21.3: Biosynthesis of Membrane Glycerolipids

Phosphatidic Acid – Occurrence and Biosynthesis

Phosphatidic acid is not an abundant lipid constituent of any living organism and is seldom present at concentrations greater than picomolar in cells. Still, it is extremely important as an intermediate in the biosynthesis of other glycerophospholipids and triacylglycerols and as a signaling molecule or a precursor of signaling molecules. Indeed, it is often overestimated in tissues, as it can arise from inadvertent enzymatic hydrolysis under inappropriate storage or extraction conditions during analysis. It is the simplest diacylglycerophospholipid and the only one with a phosphomonoester head group. The molecule is acidic and carries a negative charge, i.e., it is an anionic lipid. The structure of phosphatidic acid is shown in Figure \(\PageIndex{1}\).

At least four important biosynthetic pathways for phosphatidic acid occur in different organelles under various stimuli. The main pathway involves the sequential acylation of sn-glycerol-3-phosphate, derived from glucose catabolism, by fatty acid acyl-CoA derivatives. First, one acyltransferase catalyzes acylation at position sn-1 to form lysophosphatidic acid (1‑acyl-sn-glycerol-3-phosphate), and then a second specific acyltransferase catalyzes acylation at position sn-2 to yield phosphatidic acid. The synthesis of phosphatidic acid from glycerol-3-phosphate is shown in Figure \(\PageIndex{2}\).

In mammals, the glycerol-3-phosphate acyltransferase that catalyzes the first step exists in four isoforms, two in the mitochondrial outer membrane (designated GPAT1 and 2) and two in the endoplasmic reticulum (GPAT3 and 4); all are membrane-bound enzymes, which are believed to span the membranes. GPAT1 is highly expressed in the liver and adipose tissue and responds to changes in feeding status via sterol regulatory element-binding protein-1 (SREBP-1), a master transcriptional regulator of lipogenic enzymes. It directs fatty acyl-CoA esters toward glycerolipid synthesis rather than β-oxidation. GPAT3 is especially important for triacylglycerol storage in adipocytes, while GPAT4 is the main contributor to lysophosphatidic acid synthesis in the liver and brown adipose tissue.

For the second step in phosphatidic acid biosynthesis, five mammalian acyl-CoA:lysophosphatidic acid acyltransferases are known, of which three are in the endoplasmic reticulum (LPAAT or LPAT or AGPAT1, 2 and 3), with a further two (LPAT4 and 5) on the outer mitochondrial membrane. While LPAT1 and 2 have strict specificity for lysophosphatidic acid as an acyl acceptor, other isoforms can esterify other lysophospholipids. Human LPAT1 showed higher activity with 14:0-, 16:0-, and 18:2‑CoAs, while LPAT2 prefers 20:4-CoA and LPAT3 produces phosphatidic acid containing docosahexaenoic acid (22:6(n-3)).  The latter is especially important in the retina and testes. LPAT4 and 5 prefer oleoyl-CoA and polyunsaturated acyl-CoAs as the acyl donor, suggesting a dual role in glycerolipid synthesis and remodeling. Endoplasmic reticulum activity predominates in adipose tissue, but mitochondrial forms are believed to account for half the activity in the liver. However, because phosphatidic acid traffics between the mitochondria and endoplasmic reticulum for remodeling or synthesis of other lipids, the relative contributions of the two can be difficult to assess.

In plants, the sn-glycerol-3-phosphate pathway exists in both plastids and the endoplasmic reticulum, with multiple isoforms of the two acyltransferases and differences in the acyl substrates. In brief, most plant lipid biosynthesis begins with fatty acid biosynthesis in the chloroplasts. In plastids, the acyltransferase ATS1 transfers 18:1 acyl groups from acyl-acyl carrier protein (acyl-ACP) to position sn-1 of glycerol 3-phosphate before ATS2 transfers a palmitoyl group from ACP to position sn-2, producing phosphatidic acid at the inner leaflet of the chloroplast inner envelope membrane (IEM). Fatty acids intended for the endoplasmic reticulum are released from ACP in the chloroplast stroma by IEM-associated thioesterases, exported, and then activated by outer envelope membrane acyl-CoA synthetases to produce species with C18 fatty acids in both positions. Thus, acyl-CoAs are used for phosphatidic acid biosynthesis in the endoplasmic reticulum, with marked differences in acyl-substrate specificity. Subsequently, plastids use phosphatidic acid to synthesize galactosyldiacylglycerols, while the endoplasmic reticulum uses it to synthesize triacylglycerols and phospholipids.

In bacteria, two enzyme families catalyze acylation at the sn-1 position of glycerol-3-phosphate. For example, one in Escherichia coli utilizes the acyl-acyl carrier protein (acyl-ACP) products of fatty acid synthesis as acyl donors, as well as acyl-CoA derived from exogenous fatty acids. In a second, broader group of bacteria, including cyanobacteria, enzymes (PlsX and PlsY) use the unique acyl donors, acyl phosphates derived partly from acyl-ACP, to acylate position sn-1. In this case, a further family of enzymes (PlsC) acylates position sn-2, using acyl-ACP as the acyl donor. However, some bacterial species may also use acyl-CoA.

In animals, a second biosynthetic pathway utilizes dihydroxyacetone phosphate (DHAP) as the primary precursor for the peroxisomal enzyme DHAP acyltransferase, which produces acyl-DHAP. Acyl-DHAP reductase converts this intermediate to lysophosphatidic acid in an NADPH-dependent reaction.  It is then acylated to form phosphatidic acid by the same LPAT as in the previous mechanism. This pathway is particularly important in ether lipid biosynthesis. The synthesis of phosphatidic acid from dihydroxyacetone phosphate is shown in Figure \(\PageIndex{3}\).

A third important route to phosphatidic acid is the hydrolysis of other phospholipids, especially phosphatidylcholine, by the enzyme phospholipase D (or by a family of related enzymes). The enzyme is readily available for study in plants, where the special functions of phosphatidic acid have long been known.  Phospholipase D is now recognized in bacteria, yeasts, and most animal cells. In the latter, it exists in two main isoforms with differing specificities and cellular locations; PLD1 is found mainly in the Golgi-lysosome continuum, while PLD2 is present mainly in the plasma membrane. They are phosphoproteins whose activity is regulated by kinases and phosphatases and by binding to phosphatidylinositol-4,5-bisphosphate. In mitochondria, a distinctive enzyme of this type utilizes cardiolipin as a substrate. The mechanism uses water as the nucleophile to catalyze hydrolysis of phosphodiester bonds in phospholipids. Phospholipase D activity depends on and is regulated by neurotransmitters, hormones, small monomeric GTPases, and lipids. The hydrolysis of phosphatidylcholine to phosphatidic acids by phospholipase D is shown in Figure \(\PageIndex{4}\).

In addition to its function in generating phosphatidic acid mainly for signaling purposes but also for the maintenance of membrane composition, phospholipase D is involved in intracellular protein trafficking, cytoskeletal dynamics, cell migration, and cell proliferation, partly through protein-protein interactions; it is considered to be important in inflammation and cancer growth and metastasis as a downstream transcriptional target of proteins involved in the pathophysiology of these diseases. It also has an unusual activity as a guanine nucleotide exchange factor. A transphosphatidylation reaction with ethanol generates phosphatidylethanol, a useful biomarker for human ethanol consumption.

Under some conditions, diacylglycerol kinases can generate phosphatidic acid from 1,2-diacyl-sn-glycerols, for example, those produced from other phospholipids by phospholipase C. Such enzymes appear ubiquitous, although bacteria and yeast differ structurally from mammalian enzymes. Diacylglycerol kinases, of which at least ten isoforms (DGKα to DGKκ) exist with different subcellular locations and functions in animals, use ATP as the phosphate donor. While the epsilon isoform (DGKε) preferentially uses the 1-stearoyl-2-arachidonoyl species of diacyl-sn-glycerols to produce phosphatidic acid for phosphatidylinositol biosynthesis, other isoenzymes phosphorylate diverse diacylglycerol species. In addition to producing phosphatidic acid for phospholipid production or signaling, these enzymes may attenuate diacylglycerol signaling. For example, diacylglycerol kinases can contribute to cellular asymmetry and control cell polarity by regulating gradients in diacylglycerol and phosphatidic acid concentrations. Figure \(\PageIndex{5}\) shows the synthesis of phosphatidic acid via diacylglycerols and the reverse reaction.

The reverse reaction, hydrolysis, is catalyzed by lipins. These enzymes regulate local phosphatidic acid concentrations and, therefore, its biological activity.

Another possible route to phosphatidic acid production for signaling is via acylation of lysophosphatidic acid, which can be produced independently for signaling, as discussed below. This pathway may be especially relevant in membranes, where the protein endophilin possesses LPAT activity and is believed to generate phosphatidic acid from lysophosphatidic acid, thereby altering membrane bilayer curvature.

Phosphatidic Acid - Role as a Lipid Precursor

In summary, phosphatidic acid generated via 1-acyl-sn-glycerol-3-phosphate is the primary precursor of other glycerolipids, although other pathways may be more important for generating the lipid for signaling functions. Whether separate pools of this lipid exist for specific purposes is not certain, as its intracellular distribution changes dynamically under various cellular conditions. This may be due to inter-organelle transfer via vesicular transport or to lipid transfer proteins at membrane contact sites. Control of its concentration in membranes, especially in the endoplasmic reticulum, is therefore of great importance, and a transcriptional repressor, 'Opi1', which binds specifically to phosphatidic acid in membranes, is a key regulatory factor. However, many other phosphatidic acid-binding proteins have been identified that influence how phosphatidic acid is used, either as a biosynthetic precursor or for signaling. The mechanisms of phosphatidic acid homeostasis differ among animals, plants, yeasts, and bacteria, reflecting the differing functional requirements of these organisms. Figure \(\PageIndex{6}\) shows the pathways for biosynthesis of complex glycerolipids.

In addition to dietary, hormonal, and tissue-specific factors in animals, the extent to which fatty acids are channeled either into triacylglycerol synthesis for storage in lipid droplets and secretion in lipoproteins or into glycerophospholipids for membrane formation depends to a large extent upon the enzymes of the glycerol-3-phosphate pathway, their isoform expression, activities, and locations. On the other hand, phosphatidic acid is not only a biosynthetic precursor of other lipids but also a regulatory molecule in the transcriptional control of glycerolipid synthesis genes, and regulating its concentration in cells is similarly essential. For example, the local concentration of phosphatidic acid in the endoplasmic reticulum is an important factor in lipid droplet biogenesis.

The subsequent steps in the biosynthesis of triacylglycerols and the various glycerophospholipids using phosphatidic acid are described in separate documents on this website. Thus, hydrolysis of phosphatidic acid by phosphatidate phosphatase enzymes (including lipins 1, 2, and 3) is the source of most other glycerolipids, e.g., sn‑1,2‑diacylglycerols (DG), which are the precursors for the biosynthesis of triacylglycerols (TAG), phosphatidylcholine (PC), and phosphatidylethanolamine (PE) via the so-called Kennedy pathway (also of monogalactosyldiacylglycerols in plants). Via reaction with cytidine triphosphate, phosphatidic acid is the precursor of cytidine diphosphate diacylglycerol, which is the key intermediate in the synthesis of phosphatidylglycerol (PG), and thence of cardiolipin (CL), and of phosphatidylinositol (PI), and, in prokaryotes and yeast but not animals, phosphatidylserine (PS). Depending on the organism and other factors, phosphatidylserine can serve as a precursor for phosphatidylethanolamine, which can then give rise to phosphatidylcholine via mono- and dimethyl-phosphatidylethanolamine intermediates. Cytidine diphosphate diacylglycerol synthase is another enzyme that consumes phosphatidic acid, helping modulate phosphatidic acid levels in cells and regulate processes mediated by this lipid.

