PLA₁ (EC 3.1.1.32) catalyzes the hydrolysis of FAs exclusively at the sn-1 position of PLs. A free fatty acid (FFA) and a lysophospholipid (lysoPL) are the products of a PLA₁ reaction (Figure 2). This class of phospholipase is not well understood, and no crystal structures exist for any true PLA₁ [2,3]. The assignment of a function for any PLA₁ from any organism has yet to be firmly established. Historically, the biological role of this acyl hydrolase was often defined by its anticipated participation in the Lands Cycle, which is a deacylation-reacylation cycle that PLs are suspected to undergo in order to remodel their acyl chains to preserve a homeostatic molecular species composition of PLs in membrane bi-layers [4]. However, only one PLA₁ has ever been directly implicated in a Lands Cycle [5], though this cycle has repeatedly been shown to occur at the sn-2 position via the action of PLA₂, and studies continue to observe the phenomenon [6–10]. The assignment of a similar function for PLA₁, though potentially valid, cannot be decuded based on experimental data. Understanding of these enzymes is limited, though some progress has been made over the past 20 years [11].

What is intriguing is the observation that despite a theoretical role in the Lands Cycle, none of the PLA₁ cloned and characterized have been linked to membrane turnover and remodeling roles. On the contrary, there is some evidence to suggest that some PLA₁ enzymes are virulence factors [12] or act to generate bioactive lysoPL molecules. For example, lysoPL mediators such as lysophosphatidic acid (lysoPA) [13] are second messenger signaling components in pathways that vary in complexity and evolutionary conservation, and they have been implicated in numerous processes such as proliferation, protein transport, differentiation, invasion, and morphogenesis [14]. By activating specific G-protein coupled lysoPL receptors, they are now viewed as key factors in cell-to-cell communication [15]. The synthesis and regulation of the formation of lysoPLs are not fully understood, a reason why PLA₁ is an important enzyme to study as it could be involved in their formation.

The FA product from a PLA₁ reaction also has bioactive potential. This has been shown in plants, where a PLA₁ regulates jasmonic acid biosynthesis. Also, it has been postulated for decades that sn-2 arachidonic acid (AA) cleavage from PLs may sometimes be mediated by concerted sequential PLA₁/LysoPLA₂ activities [16–24]. No cloned PLA₁ has been implicated in this alternative route, and therefore it remains circumstantial with regards to the in vitro studies, which provide the most indirect evidence [16,22,23].

PLA₂ (EC 3.1.1.4) mediates acyl ester hydrolysis at the sn-2 position of PLs. PLA₂ are quite well conserved across taxa, and they consist of a broad range of enzymes that segregate into one of eleven groups within the superfamily [25]. Numerous PLA₂ can contribute to lysoPL signaling events [26] and can even down-regulate the bioactive lysoPL platelet activating factor (PAF) with the sn-2 acetate-specific cleavage of PAF by PAF acetyl hydrolase (PAF-AH) [27]. However, PLA₂ is most noted for its role in initiating the AA cascade (Figure 2) [28]. The sn-2 reacylation step of the Lands Cycle appears to be quite specific for AA in a number of cells [29]. Once liberated by a cytoplasmic PLA₂, AA is converted into over 150 known eicosanoids, including prostaglandins, leukotrienes, and thromboxanes, all powerful local hormones that act as mediators in many important processes such as inflammation in the higher eukaryotes [30]. Numerous crystal structures exist in this class of enzymes whose active mechanism either utilizes a catalytic histidine in a so-called dyad or a catalytic serine in either a dyad or a triad [25,27,31].

