Neuronal Expression and Neuritogenic Action of Group X Secreted Phospholipase A2*

Although individual mammalian secreted phospholipase A2 (sPLA2) enzymes exhibit unique tissue and cellular distributions, the cell type-specific functions of each enzyme remain largely unknown. In this study, we found by immunohistochemistry that group X sPLA2 (sPLA2-X) is uniquely located in the peripheral neuronal fibers, an observation that was supported by detection of its transcript and protein in the neuronal cell line PC12 and in primary dorsal root ganglia neurons. Adenoviral expression of sPLA2-X in PC12 cells facilitated neurite outgrowth, particularly when combined with a suboptimal concentration of nerve growth factor. In neuronally differentiated PC12 cells, sPLA2-X was preferentially localized in the Golgi apparatus and growth cones, and proteolytic conversion of the proenzyme to mature enzyme mainly occurred after the secretion process. The neurite-extending ability of sPLA2-X depended on the production of its catalytic product, lysophosphatidylcholine. Moreover, nerve growth factor-induced neurite extension of PC12 cells was modestly but significantly attenuated by an anti-sPLA2-X antibody or by a small interfering RNA for sPLA2-X. These observations suggest the potential contribution of sPLA2-X to neuronal differentiation, and possibly repair, under certain conditions.

The sPLA 2 family represents a group of structurally related, disulfide-rich, and Ca 2ϩ -dependent low molecular weight enzymes with a His-Asp catalytic dyad (1,2). Individual sPLA 2 s exhibit unique tissue and cellular localizations and enzymatic properties, suggesting their distinct roles in various pathophysiological events. Of these enzymes, group X sPLA 2 (sPLA 2 -X) is of particular interest, because this enzyme shows the highest capacity to hydrolyze phospholipids in intact mammalian cell membranes when overexpressed (3,4) or added exogenously (5,6). This ability of sPLA 2 -X depends primarily on its high interfacial binding affinity for phosphatidylcholine (PC), a phospholipid that is enriched in the outer leaflet of the plasma membrane. Structurally, sPLA 2 -X is similar to both group I and group II sPLA 2 s in that it has an N-terminal propeptide and a group I-specific disulfide (like sPLA 2 -IB) as well as a C-terminal extension and a group II-specific disulfide (like sPLA 2 -IIA) (7). As in the case of sPLA 2 -IB, it has been proposed that sPLA 2 -X is converted from a catalytically inactive zymogen to a mature, catalytically active form after proteolytic removal of the N-terminal propeptide (5). However, the regulatory mechanism for this proteolytic processing is still obscure, because the active enzyme is spontaneously secreted from sPLA 2 -X-transfected cells during culture (3,4). Recent immunohistochemical analyses have demonstrated the expression of sPLA 2 -X in macrophages and alveolar epithelial cells in the lung (8,9), spermatogenic cells and duct epithelium in male reproductive organs (10), and in atherosclerotic plaques (11). Although these distributions suggest the potential roles of sPLA 2 -X in airway inflammation, reproduction, and atherosclerosis, the true physiological and pathological functions of this particular sPLA 2 enzyme are still a subject of debate.
During the course of our immunohistochemical analyses of various sPLA 2 enzymes in human tissues (9,10,12), we found that sPLA 2 -X is uniquely located in the neuronal fibers in multiple peripheral tissues. On the basis of this observation, we herein examined the expression and potential functions of sPLA 2 -X in neuronal cells. We show that sPLA 2 -X has the ability to induce neurite outgrowth of neuronal cells through production of lysophosphatidylcholine (LPC). We also provide evidence that the conversion of the proenzyme to the mature enzyme mostly occurs after the secretion process, thus supporting the extracellular action of this enzyme.
Isolation and Culture of Mouse Dorsal Root Ganglia (DRG) Explants-Procedures for animal handling and preparation of mouse tissue sections were approved by the ethical committee of our Faculty. Mouse DRG explants were obtained from E12.5 to E13 of C57BL/6 mice (Shizuoka Laboratory Animal Center) (16). Mouse fetuses were cut from the upper cervical vertebrae to the tails along the spinal cords. The spinal cord sections were separated from the bodies, and DRG neurons attached to the spinal cords were picked up and harvested. DRG explants were then cultured for up to 36 h in Dulbecco's modified Eagle's medium (Nissui Pharmaceutical) supplemented with 20 ng/ml NGF in poly-D-lysine-coated glass bottom dishes (MatTek) pre-coated with 20 g/ml laminin (Biochemical Technologies).
Northern Blotting-Equal amounts (ϳ10 g) of total RNA obtained from cells by use of TRIzol reagent (Invitrogen) were applied to separate lanes of 1.2% (w/v) formaldehyde-agarose gels, electrophoresed, and transferred to Immobilon-N membranes (Millipore). The resulting blots were then probed with appropriate cDNA probes that had been labeled with [ 32 P]dCTP (Amersham Biosciences) by random priming (Takara Biomedicals). Hybridization and subsequent membrane washing were carried out as described previously (3).
