- Email: [email protected]
Phospholipase A2 enzymes: structural diversity in lipid messenger metabolism
st8211.qxd
02/29/2000
12:18
Page R15
Minireview
R15
Phospholipase A2 enzymes: structural diversity in lipid messenger metabolism A Dessen The phospholipases A2 (PLA2s) are a large family of enzymes with varied lipidic products which are involved in numerous signal transduction pathways. The structural and functional characterization of several PLA2s have revealed the various mechanisms used by these enzymes to ingeniously manipulate the phospholipidic metabolic machinery. Address: European Molecular Biology Laboratory, 6 rue Jules Horowitz, 38000 Grenoble, France. E-mail: [email protected] Structure 2000, 8:R15–R22 0969-2126/00/$ – see front matter © 2000 Elsevier Science Ltd. All rights reserved.
Introduction
The importance of lipid messengers is reflected in their involvement in fundamental, highly regulated cellular processes, such as growth, adhesion, secretion and apoptosis, and medical concerns ranging from thrombosis to asthma [1]. Of particular interest are the eicosanoids, 20-carbon fatty acids that include the cyclooxygenasederived leukotrienes and the lipooxygenase-generated prostaglandins, both powerful mediators of inflammation [2]. It is therefore of no surprise that much attention has been given to the common element of all lipid messenger pathways: the phospholipases that are essential to their biosynthesis. The characterization of the molecular details of phospholipase biochemistry has been challenging for several reasons. Phospholipases that are involved in signaling pathways are typically present in low amounts in cells, often displaying low solubility, and their hydrophobic substrates are preferred in aggregated form. These facts alone are enough to complicate assays and their analyses, and yet phospholipases have been known since the 1950’s when their activity was first discovered in pancreatic juice and cobra venom [3,4]. Since then, elegant efforts in cloning, expression, purification, sequencing, and structure determination, as well as studies of regulation and localization have helped to uncover details of many phospholipase groups [5]. Within the phospholipases, by far the most complex and well-known element is the phospholipase A2 superfamily. Phospholipase A2 enzymes (PLA2s) catalyze the hydrolysis of the sn-2 bond of phospholipids, generating free fatty acids and lysophospholipids, and are thus key enzymes which control the release of lipid mediator precursors. The
diversity within the enzyme family can be classified according to localization of the individual member (secreted versus cytosolic), sequence homology, cation requirement and molecular weight, generating at least nine major groups with varied molecular characteristics [5]. But what is the need for such diversity? The answer lies in the fact that mammalian tissues and cells, which often contain more that one PLA2, possess differently regulated pathways for fatty acid metabolism, requiring PLA2s to play crucial roles in cellular responses as diverse as phospholipid digestion and turnover, membrane remodeling, host defense, signal transduction and the production of inflammatory mediators [4]. The divergence displayed by PLA2s in terms of structure, cation requirement and catalytic mechanism is remarkable when considering that they all target phospholipidic substrates and catalyze an acyl hydrolysis reaction. This review will concentrate on the structural aspects of this important family of enzymes, with particular emphasis on the recent crystal structure of cytosolic PLA2. The new and unexpected fold of cystolic PLA2 reveals novel eccentricities employed by such enzymes in the manipulation of fatty acid metabolism. Structure and mechanism of secreted PLA2 enzymes
Secreted PLA2s (sPLA2s) are the most well-studied members of the PLA2 family and possibly of all lipidmetabolizing enzymes. Primarily on the basis of amino acid sequence and disulfide positioning, sPLA2s have been classified into three major types (I, II and III) and two different subtypes (A and B). They have been isolated from various snake and bee venoms, as well as from several mammalian sources [4]. The functions performed by sPLA2s range from the digestion of dietary fat, to cell proliferation and migration. In particular, the abundance of group IIA sPLA2 in various inflamed tissues, including the plasma of patients with sceptic shock and rheumatoid arthritis, suggests that it may have a crucial role in the pathogenesis of inflammation. Venom sPLA2s, on the other hand, seem to be directly involved in prey digestion; in addition, they have been reported to have roles in neurotoxic, myotoxic and anticoagulatory processes [6]. The sPLA2s are cysteine-rich molecules that catalyze the cleavage of the 2-acyl ester bond of phospholipids without any fatty acid selectivity and in a Ca2+-dependent manner. Disulfide bridges ranging in number from five to eight [5] guarantee a rigid three-dimensional structure, as well as stability and proteolytic resistance, features that ensure sPLA2s their full activity in the extracellular fluids where they are found [7]. X-ray crystal structures of sPLA2s from different sources [8–14] have revealed a common fold with
st8211.qxd
02/29/2000
R16
12:18
Page R16
Structure 2000, Vol 8 No 2
Figure 1 A gallery of lipase folds. (a) sPLA2. The compact 14 kDa molecule harbors an elapid loop (blue) and a tightly bound Ca2+ ion (pink) located near the active-site His—Asp dyad. (b) cPLA2. The C2 domain is in purple, with two tightly bound Ca2+ atoms in red. The catalytic domain is composed of a central β sheet (yellow) and a cap region (pink), which harbors a highly flexible lid that covers the active-site dyad. Notice the lack of interdomain contacts between the C2 and catalytic units. (c) Phospholipase Cδ1 has a vast area of contact between its C2 domain (purple) and catalytic domain (yellow). The EF-hand domain (red) is behind the C2 domain. (d) Cutinase. The smallest α/β hydrolase contains a solvent-exposed active site and thus does not undergo interfacial activation. The catalytic serine is depicted on the nucleophilic elbow.
three major α-helical segments, a double-stranded antiparallel β sheet (the β wing), and a highly evolutionarily conserved Ca2+-binding loop; diversions from commonality include a small, surface-exposed segment observed only in type I sPLA2s (the ‘elapid’ or ‘pancreatic’ loop; blue in Figure 1a) and a six amino acid C-terminal extension which characterizes type II sPLA2s. The active site of sPLA2s is located at the bottom of a channel lined with hydrophobic residues and is composed of a highly conserved His–Asp catalytic dyad and a tightly bound, heptacoordinated Ca2+ atom (Figure 2a). Crystal structures of sPLA2s in complex with inhibitors or transition-state analogs have shown the alkane chains of such molecules to lie roughly parallel in the channel, extending between 9 and 14 Å from the catalytic histidine towards the opening of the channel near the N-terminal helix [10–14]. In the absence of a catalytic triad, as observed in serine proteases, sPLA2s are incapable of hydrolyzing the 2-acyl group of diacylphospholipids through an acyl–enzyme intermediate; instead, the catalytic histidine, aided by a neighboring aspartate residue, polarizes a nearby water molecule, which then attacks the substrate sn-2 bond. The Ca2+ atom actively participates in catalysis by stabilizing the resulting transition state, and reassumes a state of heptacoordination as two water molecules move into the active site upon collapse of the tetrahedral intermediate and product release [15].
Based initially on observations that not all catalytically active venom sPLA2s are toxic and several catalytically inactive sPLA2s show myotoxic characteristics, it has now been shown that particular sPLA2s possess high-affinity cellular receptors (N-type and M-type) the recognition of which elicits toxic responses. Initially identified in brain tissue, N-type receptors are selectively recognized by neurotoxic sPLA2s. Although their molecular structure and physiological roles are still unclear, it is known that recognition occurs mostly through elements of the sPLA2 hydrophobic channel and Ca2+-binding loop [16]. Conversely, M-type receptors, which recognize different nonneurotoxic sPLA2s, have been well characterized as type I transmembrane molecules with a cysteine-rich N-terminal domain, a fibronectin type II unit, and eight repeats of a carbohydrate-recognition domain (CRD) followed by a single transmembrane region and a short cytoplasmic tail [17]. Site-directed mutagenesis and systematic CRD deletion studies of mammalian M-type receptors indicate that neither the cysteine-rich nor the fibronectin-like type II domains are involved in sPLA2 recognition, but it is CRD5 that provides the majority of binding contacts [18], although there are no carbohydrate chains on the sPLA2s. Structural elements of the sPLA2s involved in M-type receptor interactions lie within or close to the Ca2+-binding loop; however, the presence of Ca2+, which
st8211.qxd
02/29/2000
12:18
Page R17
Minireview Phospholipase A2 enzymes Dessen
is essential for enzymatic activity, does not seem to be required for receptor binding [19]. The identification of N- and M-type receptors provided the first definitive evidence that sPLA2s possess functions that are independent of their enzymatic activity [17].
R17
Figure 2
The cytosolic enzymes: Ca2+-independent PLA2s
In most tissues, the synthesis of eicosanoidal inflammatory mediators such as leukotrienes and prostaglandins is limited by the availability of their common precursor, free arachidonic acid, which must be hydrolyzed from phospholipidic membranes [20]. Leukotrienes have a role in the pathophysiology of asthma by inducing airflow obstruction through bronchoconstriction, increased secretion of mucus, and migration of inflammatory cells into air passages of asthmatic patients. Prostaglandins are mediators of both asthmatic and arthritic processes [21]. Hence, the control of the level of free arachidonic acid in cells must require a highly regulated mechanism. Although Ca2+-independent PLA2s (iPLA2s) do not seem to actively participate in the stimulation of arachidonic acid release from membranes due to their general lack of specificity for arachidonate-containing substrates, they establish their mark in arachidonic acid metabolism by playing an important role in phospholipid fatty acid remodeling [22]. The deacylation/reacylation cycle of membrane phospholipids, known as the Lands cycle, involves the cleavage of phospholipids by an intracellular phospholipase to generate a 2-lysophospholipid, which can subsequently be reacylated with a different fatty acid. The key element of this cycle is the PLA2 which generates the 2-lysophospholipid used in the acylation reaction. That is where the iPLA2s step in — by providing the lysophospholipid acceptor, they participate in the regulation of incorporation of arachidonic acid and other unsaturated fatty acids into membrane phospholipid [6,22]. To date, few intracellular iPLA2s (type VI) have been purified to homogeneity, and such enzymes have proven recalcitrant to detailed biochemical analysis because of their lability, low specific activities, and low abundance in cells. Hence, a detailed structural and mechanistic description will have to await further studies. However, efforts in cloning and purification have identified iPLA2s with diverse biochemical characteristics, ranging in molecular weight from 38 kDa to 85 kDa and being capable of forming 250–400 kDa oligomeric complexes [6]. Notably, the iPLA2s from Chinese hamster ovary (CHO) cells and P388D1 macrophage-like cells contain a lipase motif (Gly-X-Ser-X-Gly) displaying a putative active-site serine, as well as eight ankyrin repeats that have not been identified in any other PLA2; the repeats were suggested to function in the ordered association of the subunits to give the oligomeric complex [23,24].
