Review Formation and function of phosphatidylserine and phosphatidylethanolamine in mammalian cells

Biochimica et Biophysica Acta 1831 (2013) 543–554

Contents lists available at SciVerse ScienceDirect

Biochimica et Biophysica Acta

j ourna l homepage: www.e lsev ie r .com/ locate /bba l ip

Review

Formation and function of phosphatidylserine and phosphatidylethanolamine inmammalian cells☆

Jean E. Vance ⁎, Guergana TassevaGroup on the Molecular and Cell Biology of Lipids and the Department of Medicine, University of Alberta, Edmonton, Canada AB T6G 2S2

Abbreviations: CHO, Chinese hamster ovary; ER, endphosphoethanolamine cytidylyltransferase; MAM, mitochPC, phosphatidylcholine; PE, phosphatidylethanolaminephosphatidylserine decarboxylase; PSS, PS synthase☆ This article is part of a Special Issue entitled Phospholipid⁎ Corresponding author at: 328 HMRC, University of A

T6G 2S2. Tel.: +1 780 492 7250; fax: +1 780 492 3383E-mail address: [email protected] (J.E. Vance).

1388-1981/$ – see front matter © 2012 Elsevier B.V. Allhttp://dx.doi.org/10.1016/j.bbalip.2012.08.016

a b s t r a c t
a r t i c l e i n f o
Article history:Received 26 July 2012Received in revised form 20 August 2012Accepted 21 August 2012Available online 29 August 2012

Keywords:MitochondriaMitochondria-associated membranesKnock-out mice

Phosphatidylserine (PS) and phosphatidylethanolamine (PE) are metabolically related membraneaminophospholipids. In mammalian cells, PS is required for targeting and function of several intracellular sig-naling proteins. Moreover, PS is asymmetrically distributed in the plasma membrane. Although PS is highlyenriched in the cytoplasmic leaflet of plasma membranes, PS exposure on the cell surface initiates bloodclotting and removal of apoptotic cells. PS is synthesized in mammalian cells by two distinct PS synthasesthat exchange serine for choline or ethanolamine in phosphatidylcholine (PC) or PE, respectively. Targeteddisruption of each PS synthase individually in mice demonstrated that neither enzyme is required for viabil-ity whereas elimination of both synthases was embryonic lethal. Thus, mammalian cells require a thresholdamount of PS. PE is synthesized in mammalian cells by four different pathways, the quantitatively most im-portant of which are the CDP-ethanolamine pathway that produces PE in the ER, and PS decarboxylation thatoccurs in mitochondria. PS is made in ER membranes and is imported into mitochondria for decarboxylationto PE via a domain of the ER [mitochondria-associated membranes (MAM)] that transiently associates withmitochondria. Elimination of PS decarboxylase in mice caused mitochondrial defects and embryonic lethality.Global elimination of the CDP-ethanolamine pathway was also incompatible with mouse survival. Thus, PEmade by each of these pathways has independent and necessary functions. In mammals PE is a substratefor methylation to PC in the liver, a substrate for anandamide synthesis, and supplies ethanolamine forglycosylphosphatidylinositol anchors of cell-surface signaling proteins. Thus, PS and PE participate in manypreviously unanticipated facets of mammalian cell biology. This article is part of a Special Issue entitled Phos-pholipids and Phospholipid Metabolism.

© 2012 Elsevier B.V. All rights reserved.

1. Introduction

Phosphatidylserine (PS) and phosphatidylethanolamine (PE) aretwo metabolically related aminophospholipids (Fig. 1) that arepresent in membranes of all eukaryotic and prokaryotic cells(reviewed in Ref. [1]). In mammalian, plant and yeast cells phospha-tidylcholine (PC) is the most abundant phospholipid whereas PE isthe second most abundant. However, with a few exceptions, prokary-otes do not make PC so that in this class of organisms PE is usually themost abundant phospholipid. In eukaryotic cells, PE and PS accountfor approximately 20% and 3–15%, respectively, of total phospho-lipids. The majority of phospholipids in mammalian cells are madein the ER whereas mitochondria supply all of the cardiolipin and a

oplasmic reticulum; ET, CTP:ondria-associated membranes;; PS, phosphatidylserine; PSD,

s and PhospholipidMetabolism.lberta, Edmonton, Canada AB.

rights reserved.

significant fraction of PE. Since mitochondria are presumed to be de-rived from bacteria during the evolution of eukaryotic cells, it hasbeen suggested that the lack of PC synthesis in mitochondria reflectsthe (general) inability of bacteria to synthesize PC. On the other hand,PE is the major bacterial phospholipid so that mammalian mitochon-dria have retained the capacity for synthesis of PE, as well as anotherabundant bacterial phospholipid cardiolipin. Intriguingly, however,mammalian mitochondria do not make PS whereas bacteria do, albeitby a pathway different from that in mammalian cells. Thus, not all ofthe phospholipid biosynthetic capacity of bacteria has been retainedin mammalian mitochondria.

PE was first isolated from the brain as “cephalin” by LudwigThudichum in 1884. His research on this topic was for many yearsconsidered to be “relatively insignificant,” according to Thudichum'sobituaries in Nature [volume 64, page 527 (1901)] and “The Times”of London (Sept. 10, 1901). The latter article stated that “the knowl-edge yielded by these researches was hardly commensurate withthe time and cost at which it was obtained.” Unfortunately, similarsentiments are even today often applied to basic research. Almost70 years later (1952) the structure of PE was deduced by Baer andcolleagues [2]. In 1941 PS was identified as a secondary component

http://dx.doi.org/10.1016/j.bbalip.2012.08.016

mailto:[email protected]

http://dx.doi.org/10.1016/j.bbalip.2012.08.016

http://www.sciencedirect.com/science/journal/13881981

PE PC

CMP

ser

PSPS

PSS2 PSS1

Mammals Yeast/E.coli

etn cho

ser ser

CDP-DAG

Fig. 1. Phosphatidylserine (PS) biosynthesis in mammals and yeast. In mammaliancells PS is synthesized by calcium-dependent base-exchange reactions between apre-formed phospholipid and L-serine (ser). PS synthase-1 (PSS1) catalyzes the ex-change of choline (cho) in phosphatidylcholine (PC) for serine, whereas PSS2 catalyzesthe exchange of serine for ethanolamine (etn) in phosphatidylethanolamine (PE). ThePS biosynthetic pathway in yeast and prokaryotes is completely different from that inmammalian cells. CDP-diacylglycerol and serine react to produce PS in a reaction cat-alyzed by a PS synthase unrelated to that in mammalian cells.

544 J.E. Vance, G. Tasseva / Biochimica et Biophysica Acta 1831 (2013) 543–554

of cephalin which was originally thought to be pure PE. The structureof PS was elucidated by Folch in 1941 [3] and was subsequently con-firmed by chemical synthesis [4].

In this article we shall discuss the mechanisms of the biosynthesisand cellular functions of PS and PE, particularly in mammalian cells.

2. Phosphatidylserine

2.1. Functions of phosphatidylserine

PS is not equally abundant in membranes of all types of mamma-lian cells or tissues. Compared to other tissues, the brain, and partic-ularly the retina, is enriched in PS and PE. Moreover, in the humanbrain >36% of the acyl-chains of PS consist of docosahexanoylresidues [5,6] and the presence of these acyl-chains appears to be es-sential for normal functioning of the nervous system [6–9]. The con-centration of PS also varies among different organelle membranes(reviewed in Ref. [10]); in general, the PS concentration is highestin plasma membranes and endosomes, but is very low in mitochon-dria, particularly in mitochondrial inner membranes. In addition, PSis not symmetrically distributed across the two leaflets of the mem-brane bilayer: PS is normally highly enriched in the inner, comparedto the outer, leaflet of the plasma membrane, whereas the choline-containing lipids, PC and sphingomyelin, are enriched in the outerleaflet [11–13]. Studies on the erythrocyte membrane indicate that>96% of PS resides on the inner leaflet of the bilayer [14]. It shouldbe noted, however, that little conclusive information is available onthe transbilayer distribution of phospholipids such as PS in organellesother than the plasma membrane.

2.2. Transbilayer movement of PS in the plasma membrane

The initiation of several important physiological processes causesa redistribution of PS from the inner, to the outer, surface of the plas-ma membrane of mammalian cells. For example, during the blood-clotting cascade, the transbilayer asymmetry of PS in the plasmamembrane of activated platelets is markedly altered so that PSbecomes exposed on the cell surface. Consequently, the clotting fac-tors V, VIII, X and prothrombin are recruited to the surface of plateletsso that blood coagulation is promoted [14–17]. Similarly, duringsperm maturation, the asymmetric distribution of PS in the plasmamembrane is dissipated and PS becomes exposed on the surface ofthe sperm [18].

A particularly well-characterized process is the transbilayer move-ment of PS from the inner to the outer leaflet of the plasma mem-brane during the early stages of apoptosis. The exposure of PS on

the surface of apoptotic cells represents a signal for the recognitionand engulfment of these cells via PS receptors that are expressed onthe surface of phagocytic cells [19–23]. Several candidate PS receptorshave been identified on macrophages [22–27]. Consequently, the re-lease of potentially toxic molecules from the dying cells is prevented.Some of the PS exposed during apoptosis was reported to be newlysynthesized [28]. Nevertheless, in cells in which total PS biosyntheticactivity was reduced by 95% [29], apoptosis progressed normallyand PS exposure on the cell exterior during apoptosis was notcompromised [30]. In addition, the exposure of PS on the surface ofred blood cells serves as a signal for eryptosis (the clearance of redblood cells from the circulation by phagocytic cells) [31,32]. More-over, not only PS, but the PS hydrolysis product, lyso-PS, is exposedon the surface of activated and dying neutrophils thereby initiatingthe clearance of these cells during acute inflammation [33,34]. Thecell surface exposure of lyso-PS has also been reported to be a signalfor the activation of mast cells and for platelet degranulation [35,36].

The mechanisms by which the asymmetric transbilayer distributionof PS in the plasmamembrane is established, maintained and dissipatedhave been extensively investigated. ATP-dependent aminophospholipid(PS) flippase activities have recently been identified in yeast asmembersof the P-type ATPase family of transporters (i.e. Atp8a1 and Drs2[37,38]). Orthologs of these proteins are likely to play similar roles inmammals since elimination of Atp8a1 in mice dramatically increasedthe externalization of PS in hippocampal neurons, and even impairedhippocampal learning [39]. Another protein that has been proposed tomediate the transbilayer movement of PS in the plasma membrane isthe calcium-dependent protein, scramblase-1, that can randomize thedistribution of phospholipids such as PS across membrane bilayers[40]. During apoptosis, the intracellular concentration of calcium in-creases so that the aminophospholipid translocase activity is inhibitedwhereas the scramblase is activated. Consequently, it has been proposedthat these complementary changes induce the translocation of PS tothe external leaflet of the plasma membrane during apoptosis [21].However, whether or not scramblase-1 does indeed function as anaminophospholipid translocase in the plasma membrane remainsunclear since targeted deletion of scramblase-1 in mice did not reducethe transbilayer movement of phospholipids in the plasma membrane[41].

2.3. Intracellular functions of PS

Although the extracellular functions of PS have been most exten-sively studied, PS also participates in many intracellular processes. Akey function of PS is as the precursor of PE via the mitochondrial en-zyme PS decarboxylase; this function of PS will be discussed in detailin Section 5.4).

Many of the intracellular functions of PS appear to depend on itsanionic nature. For example, some key signaling proteins, such asthe tyrosine kinase Src, as well as the Ras and Rho family of GTPases,contain positively-charged motifs that bind to PS, thereby contribut-ing to the membrane targeting and activation of these proteins[42–45]. A recently discovered function of PS is its ability to targetproteins to phagosomes ([46]; reviewed in Ref. [47]). In these studies,the PS-binding C2 domains of GFP-tagged discoidin and lactadherinwere expressed intracellularly and used to assess the intracellulardistribution of PS, particularly during phagocytosis (reviewed in Ref.[47]). These studies demonstrated that the highest concentration ofPS is in the plasma membrane and endocytic organelles, and also in-dicated that PS is enriched in the luminal, compared to the cytosolic,leaflet of the ER [48].

PS also modifies the catalytic activity of several key signaling pro-teins that contain C2 domains, such as synaptotagmin, dynamin-1[49], protein kinase C [50] and Annexin V [51]. In addition, the specificbinding of PS to the PH domain of evectin-2 in recycling endosomesappears to be required for retrograde membrane trafficking [52].

ERPC

PSS1

PS PSMAM

MITO

PSS2 PSD

PE PE

LPEATlyso-PE

ERPE

etn

Fig. 2. Subcellular locations of enzymes involved in PS and PE biosynthesis. PS is made bythe exchange of serine for the choline or ethanolamine head-groups of phosphatidylcholine(PC) or PE by PSS1 and PSS2, respectively, that are located on mitochondria-associatedmembranes (MAM). MAM consist of a specific region of the ER that becomes transientlytethered to mitochondrial (MITO) outer membranes. Newly-made PS is transferred viaMAM to mitochondria where the PS is decarboxylated to PE by PS decarboxylase (PSD)on mitochondrial inner membranes. PE is also produced in MAM by acylation of lyso-PEby lyso-PE acyltransferase (LPEAT). In addition, PE can be synthesized by the Kennedy path-way in which ethanolamine (etn) is sequentially converted to phosphoethanolamine, thenCDP-ethanolamine, by cytosolic enzymes. In the final reaction of this pathway, PE is gener-ated in the ER from CDP-ethanolamine and diacylglycerol.