Reverse Transcription-PCR-Synthesis of cDNA was performed using 0.5 g of total RNA from cells and tissues and avian myeloblastosis virus-reverse transcriptase, according to the manufacturer's instructions supplied with the RNA PCR kit (Takara Biomedicals). Subsequent amplification of the cDNA fragments was performed using 0.5 l of the reverse-transcribed mixture as a template with specific primers for each sPLA 2 (Fasmac). For amplification of sPLA 2 -IID, -IIE, -IIF, -V, and -X, a set of 23-bp oligonucleotide primers corresponding to 5Ј-and 3Ј-nucleotide sequences of their open reading frames were used as primers, as described previously (9). The PCR condition for sPLA 2 -IID, -IIE, -V, and -X was 94°C for 30 s and then 30 cycles of amplification at 94°C for 5 s and 68°C for 4 min, using the Advantage cDNA polymerase mix (Clontech). The PCR condition for sPLA 2 -IIF was 94°C for 30 s and then 30 cycles of amplification at 94°C for 30 s, 58°C for 30 s, and 72°C for 30s, using ExTaq polymerase (Takara Biomedicals). The PCR products were analyzed by 1% agarose gel electrophoresis with ethidium bromide.
Immunoblotting-Lysates from 2 ϫ 10 5 cultured cells or 25-g protein equivalents of tissue homogenates in cell lysis buffer (10 mM Tris-HCl (pH 7.4) containing 150 mM NaCl (TBS), 1 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, 2 g/ml leupeptin, and 2 g/ml pepstatin) were subjected to SDS-PAGE on 15% gels under reducing conditions with 2-mercaptoethanol. Protein concentrations in the samples were determined with a BCA protein assay kit (Pierce) with bovine serum albumin (Sigma) as a standard. As required for experiments, cell culture supernatants and cells, which were lysed in a volume of cell lysis buffer equivalent to the culture supernatants, were applied to the gels. The separated proteins were electroblotted onto nitrocellulose membranes (Schleicher & Schuell) with a semi-dry blotter (MilliBlot-SDE system; Millipore). After blocking with 5% (w/v) skim milk in TBS containing 0.05% (v/v) Tween 20 (TBS-T), the membranes were probed with rabbit antibody against human or mouse sPLA 2 -X at 1:5,000 dilution in TBS-T for 2 h, followed by incubation with horseradish peroxidase-conjugated anti-rabbit IgG (Zymed Laboratories Inc.) at 1:5,000 dilution in TBS-T for 2 h, and were visualized using the ECL Western blot system (PerkinElmer Life Sciences), as described (3).
Immunohistochemistry-Studies with human tissue sections were approved by the ethical committee of our universities with informed consent from the patients. Immunohistochemical staining of human tissue sections was performed as described previously (9,12). Briefly, the tissues embedded in paraffin were sectioned and mounted on slides, deparaffinized in xylene, and rehydrated in ethanol with increasing concentrations of water. The tissue sections (4 m thickness) were incubated with Target Retrieval Solution (Dako, Carpinteria, CA) as required, incubated for 10 min with 3% (v/v) H 2 O 2 , washed three times with TBS for 5 min each, incubated with 5% (v/v) skim milk in TBS for 30 min, washed three times with TBS for 5 min each, and incubated for 2 h with anti-sPLA 2 antibodies at 1:200 dilution in TBS. Then the sections were treated with a catalyzed signal-amplified system staining kit (Dako) with diaminobenzidine substrate. The cell type was identified from conventional hematoxylin and eosin staining of serial sections adjacent to the specimen used for immunohistochemistry.
Expression of sPLA 2 s by the Adenovirus System-Adenoviruses bearing individual sPLA 2 cDNAs were prepared with the ViraPower Adenovirus Expression System (Invitrogen), as described previously (9,10,12). Briefly, the full-length cDNAs for sPLA 2 s were subcloned into the pENTER/D-TOPO vector using the pENTER Directional TOPO cloning kit (Invitrogen). After purification of the plasmids from the transformed Top10 competent cells (Invitrogen), the cDNA inserts were transferred to the pAd/CMV/V5-DEST vector (Invitrogen) by means of the Gateway system using LR Clonase (Invitrogen). The plasmids were purified and digested with PacI (New England Biolabs). The linearized plasmids (1-2 g) were then mixed with 4 l of Lipofectamine 2000 (Invitrogen) in 200 l of Opti-MEM medium (Invitrogen) and transfected into subconfluent 293A cells (Invitrogen) in 1 ml of Opti-MEM in 6-well plates (Iwaki Glass). Then 293A cells were cultured for 1-2 weeks in RPMI 1640 medium containing 10% fetal calf serum, with replacement of the medium every 2 days. When most cells became detached from the plates, the cells and culture medium were harvested together, freezethawed twice, and centrifuged to obtain the adenovirus-enriched supernatants. Then aliquots of the supernatants were added to fresh 293A cells and cultured for 2-3 days to amplify adenoviruses. After 2-4 times of amplification, the resulting adenovirus-containing media were used as virus stocks. Viral titers were determined by the plaque-forming assay with 293A cells. As a control, the pAd/CMV/V5-GW/lacZ vector (Invitrogen) was digested with PacI and transfected into 293A cells to produce lacZ-bearing adenovirus. Aliquots of the adenovirus-containing medium was added to PC12 cells and cultured for appropriate periods with or without NGF for subsequent analyses.