PLB is able to hydrolyze both the sn-1 and sn-2 FAs of PLs (Figure 1). Glycosylation is a common feature of PLBs, such as the one from Penicillium notatum, which contains numerous asparagines-linked carbohydrates and possess phospholipase and subsequent lysophospholipase activity [32]. The distinction between PLB and LysoPLA can, rightly so, often be muddled since most PLBs possess lysophospholipase activity [33]. On the other hand, some PLB enzymes have been erroneously annotated. For example, a case in which a purified hamster heart cytosol enzyme clearly displayed both sn-1 and sn-2 hydrolysis has been referred to as a PLA [34]. Also, the first so-called PLA₁ to be purified, cloned, and crystallized was the outer-membrane phospholipase A (OMPLA) from E. coli and other bacteria [35–37], but it has been known for a long time that it can cleave at both positions sn-1 and sn-2 of diacyl- or lysoPLs and thus should be considered a phospholipase B [2,24].

LysoPLA (EC 3.1.1.5) detoxifies detergent-like lysoPL intermediates in PL metabolism by removing the remaining acyl chain from the lysoPL (Figure 1). This class of enzymes is also not well characterized. LysoPLA could also be important in the regulation of the amount of bioactive lysoPLs used in receptor-mediated or other signaling mechanisms [38]. Some LysoPLA can function equally well as either a LysoPLA₁ or a LysoPLA₂ [17]. Other LysoPLA can deacylate specific species of lysoPL, like those with AA esterfied to the sn-2 position, for example [18,39]. Such in vitro specificity has helped to fuel the theory that the AA cascade could be initiated by the combined actions of a PLA₁ and a LysoPLA₂.

PLC or PI-PLC (EC 3.1.4.11) is mostly known for catalyzing the cleavage of the phosphorylated membrane lipid PI to produce, for example, the intracellular second messengers sn-1,2-diacylglycerol [23] and inositol (1,4,5)-trisphosphate (Figure 2). The phosphoinositides are a critically important class of PLs which have been extensively researched, along with the mechanisms by which they are metabolized by PLC [40]. PLC-mediated phosphoinositide production is a key regulating component of ion channels and transporters, and controllers of vesicular trafficking and the transport of lipids between intracellular membranes [41]. Other types of PLC include a secreted form in bacteria that prefers PC (E.C. 3.1.4.3), and a GPI-PLC form found in various organisms that specifically recognises non-inositol-acylated glycosylphosphatidylinositol anchors.

PLD (EC 3.1.4.4) hydrolyzes the sn-3 phosphodiester bond of mostly PC to generate a choline molecule and PA. PLD has also been studied in detail due to its link with the production of PA, an intracellular lipid messenger implicated in almost every conceivable aspect of intracellular membrane transport [42]. PLD is also a potential regulator of lyso-PA formation, whose biosynthetic formation is unclear, but could involve both PLD and PLA₁ [43]. The mode of action and structural characteristics of the various isoforms of PLD are well characterized [42,44].

One of the first phospholipases of the type A₁ to be purified by column chromatography was from bovine brain [21]. The enzyme was isolated from the soluble fraction and eluted at a molecular mass of 365 kDa from a Sephacryl S-300HR column. Lipases are known for their interfacial activation properties, and the 365 kDa molecular weight obtained could reflect the enzyme’s molecular weight upon association with buffer detergent micelles, or it could represent the enzyme in tetramer form. Bovine brain PLA₁ migrated as two bands of 112 kDa and 95 kDa by SDS-PAGE. The purified enzyme was shown to be resistant to metal chelators, PMSF, and DFP, but it was inactivated by ZnCl₂ and enhanced by Ca²⁺, Mg²⁺ and Sr²⁺. The enzyme exhibited broad substrate specificity in mixed micelles made with CHAPS, but the highest specific activity, 23.8 μmol/min/mg, was exhibited against PE(16:0/20:4), and LysoPLA₁ activity was also observed. A subsequent study showed that PC and PI could also be catalyzed at a high rate by bovine brain PLA₁ but only in the presence of other PLs like PS, PE, and PA [73]. This same group reported the first chromatographic purification of a LysoPLA₂ that showed selectivity towards deacylating arachidonate from lyso-PC(−/20:4) [39]. In addition to its LysoPLA₂ activity, this enzyme also showed PLA₂ activity, but at a slower rate. Bovine brain LysoPLA₂ was isolated from the soluble fraction and is an approximately 95 kDa polypeptide. Moreover, it became clear that prior sn-1 deacylation of PC(16:0/24:0) by bovine brain PLA₁ greatly increased the rate of arachidonate deacylation by brain LysoPLA₂. These results suggested that in bovine brain AA might possibly be generated by sequential PLA₁/LysoPLA₂ action, yet this has never been shown in vivo. Even more, the physiological substrates for these two enzymes have never been determined so it is impossible to know whether arachidonoyl-substituted PC or PE are in fact hydrolyzed by these enzymes. Unfortunately, not knowing the in vivo physiological substrates is a recurring theme in PLA₁ studies. It is also important to note that the genes encoding these two PLA activities have not been determined.