545J.E. Vance, G. Tasseva / Biochimica et Biophysica Acta 1831 (2013) 543–554

Furthermore, PS binds to the PH domains of 3-phosphoinositide-dependent kinase-1 [53] and Akt [54], and disruption of the bindingof PS this domain of Akt inhibits Akt signaling and attenuates cell sur-vival [54]. Interestingly, highly specific interactions of PS with someheat shock proteins such as Hsp70 induces formation of ion channelsin the plasma membrane [55]. Thus, although PS represents a quanti-tatively minor phospholipid in mammalian cells, the presence of PS isrequired for many fundamental cellular processes. Indeed, the essen-tial requirement of PS in mammalian cells was highlighted by theobservation that mice in which PS synthesis was completely eliminat-ed did not survive during development [56].

3. Biosynthesis of phosphatidylserine

3.1. The PS synthases

The pathway for PS biosynthesis depends upon the type of organ-ism. In prokaryotes and the yeast Saccharomyces cerevisiae, all PS is syn-thesized by a PS synthase that uses CDP-diacylglycerol and L-serine(Fig. 1). In contrast, in mammalian cells PS is synthesized solely bycalcium-dependent base-exchange reactions in which the polarhead-group (choline or ethanolamine) of a pre-existing phospholipid(PC or PE, respectively) is exchanged for L-serine [57] (Fig. 1). WhereasEscherichia coli and the yeast S. cerevisiae contain only a single PSsynthase [58,59], mammalian cells express two PS synthases (PSSs) –

PSS1 and PSS2 – that are encoded by distinct genes and have differentsubstrate specificities. PSS1 synthesizes PS via the exchange of L-serinefor the choline head-group of PC, whereas PSS2 exchanges serine forthe ethanolamine moiety of PE (Fig. 1) (reviewed in Ref. [1]). Despiteextensive investigations, no enzymatic activity that catalyzes the forma-tion of PS from CDP-diacylglycerol and serine has been detected inmammalian cells. Although plants were also originally thought tomake PS exclusively by a base-exchange reaction, a PS synthase that isencoded by a cDNA with 54% identity to the yeast PS synthase hasbeen identified in wheat plants [60]. In the parasite Trypanosoma bruceiPS ismade entirely by a base-exchange reaction of PEwith L-serine [61].

The generation of Chinese hamster ovary (CHO) mutant cells thatsynthesize PS via serine-exchange with PE, but not PC (Fig. 1), con-firmed the existence of two distinct PSSs [62,63]. The rate of PS synthe-sis in these mutant CHO cells was ~50% lower than in parental CHO-K1cells and the amounts of PS and PEwere also reduced. Furthermutagen-esis of these mutant cells generated a cell line in which total serine-exchange activity, due to residual activities of PSS1 and PSS2, was only5% of that in parental CHO-K1 cells [29]. Survival of themutant cells re-quired supplementation of the culture medium with PS. The cells didnot, however, survive when supplemented with PE in the absence ofPS. Since somePE is derived fromPS (Section 5.4), a reduction in PS syn-thesis also leads to a reduction in PE [29]. However, the observation thatsupplementation of the mutant cells with PE did not increase survivalsuggests that a reduction in PS, rather than PE, was the cause of the de-creased survival of the mutant cells. Thus, the requirement of CHO cellsfor PS appears to be distinct from the requirement for PS-derived PE. Onthe other hand, it is possible that exogenously-delivered PE does not ef-fectively reach the relevant cellular pool (perhaps mitochondria).

The yeast PS synthase has been purified and regulation of its activ-ity has been extensively characterized [64]. cDNAs encoding PSS1 andPSS2 have been cloned from hamster and mouse, and ectopicallyexpressed [65,66]. Although neither mammalian PS synthase hasbeen purified to homogeneity, a serine-exchange enzymewas partial-ly purified from rat brain [67]. This enzyme catalyzed the exchange ofserine with the ethanolamine head-group of PE, but not with the cho-line moiety of PC, and is now presumed to have been PSS2 (Fig. 2).More recently, epitope-tagged versions of PSS1 and PSS2 were puri-fied to near homogeneity [68,69]. Although little is known aboutfunctional domains or the active sites of the mammalian PSSs, severalamino acids that are required for serine-exchange activity have been

identified [70]. The predicted amino acid sequences of mammalianPSS1 and PSS2 are ~30% identical [29,65,66,71] and each PSS appearsto contain several membrane-spanning domains. Consistent withthese predictions, both mammalian PSS activities are present onmicrosomal membranes [72–76]. However, neither enzyme appearsto be primarily localized to the bulk of the ER. Instead, these enzymesare largely restricted to a specialized domain of the ER called“mitochondria-associated membranes” (MAM) [77] (Section 4).

3.2. Regulation of PS synthesis

The mechanisms that regulate PS synthesis in mammalian cells re-main poorly understood. Early experiments suggested that PS syn-thesis is regulated by a phosphorylation reaction involving proteinkinase C [78] but more recent studies have not augmented this report.Interestingly, unlike plants in which large amounts of PS accumulatewhen PSS1 is over-expressed [60], the PS content of mammaliancells appears to be tightly regulated even when the capacity for PSsynthesis is increased [65,66]. Over-expression of PSS1 activity inMcArdle rat hepatoma cells increased the rate of PS synthesis where-as over-expression of PSS2 did not, suggesting that PSS1, but notPSS2, activity can be rate-limiting for PS synthesis [65,66].

A key mode of regulation of PS synthesis is end-product feedbackinhibition in which increased levels of cellular PS reduce the rate of PSsynthesis [79]. A mutant CHO cell line was isolated in which PS syn-thesis was resistant to inhibition by PS. In these cells, PS levels are


2.5-fold higher than in parental CHO cells [62,80]. Furthermore, pointmutations were identified in PSS1 [68,79] and PSS2 [81,82] that con-ferred resistance to end-product inhibition of PS synthesis.

The abundance of Pss1 and Pss2mRNAs and their encoded PSS activ-ities varies during mouse development and among different mouse tis-sues [83,84]. For example, Pss1mRNA and activity are particularly highin the brain [56,83] whereas Pss2 is most highly expressed in Sertolicells of the testis [84]. Our laboratory has examined mechanisms bywhich expression of Pss1 is transcriptionally regulated in cells of thebrain. Pss1 mRNA is more abundant in primary astrocytes than in neu-rons isolated from the mouse cortex [85]. In intact astrocytes, and inin vitro studies of promoter activity, themurine Pss1 proximal promoterinteracted with, and was activated by, the transcription factors N-Mycand Sp1. Moreover, RNA silencing of these factors decreased promoteractivity, and disruption of Sp/DNA binding in astrocytes reduced PSS1enzymatic activity. Thus,N-Myc and Sp1 regulate Pss1 expression in as-trocytes isolated frommouse brain [85]. Clearly, in light of the emergingimportance of PS for key cellular processes,more detailed information isrequired on how cellular amounts, and the subcellular distribution, ofPS are regulated, at the levels of synthesis and degradation.

3.3. Targeted disruption of Pss1 and Pss2 in mice

The finding that mammalian cells express two PSS isoforms that areencoded by distinct genes raised several questions. For example: whydomammalian cells express two different PSSs? Can one PSS substitutefor the other? In an attempt to answer these questions my laboratoryindividually eliminated PSS1 and PSS2 in mice. Pss2−/− mice are viable,have a normal life-span and appear outwardly normal. PSS activity inthe testis was reduced by >90% with no decrease in PS or PE content[84]. Nevertheless, testis weight was reduced and some male micewere infertile [84]. Apparently, compensatory mechanisms were in-duced to maintain normal PS levels in the absence of PSS2, since inhepatocytes isolated from Pss2−/− mice, PS degradation was decreasedand PSS1 activitywas increased [86]. Thus, although hepatic Pss2mRNAlevels are ~80-fold higher during embryonic development than in adultmice [86], PSS2 is not essential for mouse development, and for themost part, PSS1 can substitute for lack of PSS2.

Targeted disruption of the Pss1 gene inmice also yielded viablemicethat were outwardly indistinguishable from their Pss1+/+ littermates[56]. Thus, PSS2 can substitute for a lack of PSS1 in mice. The findingthat the viability of Pss1−/− mice was not compromised was unexpect-ed because survival of mutant CHO cells lacking PSS1 is severelyimpaired [63,87]. Although PSS activity was markedly lower (by up to85%) in tissues of Pss1−/− mice than in Pss1+/+ mice, the amounts ofPS in tissues, with the exception of the liver, were not significantly re-duced. Intercrosses of Pss1−/− and Pss2−/− mice yielded mice with 3disrupted Pss alleles. The mice were viable and the amounts of PS andPE were modestly reduced in most tissues. Importantly, however, nodouble knock-out (Pss1−/−/Pss2−/−) mice survived [56]. Thus, micecan tolerate the elimination of either PSS1 or PSS2, and as little as 10%of normal PSS activity, but the complete absence of both PSSs is incom-patible with survival, demonstrating that mouse tissues/cells require athreshold amount of PS. However, it is not yet clear whether thisrequirement is due to solely to the important cellular functions of PSper se (Section 2), or whether part of this requirement can be attributedto the requirement of PS as the substrate for PE production inmitochon-dria via PS decarboxylation (Section 5).

4. PS import into mitochondria

4.1. Mechanism of PS transport

A major use of PS is its conversion into PE by the mitochondrial en-zyme PS decarboxylase (PSD) (Fig. 2). Although yeast contains two dis-tinct proteins that catalyze PS decarboxylation (Psd1 in mitochondria

and Psd2 in the Golgi/vacuole [88,89]), mammalian cells contain onlya single PSD that is located in mitochondria [90,91]. Since PS is synthe-sized in elements of the ER, and since PSD activity is restricted to mito-chondria in mammalian cells, the mechanism by which PS is importedinto mitochondria from the ER has been investigated in some detailin mammalian cells [92–96] and yeast [97,98] (Fig. 2). The inter-organelle transfer of PS is not inhibited when cells are permeabilized,implying that the transport does not require either small vesicles or sol-uble cytosolic lipid transfer proteins [93,95]. Instead, PS import into mi-tochondria appears to bemediated by transient membrane contact, butnot fusion, between a sub-fraction of the ER [now called mitochondria-associated membranes (MAM)] and mitochondrial outer membranes[75,94,95,99,100] (Fig. 2). In addition, a pool of newly-synthesized PSappears to be compartmentalized and channeled into mitochondriavia MAM for conversion of the PS into PE [94]. The rate-limiting stepin this conversion is the transfer of PS from the ER/MAM to the mito-chondrial outer membrane (Fig. 2). This transfer occurs with a half-time of 6.5 h [92], and in permeabilized CHO cells and baby hamsterkidney cells the transport requires ATP, although the precise functionof ATP in this translocation has not been clearly defined [101–103].

The transport of PS from its site of synthesis on MAM to mitochon-drial outer membranes is not the only lipid translocation step that isrequired for the generation of PE from PS. The PS must also undergotransbilayer movement across the mitochondrial outer membraneand subsequently be delivered from the inner leaflet of the mitochon-drial outer membrane, across the inter-membrane space, to the activesite of PSD on the outer leaflet of the mitochondrial inner membrane[72,91]. The mechanisms underlying these lipid translocations arelargely unknown, although there is evidence that PS transfer betweenthe outer and inner membranes of mitochondria occurs via mem-brane contact sites [104,105].

4.2. Characterization of mitochondria-associated membranes (MAM)

The MAMwere first isolated from rat liver as an ER-like membranefraction that co-isolated with mitochondria [75,94], suggesting thatMAM represent a specific domain of the ER that becomes transientlytethered to mitochondrial outer membranes. The MAM possess many,but not all, properties of the bulk of ER membranes [75,106]. For exam-ple, MAM are enriched, compared to the bulk of the ER, in several lipidbiosynthetic activities, such as PSS1 and PSS2 [77], as well as acyl-CoA:cholesterol acyltransferase and 1,2-diacylglycerol acyltransferase [106].On the other hand, a commonly-used ERmarker enzyme, NADPH:cyto-chrome c reductase, is relatively depleted from MAM [75]. A specificmarker protein for MAM in primary rodent hepatocytes is phosphati-dylethanolamine N-methyltransferase-2, an enzyme that sequentiallymethylates PE to form PC [107]. A membrane fraction that appears tobe equivalent to MAM has also been isolated and characterized fromseveral other types of mammalian cells [95,104,108–110], as well asS. cerevisiae [99,111].

