What is the most abundant phospholipid found in membranes?

Systemic Context of Pulmonary and Surfactant Phosphatidylcholine Metabolism

Surfactant PC maintenance must be regarded in the context of total PC metabolism in PN-II. PC synthesis is an order of magnitude higher than required for surfactant production, because PN-II express ABCA1, which mediates the export of PC and cholesterol across the basolateral membrane, as well as ABCA3 for lamellar body assembly. ABCA1 is linked to lipid transfer to apolipoprotein A1, thereby connecting surfactant with systemic high-density lipoprotein metabolism.50,334-336 ABCA1 is under the control of 9-cis-retinoic acid and 22-hydroxycholesterol. Consequently, ABCA1 defects result in virtual absence of high-density lipoprotein PC in plasma, alveolar proteinosis, pulmonary accumulation of total PC, DSPC, and cholesterol, impaired lung function, and structural lung defects.336,337

5.2 Phospholipid supply

Phosphatidylcholine (PC) is the major PL found in circulating VLDL [28]. In the liver, PC is synthesised by one of two pathways (Chapter 7): de novo biosynthesis via the CDP-choline pathway or methylation of phosphatidylethanolamine (PE) by PE methyltransferase (PEMT). Impairment of either PC biosynthetic pathway reduces VLDL secretion, with the level of reduction reflecting the contribution of the two pathways to overall PC biosynthesis: 70% via CDP-choline pathway and 30% via the PEMT pathway.

Impaired PC biosynthesis leads to nascent VLDL particles that are recognised as defective and removed from the secretory pathway, likely by post-ER autophagy (see below). Phospholipid transfer protein (PLTP) may work in a coordinated manner with MTP for delivery of PL substrates for VLDL assembly. There is also evidence that re-acylation of lysophosphatidylcholine to make PC is important in particle lipidation.

4.1 PC synthesis by the CDP-choline/Kennedy pathway

PC is a choline-containing, zwitterionic phospholipid (Figure 7.1) that constitutes between 30 and 60% of the phospholipid mass of eukaryotic cell membranes. The quaternary amine choline headgroup of PC cannot be replaced by primary or secondary amine analogues without adversely affecting cell viability. Given its abundance in mammalian cell membranes it is not surprising that depriving cells of PC by inhibiting synthesis or the availability of precursors results in growth arrest and apoptosis. In addition to an essential role in membrane biogenesis, PC is also a major component of secreted lipoproteins, bile and lung surfactant. If PC availability is limited, the assembly and secretion of these lipid–protein complexes is inhibited, causing the sequestration of lipids within cells and interfering with broader physiological functions. PC is also a reservoir for second messengers such as fatty acids, DAG, PA and lyso-PC that can be released by activation of lipases in response to primary signalling factors. Finally, PC is an important precursor for the synthesis of PS and sphingomyelin through donation of its PA and phosphocholine moieties, respectively.

In the 1950s, Kennedy and Weiss identified the de novo pathways for the synthesis of PC and PE that utilise the activated nucleotide intermediates CDP-choline and CDP-ethanolamine, respectively. The CDP-choline and CDP-ethanolamine (Kennedy) pathways involve similar reactions that are, in some cases, catalysed by enzymes that utilise both choline and ethanolamine intermediates as substrates. The three-step CDP-choline pathway is the primary source for PC in all mammalian cells and utilises choline that is salvaged from PC degradation or transported from the extracellular space (Figure 7.5). PC can also be synthesised from pre-existing phospholipids, namely the methylation of PE and acylation lyso-PC. However, these pathways make minor contributions to cellular PC synthesis or are tissue specific.

Phosphatidylcholine

Phosphatidylcholine (PC) is an important component of the mucosal layer of the colon and acts as a surfactant within the mucus to create a hydrophobic surface to prevent bacterial penetrance. PC levels have been found to be reduced in the colon of UC patients compared to controls.160,161 Without adequate PC in the mucosa, the intestinal barrier is permeable to colonic bacteria, leading to chronic intestinal inflammation and barrier defects.161-164 Four human trials have been performed to determine the efficacy of PC in the treatment of UC. Two studies on patients with chronic active UC have reported clinical, endoscopic, and histological improvements.165,166 Steroid-refractory UC patients were more successful in weaning off steroids when treated with PC compared to placebo (P < 0.01).167 A double-blinded, randomized, placebo-controlled, multicenter study found that 3.2 g daily of PC improved clinical activity scores, histological remission, and relapse rates without any significant adverse events.168