The phosphatidic acid-preferring phospholipase A₁ (PA-PLA₁) from bovine testis is well studied. The initial identification of PA-PLA₁ was from Mono Q fractions of high-speed supernatants from bovine testis, and to a lesser extent in bovine brain [53]. The enzyme displayed preference for PA in TX-100 detergent mixed micelles, and it also displayed a relatively small amount of LysoPLA₁ activity. At the time of its identification, PA was just beginning to be recognized as a second messenger that could affect a number of cellular processes. The production and, particularly, the attenuation of signaling PA can be mediated by multiple phospholipases that regulate the timing, location, amount, and various molecular species of PA [74]. A 14,000-fold purification of native PA-PLA₁ was achieved and evidence was provided that the enzyme was a homotetramer of 110 kDa subunits [75]. When the enzyme was examined under the same conditions as the bovine brain PLA₁ previously reported (i.e., 3–5 mM MgCl₂) [21], the results showed that PE(16:0/24:0) became the preferred substrate. When MgCl₂ was removed, PA-PLA₁ preference for GPA was restored. It is thus possible that the bovine testis PA-PLA₁ and the bovine brain PLA₁ are one and the same enzyme.

Further studies eventually led to the cloning of bovine testis PA-PLA₁, which was encoded by an 875 amino acid protein with a predicted molecular mass of 97.6 kDa [52]. A ~2000 fold increase in PA-PLA₁ activity was observed when the protein was expressed in COS1 cells. A lipase-like consensus sequence was identified and the active site serine residue within a SXSXG motif was experimentally established as being essential for PA-PLA₁ activity in the COS1 expression system. Apart from containing a lipase-like consensus motif, the PA-PLA₁ sequence differed from other lipases and phospholipases. Homologues were identified in the genomes of Caenorhabditis elegans, yeast, Drosophila, and in human and mouse. A homologue involved in shoot gravitropism has also been identified in A. thaliana [63]. Splice variants were proposed to possibly exist after three cDNAs were found that contained a 123-base deletion, and it was put forth that if splice variants had different substrate specificities, then the seeming discrepancy between the bovine brain and testis substrate specificities may be explained in this manner. A complex array of conditions were created to examine properties of purified PA-PLA₁ [75] using an interface composed of unilamellar lipid mixed micelles [76]. In the end, catalytic preference for PA(16:0/18:1) was reported to be observed when the unilamellar micelles contained (i) a low relative amount of PC; (ii) high relative amounts of PE, PS, and cholesterol; (iii) DAG; (iv) PE(18:0/20:4) instead of PE(16:0/18:1); and (v) 10 mol % PA per 100 mol % total phosphoglycerides. The study also showed that PA-PLA₁ could bind to membranes composed of anionic phosphoglycerides and could be stabilized by these membranes in the presence of albumin.

Other experiments with PA-PLA₁ provided information about the effects of phosphorylation and dephosphorylation on the behavior of the enzyme [77]. The first splice variant, PA-PLA₁α, was recombinantly expressed in Sf9 cells and purified, though a homogeneous fraction of enzyme was not shown. The recombinant protein was shown by mass spectrometry to be phosphorylated by protein kinase CK2, with which it also formed complexes, and by extracellular signal-regulated kinase 2. They also showed that protein phosphatase 2A could dephosphorylate some of the phosphoryl modifications, which led them to raise the possibility that the native counterparts of these enzymes could coordinately regulate the phosphorylation state of PA-PLA₁ in vivo. The physiological significance of PA-PLA₁ still remains to debatably [50], because it is most highly expressed in testis it has been suggested that PA-PLA₁ could be involved in PA signaling during spermiogenesis [74,76].