Close apposition (~20 nm) between the ER and mitochondria hasbeen observed morphologically for several decades [108,110,112–117].Indeed, up to 20% of the surface of the ER appears to be closely juxta-posed with mitochondria [108,115]. Moreover, high-resolution 3D elec-tron tomography revealed the existence of many ~15 nm contact zonesbetween ER and mitochondria [118,119]. Mitofusin2, a protein that isrequired for mitochondrial fusion, was identified in mouse embryonicfibroblasts as a “docking” protein that transiently tethers MAM to mito-chondria [117]. Consistent with this idea, the distance between mito-chondria and ER/MAM is increased in cells that lack mitofusin2 [117].Moreover, disruption of mitochondria/ER connections in mitofusin2-depleted mammalian cells markedly reduces starvation-inducedautophagosome biogenesis [120], and the liver-specific elimination ofmitofusin2 in mice disrupted ER/mitochondria contacts and impairedglucose tolerance and insulin signaling [121]. The average distancebetween ER and mitochondria was also increased in pulmonary arterial


smoothmuscle cells thatwere hypertensive [122]. In yeast, MAMappearto be tethered to mitochondria via specific docking complexes thatcontain the MET30 protein (a ubiquitin ligase) and phosphoinositides([111]; reviewed in Refs. [123,124]). The Mmm1/Mdm10/Mdm12 com-plex has also been implicated in the tethering of ER/MAM to mitochon-dria in yeast, and this complex is required for the efficient inter-organelle exchange of phospholipids and for maintaining normalcalcium homeostasis [125]. Interestingly, PS import into mitochondriain yeast requires donormembranes with a high content of PS, consistentwith the finding that the two mammalian enzymes that synthesize PS(PSS1 and PSS2) are highly enriched in MAM [77]. These observationssuggest that negatively charged, PS-rich domains promote contact be-tween MAM and mitochondria [126].

Recently, interest in the function of MAM has exploded with the dis-covery that close contact betweenMAMandmitochondria is required forcalcium exchange between these organelles [108,115,117,119,127]. Cal-cium release channels (inositol tris-phosphate receptors) have been lo-calized to MAM in living cells [108,115]. Moreover, mitochondria-MAMcontact zones are thought to generate a high local concentration of calci-um and to facilitate the exchange of calcium between the ER and mito-chondria [108,115,119,128,129]. The formation of MAM-mitochondriacontact sites is reversible and is regulated by the level of cytosolic calcium[116,117] and by the sorting protein, PACS [110,130]. MAMwere also re-cently suggested to play an important role in mitochondrial fusion [131].

The research summarized above demonstrates that tethering be-tween elements of the ER (MAM) and mitochondria is importantnot only for supplying PS for PE synthesis in mitochondria, but alsofor the regulation of fundamental biological processes such as mito-chondrial function and function, as well as calcium homeostasis.

5. Phosphatidylethanolamine

5.1. Functions of PE

PE comprises ~25% of mammalian phospholipids and is particularlyenriched in the brain where the PE content is ~45% of total phospho-lipids. PE appears to play an important role in the heart since a decreasedPE content of cardiacmyocytes causes cell damage after ischemia, and al-tered asymmetrical transbilayer distribution of PE in sarcolemmalmem-branes disrupts thesemembranes [132]. PE is the phospholipid substratefor the hepatic enzyme PE N-methyltransferase [133,134] that providesapproximately one third of the PC in the liver.Moreover, PE is a precursorfor the synthesis of anandamide (N-arachidonoylethanolamine), theligand for cannabinoid receptors in the brain [135,136]. Mitochondria,particularly the inner membranes, are enriched in PE compared toothermembranes and a decrease in themitochondrial content of PE pro-foundly alters mitochondrial morphology in yeast [137], T. brucei [138]andmammalian cells [139] (Section 5.4). Convincing in vitro experimen-tal evidence supports a role for PE, a cone-shaped lipid, in membranefusion since PE has the tendency to form non-lamellar membrane struc-tures and modulates membrane curvature [140–143]. Consistent withthese findings, PE plays an important role in contractile ring disassemblyat the cleavage furrow during cytokinesis of mammalian cells [144,145],and a lack of PE inhibits progression of the cell cycle in the parasiteT. brucei, probably by impairingmembrane fusion [138]. PE also regulatesthe fusion of mitotic Golgi membranes [146].

A unique role for PE was discovered in E. coli as a deficiency of PEin this organism prevented the correct folding, and altered the topol-ogy, of integral membrane proteins such as lactose permease. Theseobservations suggest that PE can act as a “lipid chaperone,” at leastin E. coli [147–149]. Moreover, in Drosophila, a cholesterol auxotrophin which PE is the most abundant phospholipid, PE regulates the pro-cessing of the transcription factor, sterol regulatory element bindingprotein [151]. In contrast, in mammalian cells this protein is the mas-ter regulator of cholesterol homeostasis, and processing of this pro-tein is regulated by level of cholesterol, not PE, in the ER [152]. Very

recently, and unexpectedly, PE was identified as the single endoge-nous factor in the brain that is required for the propagation andinfectivity of mammalian prions although the precise mechanism un-derlying this requirement is not yet clear [150].

PE is also the donor of the ethanolamine moiety that cova-lently modifies several proteins. For example, the synthesis of theglycosylphosphatidylinositol anchors that attach many signaling pro-teins to the surface of the plasma membrane requires the covalent at-tachment of ethanolamine molecules that are derived from PE [153].In addition, PE is the source of ethanolamine that is covalently at-tached to the eukaryotic elongation factor eEF1A [61]. Intriguingly, afunction for PE has recently been discovered in autophagy. This pro-cess requires the participation of many cytosolic proteins includingthe ubiquitin-like protein LC3 (the mammalian homolog of theyeast Atg8p), a commonly used marker of autophagosomes. As a re-quirement for autophagosome formation, LC3 is modified by the co-valent attachment of PE, thereby recruiting LC3 to membranes andmediating membrane hemifusion during autophagosome formation[154]. In the final steps of autophagy, the autophagosomes fuse withlysosomes so that the cytosolic components that were engulfed dur-ing autophagosome formation are degraded. A continuing debate iswhether autophagosomes originate from pre-existing organellemembranes, such as the ER, mitochondria and/or plasma membranes,or from a de novo pathway. Some evidence supports a role of mito-chondria and ER/mitochondria connections in providing the PE thatis used for attachment to LC3 and for autophagosome formation[120]. As discussed in Section 5.4, mitochondrial PE is largely generat-ed in mitochondrial inner membranes by the decarboxylation of PSthat is made in the ER/MAM from PC (via PSS1) and PE (via PSS2)(Fig. 2).

Thus, in addition to its structural role in membranes PE serves mul-tiple important cellular functions thatwere until recently unrecognized.

5.2. Pathways of phosphatidylethanolamine biosynthesis

Inmammalian cells PE can be synthesized by four biosynthetic path-ways (Fig. 3), two of which areminor contributors. The twomajor path-ways are (i) the Kennedy pathway (i.e. the CDP-ethanolamine pathway[155]), the final step of which takes place on ER membranes [75,156],and (ii) the PS decarboxylation pathway, which occurs onmitochondri-al inner membranes [90,91,157]. A third PE biosynthetic pathway is theacylation of lyso-PEwhich is active in yeast andmight also be importantin mammalian cells [158–160] (Fig. 3). Interestingly, the majority oflyso-PE acyltransferase activity is localized to the MAM [160] providinga possible explanation for why exogenously-added lyso-PE generatesthe pool of PE in mitochondria far more efficiently than doesexogenously-supplied ethanolamine [158–160]. PE can also be pro-duced in the ER by a calcium-dependent, base-exchange reactionbetween PS and ethanolamine, catalyzed by PSS2; this source of PE isgenerally considered to be quantitatively minor in mammalian cells[161–163] (Fig. 3).

Thus, at least two spatially distinct pools of PE can be synthesized inmammalian cells—one in mitochondria the other in the ER/MAM. Con-sistent with the idea that cellular PE is compartmentalized into func-tionally distinct pools on the basis of biosynthetic origin, the majorityof PE in mitochondria is synthesized in situ in mitochondria via PSD,whereas only a small fraction of mitochondrial PE is made in the ERby the CDP-ethanolamine pathway [95,164–167]. Stable isotope label-ing of PE made by the two pathways in hepatoma cells and CHO cellsdemonstrated that the pools of PE made by the CDP-ethanolaminepathway and the PSD pathway differ in their acyl-chain compositions.Whereas PSD preferentially synthesizes species of PE with polyunsatu-rated acyl chains, such as 20:4 and 20:5 at the sn-2 position, theCDP-ethanolamine pathway preferentially synthesizes PE containingmono- or di-unsaturated acyl chains, such as 18:2, at the sn-2 position[167]. Available data do not permit a definitive conclusion about

etn

EK

P-etn

ET

CDP-etn

EPTPE

PS

PS lyso-PE

PSS2 LPEAT

PSD

Fig. 3. Biosynthetic pathways for PE inmammalian cells. The two quantitativelymajor path-ways for PE synthesis in mammalian cells are the PS decarboxylase (PSD) pathway and theCDP-ethanolamine pathway. In the latter, ethanolamine (etn) is phosphorylated by ethanol-amine kinase (EK) to produce phosphoethanolamine which is converted into CDP-ethanolamine by CTP:phosphoethanolamine cytidylyltransferase (ET). In the final reactionof this pathway, CDP-ethanolamine:1,2-diacylglycerol ethanolaminephosphotransferase(EPT) combines diacylglycerol with CDP-ethanolamine to produce PE. Small amounts of PEare synthesized by acylation of lyso-PE with lyso-PE acyltransferase (LPEAT), and also by abase-exchange reaction inwhichPS is converted to PEby the exchangeof serine in PS for eth-anolamine by the action of PS synthase-2 (PSS2).


whether the PS used for decarboxylation to PE is derived primarily fromPC (via PSS1) or PE (via PSS2). In rat hepatoma cells, analysis of theacyl-chain composition of PS, PE and PC bymass spectrometry revealedthat the predominantmolecular species of PSwas 18:0–18:1whichwasnot the major molecular species of the bulk of either PC or PE [167].Nevertheless, it is difficult to conclude from these studies whether thePS used for decarboxylation is primarily derived from PC or PE becausedeacylation–reacylation remodeling reactions can rapidly modify theacyl-chain composition of phospholipids. In addition, it is possible thatspecific molecular species of PS are selected for import into mitochon-dria for decarboxylation.

The relative contributions of the four PE biosynthetic pathways(Fig. 3) appears to depend on the cell type. In rat liver/hepatocytes[167–170] and hamster heart [171] the CDP-ethanolamine pathwayhas been reported to provide the majority of PE; in T. brucei, PE ismade solely by this pathway [172]. On the other hand, in othercells, such as baby hamster kidney cells and CHO cells, >80% of PEis derived from the PSD pathway even when ethanolamine is sup-plied as a substrate for the CDP-ethanolamine pathway [173,174].Nevertheless, data on the quantitative contribution of the two majorPE biosynthetic pathways in mammalian cells are conflicting andcomplex and the reasons underlying these differences remain to beresolved. It is also noteworthy that when the contribution of the PEbiosynthetic pathways is assessed in radiolabeling experiments, theassumption is made that the pool of the immediate precursor of PEis homogeneously radiolabeled, an assumption that is not necessarilyvalid [175]. Moreover, another complicating factor is that differentcell types can adapt to artificial conditions that might not reflect thein vivo situation.

5.3. PE synthesis by the CDP-ethanolamine pathway

The CDP-ethanolamine pathway for PE synthesis (Fig. 3) consists ofa sequence of reactions parallel to that of the CDP-choline pathway forPC synthesis [155]. The first step of the CDP-ethanolamine pathway iscatalyzed by ethanolamine kinase, a cytosolic protein. Two genesencoding ethanolamine kinase activity have been identified inmamma-lian cells [176]. One of these kinases (ethanolamine kinase-1) phos-phorylates both ethanolamine and choline, whereas the other isoformphosphorylates only ethanolamine, but not choline [176,177]. Thetissue distribution of the two ethanolamine kinases is different: etha-nolamine kinase-1 is widely expressed throughout mouse tissueswhereas ethanolamine kinase-2 is most highly expressed in liver andreproductive tissues. To determine the functions of the two ethanol-amine kinases in intact animals, mice were generated in which thegene encoding the ethanolamine-specific kinase-2 was disrupted[176]. Fewer pups were born to the knock-out mice than to theirwild-type counterparts and a significant fraction of the newbornknock-out pups failed to survive. However, the PE content of the liversof the surviving mice was not reduced suggesting that either thedual-specificity choline/ethanolamine kinase or the PSD pathway waspartially able to substitute for the lack of the ethanolamine-specific ki-nase [176].

Ethanolamine is a required substrate for the ethanolamine kinasereaction. However, since mammalian cells cannot synthesize ethanol-amine de novo ethanolamine must be provided either from the diet orfrom the degradation of PE made by the PSD pathway. In addition,small amounts of ethanolamine phosphate are generated from thebreakdown of sphingosine phosphate by the action of sphingosinephosphate lyase [178,179]. The latter source of ethanolamine is gen-erally considered to provide only small amounts of ethanolamine inmammalian cells [179]. In plants, ethanolamine can be generated bythe direct decarboxylation of serine [180] but this pathway has notbeen detected in mammalian cells.

The rate-limiting reaction of the CDP-ethanolamine pathway for PEsynthesis is, under most metabolic conditions, catalyzed by the secondenzyme in the pathway: CTP:phosphoethanolamine cytidylyltransferase(ET), the product of the Pcyt2 gene in mice (Fig. 3) [168,181,182]. Undersome conditions, however, ethanolamine kinase has been reported tobe rate-limiting for this pathway [177]. ET catalyzes the formation ofCDP-ethanolamine from phosphoethanolamine and CTP [168]. Only asingle gene that encodes ET activity has been identified in mice[163,183,184] but the encoded protein has extensive regions of homologyto the corresponding cytidylyltransferase of the CDP-choline pathway forPC synthesis. An interesting feature of the ET protein is that whereas CTP:phosphocholine cytidylyltransferase contains only a single copy of thecatalytic domain, ET contains two copies of the presumed catalytic motif[183]. Moreover, in contrast to CTP:phosphocholine cytidylyltransferase,ET is not found in the nucleus of mammalian cells, nor is its activity regu-lated by reversible translocation between a soluble pool andmembranes.