Distinct active sites for acyl hydrolysis. Catalytic residues are highlighted in green. (a) The sPLA2 active site consists of a His–Asp pair and a heptacoordinated Ca2+ atom the function of which is to stabilize the reaction transition state. Other residues that also contact inhibitor molecules in co-complex structures are shown. (b) The active site of a neutral lipase (PDB code 1ETH) contains a classic His–Asp–Ser triad reminiscent of that of serine proteases. The oxyanion hole region is shown in yellow, as in (c). Other residues relevant for inhibitor molecule contact are also displayed. (c) The active site of cPLA2 lacks histidine and therefore differs from that of a classic lipase. Ser228 is located on the nucleophilic elbow. The sidechain of Arg200 may stabilize the phosphoryl moiety of a bound substrate.
st8211.qxd
02/29/2000
R18
12:18
Page R18
Structure 2000, Vol 8 No 2
Little is known about the catalytic mechanism of iPLA2s, except for the obvious lack of Ca2+ requirement and their selective inhibition by bromoenol lactone (BEL), a mechanism-based molecule originally proposed as a serine protease inhibitor [25]. The presence of a serine residue in a classic lipase motif together with the results of studies with BEL point to the possible employment of a serine–acyl intermediate in iPLA2 catalysis.
protein kinase C, GTPase-activating protein, synaptotagmin and phospholipase Cδ1, has been shown to bind phospholipid membranes in a Ca2+-dependent manner. A clear ubiquitous role for C2 domains in bilayer targeting is also supported by the observation that various components of cellular membranes, such as phospholipids, inositol phosphates and other membrane-associated proteins, have been identified as potential ligands [39].
Cytosolic phospholipase A2: membrane association and activation by Ca2+
C2 domains consist of an eight-stranded antiparallel β sandwich which in many cases coordinates between one and three Ca2+ atoms through a constellation of acidic residues on three Ca2+-binding loops (CBLs or ‘jaws’) [40–43]. Membrane binding is accomplished mostly by residues located on and around the apex of the CBL region. In the C2 domain of cPLA2, these amino acids have been implicated in interface recognition through the employment of methods ranging from the detection of N15/NH heteronuclear single-quantum coherence (HSQC) shifts upon incubation of C2 with dodecylphosphocholine micelles [43] to cysteine engineering followed by fluorescence probing [39]. It is conceivable that when phospholipids bind to the C2 domain they fill unsatisfied coordination sites on the bound Ca2+ ions [44], which explains why cPLA2 binding to phosphatidylcholine vesicles is enhanced tenfold in the presence of Ca2+ [45].
Of all members of the vast PLA2 family, it is the cytosolic PLA2 (cPLA2 or type IV PLA2) which has the most important role in lipid messenger biosynthesis. Although three isozymes of this group have been identified so far (α, β and γ) [26–28], it is cPLA2-α which is the best characterized and the most relevant for intracellular signaling (and will thus be referred to henceforth simply as cPLA2). cPLA2 is highly selective for phospholipidic substrates with arachidonic acid at the sn-2 position [26,29] and thus has a crucial role in eicosanoid biosynthesis. Interestingly, the cyclooxygenases (COX-1 and COX-2) which catalyze the metabolism of arachidonic acid to prostaglandins are localized in the nuclear membrane, while the leukotriene biosynthesizing enzyme 5-lipoxygenase translocates from the cytosol to the perinuclear membrane in response to an increase in intracellular Ca2+ [30], as does cPLA2 [31,32]. This chain of events is verified by the observation that activation or increased expression of cPLA2 is linked to increased arachidonate release as well as leukotriene and prostaglandin synthesis [33,34]. The most remarkable evidence regarding the strategic importance of cPLA2 in eicosanoid biosynthesis, however, comes from analyses of cPLA2 knockout mice, whose peritoneal macrophages are incapable of generating leukotrienes, prostaglandins or PAF [35,36]. The relevance of the role of these three messengers in the pathophysiology of different inflammatory processes, coupled to the high specificity of cPLA2 for arachidonate phospholipids, makes cPLA2 an attractive target for drug discovery — its specific inhibition would simultaneously block the biosynthetic pathways of all three molecules. cPLA2 is an 85 kDa molecule composed of an N-terminal C2 domain (138 amino acids; purple in Figure 1b) and a catalytic unit with 611 residues (yellow and pink Figure 1b [37]). Maximal activation of cPLA2 in intact cells requires mitogen-activated protein (MAP) kinase directed phosphorylation [34,38]; this event on its own, however, is insufficient for such activation if it occurs in the absence of an increase in intracellular Ca2+ concentration, a requirement for C2-mediated translocation of cPLA2 to the target membrane [26,31,32,37]. The C2 domain of cPLA2, like its structural counterparts observed in signal-transduction and vesicle-fusion proteins such as
Thus, for cPLA2, upon a submicromolar increase in intracellular Ca2+ concentration, the C2 domain recruits the Ca2+-independent catalytic domain to the nuclear or endoplasmic reticulum (ER) membranes, presenting it to its substrate [26,37,39]. The function of the C2 domain in cPLA2 biochemistry is thus clear: it must tether the fulllength molecule to the membrane, but must do so in a way as to present the catalytic domain to its substrate in a productive fashion. Hence, the relative interdomain orientation is of key importance for cPLA2 activity. This is also the case for inositol-specific phospholipase Cδ1 (PCLδ1; Figure 1c), which catalyzes the hydrolysis of membranebound phosphatidyl inositol-4-5-bisphosphate into two second messengers, namely D-myo-inositol-1,4,5-triphosphate and sn-1,2-diacylglycerol, in response to diverse extracellular signals. PLCδ1 is a tetradomain molecule; in addition to the C2 and catalytic units, it contains an EF hand and a pleckstrin homology (PH) domain. Based on crystallographic evidence, the mode of PLCδ1 binding to the membrane involves the initial tethering of the molecule through the PH domain, whereas the function of C2 is to fix the catalytic unit, and hence the active site, in close apposition to the substrate interface [40,41]. A comparison of the surface diagrams of full-length cPLA2 [46] and PLCδ1 [40,41] through a superposition of their C2 domains allows visualization of the critical function of C2 in orienting the catalytic domains with respect to the target membrane. In Figure 3, the CBL region of both C2
st8211.qxd
02/29/2000
12:18
Page R19
Minireview Phospholipase A2 enzymes Dessen
R19
domains (superimposed) points towards and interacts with the interface. It is clear that the C2 domains of cPLA2 and PCLδ1 place the active sites in approximately the same orientation. Interestingly, the vast hydrophobic contact surface between the C2 and catalytic domains of PLCδ1 does not allow further interdomain rotation, suggesting that this orientation is optimal for membrane interaction. Similarly, the relative orientation of the cPLA2 domains allows C2 to fulfill its function of tethering cPLA2 to the membrane, juxtaposing the active-site region and the substrate interface. In addition, due to the lack of major contacts between the cPLA2 C2 and catalytic units (see Figure 1b), an additional small interdomain rotation could be possible. The orientation of the cPLA2 domains observed in the crystal structure is biochemically productive, and is clearly an essential element of the enzyme’s cellular function.
Figure 3
The catalytic domain of cPLA2: an original fold for a new enzymatic mechanism
Superposition of surface diagrams of cPLA2 (yellow) and phospholipase Cδ1 (white) through their C2 domains. Residues in C2 that have been shown to bind lipids in N15/NH HSQC experiments (concentrated around the CBL region) [43] are depicted in dark blue. Active sites of both molecules (red) are similarly positioned and strategically oriented to interact with the target membrane. In cPLA2, the basic patch region is also available for bilayer contact. A flexible lid (pink) covers the activesite funnel and must move to allow substrate access.
The structure determination of multiple lipases have revealed that they all share a common topological feature, namely the α/β hydrolase fold, which consists primarily of a central, mostly parallel β sheet interwoven by α helices [47,48]. This α/β structure encompasses a single domain which can range from 197 residues (Fusarium solani cutinase; Figure 1d) to 582 amino acids (mouse acetylcholinesterase), indicating that variations within the fold are generated by the addition of secondary structure units without the inclusion of new domains [49]. The central β sheet has a left-handed superhelical twist, forcing the first and last strands to cross at an approximate angle of 90° to one another. Although the positions of the β strands are conserved, the twist of the β sheet varies among the different enzymes, as do the orientations of the α helices [48]. In addition to the common fold, lipases possess a serineacid-histidine catalytic triad reminiscent of the serine proteases (Figure 2b) and proceed catalytically through a serine–acyl intermediate. Within the twisted β sheet, the nucleophilic serine is located on the ‘nucleophilic elbow’, a tight turn between a central α helix and its preceding β strand [47,48]. The commonality of the α/β hydrolase fold, including identical positioning of the catalytic residues (both in terms of sequence and three-dimensional localization) [48] reveal a framework that seems to have diverged from a common ancestor, generating enzymes which differ in their substrate-binding sites but preserve the arrangement of the catalytic triad [47]. Prior to structure solution, cPLA2 showed signs that it too could be a member of the α/β hydrolases. Its mechanism proceeds through a serine–acyl intermediate [50,51]; the nucleophile, identified as Ser228 by site-specific mutagenesis, is located in the Gly-Leu-Ser-Gly-Ser sequence,
which is highly reminiscent of the classic Gly-X-Ser-X-Gly lipase motif [52]. In addition to Ser228, Asp549 and Arg200 were also shown to be essential for cPLA2 activity, but none of the 19 histidine residues were shown to have any catalytic role [53]. This fact suggested that the catalytic domain of cPLA2 might hold structural surprises. The X-ray crystal structure of cPLA2 [46] indeed reveals a catalytic domain harboring a twisted central β sheet (Figure 1b), with Ser228 located on a ‘nucleophilic elbow’. Beyond these two elements, however, there are no further similarities with α/β hydrolases. After the formation of the initial part of the β sheet, the amino acid chain is detracted from the core to form a highly basic, 180-residue region which positions itself between the active site (partly covering it) and the C2 domain — the ‘cap’ (residues 370–548; pink in Figure 1b). The cap is composed of two sets of antiparallel β strands and four α helices, a fold which possesses no similarity to any known structure. Although the specific function of the cPLA2 cap is unknown, its large number of solvent-exposed, basic amino acids (the ‘basic patch’) as well as its strategic location, facing the proposed membrane-binding region, are of note (see Figure 3). Because cPLA2 has affinity for anionic phospholipids [45,54], it is tempting to propose that this charged region in cPLA2 has a role in contacting such acidic substrates [46]. Interestingly, other membrane-targeting enzymes that also reveal a striking preference for negatively charged substrates, such as type IIb phosphatidylinositol phosphate kinase [55], several sPLA2s [56], protein kinase Cδ
st8211.qxd
02/29/2000
R20
12:18
Page R20
Structure 2000, Vol 8 No 2