While the fatty acid composition of phosphatidic acid can resemble that of the eventual products, the latter are generally much altered by remodeling after synthesis via deacylation-reacylation reactions.

Lysophosphatidic Acid

Figure \(\PageIndex{7}\) shows the structure of a lysophosphatidic acid (note the absence of an acyl group at C2).

Lysophosphatidic acid (LPA), or 1-acyl-sn-glycerol-3-phosphate, differs structurally from phosphatidic acid by having only one mole of fatty acid per mole of lipid. As such, it is one of the simplest possible glycerophospholipids. It exists in many molecular species, i.e., esterified to fatty acids ranging from 16:0 to 22:6, and preliminary evidence suggests that saturated and polyunsaturated species may differ in their biological properties under some circumstances. As the sn-1-acylated form is more thermodynamically stable, facile isomerization ensures that it tends to predominate. It lacks one fatty acid compared to phosphatidic acid, making it much more hydrophilic. At the same time, the additional hydroxyl group strengthens hydrogen bonding within membranes, properties that may be important for its function in cells.

Although lysophosphatidic acid is present at very low levels only in animal tissues, it is biologically important and influences many biochemical processes. It is a biosynthetic precursor of phosphatidic acid, but it is also of particular interest as a lipid mediator with growth factor-like activities. For example, activated platelets rapidly produce and release it to influence target cells.

Biosynthesis: In the circulation, the most important source of lysophosphatidic acid is the activity of an enzyme with lysophospholipase D-like activity, known as ‘autotaxin ’, on lysophosphatidylcholine (200 μM in plasma) to yield LPA in an albumin-bound form, mainly, although it is relatively soluble in aqueous media because of its polarity and small size. This lipid is more abundant in serum (1 to 5 μM) than in plasma (100 nM) because its main precursor, lysophosphatidylcholine, is released from activated platelets during coagulation. Autotaxin belongs to the nucleotide pyrophosphatase-phosphodiesterase family. It is also present in cerebrospinal and seminal fluids, as well as in many other tissues, including cancer cell lines from which it was first isolated and characterized. Indeed, the name derives from the finding that it promoted chemotaxis in melanoma cells in an autocrine fashion. It binds to target cells via integrins and heparan sulfate proteoglycans, which may help deliver lysophosphatidic acid to its receptors. Genetic deletion of the enzyme in mice results in aberrant vascular and neuronal development and soon leads to embryonic death. However, the overexpression of autotaxin also causes physical defects and is eventually lethal to embryos.

Figure \(\PageIndex{8}\) shows the pathways for lysophosphatidic acid synthesis.

While autotaxin is the primary source of extracellular lysophosphatidic acid, it is now established that a wide variety of cell types produce it intracellularly through various mechanisms, often using phosphatidic acid derived from other phospholipids as the primary precursor via phospholipase D. For example, the hydrolysis of phosphatidic acid by a phospholipase A₂ (PLA₂) is the main mechanism in platelets. Other cellular enzymes include a phosphatidic acid-selective phospholipase A₁ (PLA₁), which produces sn-2-acyl-lysophosphatidic acid; a monoacylglycerol kinase (utilizing monoacylglycerols produced by lipid phosphate phosphatases); and glycerol-3-phosphate acyltransferase (the first step in phosphatidic acid biosynthesis). In particular, secretory PLA₂-IIA (sPLA2-IIA) can induce the release of LPA from phosphatidic acid exposed on the surface of extracellular vesicles derived from platelets and Ca²⁺-loaded erythrocytes upon stimulation by pro-inflammatory cytokines.

Phosphatidylcholine - Structure and Occurrence

Phosphatidylcholine or 1,2-diacyl-sn-glycero-3-phosphocholine (once given the trivial name 'lecithin') is a neutral or zwitterionic phospholipid over a pH range from strongly acidic to strongly alkaline. It is usually the most abundant phospholipid in animals and plants, accounting for almost 50% of complex lipids. As such, it is a key building block of membrane bilayers. In particular, it makes up a very high proportion of the lipids of the outer leaflet of the plasma membrane in animals. Virtually all the phosphatidylcholine in human erythrocyte membranes is in the outer leaflet.  In the plasma membranes of nucleated cells, 80-90% of this lipid is located on the outer leaflet. Phosphatidylcholine is also the principal phospholipid circulating in plasma, where it is an integral component of the lipoproteins, especially the HDL. On the other hand, it is less common in bacterial membranes, perhaps ~10% of species, but it is absent in the 'model' organisms Escherichia coli and Bacillus subtilis. In animal tissues, some of its membrane functions appear to be shared with the structurally related sphingolipid, sphingomyelin, although the latter has many unique properties.

Figure \(\PageIndex{9}\) shows the structure of phosphatidylcholine

In animal tissues, phosphatidylcholine predominantly exists in the diacyl form. However, small proportions (compared with phosphatidylethanolamine and phosphatidylserine) of alkyl, acyl, and alkenylacyl forms may also be present. In general, animal phosphatidylcholine contains lower proportions of arachidonic and docosahexaenoic acids and higher proportions of C₁₈ unsaturated fatty acids than the other zwitterionic phospholipid, phosphatidylethanolamine. Saturated fatty acids are most abundant in position sn-1, while polyunsaturated components are concentrated in position sn-2. Indeed, C₂₀ and C₂₂ polyenoic acids are exclusively in position sn-2, yet in the brain and retina, the unusual very-long-chain polyunsaturated fatty acids (C₃₀ to C₃₈) of the n-6 and n-3 families occur in position sn-1. Dietary factors influence fatty acid composition, but across animal species, we expect phosphatidylcholine in the same metabolically active tissue to have similar relative distributions of fatty acids between the two positions. Table \(\PageIndex{1}\) lists some representative data.

Some exceptions exist: phosphatidylcholine in some tissues or organelles contains relatively high proportions of di-saturated molecular species. For example, lung phosphatidylcholine in most, if not all, animal species studied to date contains a high proportion (50% or more) of dipalmitoylphosphatidylcholine.

The positional distributions of fatty acids in phosphatidylcholine in representative plants and yeast are listed in Table \(\PageIndex{2}\). In the leaves of the model plant Arabidopsis thaliana, saturated fatty acids are concentrated at sn-1, but monoenoic fatty acids are distributed approximately equally between the two positions, and di- and triunsaturated fatty acids predominate at sn-2; the same is true for soybean lecithin. The pattern is somewhat similar in the yeast Lipomyces lipoferus, except that much of the 16:1 is in position sn-1.

Phosphatidylcholine – Biosynthesis

Several mechanisms support phosphatidylcholine biosynthesis in animals, plants, and microorganisms. Choline itself is not synthesized by animal cells. It is an essential nutrient for phospholipid synthesis and cholinergic neurotransmission (acetylcholine synthesis), and it provides methyl groups for numerous other metabolites. It must be obtained from dietary sources or by degradation of existing choline-containing lipids, for example, those produced by the second pathway described below. Once taken across membranes and into cells by specific transporters, choline is immediately phosphorylated by a choline kinase (1) in the cytoplasm of the cell to produce phosphocholine, which is reacted with cytidine triphosphate (CTP) by the enzyme CTP:phosphocholine cytidylyltransferase (CCT) (2) to form cytidine diphosphocholine (CDP-choline).

CTP + PC → CDP-choline + Pi

The latter enzyme exists in two isoforms, of which CCTα is the more important. It is a soluble protein found first in the nucleoplasm and then in the nucleoplasmic reticulum. This is considered the rate-limiting step in phosphatidylcholine biosynthesis, and the enzyme's activity is regulated by signals from a membrane sensor that reports on the relative abundance of the final product. However, choline kinase (ChoKα) also has regulatory functions.

Figure \(\PageIndex{10}\) shows an interactive iCn3D model of the mammalian (rat) CTP: Phosphocholine cytidylyltransferase catalytic domain (3HL4).

Figure \(\PageIndex{10}\): Mammalian (rat) CTP-Phosphocholine cytidylyltransferase catalytic domain (3HL4). Click the image for a popup or use this external link: https://structure.ncbi.nlm.nih.gov/i...tBYRYk7vPKD7f6

The biologically active homodimer is shown. The B chain is colored by secondary structure, and the A chain is shown in gray. Key active-site residues (shown only in the A chain) are labeled and shown as CPK-colored sticks. A bound CDP-choline analog ((2-cytidylate-O'-phosphonyloxy)-ethyl-trimethylammonium) is shown in spacefill CPK colors.

This enzyme (CCT) catalyzes the key regulatory and rate-limiting step in PC synthesis. The C-terminal domain binds membrane lipids and regulates the enzyme. The iCn3D model above is for the catalytic domain with the regulatory domain deleted. Two nonconserved active site side chains, His 168 and Tyr 173, interact with and position the phosphocholine. Other active site residues include Arg-196 in L6, Lys-122 in L2, and Asp-94. Figure \(\PageIndex{11}\) shows a simplified mechanism for the reaction

In plants, nematodes, and certain parasites, phospho-base N-methyltransferases synthesize most phosphocholine by sequentially methylating phosphoethanolamine. However, only a few plant species methylate phosphatidylethanolamine this way. This is also the main route to free choline and betaine in plants.

The CDP-choline produced is acted upon by the membrane-bound enzyme CDP-choline:1,2-diacylglycerol choline/ethanolamine-phosphotransferase in the endoplasmic reticulum (CEPT1) and a related choline phosphotransferase 1 (CPT1) in the trans-Golgi, which catalyze the reaction with sn-1,2-diacylglycerols to form phosphatidylcholine. The first is responsible for most phosphatidylcholine biosynthesis, but it uses a somewhat different molecular species composition than the second, which prefers 1-alkyl precursors. This is the main pathway for phosphatidylcholine synthesis in animals and plants, and it is analogous to the major route to phosphatidylethanolamine; it is also found in a few bacterial species (e.g., Sinorhizobium meliloti). Phosphatidylcholine in mitochondria is obtained by transfer from the endoplasmic reticulum.

Figure \(\PageIndex{12}\) shows the main pathways for PC synthesis in plants and animals.

The discovery of this pathway's importance depended somewhat on serendipity. In experiments in Eugene Kennedy's laboratory, samples of adenosine triphosphate (ATP) contained a small amount of cytidine triphosphate (CTP) as an impurity. However, luck is of little value without receptive minds, and Kennedy and co-workers demonstrated that the impurity was an important metabolite essential for the formation of phosphatidylcholine.

The reaction above and the biosynthetic mechanism for phosphatidylethanolamine differ significantly from those for phosphatidylglycerol, phosphatidylinositol, and cardiolipin. Both use nucleotides, but with the latter, the nucleotide is covalently linked directly to the lipid intermediate, cytidine diphosphate diacylglycerol. However, a comparable pathway for phosphatidylcholine biosynthesis occurs in bacteria (see below).