Among mouse tissues, Pcyt2 mRNA is most abundant in liver, heartand skeletal muscle [163,185]. When ET was globally eliminated inmice, all the mice died before birth (prior to embryonic day 8.5). Sincethe PSD pathway for PE synthesis was active in these mice, this sourceof PE was, apparently, unable to substitute for the absence of ET andthe CDP-ethanolamine pathway. Nevertheless, Pcyt2+/− mice wereovertly normal and their tissues contained normal amounts of PE prob-ably because PE degradation was down-regulated and/or PE synthesisby other pathways was activated [185]. However, the amounts ofdiacylglycerols and triacylglycerols were markedly increased in liversof Pcyt2+/−mice, accompanied bymarked changes in hepatic lipid me-tabolism, and increased glucose and insulin intolerance [186]. In otherstudies, the Pcyt2 gene was disrupted specifically in hepatocytes of in-tact mice. Although livers of the mice appeared normal, the acyl chaincomposition of hepatic PE was altered and reflected the compositionof PS, indicating that all of the PE had been generated from PS


decarboxylation (Fig. 3). Moreover, as was the case with the Pcyt2+/−

mice [186], the Pcyt2−/− hepatocytes accumulated large amounts oftriacylglycerols and the mice developed hepatic steatosis [163]. Thus, al-though the PSD pathway can satisfy the requirement of the liver for PE inthe absence of the CDP-ethanolamine pathway both during developmentand in the adult animal, global elimination of the CDP-ethanolaminepathway in all tissues of the mouse is embryonic lethal. These studiesindicate that PE made by the CDP-ethanolamine pathway is required innon-hepatic tissues during mouse development.

Thefinal stepof theCDP-ethanolamine pathway for PE synthesis is thereaction between diacylglycerol and CDP-ethanolamine, catalyzed byCDP-ethanolamine:1,2-diacylglycerol ethanolaminephosphotransferase,an integral protein that resides primarily in ER membranes [72,74,187].Characterization of the ethanolaminephosphotransferase, and elucidationof its molecular structure, has been hampered by difficulties in purifica-tion of this integral membrane protein. However, a yeast gene thatencodes a dual specificity ethanolamine/choline phosphotransferase ac-tivity was identified [188], and the human ortholog of this gene was sub-sequently cloned [187]. As is the case for the yeast enzyme, the humanenzyme exhibits both choline- and ethanolamine-phosphotransferaseactivities and was originally proposed to be the sole protein catalyzingthe final reaction of the CDP-ethanolamine pathway [187]. However, an-other human ethanolaminephosphotransferase, that appears to be specif-ic for CDP-ethanolamine,wasmore recently identified and is presumed toparticipate in the CDP-ethanolamine pathway for PE synthesis [189]although its distribution in mouse tissues (mainly in testis and small in-testine) is more restricted than that of the dual-specificity enzyme. It isgenerally thought that elimination of ethanolaminephosphotransferaseactivity in mice would be incompatible with life since global disruptionof the single mouse gene encoding ET, a component of the same PE bio-synthetic pathway, is embryonic lethal [185].

5.4. PE synthesis via PS decarboxylation

In mammalian cells, PE can be synthesized from PS by a single PSDactivity that is present on mitochondrial inner membranes [157].S. cerevisiae, on the other hand, contains two PSDs that are encodedby distinct genes. The yeast Psd1 protein is similar to the mammalianPSD and is located on mitochondrial inner membranes, whereas theyeast Psd2 protein has a sequence that is completely different fromthat of PSD1/Psd1 and is present in the Golgi/vacuole [88,89]. ThePSDs in mammalian and yeast cells are membrane-bound proteins.However, an unusual soluble, secreted PSD protein with limited se-quence homology to the mitochondrial PSD was recently identifiedin the parasite Toxoplasma gondii [190]. In addition, soluble truncatedPSD proteins have been detected in the parasite Plasmodium knowlesi[191]. Interestingly, another parasite, T. brucei, synthesizes PE entirelyby the CDP-ethanolamine pathway [172]. It is noteworthy that the PSthat is used for decarboxylation to PE in mitochondria is synthesizedin the ER/MAM from PC (via PSS1) and PE (via PSS2) (Fig. 2).

The mammalian PSD belongs to a small group of decarboxylasesthat contain a pyruvoyl group at the active site [192]. The mature, ac-tive form of PSD is a heterodimer consisting of two subunits, α and β,that are generated from a single large precursor molecule by an auto-catalytic cleavage reaction that occurs within a LGST motif [193–195].During this autocatalytic proteolysis, the serine residue in this motif isconverted into a pyruvoyl group that becomes the N-terminus of theα subunit. cDNAs encoding PSD activity were identified in CHO cells[196] and S. cerevisiae [197]. The corresponding mouse Pisd genewas subsequently disrupted to determine the requirement of thePSD pathway for PE synthesis in intact animals [139]. No Pisd−/−

mice survived beyond embryonic day 9.5 of development demon-strating that the PSD pathway is essential for mouse development.This observation, combined with studies showing that global disrup-tion of the CDP-ethanolamine pathway is also embryonic lethal [185],strongly implies that each of the two major pathways for mammalian

PE biosynthesis is required for specific functions and for mouse viabil-ity. Alternatively, it is possible that insufficient PE is available unlessboth pathways are operational.

Immunofluorescence microscopy studies in Pisd−/− mouse embry-onic fibroblasts, and electron microscopy studies in Pisd−/− embryos,revealed that elimination of PSD profoundly altered mitochondrialmorphology and caused extensive mitochondrial fragmentation [139].Our recent studies in CHO cells, in which Pisd expression was reducedby ~90% by RNA silencing, have also shown widespread changes in mi-tochondrialmorphology and defects inmitochondrial functions, such asATP production, oxygen consumption and activities of components ofthe electron transport chain, in response tomitochondrial PE deficiency(G. Tassseva and J.E. Vance, unpublished observations). Pisd+/−mice, incontrast to Pisd−/− mice, appear outwardly normal, and the morpholo-gy of mitochondria and the PE content of tissues of these mice areunaffected. However, the amount and activity of ET in hepatocytesfrom the Pisd+/− mice were higher than in their Pisd+/+ littermates,presumably because of a compensatory increase in flux through theCDP-ethanolamine pathway in response to the lack of PE synthesis viaPSD. Nevertheless, the addition of ethanolamine to the diet of pregnantfemale Pisd+/− mice did not permit survival of any Pisd−/− pups [139].

The embryonic lethality and mitochondrial abnormalities causedby elimination of PSD activity in mice underscore similar findings inother organisms, such as yeast in which PSD activity is required fornormal membrane biogenesis [164,165,198,199]. Remarkably, how-ever, E. coli is viable even when the levels of PE, normally the mostabundant phospholipid in this organism, are genetically reduced toas little as 0.007% of total phospholipids [59]. In yeast, the mitochon-drial Psd1 is responsible for ~90% of total PSD activity. Mutants lack-ing either Psd1 or Psd2 are viable, but when the mitochondrial PEcontent was reduced to b4% of total phospholipids (i.e. in cells inwhich PE was synthesized only by the CDP-ethanolamine pathwayor only by Psd2 in the Golgi/vacuole) survival of the yeast was severe-ly compromised [89,158,165,200]. Thus, PE production by the alter-native pathways does not appear to satisfy the mitochondrialrequirement for PE derived from Psd1, showing that a critical levelof mitochondrial PE is required for viability and normal mito-chondrial function. PE and the mitochondria-specific phospholipidcardiolipin are non-bilayer-forming lipids that in yeast haveoverlapping functions [201] and are involved in maintaining normalmitochondrial morphology [202]. Elimination of both PE andcardiolipin in yeast causes massive fragmentation of mitochondriaand loss of mitochondrial membrane potential [137]. Mitochondrialmorphology is also altered by depletion of PE in T. brucei althoughin this organism PE is made entirely by the non-mitochondrialCDP-ethanolamine pathway [138]. Thus, evidence is accumulatingthat the PE content of mitochondria is crucial for maintaining normalmitochondrial function.

The mechanism by which loss of mitochondrial PE causes such pro-found changes in mitochondrial morphology and function has not yetbeen firmly established. A factor that might contribute to these changesis the finding that PE plays a role in membrane fusion events, whichmight include mitochondrial fusion. In addition, a crystal structure ofthe ubiquitinol:cytochrome c reductase (Complex III) of the mitochon-drial electron transport chain revealed that PE is associated with thiscomplex, suggesting that PEmightmediate protein:protein interactionswithin this complex [203]. This suggestion is supported by the findingthat inclusion of PE in proteoliposomes reconstitutedwith this ComplexIII protein increased the respiratory control ratio [204,205]. Other stud-ies show that PE is also a component of crystallized bovine heart cyto-chrome c oxidase [206], and that NADH:ubiquinone oxidoreductase(Complex I) activity depends on the presence of bound PE [207]. Conse-quently, some of the detrimental effects on mitochondrial function thatare caused by depletion of mitochondrial PE might be attributable tochanges in the conformation of proteins of the electron transportchain caused by disruption of PE:protein interactions.


6. Ether-linked ethanolamine phospholipids

Approximately 20% of human phospholipids are ether-linked lipids(reviewed in Ref. [208]). Whereas the diacyl lipids, including PE, consistof glycerol-3-phosphate esterifed to acyl residues at the sn-1 and sn-2 po-sitions, the ether lipids contain an ether linkage at the sn-1 position. In ad-dition, the choline and ethanolamine plasmalogens (plasmenylcholineand plasmenylethanolamine, respectively) contain a cis double bond ad-jacent to the sn-1 ether linkage, whereas the 1-alkyl-2-acyl phospholipidsplasmanylcholine and plasmanylethanolamine contain the sn-1 etherlinkage but lack the cis-double bond. In general, the O-alk-1′-enyl groupis present in the ethanolamine plasmalogens whereas the O-alkyl groupis present in the plasmanylcholines. An exception is the heart in whichmost of the choline ether lipids consist of plasmenylcholines [209,210].In mammals, the heart, nervous tissues and inflammatory cells containthe highest concentrations of ether lipids [209]. For example, up to70% of the ethanolamine phospholipids in inflammatory cells areplasmenylethanolamines. In contrast, the diacyl phospholipids accountfor 95% of the phospholipids in the liver. The precise physiological func-tions of the ether lipids are still unclear.

The pathway for biosynthesis of the ether lipids is unusual in severalrespects (reviewed in Ref. [208]). For example, the generation of theether linkage at the sn-1 position from a fatty alcohol, and the insertionof the cis double bond, both involve reactions that are unique to etherlipid synthesis. Several reactions in thebiosynthetic pathwayoccur in per-oxisomes whereas other reactions occur in the ER. The final reaction ofthe ether phospholipid biosynthetic pathway is catalyzed by the same en-zymes that are used for the synthesis of the diacylglycerophospholipidsi.e. CDP-choline:1,2-diacylglycerol cholinephosphotransferase and CDP-ethanolamine:1,2-diacylglycerol ethanolaminephosphotransferase.

An additional point should be made regarding ethanolamineplasmalogen synthesis. In rat hepatoma cells the composition ofthe aliphatic chains of the ethanolamine plasmalogens resemblesclosely that of PE made from the CDP-ethanolamine pathway ratherthan the PSD pathway [167]. This finding was interpreted to demon-strate that ethanolamine plasmalogens are exclusively synthesizedvia the CDP-ethanolamine pathway [186]. However, the rationaleon which this conclusion is based is unlikely to be valid. TheCDP-ethanolamine pathway provides the CDP-ethanolamine that isutilized in the final reaction of plasmalogen synthesis (catalyzed byethanolaminephosphotransferase in the ER), but does not dictatethe nature of the substituents at the sn-1 and sn-2 positions ofthe plasmalogen molecule. The alkyl-acylglycerol moiety of theplasmalogens is made by a pathway that operates in peroxisomes byunique reactions in which a specific pool of 1-acyldihydroxyacetonephosphate is converted into 1-alkyldihydroxyacetone phosphate by re-action with a fatty alcohol (reviewed in Ref. [208]). This pathway iscompletely different, and is spatially separated, from the one that pro-duces the diacylglycerol that is incorporated into PE made via theCDP-ethanolamine pathway (reviewed in Ref. [208]).

7. Summary and future directions

In addition to contributing tomembrane structure, themetabolically-related aminophospholipids PS and PE participate in multiple facets ofmetabolism and cell biology that were, until recently, completely unan-ticipated. Many of the key experiments that revealed these functionswere performed in genetically modified mice and in mammalian cellmutants, as well as in other eukaryotic organisms such as yeast and par-asites. Recent studies have demonstrated the crucial requirement of PE inregulating mitochondrial function and morphology. Other studies haveindicated that PE plays an important role in autophagy. The finding thatMAM mediate the import of PS into mitochondria for decarboxylationto PE has resulted in the exciting discovery that these ER-mitochondriaconnections also regulate cellular calcium homeostasis andmitochondri-al fusion. Furthermore, PS is involved in the fundamental process of

apoptosis inwhich PS exposure on the surface of apoptotic cellsmediatesthe recognition and engulfment of these cells by PS receptors on macro-phages. A previously unrecognized function of PS is based on its anionicnature which is responsible for targeting specific proteins to defined in-tracellular membranes such as phagosomes. Clearly, additional studiesare required to establish the precise mechanisms underlying the require-ments for PS and PE in these cellular processes and also in uncovering thelargely unknown mechanisms that regulate the biosynthesis and degra-dation of PS and PE.