[57] and PLCδ1 itself [40], also possess basic surface regions. A basic patch observed in the trimeric matrix protein of both human immunodeficiency virus 1 and Simian immunodeficiency virus 1 has also been implicated in stabilization of the molecule on the internal side of the acidic viral membrane [58,59]; these observations suggest a ubiquitous interface-association role for such charged regions on membrane-binding proteins. In addition to contacting the membrane interface, the multifunctional cap region of cPLA2 also harbors a flexible lid which, in the crystal structure, covers the buried active site (pink in Figure 3). Analysis of this active site reveals yet another difference between cPLA2 and enzymes of the α/β hydrolase family: the lack of any histidines in a 10 Å radius around the nucleophilic Ser228 (Figure 2c). Since mutagenesis studies failed to identify any histidine residue that could act in a classic catalytic triad as seen in serine proteases [46], this is not completely surprising. It is of note, however, that the active site of cPLA2 lacks any sidechains that could replace the histidine residue identified in all α/β hydrolases to date [53]. Even Arg200, the relevance of which in catalysis was highlighted by the lack of activity in a Arg200→Lys mutant [53], positions its guanidinium group 9 Å away from the Oγ atom of Ser228, and is thus not a candidate. These observations led to the proposal of a catalytic mechanism for cPLA2 which obviates the participation of a histidine residue [46]. In this model, Arg200 has a role in stabilizing the phosphoryl moiety of an incoming phospholipidic substrate. Ser228, activated by Asp549, attacks the sn-2 ester of a bound substrate molecule. The oxyanion hole, formed by the backbone atoms of Gly197 and Gly198 (located on the yellow turn in Figure 2c), polarizes the ester and stabilizes the transition state of the developing tetrahedral intermediate. Hydrolysis of the formed serine–acyl intermediate is then promoted by a bound water molecule, leaving cPLA2 free to bind another phospholipid substrate and ready to repeat the cycle. Conformational change and interfacial activation
By far the most fascinating phenomenon associated with lipid metabolism and lipase structure is that of ‘interfacial activation’, which describes the dramatic increase in lipase activity when substrate levels exceed the critical micellar concentration (CMC) [60]. Although distinct in structure and catalytic mechanism, sPLA2s, cPLA2, and various lipases of the α/β hydrolase family all undergo interfacial activation. Studies into the molecular basis of interfacial activation of lipases have resulted in the formulation of two main theories. The ‘substrate hypothesis’ suggests that factors such as concentration gradient and pressure at the membrane surface have important roles in the generation of this phenomenon. Conversely, the ‘enzyme hypothesis’ suggests that lipases undergo localized conformational changes upon membrane interaction. It is now
apparent that both the nature of the substrate as well as conformational changes in the protein may contribute to enzymatic activation at the interface [61]. X-ray crystallographic studies of secreted PLA2s, both in apo form and complexed to inhibitors, suggest rigid enzymes in which no conformational changes can occur upon interaction with the lipid–water interface. These studies, in conjunction with other biochemical evidence, suggest that catalysis at the interface occurs as a result of facilitated substrate diffusion from the binding surface through a hydrophobic channel that leads to the sPLA2 active site [15]. Recently, electron paramagnetic resonance (EPR) based membrane-docking methods have shown that the surface region of sPLA2s which surrounds the opening of the hydrophobic channel in fact constitutes the interfacial binding surface (i-face) [62]. The i-face participates in juxtaposition between enzyme and membrane (obviating membrane penetration) so that phospholipid transfer may occur without solvation of the hydrophobic alkyl groups [15,56] and in the absence of conformational changes. Notably, elegant charge-reversal mutagenesis studies have identified a patch of basic residues located on the i-face that has an important role in anionic phospholipid recognition and interfacial binding [63]. The explanation for the activation phenomenon of neutral lipases, however, has had a different history. Crystal structures of several lipases of the α/β hydrolase family have revealed active sites occluded by surface loops (or ‘lids’), which must undergo a conformational change in order for catalytic residues to be exposed to the substrate. Notably, lid movement also exposes vast hydrophobic areas which surround the site itself, and are thus likely to be stabilized by a lipidic interface. In several cases, this conformational change is necessary to strategically localize the oxyanion hole within the active-site region. The crucial role played in interfacial activation by structural elements of the lipases themselves is also clear in the crystal structure of cutinase, a small lipase which contains a solvent-exposed active site and thus does not undergo interfacial activation [64] (Figure 1d). cPLA2 also undergoes the phenomenon of interfacial activation by displaying a 1500-fold higher specific activity in the presence of micellar substrate [37]. Again, a structural explanation for this effect can be elicited. The crystal structure of human cPLA2 [46] shows that an amphipathic lid covers the active site of the enzyme; short flexible regions are ideally positioned to serve as hinges for the moveable lid (see Figure 3). An open cPLA2 lid would reveal a 30 Å wide hydrophobic platform funneling down into a deep active site, with members of the catalytic dyad located at the bottom of a 7 Å funnel stalk. The surface of the funnel should provide an optimal binding cradle for fatty acyl elements of the target interface, certainly a stabilizing factor of
st8211.qxd
02/29/2000
12:18
Page R21
Minireview Phospholipase A2 enzymes Dessen
major importance for cPLA2 activation. The location of the Ser228–Asp549 dyad partway into the body of the catalytic domain of cPLA2 also suggests that, in order for the substrate to come into contact with the active site, it is necessary for cPLA2 to become imbedded in the membrane itself. This hypothesis is supported by the existence of a 7 Å hydrophobic rim protruding from the base of the active-site funnel, which could be involved in interface recognition, as well by the observation that cPLA2 activity is inhibited in the presence of sphingomyelin-rich membranes [54]. It is of interest that PLCδ1, which also displays a hydrophobic rim around the active site and the interfacerecognition mechanism of which has been hypothesized to involve partial insertion of the catalytic domain into the membrane [40], is also inhibited by sphingomyelin [65]. Sphingomyelin has been reported to decrease membrane hydration through an increase in interlipid hydrogen bonding, causing the interface to become more rigid; this could inhibit protein penetration, and in the case of cPLA2, hinder access of substrate into the deep active site [46]. Conclusions
PLA2 enzymes owe the great interest they have generated over the past decades to their key activities in lipid mediator biosynthesis. Eicosanoids, powerful inflammatory signaling molecules, metabolically depend on the release of arachidonic acid from phospholipidic membranes by PLA2s, ensuring that these enzymes are under tight regulatory control and thus making them fascinating targets for biochemical studies. Due to their low abundance in cells and preference for substrates in aggregated forms, however, PLA2s have been difficult to characterize; in addition, the high molecular weight PLA2s have provided crystallographic challenges. The relationship between the structural diversity of the PLA2 family and the cellular processes in which their diverse lipid messenger products are involved is finally coming into perspective. Low molecular weight sPLA2s target phospholipidic substrates without fatty acid selectivity and have biological functions that reflect either their catalytic activities or interactions with specific receptors; their compact three-dimensional structures reveal a framework that must be sealed onto its bilayersequestered substrate, allowing it to diffuse into a deep hydrophobic catalytic channel. On the other hand, cPLA2 is the most specific enzyme targeting arachidonyl phospholipids, and does so by employing its C2 domain in an initial tethering step which positions the catalytic unit appropriately for partial membrane embedding, lid opening and active-site access by the substrate. The essential nature of PLA2s for the generation of lipidic messengers holds the key to the most exciting use for the knowledge gained in their structural and functional analysis: the design of novel, specific metabolic inhibitors for the control of a wide range of cellular processes.
R21
Acknowledgements The author wishes to thank former colleagues at Wyeth Ayerst Biochemistry for many helpful discussions.
References 1. Prescott, S. (1997). A thematic series on phospholipases. J. Biol. Chem. 272, 15043. 2. Goetzl E.J., An, S. & Smith, W.L. (1995). Specificity of expression and effects of eicosanoid mediators in normal physiology and human diseases. FASEB J. 9, 1051-1058. 3. Wittcoff, H. (1951). The Phosphatides. pp. 99-115, Reinhold Publishing Corp., New York. 4. Dennis, E.A. (1994). Diversity of group types, regulation, and function of phospholipase A2. J. Biol. Chem. 269, 13057-13060. 5. Dennis, E.A. (1997). The growing phospholipase A2 superfamily of signal transduction enzymes. Trends Biochem. Sci. 22, 1-2. 6. Murakami, M., Nakatani, Y., Atsumi, G., Inoue, K. & Kudo, I. (1997). Regulatory functions of phospholipase A2. Crit. Rev. Immunol. 17, 225-283. 7. Balsinde, J., Balboa, M.A., Insel, P.A. & Dennis, E.A. (1999). Regulation and inhibition of phospholipase A2. Annu. Rev. Pharmacol. Toxicol. 39, 175-189. 8. Fremont, D.H., Anderson, D.H., Wilson, I.A., Dennis, E.A. & Xuong, N.-H. (1993). Crystal structure of phospholipase A2 from Indian cobra reveals a trimeric association. Proc. Natl Acad. Sci. USA 90, 342-346. 9. Wery, J.-P., et al., & N.D. Jones (1991). Structure of recombinant human rheumatoid arthritic synovial fluid phospholipase A2 at 2.2 Å resolution. Nature 352, 79-82. 10. Scott, D.L., Otwinowski, Z., Gelb, M.H. & Sigler, P.B. (1990). Crystal structure of bee-venom phospholipase A2 in a complex with a transition-state analogue. Science 250, 1563-1566. 11. Scott, D.L., White, S.P., Browning, J.L., Rosa, J.J., Gelb, M.H. & Sigler, P.B. (1991). Structures of free and inhibited human secretory phospholipase A2 from inflammatory exudate. Science 254, 1007-1009. 12. Thunnissen, M.M., et al., & Verheij, H.M. (1990). X-ray structure of phospholipase A2 complexed with a substrate-derived inhibitor. Nature 347, 689-691. 13. Sekar, K., Eswaramoorthy, S., Jain, M.K. & Sundaralingam, M. (1997). Crystal structure of the complex of bovine pancreatic phospholipase A2 with the inhibitor 1-hexadecyl-3-(trifluoroethyl)-sn-glycero-2phosphomethanol. Biochemistry 36, 14186-14191. 14. White, S.P., Scott, D.L., Otwinowski, Z., Gelb, M.H. & Sigler, P.B. (1990). Crystal structure of cobre-venom phospholipase A2 in a complex with a transition-state analogue. Science 250, 1560-1562. 15. Scott, D.L., White, S.P., Otwinowski, Z., Yuan, W., Gelb, M.H. & Sigler, P.B. (1990). Interfacial catalysis: the mechanism of phospholipase A2. Science 250, 1541-1546. 16. Nicolas, J.-P., Ying, L., Lambeau, G., Ghomashchi, F., Lazdunski, M. & Gelb, M.H. (1997). Localization of structural elements of bee venom phospholipase A2 involved in N-type receptor binding and neurotoxicity. J. Biol. Chem. 272, 7173-7181. 17. Lambeau, G. & Lazdunski, M. (1999). Receptors for a growing family of secreted phospholipases A2. Trends Pharm. Sci. 20, 162-170. 18. Nicolas, J.-P., Lambeau, G. & Lazdunski, M. (1995). Identification of the binding domain for secretory phospholipases A2 on their M-type 180-kDa membrane receptor. J. Biol. Chem. 270, 28869-28873. 19. Lambeau, G., et al., & Lazdunski, M. (1995). Structural elements of secretory phospholipases A2 involved in the binding to M-type receptors. J. Biol. Chem. 270, 5534-5540. 20. Irvine, R.F. (1982). How is the level of free arachidonic acid controlled in mammalian cells? Biochem. J. 204, 3-16. 21. O'Byrne, P.M. (1997). Leukotrienes in the pathogenesis of asthma. Chest 111, 275-345. 22. Balsinde, J. & Dennis, E.A. (1997). Function and inhibition of intracellular calcium-independent phospholipase A2. J. Biol. Chem. 272, 16069-16072. 23. Tang, J., Kriz, R.W., Wolfman, N., Shaffer, M., Seehra, J. & Jones, S.S. (1997). A novel cytosolic calcium-independent phospholipase A2 contains eight ankyrin motifs. J. Biol. Chem. 272, 8567-8575. 24. Balboa, M.A., Balsinde, J., Jones, S.S. & Dennis, E.A. (1997). Identity between the Ca2+-independent phospholipase A2 enzymes from P388D1 macrophages and chinese hamster ovary cells. J. Biol. Chem. 272, 8676-8580. 25. Ackermann, E.J., Conde-Frieboes, K. & Dennis, E.A. (1995). Inhibition of macrophage Ca2+-independent phospholipase A2 by bromoenol lactone and trifluoromethyl ketones. J. Biol. Chem. 270, 445-450.