The source of the sn-1,2-diacylglycerol precursor, which is also a key intermediate in forming phosphatidylethanolamine and phosphatidylserine, and triacylglycerols, is phosphatidic acid. In this instance, the important enzyme is phosphatidic acid phosphatase (also known as lipin, phosphatidate phosphatase, lipid phosphate phosphatase, or phosphatidate phosphohydrolase).

Figure \(\PageIndex{13}\) shows the biosynthesis of the diacyl precursor of PC

This enzyme is also important for producing diacylglycerols, which are essential intermediates in the biosynthesis of triacylglycerols and phosphatidylethanolamine. Yeasts contain two such enzymes: one Mg²⁺-dependent (PAP1) and the other Mg²⁺-independent (PAP2). Much of the phosphatidic acid phosphatase activity in mammals resides in three related cytoplasmic proteins, lipins-1, -2, and -3. Lipin-1 is found mainly in adipose tissue, while lipin-2 is present mainly in the liver. They are unique among glycerolipid biosynthetic enzymes in that they can move between cellular membranes rather than remain tethered. Of these, lipin-1 is the most important and exists in three isoforms, lipin-1α, lipin-1β, and lipin-1γ, with lipin-1α located mainly in the nucleus and lipin-1β in the cytoplasm. Lipin-1γ is present primarily in the brain.

The second pathway for phosphatidylcholine biosynthesis involves sequential methylation of phosphatidylethanolamine, with S-adenosylmethionine (SAM) as the methyl donor, mono- and dimethylphosphatidylethanolamine as intermediates, and phosphatidylethanolamine N‑methyltransferase as the catalyzing enzyme. A single enzyme (~20 kDa) in two isoforms catalyzes all three reactions in hepatocytes; the main form is located in the endoplasmic reticulum (ER), where it spans the membrane, while the second is found in the mitochondria-associated ER membrane. Yeast has at least two N-methyltransferases. This is a major pathway in the liver, generating one-third of the phosphatidylcholine in this organ but not in other animal tissues or, in general, in higher organisms. It may be the main route to phosphatidylcholine in those bacterial species that produce this lipid and in yeasts, but it appears to operate in only a few species of higher plants. This liver pathway is especially important when dietary choline is deficient.

Figure \(\PageIndex{14}\) shows PC synthesis via PE methylation.

A by-product of phosphatidylcholine biosynthesis from phosphatidylethanolamine is the conversion of S‑adenosylmethionine to S‑adenosylhomocysteine, which the liver hydrolyzes to adenosine and homocysteine. An elevated level of the latter in plasma is a risk factor for cardiovascular disease and myocardial infarction.

Phosphatidylcholine biosynthesis via both pathways in the liver is necessary for normal secretion of plasma lipoproteins (VLDL and HDL) and is relevant to many human physiological conditions. These pathways for diacylphosphatidylcholine biosynthesis differ and are spatially separated from those producing alkyl-, acyl-, and alkenylacyl-phosphatidylcholines de novo. Phosphatidylcholine synthesis also does not occur uniformly throughout the endoplasmic reticulum.  It occurs at membrane interfaces or where membranes meet other organelles, especially during dynamic membrane expansion.

The enzymes in the endoplasmic reticulum that synthesize all phospholipids are oriented, so their active sites face exclusively toward the cytosol. Problems would arise if the cytosolic leaflet expanded rapidly while the luminal leaflet did not, but a phospholipid transporter known as a scramblase enables rapid, bidirectional flip-flop of phospholipids between leaflets of the bilayer in an energy-independent manner. Compositional asymmetry first appears in the trans-Golgi and is completed before the plasma membrane forms, with phosphatidylcholine and sphingolipids present mainly in the exofacial (outer) leaflet.  Phosphatidylethanolamine and phosphatidylserine are enriched in the cytosolic leaflet.

Dietary phosphatidylcholine is rapidly hydrolyzed in the proximal small intestine by pancreatic enzymes with lysophosphatidylcholine (and free fatty acids) formation. Further hydrolysis can occur in the jejuno-ileal brush-border membrane by membrane phospholipases, releasing glycerophosphocholine. Still, much of the lysophosphatidylcholine is reacylated by the lyso-PC-acyl-CoA-acyltransferase 3 for export in chylomicrons.

In plant cells, phosphatidylcholine biosynthesis occurs mainly in the endoplasmic reticulum, and it is a major component of most membranes other than the internal membranes of plastids; it is absent from the thylakoids and the inner envelope membrane but is the main glycerolipid of the outer monolayer of the outer envelope membrane. Further complications arise in plants, in that turnover or partial synthesis via lysophosphatidylcholine occurs in different organelles, using fatty acid pools or enzymes with differing specificities. In addition, fatty acids esterified to phosphatidylcholine serve as substrates for desaturases. As a result, an appreciable pool of diacylglycerols for the biosynthesis of triacylglycerols, galactosyldiacylglycerols, and other glycerolipids passes through phosphatidylcholine as an intermediate, so fatty acid compositions in different membranes change after the initial synthetic process. This mechanism differs from the remodeling of molecular species in animal tissues, which we discuss next. However, a comparable exchange of acyl groups does occur, in part catalyzed by acyltransferases (see the next section). Some phosphatidylcholine transfer per se from the endoplasmic reticulum to plastids may occur via contact points between the two membranes or may be facilitated by specific transport proteins.

While phosphatidylcholine is a major lipid in yeasts, recent work suggests that it is not essential if suitable alternative growth substrates are available, unlike in higher organisms, where perturbation of phosphatidylcholine synthesis can lead to growth inhibition or even cell death.

Remodeling of Phosphatidylcholine - the Lands' cycle

Whatever the mechanism of biosynthesis of phosphatidylcholine in animal tissues, it is apparent that the fatty acid compositions and positional distributions on the glycerol moiety are determined post-synthesis by extensive re-modeling involving orchestrated reactions of hydrolysis (phospholipase A₂ mainly) to lysophosphatidylcholine, acyl-CoA synthesis and re-acylation by lysophospholipid acyltransferases or transacylases, a series of reactions that is sometimes termed the 'Lands' Cycle' after its discoverer W.E.M. (Bill) Lands. Similar processes occur with all glycerophospholipid classes.

The final lipid composition results from a mixture of de novo synthesis and remodeling. At least 15 groups of enzymes in the phospholipase A₂ superfamily differ in calcium dependence, cellular location, and structure. All hydrolyze the sn-2 ester bond of phospholipids specifically, generating a fatty acid and a lysophospholipid, which have important functions beyond their role in the Lands cycle. A phospholipase A1 family of enzymes are esterases that can cleave the sn-1 ester bond but are less important in this context.

Figure \(\PageIndex{15}\) shows Land's cycle

Membrane-bound, coenzyme A-dependent lysophosphatidylcholine acyltransferases such as LPCAT3 (also designated ‘MBOAT5’) catalyze the re-acylation step; it is located chiefly in the endoplasmic reticulum, but also in mitochondria and the plasma membrane in organs such as the liver, adipose tissue, and pancreas. It maintains systemic lipid homeostasis by regulating lipid absorption and composition in the intestines, lipoprotein secretion, and de novo lipogenesis in the liver; it also incorporates linoleoyl and arachidonoyl chains specifically into lysophosphatidylcholine (as does the related enzyme LPCAT2). Inflammatory cells also express a CoA-independent acyltransferase that transfers arachidonic acid from phosphatidylcholine to ethanolamine-containing phospholipids. While LPCAT3 prefers 1-acyl lysophosphatidylcholine as an acyl acceptor, LPCAT2 utilizes 1-acyl and 1-alkyl precursors. LPCAT2 is highly expressed in inflammatory cells such as macrophages and neutrophils, which contain ether-phospholipids and contribute to the production of eicosanoid lipid mediators. The highly saturated molecular species of phosphatidylcholine found in lung surfactant are formed from species with a more conventional composition by remodeling by an acyltransferase with a high specificity for palmitoyl-CoA acid (LPCAT1). In other tissues, those species containing high proportions of polyunsaturated fatty acids depend more on de novo synthesis. These and further related enzymes are involved in remodeling all other phospholipids. Overexpression of the genes for these enzymes is associated with the progression of many different cancers and may be involved in other pathological conditions.

Phosphatidylcholine plays a central role in plant glycerolipid metabolism, and remodeling occurs for different reasons and by mechanisms that differ from those in animal cells, as briefly described above. For example, extensive remodeling involves fatty acid desaturation, the main entry point for acyl groups exported from the plastid into the endoplasmic reticulum. In addition, phosphatidylcholine remodeling provides fatty acids for triacylglycerol synthesis in developing seeds and diacylglycerols for synthesizing thylakoid lipids such as galactosyldiacylglycerols. In Arabidopsis, two lysophosphatidylcholine acyltransferases, LPCAT1 and LPCAT2, remodel phosphatidylcholine in developing seeds and leaves, preferring the sn-2 position and using fatty acids exported from the plastid. In some plant species, there is a strong preference for C₁₈‑unsaturated acyl chains over 16:0. However, the lipases that generate lysophosphatidylcholine from phosphatidylcholine for this purpose are unknown. Some remodeling in plant membranes occurs in response to stress.

The yeast Saccharomyces cerevisiae can reacylate glycerophosphocholine, generated endogenously by the action of phospholipase B (an enzyme with both phospholipase A₁ and A₂ activities) on phosphatidylcholine, with acyl-CoA in the microsomal membranes using a glycerophosphocholine acyltransferase (Gpc1) to produce lysophosphatidylcholine, which can be converted back to phosphatidylcholine by the lysophospholipid acyltransferase (Ale1) with appreciable changes in the molecular species composition. This process is regulated in coordination with other major lipid pathways and affects yeast growth. The enzyme Gpc1 does not affect other phospholipids in yeast. A similar mechanism appears to operate in some plant species. Figure \(\PageIndex{16}\) shows the Land's cycle variants.

Catabolism

Phosphatidylcholine (and most other glycerophospholipids) in membranes can be metabolized by lipolytic enzymes, especially phospholipases, some of which are specific for particular lipid classes in humans. For example, in addition to the action of phospholipase A (discussed above), phospholipase C (six families in mammals differing in expression and subcellular distribution) yields diacylglycerols together with phosphocholine by hydrolyzing glycerophospholipids at the phosphodiester bond. This process is especially important in phosphoinositide metabolism. The sphingomyelin synthases also have phospholipase C activity (in the absence of ceramide). Phospholipase D generates phosphatidic acid and choline, while phospholipase B removes both fatty acids to yield glycerophosphocholine.

Figure \(\PageIndex{17}\) shows the activities of phospholipases on phosphatidylcholine.

In this way, catabolism recycles lipid components or confers signaling functions. At the same time, much of the choline is reused for phosphatidylcholine biosynthesis, often after being returned to the liver (the CDP-choline cycle). Some choline is oxidized in the kidney and liver to betaine, which serves as a methyl group donor for S-adenosylmethionine production, and some is lost through the excretion of phosphatidylcholine in bile. A proportion is used in nervous tissues to produce acetylcholine, a neurotransmitter important to learning, memory, and sleep. The liver takes up phosphatidylcholine from plasma high-density lipoproteins and, perhaps surprisingly, eventually converts a large proportion of it to triacylglycerols via diacylglycerol intermediates.