Experiments with Small Interfering RNA (siRNA)-A set of synthetic hairpin-forming oligonucleotides directed to rat sPLA 2 -X (5Ј-CACCGA-TACCTCTTCTTCCCCTCGTTCAAGAGACGAGGGGAAGAAGAGGT-AT-3Ј (sense) and 5Ј-AAAAATACCTCTTCTTCCCCTCGTCTCTTGAA-CGAGGGGAAGAAGAGGTATC-3Ј (antisense)) and that directed to rat sPLA 2 -IIA (5Ј-CACCGGACAGGAAAGAGAGCTGATTTCAAGAGAAT-CAGCTCTCTTTCCTGTC-3Ј (sense) and 5Ј-AAAAGACAGGAAAGAG-AGCTGATTCTCTTGAAATCAGCTCTCTTTCCTGTCC-3Ј (antisense)) were prepared. After annealing, the oligonucleotides were subcloned into pENTER TM /U6 vector (Invitrogen) using BLOCK-iT TM U6 RNAi Entry Vector kit (Invitrogen). After transformation into Top10 competent cells and plasmid purification, the insert was transferred to the pAd/BLOCK-iT TM -DEST vector (Invitrogen) by means of the Gateway system using LR Clonase. After transformation into Top10 competent cells and plasmid purification, the plasmid was linearized with PacI and transfected into 293A cells with Lipofectamine 2000 to produce adenovirus bearing the sPLA 2 -X or SPLA 2 -IIA siRNA construct. After 2-4 times of amplification, the resulting adenovirus-containing media were used for subsequent analyses.
Measurement of sPLA 2 Activity-sPLA 2 activities in cell lysates and culture supernatants were assayed by measuring the amounts of radiolabeled linoleic acid released from the substrate 1-palmitoyl-2-[ 14 C]linoleoyl-phosphatidylethanolamine (Amersham Biosciences). For preparation of cell lysates, PC12 cells were harvested with cell scrapers, suspended in cell lysis buffer, and disrupted by sonication followed by two freeze-thawing cycles. The substrate in ethanol was dried up under N 2 stream and was dispersed in water by sonication. Each reaction mixture (total volume 250 l) consisted of appropriate amounts of the required samples, 100 mM Tris-HCl (pH 7.4), 4 mM CaCl 2 , and 2 M substrate. After incubation for 30 min at 37°C, [ 14 C]linoleic acid was extracted, and the radioactivity was quantified by liquid scintillation counter, as described previously (3).
Confocal Laser Microscopy-Cells grown in subconfluency on glass bottom dishes (Matsunami) pre-coated with 5 g/ml fibronectin (Sigma) were fixed with 3% paraformaldehyde in TBS for 1 h. After three washes with TBS, the fixed cells were treated with 1% (w/v) bovine serum albumin and 0.5% (w/v) saponin in TBS (blocking solution) for 1 h, with rabbit anti-sPLA 2 antibodies or goat anti-GM130 antibody (BD Biosciences) at 1:200 dilution in blocking solution for 2 h, and then with species-matched fluorescein isothiocyanate-conjugated or Cy3-conjugated second antibody at 1:200 dilution in blocking solution for 2 h, with three washes in each interval. After six washes with TBS, specific immunofluorescent signals were visualized with a laser-scanning confocal microscope (IX70; Olympus).
TLC-PC12 cells (10 7 cells) infected for 3 days with mock-or sPLA 2containing adenovirus in the presence of 0.2 ng/ml NGF were incubated for an additional 12 h with [ 14 C]choline (0.5 Ci/ml). The cells were then washed with TBS and harvested, and phospholipids were extracted by the method of Bligh and Dyer (17). Individual phospholipids were separated on silica gel TLC plates (Sigma) with a solvent system of chloroform/methanol/acetic acid (65:25:10, v/v) (18). After exposure to iodine vapor, visible spots were identified by comparing their positions with those of standard phospholipids. The plates were then exposed to Kodak BioMax XAR films to visualize the incorporation of radioactivity into choline-containing phospholipids.
Electrospray Ionization/Mass Spectrometry (ESI/MS)-MS spectra were obtained on a Quattro Micro tandem quadrupole mass spectrometer (Micromass) equipped with an ESI, as described previously (19). Lipid extracts obtained from PC12 cells (100-g protein equivalents) were reconstituted in 2:1 chloroform/methanol (100 -300 mol phosphorus/liter), and 2 l of the sample was injected per run. As internal standards, 100 pmol of PC (C14:0 -C14:1, diacyl) and LPC (C12:0) were added to the samples. The samples were introduced by means of a flow injector into the ESI chamber at a flow rate of 4 l/min in a solvent system of acetonitrile/methanol/water (2:3:1, v/v) containing 0.1% (v/v) ammonium formate (pH 6.4). The mass spectrometer was operated in the positive and negative scan modes. The flow rate of the nitrogen drying gas was 12 liters/min at 80°C. The capillary and cone voltages were set at 3.7 kV and 30 V, respectively; argon at 3-4 ϫ 10 4 torr was used as the collision gas, and a collision energy of 30 -40 V was used for obtaining fragment ions for precursor ions.

sPLA 2 -X Is Expressed in Peripheral Neuronal Fibers in Var-
ious Tissues-While our immunohistochemical analyses of sPLA 2 enzymes in human pathologic tissues were underway (9, 10, 12), we found that small round slices of the peripheral neuronal fibers in various tissues, including the stomach (Fig.  1A, a), lung (b), seminal vesicles (c and d), prostate (e), and uterus (f), showed intense immunoreactivity for sPLA 2 -X. On the other hand, no appreciable staining for sPLA 2 -IIA ( Fig. 1B,  a), -IID (b), -V (c), -IIE, and -IIF (data not shown) was observed in the neuronal fibers in the stomach as well as in other tissues (data not shown). In the lung, sPLA 2 -X staining was also found in alveolar epithelial cells (Fig. 1A, b), as reported recently (5,9). Staining for sPLA 2 -X in human cerebrum neurons was rather weak (data not shown), suggesting that its expression is generally restricted to neuronal fibers in peripheral tissues.