st8211.qxd
02/29/2000
R22
12:18
Page R22
Structure 2000, Vol 8 No 2
26. Clark, J.D., et al., & Knopf, J.L. (1991). A novel arachidonic acidselective PLA2 contains a Ca2+-dependent translocation domain with homology to PKC and GAP. Cell 65, 1043-1051. 27. Song, C., Chang, X.-J., Bean, K.M., Proia, M.S., Knopf, J.L. & Kriz, R.W. (1999). Molecular characterization of cytosolic phospholipase A2-β. J. Biol. Chem. 274, 17063-17067. 28. Underwood, K.W., Song, C., Kriz, R.W., Chang, X.J., Knopf, J.L. & Lin, L.-L. (1998). A novel calcium-independent phospholipase A2, cPLA2-γ, that is prenylated and contains homology to cPLA2. J. Biol. Chem. 273, 21926-21932. 29. Hanel, A.M. & Gelb, M.H. (1993). Processive interfacial catalysis by mammalian 85-kilodalton phospholipase A2 enzymes on productcontaining vesicles: application to the determination of substrate preferences. Biochemistry 32, 5949-5958. 30. Peters-Golden, L. & McNish, R.W. (1993). Redistribution of 5lipoxygenase and cytosolic phospholipase A2 to the nuclear fraction upon macrophage activation. Biochem. Biophys. Res. Commun. 196, 147-153. 31. Glover, S., et al., & Gelb, M.H. (1995). Translocation of the 85-kDa phospholipase A2 from cytosol to the nuclear envelope in rat basophilic leukemia cells stimulated with calcium ionophore or IgE/antigen. J. Biol. Chem. 270, 15359-15367. 32. Schievella, A.R., Regier, M.K., Smith, W.L. & Lin, L.-L. (1995). Calcium-mediated translocation of cytosolic phospholipase A2 to the nuclear envelope and endoplasmic reticulum. J. Biol. Chem. 270, 30749-30754. 33. Lin, L.-L., Lin, A.Y. & Knopf, J.L. (1992). Cytosolic phospholipase A2 is coupled to hormonally regulated release of arachidonic acid. Proc. Natl Acad. Sci. USA 89, 6147-6151. 34. Lin, L.-L., Wartmann, M., Lin, A.Y., Knopf, J.L., Seth, A. & Davis, R.J. (1993). cPLA2 is phosphorylated and activated by MAP kinase. Cell 72, 269-278. 35. Uozumi, N., et al., & Shimizu, T. (1997). Role of cytosolic phospholipase A2 in allergic response and parturition. Nature 390, 618-622. 36. Bonventre, J.V., et al., & Sapirstein, A. (1997). Reduced fertility and postischaemic brain injury in mice deficient in cytosolic phospholipase A2. Nature 390, 622-625. 37. Nalefski, E.A., et al., & Clark, J.D. (1994). Delineation of two functionally distinct domains of cytosolic phospholipase A2, a regulatory Ca2+-dependent lipid-binding domain and a Ca2+independent catalytic domain. J. Biol. Chem. 269, 18239-18249. 38. Qiu, Z.-H., Gijon, M.A., de Carvalho, M.S., Spencer, D.M. & Leslie, C.C. (1998). The role of calcium and phosphorylation of cytosolic phospholipase A2 in regulating arachidonic acid release in macrophages. J. Biol. Chem. 273, 8203-8211. 39. Nalefski, E.A. & Falke, J.J. (1998). Location of the membrane-docking face on the Ca2+-activated C2 domain of cytosolic phospholipase A2. Biochemistry 37, 17642-17650. 40. Essen, L.-O., Perisic, O., Cheung, R., Katan, M. & Williams, R.L. (1996). Crystal structure of a mammalian phosphoinositide-specific phospholipase Cδ. Nature 380, 595-602. 41. Grobler, J.A., Essen, L.-O., Williams, R.L. & Hurley, J.H. (1996). C2 domain conformational changes in phospholipase Cδ1. Nat. Struct. Biol. 3, 788-795. 42. Perisic, O., Fong, S., Lynch, D.E., Bycroft, M. & Williams, R.L. (1998). Crystal structure of a calcium-phospholipid binding domain from cytosolic phospholipase A2. J. Biol. Chem. 273, 1596-1604. 43. Xu, G.-Y., McDonagh, T., Yu, H.-A., Nalefski, E., Clark, J.D. & Cumming, D.A. (1998). Solution structure and membrane interactions of the C2 domain of cytosolic phospholipase A2. J. Mol. Biol. 280, 485-500. 44. Rizo, J. & Südhof, T.C. (1998). C2-domains, structure and function of a universal Ca2+-binding domain. J. Biol. Chem. 273, 15879-15882. 45. Hixon, M.S., Ball, A. & Gelb, M.H. (1998). Calcium-dependent and independent interfacial binding and catalysis of cytosolic group IV phospholipase. Biochemistry 37, 8516-8526. 46. Dessen, A., et al., & Somers, W.S. (1999). Crystal structure of human cytosolic phospholipase A2 reveals a novel topology and catalytic mechanism. Cell 97, 349-360. 47. Ollis, D.L., et al., & Goldman, A. (1992). The α/β hydrolase fold. Protein Eng. 5, 197-211. 48. Schrag, J.D. & Cygler, M. (1997). Lipases and the α/β hydrolase fold. Methods Enzymol. 284, 85-107. 49. Heikinheimo, P., Goldman, A., Jeffries, C. & Ollis, D.L. (1999). Of barn owls and bankers: a lush variety of α/β hydrolases. Structure 7, R141-R146.
50. Trimble, L.A., Street, I.P., Perrier, H., Tremblay, N.M., Weech, P.K. & Bernstein, M.A. (1993). NMR structural studies of the tight complex between a trifluoromethyl ketone inhibitor and the 85-kDa human phospholipase A2. Biochemistry 32, 12560-12565. 51. Hanel, A.M. & Gelb, M.H. (1995). Multiple enzymatic activities of the human cytosolic 85-kDa phospholipase A2: hydrolytic reactions and acyl transfer to glycerol. Biochemistry 34, 7807-7818. 52. Sharp, J.D., et al., & Kramer, R. (1994). Serine 228 is essential for catalytic activities of 85-kDa cytosolic phospholipase A2. J. Biol. Chem. 269, 23250-23254. 53. Pickard, R.T., et al., & Sharp, J.D. (1996). Identification of essential residues for the catalytic function of 85-kDa cytosolic phospholipase A2. J. Biol. Chem. 271, 19225-19231. 54. Leslie, C.C. & Channon, J.Y. (1990). Anionic phospholipids stimulate an arachidonoyl-hydrolyzing phospholipase A2 from macrophages and reduce the calcium requirement for activity. Biochim. Biophys. Acta 1045, 261-270. 55. Rao, V.D., Misra, S., Boronekov, I.V., Anderson, R.A. & Hurley, J.H. (1998). Structure of type IIb phosphatidylinositol phosphate kinase: a protein kinase fold flattened for interfacial phosphorylation. Cell 94, 829-839. 56. Gelb, M.H., Cho, W. & Wilton, D.C. (1999). Interfacial binding of secreted phospholipases A2: more than electrostatics and a major role for tryptophan. Curr. Opin. Struct. Biol. 9, 428-432. 57. Zhang, G., Kazanietz, M.G., Blumberg, P.M. & Hurley, J.H. (1995). Crystal structure of the Cys2 activator-binding domain of protein kinase Cδ in complex with phorbol ester. Cell 81, 917-924. 58. Rao, Z., Belyaev, A.S., Fry, E., Roy, P., Jones, I.M. & Stuart, D.I. (1995). Crystal structure of SIV matrix antigen and implications for virus assembly. Nature 378, 743-747. 59. Hill, C.P., Worthylake, D., Bancroft, D.P., Christensen, A.M. & Sundquist, W. (1996). Crystal structures of the trimeric human immunodeficiency virus type 1 matrix protein: implications for membrane association and assembly. Proc. Natl Acad. Sci. USA 93, 3099-3104. 60. Sarda, L. & Desnuelle, P. (1958). Action de la lipase pancreatique sur les esters en emulsion. Biochim. Biophys. Acta 30, 513-520. 61. Derewenda, Z.S. (1995). A twist in the tale of lipolytic enzymes. Nat. Struct. Biol. 5, 347-349. 62. Lin, Y., et al., & Gelb, M.H. (1998). Docking of phospholipase A2 on membranes using electrostatic potential-modulated spin relaxation magnetic resonance. Science 279, 1925-1929. 63. Han, S.K., Yoon, E.T., Scott, D.L., Sigler, P.B. & Cho, W. (1997). Structural aspects of interfacial absorption. J. Biol. Chem. 272, 3573-3582. 64. Longhi, S., Czjzek, M., Lamzin, V., Nicolas, A. & Cambillau, C. (1997). Atomic resolution (1.0 Å) crystal structure of Fusarium solani cutinase: stereochemical analysis. J. Mol. Biol. 268, 779-799. 65. Scarlata, S., et al., & Rebecchi, M.J. (1996). Inhibition of phospholipase Cδ1 catalytic activity by sphingomyelin. Biochemistry 35, 14882-14888.
