Integrated Lipidomics in the Secreted Phospholipase A2 Biology

Mammalian genomes encode genes for more than 30 phospholipase A2s (PLA2s) or related enzymes, which are subdivided into several subgroups based on their structures, catalytic mechanisms, localizations and evolutionary relationships. More than one third of the PLA2 enzymes belong to the secreted PLA2 (sPLA2) family, which consists of low-molecular-weight, Ca2+-requiring extracellular enzymes, with a His-Asp catalytic dyad. Individual sPLA2 isoforms exhibit unique tissue and cellular localizations and enzymatic properties, suggesting their distinct pathophysiological roles. Recent studies using transgenic and knockout mice for several sPLA2 isoforms, in combination with lipidomics approaches, have revealed their distinct contributions to various biological events. Herein, we will describe several examples of sPLA2-mediated phospholipid metabolism in vivo, as revealed by integrated analysis of sPLA2 transgenic/knockout mice and lipid mass spectrometry. Knowledge obtained from this approach greatly contributes to expanding our understanding of the sPLA2 biology and pathophysiology.


Introduction
Phospholipase A 2 (PLA 2 ) hydrolyzes the sn-2 position of glycerophospholipids to yield fatty acids and lysophospholipids. In the view of signal transduction, the PLA 2 reaction has been considered to be of particular importance since arachidonic acid, one of the polyunsaturated fatty acids (PUFAs) released by PLA 2 , is metabolized to various lipid mediators such as prostaglandins and leukotrienes. In addition, lysophospholipids or its metabolites, such as lysophosphatidic acid and platelet-activating factor, also represent another class of lipid mediators. These lipid mediators exert numerous biological actions through their cognate G protein-coupled receptors on target cells. PLA 2 has also been implicated in membrane glycerophospholipid remodeling, thereby contributing to cellular homeostasis.
Mammalian genomes encode more than 30 PLA 2 s or related enzymes, which are classified into several subgroups on the basis of their primary structures and functions. Critical contributions of the intracellular PLA 2 families, namely cytosolic PLA 2 s (cPLA 2 s) and Ca 2+ -independent PLA 2 s (iPLA 2 s), to aracidnonic acid metabolism and membrane homeostasis, respectively, have been well established by numerous studies [1,2]. The secreted PLA 2 (sPLA 2 ) family represents structurally related, disulfide-rich, low molecular weight, lipolytic enzymes with a His-Asp catalytic dyad. sPLA 2 s occur in a wide variety of vertebrate and invertebrate animals, plants, fungus, bacteria, and viruses, and 11 sPLA 2 isozymes (IB, IIA, IIC, IID, IIE, IIF, III, V, X, XIIA and XIIB) have been identified in mammals [2][3][4]. Of these, sPLA 2 s belonging to the group I/II/V/X collection (conventional sPLA 2 s) are closely related, 14-19-kDa secreted enzymes with a highly conserved Ca 2+ -binding loop (XCGXGG) and a catalytic site (DXCCXXHD). In addition to these elements, there are six absolutely conserved disulfide bonds and up to two additional unique disulfide bonds, which contribute to the high degree of stability of these enzymes. Group III and group XII sPLA 2 s (atypical sPLA 2 s) share homology with the I/II/V/X collection of sPLA 2 s only in the Ca 2+ -binding loop and catalytic site, thereby representing the group III and XII collections, respectively. sPLA 2 enzymes hydrolyze the ester bond at the sn-2 position of glycerophospholipids with distinct selectivity toward sn-2 fatty acids and polar head groups in the presence of mM concentrations of Ca 2+ . Since individual sPLA 2 s display distinct cellular/tissue distributions and substrate head group specificities, they may play non-redundant, isoform-specific roles in vivo.
Although many potential functions of sPLA 2 s have been proposed on the basis of in vitro studies, the precise biological roles and relevant target membranes of these enzymes in vivo have remained elusive until recently. Several, if not all, sPLA 2 s are capable of releasing arachidonic acid from cultured cell membranes when overexpressed or added exogenously at excess amounts in vitro [2][3][4]. However, it still remains controversial whether this function could indeed be operated by sPLA 2 s in vivo. The reason why sPLA 2 s are secreted is most probably because sPLA 2 s participate in pathophysiology by regulating extracellular phospholipid metabolism, which include adjacent cell membranes (plasma membranes or microvesicles shed from cells), non-cellular lipid components such as lipoproteins and pulmonary surfactant, and foreign phospholipids such as microbe membranes and dietary lipids. The in vitro actions of individual sPLA 2 s on various target membranes are summarized in Table 1. This target variation may explain the molecular evolution of a number of sPLA 2 s with distinct localizations and substrate specificities. Therefore, once some phenotypes appear in sPLA 2 -knockout or -transgenic mice, this could be attributable to a combination of these varied actions rather than only by alterations in lipid mediator levels.
