
DeLong
,
C.J.
,
Thomas
,
M.J.
,
Samuel
,
M.
and
Cui
,
Z.
(
2006
)
Unique molecular signatures of glycerophospholipid species in different rat tissues analyzed by tandem mass spectrometry
.
Biochim. Biophys. Acta Mol. Cell Biol. Lipids
1761
,
1022
1029
108
Mujalli
,
A.
,
Chicanne
,
G.
,
Bertrand-Michel
,
J.
,
Viars
,
F.
,
Stephens
,
L.
,
Hawkins
,
P.
et al.  (
2018
)
Profiling of phosphoinositide molecular species in human and mouse platelets identifies new species increasing following stimulation
.
Biochim. Biophys. Acta Mol. Cell Biol. Lipids
1863
,
1121
1131
109
Ren
,
M.
,
Xu
,
Y.
,
Erdjument-Bromage
,
H.
,
Donelian
,
A.
,
Phoon
,
C.K.L.
,
Terada
,
N.
et al.  (
2019
)
Extramitochondrial cardiolipin suggests a novel function of mitochondria in spermatogenesis
.
J. Cell Biol.
218
,
1491
1502
110
Llorente
,
A.
,
Skotland
,
T.
,
Sylvänne
,
T.
,
Kauhanen
,
D.
,
Róg
,
T.
,
Orłowski
,
A.
et al.  (
2013
)
Molecular lipidomics of exosomes released by PC-3 prostate cancer cells
.
Biochim. Biophys. Acta Mol. Cell Biol. Lipids
1831
,
1302
1309
111
Brzozowski
,
J.S.
,
Jankowski
,
H.
,
Bond
,
D.R.
,
McCague
,
S.B.
,
Munro
,
B.R.
,
Predebon
,
M.J.
et al.  (
2018
)
Lipidomic profiling of extracellular vesicles derived from prostate and prostate cancer cell lines
.
Lipids Health Dis.
17
,
211
112
Durcin
,
M.
,
Fleury
,
A.
,
Taillebois
,
E.
,
Hilairet
,
G.
,
Krupova
,
Z.
,
Henry
,
C.
et al.  (
2017
)
Characterisation of adipocyte-derived extracellular vesicle subtypes identifies distinct protein and lipid signatures for large and small extracellular vesicles
.
J. Extracell. Vesicles
6
,
1305677
113
Rüstow
,
B.
,
Nakagawa
,
Y.
,
Rabe
,
H.
,
Waku
,
K.
and
Kunze
,
D.
(
1988
)
Species pattern of phosphatidylinositol from lung surfactant and a comparison of the species pattern of phosphatidylinositol and phosphatidylglycerol synthesized de novo in lung microsomal fractions
.
Biochem. J.
254
,
67
71
114
van den Bogaart
,
G.
,
Meyenberg
,
K.
,
Risselada
,
H.J.
,
Amin
,
H.
,
Willig
,
K.I.
,
Hubrich
,
B.E.
et al.  (
2011
)
Membrane protein sequestering by ionic protein–lipid interactions
.
Nature
479
,
552
115
Traynor-Kaplan
,
A.
,
Kruse
,
M.
,
Dickson
,
E.J.
,
Dai
,
G.
,
Vivas
,
O.
,
Yu
,
H.
, et al.  (
2017
)
Fatty-acyl chain profiles of cellular phosphoinositides
.
Biochim. Biophys. Acta Mol. Cell Biol. Lipids
1862
,
513
522
116
Meyer
,
S.
,
Gessner
,
D.K.
,
Wen
,
G.
,
Most
,
E.
,
Liebisch
,
G.
,
Zorn
,
H.
et al.  (
2019
)
The antisteatotic and hypolipidemic effect of insect meal in obese Zucker rats is accompanied by profound changes in hepatic phospholipid and 1-carbon metabolism
.
Mol. Nutr. Food Res.
,
1801305
117
He
,
M.