02/29/2000
12:18
Page R15
Minireview
R15
Phospholipase A2 enzymes: structural diversity in lipid messenger metabolism A Dessen The phospholipases A2 (PLA2s) are a large family of enzymes with varied lipidic products which are involved in numerous signal transduction pathways. The structural and functional characterization of several PLA2s have revealed the various mechanisms used by these enzymes to ingeniously manipulate the phospholipidic metabolic machinery. Address: European Molecular Biology Laboratory, 6 rue Jules Horowitz, 38000 Grenoble, France. E-mail: [email protected] Structure 2000, 8:R15–R22 0969-2126/00/$ – see front matter © 2000 Elsevier Science Ltd. All rights reserved.
Introduction
The importance of lipid messengers is reflected in their involvement in fundamental, highly regulated cellular processes, such as growth, adhesion, secretion and apoptosis, and medical concerns ranging from thrombosis to asthma [1]. Of particular interest are the eicosanoids, 20-carbon fatty acids that include the cyclooxygenasederived leukotrienes and the lipooxygenase-generated prostaglandins, both powerful mediators of inflammation [2]. It is therefore of no surprise that much attention has been given to the common element of all lipid messenger pathways: the phospholipases that are essential to their biosynthesis. The characterization of the molecular details of phospholipase biochemistry has been challenging for several reasons. Phospholipases that are involved in signaling pathways are typically present in low amounts in cells, often displaying low solubility, and their hydrophobic substrates are preferred in aggregated form. These facts alone are enough to complicate assays and their analyses, and yet phospholipases have been known since the 1950’s when their activity was first discovered in pancreatic juice and cobra venom [3,4]. Since then, elegant efforts in cloning, expression, purification, sequencing, and structure determination, as well as studies of regulation and localization have helped to uncover details of many phospholipase groups [5]. Within the phospholipases, by far the most complex and well-known element is the phospholipase A2 superfamily. Phospholipase A2 enzymes (PLA2s) catalyze the hydrolysis of the sn-2 bond of phospholipids, generating free fatty acids and lysophospholipids, and are thus key enzymes which control the release of lipid mediator precursors. The
diversity within the enzyme family can be classified according to localization of the individual member (secreted versus cytosolic), sequence homology, cation requirement and molecular weight, generating at least nine major groups with varied molecular characteristics [5]. But what is the need for such diversity? The answer lies in the fact that mammalian tissues and cells, which often contain more that one PLA2, possess differently regulated pathways for fatty acid metabolism, requiring PLA2s to play crucial roles in cellular responses as diverse as phospholipid digestion and turnover, membrane remodeling, host defense, signal transduction and the production of inflammatory mediators [4]. The divergence displayed by PLA2s in terms of structure, cation requirement and catalytic mechanism is remarkable when considering that they all target phospholipidic substrates and catalyze an acyl hydrolysis reaction. This review will concentrate on the structural aspects of this important family of enzymes, with particular emphasis on the recent crystal structure of cytosolic PLA2. The new and unexpected fold of cystolic PLA2 reveals novel eccentricities employed by such enzymes in the manipulation of fatty acid metabolism. Structure and mechanism of secreted PLA2 enzymes
Secreted PLA2s (sPLA2s) are the most well-studied members of the PLA2 family and possibly of all lipidmetabolizing enzymes. Primarily on the basis of amino acid sequence and disulfide positioning, sPLA2s have been classified into three major types (I, II and III) and two different subtypes (A and B). They have been isolated from various snake and bee venoms, as well as from several mammalian sources [4]. The functions performed by sPLA2s range from the digestion of dietary fat, to cell proliferation and migration. In particular, the abundance of group IIA sPLA2 in various inflamed tissues, including the plasma of patients with sceptic shock and rheumatoid arthritis, suggests that it may have a crucial role in the pathogenesis of inflammation. Venom sPLA2s, on the other hand, seem to be directly involved in prey digestion; in addition, they have been reported to have roles in neurotoxic, myotoxic and anticoagulatory processes [6]. The sPLA2s are cysteine-rich molecules that catalyze the cleavage of the 2-acyl ester bond of phospholipids without any fatty acid selectivity and in a Ca2+-dependent manner. Disulfide bridges ranging in number from five to eight [5] guarantee a rigid three-dimensional structure, as well as stability and proteolytic resistance, features that ensure sPLA2s their full activity in the extracellular fluids where they are found [7]. X-ray crystal structures of sPLA2s from different sources [8–14] have revealed a common fold with
st8211.qxd
02/29/2000
R16
12:18
Page R16
Structure 2000, Vol 8 No 2
Figure 1 A gallery of lipase folds. (a) sPLA2. The compact 14 kDa molecule harbors an elapid loop (blue) and a tightly bound Ca2+ ion (pink) located near the active-site His—Asp dyad. (b) cPLA2. The C2 domain is in purple, with two tightly bound Ca2+ atoms in red. The catalytic domain is composed of a central β sheet (yellow) and a cap region (pink), which harbors a highly flexible lid that covers the active-site dyad. Notice the lack of interdomain contacts between the C2 and catalytic units. (c) Phospholipase Cδ1 has a vast area of contact between its C2 domain (purple) and catalytic domain (yellow). The EF-hand domain (red) is behind the C2 domain. (d) Cutinase. The smallest α/β hydrolase contains a solvent-exposed active site and thus does not undergo interfacial activation. The catalytic serine is depicted on the nucleophilic elbow.
three major α-helical segments, a double-stranded antiparallel β sheet (the β wing), and a highly evolutionarily conserved Ca2+-binding loop; diversions from commonality include a small, surface-exposed segment observed only in type I sPLA2s (the ‘elapid’ or ‘pancreatic’ loop; blue in Figure 1a) and a six amino acid C-terminal extension which characterizes type II sPLA2s. The active site of sPLA2s is located at the bottom of a channel lined with hydrophobic residues and is composed of a highly conserved His–Asp catalytic dyad and a tightly bound, heptacoordinated Ca2+ atom (Figure 2a). Crystal structures of sPLA2s in complex with inhibitors or transition-state analogs have shown the alkane chains of such molecules to lie roughly parallel in the channel, extending between 9 and 14 Å from the catalytic histidine towards the opening of the channel near the N-terminal helix [10–14]. In the absence of a catalytic triad, as observed in serine proteases, sPLA2s are incapable of hydrolyzing the 2-acyl group of diacylphospholipids through an acyl–enzyme intermediate; instead, the catalytic histidine, aided by a neighboring aspartate residue, polarizes a nearby water molecule, which then attacks the substrate sn-2 bond. The Ca2+ atom actively participates in catalysis by stabilizing the resulting transition state, and reassumes a state of heptacoordination as two water molecules move into the active site upon collapse of the tetrahedral intermediate and product release [15].
Based initially on observations that not all catalytically active venom sPLA2s are toxic and several catalytically inactive sPLA2s show myotoxic characteristics, it has now been shown that particular sPLA2s possess high-affinity cellular receptors (N-type and M-type) the recognition of which elicits toxic responses. Initially identified in brain tissue, N-type receptors are selectively recognized by neurotoxic sPLA2s. Although their molecular structure and physiological roles are still unclear, it is known that recognition occurs mostly through elements of the sPLA2 hydrophobic channel and Ca2+-binding loop [16]. Conversely, M-type receptors, which recognize different nonneurotoxic sPLA2s, have been well characterized as type I transmembrane molecules with a cysteine-rich N-terminal domain, a fibronectin type II unit, and eight repeats of a carbohydrate-recognition domain (CRD) followed by a single transmembrane region and a short cytoplasmic tail [17]. Site-directed mutagenesis and systematic CRD deletion studies of mammalian M-type receptors indicate that neither the cysteine-rich nor the fibronectin-like type II domains are involved in sPLA2 recognition, but it is CRD5 that provides the majority of binding contacts [18], although there are no carbohydrate chains on the sPLA2s. Structural elements of the sPLA2s involved in M-type receptor interactions lie within or close to the Ca2+-binding loop; however, the presence of Ca2+, which
st8211.qxd
02/29/2000
12:18
Page R17
Minireview Phospholipase A2 enzymes Dessen
is essential for enzymatic activity, does not seem to be required for receptor binding [19]. The identification of N- and M-type receptors provided the first definitive evidence that sPLA2s possess functions that are independent of their enzymatic activity [17].
R17
Figure 2
The cytosolic enzymes: Ca2+-independent PLA2s
In most tissues, the synthesis of eicosanoidal inflammatory mediators such as leukotrienes and prostaglandins is limited by the availability of their common precursor, free arachidonic acid, which must be hydrolyzed from phospholipidic membranes [20]. Leukotrienes have a role in the pathophysiology of asthma by inducing airflow obstruction through bronchoconstriction, increased secretion of mucus, and migration of inflammatory cells into air passages of asthmatic patients. Prostaglandins are mediators of both asthmatic and arthritic processes [21]. Hence, the control of the level of free arachidonic acid in cells must require a highly regulated mechanism. Although Ca2+-independent PLA2s (iPLA2s) do not seem to actively participate in the stimulation of arachidonic acid release from membranes due to their general lack of specificity for arachidonate-containing substrates, they establish their mark in arachidonic acid metabolism by playing an important role in phospholipid fatty acid remodeling [22]. The deacylation/reacylation cycle of membrane phospholipids, known as the Lands cycle, involves the cleavage of phospholipids by an intracellular phospholipase to generate a 2-lysophospholipid, which can subsequently be reacylated with a different fatty acid. The key element of this cycle is the PLA2 which generates the 2-lysophospholipid used in the acylation reaction. That is where the iPLA2s step in — by providing the lysophospholipid acceptor, they participate in the regulation of incorporation of arachidonic acid and other unsaturated fatty acids into membrane phospholipid [6,22]. To date, few intracellular iPLA2s (type VI) have been purified to homogeneity, and such enzymes have proven recalcitrant to detailed biochemical analysis because of their lability, low specific activities, and low abundance in cells. Hence, a detailed structural and mechanistic description will have to await further studies. However, efforts in cloning and purification have identified iPLA2s with diverse biochemical characteristics, ranging in molecular weight from 38 kDa to 85 kDa and being capable of forming 250–400 kDa oligomeric complexes [6]. Notably, the iPLA2s from Chinese hamster ovary (CHO) cells and P388D1 macrophage-like cells contain a lipase motif (Gly-X-Ser-X-Gly) displaying a putative active-site serine, as well as eight ankyrin repeats that have not been identified in any other PLA2; the repeats were suggested to function in the ordered association of the subunits to give the oligomeric complex [23,24].