To assess further the expression of sPLA 2 -X in peripheral neurons, DRG explants isolated from fetal C57/BL6 mice (E12.5-13) were subjected to RT-PCR for various mouse sPLA 2 s. As shown in Fig. 2A, the transcript for sPLA 2 -X and, to a lesser extent, for sPLA 2 -IIE, but not for sPLA 2 -IID, -IIF, or -V (sPLA 2 -IIA is intrinsically deficient in C57BL/6 mice (20)), was detected by RT-PCR in DRG neurons. Immunoblotting revealed the presence of sPLA 2 -X protein in the DRG neurons, where two bands (18 and 23 kDa) became apparent (Fig. 2B). As described below (see Fig. 9), the lower and upper immuno- reactive bands appeared to correspond to the proenzyme of sPLA 2 -X and its N-glycosylated form, respectively.
After DRG explants were cultured with 20 ng/ml NGF for 24 h on laminin-coated plates, there was a marked extension of neuronal fibers, as described previously (16). When these cells were fixed and subjected to immunocytostaining with antibodies specific for individual sPLA 2 s, sPLA 2 -X (Fig. 2C, e and f) but not sPLA 2 -IIA (a), -IID (b), -IIE (c), -V (d), or -IIF (data not shown) was positively stained in most neurons, in which cell bodies as well as neuronal filaments yielded significant sPLA 2 -X immunoreactivity. In some neuronal filaments, staining of sPLA 2 -X in the growth cones was evident (Fig. 2C, f). Signal for sPLA 2 -X in the cell bodies partially, if not completely, overlapped with that for GM130, a Golgi marker (Fig.  2D), indicating that sPLA 2 -X accumulates in the Golgi prior to secretion. sPLA 2 -X Induces Neurite Outgrowth in PC12 Cells-In an attempt to identify some functional aspects of sPLA 2 -X in peripheral neuronal cells, we took advantage of PC12 cells, which differentiate into sympathetic neuron-like cells following culture with 100 ng/ml NGF (14), as shown in Fig. 3A. Northern blot analysis showed the expression of endogenous sPLA 2 -X mRNA, which gradually increased over 24 -72 h of culture with NGF (Fig. 3B). When cell lysates were subjected to immunoblotting with anti-sPLA 2 -X antibody, we detected two immunoreactive bands with molecular masses of 18 and 23 kDa with or without culture for 3 days with NGF, with the former band being increased modestly following NGF treatment (Fig. 3C). The PLA 2 activity in the culture supernatants was gradually elevated over 3 days of culture with NGF (Fig. 3D). A subtle difference between RNA blotting (Fig. 3B) and immunoblotting ( Fig. 3C) (i.e. mRNA showed more induction than protein) might be due to the different stability of sPLA 2 -X mRNA and protein in PC12 cells, as has been suggested for sPLA 2 s in other cell types (13). Nonetheless, these results suggest that trace levels of sPLA 2 -X are endogenously expressed in PC12 cells.
Next, we investigated the impact of sPLA 2 -X on cellular functions by using the adenovirus method to transfect the enzyme into undifferentiated PC12 cells. Effective expression of the transfected sPLA 2 -X was verified by immunoblotting; lysates of cells infected with adenovirus for sPLA 2 -X provided intense 18-and 23-kDa bands, which were exactly the same sizes as those detected faintly in cells infected with control adenovirus (endogenous sPLA 2 -X; also see Fig. 3C) (Fig. 4A). At a high dose of sPLA 2 -X adenovirus (m.o.i. ϭ 10), the amounts of active sPLA 2 -X protein released from the cells into the culture supernatants were estimated to be 5-10 ng/ml, as judged by the secreted enzymatic activity in comparison with the activity of authentic pure sPLA 2 -X protein diluted in culture medium (data not shown). This value was about 25-50 times higher than that with mock-transfected cells. Cell proliferation and viability were unaffected by overexpression of sPLA 2 -X throughout the experimental periods.
Immunocytostaining with anti-sPLA 2 -X antibody revealed that, although the signal for endogenous sPLA 2 -X was rather faint under the present fluorescent conditions (Fig. 4B, top  panel), sPLA 2 -X signal was markedly increased in accordance with increasing doses of sPLA 2 -X adenovirus (Fig. 4B, middle and bottom panels). Of interest, we found that several cells displaying high expression of sPLA 2 -X began to extend neurites (Fig. 4B). As revealed by higher magnification, only the cells showing intense sPLA 2 -X immunoreactivity extended neurites, particularly in the perinuclear region, whereas adjacent cells without appreciable neurite outgrowth did not exhibit perinuclear staining for sPLA 2 -X (Fig. 4C). The perinuclear sPLA 2 -X signal was colocalized with that of GM130 (Fig. 4D), implying that the main localization of this enzyme is in the Golgi apparatus. Weak sPLA 2 -X signal diffused in the cytoplasm with a trend to a reticular pattern may reflect its localization in the ER (Fig. 4, C and D).