3.1 Historical background

PC was first described by Gobley in 1847 as a component of egg yolk and was named ‘lecithin’ after the Greek equivalent for egg yolk (lekithos). In the 1860s Diakonow and Strecker demonstrated that lecithin contained two fatty acids linked to glycerol and that choline was attached to the third hydroxyl by a phosphodiester linkage. The first significant advance in understanding PC biosynthesis occurred in 1932 with the discovery by Charles Best that animals have a dietary requirement for choline. In the 1950s, the CDP-choline pathway for PC biosynthesis (Fig. 3) was described by Eugene Kennedy and co-workers. A key observation was that CTP, rather than ATP, was the activating nucleotide for PC biosynthesis [4]. CTP is required not only for PC biosynthesis but also for the de novo synthesis of all phospholipids (prokaryotic and eukaryotic) excluding PA which can be considered as an intermediate in glycerolipid biosynthesis).

An alternative pathway for PC biosynthesis, of quantitative significance only in liver, is the conversion of PE to PC via PE methylation (Fig. 4). This pathway was first observed in 1941 when Stetten fed [15N]ethanolamine to rats and isolated [15N]choline. Two decades later Bremer and Greenberg detected a microsomal enzyme activity that converted PE to PC via transfer of methyl groups from S-adenosylmethionine.

3.1 Phosphatidylcholine and sphingomyelin

PC is the main component of mammalian membranes and lipoproteins. PC is also keenly involved in cell signaling [41]. SM is specifically enriched to the plasma membranes of cells and is also abundant in lipoproteins. Notably, SM and cholesterol are thought to form segregated, ordered domains within the cellular membranes [42]. Such domains, also referred as “rafts,” are presently under intensive investigation due to their putative roles in intracellular lipid and protein sorting, cellular signaling molecules, and various diseases [43].

Both PC and SM contain a phosphocholine head group, which makes the molecules zwitterionic and largely dictates their ionization and fragmentation behavior. PC and SM readily form [M+H]+ ions which upon CAD yield an abundant phosphocholine fragment with m/z 184 (cf. Table 1) and can thus be selectively detected by scanning for parents of this ion. In the presence of different salts, PC and SM form both cation and anion adducts, which may be utilized for the elucidation of the fatty acids present in the molecules [19,29,34,44,45]. In case of SM, also a fragment indicative of the long-chain base is found [45].

Quantification of SM species is often hampered by spectral overlap due to the much more abundant PC species. This problem can be solved either by removing the PCs with alkaline hydrolysis in silico using a spectral subtraction protocol [37] or by employing LC–MS [19]. Also, NL scanning in the negative mode with properly adjusted instrument settings seems to allow selective detection of SM in the presence of PC [46]. Many tissues contain PC species with an ether-linked alkyl chain in the sn-1 position. At present, these molecules are probably most readily analyzed by LC–MS [19,47].

Phosphatidylcholine-Specific Phospholipase C

Phosphatidylcholine Hydrolysis. As work on the PPI pathway became more sophisticated and quantitative, certain discrepancies began to arise. In particular, the data showed that there was far more DG produced than PIP2 hydrolyzed; for example, in hepatocytes, vasopressin hydrolyzed 9 ng of PIP2 per milligram of tissue but generated 400 ng of DG per milligram. Furthermore, the discrepancies were greatest at later time points. With phospholipids tagged with different fatty acids, it was determined that PIP2 supplied the DG acutely, but that PC was the source of DG after the first few minutes of stimulation. There are two ways that PC can release DG: (1) a PC-specific PLC (PC-PLC) can hydrolyze PC in a manner analogous to the PPI pathway, or (2) PC could be acted on by a PC-specific phospholipase D (PC-PLD). The latter reaction would actually produce phosphatidic acid (PA) and choline (see Fig. 10-3); the former would have to be degraded further by a PA phosphohydrolase to produce DG. Both pathways exist; the contribution by each will vary depending on the specific system.

Phosphocholine, the other product of the first hydrolytic pathway, has not been well studied. However, there are a few reports that suggest that it can transduce signals. For example, phosphocholine triggers mitosis, activates the transcription factor NF-κB, and synergizes with sphingosine-1-phosphate in the stimulation of Raf.