Distinct active sites for acyl hydrolysis. Catalytic residues are highlighted in green. (a) The sPLA2 active site consists of a His–Asp pair and a heptacoordinated Ca2+ atom the function of which is to stabilize the reaction transition state. Other residues that also contact inhibitor molecules in co-complex structures are shown. (b) The active site of a neutral lipase (PDB code 1ETH) contains a classic His–Asp–Ser triad reminiscent of that of serine proteases. The oxyanion hole region is shown in yellow, as in (c). Other residues relevant for inhibitor molecule contact are also displayed. (c) The active site of cPLA2 lacks histidine and therefore differs from that of a classic lipase. Ser228 is located on the nucleophilic elbow. The sidechain of Arg200 may stabilize the phosphoryl moiety of a bound substrate.
st8211.qxd
02/29/2000
R18
12:18
Page R18
Structure 2000, Vol 8 No 2
Little is known about the catalytic mechanism of iPLA2s, except for the obvious lack of Ca2+ requirement and their selective inhibition by bromoenol lactone (BEL), a mechanism-based molecule originally proposed as a serine protease inhibitor [25]. The presence of a serine residue in a classic lipase motif together with the results of studies with BEL point to the possible employment of a serine–acyl intermediate in iPLA2 catalysis.
protein kinase C, GTPase-activating protein, synaptotagmin and phospholipase Cδ1, has been shown to bind phospholipid membranes in a Ca2+-dependent manner. A clear ubiquitous role for C2 domains in bilayer targeting is also supported by the observation that various components of cellular membranes, such as phospholipids, inositol phosphates and other membrane-associated proteins, have been identified as potential ligands [39].
Cytosolic phospholipase A2: membrane association and activation by Ca2+
C2 domains consist of an eight-stranded antiparallel β sandwich which in many cases coordinates between one and three Ca2+ atoms through a constellation of acidic residues on three Ca2+-binding loops (CBLs or ‘jaws’) [40–43]. Membrane binding is accomplished mostly by residues located on and around the apex of the CBL region. In the C2 domain of cPLA2, these amino acids have been implicated in interface recognition through the employment of methods ranging from the detection of N15/NH heteronuclear single-quantum coherence (HSQC) shifts upon incubation of C2 with dodecylphosphocholine micelles [43] to cysteine engineering followed by fluorescence probing [39]. It is conceivable that when phospholipids bind to the C2 domain they fill unsatisfied coordination sites on the bound Ca2+ ions [44], which explains why cPLA2 binding to phosphatidylcholine vesicles is enhanced tenfold in the presence of Ca2+ [45].
Of all members of the vast PLA2 family, it is the cytosolic PLA2 (cPLA2 or type IV PLA2) which has the most important role in lipid messenger biosynthesis. Although three isozymes of this group have been identified so far (α, β and γ) [26–28], it is cPLA2-α which is the best characterized and the most relevant for intracellular signaling (and will thus be referred to henceforth simply as cPLA2). cPLA2 is highly selective for phospholipidic substrates with arachidonic acid at the sn-2 position [26,29] and thus has a crucial role in eicosanoid biosynthesis. Interestingly, the cyclooxygenases (COX-1 and COX-2) which catalyze the metabolism of arachidonic acid to prostaglandins are localized in the nuclear membrane, while the leukotriene biosynthesizing enzyme 5-lipoxygenase translocates from the cytosol to the perinuclear membrane in response to an increase in intracellular Ca2+ [30], as does cPLA2 [31,32]. This chain of events is verified by the observation that activation or increased expression of cPLA2 is linked to increased arachidonate release as well as leukotriene and prostaglandin synthesis [33,34]. The most remarkable evidence regarding the strategic importance of cPLA2 in eicosanoid biosynthesis, however, comes from analyses of cPLA2 knockout mice, whose peritoneal macrophages are incapable of generating leukotrienes, prostaglandins or PAF [35,36]. The relevance of the role of these three messengers in the pathophysiology of different inflammatory processes, coupled to the high specificity of cPLA2 for arachidonate phospholipids, makes cPLA2 an attractive target for drug discovery — its specific inhibition would simultaneously block the biosynthetic pathways of all three molecules. cPLA2 is an 85 kDa molecule composed of an N-terminal C2 domain (138 amino acids; purple in Figure 1b) and a catalytic unit with 611 residues (yellow and pink Figure 1b [37]). Maximal activation of cPLA2 in intact cells requires mitogen-activated protein (MAP) kinase directed phosphorylation [34,38]; this event on its own, however, is insufficient for such activation if it occurs in the absence of an increase in intracellular Ca2+ concentration, a requirement for C2-mediated translocation of cPLA2 to the target membrane [26,31,32,37]. The C2 domain of cPLA2, like its structural counterparts observed in signal-transduction and vesicle-fusion proteins such as
Thus, for cPLA2, upon a submicromolar increase in intracellular Ca2+ concentration, the C2 domain recruits the Ca2+-independent catalytic domain to the nuclear or endoplasmic reticulum (ER) membranes, presenting it to its substrate [26,37,39]. The function of the C2 domain in cPLA2 biochemistry is thus clear: it must tether the fulllength molecule to the membrane, but must do so in a way as to present the catalytic domain to its substrate in a productive fashion. Hence, the relative interdomain orientation is of key importance for cPLA2 activity. This is also the case for inositol-specific phospholipase Cδ1 (PCLδ1; Figure 1c), which catalyzes the hydrolysis of membranebound phosphatidyl inositol-4-5-bisphosphate into two second messengers, namely D-myo-inositol-1,4,5-triphosphate and sn-1,2-diacylglycerol, in response to diverse extracellular signals. PLCδ1 is a tetradomain molecule; in addition to the C2 and catalytic units, it contains an EF hand and a pleckstrin homology (PH) domain. Based on crystallographic evidence, the mode of PLCδ1 binding to the membrane involves the initial tethering of the molecule through the PH domain, whereas the function of C2 is to fix the catalytic unit, and hence the active site, in close apposition to the substrate interface [40,41]. A comparison of the surface diagrams of full-length cPLA2 [46] and PLCδ1 [40,41] through a superposition of their C2 domains allows visualization of the critical function of C2 in orienting the catalytic domains with respect to the target membrane. In Figure 3, the CBL region of both C2
st8211.qxd
02/29/2000
12:18
Page R19
Minireview Phospholipase A2 enzymes Dessen
R19
domains (superimposed) points towards and interacts with the interface. It is clear that the C2 domains of cPLA2 and PCLδ1 place the active sites in approximately the same orientation. Interestingly, the vast hydrophobic contact surface between the C2 and catalytic domains of PLCδ1 does not allow further interdomain rotation, suggesting that this orientation is optimal for membrane interaction. Similarly, the relative orientation of the cPLA2 domains allows C2 to fulfill its function of tethering cPLA2 to the membrane, juxtaposing the active-site region and the substrate interface. In addition, due to the lack of major contacts between the cPLA2 C2 and catalytic units (see Figure 1b), an additional small interdomain rotation could be possible. The orientation of the cPLA2 domains observed in the crystal structure is biochemically productive, and is clearly an essential element of the enzyme’s cellular function.
Figure 3
The catalytic domain of cPLA2: an original fold for a new enzymatic mechanism
Superposition of surface diagrams of cPLA2 (yellow) and phospholipase Cδ1 (white) through their C2 domains. Residues in C2 that have been shown to bind lipids in N15/NH HSQC experiments (concentrated around the CBL region) [43] are depicted in dark blue. Active sites of both molecules (red) are similarly positioned and strategically oriented to interact with the target membrane. In cPLA2, the basic patch region is also available for bilayer contact. A flexible lid (pink) covers the activesite funnel and must move to allow substrate access.
The structure determination of multiple lipases have revealed that they all share a common topological feature, namely the α/β hydrolase fold, which consists primarily of a central, mostly parallel β sheet interwoven by α helices [47,48]. This α/β structure encompasses a single domain which can range from 197 residues (Fusarium solani cutinase; Figure 1d) to 582 amino acids (mouse acetylcholinesterase), indicating that variations within the fold are generated by the addition of secondary structure units without the inclusion of new domains [49]. The central β sheet has a left-handed superhelical twist, forcing the first and last strands to cross at an approximate angle of 90° to one another. Although the positions of the β strands are conserved, the twist of the β sheet varies among the different enzymes, as do the orientations of the α helices [48]. In addition to the common fold, lipases possess a serineacid-histidine catalytic triad reminiscent of the serine proteases (Figure 2b) and proceed catalytically through a serine–acyl intermediate. Within the twisted β sheet, the nucleophilic serine is located on the ‘nucleophilic elbow’, a tight turn between a central α helix and its preceding β strand [47,48]. The commonality of the α/β hydrolase fold, including identical positioning of the catalytic residues (both in terms of sequence and three-dimensional localization) [48] reveal a framework that seems to have diverged from a common ancestor, generating enzymes which differ in their substrate-binding sites but preserve the arrangement of the catalytic triad [47]. Prior to structure solution, cPLA2 showed signs that it too could be a member of the α/β hydrolases. Its mechanism proceeds through a serine–acyl intermediate [50,51]; the nucleophile, identified as Ser228 by site-specific mutagenesis, is located in the Gly-Leu-Ser-Gly-Ser sequence,
which is highly reminiscent of the classic Gly-X-Ser-X-Gly lipase motif [52]. In addition to Ser228, Asp549 and Arg200 were also shown to be essential for cPLA2 activity, but none of the 19 histidine residues were shown to have any catalytic role [53]. This fact suggested that the catalytic domain of cPLA2 might hold structural surprises. The X-ray crystal structure of cPLA2 [46] indeed reveals a catalytic domain harboring a twisted central β sheet (Figure 1b), with Ser228 located on a ‘nucleophilic elbow’. Beyond these two elements, however, there are no further similarities with α/β hydrolases. After the formation of the initial part of the β sheet, the amino acid chain is detracted from the core to form a highly basic, 180-residue region which positions itself between the active site (partly covering it) and the C2 domain — the ‘cap’ (residues 370–548; pink in Figure 1b). The cap is composed of two sets of antiparallel β strands and four α helices, a fold which possesses no similarity to any known structure. Although the specific function of the cPLA2 cap is unknown, its large number of solvent-exposed, basic amino acids (the ‘basic patch’) as well as its strategic location, facing the proposed membrane-binding region, are of note (see Figure 3). Because cPLA2 has affinity for anionic phospholipids [45,54], it is tempting to propose that this charged region in cPLA2 has a role in contacting such acidic substrates [46]. Interestingly, other membrane-targeting enzymes that also reveal a striking preference for negatively charged substrates, such as type IIb phosphatidylinositol phosphate kinase [55], several sPLA2s [56], protein kinase Cδ
st8211.qxd
02/29/2000
R20
12:18
Page R20
Structure 2000, Vol 8 No 2