We found that the neurite-extending effect of sPLA 2 -X was more obvious when PC12 cells were cultured in the presence of a suboptimal concentration (0.2 ng/ml) of NGF (Fig. 5A). Again, marked extension of the neurites was preferentially observed in cells that exhibited intense sPLA 2 -X signal in the perinuclear Golgi (i.e. cells into which sPLA 2 -X adenovirus had been efficiently infected) (Fig. 5, B and C). Although weak staining for sPLA 2 -X was also found on the plasma membrane but not in the Golgi of neighboring cells, an event that may be a reflection of the paracrine transfer of sPLA 2 -X from sPLA 2 -X-transfected cells, enhanced neurite outgrowth was sparse in these neighboring cells (Fig. 5, B and C). These results suggest that . The Golgi (white arrowheads) as well as the growth cones of elongating neurites (yellow arrowheads) showed sPLA 2 -X immunoreactivity. Magnification: ϫ400. C, double immunostaining with anti-sPLA 2 -X (green) and anti-GM130 (red) antibodies and their merged image (yellow) are shown. Localization of sPLA 2 -X in the Golgi and growth cones of neurites of sPLA 2 -X-expressing cells was obvious. In addition, the plasma membranes (red arrowheads) of adjacent cells, in which sPLA 2 -X signal was absent from the Golgi, were faintly stained for sPLA 2 -X. Magnification: ϫ400. D, immunostaining of sPLA 2 -X in PC12 cells infected for 3 days with sPLA 2 -X adenovirus in the presence of 100 ng/ml NGF. The Golgi and growth cones of multiple elongating neurites showed intense sPLA 2 -X immunoreactivity (a and b). When the growth cone was magnified, sPLA 2 -X signal was preferentially localized at the leading edge (yellow arrow) (c). Magnification: (a) ϫ400; (b) ϫ200; (c) ϫ800. sPLA 2 -X promotes neuritogenesis preferentially in cells that express this enzyme. It is notable that significant sPLA 2 -X signal was also detected in the growth cones of elongating neurites in cells cultured with suboptimal NGF (Fig. 5, B and C). This unique localization was more evident when the cells infected with sPLA 2 -X adenovirus were cultured with 100 ng/ml NGF, where intense sPLA 2 -X signal was located in the Golgi as well as in the growth cones of multiple elongating neurites (Fig. 5D, a and b). Within the growth cones, intense signal for sPLA 2 -X was preferentially distributed at the leading edges (Fig. 5D, c).
The neurite-extending effect of sPLA 2 -X required its catalytic activity, because no neurite outgrowth was induced when the catalytically inactive sPLA 2 -X mutant G30S (a mutation in the Ca 2ϩ -binding loop (3)) was adenovirally transfected into PC12 cells (Fig. 6, A and B). Furthermore, neurite extension was sparse in cells infected with adenovirus for sPLA 2 -IIA (Fig.  6, A and B), even in culture with suboptimal NGF (data not shown). Adenoviral transfection of sPLA 2 -V showed a trend to modest neurite extension, yet this effect was rather weaker than that of sPLA 2 -X (Fig. 6, A and B). Other sPLA 2 s, including sPLA 2 -IID, -IIE, and -IIF, failed to induce neurite outgrowth (data not shown). Thus, among mammalian sPLA 2 s tested so far, the neuritogenic action is fairly specific for sPLA 2 -X. sPLA 2 -X Promotes LPC Production in PC12 Cells-The observation that the enzymatic activity of sPLA 2 -X is required for neurite outgrowth (Fig. 6) suggests that some catalytic product(s) that are preferentially generated by sPLA 2 -X in comparison with other sPLA 2 s mediates the neuritogenic effect of this enzyme. To address this possibility, we examined the effects of various lysophospholipids, fatty acids, and their derivatives on the neuritogenesis of PC12 cells. We found that 1-palmitoyl-LPC alone had the ability to induce neurite extension at 50 M (Fig. 7, A and B), whereas other lysophospholipids, including lysophosphatidic acid, lysophosphatidylinositol, lysophosphatidylethanolamine, and lysophosphatidylserine with sn-1 palmitate, fatty acids such as AA and oleic acid, and PAF failed to do so (Fig. 7B). Of the LPC molecular species, those bearing long sn-1 fatty acyl chains, including myristate (C14:0), palmitate (C16:0), and stearate (C18:0), were equally active in inducing neurite outgrowth (Fig. 7B, inset).