PC-PLC is stimulated by MAPK phosphorylation and inhibited by Go. The effects of this enzyme can also be potentiated by factors that interfere with PC recycling. CTP:phosphocholine cytidylyltransferase is involved in the reutilization of DG and can be regulated by hormone mediators. For example, cholecystokinin (CCK), a-adrenergic agonists, and bombesin inhibit this enzyme in the pancreas by elevating calcium and activating CaM. When cytidylyltransferase activity is suppressed, DG recycling is delayed and DG accumulates.

Phosphatidylcholine Synthesis. In addition to being an alternate source for DG, the accumulation of PC can alter membrane properties in a way that enhances cAMP production. PC can be synthesized from phosphatidylserine through a decarboxylation to phosphatidylethanolamine followed by successive methylations (Fig. 10-12). Although two different phospholipid methyltransferases were originally postulated, recent data suggest that there may be only one. Because phosphatidylserine is predominantly located on the cytoplasmic face of the plasmalemma and PC is usually on the extracellular side, this conversion to PC is accompanied by a transverse migration across the membrane. The source of the methyl groups is S-adenosylmethionine, which is converted to S-adenosylhomocysteine (SAH). The methyltransferase is subject to product inhibition by SAH, and several drugs have been developed to take advantage of this fact. For example, 3-deazaadenosine and its structural variants either are or can be metabolized to SAH analogs and are potent inhibitors of the methyltransferase.

Phospholipid methylation has been closely associated with cAMP production in several systems. In fibroblasts, bradykinin stimulates phospholipid methylation before cAMP content rises; in most systems, an elevation in cAMP concentrations is the most common response to PC synthesis. However, in Xenopus oocytes, progesterone stimulation of phospholipid methylation is associated with a decline in cAMP levels. In both systems, the methylation peaks at 15 seconds, although the changes in cAMP content require 2 to 5 minutes, suggesting a cause-and-effect relationship. This is supported by the use of methyltransferase inhibitors, which also inhibit the changes in cAMP concentrations. Finally, cholera toxin and fluoride can stimulate the adenylate cyclase through Gs without affecting the PC levels; this finding, along with the time courses, would eliminate the possibility that the changes in the PC metabolism are a secondary event.

How might phospholipid methylation influence cAMP production? The first argument is that it increases membrane fluidity, thereby facilitating the coupling of receptor, G proteins, and adenylate cyclase. The ability of isoproterenol, a β-adrenergic agonist, to stimulate adenylate cyclase in turkey erythrocytes is influenced by membrane fluidity; loading the membranes with cholesterol decreases membrane fluidity and dampens isoproterenol-induced cyclase activity. Conversely, loading the membranes with vaccenic acid increases fluidity and enhances cyclase activity. Likewise, increasing the PC content of these membranes increases their fluidity and coupling efficiency.

Second, phospholipid methylation may act through calcium. Calcium influxes are stimulated after phospholipid methylation but before changes in cAMP content are observed; these fluxes usually occur in 0.5 to 2 minutes depending on the system. Furthermore, methylation inhibitors also inhibit these fluxes, suggesting that they were evoked by the methylation. Because bradykinin has been shown to stimulate the PPI pathway, it is possible that the changes in membrane fluidity induced by changes in PC metabolism could also have facilitated PPI hydrolysis, which then led to the calcium fluxes. Regardless of the exact mechanism, the resulting calcium fluctuations would alter adenylate cyclase activity (see later).

Third, the increase in PC may stimulate a PC-specific phospholipase A2(PC-PLA2), which would release arachidonic acid for eicosanoid synthesis (see later); many receptors for the eicosanoids are coupled to adenylate cyclase. Methyltransferase inhibitors block the release of arachidonic acid and cAMP elevation; mepacrine (also called quinacrine) is an inhibitor of PLA2 and has the same effect. Alternatively, the active agent may not be arachidonic acid or its metabolites but the other hydrolytic product, lysophosphatidylcholine. Lysophosphatidylcholine, like the eicosanoids, is a parahormone that binds a GPCR. In addition, lysophospholipids, which lack a fatty acid in the second position, are strong detergents and, in sufficiently large concentrations, can lyse cells (see later). Indeed, the active ingredient in several snake toxins is a PLA2, and the toxicity of the venom can be directly attributed to this lytic effect. In smaller amounts, lysophosphatidylcholine might act as a membrane fusogen and aid in secretion; phospholipid methylation has been implicated in a number of secretory systems. Lysophospholipids have also been implicated in the regulation of the (Na+, H+)antiport system that influences cellular pH.