In the past few years, we have analyzed the phenotypes of transgenic or knockout mice for several sPLA 2 isozymes, in combination with a lipid profiling technique by mass spectrometry. This integrated approach, together with studies using these mice by other research groups, has helped us understand the potential action of a given sPLA 2 on particular target membranes and its impact on pathophysiology in vivo. In this article, we will give an overview of current analyses on transgenic or knockout mice for two particular conventional sPLA 2 s, group V and X, and an atypical sPLA 2 , group III. Also, we will give a brief summary of pathophysiological functions of other sPLA 2 s that have been clarified to date.

Biological Functions of sPLA 2 s in Vivo
Among the conventional sPLA 2 s, sPLA 2 -V has the simplest structure. It lacks group I-and II-specific disulfide bonds, group II-specific C-terminal extension, and group I-or X-specific N-terminal propeptide [5]. However, sPLA 2 -V is evolutionally close to group II sPLA 2 s, since the Pla2g5 gene resides within the gene cluster for the group II subfamily of sPLA 2 s. sPLA 2 -V is expressed at the highest level in the heart, followed by the lung. In the lung, sPLA 2 -V is expressed in the airway epithelium and alveolar macrophages, and its expression is markedly elevated in mice receiving models of asthma or acute respiratory distress syndrome (ARDS) [6]. Immunohistochemistry and in situ hybridization of sPLA 2 -V clearly indicate its location in the bronchial epithelium of patients with severe pneumonia [7].
To assess the biological functions of sPLA 2 -V, we generated transgenic mice overexpressing this enzyme in the whole body (Pla2g5-Tg) [8]. We found that Pla2g5-Tg mice died in the neonatal period (within 8 h after birth) due to respiratory failure. The lungs of Pla2g5-Tg mice exhibited atelectasis with thickened alveolar walls and narrow air spaces, accompanied by infiltration of macrophages and only modest elevations in eicosanoid levels. This severe pulmonary defect in Pla2g5-Tg mice was attributable to marked reduction of the lung surfactant phospholipids, phosphatidylcholine (PC) ( Figure 1a) and phosphatidylglycerol (PG) (Figure 1b), as demonstrated by ESI-MS (electrospray ionization mass spectrometry) analysis.
The principal function of lung surfactant, which is a mixture of phospholipids (90%) and surfactant proteins (10%), is to reduce the work of breathing by lowering alveolar surface tension during respiration. PC with saturated and monosaturated fatty acids (palmitic acid in particular) is predominant in surfactant phospholipids (~80% of total lipid) [9], and PG (~10% of total lipid) plays a role in phospholipid-protein interactions that maintain the alveolar surfactant layer, especially through interaction with the surfactant protein SP-B [10]. Since changes in the ratio of the surfactant components can dramatically alter the surface tension in small airways and alveoli, compromising airway patency, gas exchange and host defense, any surfactant abnormality can have severe pathological consequences in terms of lung function. Indeed, previous studies have indicated a role of sPLA 2 s in inflammation-mediated surfactant malfunction through hydrolysis of surfactant phospholipids [11][12][13][14][15][16][17][18]. Aberrant hydrolysis of surfactant phospholipids leads to ARDS, which is a clinically and pathologically complex syndrome due to acute life-threatening lung injury involving an alteration of pulmonary surfactant, and numerous predisposing factors can be involved in the etiology, including pneumonia and sepsis [9]. Hydrolysis of surfactant phospholipids is an early pathological event of ARDS, and hydrolysis of as little as 10-15% of surfactant can cause this serious condition. Levels of sPLA 2 activity in bronchoalveolar fluid (BALF) of patients with ARDS are often positively correlated with disease severity [11][12][13][14], and chemical sPLA 2 inhibitors that block classical sPLA 2 s protect animals against experimental ARDS or related lung injury [15]. Given that the expression of sPLA 2 -V is greatly elevated in human lungs with severe inflammation [7] and in cultured human bronchial epithelial cells stimulated with pro-inflammatory