Phosphatidylethanolamine – Structure and Occurrence

Phosphatidylethanolamine, or 1,2-diacyl-sn-glycero-3-phosphoethanolamine (once given the trivial name 'cephalin'), is usually the second most abundant phospholipid in animal and plant lipids, after phosphatidylcholine, and is frequently the main lipid component of microbial membranes. It can account for 20% of liver phospholipids and as much as 45% of those in the brain; mitochondria contain higher proportions than other organelles. As such, it is a key building block of membrane bilayers, and it is present exclusively in the inner leaflet of the plasma membrane in animal cells, for example. It is a neutral or zwitterionic phospholipid (at least in the pH range 2 to 7), with the structure shown (a specific molecular species is illustrated as an example).

Figure \(\PageIndex{19}\) shows the structure of phosphatidylethanolamine.

In animal tissues, phosphatidylethanolamine exists in diacyl, alkyl, alkenyl, and acyl forms. In some cell types (especially inflammatory cells, neurons, and tumor cells), up to 70% of phosphatidylethanolamine may have an ether linkage. Still, in the liver, the plasmalogen form of phosphatidylethanolamine, i.e., with an O‑alk-1’-enyl linkage, accounts for only 0.8% of total phospholipids. Generally, PE has a much higher proportion of ether linkages than PC. If physiological conditions inhibit plasmalogen biosynthesis, the diacyl form replaces it, thereby maintaining overall phospholipid content.

Animal phosphatidylethanolamine tends to contain higher proportions of arachidonic and docosahexaenoic acids than the other zwitterionic phospholipid, phosphatidylcholine. These polyunsaturated components are concentrated in position sn-2, with saturated fatty acids most abundant in position sn-1, as illustrated for rat liver and chicken egg in Table \(\PageIndex{3}\). In most other species, phosphatidylethanolamine in the same metabolically active tissues would be expected to show similar features.

The O-alkyl and O-alkenyl chains at the sn-1 position of the analogous ether lipids generally consist of 16:0, 18:0 or 18:1 chains. In contrast, arachidonic and docosahexaenoic acids are the most abundant components at the sn-2 position.

The positional distributions of fatty acids in phosphatidylethanolamine from the model plant Arabidopsis thaliana leaves are listed in Table \(\PageIndex{4}\). Here, saturated fatty acids are also concentrated at position sn-1, and di- and tri-unsaturated fatty acids predominate at position sn-2. The pattern differs somewhat in the yeast Lipomyces lipoferus, where the compositions of the two positions are relatively similar.

Phosphatidylethanolamine – Biosynthesis

The two main pathways employed by mammalian cells for the biosynthesis of phosphatidylethanolamine are the CDP-ethanolamine pathway, i.e., one of the general routes to phospholipid biosynthesis de novo in plants and animals, and the phosphatidylserine decarboxylase pathway, which occurs in two spatially separated organelles - the endoplasmic reticulum and mitochondria, respectively. Plants produce ethanolamine by decarboxylating serine, whereas in animals most comes from dietary sources and requires facilitated transport into cells. A small amount of ethanolamine phosphate comes from the catabolism of sphingosine-1-phosphate, which is essential for the survival of the protozoan Trypanosoma brucei. The initial steps in phosphatidylethanolamine biosynthesis occur in the cytosol, where two specific ethanolamine kinases phosphorylate ethanolamine to produce ethanolamine phosphate; ethanolamine phosphate phosphorylase catalyzes the reverse reaction and may regulate the pathway in some tissues. The second step is rate-limiting: the product reacts with cytidine triphosphate (CTP) to form cytidine diphosphoethanolamine, catalyzed by CTP:phosphoethanolamine cytidylyltransferase.

In the final step, a membrane-bound enzyme, CDP-ethanolamine:diacylglycerol ethanolaminephosphotransferase, catalyzes the reaction of cytidine diphosphoethanolamine with diacylglycerol to form phosphatidylethanolamine. Two such enzymes exist: ethanolamine phosphotransferase 1 (EPT1) in the Golgi and choline/EPT1 (CEPT1) in the endoplasmic reticulum, but EPT1 is more important for the biosynthesis of the plasmalogen form, 1-alkenyl-2-acyl-glycerophosphoethanolamine, especially molecular species containing polyunsaturated fatty acids. At the same time, CEPT1 produced species with shorter-chain fatty acids. The diacylglycerol precursor is formed from phosphatidic acid by the enzyme phosphatidic acid phosphohydrolase.

Figure \(\PageIndex{20}\) shows the synthesis of phosphatidylethanolamine in the ER/Golgi.

The second major pathway is the mitochondrial conversion of phosphatidylserine (decarboxylation) to phosphatidylethanolamine. The conservation of this pathway from bacteria to humans suggests that it has been preserved to optimize mitochondrial performance. Mitochondria are not connected with the rest of the cell's membrane network by classical vesicular routes but must receive and export small molecules through nonvesicular transport at zones of proximity with other organelles at membrane contact sites, such as a specific domain of the endoplasmic reticulum termed the mitochondria-associated membrane (MAM). Lipid transport of phosphatidylserine is then enabled by tethers that bridge two membranes, lipid transfer proteins, and recruitment proteins. In this process, the lipid must traverse two aqueous compartments, the cytosol and the mitochondrial intermembrane space, to reach the inner mitochondrial membrane.

Phosphatidylserine decarboxylase is located on the external surface of the mitochondrial inner membrane, and most mitochondrial phosphatidylethanolamine is derived from this pathway. While various isoforms of phosphatidylserine decarboxylase exist in prokaryotes, yeasts, and mammals, the main forms designated 'PSD1' are found only in mitochondria and are structurally related. An isoform designated 'PSD2' is located in the endosomal membranes of yeasts, and the phosphatidylethanolamine formed is the preferred substrate for phosphatidylcholine biosynthesis. Cellular concentrations of phosphatidylethanolamine and phosphatidylserine are closely related and tightly regulated.

Figure \(\PageIndex{21}\) shows the synthesis of phosphatidylethanolamine in mitochondria.

In prokaryotic cells, such as E. coli, where phosphatidylethanolamine is the most abundant membrane phospholipid, it is derived entirely from phosphatidylserine decarboxylation. In this instance, the enzyme undergoes autocleavage for activation and utilizes a pyruvoyl moiety to form a Schiff base intermediate with phosphatidylserine, facilitating decarboxylation. This pathway is also important in mammalian cells and yeasts. However, the relative contributions of the two main pathways of phosphatidylethanolamine synthesis in mammalian cells appear to vary with cell type. In cultured cells, more than 80% of phosphatidylethanolamine is reported to be derived from the phosphatidylserine decarboxylase pathway.  In hamster hearts and rat hepatocytes, only ~5% of phosphatidylethanolamine is synthesized via this route, and most is made via the CDP-ethanolamine pathway. In yeast, PSD1 generates 70% of phosphatidylethanolamine. The spatially distinct pools within the cell are functionally distinct, but both are essential to life. For example, disruption of the phosphatidylserine decarboxylase gene causes misshapen mitochondria and has lethal consequences in embryonic mice, although phosphatidylethanolamine synthesis continues for a time in other cellular regions. Similarly, eliminating the endoplasmic reticulum route is embryonically lethal.

Three additional minor biosynthetic pathways are known. Phosphatidylethanolamine can form by enzymatic exchange of ethanolamine with phosphatidylserine or by re-acylation of lysophosphatidylethanolamine. The second is associated with the mitochondria-associated membrane, where phosphatidylserine synthase II is located. Finally, the bacterial plant pathogen Xanthomonas campestris can synthesize phosphatidylethanolamine by condensing cytidine diphosphate diacylglycerol with ethanolamine.

Ether lipids: These pathways for diacyl-phosphatidylethanolamine biosynthesis are distinct and spatially separated from those producing alkyl-, acyl-, and alkenyl-acyl-phosphatidylethanolamines, suggesting functional differences. In the protozoan T. brucei, for example, diacyl and ether phosphatidylethanolamine pools have separate functions and cannot substitute for each other.

Lands cycle: The various mechanisms produce different pools of phosphatidylethanolamine species, often in different cellular compartments with distinctive compositions. Studies with mammalian cell types in vitro suggest that the CDP-ethanolamine pathway preferentially produces molecular species with mono- or di-unsaturated fatty acids at the sn-2 position. At the same time, the phosphatidylserine decarboxylation reaction mainly generates species with polyunsaturated fatty acids at the sn-2 position. However, as with other phospholipids, animal tissues attain their final fatty acid composition through a remodeling process known as the Lands cycle. The first step is hydrolysis by a phospholipase A₂ to lysophosphatidylethanolamine, followed by reacylation using various acyl-CoA:lysophospholipid acyltransferases. The enzymes LPCAT1, 2, and 3, involved in phosphatidylcholine biosynthesis, are also active on phosphatidylethanolamine, while LPEAT1 mainly uses lysophosphatidylethanolamine and is specific for oleoyl-CoA. Some of these isoforms appear to be confined to particular tissues. A CoA-independent acyltransferase in macrophages also transfers arachidonic acid from phosphatidylcholine to ethanolamine-containing phospholipids.

Phosphatidylserine - Structure and Occurrence

Although phosphatidylserine is widely distributed among animals, plants, and microorganisms, it usually accounts for less than 10% of total phospholipids, with the highest concentration in myelin from brain tissue. For example, the brain and liver of mice contain 14% and 3% phosphatidylserine, respectively. However, it may comprise 10 to 20 mol% of total phospholipids in the plasma membrane, where, under normal conditions, it is concentrated in the inner leaflet and in the endoplasmic reticulum. In the yeast Saccharomyces cerevisiae, it is a minor component of most cellular organelles other than the plasma membrane, where it can, surprisingly, account for more than 30% of total lipids. It is a minor membrane constituent in most bacteria, although it is important as an intermediate in phosphatidylethanolamine biosynthesis. The 1‑octadecanoyl-2-docosahexaenoyl molecular species, especially important in brain tissue, is illustrated here.

Figure \(\PageIndex{24}\) shows the structure of phosphatidylserine

Phosphatidylserine is an acidic (anionic) phospholipid with three ionizable groups, i.e., the phosphate moiety, the amino group, and the carboxyl function. As with other acidic lipids, it occurs naturally as a salt. However, it strongly tends to chelate calcium via the charged oxygen atoms of both the carboxyl and phosphate moieties, thereby altering the conformation of the polar head group. This interaction may be highly relevant to phosphatidylserine's biological function, especially during bone formation.

In animal cells, the fatty acid composition of phosphatidylserine varies by tissue. Still, it does not appear to resemble the precursor phospholipids, either because specific molecular species are selectively used for biosynthesis or because lipid remodeling via deacylation-reacylation reactions with lysophosphatidylserine as an intermediate occurs (see below). In human plasma, 1-stearoyl-2-oleoyl and 1-stearoyl-2-arachidonoyl species predominate, but in the brain (especially grey matter), retina, and many other tissues, 1-stearoyl-2-docosahexaenoyl species are especially abundant and appear essential for normal nervous system function. Indeed, the n-3-to-n-6 fatty acid ratio in brain phosphatidylserine is much higher than in most other lipids. The positional distribution of fatty acids in phosphatidylserine from rat liver and bovine brain is listed in Table \(\PageIndex{5}\). As with most phospholipids, saturated fatty acids are concentrated at sn-1 and polyunsaturated fatty acids at sn-2.