,
Guo
,
S.
and
Li
,
Z.
(
2015
)
In situ characterizing membrane lipid phenotype of breast cancer cells using mass spectrometry profiling
.
Sci. Rep.
5
,
11298
118
Goto
,
T.
,
Terada
,
N.
,
Inoue
,
T.
,
Nakayama
,
K.
,
Okada
,
Y.
,
Yoshikawa
,
T.
et al.  (
2014
)
The expression profile of phosphatidylinositol in high spatial resolution imaging mass spectrometry as a potential biomarker for prostate cancer
.
PLoS ONE
9
,
e90242
119
Kawashima
,
M.
,
Iwamoto
,
N.
,
Kawaguchi-Sakita
,
N.
,
Sugimoto
,
M.
,
Ueno
,
T.
,
Mikami
,
Y.
et al.  (
2013
)
High-resolution imaging mass spectrometry reveals detailed spatial distribution of phosphatidylinositols in human breast cancer
.
Cancer Sci.
104
,
1372
1379
120
Naguib
,
A.
,
Bencze
,
G.
,
Engle
,
D.D.
,
Chio
,
I.I.C.
,
Herzka
,
T.
,
Watrud
,
K.
et al.  (
2015
)
P53 mutations change phosphatidylinositol acyl chain composition
.
Cell Rep.
10
,
8
19
121
Rueda-Rincon
,
N.
,
Bloch
,
K.
,
Derua
,
R.
,
Vyas
,
R.
,
Harms
,
A.
,
Hankemeier
,
T.
et al.  (
2015
)
P53 attenuates AKT signaling by modulating membrane phospholipid composition
.
Oncotarget
6
,
21240
122
Bui
,
H.H.
,
Sanders
,
P.E.
,
Bodenmiller
,
D.
,
Kuo
,
M.S.
,
Donoho
,
G.P.
and
Fischl
,
A.S.
(
2018
)
Direct analysis of PI(3,4,5)P3 using liquid chromatography electrospray ionization tandem mass spectrometry
.
Anal. Biochem.
547
,
66
76
123
Kielkowska
,
A.
,
Niewczas
,
I.
,
Anderson
,
K.E.
,
Durrant
,
T.N.
,
Clark
,
J.
,
Stephens
,
L.R.
et al.  (
2014
)
A new approach to measuring phosphoinositides in cells by mass spectrometry
.
Adv. Biol. Regul.
54
,
131
141
124
Bravo
,
J.
,
Karathanassis
,
D.
,
Pacold
,
C.M.
,
Pacold
,
M.E.
,
Ellson
,
C.D.
,
Anderson
,
K.E.
et al.  (
2001
)
The crystal structure of the PX domain from p40phox bound to phosphatidylinositol 3-phosphate
.
Mol. Cell
8
,
829
839
125
Tilley
,
S.J.
,
Skippen
,
A.
,
Murray-Rust
,
J.
,
Swigart
,
P.M.
,
Stewart
,
A.
,
Morgan
,
C.P.
et al.  (
2004
)
Structure-function analysis of phosphatidylinositol transfer protein α bound to human phosphatidylinositol
.
Structure
12
,
317
326
126
Whorton
,
M.R.
and
MacKinnon
,
R.
(
2011
)
Crystal structure of the mammalian GIRK2K+ channel and gating regulation by G proteins, PIP2, and sodium
.
Cell
147
,
199
208
127
von Filseck
,
J.M.
,
Čopič
,
A.
,
Delfosse
,
V.
,
Vanni
,
S.
,
Jackson
,
C.L.
,
Bourguet
,
W.
et al.  (
2015
)
Phosphatidylserine transport by ORP/Osh proteins is driven by phosphatidylinositol 4-phosphate
.
Science
349
,
432
436
This is an open access article published by Portland Press Limited on behalf of the Biochemical Society and distributed under the Creative Commons Attribution License 4.0 (CC BY).