cytokines (Figure 1c), that sPLA 2 -V can efficiently hydrolyze lung surfactant phospholipids in vitro [16,18], and that the neonatal death of Pla2g5-Tg mice is in many aspects similar to that observed in mutant mice manipulated for a panel of genes that have been implicated in surfactant homeostasis [19,20], sPLA 2 -V may contribute to ongoing surfactant hydrolysis often observed in the lungs of patients with ARDS ( Figure 1d). Following our study using Pla2g5-Tg mice as shown above, three studies using mice null for sPLA 2 -V (Pla2g5 -/-) by other groups have delineated the crucial contribution of this sPLA 2 isoform to mouse airway disease models [6,21,22]. Thus, the allergen (ovalbumin or house dust mite)-induced, Th2-dependent asthmatic models, as well as the LPS-induced ARDS model, were significantly reduced in Pla2g5 -/mice compared with replicate Pla2g5 +/+ mice [6,21]. In the asthmatic models, the action of sPLA 2 -V occurs in two regulatory steps; one at the step of antigen uptake and processing by dendritic cells leading to the initiation of the Th2 response, and the other at the step of airway-resident cells which may contribute to the propagation of airway inflammation [22]. The airway-resident cell-dependent pathway requires catalytic activity of sPLA 2 -V, since aerosolized intake of native, but not mutant, sPLA 2 -V caused a dose-related increase of airway resistance, persistent airway narrowing, and leukocyte migration, and since intratracheal application of a neutralizing antibody against sPLA 2 -V ameliorated the asthmatic response [6]. However, eicosanoid levels in BALF were unchanged in this model, suggesting that the airway action of sPLA 2 -V does not profoundly depend on lipid mediators. Although the molecular mechanism underlying the airway-resident cell-dependent pathway has not yet been clarified, we speculate that the protection from disease-associated surfactant hydrolysis by the absence of sPLA 2 -V may be a likely explanation for this event. Thus, blockade of endogenous sPLA 2 -V could provide a potential new therapeutic approach for treating diverse phenotypes of human asthma.
Studies using Pla2g5 -/mice have also revealed unique functions of sPLA 2 -V in inflammation, host defense, and atherosclerosis. Pla2g5 -/mice displayed reduced zymosan-induced peritonitis since peritoneal macrophages produced less eicosanoids [23], were protected from Candida albicans infection since phagocytic killing of the fungi by macrophage was reduced [24,25], and were more sensitive to inflammatory arthritis since phagocytosis of the pro-inflammatory immune-complex by macrophages was reduced in the joints [26]. sPLA 2 -V can also potently hydrolyze phospholipids in low-density (LDL) and high-density (HDL) lipoprotein particles, and LDL receptor-deficient mice transplanted with Pla2g5 -/bone marrow cells are partially protected from atherosclerosis development [27]. Furthermore, a recent single nucleotide polymorphism analysis has revealed an association of the human sPLA 2 -V gene haplotype with plasma LDL levels in patients with type 2 diabetes [28], suggesting its metabolic role.

Group X sPLA 2 (sPLA 2 -X)
Structurally, sPLA 2 -X has both the group I-and II-specific properties. Unlike sPLA 2 -V, which is constitutively active once synthesized, sPLA 2 -X is synthesized as an inactive zymogen and converted to an active enzyme by proteolytic removal of the N-terminal propeptide [29]. Amongst the sPLA 2 members, sPLA 2 -X shows the highest affinity for PC and thereby for the PC-rich outer leaflet in the plasma membrane of mammalian cells [30,31]. Accordingly, supplementation or forcible transfection of exogenous sPLA 2 -X results in increased release of arachidonic acid and its oxygenated metabolites in many cell types. However, these results should be carefully interpreted, because unlike cPLA 2 , which is ubiquitously expressed and is a central player of arachidonic acid release [1], the expression of sPLA 2 -X is tissue-or cell-specific. In fact, sPLA 2 -X is constitutively expressed at high levels in the genital and digestive organs, where they play roles in sperm activation and gastrointestinal phospholipid digestion, respectively, independently of lipid mediator production [32,33].