This scheme is not without its critics; there are, in fact, three basic problems with this hypothesis. First, in many systems the elevation in cAMP precedes phospholipid methylation, and the methylation can be induced by cAMP. These systems include glucagon in the liver, ACTH in adipocytes, and hCG in Leydig cells. This is in contrast to the findings in turkey erythrocytes and Xenopus oocytes noted earlier. Second, the various inhibitors used all have side effects; the methyltransferase inhibitors can also inhibit other types of methylation reactions, and mepacrine can bind to chromatin and inhibits both oxidative phosphorylation and the mitochondrial ATPase. Furthermore, the effects of these drugs can be inconsistent; some methyltransferase inhibitors will block the actions of certain hormones in a particular system, whereas other inhibitors cannot, even though both effectively suppress methyltransferase activity.

Third, methyltransferase activity does not necessarily correspond to PC content. Phosphatidylcholine can be synthesized by two separate pathways: the methylation pathway (Fig. 10-12) and the salvage pathway. The latter pathway activates phosphocholine with CTP to form CDP-choline, which is then coupled to DG to form PC. In the liver, the two pathways are reciprocally controlled so as to maintain a constant PC content. For example, glucagon, β-agonists, and vasopressin stimulate the methyltransferase but inhibit the salvage pathway; 3-deazaadenosine inhibits the methyltransferase but stimulates the salvage pathway. Similarly, exogenous choline activates the salvage pathway and suppresses methylation; choline deficiency has the opposite effect. If this type of control operated in all tissues, PC content would remain constant regardless of the methyltransferase activity, and membrane fluidity would not be altered. However, PC could still be a significant source of arachidonic acid, DG, and lysophosphatidylcholine, even if total PC content and membrane fluidity does not fluctuate.

Phosphatidylcholine

Phosphatidylcholine (PPC) is an essential component of all cell membranes and is vulnerable to attack by lipid peroxidation. Through mechanisms that are, as yet, unclear, dietary supplementation with phosphatidylcholine has been shown to attenuate ethanol-induced fibrosis in baboons.412 Potential mechanisms of action include stimulation of collagenase413 and acting as a “sink” for free radicals.414 A long-term trial in patients with alcoholic cirrhosis has just been completed in the United States. Although there was a trend to improvement in transaminases and bilirubin in the PPC group in certain patient subgroups (heavy drinkers and those with hepatitis C), overall there was no improvement in liver histology as determined by liver biopsies 24 months apart. The potential benefits of the drug may not have been evaluated appropriately because of the dramatic reduction in drinking seen in the treated and placebo groups of patients that were followed up to completion.415

6.2 Phospholipids

Phosphatidylcholine is the major phospholipid on the surface monolayer of all lipoproteins, including VLDLs. In the liver, phosphatidylcholine is synthesized by two biosynthetic pathways: the CDP-choline pathway and the phosphatidylethanolamine N-methyltransferase pathway (Chapter 8). Choline is an essential biosynthetic precursor of phosphatidylcholine via the CDP-choline pathway. When cells or animals are deprived of choline, plasma levels of TG and apo B are markedly reduced and TG accumulates in the liver, resulting in fatty liver. These observations led to the widely held view that the fatty liver caused by choline deficiency is due to inhibition of PC synthesis, which in turn would decrease VLDL secretion. This hypothesis was tested in primary rat hepatocytes cultured in medium lacking choline. Upon removal of choline/methionine from culture medium, the TG content of hepatocytes was increased ~6-fold, and the secretion of TG and apo B in VLDL was markedly reduced. The interpretation of these experiments was that hepatic VLDL secretion requires the synthesis of phosphatidylcholine from either the CDP-choline or methylation pathways which require choline or methionine, respectively, as precursors (D.E. Vance, 1988). However, since choline deprivation was induced in a background of methionine insufficiency, it was not clear whether the lack of choline per se, and inhibition of the choline pathway for phosphatidylcholine synthesis, decreased VLDL secretion. More recent experiments have shown, surprisingly, that deficiency of choline in primary mouse hepatocytes does not reduce, but increases, phosphatidylcholine synthesis via the CDP-choline pathway, and does not decrease VLDL secretion (J.E. Vance, 2004). Thus, a deficiency of dietary choline reduces plasma TG and apo B levels by a mechanism that does not involve reduction of phosphatidylcholine synthesis.