Assuming that the neuritogenic effect of sPLA 2 -X would be mediated by the production of LPC, we aimed to assess whether sPLA 2 -X could indeed generate LPC species (those having myristate, palmitate, and/or stearate in particular) in PC12 cell membranes. To this end, we chose a culture condition employing suboptimal NGF, under which the neuritogenic effect of sPLA 2 -X was particularly obvious (see above). When phospholipids extracted from PC12 cells that had been infected for 3 days with adenovirus for sPLA 2 -IIA, -V, or -X or control adenovirus were separated on TLC, a marked increase in LPC was observed in cells expressing sPLA 2 -X as compared with control cells (Fig. 8A). In contrast, accumulation of LPC in sPLA 2 -IIA-transfected cells was minimal, and there was only a modest increase in LPC in sPLA 2 -V-transfected cells (Fig. 8A). Generation of LPC by the three sPLA 2 s observed here seems to agree with their PC hydrolytic capacity in the in vitro PLA 2 assay (3, 5, 6). We further evaluated LPC production by sPLA 2 -X in PC12 cells by means of the ESI/MS technique. As shown in Fig. 8B, peaks of PC subclasses with C34:1 (16:0 -18:1), C36:2 (18:0 -18:2), and C38:4 (18:0 -20:4) were significantly reduced in sPLA 2 -X-transfected cells relative to those in control cells. Concomitantly, LPC subclasses with C16:0 and C18:0, which could induce neuritogenesis (Fig. 7B, inset), were increased in sPLA 2 -X-transfected cells compared with those in control cells (Fig.  8B). Comparing the peak areas, ϳ15-20% of total PC was converted to LPC by sPLA 2 -X in this experimental setting. There were no differences in other PC and LPC species, or in sphingomyelin, between control and sPLA 2 -X-expressing cells (Fig. 8B). sPLA 2 -X Undergoes Proteolytic Processing Mainly after Secretion-When culture supernatants and lysates of sPLA 2 -X adenovirus-infected PC12 cells were subjected to the PLA 2 enzyme assay, we unexpectedly found that the majority of the activity was present in the supernatants, and the activity in the lysates was rather low (Fig. 9A), despite the fact that intense sPLA 2 -X immunoreactivity was observed in the Golgi of the cells (Figs. 4 and 5). Immunoblotting with anti-sPLA 2 -X antibody revealed that both fractions contained more than two immunoreactive bands (Fig. 9B). Thus, 18-and 23-kDa bands appeared in the lysates (as noted above), whereas additional 16-kDa and broad 22-28-kDa bands were detected in the supernatants. Considering that sPLA 2 -X has been reported to be activated by the proteolytic cleavage of the N-terminal propeptide (5), these results may indicate that the 16-kDa band in the supernatants corresponds to the N-terminally cleaved mature form of sPLA 2 -X, and the 18-kDa band in both the supernatants and cells represents the proform (pro-X). Moreover, when the cells were cultured in the presence of tunicamycin, an N-glycosylation inhibitor, the upper bands in the supernatants and lysates were decreased in a dose-dependent manner, with concomitant accumulation of lower bands (pro-X and mature sPLA 2 -X) in both fractions (Fig. 9B), implying that the upper bands detected in the supernatant and lysate fractions repre-sent the N-glycosylated forms of sPLA 2 -X. As shown in Fig. 9C, PLA 2 activity in the supernatants was not altered following tunicamycin treatment, indicating that the N-glycosylation does not affect the enzymatic activity of sPLA 2 -X.
The above result argues that the proteolytic conversion of pro-X to mature sPLA 2 -X occurs mainly after secretion from PC12 cells. To assess whether this is a general or cell typespecific event, we also infected sPLA 2 -X adenovirus into NHPF (Fig. 9D) and BEAS-2B (Fig. 9E), cell types that intrinsically express sPLA 2 -X (9). In these cells, sPLA 2 -X is mainly localized in the Golgi and ER, as reported (9). Again, sPLA 2 -X appeared as multiple bands with distinct sizes in supernatants and cell lysates, where the 16-kDa band for mature sPLA 2 -X was detected only in the supernatants, and the 18-kDa band for pro-X was mainly detected in the cell lysates (Fig. 9, D and E). Moreover, tunicamycin treatment of NHPF resulted in a dosedependent disappearance of the upper, 22-28-kDa broad band from the supernatants as well as of the 23-kDa band from the lysates (Fig. 9D), indicating that these larger species represent N-glycosylated sPLA 2 -X. Because of the appearance of a broader range of molecular masses for sPLA 2 -X in the supernatants of PC12 (Fig. 9B), NHPF (Fig. 9D) and BEAS-2B (Fig.  9E), the N-glycosylation chain of mature sPLA 2 -X may have a more heterogeneous structure than that of pro-X residing in the cells. Unlike in PC12 cells (Fig. 9C), however, the amounts of the 16-kDa nonglycosylated form of mature sPLA 2 -X in the supernatants of NHPF were decreased after treatment with increasing concentrations of tunicamycin (Fig. 9D). Accordingly, PLA 2 activity in the supernatants of NHPF was also decreased after treatment with increasing concentrations of tunicamycin (Fig. 9C). These results suggest that the N-glycosylation of sPLA 2 -X facilitates its secretion in a cell type-specific manner.
Participation of Endogenous sPLA 2 -X in Neuritogenesis-Although the functional studies of sPLA 2 -X by the overexpression strategy revealed its potential role in neurite outgrowth of PC12 cells as shown above, it still remained unclear whether endogenous sPLA 2 -X could participate in this event. Given the results shown in Fig. 9, we reasoned that the neuritogenic effect of sPLA 2 -X in PC12 cells would occur through its action on the plasma membrane after secretion and, if so, that neutralization of the extracellular sPLA 2 -X by anti-sPLA 2 -X antibody would prevent the neuritogenic function of this enzyme. Thus, we incubated PC12 cells with 100 ng/ml NGF in the presence or absence of anti-sPLA 2 -X antibody at a concentration that could completely neutralize the enzymatic activity of sPLA 2 -X in the culture supernatants. Under the condition where NGF elicited marked neurite outgrowth of PC12 cells, further addition of anti-sPLA 2 -X antibody to the culture resulted in slight, but significant, reduction of the length of the neurites without affecting the total neurite number (Fig. 10, A  and B). In contrast, addition of anti-sPLA 2 -IIA antibody, which was used as a negative control, failed to influence NGF-induced neurite outgrowth (Fig. 10, A and B). These results suggest that endogenous sPLA 2 -X contributes, even if partially, to elongation of the NGF-induced neurites under the present experimental conditions.