In the lung, sPLA 2 -X is focally expressed in airway epithelial cells, and its expression is elevated in the epithelium as well as in alveolar macrophages following asthmatic challenge in both mice and humans [34,35]. The contribution of sPLA 2 -X to airway inflammation was confirmed by a study using mice lacking this enzyme (Pla2g10 -/-), in which the ovalbumin-induced, Th2-dependent asthmatic responses in the airway, including infiltrations of CD4 + and CD8 + T cells and eosinophils, mucus secretion, elevation of Th2 cytokines, and production of pro-asthmatic lipid mediators such as cysteinyl leukotrienes and prostaglandin D 2 (PGD 2 ), were markedly reduced [34]. Taken together with the evidence from Pla2g5 -/mice (see above), it has become obvious that the two particular sPLA 2 s, sPLA 2 -V and -X, participate in the asthma pathology. In addition, Pla2g10 -/mice are protected from neutrophil-induced myocardial damage following ischemia-reperfusion, where sPLA 2 -X is involved in the production of leukotriene B 4 (LTB 4 ) by neutrophils [36].
In order to address the in vivo action of sPLA 2 -X, we produced transgenic mice overexpressing this enzyme in the whole body (Pla2g10-Tg) [8]. Unexpectedly, in contrast to Pla2g5-Tg neonates that exhibited fatal respiratory failure (see above), systemic Pla2g10-Tg mice displayed no apparent abnormality of the respiratory tract with normal alveolar architecture and surfactant composition [8], despite the fact that sPLA 2 -X can potently hydrolyze surfactant PC in vitro [18]. This surprising result turned out to be because sPLA 2 -X protein existed as an inactive zymogen in most tissues. The active form of sPLA 2 -X was produced at inflamed sites in Pla2g10-Tg mice [8]. These results suggest that sPLA 2 -X mostly exists as an inactive zymogen under physiological conditions and that its proteolytic activation occurs during inflammation. In contrast, macrophage-specific Pla2g10-Tg mice developed severe lung inflammation which led to early death by 2~3-weeks of age [37]. Although the discrepancy between systemic and macrophage-specific Pla2g10-Tg mice is unclear, sPLA 2 -X expressed in alveolar macrophages might be efficiently converted by proteolytic processing to an active form.
Although systemic Pla2g10-Tg mice did not have any alveolar injury, we found a remarkable phenotype in these mice before weaning: they developed alopecia [38]. Although pelage hairs of Pla2g10-Tg mice initially grew, complete but transient hair loss was observed at 3-4 weeks of age, a period corresponding to the late stage of the initial hair cycle (Figure 2a). Proteolytic activation of sPLA 2 -X in Pla2g10-Tg skin temporally preceded hair loss. Histological analyses of the alopecic Pla2g10-Tg skin revealed hair follicle distortion, hyperkeratosis and sebaceous gland hyperplasia (Figure 2b), which were accompanied by increased expression of genes related to terminal differentiation of epidermis and reduced expression of genes related to hair development. ESI-MS analysis of Pla2g10-Tg skin revealed that sPLA 2 -X hydrolyzed phosphatidylethanolamine (PE), but not PC, molecular species to yield PUFAs (Figure 2c), which were further converted to some if not all eicosanoids. A schematic model for the action of sPLA 2 -X in Pla2g10-Tg skin is illustrated in Figure 2d. These results, together with the finding that endogenous sPLA 2 -X shows a hair cycle-dependent periodic expression in the outer root sheath of hair follicles in mouse skin (Figure 2e) [38], suggest a potential functional link between sPLA 2 -X and skin biology, and may provide a molecular explanation for the skin abnormality induced by aberrant expression of other sPLA 2 s such as sPLA 2 -IIA, whose transgenic mice also developed alopecia [39]. Importantly, in Pla2g10 -/mice, hair growth in the anagen phase was significantly delayed, and this was caused by growth retardation of the outer root sheath in hair follicles [40]. Thus, sPLA 2 -X intrinsically functions in the hair quality control.
The ability of sPLA 2 -X to potently hydrolyze phospholipids in LDL and HDL in vitro has led to the hypothesis that, as in the case of sPLA 2 -V (see above), sPLA 2 -X may also participate in atherosclerosis. Indeed, sPLA 2 -X-hydrolyzed LDL particles promote foam cell formation from mouse peritoneal macrophages [41]. These in vitro observations may be relevant to cardiovascular pathology, since Pla2g10 -/mice are protected from angiotensin-II-induced aortic aneurysm and atherosclerosis [42]. sPLA 2 -X-released PUFAs negatively regulates liver X receptor (LXR), and accordingly, deficiency of sPLA 2 -X results in augmented LXR activation leading to increased expression of LXR-target genes. Thus, in Pla2g10 -/mice, elevated expression of the ATB-binding cassette (ABC) transporters ABCA1 and ABCG1 led to increased cholesterol efflux by macrophages [43], that of the steroidogenesis acute regulatory protein StAR resulted in increased corticosterone production by adrenal cells [44], and that of PPAR (peroxisome proliferator-activated receptor ) facilitated adipogenesis and adiposity [45].