In the leaves of Arabidopsis thaliana, used as a 'model' plant in many studies, the fatty acid composition of phosphatidylserine resembles that of phosphatidylethanolamine. An intriguing report states that the acyl chain lengths of phosphatidylserine increase with age and stress, and that 22:0 and 24:0 fatty acids have been reported to occur in this lipid in the plasma membrane of some plant species.

Compared to PE, PS with ether-linked moieties (alkyl and alkenyl) is uncommon in animal tissues. However, it is reported to be relatively abundant in the human retina and macrophages (first found in rat lung). In general, phosphatidylserine is most concentrated in plasma membranes and endosomes but is very low in mitochondria. Because it lies entirely on the inner monolayer of the plasma membrane (and other cellular membranes) and is the most abundant anionic phospholipid, it may contribute to interfacial effects through nonspecific electrostatic interactions. This normal distribution is disturbed during platelet activation and cellular apoptosis.

N-Acylphosphatidylserine is reportedly present in the frontal cortex of patients with schizophrenia as a minor component of the lipids of sheep erythrocytes, bovine brain, and the central nervous system of freshwater fish and Bryozoans, amongst others. The N-arachidonoyl form may be the precursor of the endocannabinoid N-arachidonoylserine.

Biosynthesis of Phosphatidylserine

L-serine is a non-essential amino acid that most organisms actively synthesize. In animals, nearly all cell types produce it, though in the brain astrocytes synthesize it, not neurons. Neurons must be supplied with this amino acid for the biosynthesis of phosphatidylserine (and of sphingoid bases).

In animal tissues, phosphatidylserine is synthesized solely by calcium-dependent base-exchange reactions in which the polar head-group of an existing phospholipid is exchanged for L-serine. Two routes involve distinct enzymes (PS synthases I and II) that share 30% homology and several membrane-spanning domains and can use different substrates. Both enzymes synthesize phosphatidylserine on the cytosolic face of the endoplasmic reticulum (ER), mainly in a specific domain termed the mitochondria-associated membrane ('MAM'), because it is transiently tethered to the mitochondrial outer membrane, presumably by a protein bridge. A complex of integrated proteins ('ERMES') has been characterized in yeast with a similar function. The reaction involves exchanging L-serine with either phosphatidylcholine or phosphatidylethanolamine, catalyzed by PS synthase I (long thought to accept only phosphatidylcholine as a substrate).

In contrast, PS synthase II catalyzes a similar exchange with diacyl-phosphatidylethanolamine and its plasmalogen form. Both enzymes are subject to feedback regulation by their product phosphatidylserine, thereby maintaining the correct amounts of this lipid. Figure \(\PageIndex{25}\) shows the synthesis and metabolism of phosphatidylserine in animals.

Phosphatidylserine synthase I is expressed in all mouse tissues, especially the kidney, liver, and brain, while phosphatidylserine synthase II is most active in the brain and testis and much less so in other tissues. The latter enzyme has a high specificity for molecular species containing docosahexaenoic acid. It is unknown why such a complex series of coupled reactions is necessary or why two enzymes are required. Still, one virtue is that the free ethanolamine and choline formed are rapidly re-utilized for phospholipid synthesis. Thus, phosphatidylserine and phosphatidylethanolamine are produced without reducing phosphatidylcholine levels. Eliminating both enzymes is embryonically lethal in knockout mice. Still, each can be knocked out separately, and the mice survive, even though they have substantially reduced levels of phosphatidylserine and phosphatidylethanolamine.

As with other phospholipids, animal tissues attain their final fatty acid composition through remodeling, known as the Lands cycle. The first step is hydrolysis by a phospholipase A₂ to lysophosphatidylserine, followed by reacylation by various acyl-CoA:lysophospholipid acyltransferases. One membrane-bound O-acyltransferase (LPCAT4 or MBOAT2) prefers oleoyl-CoA, while a second (LPCAT3 or MBOAT5) incorporates linoleoyl and arachidonoyl chains (and also uses lysophosphatidylcholine).

Following synthesis, phosphatidylserine molecules can diffuse laterally in a concentration-dependent manner to different membrane regions to fulfill their physiological functions. In humans, cytosolic transport proteins transfer phosphatidylserine and other acidic phospholipids between membranes, and this can also occur via vesicular transport.

Some of the newly synthesized phosphatidylserine is transferred to the plasma membrane. At the same time, a proportion is transported to the mitochondria, probably again via transient membrane contact (MAM), where a specific decarboxylase in the inner mitochondrial membrane decarboxylates it to produce phosphatidylethanolamine. In yeast, molecular species containing two monoenoic fatty acids are preferred for transport and metabolism; this process also occurs at the Golgi/endosome membranes. All the phosphatidylethanolamine in mitochondria is formed this way. Still, some can return to the endoplasmic reticulum, where it may be converted back to phosphatidylserine by the action of the PS synthases. The CDP-ethanolamine pathway does not fully complement the mitochondrial production of phosphatidylethanolamine from phosphatidylserine, as mice lacking the enzyme do not survive for long. Cellular concentrations of these two lipids are intimately related and tightly regulated. Figure \(\PageIndex{26}\) shows the mitochondrial metabolism of phosphatidylserine

Thus, much of the phosphatidylserine is decarboxylated to phosphatidylethanolamine, which may be the major route to the latter in bacteria. As phosphatidylcholine in yeast is produced via methylation of phosphatidylethanolamine, phosphatidylserine is the primary precursor for this phospholipid in these organisms.

Bacteria and plants: In bacteria, other prokaryotic organisms and yeast, phosphatidylserine is synthesized by a mechanism comparable to that of most other phospholipids, i.e., by reaction of L-serine with CDP-diacylglycerol, and depends on Mg²⁺ or Mn²⁺. Phosphatidylserine synthases belong to two different families: type I (non-integral membrane form) in the phospholipase D-like family, as in E. coli, and type II (integral membrane form) in the CDP-alcohol phosphotransferase family, as in Bacillus sp. and the yeast S. cerevisiae. However, the latter shows no homology with the bacterial enzymes.

In many plants, including the model plant Arabidopsis, much phosphatidylserine is produced by a calcium-dependent base-exchange reaction in which the head group of an existing phospholipid is exchanged for L-serine in the luminal leaflet of the endoplasmic reticulum (i.e., mechanistically similar to PS synthase I). Flippases transfer it to the cytoplasmic membrane leaflet, then to post-Golgi compartments, before it eventually accumulates at the plasma membrane. However, some vesicular transport may occur, or direct transfer may occur at membrane contact sites. A CDP-diacylglycerol (prokaryotic-like) biosynthetic pathway exists in some species, e.g., wheat.

Let's explore the mechanism of the Methanocaldococcus jannaschii phosphatidylserine synthase (MjPSS). The organism is a hyperthermophilic methanogen. Figure \(\PageIndex{27}\) shows substrate binding by MjPSS.

The large binding pocket for CDP-DAG in MjPSS extends from the hydrophobic membrane core to the active site near the cytoplasmic surface in the center of the dimer. In the closed structures (left), both CDP-DAG alkyl chains adopt similar conformations within the binding pocket. In contrast, helices 7 and 8 positions in the open structures (right) allow one alkyl chain to reach the membrane via a different path. Serine molecules are only found in open structures (right). In one closed conformation, a citrate is near the substrate-binding site, whereas in the other closed structures, this position is empty. A chain of three chloride ions extends parallel to the dimer interface from the active site to the cytoplasmic interface with the N-terminal helix hH from the other protomer.

Figure \(\PageIndex{28}\) shows the reaction cycle of MjPSS during the synthesis of PS from CDP-DAG and serine in the presence of Mg²⁺ and Ca²⁺.

The CDPDAG binding site in MjPSS forms and is stabilized by the divalent cations Mg²⁺ and Ca²⁺ (a). In the absence of CDP-DAG, Ca²⁺ most likely is coordinated by water molecules, as shown in a, or by residues in nearby loops that would be flexible in the absence of CDP-DAG. CDP-DAG binding (b) is driven by coordination of Ca²⁺ by the negatively charged phosphates. Serine binds to the binding pocket after CDP-DAG (c). For nucleophilic attack, the serine molecule positions its hydroxyl group near the β-phosphate of CDP-DAG. The serine molecule is activated by deprotonation, attacks the β-phosphate, and forms the penta-coordinated transition state (d). A water molecule in the interaction network of Asp66, Arg101, and nearby chloride ions likely removes the serine proton. Hydrolysis of the CDP-DAG/serine complex at the transition state forms the MjPSS complex with the products PS and CMP (e). The next cycle starts after product release and CDP-DAG binding. Structural data are available for the state with bound CDP-DAG (b), CDP-DAG and serine (c), and for the transition state of the CDP-DAG/serine complex (d).

Figure \(\PageIndex{29}\) shows an interactive iCn3D model of the Methanocaldococcus jannaschii phosphatidylserine synthase (PSS) in the open state with bound CDP-DAG and serine (7B1L). The enzyme is also named CDP-diacylglycerol--serine O-phosphatidyltransferase.

Figure \(\PageIndex{29}\): Bacterial phosphatidylserine synthase (PSS) in the open state with bound CDP-DAG and serine (7B1L) (Note the actual PDB file title is for the closed state.) Click the image for a popup or use this external link: https://structure.ncbi.nlm.nih.gov/i...B3Rn3gVN4c4ge6

The A chain of the homodimer is shown in gray and the B in plum. The active site residues in the A chain in the above mechanism are shown in stick representation, with CPK colors and labels. Hover over the large ligand (58A) in the gray subunit. Two free serines are shown near it, but only one is probably the substrate.

Phosphatidylinositol

Structure and Occurrence: Phosphatidylinositol is an important lipid, both as a membrane constituent and as a participant in essential metabolic processes in all plants and animals, directly and via several metabolites. It is an acidic (anionic) phospholipid with a phosphatidic acid backbone, linked via the phosphate group to inositol (hexahydroxycyclohexane). In most organisms, the stereochemical form is myo-D-inositol (with one axial hydroxyl at position 2 and the remainder equatorial, i.e., a chair-like structure). However, other forms (scyllo- and chiro-) have occasionally been found in plants. The 1‑stearoyl,2-arachidonoyl molecular species, which is biologically important in animals, is illustrated.

Figure \(\PageIndex{31}\) shows the structure of phosphatidylinositol.

Phosphatidylinositol is especially abundant in brain tissue, where it can account for 10% of phospholipids.  It is present in all tissues, cell types, and membranes at relatively low levels compared to many other phospholipids. In rat liver, it amounts to 1.7 micromoles/g, i.e., less than phosphatidylcholine, phosphatidylethanolamine, and phosphatidylserine. Under normal conditions, it is present entirely in the inner leaflet of the erythrocyte membrane and in the plasma membrane of nucleated cells. Phosphatidylinositol per se is rarely found in prokaryotes other than the Actinomycetales, although the thermophilic α-proteobacterium Rhodothermus marinus contains dialkylether glycerophosphoinositides.