Budding vesicles mediate transport of proteins between intracellular compartments. One of the components of ER to Golgi COPII-coated vesicles of the secretory pathway is the Sec23p-Sec24p complex. In a study, that attempted to affinity isolate and identify new Sec23p-interacting proteins a novel protein, p125 (111 kDa), was identified. The predicted peptide sequence of p125 was homologous to bovine PA-PLA₁ [54]. In addition, the N-terminal region of p125 contains a proline-rich region that was shown to be the functional region of the enzyme in its association with Sec23p. Detection of low levels of recombinant p125 expression showed co-localization with β-COP and ERGIC53, an ER-Golgi intermediate compartment marker. In contrast, detection of high levels of p125 over-expression revealed p125 dispersed throughout the cells, not only in membranes, but also in the cytosol, causing disorganization of the ER-Golgi intermediate compartments. ERGIC-53 also showed a dispersed staining pattern in these cells, and a second type of staining showed that the Golgi itself was dispersed. p125 was also shown to co-localize with p115 and GM130 proteins [78], both of which play a role in vesicle tethering to Golgi membranes [79]. Depletion of p125 by RNAi suggested that p125 is needed for the proper organization and distribution of ER exit sites; however, p125-depleted cells maintained regular rates of protein transport from the ER [80]. Despite these detailed studies, no examination of stereo-specific enzyme activity was undertaken, so p125 may indeed not even be a PLA₁.

However, another protein, KIAA0725p, has also been identified as a homologue of both bovine PA-PLA₁ and p125 [55]. KIAA0725p is ubiquitously expressed in mammalian cells and is predominantly localized in the cytoplasm. Consistent with its bovine PA-PLA₁ homologue, KIAA0725p preferentially cleaves the sn-1 ester linkage of PA; however, in the absence of detergent it also hydrolyzed PE. PA specificity was determined using homogenates of KIAA0725p-expressing cells, and homogenates with the GXSXG seryl residue mutated to alanine showed no PLA₁ activity. Like p125, results showed that over-expression of KIAA0725p in cultured cells provoked dispersion of ERGIC53 and β-COP; but, unlike p125, over-expression also caused dispersion of p115 and GM130. This morphological phenotype was not due to PLA₁ activity because over-expression of the lipase-inactive mutant caused the same phenotype. However, over-expression of KIAA0725p, but not its lipase-inactive mutant, also caused aggregation of the ER, which was thus determined to be dependent on PLA₁ activity. The authors proposed that KIAA0725p may promote fusion of ER membranes by changing cone-shaped PA to inverted-cone shaped lyso-PA, which has been suggested to promote fusion pore formation, the last step of membrane fusion [81]. Recently it has been shown that the isoform iPLA(1) g is a novel membrane transport factor that mediates a membrane transport pathway between the ER and the Golgi that is distinct from the previously characterized COPI- and Rab6-dependent pathways [82].