On the contrary, Pla2g10 -/mice maintained on a chow diet over one year gradually lost body weight, most likely because dietary phospholipid digestion and thereby lipid absorption in the gastrointestinal tract was perturbed [38]. Collectively, these observations have highlighted a novel role of sPLA 2 -X in the regulation of metabolic states.  6)), but not PC, was markedly decreased in Pla2g10-Tg mice compared with WT mice; (d) A schematic model of the sPLA 2 -X action in the skin. sPLA 2 -X is converted by certain skin proteases to its active form, which then hydrolyzes PE in skin membranes to liberate PUFA that is further metabolized to skin-acting eicosanoids; (e) Endogenous sPLA 2 -X is localized in the ORS of hair follicles (shown in red). ORS, outer root sheath; IRS, inner root sheath. For details, see [38].
Finally, sPLA 2 -X is abundantly expressed in testicular spermatigenic cells and is released from the acrosome of capacitative (activated) sperm. Pla2g10 -/spermatozoa displayed reduced acrosome reaction and thereby reduced fertility, and this defect could be restored by LPC, a potential sPLA 2 -X-generated lipid product [32,38]. sPLA 2 -X is also expressed in peripheral neurons such as dorsal ganglion (DRG) neurons, and DRG from Pla2g10 -/mice showed reduced, whereas that from Pla2g10-Tg mice showed increased, ex vivo neuritogenesis [38]. Probably because of the altered neuritogenesis, pain nociception in the acetic acid writhing test was partially ameliorated in Pla2g10 -/mice, whereas it was augmented in Pla2g10-Tg mice, compared with that in littermate control mice [38].

Group III sPLA 2 (sPLA 2 -III)
sPLA 2 -III is the only enzyme belonging to the group III collection. It is an unusually large protein (55 kDa) among the sPLA 2 family and consists of three domains, in which a central sPLA 2 domain displaying all the features of group III bee venom sPLA 2 , including 10 cysteines and the key residues of the Ca 2+ loop and catalytic site, is flanked by large and unique N-and C-terminal region [46]. sPLA 2 -III is processed to a sPLA 2 domain-only form (devoid of the N-and C-terminal domains), which is sufficient for its catalytic function [47,48]. sPLA 2 -III undergoes N-glycosylation and can hydrolyze PC and PE equally and augment arachidonate release from cell membranes more efficiently than sPLA 2 -IIA, and less efficiently than sPLA 2 -X and sPLA 2 -V. sPLA 2 -III is immunohistochemically detected in the vascular endothelium of various tissues, peripheral and central nervous systems, male reproductive tracts, and several types of cancer [48,49]. Implantation of sPLA 2 -III-transfected colorectal adenocarcinoma cells into nude mice promotes the growth of tumor xenografts [48]. Expression profiling of the full set of sPLA 2 s in human colon suggests that sPLA 2 -III might be a good candidate as a novel biomarker for colon cancers [50]. In the central nervous system, Pla2g3 mRNA is localized in DRG neurons in mice, and overexpression of human sPLA 2 -III in cultured neuronal cells facilitates neurite outgrowth and survival in correlation with the production of LPC, whereas knockdown of endogenous sPLA 2 -III by siRNA partially suppresses these processes [49].
To address the potential in vivo action of sPLA 2 -III, we produced transgenic mice overexpressing this enzyme in the whole body (Pla2g3-Tg). Unlike Pla2g5-Tg mice, which die shortly after birth due to a lung disorder resulting from aberrant hydrolysis of the lung surfactant phospholipids (see above), Pla2g3-Tg mice showed no respiratory disorder, and lung surfactant phospholipids did not show appreciable difference between control and Pla2g3-Tg mice [51]. Furthermore, although Pla2g10-Tg mice show alopecia (see above), Pla2g3-Tg mice had normal pelage hairs up to nine months of age. Later on, however, Pla2g3-Tg mice spontaneously developed inflammation such as dermatitis, lymphocytic sialadenitis and splenomegaly [51]. The dermatitis was accompanied by hyperkeratosis, acanthosis, parakeratosis, erosion, ulcer, neutrophil infiltration, and increased production of proinflammatory cytokines, chemokines and prostaglandin E 2 (PGE 2 ). It is thus likely that overexpression of sPLA 2 -III facilitates the production of pro-inflammatory lipid mediators in the whole body, leading to systemic inflammation.