The fatty acid composition of phosphatidylinositol is rather distinctive, as shown in Table \(\PageIndex{6}\). Thus, in almost all animal tissues, the characteristic feature is a high content of stearic and arachidonic acids. All stearic acid is linked to position sn-1, and all arachidonic acid to position sn-2. As much as 78% of the total lipid may consist of the single molecular species sn-1-stearoyl-sn-2-arachidonoyl-glycerophosphorylinositol (see Table \(\PageIndex{7}\) below). Although 1-alkyl- and alkenyl- forms of phosphatidylinositol are known, they are much less abundant than the diacyl form. In plant phosphatidylinositol (e.g., Arabidopsis thaliana), palmitic acid is the main saturated fatty acid at sn-1, while linoleic and linolenic acids are the main unsaturated components at sn-2. Similarly, in yeast, palmitic acid is predominantly at sn-1, with oleic and palmitoleic acids at sn-2; in Amoebozoa, a C16 alkyl group is at sn-1 and cis-vaccenic acid at sn-2.

Biosynthesis: The basic mechanism for the biosynthesis of phosphatidylinositol and phosphatidylglycerol is sometimes termed a branch point in phospholipid synthesis, as phosphatidylcholine and phosphatidylethanolamine are produced by a somewhat different route.

Phosphatidylinositol is found in all eukaryotes, which can generally synthesize inositol de novo from glucose-6-phosphate. As with phosphatidylglycerol (and hence cardiolipin), phosphatidylinositol is formed biosynthetically from phosphatidic acid via the intermediate cytidine diphosphate diacylglycerol, which is produced by the action of a CDP-diacylglycerol synthase believed to be the rate-limiting enzyme in phosphatidylinositol biosynthesis. Then, the enzyme CDP-diacylglycerol inositol phosphatidyltransferase ('phosphatidylinositol synthase' or 'PIS') catalyzes a reaction with myo-inositol to produce phosphatidylinositol.

Figure \(\PageIndex{32}\) shows the synthesis of phosphatidylinositol in eukaryotes.

In mammals, PIS is located in the endoplasmic reticulum, in part in a subcompartment associated with mitochondria (mitochondria-associated membranes, MAM) and in mitochondria per se. Indeed, PIS is present in a mobile ER-derived subcompartment that transiently contacts other organelles, including the plasma membrane, and facilitates the distribution of phosphatidylinositol to other subcellular compartments. The other product is cytidine monophosphate (CMP). Because PIS catalyzes the reverse reaction, the phosphatidylinositol synthesis rate depends on the relative concentrations of the precursors and product, and the latter must be transported away from the site of synthesis for the reaction to continue. Vesicular transport delivers much of the phosphatidylinositol to other membranes.  A family of soluble phosphatidylinositol transfer proteins (PITPα, PITPβ, and PITPNC1) provides phosphatidylinositol from the ER to kinases for phosphorylation (see below).

Molecular species specificity: The phosphatidylinositol synthase per se does not exhibit the fatty acyl specificity observed in the final product, but earlier in the biosynthetic process, 1-stearoyl-2-arachidonoyl species of diacyl-sn-glycerols are converted preferentially into phosphatidic acid by the epsilon isoform of diacylglycerol kinase (DGKε), anchored to the membrane via its N-terminal hydrophobic helix segment; ATP is the phosphate donor. In addition, one CDP-diacylglycerol synthase (CDS2) has similar specificity and generates the immediate precursor CDP-diacylglycerol from phosphatidic acid.  Some specificity may be introduced via lysophosphatidylinositol, formed as a by-product of eicosanoid formation (see below) or as an intermediate as part of the normal cycle of deacylation-acylation of phosphatidylinositol in tissues in which the fatty acid composition is remodeled to give the final distinctive composition. A membrane-bound O-acyltransferase (MBOAT7 or LPIAT1) specific for position sn-2 of lysophosphatidylinositol, with a marked preference for arachidonoyl-CoA, is ubiquitously expressed in animal tissues and may regulate free arachidonic acid and eicosanoid levels.

In macrophages subjected to inflammatory stimuli, remodeling reactions produce phosphatidylinositol containing two molecules of arachidonate, and evidence suggests it is a novel bioactive phospholipid that regulates innate immune responses in these cells. Further specificity may be introduced by lysocardiolipin acyltransferase (LYCAT; also known as LCLAT1 or ALCAT1), which exhibits a preference for lysophosphatidylinositol and lysophosphatidylglycerol over other phospholipids in vitro, and incorporates 18:0 rather than shorter chain fatty acids into position sn-1 of phosphatidylinositol and other phosphoinositides, especially phosphatidylinositol-4,5-bisphosphate and phosphatidylinositol-3-phosphate; this enzyme may be located adjacent to the phosphatidylinositol synthase in the endoplasmic reticulum. Some of the phosphatidylinositol in membranes is derived from recycling polyphosphoinositides via the phosphatidylinositol cycle, which could influence the molecular species composition (see below).

The highly specific distribution of fatty acids on the glycerol moiety of phosphatidylinositol is disrupted in some cancer cells, especially those with a mutation in the p53 gene, one of the most frequently mutated genes in cancer.

Plants and bacteria: In contrast to animals, plants have two phosphatidylinositol synthase isoforms, PIS1 and PIS2, which display specificities for particular species of the CDP-diacylglycerol substrate. PIS1 preferentially generates phosphatidylinositol with saturated or monounsaturated fatty acids, while PIS2 generates polyunsaturated species; the two forms may have different functions. In protozoan parasites, such as Trypanosoma brucei, phosphatidylinositol synthase's active site may be located in the endoplasmic reticulum and the Golgi lumen. Evidence suggests two distinct product pools in this organism: a bulk membrane form derived from inositol imported from the environment, and a second used to synthesize GPI anchors, which uses myo-inositol synthesized de novo. In yeast, some biosynthesis may occur on the cytosolic side of the plasma membrane.

The enzyme is a transmembrane protein and uses CDP-diacylglycerol as a donor and either inositol (eukaryotes) or inositol phosphate (prokaryotes) as the acceptor alcohol. The structure of a similar enzyme, phosphatidylinositol-phosphate synthase from Renibacterium salmoninarum, is shown in Figure \(\PageIndex{33}\).

Figure \(\PageIndex{33}\): Structure and reaction of phosphatidylinositol-phosphate synthase from Renibacterium salmoninarum. Clarke, O., Tomasek, D., Jorge, C. et al. Structural basis for phosphatidylinositol-phosphate biosynthesis. Nat Commun 6, 8505 (2015). https://doi.org/10.1038/ncomms9505. Creative Commons Attribution 4.0 International License. http://creativecommons.org/licenses/by/4.0/

Panel A shows the reaction for PIP synthases, which transfer a diacylglycerol-substituted phosphate group (purple/red) from the CDP-DAG donor to the inositol phosphate acceptor (green), generating PIP and CMP. Panel B shows the structure of the RsPIPS-Δ6N homodimer in ribbon representation viewed from two orthogonal orientations (in the plane of the membrane on the left; towards the cytosol down the dimer axis on the right). One protomer is colored grey, and the helices of the other are depicted in spectral coloring, from blue (JM1) to red (TM6). The Af2299 extramembrane domain used to facilitate crystallization is not shown here.

Figure \(\PageIndex{34}\) shows a large cavity that contains the active site of RsPIPS.

Panel (a) shows the structure of RsPIPS-Δ6N in ribbon representation, with one protomer colored grey and the other colored by the Kyte–Doolittle hydrophobicity scale, from −4.5 (most polar, light blue) to 4.5 (most hydrophobic, orange). Two orthogonal representations are shown; on the left is a view in the plane of the membrane, and on the right is a view from the cytosol along the dimer axis. A transparent purple surface delineates the borders of the interfacial cavity, which contains three subregions: 1, the inositol phosphate acceptor-binding pocket; 2, the nucleotide-binding pocket between TM2 and TM3; and 3, a hydrophobic groove between TM2 and JM1. (b) Detail of the active site viewed in the plane of the membrane, with side chains that contact the bound Mg²⁺ and SO₄^2- ions labeled and depicted in stick representation.

A nucleotide-binding site formed from transmembrane segments 1, 2, and 3 contains 8 conserved residues (D₁xxD₂G₁xxAR…G₂xxxD₃xxxD₄). The first three aspartic acid side chains coordinate a metal ion, while the fourth likely acts as a general base in catalysis.

Figure \(\PageIndex{35}\) shows an interactive iCn3D model of the phosphatidylinositolphosphate (PIP) synthase with bound CDP-DAG from Renibacterium salmoninarum (5D92).

Figure \(\PageIndex{35}\): Phosphatidylinositolphosphate (PIP) synthase with bound CDP-DAG from Renibacterium salmoninarum (5D92).. Click the image for a popup or use this external link: https://structure.ncbi.nlm.nih.gov/i...i9nhSw9U32Man8

The two identical subunits of the homodimer are shown in gray and plum. The two CDP-DAGs are shown in spacefill and CPK colors.

Function: In addition to functioning as negatively charged building blocks of membranes, the inositol phospholipids (including the phosphatidylinositol phosphates or 'polyphosphoinositides' discussed below) have crucial roles in the interfacial binding of proteins and the regulation of protein activity at the cell interface. Because phosphoinositides are polyanionic, they can engage in strong, nonspecific electrostatic interactions with proteins. However, they bind especially efficiently to so-called ‘PH’ domains of cellular proteins. For example, at least three phosphatidylinositol molecules are present in the crystal structure of human erythrocyte glycophorin, and they are believed to influence binding to other proteins via their head groups. The lipid is a structural component of the yeast cytochrome bc1 complex.

In animal tissues, phosphatidylinositol is the primary source of arachidonic acid required for eicosanoid biosynthesis, including prostaglandins, via phospholipase A2, which releases fatty acids from the sn-2 position. The reverse reaction also occurs.

Figure \(\PageIndex{36}\)s shows the generation of arachidonic acid and eicosanoids from PI using phospholipase A2

Similarly, phosphatidylinositol and its phosphates are the main sources of diacylglycerol, a signaling molecule in animal and plant cells, via a family of highly specific enzymes collectively known as phospholipase C. In brief, diacylglycerols regulate the activity of a group of at least a dozen related enzymes known as protein kinase C, which control many key cellular functions, including differentiation, proliferation, metabolism, and apoptosis. Indeed, the biological actions of the various components released have been the subject of intensive study over many years. 2‑Arachidonoylglycerol, an endogenous cannabinoid receptor ligand, may also be a product of phosphatidylinositol catabolism.

Phosphatidylinositol Phosphates (Polyphosphoinositides) in Animals

Structure and Occurrence: Pioneering work by Mable and Lowell Hokin in the 1950s showed that phosphatidylinositol is converted to polyphosphoinositides, which have important signaling and other functional activities, including cell communication via signal transduction, cell survival and proliferation, membrane trafficking, and modulation of gene expression. Many substrate-selective kinases phosphorylate phosphatidylinositol at positions 3, 4, and 5 of the inositol ring, with distinct phosphatases and phospholipases maintaining the balance. Seven different isomers are known (mono-, bis-, and tris-phosphorylated), produced in a tightly coordinated manner, and each has characteristic biological activities. Each turns over much more rapidly than the parent phosphatidylinositol molecule. In addition, the molecular species of each isomer differ in the nature of their fatty acyl groups. Although phosphatidylinositol 4-phosphate and phosphatidylinositol 4,5-bisphosphate were long thought to be the most significant in quantitative and possibly biological terms, phosphatidylinositol 3-phosphate and its metabolites are now recognized as biologically important.