References

[1] J.E. Vance, Phosphatidylserine and phosphatidylethanolamine in mammaliancells: two metabolically-related aminophospholipids, J. Lipid Res. 49 (2008)1337–1387.

[2] E. Baer, J.Maurukas,M. Russell, Synthesis of enantiomeric a-cephalins, J. Am. Chem.Soc. 74 (1952) 152–157.

[3] J. Folch, Isolation of phosphatidylserine from brain cephalin, and identificationof the serine component, J. Biol. Chem. 139 (1941) 973–974.

[4] F. Baer, J. Maurukas, Phosphatidylserine, J. Biol. Chem. 212 (1955) 25–38.[5] A.M. Hicks, C.J. DeLong, M.J. Thomas, M. Samuel, Z. Cui, Unique molecular signatures

of glycerophospholipid species in different rat tissues analyzed by tandem massspectrometry, Biochim. Biophys. Acta 1761 (2006) 1022–1029.

[6] H.Y. Kim, Novel metabolism of docosahexaenoic acid in neural cells, J. Biol. Chem.282 (2007) 18661–18665.

[7] S. Suzuki, H. Yamatoya,M. Sakai, A. Kataoka,M. Furushiro, S. Kudo, Oral administra-tion of soybean lecithin transphosphatidylated phosphatidylserine improvesmemory impairment in aged rats, J. Nutr. 131 (2001) 2951–2956.

[8] L. Piccotti, C. Marchetti, G. Migliorati, R. Roberti, L. Corazzi, Exogenous phospho-lipids specifically affect transmembrane potential of brain mitochondria andcytochrome C release, J. Biol. Chem. 277 (2002) 12075–12081.

[9] R. Mozzi, S. Buratta, G. Goracci, Metabolism and functions of phosphatidylserinein mammalian brain, Neurochem. Res. 28 (2003) 195–214.

[10] J.E. Vance, R. Steenbergen, Metabolism and functions of phosphatidylserine,Prog. Lipid Res. 44 (2005) 207–234.

[11] P.K. Schick, K.B. Kurica, G.K. Chacko, Location of phosphatidylethanolamine andphosphatidylserine in the human platelet plasma membrane, J. Clin. Invest. 57(1976) 1221–1226.

[12] J.A. Higgins, W.H. Evans, Transverse organization of phospholipids across thebilayer of plasma-membrane subfractions of rat hepatocytes, Biochem. J. 174(1978) 563–567.

[13] C. Venien, C. Le Grimellec, The involvement of cytoskeletal proteins in the main-tenance of phospholipid topology in renal brush-border membranes, Biochim.Biophys. Acta 946 (1988) 307–314.

[14] A. Zachowski, Phospholipids in animal eukaryotic membranes: transverse asym-metry and movement, Biochem. J. 294 (1993) 1–14.

[15] E. Bevers, P. Comfurius, J. van Rijn, H. Hemker, R. Zwaal, Generation ofprothrombin-converting activity and the exposure of phosphatidylserine atthe outer surface of platelets, Eur. J. Biochem. 122 (1982) 429–436.

[16] P. Williamson, E.M. Bevers, E.F. Smeets, P. Comfurius, R.A. Schlegel, R.F.A. Zwaal,Continuous analysis of the mechanism of activated transbilayer lipid movementin platelets, Biochemistry 34 (1995) 10448–10455.

[17] R. Majumder, M.A. Quinn-Allen, W.H. Kane, B.R. Lentz, A phosphatidylserinebinding site in factor Va C1 domain regulates both assembly and activity ofthe prothrombinase complex, Blood 112 (2008) 2795–2802.

[18] B.M. Gadella, R.A. Harrison, The capacitating agent bicarbonate induces proteinkinase A-dependent changes in phospholipid transbilayer behavior in thesperm plasma membrane, Development 127 (2000) 2407–2420.

[19] V.A. Fadok, D.R. Voelker, P.A. Campbell, J.J. Cohen, D.L. Bratton, P.M. Henson,Exposure of phosphatidylserine on the surface of apoptotic lymphocytes trig-gers specific recognition and removal by macrophages, J. Immunol. 148 (1992)2207–2216.

[20] V.A. Fadok, A. de Cathelineau, D.L. Daleke, P.M. Henson, D.L. Bratton, Loss ofphospholipid asymmetry and surface exposure of phosphatidylserine is re-quired for phagocytosis of apoptotic cells by macrophages and fibroblasts,J. Biol. Chem. 276 (2001) 1071–1077.

[21] K. Balasubramanian, B. Mirnikjoo, A.J. Schroit, Regulated externalization ofphosphatidylserine at the cell surface: implications for apoptosis, J. Biol. Chem.282 (2007) 18357–18364.

[22] M. Miyanishi, K. Tada, M. Koike, Y. Uchiyama, T. Kitamura, S. Nagata, Identifica-tion of Tim4 as a phosphatidylserine receptor, Nature 450 (2007) 435–439.

[23] D. Park, A. Hochreiter-Hufford, K.S. Ravichandran, The phosphatidylserine re-ceptor TIM-4 does not mediate direct signaling, Curr. Biol. 19 (2009) 346–351.

[24] R.S. Scott, E.J. McMahon, S.M. Pop, E.A. Reap, R. Caricchio, P.L. Cohen, H.S. Earp,G.K. Matsushima, Phagocytosis and clearance of apoptotic cells is mediated byMER, Nature 411 (2001) 207–211.

[25] M.O. Li, M.R. Sarkisian, W.Z. Mehal, P. Rakic, R.A. Flavell, Phosphatidylserinereceptor is required for clearance of apoptotic cells, Science 302 (2003)1560–1563.

[26] J.I. Elliott, A. Surprenant, F.M.Marelli-Berg, J.C. Cooper, R.L. Cassady-Cain, C.Wooding,K. Linton, D.R. Alexander, C.F. Higgins, Membrane phosphatidylserine distribution asa non-apoptotic signalling mechanism in lymphocytes, Nat. Cell Biol. 7 (2005)808–816.


[27] D.L. Bratton, P.M. Henson, Apoptotic cell recognition:will the real phosphatidylserinereceptor(s) please stand up? Curr. Biol. 18 (2008) 76–79.

[28] A. Yu, C.R. McMaster, D.M. Byers, N.D. Ridgway, H.W. Cook, Stimulation ofphosphatidylserine biosynthesis and facilitation of UV-induced apoptosis in Chinesehamster ovary cells over-expressing phospholipid scramblase 1, J. Biol. Chem. 278(2003) 9706–9714.

[29] K. Saito, M. Nishijima, O. Kuge, Genetic evidence that phosphatidylserine synthaseII catalyzes the conversion of phosphatidylethanolamine to phosphatidylserine inChinese hamster ovary cells, J. Biol. Chem. 273 (1998) 17199–17205.

[30] P.A. Grandmaison, T.S. Nanowski, J.E. Vance, Externalization of phosphatidylserineduring apoptosis does not specifically require either isoform of phosphatidylserinesynthase, Biochim. Biophys. Acta 1636 (2004) 1–11.

[31] M. Valenza, V. Leoni, J.M. Karasinska, L. Petricca, J. Fan, J. Carroll, M.A. Pouladi,E. Fossale, H.P. Nguyen, O. Riess, M. MacDonald, C. Wellington, S. DiDonato,M. Hayden, E. Cattaneo, Cholesterol defect is marked across multiple rodentmodels of Huntington's disease and is manifest in astrocytes, J. Neurosci. 30(2011) 10844–10850.

[32] C. Zelenak, V. Pasham, K. Jilani, P.M. Tripodi, L. Rosaclerio, G. Pathare, A. Lupescu, C.Faggio, S.M. Qadri, F. Lang, Tanshinone IIA stimulates erythrocyte phosphatidylserineexposure, Cell. Physiol. Biochem. 30 (2012) 282–294.

[33] S.C. Frasch, R.F. Fernandez-Boyanapalli, K.Z. Berry, C.C. Leslie, J.V. Bonventre,R.C. Murphy, P.M. Henson, D.L. Bratton, Signaling via macrophage G2A en-hances efferocytosis of dying neutrophils by augmentation of Rac activity,J. Biol. Chem. 286 (2011) 12108–12122.

[34] D.L. Bratton, P.M. Henson, Neutrophil clearance: when the party is over,clean-up begins, Trends Immunol. 32 (2011) 350–357.

[35] H. Hosono, J. Aoki, Y. Nagai, K. Bandoh, M. Ishida, R. Taguchi, H. Arai, K. Inoue,Phosphatidylserine-specific phospholipase A1 stimulates histamine release fromrat peritoneal mast cells through production of 2-acyl-1-lysophosphatidylserine,J. Biol. Chem. 276 (2001) 29664–29670.

[36] J. Aoki, Y. Nagai, H. Hosono, K. Inoue, H. Arai, Structure and function ofphosphatidylserine-specific phospholipase A1, Biochim. Biophys. Acta 1582(2002) 26–32.

[37] D.L. Daleke, Phospholipid flippases, J. Biol. Chem. 282 (2007) 821–825.[38] R.D. Baldridge, T.R. Graham, Identification of residues defining phospholipid

flippase substrate specificity of type IV P-type ATPases, Proc. Natl. Acad. Sci.U. S. A. 109 (2012) E290–E298.

[39] K. Levano, V. Punia, M. Raghunath, P.R. Debata, G.M. Curcio, A. Mogha, S. Purkayastha,D. McCloskey, J. Fata, P. Banerjee, Atp8a1 deficiency is associated withphosphatidylserine externalization in hippocampus and delayed hippocampus-dependent learning, J. Neurochem. 120 (2011) 302–313.

[40] Q. Zhou, J. Zhao, J.G. Stout, R.A. Luhm, T. Wiedmer, P.J. Sims, Molecular cloning ofhuman plasma membrane phospholipid scramblase: a protein mediatingtransbilayer movement of plasma membrane phospholipids, J. Biol. Chem. 272(1997) 18240–18244.

[41] Q. Zhou, J. Zhao, T. Wiedmer, P.J. Sims, Normal hemostasis but defective hema-topoietic response to growth factors in mice deficient in phospholipidscramblase 1, Blood 99 (2002) 4030–4038.

[42] C.T. Sigal, W. Zhou, C.A. Buser, S. McLaughlin, M.D. Resh, Amino-terminal basicresidues of Src mediate membrane binding through electrostatic interactionwith acidic phospholipids, Proc. Natl. Acad. Sci. U. S. A. 91 (1994) 12253–12257.

[43] C.V. Finkielstein, M. Overduin, D.G. Capelluto, Cell migration and signaling spec-ificity is determined by the phosphatidylserine recognition motif of Rac1, J. Biol.Chem. 281 (2006) 27317–27326.

[44] M.A. Lemmon, Membrane recognition by phospholipid-binding domains, Nat.Rev. Mol. Cell Biol. 9 (2008) 99–111.

[45] T. Yeung, B. Heit, J.F. Dubuisson, G.D. Fairn, B. Chiu, R. Inman, A. Kapus, M. Swanson,S. Grinstein, Contribution of phosphatidylserine to membrane surface chargeand protein targeting during phagosome maturation, J. Cell Biol. 185 (2009)917–928.

[46] T. Yeung, G.E. Gilbert, J. Shi, J. Silvius, A. Kapus, S. Grinstein, Membranephosphatidylserine regulates surface charge and protein localization, Science319 (2008) 210–213.

[47] P.A. Leventis, S. Grinstein, The distribution and function of phosphatidylserine incellular membranes, Annu. Rev. Biophys. 39 (2010) 407–427.

[48] G.D. Fairn, N.L. Schieber, N. Ariotti, S.Murphy, L. Kuerschner, R.I.Webb, S. Grinstein,R.G. Parton, High-resolutionmapping reveals topologically distinct cellular pools ofphosphatidylserine, J. Cell Biol. 194 (2011) 257–275.

[49] K.A. Powell, V.A. Valova, C.S. Malladi, O.N. Jensen, M.R. Larsen, P.J. Robinson,Phosphorylation of dynamin I on Ser-795 by protein kinase C blocks its associa-tion with phospholipids, J. Biol. Chem. 275 (2000) 11610–11617.

[50] N. Verdaguer, S. Corbalan-Garcia, W.F. Ochoa, I. Fita, J.C. Gomez-Fernandez,Ca(2+) bridges the C2 membrane-binding domain of protein kinase Calphadirectly to phosphatidylserine, EMBO J. 18 (1999) 6329–6338.

[51] M.A. Swairjo, N.O. Concha, M.A. Kaetzel, J.R. Dedman, B.A. Seaton, Ca(2+)-bridgingmechanism and phospholipid head group recognition in the membrane-bindingprotein annexin V, Nat. Struct. Biol. 2 (1995) 968–974.

[52] Y. Uchida, J. Hasegawa, D. Chinnapen, T. Inoue, S. Okazaki, R. Kato, S. Wakatsuki,R. Misaki, M. Koike, Y. Uchiyama, S. Iemura, T. Natsume, R. Kuwahara, T. Nakagawa,K. Nishikawa, K. Mukai, E. Miyoshi, N. Taniguchi, D. Sheff, W.I. Lencer, T. Taguchi,H. Arai, Intracellular phosphatidylserine is essential for retrograde mem-brane traffic through endosomes, Proc. Natl. Acad. Sci. U. S. A. 108 (2011)15846–15851.