To look for potential substrates for sPLA 2 -III in Pla2g3-Tg mice, lipids extracted from splenocytes of aged Pla2g3-Tg and littermate control mice were subjected to ESI-MS analysis. Several PC molecular species were detected in splenocytes of wild-type (WT) mice, and their composition did not differ appreciably from those of Pla2g3-Tg mice (Figure 3a, Upper). However, a notable difference that could account for the PLA 2 -mediated lipolysis was seen in PS; of the detectable PS molecular species, PS with C18:0-18:1 (m/z = 790.6) were ~50% less in Pla2g3-Tg mice than in control mice (Figure 3a, Lower). These results suggest that, in splenocyte membranes, PS with C18:0-18:1 may represent a major target substrate of sPLA 2 -III. Since PS is mainly present in the inner leaflet of the plasma membrane of live cells and exposed on apoptotic cell surfaces [52,53], extracellular sPLA 2 -III might preferentially hydrolyze PS with C18:0-18:1 on apoptotic cells and thereby modulate the life-span of inflammatory cells. In support of this idea, susceptibility of cell membranes to sPLA 2 s increases in apoptotic cells [54].
As in the case of sPLA 2 -V and -X, sPLA 2 -III can potently hydrolyze phospholipids in plasma lipoprtein particles [55]. Indeed, the decreased level of plasma lipoproteins, HDL in particular, was obvious in Pla2g3-Tg mice in comparison with WT mice (Figure 3b), suggesting HDL hydrolysis by overexpressed sPLA 2 -III. LDL treated with sPLA 2 -III in vitro was pro-atherogenic, promoting foam cell formation from macrophages. When Pla2g3-Tg mice that had been crossed with ApoE -/mice (Pla2g3 tg /ApoE -/-) were fed a high-cholesterol diet, lipid accumulation in the aortic walls was markedly increased as compared with replicate ApoE -/mice (Figure 3c). Immunohistochemistry and in situ hybridization revealed the presence of sPLA 2 -III in human atherosclerotic plaques, particularly in macrophages and smooth muscle cells [55,56]. These results suggest that sPLA 2 -III may have a role in acceleration of atherosclerosis development [55].  (c) The roles of sPLA 2 -III in epididymal sperm maturation are summarized. For details, see [57]. sPLA 2 -III is expressed in the testis and epididymis, and in the latter tissue the mature form of sPLA 2 -III is secreted from the proximal epididymal epithelium into the lumen [57]. We have recently succeeded in generating Pla2g3 -/mice and found that they displayed male infertility [57]. Although testicular spermatogenesis in Pla2g3 -/mice was grossly normal, spermatozoa from the cauda (tail) epididymidis displayed hypomotility, and their ability to fertilize intact eggs was markedly impaired. Epididymal spermatozoa in Pla2g3 -/mice had aberrant acrosomal structures and flagella with abnormal axonemes. These results revealed an unexplored role of this atypical sPLA 2 in epididymal lipid homeostasis, whose perturbation led to sperm dysfunction.
After the complex differentiation process of male germ cells, spermatozoa exit the seminiferous tubules of the testis through the efferent ducts toward the epididymis. During their transit from the caput (head) to the cauda (tail) epididymidis, sperm cells undergo significant morphological and biochemical modifications, which lead to acquisition of their forward motility and ability to recognize and fertilize oocytes [58]. Unique to mammalian sperm cells is the abundance of phospholipid species with C22-PUFAs, particularly docosahexaenoic acid (DHA) and docosapentaenoic acid (DPA), whose proportion in membrane phospholipids appears to correlate with sperm maturity, motility and fertility [59][60][61][62]. The percentage of DHA relative to total fatty acids is correlated with the normal morphology of sperm cells [61], and sperm from subfertile men with low sperm motility or counts contain a percentage of DHA lower than that from normal men [63]. Sperm maturation involves the remodeling of membrane phospholipids toward the acquisition of motility and fertility during sperm migration through the epididymis. Indeed, the increase in C22-PUFAs such as DHA and DPA and the reciprocal decrease in arachidonic acid (C20:4) favor an increase in the unsaturation degree of fatty acids in mouse sperm membrane during epididymal transit [59], which could consequently contribute to increasing the mouse sperm membranous fluidity [64,65]. Interestingly, ESI-MS analysis of sperm membrane phospholipids revealed that, during epididymal transit, PC in WT sperm underwent a dramatic shift in its acyl groups from oleic, linoleic and arachidonic acids to DPA and DHA, whereas this membrane lipid remodeling was compromised in Pla2g3 -/sperm [57]. Accordingly, cauda epididymal spermatozoa in Pla2g3 -/mice had PC species containing more oleate and less DHA/DPA than did those in Pla2g3 +/+ mice, a finding that appears to be consistent with the aforementioned notion that sperm with higher DHA percentages have better motility and fertility. Thus, sPLA 2 -III may participate in the hydrolysis of PC with oleic, linoleic and arachidonic acids in the sperm membrane during epididymal transit and that this event may be followed by reacylation of LPC, a PLA 2 reaction product, with DHA and DPA, leading to an increase of PC with DPA/DHA in mature spermatozoa. In the Pla2g3 -/epididymis, impairment of the deacylation step may eventually perturb the subsequent reconstitution of DPA/DHA in the sperm membrane, culminating in the asthenozoospermia phenotype.