We further assessed the contribution of endogenous sPLA 2 -X to neuritogenesis by means of the siRNA technique. When adenovirus harboring sPLA 2 -X-directed siRNA was infected into PC12 cells, the expression of sPLA 2 -X protein was considerably decreased as compared with that in cells infected with control adenovirus for lacZ or sPLA 2 -IIA-directed siRNA (Fig.   11A). Under this condition, NGF-induced neurite outgrowth was partially but significantly reduced in cells adenovirally transfected with sPLA 2 -X-directed siRNA relative to that in cells transfected with lacZ or sPLA 2 -IIA-directed siRNA (Fig.  11, B and C), results being essentially consistent with the antibody neutralization experiments (Fig. 10). DISCUSSION Current evidence suggests that group IVA cPLA 2 ␣ plays a central role in lipid mediator production, as shown by studies using cPLA 2 ␣-deficient mice as well as by many cell biological studies (21)(22)(23). On the other hand, the contribution of other PLA 2 enzymes, sPLA 2 s in particular, to lipid mediator production under pathophysiological conditions has remained controversial, despite the facts that several sPLA 2 s can augment AA release in cell culture studies (3-6, 24 -27) and that pharmacological inhibition of sPLA 2 s can often attenuate inflammation in experimental animals (28 -30). A recent gene targeting study of sPLA 2 -V has provided unequivocal evidence for the role of this enzyme in eicosanoid production in vivo (31). Nonetheless, considering that the expression of individual sPLA 2 enzymes is tissue-and cell-specific (32,33), we aimed to determine specific cell types that intrinsically express particular sPLA 2 enzymes in order to approach their potential cell typerelated functions. In the present study, we have provided evidence that sPLA 2 -X is uniquely expressed in peripheral neurons. Furthermore, we have shown that sPLA 2 -X is capable of inducing neuritogenesis through the production of LPC. Thus, this study delineates an unexplored, lysophospholipid-dependent, neurotrophin-like action of a particular sPLA 2 isozyme that is intrinsically expressed in the nervous system. Immunohistochemistry of various human tissues revealed the localization of sPLA 2 -X but not of several other sPLA 2 s in peripheral neuronal fibers (Fig. 1). Expression of sPLA 2 -X in peripheral neurons is further supported by detection of its protein and mRNA in mouse DRG neurons (Fig. 2) and in rat pheochromocytoma PC12 cells undergoing NGF-induced neuronal differentiation (Fig. 3). We further found that sPLA 2 -X is capable of inducing neurite outgrowth in PC12 cells . Although the neuronal actions of some venom sPLA 2 s have been reported to be mediated by their binding to receptorlike molecules but not by their catalytic activity (34 -38), several lines of evidence support the idea that the neuritogenic action of sPLA 2 -X observed in this study requires its catalytic activity to hydrolyze PC and thereby to produce LPC.
First, under conditions where sPLA 2 -X exhibited a neuritogenic effect, neither the catalytically inactive sPLA 2 -X mutant G30S nor other sPLA 2 s (such as sPLA 2 -IIA), which exhibit poor PC-hydrolytic activity, induced neurite extension (Fig. 6). Indeed, of the mammalian sPLA 2 enzymes tested so far, sPLA 2 -X has the most potent ability to release fatty acids, including AA, from intact cellular membranes, and this action depends on its high capacity to bind and hydrolyze PC, which is enriched in the outer leaflet of the plasma membrane (3)(4)(5)(6). In the experimental setting employed here, sPLA 2 -V, another sPLA 2 that has the ability to hydrolyze PC (even though less efficiently than does sPLA 2 -X) (6, 24 -26, 39), exhibited only modest neuritogenic effect. Although neurite outgrowth was more obvious following adenoviral expression of excess sPLA 2 -V (data not shown), the importance of this observation is uncertain because sPLA 2 -V is scarcely expressed in peripheral neurons (Figs. 1  and 2). Second, of the PLA 2 reaction products tested so far, only LPC species with long sn-1 acyl chains (C14:0, 16:0, and 18:0), but not other lysophospholipids, fatty acids, or PAF, induced neurite outgrowth (Fig. 7). Although some AA metabolites have been reported to exert neuronal functions (40), the observation that cyclooxygenase and lipoxygenase inhibitors do not prevent neurite outgrowth of PC12 cells (41) argues against the contribution of AA metabolites to this event. Third, TLC and ESI/MS analyses demonstrated that sPLA 2 -X produces LPC with C16:0 and 18:0 in PC12 cells (Fig. 8). In fact, the rank order of neuritogenic potency of several sPLA 2 s (X Ͼ Ͼ V Ͼ IIA; Fig. 6B) appears to parallel that of LPC accumulation by these enzymes (Fig. 8A). Finally, in support of the present results, our collaborators have shown recently that exogenous bee venom sPLA 2 (group III) (42) and fungal sPLA 2 (group XIV) (41) induce neurite outgrowth in PC12 cells in a manner dependent upon their catalytic activity.
Accumulating evidence suggests that LPC is involved in a variety of cellular responses (43), and LPC-induced neuritogenesis in PC12 cells may involve L-type Ca 2ϩ channels (41). Increased intracellular Ca 2ϩ concentration via L-type Ca 2ϩ channels leads to activation of protein kinases C and mitogenactivated protein kinases. Ca 2ϩ influx through the voltagegated channels also triggers a variety of cellular events leading to neurite outgrowth via the concerted actions of various Ca 2ϩbinding proteins (44) and cell adhesion molecules (45,46). The finding that sPLA 2 -X and suboptimal NGF act in synergy in inducing neurite outgrowth (Fig. 5) may indicate that LPCinduced Ca 2ϩ signaling and NGF receptor (TrkA)-mediated tyrosine kinase signaling converge on activation of mitogenactivated protein kinases, an event that is crucial for neurite extension (43,(47)(48)(49). Indeed, neuritogenesis induced by fungal sPLA 2 is blocked by an inhibitor of the mitogen-activated protein kinase pathway (50). It remains unclear whether the G protein-coupled receptors G2A and GPR4, of which LPC has been proposed to be a ligand (51)(52)(53), could be involved in the neuritogenic action of LPC. Because G2A is expressed in PC12 cells, 2 it would be interesting to examine whether this receptor is involved in sPLA 2 -X-or LPC-induced neurite outgrowth.