Figure \(\PageIndex{37}\) shows the structures of phosphatidylinositol phosphates

These lipids are usually present at low levels in tissues, typically about 0.5 to 1% of the total lipids in the inner leaflet of the plasma membrane, so they are unlikely to have an appreciable structural role. On the other hand, static measurements of rapidly turning-over lipids do not meaningfully assess their cellular functions. The positional distributions of fatty acids in the phosphatidylinositol, phosphatidylinositol 4-phosphate, and phosphatidylinositol 4,5-bisphosphate of ox brain are listed in Table \(\PageIndex{7}\). In each, saturated fatty acids are concentrated at sn-1 and polyunsaturated fatty acids, especially arachidonate, at sn-2; differences among the three lipids are minimal.

Biosynthesis: Phosphatidylinositol per se is the ultimate precursor of all phosphoinositides; their head groups have different charges and structures that directly affect membrane properties and, via metabolic interactions, can act as chemical switches. Membranes maintain individual phosphoinositides at steady-state levels through a continuous, sequential series of phosphorylation and dephosphorylation reactions catalyzed by specific kinases, phosphatases, and phospholipase C enzymes, which are regulated and/or relocated via cell-surface receptors for extracellular ligands. While this has been termed a ‘futile cycle’, which can consume a significant proportion of cellular ATP production, it is only part of a wider pattern of reactions - the phosphatidylinositol cycle (see below). Controlled synthesis of these phosphoinositides occurs in distinct intracellular compartments, supporting distinct, independently regulated functions and spatially distinct target enzymes or receptors. In mammals, researchers have identified 18 phosphoinositide interconversion reactions to date, mediated by at least 20 phosphoinositide kinases and 34 phosphoinositide phosphatases spanning 8 and 10 classes, respectively; some remain uncharacterized.  Most of these enzymes are conserved across all eukaryotes, and each has distinct functions and specificities that related isoforms cannot replace.

Generally, most monophosphorylations occur in endomembranes, such as the endosomes and the Golgi network. In contrast, the second and third phosphorylations occur primarily at the plasma membrane, as reflected in each membrane's lipid composition. While these enzymes are believed to act independently and sequentially to produce a specific product, some may participate in protein complexes to coordinate their activities. Specific transporters, especially the 'Nir2' protein, facilitate phosphoinositide exchange between membranes. Phosphatidylcholine metabolism is also linked, as phospholipase D can hydrolyze it to phosphatidic acid, an important activator of key kinases. Figure \(\PageIndex{38}\) overviews polyphosphoinositide metabolism in animal tissues.

Thus, as an example, phosphatidylinositol 4-phosphate (PI(4)P) is produced by the action of a phosphatidylinositol 4-kinase (PI4K) in the Golgi, and is in turn phosphorylated by a phosphatidylinositol phosphate 5-kinase (PIPK I) to form phosphatidylinositol 4,5-bisphosphate (PI(4,5)P) at the plasma membrane. However, a specific 4-kinase (PIPK II) can also form it by phosphorylating phosphatidylinositol 5-phosphate. Four PI4K isoforms in two structural families operate in different subcellular membrane compartments to produce phosphatidylinositol 4-phosphate for specific signaling functions. Some selectivity in forming molecular species or remodeling may further enrich arachidonic acid content.

Researchers later discovered that a 3-kinase (PI3K III, or the VPS 34 complex) also phosphorylates phosphatidylinositol to produce phosphatidylinositol 3-phosphate (PI(3)P) in early endosomes. Three phosphatidylinositol 3-kinase families (eight isoforms) have been described, each with distinct substrate specificities. A second phosphoinositide signaling pathway involves activation of two of these 3‑kinases, stimulated by growth factors and hormones, which phosphorylate phosphatidylinositol 4,5-bisphosphate (by PI3K I - four isoforms) and phosphatidylinositol 4‑phosphate (by PI3K II - three isoforms) to produce phosphatidylinositol 3,4,5-trisphosphate (PI(3,4,5)P) and phosphatidylinositol 3,4‑bisphosphate (PI(3,4)P), respectively. Although phosphatidylinositol 3-phosphate and other 3‑phosphorylated metabolites account for only about 0.5% of total phosphoinositides in resting mammalian cells, they are now recognized as profoundly important for cellular metabolism.

In addition to kinase activity, specific phosphoinositide phosphatases regulate these metabolite levels; these enzymes are highly conserved in eukaryotes and dephosphorylate phosphoinositides at the 3, 4, and 5 positions of the inositol ring. For example, so-called ‘SHIP’ phosphatases convert phosphatidylinositol 4,5‑bisphosphate back to phosphatidylinositol 4-phosphate by hydrolysis of the 5-phosphate group. 3‑Phosphorylated phosphoinositides are only degraded by phosphatases, especially those of the PTEN family, and not by phospholipase C (see below).

Cell organelles have membranes with distinct functions and molecular compositions. Yet, all phosphatidylinositol precursors are primarily formed in the endoplasmic reticulum, and membrane lipids must be transported between sites via specific trafficking processes/proteins. Effector proteins are selectively recruited to particular membranes by binding a single phosphoinositide, then interact with phosphoinositide-binding proteins and enzymes that channel phosphoinositide production toward the required biological outcomes and regulate signaling. For example, much of the phosphatidylinositol 4‑phosphate and phosphatidylinositol 4,5-bisphosphate involved in signaling is believed to be formed at contact sites between the endoplasmic reticulum and plasma membrane.

A concept has emerged in which each phosphoinositide has its own role – the ‘lipid code’ hypothesis, in which defined lipids act as labels for each cellular membrane to organize cells into dynamic and responsive membrane-bound compartments and maintain the orderly flow required for the complexities of membrane trafficking and spatio-temporal signaling reactions. Thus, phosphatidylinositol 4-phosphate, phosphatidylinositol 4,5‑bisphosphate, phosphatidylinositol 3-phosphate, and phosphatidylinositol 3,5‑bisphosphate are found mainly on the Golgi, plasma membrane, early endosomes, and late endocytic organelles, respectively, where they are sometimes regarded as landmarks for these compartments. For example, phosphatidylinositol 4,5‑bisphosphate is present throughout the plasma membrane and is considered a general marker. In contrast, phosphatidylinositol 3,4,5-triphosphate is a characteristic component of the basolateral region of this membrane in a polarized cell but is absent from the apical part. However, this map of phosphoinositides to specific organelles is derived from their steady-state distributions. Still, the highly dynamic generation and consumption of different phosphoinositides in response to stimuli across various subcellular compartments in living cells, driven by kinases and phosphatases, together with lipase reactions, may lead to transient pools of distinct molecular forms. The precursors must be continuously replenished by new synthesis.

Water-Soluble Inositol Phosphates

As mentioned briefly above, hydrolysis of phosphatidylinositol phosphates by calcium-dependent phospholipase C (or 'phosphoinositidase C') leads to the generation of sn‑1,2‑diacylglycerols, which act as second messengers in animal cells and are of enormous metabolic importance. Many enzymes of this type exist, but phosphoinositide-specific phospholipase C activity is an essential step in inositol signaling pathways. The enzyme exists in six families, each with at least 13 isoenzymes, all of which share conserved regions, such as the pleckstrin homology (PH) binding domain. Each has a distinctive role and can have a characteristic cell distribution linked to a specific function. Signaling molecules such as G-protein-coupled receptors, receptor tyrosine kinases, Ras-like GTPases, and calcium ions stimulate these enzymes, linking phosphatidylinositol phosphate hydrolysis to a wide range of other cellular signals. Because phospholipase C is a soluble cytosolic protein, translocation to the plasma membrane is crucial for signal transduction. Regulation of these isoenzymes, and PLCγ1 in particular, is vital for health, as they can activate or inhibit important pathophysiological processes, especially cancer.

Some phosphatidic acid is synthesized from diacylglycerols produced within the plasma membrane by diacylglycerol kinases. This acid is transported back to the endoplasmic reticulum and can ultimately be reutilized for phosphatidylinositol biosynthesis.

The other products of the phospholipase C reaction that are especially relevant because of their many essential functions are water-soluble inositol phosphates. Up to 60 different compounds of this type are possible, and at least 37 have been found in nature, all of which are biologically important. However, polyphosphoinositides with a phosphate in position 3 are not substrates for phospholipase C.

Figure \(\PageIndex{39}\) shows the generation of inositol phosphates by phospholipase C.

For example, in response to various physiological stimuli in animals, phospholipase C hydrolyzes phosphatidylinositol 4,5-bisphosphate in the plasma membrane, releasing inositol 1,4,5-trisphosphate. This important cellular messenger diffuses into the cytosol and stimulates calcium release from an ATP-loaded store in the endoplasmic reticulum via ligand-gated calcium channels (the diacylglycerols remain in the membrane to recruit and activate members of the protein kinase C family). The increase in calcium concentration, together with the altered phosphorylation status, activates or deactivates many different protein targets, enabling cells to respond appropriately to the extracellular stimulus. To enable rapid replenishment of the phosphatidylinositol 4,5‑bisphosphate used in this way, a cycle of reactions - the phosphatidylinositol cycle - must occur (see below). In contrast, a recent publication suggests that phosphatidylinositol 4-phosphate in the plasma membrane may be a more important source of diacylglycerol following stimulation of G protein–coupled receptors.

The various inositol phosphates appear to control cellular events in specific ways, especially by organizing key signaling pathways, rearranging the actin cytoskeleton, or regulating intracellular vesicle trafficking. They have been implicated in gene transcription, RNA editing, nuclear export, and protein phosphorylation. These remarkable compounds can be rapidly synthesized and degraded in discrete membrane domains or even subnuclear structures, making them ideal regulators of dynamic cellular mechanisms. Structural studies of inositol polyphosphate-binding proteins suggest that inositides may, at least in part, modify protein function as structural cofactors, ensuring proteins adopt optimal conformations. In addition, phosphoinositides and inositol polyphosphates are key components of the cell's nucleus, where they have many essential functions, including DNA repair, transcription regulation, and RNA dynamics. They may serve as activity switches for the nuclear complexes responsible for these processes, with the phosphorylation state of the inositol ring being of primary importance. As different isomers appear to have specific functions at each gene expression level, extracellular events must coordinate the production of these compounds in a highly synchronous manner.

From plants to mammals, kinases add an extra regulatory tier by generating energetic diphosphate (pyrophosphate)-containing molecules from inositol phosphates. Conversely, polyphosphate phosphohydrolase enzymes can dephosphorylate them, regenerating the original inositol phosphates. These inositol pyrophosphates and the enzymes involved in their metabolism also regulate cellular processes by modulating protein activity through various mechanisms.

Phospholipase C isoenzymes regulate the concentration of phosphatidylinositol 4,5-bisphosphate and related lipids and, hence, their activities, in addition to generating new biologically active metabolites.

Phosphatidylinositides in Plants

In plants, as in animals, phosphatidylinositol and polyphosphoinositides have essential biological functions, exerting regulatory effects by acting as ligands that bind protein targets via specific lipid-binding domains and thereby altering protein localization and enzymatic activity. However, polyphosphoinositide metabolism appears to have evolved differently after the divergence of the animal and plant kingdoms. As a result, the details of each process differ substantially, not least because phosphoinositides are localized differently in plants and animals. Phosphatidylinositol per se is the precursor of the phosphorylated forms and determines their fatty acid compositions. It also inhibits programmed cell death by serving as the biosynthetic precursor of the sphingolipid ceramide phosphoinositol and by reducing ceramide levels.