We also found a notable change in the ESI-MS/MS profile of lipid mediators in the epididymis of Pla2g3 -/mice [57]. Thus, arachidonate/linoleate metabolites of the 12/15-lipoxygenase (LOX) and cytochrome P450 (CYP450) pathways, but not those of the cyclooxygenase (COX) and 5-LOX pathways, were substantially reduced in the epididymis of Pla2g3 -/mice compared with that of Pla2g3 +/+ mice (Figure 4a). Although the role of 12/15-LOX or CYP450 metabolites in male fertility has not yet been fully established, expression of 12/15-LOX in spermatogenic cells has led to the suggestion that it may participate in sperm maturation [62]. In this context, the possibility that certain 12/15-LOX-or CYP450-derived lipid mediator(s) may be at least partly responsible for the regulation of sperm maturation by sPLA 2 -III should be taken into account.
Additionally, sPLA 2 -III may also affect lipid transport between sperm and epididymal epithelial cells. Several lipoprotein components are secreted by epididymal epithelial cells [67] and associate with and dissociate from sperm membranes scheduled for endocytosis by epididymal principal cells [68]. Male fertility can be impaired to various degrees by inactivation of the genes involved in lipoprotein metabolism [69][70][71]. In fact, membrane transport by epididymosomes, a particular lipoprotein membrane particle emitted from caput epididymal principal cells into the lumen, is fundamental for the process of sperm cell maturation in the epididymis [59,72]. Our speculation that sPLA 2 -III may also affect this epididymal lipid transport is supported by the finding that, as assessed by ESI-MS, the epididymal fluid from Pla2g3 -/mice contained PC more abundantly than that from Pla2g3 +/+ mice (Figure 4b). Taken together, we conclude that sPLA 2 -III may regulate epididymal sperm maturation through (i) regulation of phospholipid remodeling in sperm membranes, (ii) production of 12/15-LOX and CYP450 metabolites, and (iii) modification of lipid transport between sperm and epididymal epithelial cells (Figure 4c).

Other sPLA 2 s: Classical and Novel Enzymes
Group IB pancreatic sPLA 2 (sPLA 2 -IB) is synthesized in the pancreatic acinar cells, and after secretion into the pancreatic juice, an N-terminal heptapeptide of the inactive zymogen is cleaved by trypsin to yield an active enzyme in the duodenum [74]. The main role of sPLA 2 -IB is digestion of dietary and biliary phospholipids. Thus, perturbation of this process by gene disruption (Pla2g1b -/-) or pharmacological inhibition of sPLA 2 -IB led to protection from diet-induced obesity and insulin resistance due to decreased lipid digestion and absorption in the gut [75,76]. In agreement, the PLA2G1B gene resides within a locus for obesity susceptibility in humans [77].