[53] N. Lucas, W. Cho, Phosphatidylserine binding is essential for plasma membranerecruitment and signaling function of 3-phosphoinositide-dependent kinase-1,J. Biol. Chem. 286 (2011) 41265–41272.

[54] B.X. Huang, M. Akbar, K. Kevala, H.Y. Kim, Phosphatidylserine is a critical modu-lator for Akt activation, J. Cell Biol. 192 (2011) 979–992.

[55] N. Arispe, M. Doh, O. Simakova, B. Kurganov, A. De Maio, Hsc70 and Hsp70 inter-act with phosphatidylserine on the surface of PC12 cells resulting in a decreaseof viability, FASEB J. 18 (2004) 1636–1645.

[56] D. Arikketh, R. Nelson, J.E. Vance, Defining the importance of phosphatidylserinesynthase-1 (PSS1): unexpected viability of PSS1-deficient mice, J. Biol. Chem.283 (2008) 12888–12897.

[57] H.G. Hübscher, R.R. Dils, W.F.R. Pover, Studies on the biosynthesis ofphosphatidylserine, Biochim. Biophys. Acta 36 (1959) 518–525.

[58] M.S. Bae-Lee, G.M. Carman, Phosphatidylserine synthesis in Saccharomyces cerevisiae.Purification and characterization of membrane-associated phosphatidylserinesynthase, J. Biol. Chem. 259 (1984) 10857–10862.

[59] A. DeChavigny, P.N. Heacock, W. Dowhan, Sequence and inactivation of the pssgene of Escherichia coli. Phosphatidylethanolamine may not be essential for cellviability, J. Biol. Chem. 266 (1991) 5323–5332.

[60] R.G. Gardner, R.Y. Hampton, A highly conserved signal controls degradation of3-hydroxy-3-methylglutaryl-coenzyme A (HMG-CoA) reductase in eukaryotes,J. Biol. Chem. 274 (1999) 31671–31678.

[61] A. Signorell, J. Jelk, M. Rauch, P. Butikofer, Phosphatidylethanolamine is the precur-sor of the ethanolamine phosphoglycerol moiety bound to eukaryotic elongationfactor 1A, J. Biol. Chem. 283 (2008) 20320–20329.

[62] O. Kuge, M. Nishijima, Y. Akamatsu, Isolation of a somatic cell mutant defective inphosphatidylserine biosynthesis, Proc. Natl. Acad. Sci. U. S. A. 82 (1985) 1926–1930.

[63] D.R. Voelker, J.L. Frazier, Isolation and characterization of a Chinese hamsterovary cell line requiring ethanolamine or phosphatidylserine for growth andexhibiting defective phosphatidylserine synthase activity, J. Biol. Chem. 261(1986) 1002–1008.

[64] G.M. Carman, G.S. Han, Regulation of phospholipid synthesis in the yeastSaccharomyces cerevisiae, Annu. Rev. Biochem. 80 (2011) 859–883.

[65] S.J. Stone, Z. Cui, J.E. Vance, Cloning and expression of mouse liverphosphatidylserine synthase-1 cDNA: overexpression in rat hepatoma cells in-hibits the CDP-ethanolamine pathway for phosphatidylethanolamine biosyn-thesis, J. Biol. Chem. 273 (1998) 7293–7302.

[66] S.J. Stone, J.E. Vance, Cloning and expression of murine liver phosphatidylserinesynthase (PSS)-2: differential regulation of phospholipid metabolism by PSS1and PSS2, Biochem. J. 342 (1999) 57–64.

[67] T.T. Suzuki, J.N. Kanfer, Purification and properties of an ethanolamine-serinebase exchange enzyme of rat brain microsomes, J. Biol. Chem. 260 (1985)1394–1399.

[68] O. Kuge, K. Hasegawa, T. Ohsawa, K. Saito, M. Nishijima, Purification and charac-terization of Chinese hamster phosphatidylserine synthase 2, J. Biol. Chem. 278(2003) 42692–42698.

[69] S. Tomohiro, A. Kawaguti, Y. Kawabe, S. Kitada, O. Kuge, Purification and characteriza-tion of humanphosphatidylserine synthase 1 and 2, Biochem. J. 418 (2008) 421–429.

[70] T. Ohsawa, M. Nishijima, O. Kuge, Functional analysis of Chinese hamsterphosphatidylserine synthase 1 through systematic alanine mutagenesis,Biochem. J. 381 (2004) 853–859.

[71] O. Kuge, K. Saito, M. Nishijima, Cloning of a Chinese hamster ovary (CHO) cDNAencoding phosphatidylserine synthase (PSS) II, overexpression of which sup-presses the phosphatidylserine biosynthetic defect of a PSS I-lacking mutant ofCHO-K1 cells, J. Biol. Chem. 272 (1997) 19133–19139.

[72] L.M.G. van Golde, J. Raben, J.J. Batenburg, B. Fleischer, F. Zambrano, S. Fleischer,Biosynthesis of lipids in Golgi complex and other subcellular fractions from ratliver, Biochim. Biophys. Acta 360 (1974) 179–192.

[73] C.J. Jelsema, D.J. Morré, Distribution of phospholipid biosynthetic enzymesamong cell components of rat liver, J. Biol. Chem. 253 (1978) 7960–7971.

[74] J.E. Vance, D.E. Vance, Does rat liver Golgi have the capacity to synthesize phos-pholipids for lipoprotein secretion? J. Biol. Chem. 263 (1988) 5898–5908.

[75] J.E. Vance, Phospholipid synthesis in a membrane fraction associated with mito-chondria, J. Biol. Chem. 265 (1990) 7248–7256.

[76] K. Saito, O. Kuge, Y. Akamatsu, M. Nishijima, Immunochemical identification ofthe pssA gene product as phosphatidylserine synthase I of Chinese hamsterovary cells, FEBS Lett. 395 (1996) 262–266.

[77] S.J. Stone, J.E. Vance, Phosphatidylserine synthase-1 and ‐2 are localized tomitochondria-associated membranes, J. Biol. Chem. 275 (2000) 34534–34540.

[78] J.N. Kanfer, D. McCartney, H. Hattori, Regulation of the choline, ethanolamineand serine base exchange enzyme activities of rat brain microsomes by phos-phorylation and dephosphorylation, FEBS Lett. 240 (1988) 101–104.

[79] O. Kuge, K. Hasegawa, K. Saito, M. Nishijima, Control of phosphatidylserine biosyn-thesis through phosphatidylserine-mediated inhibition of phosphatidylserinesynthase I in Chinese hamster ovary cells, Proc. Natl. Acad. Sci. U. S. A. 95 (1998)4199–4203.

[80] K. Hasegawa, O. Kuge, M. Nishijima, Y. Akamatsu, Isolation and characterizationof a Chinese hamster ovary cell mutant with altered regulation ofphosphatidylserine biosynthesis, J. Biol. Chem. 264 (1989) 19887–19892.

[81] O. Kuge, K. Saito, M. Nishijima, Control of phosphatidylserine synthase II activityin Chinese hamster ovary cells, J. Biol. Chem. 274 (1999) 23844–23849.

[82] O. Kuge, M. Nishijima, Biosynthetic regulation and intracellular transport ofphosphatidylserine in mammalian cells, J. Biochem. (Tokyo) 133 (2003) 397–403.

[83] B. Sturbois-Balcerzak, S.J. Stone, A. Sreenivas, J.E. Vance, Structure and expres-sion of the murine phosphatidylserine synthase-1 gene, J. Biol. Chem. 276(2001) 8205–8212.

[84] M.O. Bergo, B.J. Gavino, R. Steenbergen, B. Sturbois, A.F. Parlow, D.A. Sanan,W.C. Skarnes, J.E. Vance, S.G. Young, Defining the importance of phosphatidylserinesynthase 2 (Ptdss2) in Mice, J. Biol. Chem. 277 (2002) 47701–47708.


[85] G. Tasseva, L. Cole, J.E. Vance, N-Myc and SP regulate phosphatidylserine synthase-1expression in brain and glial cells, J. Biol. Chem. 286 (2011) 1061–1073.

[86] R. Steenbergen, T.S. Nanowski, R. Nelson, S.G. Young, J.E. Vance, Phospholipidhomeostasis in phosphatidylserine synthase-2-deficient mice, Biochim. Biophys.Acta 1761 (2006) 313–323.

[87] O. Kuge, M. Nishijima, Y. Akamatsu, Phosphatidylserine biosynthesis in culturedChinese hamster ovary cells. III. Genetic evidence for utilization of phosphatidylcho-line and phosphatidylethanolamine as precursors, J. Biol. Chem. 261 (1986)5795–5798.

[88] P.J. Trotter, D.R. Voelker, Identification of a non-mitochondrial phosphatidylserinedecarboxylase activity (PSD2) in the yeast Saccharomyces cerevisiae, J. Biol. Chem.270 (1995) 6062–6070.

[89] P.J. Trotter, J. Pedretti, R. Yates, D.R. Voelker, Phosphatidylserine decarboxylase 2of Saccharomyces cerevisiae. Cloning and mapping of the gene, heterologous ex-pression and creation of the null allele, J. Biol. Chem. 270 (1995) 6071–6080.

[90] A.K. Percy, J.F. Moore, M.A. Carson, C.J. Waechter, Characterization of brainphosphatidylserine decarboxylase: localization in the mitochondrial innermembrane, Arch. Biochem. Biophys. 223 (1983) 484–494.

[91] J. Zborowski, A. Dygas, L. Wojtczak, Phosphatidylserine decarboxylase is located onthe external side of the inner mitochondrial membrane, FEBS Lett. 157 (1983)179–182.

[92] D.R. Voelker, Reconstitution of phosphatidylserine import into rat liver mito-chondria, J. Biol. Chem. 264 (1989) 8019–8025.

[93] D.R. Voelker, Characterization of phosphatidylserine synthesis and translocationin permeabilized animal cells, J. Biol. Chem. 265 (1990) 14340–14346.

[94] J.E. Vance, Newly made phosphatidylserine and phosphatidylethanolamine arepreferentially translocated between rat liver mitochondria and endoplasmic re-ticulum, J. Biol. Chem. 266 (1991) 89–97.

[95] Y.-J. Shiao, G. Lupo, J.E. Vance, Evidence that phosphatidylserine is imported intomitochondria via a mitochondria-associated membrane and that the majority ofmitochondrial phosphatidylethanolamine is derived from decarboxylation ofphosphatidylserine, J. Biol. Chem. 270 (1995) 11190–11198.

[96] Y.-J. Shiao, B. Balcerzak, J.E. Vance, A mitochondrial membrane protein is re-quired for translocation of phosphatidylserine from mitochondria-associatedmembranes to mitochondria, Biochem. J. 331 (1998) 217–223.

[97] G. Achleitner, D. Zweytick, P.J. Trotter, D.R. Voelker, G. Daum, Synthesis and in-tracellular transport of aminoglycerophospholipids in permeabilized cells ofthe yeast, Saccharomyces cerevisiae, J. Biol. Chem. 270 (1995) 29836–29842.

[98] W.I.Wu, D.R. Voelker, Characterization of phosphatidylserine transport to the locusof phosphatidylserine decarboxylase 2 in permeabilized yeast, J. Biol. Chem. 276(2001) 7114–7121.

[99] R. Simbeni, L. Pon, E. Zinser, F. Paltauf, G. Daum, Mitochondrial membrane con-tact sites of yeast. Characterization of lipid components and possible involve-ment in intramitochondrial translocation of phospholipids, J. Biol. Chem. 266(1991) 10047–10049.

[100] G. Achleitner, B. Gaigg, A. Krasser, E. Kainersdorfer, S.D. Kohlwein, A. Perktold,G. Zellnig, G. Daum, Association between the endoplasmic reticulum andmito-chondria of yeast facilitates interorganelle transport of phospholipids throughmembrane contact, Eur. J. Biochem. 264 (1999) 545–553.

[101] D.R. Voelker, Disruption of phosphatidylserine translocation to the mitochon-dria in baby hamster kidney cells, J. Biol. Chem. 260 (1985) 14671–14676.

[102] D.R. Voelker, Phosphatidylserine translocation to the mitochondrion is anATP-dependent process in permeabilized animal cells, Proc. Natl. Acad. Sci.U. S. A. 86 (1989) 9921–9925.

[103] D.R. Voelker, The ATP-dependent translocation of phosphatidylserine to the mi-tochondria is a process that is restricted to the autologous organelle, J. Biol.Chem. 268 (1993) 7069–7074.

[104] D. Ardail, F. Lerme, P. Louisot, Involvement of contact sites in phosphatidylserineimport into liver mitochondria, J. Biol. Chem. 266 (1991) 7978–7981.

[105] D.R. Voelker, Adriamycin disrupts phosphatidylserine import into themitochondria of permeabilized CHO-K1 cells, J. Biol. Chem. 266 (1991)12185–12188.

[106] A.E. Rusiñol, Z. Cui, M.H. Chen, J.E. Vance, A unique mitochondria-associatedmembrane fraction from rat liver has a high capacity for lipid synthesis and con-tains pre-Golgi secretory proteins including nascent lipoproteins, J. Biol. Chem.269 (1994) 27494–27502.

[107] Z. Cui, J.E. Vance, M.H. Chen, D.R. Voelker, D.E. Vance, Cloning and expression of anovel phosphatidylethanolamine N-methyltransferase, J. Biol. Chem. 268 (1993)16655–16663.