A novel PLA₁ secreted from rat platelets was the first PLA₁ discovered which preferentially cleaves FAs from PS [46]. The predicted peptide sequence from the cloned cDNA of PS-PLA₁ was found to show significant sequence homology to hepatic lipase, lipoprotein lipase, pancreatic lipase, and endothelial lipase, but the enzyme does not possess appreciable neutral lipase activity. The enzyme was partially purified by sequential column chromatography from the culture medium of thrombin-activated platelets, and from medium of insect Sf9 cells expressing recombinant PS-PLA₁ in a baculovirus system. The resulting enzyme was shown to be a 50–55 kDa protein that was equally active towards PS and lyso-PS. PS-PLA₁ is also believed to be a glycoprotein. Due to its sequence similarity with lipases, the primary PS-PLA₁ peptide sequence reveals that is has a β-9 loop and a lid domain; but PS-PLA₁ has shorter versions of each [46], which could explain why diisopropylphosphofluoridate (DFP) inhibits this enzyme, if the truncated lid domain ineffectively covers the active site serine. A subsequent study revealed that PS-PLA₁ might be the synthetic route in the production of bioactive lyso-PS since it was shown that it has the ability to efficiently stimulate histamine release from rat peritoneal mast cells, especially when in the presence of apoptotic Jurkat cells [83]. In this system, 2-acyl-1-lysoPS was released from apoptotic cells exposed to PS-PLA₁ and it was proposed that PS-PLA₁ may play an in vivo role in hydrolyzing PS exposed on plasma membranes of apoptotic, dead, and cytokine-stimulated cells, as a means to transduce mast cell activation mediated by 2-acyl-1-lysoPS. PS-PLA₁ΔC, a splice variant of the human version of PS-PLA₁, shows poor catalytic activity towards PS and the other diacylPLs but is able to deacylate lyso-PS effectively and is therefore a LysoPLA [84]. PS-PLA₁ΔC possesses a truncated C-terminal region; therefore peptides on the complete C-terminal domain of PS-PLA₁ are thought to play a role in substrate recognition. In a very recent study PS-PLA₁ expression in human THP-1-derived macrophages, responsible for the immune responses to allograft rejection, are activated via TLR4. This activation can be inhibited by corticosteroids, which are used at high dosages to suppress chronic allograft rejection [57].

A sequence similarity search of the PS-PLA₁ detected another PLA₁ enzyme which was homologous to neutral lipases, that of membrane-associated phosphatidic acid-selective phospholipase A₁ alpha and beta (mPA-PLA₁α/β) [43,56]. Both recombinant forms show PA-specific substrate specificity. Attempts to purify the recombinant membrane proteins from Sf9 cells failed but medium from mPA-PLA₁α-expressing cells was shown to activate a lyso-PA receptor family member, LPA3/EDG7. Medium from cells expressing an active site serine mutant failed to induce a receptor response. These results indicated that cells expressing recombinant mPA-PLA₁α were able to produce and release bioactive lyso-PA. When expressed in HeLa cells mPA-PLA₁α was recovered from the cell supernatant whereas mPA-PLA₁β was still membrane associated [56]. A set of elegant experiments also showed that mPA-PLA₁α/β acted after and in concert with both endogenous and exogenous PLD to form lyso-PA. This report thus suggests a possible in vivo metabolic pathway for lyso-PA production involving the sequential action of a PLD and mPA-PLA₁α/β. Thus, both PS-PLA₁ and mPA-PLA₁α/β thus have evolved from lipases to specialize in producing bioactive lyso-PL mediators [5].

It has been reported that microsomes from guinea-pig heart possess PLA₁ activity on PE and PC, and that sn-1 cleavage could be influenced by the sn-2 FA since more efficient hydrolysis was observed when the sn-2 FA was polyunsaturated [85]. The enzyme(s) responsible for PLA₁ activity from guinea-pig heart microsomes have not been identified and purified to any extent other than into subcellular fractions, but a few studies are worth mentioning since they are devoted to possible regulatory mechanisms of PLA₁ activity in this tissue, and very few other studies exist that begin to explain how some PLA₁ activity could be controlled. No PLA₁-activating agonists are known, so a possibility that PLA₁ could be receptor activated by G-proteins was investigated by measuring the response of PLA₁ activity towards guanine nucleotides, which can activate G-proteins. PLA₁ activity towards PC(16:0/18:2) was partially inhibited by guanosine 5′-[γ-thio]triphosphate (GTP[S]), but not by GTP, guanosine 5′-[γ-thio]diphosphate (GDP[S]), GDP, ATP or adenine 5′-[γ-thio]diphosphate (ATP[S]) [86]. On the other hand, PLA₁ activity on PE(16:0/18:2) was stimulated by GTP[S] but not GDP[S] or ATP[S] [86]. PE(16:0/18:2) hydrolysis by guinea-pig heart microsomes was also shown to be stimulated 40–60% by isoprenaline. It has still yet to be shown in vivo if PLA₁ activity in guinea-pig heart microsomes is activated by the binding of isoprenaline to adrenergic receptors and mediated via activation of G-proteins. Without either the purification or cloning of the enzymes responsible for this activity in guinea-pig heart microsomes, it is unlikely that any resolution of the issue will be made. It is worth recalling, however, that a ubiquitously expressed mammalian PA-PLA₁ homologue, p125, is localized to the ER and interacts with Sec23p, which acts as a GTPase-activating protein that plays a role in uncoating budding vesicles [87]. Whether these PLA₁ activities are due to the same or different enzymes is unknown, but it does stress the importance of how substrate specificity is analyzed in vitro, and the caution with which to interpret such data. To confuse the situation even further, guinea-pig heart microsomes were also shown to possess a LysoPLA₂ that could hydrolyze lyso-PC(−/18:2 or 20:4) more efficiently than lyso-PC(−/18:1 or 16:0) [18].This result lent some support to the standing possibility that arachidonic and linoleic acids could be released by LysoPLA₂ acting in concert with a PLA₁.