Group IIA sPLA 2 (sPLA 2 -IIA) is often referred to as an inflammatory sPLA 2 , since its expression is markedly induced during inflammation, cardiovascular diseases, and tissue damages [78]. When overexpressed, sPLA 2 -IIA is capable of augmenting arachidonic acid release in cytokine-stimulated cells, albeit more weakly than sPLA 2 -V, -X and -III [79]. Despite these facts, the contribution of sPLA 2 -IIA to inflammation has remained a subject of debate until recently, since a natural mutation of its gene in C57BL/6 and 129Sv mice [80] prevents the proper assessment of its functions by a classical gene targeting strategy. Intrinsic deficiency of sPLA 2 -IIA in these mouse strains is associated with increased incidence of intestinal polyposis and tumorigengesis [80], a phenotype that is reversed by transgenic expression of the Pla2g2a gene [81]. A recent study using Pla2g2a-deficient BALB/c mice as well as Pla2g2a-Tg mice has provided compelling evidence that the enzyme plays an exacerbating role in inflammatory arthritis [26]. The best-recognized physiologic function of sPLA 2 -IIA is the degradation of Gram-positive bacterial membrane, thereby providing the first line of antimicrobial defense of the host [82][83][84][85]. The serum level of sPLA 2 -IIA also shows correlation with the risk of cardiovascular diseases [86], and Pla2g2a-Tg mice fed an atherogenic diet developed atherosclerosis [87,88]. This effect is probably because sPLA 2 -mediated hydrolysis of LDL phospholipids leads to generation of smalldense, pro-atherogenic LDL particles that facilitate macrophage foam cell formation, even though the hydrolytic activity of sPLA 2 -IIA toward lipoportein particles is much weaker than that of sPLA 2 -V, -X and -III. Pla2g2a-Tg mice also displayed permanent alopecia and were susceptible to carcinogeninduced skin tumorigenesis [39,40].
The roles of other group II subfamily sPLA 2 isoforms remain elusive, since knockout or transgenic mice for these enzymes have not yet been reported. sPLA 2 -IIC is expressed in rodent testis, but not in humans [89]. sPLA 2 -IID is structurally most similar to sPLA 2 -IIA, and its transcript is constitutively detected in the lymphoid organs [90]. This enzyme may have immuno-suppressive functions, since it is expressed in regulatory T cells and its fusion protein has the ability to suppress inflammatory bowel disease and experimental autoimmune encephalomyelitis in mice [91]. sPLA 2 -IIE, another group IIA-related enzyme, is expressed constitutively in several tissues at low levels and has a lower catalytic activity than other group II sPLA 2 s [92]. sPLA 2 -IIF possesses a unique 30-amino acid C-terminal extension that contains an additional Cys residue, which might contribute to formation of a homodimer or a heterodimer with a second protein [93,94]. This enzyme is expressed most abundantly in the skin [100].
Lastly, group XII sPLA 2 s (sPLA 2 -XIIA and -XIIB) represent a unique collection of the sPLA 2 family. sPLA 2 -XIIA has the central catalytic domain with a His/Asp catalytic dyad, yet the location of Cys residues outside the catalytic domain is rather distinct from that of other sPLA 2 s [95]. High expression of this enzyme is found in many tissues, although its enzymatic activity is very weak. A study using Xenopus suggests the role of this enzyme in early neuronal development [96]. sPLA 2 -XIIB is structurally related to sPLA 2 -XIIA and is expressed in liver and intestine [97]. A recent study has demonstrated that the transcription of Pla2g12b was regulated by the transcription factor HNF-4 and its co-activator PGC-1, and deletion of the Pla2g12b gene resulted in increased fat accumulation in the liver leading to steatohepatitis, a phenotype similar to that seen in Hnf4a -/mice [98]. The aberrant fat accumulation in Pla2g12b -/liver was ascribed to impaired hepatic secretion of VLDL. However, because sPLA 2 -XIIB lacks the catalytic activity since the catalytic center His is replaced with Leu [97], the molecular mechanism whereby this sPLA 2 isoform regulates VLDL secretion remains unknown.

Conclusions
During the past decade, the biological functions of several sPLA 2 s and their target substrates have been clarified by studies using transgenic and knockout mice in combination with lipidomics. Nevertheless, full understanding of the biological roles of all sPLA 2 isoforms is still a challenging area of research. The control of particular sPLA 2 s, alone or in combination of multiple isoforms, should have advantages over the inhibition of selective lipid metabolic pathways in the treatment of various diseases. Interestingly, the pan-sPLA 2 inhibitor A-002 (varespladib), which inhibits the conventional class of sPLA 2 s, can markedly reduce the atherosclerotic lesion area in experimental animals and even in humans in early-phase clinical studies [99]. This fact points to the sPLA 2 family as a potential therapeutic target for atherosclerosis, and probably other diseases in which one or more sPLA 2 s are involved, such as asthma, arthritis, and metabolic syndrome.