It is notable that sPLA 2 -X is preferentially distributed in the growth cones of neuronally differentiated PC12 cells (Fig. 5). This fact, together with the finding that the neutralization of sPLA 2 -X by anti-sPLA 2 -X antibody or knockdown of its expression by siRNA resulted in substantial reduction of neurite length (Figs. 10 and 11), raises the intriguing possibility that sPLA 2 -X-promoted LPC production and resultant LPC signaling may occur spatiotemporally at the growth cones during neuritogenesis. Because LPC binds to serum albumin and because serum in medium already contains substantial amounts of albumin-bound LPC (43,55), it was difficult to evaluate precisely the actual working concentration of LPC in our experimental system. Nevertheless, considering that a substantial level of cellular PC was metabolized to LPC in sPLA 2 -Xtransfected cells (Fig. 8B), the local concentration of LPC produced by sPLA 2 -X in the putative microdomains may be sufficient to facilitate neuritogenesis.
In PC12 cells as well as in lung fibroblasts and bronchial 2 M. Arioka, unpublished observations.
FIG. 11. Effect of siRNA for sPLA 2 -X on NGF-induced neuritogenesis of PC12 cells. A, expression of sPLA 2 -X protein in PC12 cells infected with adenovirus for sPLA 2 -X siRNA (si-X), sPLA 2 -IIA siRNA (si-IIA), or lacZ, as assessed by immunoblotting. B, morphology of PC12 cells infected with adenovirus for sPLA 2 -X siRNA, sPLA 2 -IIA siRNA, or lacZ after treatment for 3 days with 100 ng/ml NGF. Magnification: ϫ100. C, under the culture conditions shown in B, neurite length (left panel) and number (right panel) of all the cells in 3-4 randomly selected photographs were evaluated. *, p Ͻ 0.05 versus lacZ and sPLA 2 -IIA siRNA. For evaluation of neurite length, the average length of neurites in cells treated with NGF plus lacZ adenovirus was considered as 1. epithelial cells, sPLA 2 -X is mainly located in the Golgi as a zymogen and is converted to an active mature enzyme after (or just prior to) secretion (Fig. 9). Although the precise mechanism for proteolytic conversion of sPLA 2 -X is under investigation, certain protease(s) present in serum, secreted from cells, or on cell surfaces might participate in this process. In this regard, sPLA 2 -IB, another sPLA 2 that possesses an N-terminal propeptide, can be converted to its active form by serum plasmin (54). Given that the mature sPLA 2 -X exists predominantly in the culture supernatant (Fig. 9), it is conceivable that PC hydrolysis and attendant LPC production by sPLA 2 -X may take place on the plasma membrane surface of PC12 cells (the external plasma membrane pathway). The observations that the addition of cell-impermeable anti-sPLA 2 -X antibody to the culture resulted in partial reduction of neuritogenesis to a level similar to that achieved by treatment with sPLA 2 -X siRNA (Figs. 10 and 11) and that excess sPLA 2 -X added exogenously elicits neurite outgrowth (41) lend support for this notion. Because neurite outgrowth was evident in sPLA 2 -X-expressing cells but not appreciably in neighboring cells (indicative of autocrine action) (Figs. 4 and 5), LPC production by this enzyme may occur immediately after secretion, at which time the local concentration of active sPLA 2 -X may reach a level high enough to promote the hydrolysis of PC in the plasma membrane of sPLA 2 -X-releasing cells. The concentration of sPLA 2 -X dispersed into the culture medium (Ͻ10 ng/ml even when maximally overexpressed in this study) may be insufficient to induce neuritogenesis in neighboring cells. It is also possible that the membrane microdomains in the elongating growth cones, from which sPLA 2 -X is supposed to be released, may be more susceptible to this enzyme. However, the possibility that a low level of intracellular mature sPLA 2 -X could function prior to secretion, a pathway that has been proposed recently (4), cannot be ruled out. In this case, a trace level of mature sPLA 2 -X might be enriched in the growth cones and acts in situ to give rise to the neuritogenic LPC.
Taken together, we have shown that sPLA 2 -X is expressed in peripheral neurons and exhibits a neuritogenic effect. Besides this neurotrophic action, sPLA 2 s have been implicated in neurotoxicity, particularly in neurons isolated from the central nervous system (35-37, 56 -58). For instance, sPLA 2 -IIA is induced in rat brain undergoing ischemic damage (56), and this enzyme can promote apoptotic death of primary rat cortical neurons (57). sPLA 2 -IB also promotes neuronal cell death, an event that depends on its binding to the M-type sPLA 2 receptor (38). Bee venom sPLA 2 , in synergy with glutamate, exerts a neurotoxic effect through the N-type sPLA 2 receptor (36,58). In addition, sPLA 2 -IIA enhances neurotransmitter release from PC12 cells and cultured rat hippocampal neurons (14,59). Further studies will be needed to gain a more detailed understanding of the mode of action and of the full complement of physiological roles of individual sPLA 2 s in neurons as well as in other mammalian cells. Studies using sPLA 2 isozyme-specific inhibitors or gene-manipulated mice may help to clarify the physiological relevance of these observations.