[108] R. Rizzuto, P. Pinton, W. Carrington, F.S. Fay, K.E. Fogarty, L.M. Lifshitz, R.A. Tuft,T. Pozzan, Close contacts with the endoplasmic reticulum as determinants of mi-tochondrial Ca2+ responses, Science 280 (1998) 1763–1765.

[109] C.J. Loewen, T.P. Levine, A highly conserved binding site in vesicle-associatedmem-brane protein-associated protein (VAP) for the FFAT motif of lipid-binding pro-teins, J. Biol. Chem. 280 (2005) 14097–14104.

[110] T. Simmen, J.E. Aslan, A.D. Blagoveshchenskaya, L. Thomas, L. Wan, Y. Xiang, S.F.Feliciangeli, C.H. Hung, C.M. Crump, G. Thomas, PACS-2 controls endoplasmicreticulum-mitochondria communication and Bid-mediated apoptosis, EMBO J.24 (2005) 717–729.

[111] M.M. Schumacher, J.Y. Choi, D.R. Voelker, Phosphatidylserine transport to themitochondria is regulated by ubiquitination, J. Biol. Chem. 277 (2002)51033–51042.

[112] W.W. Franke, J. Kartenbeck, Outer mitochondrial membrane continuous withendoplasmic reticulum, Protoplasma 73 (1971) 35–41.

[113] G.C. Shore, J.R. Tata, Two fractions of rough endoplasmic reticulum from ratliver, J. Cell Biol. 72 (1977) 714–725.

[114] G. Szabadkai, K. Bianchi, P. Varnai, D. De Stefani, M.R. Wieckowski, D. Cavagna,A.I. Nagy, T. Balla, R. Rizzuto, Chaperone-mediated coupling of endoplasmicreticulum and mitochondrial Ca2+ channels, J. Cell Biol. 175 (2006) 901–911.

[115] R. Rizzuto, T. Pozzan, Microdomains of intracellular Ca2+: molecular determi-nants and functional consequences, Physiol. Rev. 86 (2006) 369–408.

[116] J.G. Goetz, I.R. Nabi, Interaction of the smooth endoplasmic reticulum andmitochondria, Biochem. Soc. Trans. 34 (2006) 370–373.

[117] O.M. deBrito, L. Scorrano,Mitofusin 2 tethers endoplasmic reticulum tomitochondria,Nature 456 (2008) 605–610.

[118] G. Perkins, C. Renken, M.E. Martone, S.J. Young, M. Ellisman, T. Frey, Electron to-mography of neuronal mitochondria: three-dimensional structure and organi-zation of cristae and membrane contacts, J. Struct. Biol. 119 (1997) 260–272.

[119] G. Csordas, C. Renken, P. Varnai, L. Walter, D. Weaver, K.F. Buttle, T. Balla, C.A.Mannella, G. Hajnoczky, Structural and functional features and significance ofthe physical linkage between ER and mitochondria, J. Cell Biol. 174 (2006)915–921.

[120] D.W. Hailey, A.S. Rambold, P. Satpute-Krishnan, K. Mitra, R. Sougrat, P.K. Kim,J. Lippincott-Schwartz, Mitochondria supply membranes for autophagosomebiogenesis during starvation, Cell 141 (2010) 656–667.

[121] D. Sebastian, M.I. Hernandez-Alvarez, J. Segales, E. Sorianello, J.P. Munoz, D. Sala,A. Waget, M. Liesa, J.C. Paz, P. Gopalacharyulu, M. Oresic, S. Pich, R. Burcelin,M. Palacin, A. Zorzano, Mitofusin 2 (Mfn2) links mitochondrial andendoplasmic reticulum function with insulin signaling and is essential fornormal glucose homeostasis, Proc. Natl. Acad. Sci. U. S. A. 109 (2012)5523–5528.

[122] G. Sutendra, P. Dromparis, P. Wright, S. Bonnet, A. Haromy, Z. Hao, M.S. McMurtry,M. Michalak, J.E. Vance, W.C. Sessa, E.D. Michelakis, The role of Nogo and themitochondria-endoplasmic reticulum unit in pulmonary hypertension, Sci. Transl.Med. 3 (2011) 88ra55.

[123] J.Y. Choi, W.R. Riekhof, W.I. Wu, D.R. Voelker, Macromolecular assemblies regu-late nonvesicular phosphatidylserine traffic in yeast, Biochem. Soc. Trans. 34(2006) 404–408.

[124] D.R. Voelker, Genetic and biochemical analysis of non-vesicular lipid traffic,Annu. Rev. Biochem. 78 (2009) 827–856.

[125] B. Kornmann, E. Currie, S.R. Collins, M. Schuldiner, J. Nunnari, J.S. Weissman,P. Walter, An ER-mitochondria tethering complex revealed by a syntheticbiology screen, Science 325 (2009) 477–481.

[126] W.I. Wu, D.R. Voelker, Reconstitution of phosphatidylserine transport fromchemically defined donor membranes to phosphatidylserine decarboxylase 2implicates specific lipid domains in the process, J. Biol. Chem. 279 (2004)6635–6642.

[127] L. Filippin, P.J. Magalhaes, G. Di Benedetto, M. Colella, T. Pozzan, Stable interac-tions between mitochondria and endoplasmic reticulum allow rapid accumula-tion of calcium in a subpopulation of mitochondria, J. Biol. Chem. 278 (2003)39224–39234.

[128] S. Leo, K. Bianchi, M. Brini, R. Rizzuto, Mitochondrial calcium signalling in celldeath, FEBS J. 272 (2005) 4013–4022.

[129] V.M. Shkryl, N. Shirokova, Transfer and tunneling of Ca2+ from sarcoplasmicreticulum to mitochondria in skeletal muscle, J. Biol. Chem. 281 (2006)1547–1554.

[130] N. Myhill, E.M. Lynes, J.A. Nanji, A.D. Blagoveshchenskaya, H. Fei, K. CarmineSimmen, T.J. Cooper, G. Thomas, T. Simmen, The subcellular distribution ofcalnexin is mediated by PACS-2, Mol. Biol. Cell 19 (2008) 2777–2788.

[131] J.R. Friedman, L.L. Lackner, M. West, J.R. Dibenedetto, J. Nunnari, G.K. Voeltz, ERtubules mark sites of mitochondrial division, Science 21 (2011) 358–362.

[132] J.A. Post, J.J. Bijvelt, A.J. Verkleij, Phosphatidylethanolamine and sarcolemmal dam-age during ischemia or metabolic inhibition of heart myocytes, Am. J. Physiol. 268(1995) H773–H780.

[133] C.J. Walkey, L.R. Donohue, R. Bronson, L.B. Agellon, D.E. Vance, Disruption of themurine gene encoding phosphatidylethanolamine N-methyltransferase, Proc.Natl. Acad. Sci. U. S. A. 94 (1997) 12880–12885.

[134] Z. Li, L.B. Agellon, T.M. Allen, M. Umeda, L. Jewell, A. Mason, D.E. Vance, The ratioof phosphatidylcholine to phosphatidylethanolamine influences membrane in-tegrity and steatohepatitis, Cell Metab. 3 (2006) 321–331.

[135] X.H. Jin, Y. Okamoto, J. Morishita, K. Tsuboi, T. Tonai, N. Ueda, Discovery and charac-terization of a Ca2+‐independent phosphatidylethanolamine N-acyltransferasegenerating the anandamide precursor and its congeners, J. Biol. Chem. 282 (2007)3614–3623.

[136] W.A. Devane, L. Hanus, A. Breuer, R.G. Pertwee, L.A. Stevenson, G. Griffin, D.Gibson, A. Mandelbaum, A. Etinger, R. Mechoulam, Isolation and structure of abrain constituent that binds to the cannabinoid receptor, Science 258 (1992)1946–1949.

[137] A.S. Joshi, M.N. Thompson, N. Fei, M. Huttemann, M.L. Greenberg, Cardiolipinand mitochondrial phosphatidylethanolamine have overlapping functions inmitochondrial fusion in Saccharomyces cerevisiae, J. Biol. Chem. 287 (2012)17589–17597.

[138] A. Signorell, E. Gluenz, J. Rettig, A. Schneider, M.K. Shaw, K. Gull, P. Butikofer,Perturbation of phosphatidylethanolamine synthesis affects mitochondrial mor-phology and cell-cycle progression in procyclic-form Trypanosoma brucei, Mol.Microbiol. 72 (2009) 1068–1079.

[139] R. Steenbergen, T.S. Nanowski, A. Beigneux, A. Kulinski, S.G. Young, J.E. Vance,Disruption of the phosphatidylserine decarboxylase gene in mice causesembryonic lethality and mitochondrial defects, J. Biol. Chem. 280 (2005)40032–40040.

[140] P.R. Cullis, B. De Kruijff, Polymorphic phase behaviour of lipid mixtures asdetected by 31P NMR. Evidence that cholesterol may destabilize bilayer


structure in membrane systems containing phosphatidylethanolamine,Biochim. Biophys. Acta 507 (1978) 207–218.

[141] P.R. Cullis, B. de Kruijff, M. Hope, A.J. Verkleij, R. Nayar, S.B. Farren, C. Tildock,T.D. Madden, M.B. Bally, Structural properties of lipids and their functionalrole in biological membranes, in: R.C. Aloia (Ed.), Membrane Fluidity in Biology,vol. 1, Academic Press, New York, 1983, pp. 39–81.

[142] A.J. Verkleij, J. Leunissen-Bijvelt, B. de Kruijff, M. Hope, P.R. Cullis, Non-bilayerstructures in membrane fusion, Ciba Found. Symp. 103 (1984) 45–59.

[143] S. Martens, H.T. McMahon, Mechanisms of membrane fusion: disparate playersand common principles, Nat. Rev. Mol. Cell Biol. 9 (2008) 543–556.

[144] K. Emoto, T. Kobayashi, A. Yamaji, H. Aizawa, I. Yahara, K. Inoue, M. Umeda,Redistribution of phosphatidylethanolamine at the cleavage furrow of dividingcells during cytokinesis, Proc. Natl. Acad. Sci. U. S. A. 93 (1996) 12867–12872.

[145] K. Emoto, M. Umeda, An essential role for a membrane lipid in cytokinesis: reg-ulation of contractile ring disassembly by redistribution of phosphatidylethanol-amine, J. Cell Biol. 149 (2000) 1215–1224.

[146] E.I. Pecheur, I. Martin, O. Maier, U. Bakowsky, J.M. Ruysschaert, D. Hoekstra,Phospholipid species act as modulators in p97/p47-mediated fusion of Golgimembranes, Biochemistry 41 (2002) 9813–9823.

[147] M. Bogdanov, J. Sun, H.R. Kaback,W. Dowhan, A phospholipid acts as a chaperone inassembly of a membrane transport protein, J. Biol. Chem. 271 (1996) 11615–11618.

[148] M. Bogdanov, M. Umeda, W. Dowhan, Phospholipid-assisted refolding of an integralmembrane protein. Minimum structural features for phosphatidylethanolamine toact as a molecular chaperone, J. Biol. Chem. 274 (1999) 12339–12345.

[149] W. Dowhan, M. Bogdanov, Molecular genetic and biochemical approaches fordefining lipid-dependent membrane protein folding, Biochim. Biophys. Acta1818 (2012) 1097–1107.

[150] N.R. Deleault, J.R. Piro, D.J. Walsh, F. Wang, J. Ma, J.C. Geoghegan, S. Supattapone,Isolation of phosphatidylethanolamine as a solitary cofactor for prion formationin the absence of nucleic acids, Proc. Natl. Acad. Sci. U. S. A. 109 (2012) 8546–8551.

[151] I.Y. Dobrosotskaya, A.C. Seegmiller, M.S. Brown, J.L. Goldstein, R.B. Rawson,Regulation of SREBP processing and membrane lipid production by phospho-lipids in Drosophila, Science 296 (2002) 879–883.

[152] M.S. Brown, J. Ye, R.B. Rawson, J.L. Goldstein, Regulated intramembraneproteolysis:a control mechanism conserved from bacteria to humans, Cell 100 (2000)391–398.

[153] A.K. Menon, V.L. Stevens, Phosphatidylethanolamine is the donor of the ethanol-amine residue linking a glycosylphosphatidylinositol anchor to protein, J. Biol.Chem. 267 (1992) 15277–15280.

[154] R. Nebauer, S. Rosenberger, G. Daum, Phosphatidylethanolamine, a limiting fac-tor of autophagy in yeast strains bearing a defect in the carboxypeptidase Ypathway of vacuolar targeting, J. Biol. Chem. 282 (2007) 16736–16743.

[155] E.P. Kennedy, S.B. Weiss, The function of cytidine coenzymes in the biosynthesisof phospholipides, J. Biol. Chem. 222 (1956) 193–214.

[156] A.L. Henneberry, M.M. Wright, C.R. McMaster, The major sites of cellular phos-pholipid synthesis and molecular determinants of Fatty Acid and lipid headgroup specificity, Mol. Biol. Cell 13 (2002) 3148–3161.

[157] L.F. Borkenhagen, E.P. Kennedy, L. Fielding, Enzymatic formation and decarbox-ylation of phosphatidylserine, J. Biol. Chem. 236 (1961) 28–32.

[158] W.R. Riekhof, D.R. Voelker, Uptake andutilization of lyso-phosphatidylethanolamineby Saccharomyces cerevisiae, J. Biol. Chem. 281 (2006) 36588–36596.