As in animals, the various phosphoinositides (five in total) are produced and interconverted rapidly by a series of kinases and phosphatases (in many isoforms) in different cellular membranes in response to environmental or developmental cues. For example, the endoplasmic reticulum generates phosphatidylinositol. At the same time, PI 4-kinases and their products are located in the trans-Golgi network and nucleus, and PI4P 5-kinases and their products are present at the plasma membrane. During polyphosphoinositide biosynthesis, the first phosphorylation occurs at the hydroxyl group at position 3 or 4 of the inositol ring, catalyzed by the appropriate kinases, while the second phosphorylation then occurs at position 5; phosphatase action on PI 3,5‑bisphosphate produces PI 5-phosphate. Most other metabolites are produced via phosphatidylinositol 3-phosphate, and reports that some phosphatidylinositol 3,4,5-trisphosphate may be produced from phosphatidylinositol 4,5‑bisphosphate require confirmation. In contrast to mammalian phosphatidylinositol 3-kinases, which accept phosphatidylinositol and its monophosphates as substrates, the plant enzyme acts only on the former.

Figure \(\PageIndex{40}\) shows polyphosphoinositide metabolism in plants

The reverse reaction in plants is carried out by phosphoinositide phosphatases, which fall into three main families: the phosphatase/tensin (PTEN) family, 5-phosphatases (5-PTases), and phosphatases containing Suppressor of Actin (SAC) domains, each with distinct subcellular locations, substrate specificities, and regulatory mechanisms.

Although plants have what might be considered normal levels of phosphatidylinositol 4-phosphate, the concentrations of phosphatidylinositol 4,5‑bisphosphate and other phosphoinositides are extremely low (10 to 20-fold lower than in mammalian cells), yet they still have vital functions. Cell types differ, but in Arabidopsis epidermal root cells, PI(4,5)P₂ is most abundant in the plasma membrane (apex region) and nucleus. At the same time, PI4P slowly distributes between the plasma membrane and Golgi, with the highest concentration in the former. Multivesicular bodies/late endosomes accumulate PI3P and PI(3,5)P₂; the tonoplast and autophagosomes contain PI3P. How the various metabolites are transported across membranes remains to be determined.  Non-vesicular transport is believed to occur at membrane contact sites, and vesicular transport probably occurs as well.

Highly polarized distributions of phosphoinositides are found within membranes, generally oriented toward the cytosolic leaflet, and they are believed to be organized in nanoclusters with other lipids and proteins. For example, phosphatidylinositol-4-phosphate is an important constituent of the plasma membrane in plant cells, where it controls the electrostatic state and is involved in cell division. It may control the location and function of many membrane proteins, including those required for development, reproduction, immunity, nutrition, and signaling. PI(4)P is the only phosphoinositide present at the cell plate, i.e., the membrane separating two daughter cells during cell division. In addition, PI(4)P may interact with salicylic acid in the plant immune response, and it is produced during salt stress. However, researchers are now discovering specific functions for each plant phosphoinositide; these lipids are rapidly produced in response to osmotic and heat stress, and continuous turnover is essential for cell growth and development. For example, they strongly affect the growth of many cell types and guard cell function. In the nucleus, proteins have been identified that bind phosphoinositides via their acyl chains, leaving the head group exposed for enzymatic modification and signal transduction.

Phosphoinositides are especially important in microdomains at the tips of growing tissues such as the shoot apical meristem, pollen tubes, and root hairs, where phosphatidylinositol 4,5-bisphosphate functions in stem cell maintenance and organogenesis. In the plasma membrane, it is enriched in the detergent-resistant component, which is commonly equated with rafts. Although its concentration is low, PI(4,5)P₂ has been shown to signal by binding many target proteins with characteristic binding domains. For example, with phosphatidic acid, PI(4,5)P₂ regulates the activity of many actin-binding proteins, which in turn control actin cytoskeleton activity. This plays a key role in plant growth, subcellular organelle movement, cell division and differentiation, and plant defense. In addition, this lipid regulates ion channels, ATPases, and phospholipase C-mediated lipid degradation, as well as the production of additional second messengers. It is an important factor in both clathrin-mediated endocytosis and exocytosis. The specificity of these interactions may depend on the fatty acid composition of the lipid and on the activity of phosphatidylinositol 4-phosphate 5-kinase.

As in animals, phosphoinositides have a role in endosomal sorting through the central vacuole, a plant-specific organelle with both lytic and storage functions. Phosphatidylinositol 3,5-bisphosphate is the least abundant phosphoinositide, but it is a crucial lipid for membrane trafficking. The PI to PI(3)P to PI(3,5)P₂ cascade, whose second step requires the kinase FAB1, is required for endosomal sorting events leading to membrane protein degradation or retrieval, vacuolar morphogenesis, and autophagy. PI(3,5)P₂ is involved in stomatal closure and root hair growth, and it is also induced by salt stress.

Several phospholipase C enzymes specific for polyphosphoinositides have been isolated from higher plants; Ca²⁺ activates them, and, unlike their mammalian counterparts, G proteins do not regulate them. It is uncertain whether phosphatidylinositol is a substrate for these enzymes in vivo. Less is known about the metabolism of the water-soluble inositol phosphates produced compared to animals, and plants appear to lack a receptor for inositol 1,4,5-trisphosphate (IP₃). However, it is the most abundant metabolite of this type and has been reported to induce calcium ion release, triggering stomatal closure. However, increasing evidence supports lipid signaling mediated by phospholipase C in plant abiotic stress tolerance and development. There is a general, if contested, belief that inositol hexakisphosphate (phytic acid or IP6), produced at least in part by sequential phosphorylation of inositol 1,4,5-trisphosphate, is a more important cellular messenger in plants and mobilizes an endomembrane store of calcium ions. Inositol-1,2,4,5,6-pentakisphosphate (IP₅) is a structural co-factor of the jasmonic acid receptor coronatine insensitive 1, linking phosphoinositide signaling with phytohormone-controlled pathways.

In contrast to animals, diacylglycerols, the other product of phospholipase C hydrolysis of phosphoinositides, are rapidly converted to phosphatidic acid by diacylglycerol kinases and have not been considered important in signal transduction in plants. Plants lack protein kinase C, but they do have proteins with related properties that appear to be influenced by diacylglycerols. Via phospholipase D, inositol phospholipids provide phosphatidic acid, which has well-characterized signaling functions in plants, especially in defense.

Lyso-Phosphoinositides

Figure \(\PageIndex{41}\) shows the structure of lysophosphatidylinositol

Lysophosphatidylinositols: Lysophosphatidylinositols (LPI), i.e., with a single fatty acid only linked to the glycerol moiety, are formed as intermediates in the remodeling of the fatty acid compositions of the lipids by the action of phospholipase A₁ or phospholipase A₂ (e.g., cPLA₂α), and when arachidonic acid is released for eicosanoid biosynthesis (see above). In ovarian cancer, LPI is elevated appreciably to around 15µM in ascites, and it is also present at high levels in obese subjects.

It has become apparent relatively recently that, like other lysophospholipids, phosphatidylinositol and polyphospho-analogs may have messenger functions. For example, it has long been known to stimulate insulin release from pancreatic cells, suggesting a role in glucose homeostasis. sn-2-Arachidonoyl-lysophosphatidylinositol, in particular, is an endogenous ligand for the G protein-coupled receptor GPR55 and can thereby induce rapid phosphorylation of certain enzymes, including a protein kinase, which promotes cancer cell proliferation, migration, and metastasis. Indeed, lysophosphatidylinositol is a biomarker for poor prognosis in cancer patients, and its concentration is significantly elevated in highly proliferative cancer cells in vitro. GPR55 is expressed in many brain regions, the intestines, endocrine pancreas, and islets (where it may stimulate insulin release). It has been implicated in macrophage activation and inflammation. Beyond its role in cancer, lysophosphatidylinositol has been implicated in many metabolic diseases. It is reported to be a precursor of the endocannabinoid 2‑arachidonoylglycerol via human glycerophosphodiesterase 3, a lysophospholipase C. This enzyme suppresses the lysophosphatidylinositol receptor, thereby acting as a switch between GPR55 and endocannabinoid (CB2) signaling.

Glycerophosphoinositol: Sequential removal of both fatty acids from phosphatidylinositol by a specific phospholipase A₂ (PLA₂IVα) with both phospholipase A₂ and lysophospholipase activities releases water-soluble glycerophosphoinositol. A glycerophosphodiester phosphodiesterase can hydrolyze this to inositol 1-phosphate; however, glycerophosphoinositol per se has distinctive biological activities and functions, as do related compounds derived from phosphatidylinositol phosphates. In particular, glycerophosphoinositol has anti-inflammatory activity, as it inhibits the inflammatory and thrombotic responses induced by bacterial lipopolysaccharides (endotoxins).

Figure \(\PageIndex{42}\) shows the structure of glycerophosphoinositol

The Phosphatidylinositol Cycle

Phosphatidylinositol can be considered the center of a cycle of reactions and intermediates involved in innumerable aspects of cellular signaling in animals (a similar cycle can be described for plants). We discuss these in detail above, but it is useful to note how each component fits into a larger pattern. In brief, as illustrated, the various synthetic and hydrolytic reactions involved in phosphoinositide metabolism can be considered to constitute a phosphatidylinositol cycle with enzymes located both in the endoplasmic reticulum and plasma membrane, so lipids have to be transferred across the cytosol in both directions between the two to complete the cycle, probably via adjacent membrane structures and facilitated by proteins of the phosphatidylinositol transfer protein membrane-associated family (PITPNM or nir2), which may channel phosphoinositide production to specific biological outcomes. Phospholipase C and phosphatidylinositol-4-phosphate 5-kinase (PI4P 5K) are located in the plasma membrane, while the cytidine diphosphate-diacylglycerol synthase (CDS2) and phosphatidylinositol synthase are in the endoplasmic reticulum. The epsilon isoform of diacylglycerol kinase (DGKε) is located at contact sites between the endoplasmic reticulum and plasma membrane. Still, nine additional isoforms with distinct cellular and subcellular localizations may be involved in the cycle. Each turn of the cycle consumes a great deal of energy and requires three moles of ATP, as well as cytidine triphosphate and inositol.  If endogenous pyrophosphatases hydrolyze pyrophosphate to inorganic phosphate, the cycle can proceed in only one direction.

Figure \(\PageIndex{43}\) shows the phosphatidylinositol cycle.

Factors such as membrane curvature must be taken into account, and the diagram is necessarily a considerable oversimplification. In addition to participating in this cycle, many lipid intermediates can serve as precursors for other lipids; for example, diacylglycerols can be precursors for triacylglycerols, while phosphatidic acid is a precursor for phosphatidylcholine and phosphatidylethanolamine. Each lipid intermediate undergoes acyl-chain remodeling via the Lands cycle, and the released polyunsaturated fatty acids can be used for eicosanoid production. A further byproduct of the cycle is inositol triphosphate, which regulates intracellular calcium levels.

It has been suggested that the unique molecular species composition of phosphoinositides (18:0-20:4) could influence their selective recycling back into phosphatidylinositol, as many of the enzymes involved prefer this substrate. A further proposal is that the phosphatidylinositol cycle could enrich this species through multiple rounds of cycling.