[57] and PLCδ1 itself [40], also possess basic surface regions. A basic patch observed in the trimeric matrix protein of both human immunodeficiency virus 1 and Simian immunodeficiency virus 1 has also been implicated in stabilization of the molecule on the internal side of the acidic viral membrane [58,59]; these observations suggest a ubiquitous interface-association role for such charged regions on membrane-binding proteins. In addition to contacting the membrane interface, the multifunctional cap region of cPLA2 also harbors a flexible lid which, in the crystal structure, covers the buried active site (pink in Figure 3). Analysis of this active site reveals yet another difference between cPLA2 and enzymes of the α/β hydrolase family: the lack of any histidines in a 10 Å radius around the nucleophilic Ser228 (Figure 2c). Since mutagenesis studies failed to identify any histidine residue that could act in a classic catalytic triad as seen in serine proteases [46], this is not completely surprising. It is of note, however, that the active site of cPLA2 lacks any sidechains that could replace the histidine residue identified in all α/β hydrolases to date [53]. Even Arg200, the relevance of which in catalysis was highlighted by the lack of activity in a Arg200→Lys mutant [53], positions its guanidinium group 9 Å away from the Oγ atom of Ser228, and is thus not a candidate. These observations led to the proposal of a catalytic mechanism for cPLA2 which obviates the participation of a histidine residue [46]. In this model, Arg200 has a role in stabilizing the phosphoryl moiety of an incoming phospholipidic substrate. Ser228, activated by Asp549, attacks the sn-2 ester of a bound substrate molecule. The oxyanion hole, formed by the backbone atoms of Gly197 and Gly198 (located on the yellow turn in Figure 2c), polarizes the ester and stabilizes the transition state of the developing tetrahedral intermediate. Hydrolysis of the formed serine–acyl intermediate is then promoted by a bound water molecule, leaving cPLA2 free to bind another phospholipid substrate and ready to repeat the cycle. Conformational change and interfacial activation
By far the most fascinating phenomenon associated with lipid metabolism and lipase structure is that of ‘interfacial activation’, which describes the dramatic increase in lipase activity when substrate levels exceed the critical micellar concentration (CMC) [60]. Although distinct in structure and catalytic mechanism, sPLA2s, cPLA2, and various lipases of the α/β hydrolase family all undergo interfacial activation. Studies into the molecular basis of interfacial activation of lipases have resulted in the formulation of two main theories. The ‘substrate hypothesis’ suggests that factors such as concentration gradient and pressure at the membrane surface have important roles in the generation of this phenomenon. Conversely, the ‘enzyme hypothesis’ suggests that lipases undergo localized conformational changes upon membrane interaction. It is now
apparent that both the nature of the substrate as well as conformational changes in the protein may contribute to enzymatic activation at the interface [61]. X-ray crystallographic studies of secreted PLA2s, both in apo form and complexed to inhibitors, suggest rigid enzymes in which no conformational changes can occur upon interaction with the lipid–water interface. These studies, in conjunction with other biochemical evidence, suggest that catalysis at the interface occurs as a result of facilitated substrate diffusion from the binding surface through a hydrophobic channel that leads to the sPLA2 active site [15]. Recently, electron paramagnetic resonance (EPR) based membrane-docking methods have shown that the surface region of sPLA2s which surrounds the opening of the hydrophobic channel in fact constitutes the interfacial binding surface (i-face) [62]. The i-face participates in juxtaposition between enzyme and membrane (obviating membrane penetration) so that phospholipid transfer may occur without solvation of the hydrophobic alkyl groups [15,56] and in the absence of conformational changes. Notably, elegant charge-reversal mutagenesis studies have identified a patch of basic residues located on the i-face that has an important role in anionic phospholipid recognition and interfacial binding [63]. The explanation for the activation phenomenon of neutral lipases, however, has had a different history. Crystal structures of several lipases of the α/β hydrolase family have revealed active sites occluded by surface loops (or ‘lids’), which must undergo a conformational change in order for catalytic residues to be exposed to the substrate. Notably, lid movement also exposes vast hydrophobic areas which surround the site itself, and are thus likely to be stabilized by a lipidic interface. In several cases, this conformational change is necessary to strategically localize the oxyanion hole within the active-site region. The crucial role played in interfacial activation by structural elements of the lipases themselves is also clear in the crystal structure of cutinase, a small lipase which contains a solvent-exposed active site and thus does not undergo interfacial activation [64] (Figure 1d). cPLA2 also undergoes the phenomenon of interfacial activation by displaying a 1500-fold higher specific activity in the presence of micellar substrate [37]. Again, a structural explanation for this effect can be elicited. The crystal structure of human cPLA2 [46] shows that an amphipathic lid covers the active site of the enzyme; short flexible regions are ideally positioned to serve as hinges for the moveable lid (see Figure 3). An open cPLA2 lid would reveal a 30 Å wide hydrophobic platform funneling down into a deep active site, with members of the catalytic dyad located at the bottom of a 7 Å funnel stalk. The surface of the funnel should provide an optimal binding cradle for fatty acyl elements of the target interface, certainly a stabilizing factor of
st8211.qxd
02/29/2000
12:18
Page R21
Minireview Phospholipase A2 enzymes Dessen
major importance for cPLA2 activation. The location of the Ser228–Asp549 dyad partway into the body of the catalytic domain of cPLA2 also suggests that, in order for the substrate to come into contact with the active site, it is necessary for cPLA2 to become imbedded in the membrane itself. This hypothesis is supported by the existence of a 7 Å hydrophobic rim protruding from the base of the active-site funnel, which could be involved in interface recognition, as well by the observation that cPLA2 activity is inhibited in the presence of sphingomyelin-rich membranes [54]. It is of interest that PLCδ1, which also displays a hydrophobic rim around the active site and the interfacerecognition mechanism of which has been hypothesized to involve partial insertion of the catalytic domain into the membrane [40], is also inhibited by sphingomyelin [65]. Sphingomyelin has been reported to decrease membrane hydration through an increase in interlipid hydrogen bonding, causing the interface to become more rigid; this could inhibit protein penetration, and in the case of cPLA2, hinder access of substrate into the deep active site [46]. Conclusions
PLA2 enzymes owe the great interest they have generated over the past decades to their key activities in lipid mediator biosynthesis. Eicosanoids, powerful inflammatory signaling molecules, metabolically depend on the release of arachidonic acid from phospholipidic membranes by PLA2s, ensuring that these enzymes are under tight regulatory control and thus making them fascinating targets for biochemical studies. Due to their low abundance in cells and preference for substrates in aggregated forms, however, PLA2s have been difficult to characterize; in addition, the high molecular weight PLA2s have provided crystallographic challenges. The relationship between the structural diversity of the PLA2 family and the cellular processes in which their diverse lipid messenger products are involved is finally coming into perspective. Low molecular weight sPLA2s target phospholipidic substrates without fatty acid selectivity and have biological functions that reflect either their catalytic activities or interactions with specific receptors; their compact three-dimensional structures reveal a framework that must be sealed onto its bilayersequestered substrate, allowing it to diffuse into a deep hydrophobic catalytic channel. On the other hand, cPLA2 is the most specific enzyme targeting arachidonyl phospholipids, and does so by employing its C2 domain in an initial tethering step which positions the catalytic unit appropriately for partial membrane embedding, lid opening and active-site access by the substrate. The essential nature of PLA2s for the generation of lipidic messengers holds the key to the most exciting use for the knowledge gained in their structural and functional analysis: the design of novel, specific metabolic inhibitors for the control of a wide range of cellular processes.
R21
Acknowledgements The author wishes to thank former colleagues at Wyeth Ayerst Biochemistry for many helpful discussions.
References 1. Prescott, S. (1997). A thematic series on phospholipases. J. Biol. Chem. 272, 15043. 2. Goetzl E.J., An, S. & Smith, W.L. (1995). Specificity of expression and effects of eicosanoid mediators in normal physiology and human diseases. FASEB J. 9, 1051-1058. 3. Wittcoff, H. (1951). The Phosphatides. pp. 99-115, Reinhold Publishing Corp., New York. 4. Dennis, E.A. (1994). Diversity of group types, regulation, and function of phospholipase A2. J. Biol. Chem. 269, 13057-13060. 5. Dennis, E.A. (1997). The growing phospholipase A2 superfamily of signal transduction enzymes. Trends Biochem. Sci. 22, 1-2. 6. Murakami, M., Nakatani, Y., Atsumi, G., Inoue, K. & Kudo, I. (1997). Regulatory functions of phospholipase A2. Crit. Rev. Immunol. 17, 225-283. 7. Balsinde, J., Balboa, M.A., Insel, P.A. & Dennis, E.A. (1999). Regulation and inhibition of phospholipase A2. Annu. Rev. Pharmacol. Toxicol. 39, 175-189. 8. Fremont, D.H., Anderson, D.H., Wilson, I.A., Dennis, E.A. & Xuong, N.-H. (1993). Crystal structure of phospholipase A2 from Indian cobra reveals a trimeric association. Proc. Natl Acad. Sci. USA 90, 342-346. 9. Wery, J.-P., et al., & N.D. Jones (1991). Structure of recombinant human rheumatoid arthritic synovial fluid phospholipase A2 at 2.2 Å resolution. Nature 352, 79-82. 10. Scott, D.L., Otwinowski, Z., Gelb, M.H. & Sigler, P.B. (1990). Crystal structure of bee-venom phospholipase A2 in a complex with a transition-state analogue. Science 250, 1563-1566. 11. Scott, D.L., White, S.P., Browning, J.L., Rosa, J.J., Gelb, M.H. & Sigler, P.B. (1991). Structures of free and inhibited human secretory phospholipase A2 from inflammatory exudate. Science 254, 1007-1009. 12. Thunnissen, M.M., et al., & Verheij, H.M. (1990). X-ray structure of phospholipase A2 complexed with a substrate-derived inhibitor. Nature 347, 689-691. 13. Sekar, K., Eswaramoorthy, S., Jain, M.K. & Sundaralingam, M. (1997). Crystal structure of the complex of bovine pancreatic phospholipase A2 with the inhibitor 1-hexadecyl-3-(trifluoroethyl)-sn-glycero-2phosphomethanol. Biochemistry 36, 14186-14191. 14. White, S.P., Scott, D.L., Otwinowski, Z., Gelb, M.H. & Sigler, P.B. (1990). Crystal structure of cobre-venom phospholipase A2 in a complex with a transition-state analogue. Science 250, 1560-1562. 15. Scott, D.L., White, S.P., Otwinowski, Z., Yuan, W., Gelb, M.H. & Sigler, P.B. (1990). Interfacial catalysis: the mechanism of phospholipase A2. Science 250, 1541-1546. 16. Nicolas, J.-P., Ying, L., Lambeau, G., Ghomashchi, F., Lazdunski, M. & Gelb, M.H. (1997). Localization of structural elements of bee venom phospholipase A2 involved in N-type receptor binding and neurotoxicity. J. Biol. Chem. 272, 7173-7181. 17. Lambeau, G. & Lazdunski, M. (1999). Receptors for a growing family of secreted phospholipases A2. Trends Pharm. Sci. 20, 162-170. 18. Nicolas, J.-P., Lambeau, G. & Lazdunski, M. (1995). Identification of the binding domain for secretory phospholipases A2 on their M-type 180-kDa membrane receptor. J. Biol. Chem. 270, 28869-28873. 19. Lambeau, G., et al., & Lazdunski, M. (1995). Structural elements of secretory phospholipases A2 involved in the binding to M-type receptors. J. Biol. Chem. 270, 5534-5540. 20. Irvine, R.F. (1982). How is the level of free arachidonic acid controlled in mammalian cells? Biochem. J. 204, 3-16. 21. O'Byrne, P.M. (1997). Leukotrienes in the pathogenesis of asthma. Chest 111, 275-345. 22. Balsinde, J. & Dennis, E.A. (1997). Function and inhibition of intracellular calcium-independent phospholipase A2. J. Biol. Chem. 272, 16069-16072. 23. Tang, J., Kriz, R.W., Wolfman, N., Shaffer, M., Seehra, J. & Jones, S.S. (1997). A novel cytosolic calcium-independent phospholipase A2 contains eight ankyrin motifs. J. Biol. Chem. 272, 8567-8575. 24. Balboa, M.A., Balsinde, J., Jones, S.S. & Dennis, E.A. (1997). Identity between the Ca2+-independent phospholipase A2 enzymes from P388D1 macrophages and chinese hamster ovary cells. J. Biol. Chem. 272, 8676-8580. 25. Ackermann, E.J., Conde-Frieboes, K. & Dennis, E.A. (1995). Inhibition of macrophage Ca2+-independent phospholipase A2 by bromoenol lactone and trifluoromethyl ketones. J. Biol. Chem. 270, 445-450.