[159] W.R. Riekhof, J. Wu, M.A. Gijon, S. Zarini, R.C. Murphy, D.R. Voelker,Lysophosphatidylcholine metabolism in Saccharomyces cerevisiae: the role ofP-type ATPases in transport and a broad specificity acyltransferase in acylation,J. Biol. Chem. 282 (2007) 36853–36861.

[160] W.R. Riekhof, J. Wu, J.L. Jones, D.R. Voelker, Identification and characterization ofthe major lysophosphatidylethanolamine acyltransferase in Saccharomycescerevisiae, J. Biol. Chem. 282 (2007) 28344–28352.

[161] R. Sundler, B. Akesson, A. Nilsson, Quantitative role of base exchange in phosphatidyl-ethanolamine synthesis in isolated rat hepatocytes, FEBS Lett. 43 (1974) 303–307.

[162] K.S. Bjerve, Phospholipid substrate-specificity of the L-serine base exchange en-zyme in rat liver microsomal fraction, Biochem. J. 219 (1984) 781–784.

[163] R. Leonardi, M.W. Frank, P.D. Jackson, C.O. Rock, S. Jackowski, Elimination of theCDP-ethanolamine pathway disrupts hepatic lipid homeostasis, J. Biol. Chem.284 (2009) 27077–27089.

[164] R. Birner, M. Burgermeister, R. Schneiter, G. Daum, Roles of phosphatidyletha-nolamine and of its several biosynthetic pathways in Saccharomyces cerevisiae,Mol. Biol. Cell 12 (2001) 997–1007.

[165] M.K. Storey, K.L. Clay, T. Kutateladze, R.C. Murphy, M. Overduin, D.R. Voelker,Phosphatidylethanolamine has an essential role in Saccharomyces cerevisiaethat is independent of its ability to form hexagonal phase structures, J. Biol.Chem. 276 (2001) 48539–48548.

[166] M. Burgermeister, R. Birner-Grunberger, R. Nebauer, G. Daum, Contribution ofdifferent pathways to the supply of phosphatidylethanolamine and phosphati-dylcholine to mitochondrial membranes of the yeast Saccharomyces cerevisiae,Biochim. Biophys. Acta 1686 (2004) 161–168.

[167] O.B. Bleijerveld, J.F. Brouwers, A.B. Vaandrager, J.B. Helms, M. Houweling, TheCDP-ethanolamine pathway and phosphatidylserine decarboxylation generate dif-ferent phosphatidylethanolamine molecular species, J. Biol. Chem. 282 (2007)28362–28372.

[168] R. Sundler, B. Akesson, Biosynthesis of phosphatidylethanolamines and phos-phatidylcholines from ethanolamine and choline in rat liver, Biochem. J. 146(1975) 309–315.

[169] L.B.M. Tijburg, M.J.H. Geelen, L.M.G. Van Golde, Biosynthesis of phosphatidyletha-nolamine via the CDP-ethanolamine route is an important pathway in isolated rathepatocytes, Biochem. Biophys. Res. Commun. 160 (1989) 1275–1280.

[170] L.B.M. Tijburg, M. Houweling, M.J.H. Geelen, L.M.G. van Golde, Effects of dietaryconditions on the pool sizes of precursors of phosphatidylcholine and phosphati-dylethanolamine synthesis in rat liver, Biochim. Biophys. Acta 959 (1988) 1–8.

[171] T.A. Zelinski, P.C. Choy, Phosphatidylethanolamine biosynthesis in isolated ham-ster heart, Can. J. Biochem. 60 (1982) 817–823.

[172] A. Signorell, M. Rauch, J. Jelk, M.A. Ferguson, P. Butikofer, Phosphatidylethanolaminein Trypanosoma brucei is organized in two separate pools and is synthesized exclu-sively by the Kennedy pathway, J. Biol. Chem. 283 (2008) 23636–23644.

[173] D.R. Voelker, Phosphatidylserine functions as themajor precursor of phosphatidyleth-anolamine in culturedBHK-21 cells, Proc.Natl. Acad. Sci. U. S. A. 81 (1984)2669–2673.

[174] M.A. Miller, C. Kent, Characterization of the pathways for phosphatidylethanol-amine biosynthesis in Chinese hamster ovary mutant and parental cell lines,J. Biol. Chem. 261 (1986) 9753–9761.

[175] K.S. Bjerve, The biosynthesis of phosphatidylserine and phosphatidylethanol-amine from L-[3-14C]serine in isolated rat hepatocytes, Biochim. Biophys. Acta833 (1985) 396–405.

[176] Y. Tian, P. Jackson, C. Gunter, J. Wang, C.O. Rock, S. Jackowski, Placental thrombo-sis and spontaneous fetal death in mice deficient in ethanolamine kinase 2,J. Biol. Chem. 281 (2006) 28438–28449.

[177] A. Lykidis, J. Wang, M.A. Karim, S. Jackowski, Overexpression of a mammalianethanolamine-specific kinase accelerates the CDP-ethanolamine pathway,J. Biol. Chem. 276 (2001) 2174–2179.

[178] J.D. Saba, F. Nara, A. Bielawska, S. Garrett, Y.A. Hannun, The BST1 gene ofSaccharomyces cerevisiae is the sphingosine-1-phosphate lyase, J. Biol. Chem.272 (1997) 26087–26090.

[179] Y.A. Hannun, C. Luberto, K.M. Argraves, Enzymes of sphingolipid metabolism:from modular to integrative signaling, Biochemistry 40 (2001) 4893–4903.

[180] D. Rontein, I. Nishida, G. Tashiro, K. Yoshioka, W.I. Wu, D.R. Voelker, G. Basset,A.D. Hanson, Plants synthesize ethanolamine by direct decarboxylation ofserine using a pyridoxal phosphate enzyme, J. Biol. Chem. 276 (2001)35523–35529.

[181] R. Sundler, Ethanolaminephosphate cytidylyltransferase. Purification and char-acterization of the enzyme from rat liver, J. Biol. Chem. 250 (1975) 8585–8590.

[182] L.B.M. Tijburg, M. Houweling, M.J.H. Geelen, L.M.G. van Golde, Stimulation ofphosphatidylethanolamine synthesis in isolated rat hepatocytes by phorbol12-myristate 13-aceate, Biochim. Biophys. Acta 922 (1987) 184–190.

[183] A. Nakashima, K. Hosaka, J.-I. Nikawa, Cloning of a human cDNA for CTP-phosphoethanolamine cytidylyltransferase by complementation in vivo of a yeastmutant, J. Biol. Chem. 272 (1997) 9567–9572.

[184] A. Poloumienko, A. Cote, A.-T. Quee, L. Zhu, M. Bakovic, Genomic organizationand differential splicing of the mouse and human Pcyt2 genes, Gene 325(2004) 145–155.

[185] M.D. Fullerton, F. Hakimuddin, M. Bakovic, Developmental and metabolic effectsof disruption of the mouse CTP:phosphoethanolamine cytidylyltransferase gene(Pcyt2), Mol. Cell. Biol. 27 (2007) 3327–3336.

[186] M.D. Fullerton, F. Hakimuddin, A. Bonen,M. Bakovic, Thedevelopmentof ametabolicdisease phenotype in CTP:phosphoethanolamine cytidylyltransferase-deficientmice, J. Biol. Chem. 284 (2009) 25704–25713.

[187] A.L. Henneberry, C.R. McMaster, Cloning and expression of a humancholine/ethanolaminephosphotransferase: synthesis of phosphatidylcholineand phosphatidylethanolamine, Biochem. J. 339 (1999) 291–298.

[188] C.R. McMaster, R.M. Bell, CDP-ethanolamine:1,2-diacylglycerol ethanolamine-phosphotransferase, Biochim. Biophys. Acta 1348 (1997) 117–123.

[189] Y. Horibata, Y. Hirabayashi, Identification and characterization of humanethanolaminephosphotransferase1, J. Lipid Res. 48 (2007) 503–508.

[190] N. Gupta, A. Hartmann, R. Lucius, D.R. Voelker, The obligate intracellular parasiteToxoplasma gondii secretes a soluble phosphatidylserine decarboxylase, J. Biol.Chem. 287 (2012) 22938–22947.

[191] J.Y. Choi, Y. Augagneur, C. Ben Mamoun, D.R. Voelker, Identification of geneencoding Plasmodium knowlesi phosphatidylserine decarboxylase by geneticcomplementation in yeast and characterization of in vitro maturation ofencoded enzyme, J. Biol. Chem. 287 (2011) 222–232.

[192] E.E. Snell, Pyruvate-containing enzymes, Trends Biochem. Sci. 2 (1977) 131–135.[193] W. Dowhan, W.T. Wickner, E.P. Kennedy, Purification and properties of

phosphatidylserine decarboxylase from Escherichia coli, J. Biol. Chem. 249 (1974)3079–3084.

[194] Q.-X. Li,W. Dowhan, Structural characterization of Escherichia coli phosphatidylserinedecarboxylase, J. Biol. Chem. 263 (1988) 11516–11522.

[195] O. Kuge, K. Saito, M. Kojima, Y. Akamatsu, M. Nishijima, Post-translational pro-cessing of the phosphatidylserine decarboxylase gene product in Chinese ham-ster ovary cells, Biochem. J. 319 (1996) 33–38.

[196] O. Kuge, M. Nishijima, Y. Akamatsu, A cloned gene encoding phosphatidylserinedecarboxylase complements the phosphatidylserine biosynthetic defect of aChinese hamster ovary cell mutant, J. Biol. Chem. 266 (1991) 6370–6376.

[197] C.J. Clancey, S.-C. Chang, W. Dowhan, Cloning of a gene (PSD1) encodingphosphatidylserine decarboxylase from Saccharomyces cerevisiae bycomplementation of an Escherichia coli mutant, J. Biol. Chem. 268 (1993)24580–24590.

[198] D.R. Voelker, Phosphatidylserine decarboxylase, Biochim. Biophys. Acta 1348(1997) 236–244.

[199] I. Schuiki, M. Schnabl, T. Czabany, C. Hrastnik, G. Daum, Phosphatidylethanolaminesynthesized by four different pathways is supplied to the plasmamembrane of theyeast Saccharomyces cerevisiae, Biochim. Biophys. Acta 1801 (2009) 480–486.

[200] P.J. Trotter, J. Pedretti, D.R. Voelker, Phosphatidylserine decarboxylase fromSaccharomyces cerevisiae: isolation of mutants, cloning of the gene, and creationof a null allele, J. Biol. Chem. 268 (1993) 21416–21424.


[201] V.M. Gohil, M.N. Thompson, M.L. Greenberg, Synthetic lethal interaction of themitochondrial phosphatidylethanolamine and cardiolipin biosynthetic path-ways in Saccharomyces cerevisiae, J. Biol. Chem. 280 (2005) 35410–35416.

[202] C. Osman, M. Haag, C. Potting, J. Rodenfels, P.V. Dip, F.T. Wieland, B. Brugger,B. Westermann, T. Langer, The genetic interactome of prohibitins: coordinat-ed control of cardiolipin and phosphatidylethanolamine by conserved regu-lators in mitochondria, J. Cell Biol. 184 (2009) 583–596.

[203] K. Pfeiffer, V. Gohil, R.A. Stuart, C. Hunte, U. Brandt, M.L. Greenberg, H. Schagger,Cardiolipin stabilizes respiratory chain supercomplexes, J. Biol. Chem. 278(2003) 52873–52880.

[204] H. Schagger, T. Hagen, B. Roth, U. Brandt, T.A. Link, G. von Jagow, Phospho-lipid specificity of bovine heart bc1 complex, Eur. J. Biochem. 190 (1990)123–130.

[205] M. Hayer-Hartl, H. Schagger, G. von Jagow, K. Beyer, Interactions of phospho-lipids with the mitochondrial cytochrome-c reductase studied by spin-labelESR and NMR spectroscopy, Eur. J. Biochem. 209 (1992) 423–430.

[206] K. Shinzawa-Itoh, H. Aoyama, K. Muramoto, H. Terada, T. Kurauchi, Y. Tadehara, A.Yamasaki, T. Sugimura, S. Kurono, K. Tsujimoto, T. Mizushima, E. Yamashita, T.Tsukihara, S. Yoshikawa, Structures and physiological roles of 13 integral lipids of bo-vine heart cytochrome c oxidase, EMBO J. 26 (2007) 1713–1725.

[207] M.S. Sharpley, R.J. Shannon, F. Draghi, J. Hirst, Interactions between phospho-lipids and NADH:ubiquinone oxidoreductase (complex I) from bovine mito-chondria, Biochemistry 45 (2006) 241–248.

[208] T.M. McIntyre, F. Snyder, G.K. Marathe, Ether-linked lipids and theirbioactive species, in: D.E.V.a.J.E. Vance (Ed.), Biochemistry of Lipids,Lipoproteins and Membranes, 5th edn., Elsevier, Amsterdam, 2008,pp. 245–276.

[209] A. Diagne, J. Fauvel, M. Record, H. Chap, L. Douste-Blazy, Studies on ether phos-pholipids. II. Comparative composition of various tissues from human, rat andguinea pig, Biochim. Biophys. Acta 793 (1984) 221–231.

[210] P.C. Choy, K. Tran, G.M. Hatch, E.A. Kroeger, Phospholipid metabolism in themammalian heart, Prog. Lipid Res. 36 (1997) 85–101.

Review Formation and function of phosphatidylserine and phosphatidylethanolamine in mammalian cells

Documents

Transcript of Review Formation and function of phosphatidylserine and phosphatidylethanolamine in mammalian cells