To determine the role of the CDP-choline pathway of phosphatidylcholine synthesis for VLDL secretion, knockout mice were generated in which the gene encoding CTP:phosphocholine cytidylyltransferase-α (Pcyt1a), a key enzyme in phosphatidylcholine synthesis via the CDP-choline pathway (Chapter 8), was disrupted only in the liver. TG accumulated in livers of these mice and the secretion of TG and apo B was decreased (D.E. Vance, 2004). Thus, elimination of the CDP-choline pathway in the liver inhibits VLDL secretion. In addition, disruption of the gene encoding the liver-specific enzyme, phosphatidylethanolamine N-methyltransferase, in mice fed a high-fat/high-cholesterol diet, also markedly reduces the secretion of TG and apo B in VLDLs (D.E. Vance, 2003). Thus, the hepatic synthesis of phosphatidylcholine via both the CDP-choline and methylation pathways appears to be required for normal VLDL secretion.

In addition to phosphatidylcholine, smaller amounts of other phospholipids such as phosphatidylethanolamine, phosphatidylserine, phosphatidylinositol, and sphingomyelin are components of VLDL, but the physiological relevance of these phospholipids is unknown. Inhibition of sphingomyelin synthesis in rat hepatocytes by >90% by fumonisin B did not reduce VLDL secretion (A.H. Merrill, 1995). Thus, normal amounts of hepatic sphingomyelin are apparently not required for VLDL secretion. Interestingly, the phosphatidylethanolamine content of newly secreted VLDLs and VLDLs isolated from the Golgi of rat liver is several-fold higher than that in circulating VLDLs (P.E. Fielding, 1989; A. Kuksis, 2005), suggesting that either the initial assembly of phosphatidylethanolamine into VLDLs or the specific hydrolysis of this phospholipid in newly secreted VLDLs might have functional significance.

2.1.1 Phospholipids

Phosphatidylcholine (PC) and sphingomyelin (SM) are the two major phospholipids of HDL. These lipids are located in the surface monolayer of the particle together with the free cholesterol and apoA-I. A significant portion of the surface constituents of HDL is derived from the hydrolysis of fasting and postprandial triglyceride-rich lipoproteins (TGRL), and transfer of redundant constituents is enhanced by phospholipid transfer protein (PLTP). PC represents the main phospholipid subclass present in HDL. Phospholipids on the lipoprotein surface serve to solubilize cholesterol. ABCG1 has been reported to mediate cellular efflux of both cholesterol and phospholipids (PC and SM) to preβ and mature HDL. It is likely that ABCG1 preferentially effluxes SM as compared to PC [1]. There are several studies supporting the concept that HDL-PLs play a key role in HDL function, that is, FC efflux and CE uptake in RCT. Indeed, the ability of a given particle to accept FC is related to the amount of PL and to the types of PL present in the HDL. In addition, HDL-PLs strongly correlate with FC efflux to serum. Moreover, total serum from hypertriglyceridemic subjects displayed a similar FC efflux capacity as compared to control normolipidemic subjects despite low HDL-C level but normal HDL-PL level [2]. The capacity of total serum to promote FC efflux via CLA-1/SR-BI can be markedly enhanced by the enrichment with PL. The phospholipid’s fatty acyl composition of lipoproteins is known to have subtle but measurable effects on the fluidity of the lipoprotein phospholipidic layer as a result of its impact on apoA-I conformation [3]. These changes may affect the ability of HDL particles to accommodate FC molecules that have desorbed from peripheral cells. Sola and colleagues [4] demonstrated that increasing the percentage of saturated acyl chains in HDL reduces its ability to act as an FC acceptor as a consequence of a decrease in fluidity of the particle. Enrichment of HDL with sphingomyelin causes a decreased rate of FC desorption from the surface of the HDL, resulting from the high affinity of SM for cholesterol. Finally, it has been shown that degradation of SM in HDL favors LCAT activity, suggesting that SM may represent an important factor in influencing RCT at different levels [5].