st8211.qxd
02/29/2000
R22
12:18
Page R22
Structure 2000, Vol 8 No 2
26. Clark, J.D., et al., & Knopf, J.L. (1991). A novel arachidonic acidselective PLA2 contains a Ca2+-dependent translocation domain with homology to PKC and GAP. Cell 65, 1043-1051. 27. Song, C., Chang, X.-J., Bean, K.M., Proia, M.S., Knopf, J.L. & Kriz, R.W. (1999). Molecular characterization of cytosolic phospholipase A2-β. J. Biol. Chem. 274, 17063-17067. 28. Underwood, K.W., Song, C., Kriz, R.W., Chang, X.J., Knopf, J.L. & Lin, L.-L. (1998). A novel calcium-independent phospholipase A2, cPLA2-γ, that is prenylated and contains homology to cPLA2. J. Biol. Chem. 273, 21926-21932. 29. Hanel, A.M. & Gelb, M.H. (1993). Processive interfacial catalysis by mammalian 85-kilodalton phospholipase A2 enzymes on productcontaining vesicles: application to the determination of substrate preferences. Biochemistry 32, 5949-5958. 30. Peters-Golden, L. & McNish, R.W. (1993). Redistribution of 5lipoxygenase and cytosolic phospholipase A2 to the nuclear fraction upon macrophage activation. Biochem. Biophys. Res. Commun. 196, 147-153. 31. Glover, S., et al., & Gelb, M.H. (1995). Translocation of the 85-kDa phospholipase A2 from cytosol to the nuclear envelope in rat basophilic leukemia cells stimulated with calcium ionophore or IgE/antigen. J. Biol. Chem. 270, 15359-15367. 32. Schievella, A.R., Regier, M.K., Smith, W.L. & Lin, L.-L. (1995). Calcium-mediated translocation of cytosolic phospholipase A2 to the nuclear envelope and endoplasmic reticulum. J. Biol. Chem. 270, 30749-30754. 33. Lin, L.-L., Lin, A.Y. & Knopf, J.L. (1992). Cytosolic phospholipase A2 is coupled to hormonally regulated release of arachidonic acid. Proc. Natl Acad. Sci. USA 89, 6147-6151. 34. Lin, L.-L., Wartmann, M., Lin, A.Y., Knopf, J.L., Seth, A. & Davis, R.J. (1993). cPLA2 is phosphorylated and activated by MAP kinase. Cell 72, 269-278. 35. Uozumi, N., et al., & Shimizu, T. (1997). Role of cytosolic phospholipase A2 in allergic response and parturition. Nature 390, 618-622. 36. Bonventre, J.V., et al., & Sapirstein, A. (1997). Reduced fertility and postischaemic brain injury in mice deficient in cytosolic phospholipase A2. Nature 390, 622-625. 37. Nalefski, E.A., et al., & Clark, J.D. (1994). Delineation of two functionally distinct domains of cytosolic phospholipase A2, a regulatory Ca2+-dependent lipid-binding domain and a Ca2+independent catalytic domain. J. Biol. Chem. 269, 18239-18249. 38. Qiu, Z.-H., Gijon, M.A., de Carvalho, M.S., Spencer, D.M. & Leslie, C.C. (1998). The role of calcium and phosphorylation of cytosolic phospholipase A2 in regulating arachidonic acid release in macrophages. J. Biol. Chem. 273, 8203-8211. 39. Nalefski, E.A. & Falke, J.J. (1998). Location of the membrane-docking face on the Ca2+-activated C2 domain of cytosolic phospholipase A2. Biochemistry 37, 17642-17650. 40. Essen, L.-O., Perisic, O., Cheung, R., Katan, M. & Williams, R.L. (1996). Crystal structure of a mammalian phosphoinositide-specific phospholipase Cδ. Nature 380, 595-602. 41. Grobler, J.A., Essen, L.-O., Williams, R.L. & Hurley, J.H. (1996). C2 domain conformational changes in phospholipase Cδ1. Nat. Struct. Biol. 3, 788-795. 42. Perisic, O., Fong, S., Lynch, D.E., Bycroft, M. & Williams, R.L. (1998). Crystal structure of a calcium-phospholipid binding domain from cytosolic phospholipase A2. J. Biol. Chem. 273, 1596-1604. 43. Xu, G.-Y., McDonagh, T., Yu, H.-A., Nalefski, E., Clark, J.D. & Cumming, D.A. (1998). Solution structure and membrane interactions of the C2 domain of cytosolic phospholipase A2. J. Mol. Biol. 280, 485-500. 44. Rizo, J. & Südhof, T.C. (1998). C2-domains, structure and function of a universal Ca2+-binding domain. J. Biol. Chem. 273, 15879-15882. 45. Hixon, M.S., Ball, A. & Gelb, M.H. (1998). Calcium-dependent and independent interfacial binding and catalysis of cytosolic group IV phospholipase. Biochemistry 37, 8516-8526. 46. Dessen, A., et al., & Somers, W.S. (1999). Crystal structure of human cytosolic phospholipase A2 reveals a novel topology and catalytic mechanism. Cell 97, 349-360. 47. Ollis, D.L., et al., & Goldman, A. (1992). The α/β hydrolase fold. Protein Eng. 5, 197-211. 48. Schrag, J.D. & Cygler, M. (1997). Lipases and the α/β hydrolase fold. Methods Enzymol. 284, 85-107. 49. Heikinheimo, P., Goldman, A., Jeffries, C. & Ollis, D.L. (1999). Of barn owls and bankers: a lush variety of α/β hydrolases. Structure 7, R141-R146.
50. Trimble, L.A., Street, I.P., Perrier, H., Tremblay, N.M., Weech, P.K. & Bernstein, M.A. (1993). NMR structural studies of the tight complex between a trifluoromethyl ketone inhibitor and the 85-kDa human phospholipase A2. Biochemistry 32, 12560-12565. 51. Hanel, A.M. & Gelb, M.H. (1995). Multiple enzymatic activities of the human cytosolic 85-kDa phospholipase A2: hydrolytic reactions and acyl transfer to glycerol. Biochemistry 34, 7807-7818. 52. Sharp, J.D., et al., & Kramer, R. (1994). Serine 228 is essential for catalytic activities of 85-kDa cytosolic phospholipase A2. J. Biol. Chem. 269, 23250-23254. 53. Pickard, R.T., et al., & Sharp, J.D. (1996). Identification of essential residues for the catalytic function of 85-kDa cytosolic phospholipase A2. J. Biol. Chem. 271, 19225-19231. 54. Leslie, C.C. & Channon, J.Y. (1990). Anionic phospholipids stimulate an arachidonoyl-hydrolyzing phospholipase A2 from macrophages and reduce the calcium requirement for activity. Biochim. Biophys. Acta 1045, 261-270. 55. Rao, V.D., Misra, S., Boronekov, I.V., Anderson, R.A. & Hurley, J.H. (1998). Structure of type IIb phosphatidylinositol phosphate kinase: a protein kinase fold flattened for interfacial phosphorylation. Cell 94, 829-839. 56. Gelb, M.H., Cho, W. & Wilton, D.C. (1999). Interfacial binding of secreted phospholipases A2: more than electrostatics and a major role for tryptophan. Curr. Opin. Struct. Biol. 9, 428-432. 57. Zhang, G., Kazanietz, M.G., Blumberg, P.M. & Hurley, J.H. (1995). Crystal structure of the Cys2 activator-binding domain of protein kinase Cδ in complex with phorbol ester. Cell 81, 917-924. 58. Rao, Z., Belyaev, A.S., Fry, E., Roy, P., Jones, I.M. & Stuart, D.I. (1995). Crystal structure of SIV matrix antigen and implications for virus assembly. Nature 378, 743-747. 59. Hill, C.P., Worthylake, D., Bancroft, D.P., Christensen, A.M. & Sundquist, W. (1996). Crystal structures of the trimeric human immunodeficiency virus type 1 matrix protein: implications for membrane association and assembly. Proc. Natl Acad. Sci. USA 93, 3099-3104. 60. Sarda, L. & Desnuelle, P. (1958). Action de la lipase pancreatique sur les esters en emulsion. Biochim. Biophys. Acta 30, 513-520. 61. Derewenda, Z.S. (1995). A twist in the tale of lipolytic enzymes. Nat. Struct. Biol. 5, 347-349. 62. Lin, Y., et al., & Gelb, M.H. (1998). Docking of phospholipase A2 on membranes using electrostatic potential-modulated spin relaxation magnetic resonance. Science 279, 1925-1929. 63. Han, S.K., Yoon, E.T., Scott, D.L., Sigler, P.B. & Cho, W. (1997). Structural aspects of interfacial absorption. J. Biol. Chem. 272, 3573-3582. 64. Longhi, S., Czjzek, M., Lamzin, V., Nicolas, A. & Cambillau, C. (1997). Atomic resolution (1.0 Å) crystal structure of Fusarium solani cutinase: stereochemical analysis. J. Mol. Biol. 268, 779-799. 65. Scarlata, S., et al., & Rebecchi, M.J. (1996). Inhibition of phospholipase Cδ1 catalytic activity by sphingomyelin. Biochemistry 35, 14882-14888.