The Caenorhabditis elegans, IPLA-1 and ACL-10 have phospholipase A₁ and acyltransferase activity respectively, both of which recognize the sn-1 position of their PI substrate [88]. The PI sn-1 fatty acid remodeling by sequential deacylation and reacylation, which resulted in stearic acid as the major fatty acid at the sn-1 position, has been proposed to be crucial for asymmetric division [5].

The cDNA of one allergen component of Dolichovespula maculate (white-faced hornet) venom has been cloned and shown to encode a PLA₁ (Dol m I) with weak lipase activity [47]. In fact, at the time of the study, the derived peptide sequence of Dol m I had no homology to other PLA₁ genes because it was similar to mammalian lipases. Now, not only is Dol m I 40% identical to pancreatic, hepatic, lipoprotein, and endothelial lipase, but it also shows homology with PS-PLA₁ [46] and mPA-PLA₁ [43]. Characterization of the 34–37 kDa enzyme has not been undertaken presumably due to a priority interest in understanding its immunochemical properties as it relates to its contribution to allergenicity. Along these same lines, two homologues of Dol m I have been cloned from Vespula spp. (yellow jacket) (Ves v I and Ves m I) which show 67% sequence identity with the white-faced hornet PLA₁ [58,59], and recent success in recombinant expression of PLA₁ (Ves v I) from the yellow jacket will allow a more detailed understanding of the molecular and allergological mechanisms of insect venoms, providing a valuable tool for diagnostic and therapeutic approaches in hymenoptera venom allergy [60].

The PLA₁/LysoPLA theorized coordinated route towards the release of unsaturated FAs, (e.g., arachidonic acid (AA)), has been supported by in vitro studies using homogenates from the kinetoplastid protozoan parasite Trypanosoma brucei, the infective agent of African sleeping sickness [23]. These studies observed robust and optimal activity at pH 6.0–8.5 and a requirement of sn-1 acyl cleavage prior to the release of AA from sn-1-palmitoyl-sn-2-arachidonoyl-sn-glycero-3-phosphatidylcholine [PC (16:0/20:4)] [89–91]. The liberated unsaturated FAs such as AA from lipids in T. brucei have been implicated in regulating calcium mobilization [92,93], and as a precursor for prostaglandin biosynthesis [94].

Two recent detailed studies describe the cloning and characterization of the cytosolic T. brucei PLA₁. TbPLA₁ is unique from other eukaryotic PLA₁ because it is phylogenetically related to bacterial secreted PLA₁ [49,66]. TbPLA₁ was most likely acquired by a prokaryotic-to-eukaryotic horizontal gene transfer event of a PLA₁ from Sodalis glossinidius, a bacterial endosymbiont of the insect vector of the protozoan parasite, the tsetse fly. These studies employed both in vitro and in vivo analytical techniques to establish that PC is the preferred substrate. The TbPLA₁ homozygous null mutants constitute the only PLA₁ double knockouts from any organism and helped to established that the enzyme functions in vivo to synthesize lyso-PC metabolites containing long-chain mostly polyunsaturated and highly unsaturated fatty acids.

One report described the existence of PLA₁ activity in Trypanosoma cruzi, the etiologic agent of Chagas disease [95]. When trypomastigote and amastigote suspensions were radiolabeled with oleic acid and lysed, the radiolabeled PC content decreased over time. In addition, when PC(16:0/18:1) was incubated with epimastigote homogenates at pH 4.7, lyso-PC(−/18:1) was produced, suggesting sn-1 hydrolytic specificity. T. cruzi PLA₁ was partially purified (1900-fold by specific activity) from epimastigote homogenate supernatants and a protein band of 38 kDa could be seen by SDS-PAGE, and size exclusion chromatography suggested PLA₁ activity eluted with an apparent molecular mass of 40 kDa. T. cruzi PLA₁ seems to be a glycoprotein due to its binding to ConA-Sepharose, possesses no divalent cation requirements, and is released from the cell by digitonin together with lysosomal markers. The physiological role played by lysosomal T. cruzi PLA₁ has not been determined, though the authors suggest that the FA and lyso-PL products could have a possible role in pathogenesis of the disease.

Among the plant hormones jasmonic acid is considered a multifunctional growth and stress regulator [97], and it is structurally similar to “animal” eicosanoids. Jasmonic acid is an oxylipin signaling molecule and a derivative of linolenic acid (C18:3). Jasmonate has been shown to regulate or co-regulate a variety of processes in plants, such as responses to biotic and abiotic stresses, tendril coiling, fruit ripening, and the developmental maturation of stamens and pollen in Arabidopsis. The enzymes involved in the jasmonic acid biosynthetic pathway have been elucidated, but the last one discovered, a phospholipase A₁ called Defective in Anther Dehiscence 1 (DAD1), was one of the most important because it is the enzyme responsible for the initial release of C18:3 from cellular lipids [61]. Starting from an Arabidopsis thaliana DAD1 mutant, the WT DAD1 gene was isolated and found to encode a chloroplastic 45 kDa PLA₁ lipolytic enzyme. A rescued phenotype was attained by either complementing the mutant with WT DAD1 or after supplying exogenous jasmonic acid. The DAD1 protein sequence showed some apparent similarities with fungal lipases, it possessed a consensus GHSLG motif, and eleven homologous proteins were identified in Arabidopsis alone. A maltose binding-DAD1 fusion protein lacking 72 amino acid residues (an N-terminal transit peptide) was recombinantly expressed and used to examine lipase activity. PC was the only PL tested, and the activity on it was more than 80% greater than on TAG, which suggested that the C18:3 precursor for jasmonic acid biosynthesis is stored in cellular PLs. DAD1 is thought to be of prime importance for the regulation of jasmonic acid levels. A PLA₁ from Capsicum annuum (hot pepper) showed a considerable degree of overall sequence identity to Arabidopsis [98].

In a search for A. thaliana sterol acyltransferases, one of the four enzymes found to be homologous to lecithin (PC): Cholesterol acyltransferases was determined to possess PLA₁ activity [62,63]. AtLCAT3 (Arabidopsis thaliana lecithin:cholesterol acyltransferase 3) heterologous expression in yeast resulted in the PC, PS, and PE content to be half as much as those species in control yeast, while lyso-PC, lyso-PE and FFA were strongly increased. There was also a higher TAG content in the cells expressing AtLCAT3. AtLCAT3 fractionated with yeast microsomes, which subsequently shown to be able to hydrolyze various PL species, including lyso-PC and PA. However, neither acyltransferase activity nor TAG activity was observed with yeast microsomes expressing the enzyme. The analogous serine, histidine and aspartic acid residues that are part of the conserved catalytic triad of Homo sapiens LCAT were shown to be essential for activity of AtLCAT3. The physiological role of AtLCAT3 is as yet unknown.