Prostaglandin E2 Amplifies Cytosolic Phospholipase A2- and Cyclooxygenase-2-dependent Delayed Prostaglandin E2 Generation in Mouse Osteoblastic Cells
ENHANCEMENT BY SECRETORY PHOSPHOLIPASE A2*

(Received for publication, May 12, 1997, and in revised form, June 3, 1997)

Makoto Murakami , Hiroshi Kuwata , Yoshihisa Amakasu , Satoko Shimbara , Yoshihito Nakatani , Gen-ichi Atsumi and Ichiro Kudo Dagger

From the Department of Health Chemistry, School of Pharmaceutical Sciences, Showa University, Shinagawa-ku, Tokyo 142, Japan

ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES


ABSTRACT

We used the MC3T3-E1 cell line, which originates from C57BL/6J mouse that is genetically type IIA secretory phospholipase A2 (sPLA2)-deficient, to reveal the type IIA sPLA2-independent route of the prostanglandin (PG) biosynthetic pathway. Kinetic and pharmacological studies showed that delayed PGE2 generation by this cell line in response to interleukin (IL)-1beta and tumor necrosis factor alpha  (TNFalpha ) was dependent upon cytosolic phospholipase A2 (cPLA2) and cyclooxygenase (COX)-2. Expression of these two enzymes was reduced by cPLA2 or COX-2 inhibitors and restored by adding exogenous arachidonic acid or PGE2, indicating that PGE2 produced by these cells acted as an autocrine amplifier of delayed PGE2 generation through enhanced cPLA2 and COX-2 expression. Exogenous addition or enforced expression of type IIA sPLA2 significantly increased IL-1beta /TNFalpha -initiated PGE2 generation, which was accompanied by increased expression of both cPLA2 and COX-2 and suppressed by inhibitors of these enzymes. Thus, our results revealed a particular cross-talk between the two PLA2 enzymes and COX-2 for delayed PGE2 biosynthesis by a type IIA sPLA2-deficient cell line. cPLA2 is responsible for initiating COX-2-dependent delayed PGE2 generation, and sPLA2, if introduced, enhances PGE2 generation by increasing cPLA2 and COX-2 expression via endogenous PGE2.


INTRODUCTION

In contrast to immediate prostaglandin (PG)1 generation that is regulated by post-translational activation of the constitutively expressed biosynthetic enzymes, delayed PG generation, which occurs several hours after stimulation with cytokines and growth factors, generally requires de novo protein synthesis. Discovery of the inducible isoform of cyclooxygenase (COX), COX-2, has provided new insight into the mechanism that regulates delayed PG generation (1). Thus, pharmacological and genetic studies have revealed that the inducible COX-2, rather than the constitutive COX-1, is the dominant enzyme involved in the delayed, prolonged phase of PG biosynthesis (2-4). Despite an apparent functional segregation of the two COX isoforms, however, whether the phospholipase A2 (PLA2) enzymes, which lie upstream of the COXs in the PG-generating pathway, are functionally segregated is still rather controversial.

The properties of type IV cytosolic PLA2 (cPLA2) support its crucial role in the burst release of arachidonic acid in the immediate response following receptor ligation (5, 6). Several lines of evidence have shown that an increase in cytosolic Ca2+ levels is a prerequisite for cPLA2 activation, demonstrating its dissociation from the delayed response that is not accompanied by intracellular Ca2+ mobilization (7, 8). Conversely, others have demonstrated the involvement of cPLA2 in delayed PG generation (9, 10), although the molecular mechanisms leading to cPLA2 activation under conditions with no Ca2+ signaling are poorly understood. Type IIA secretory PLA2 (sPLA2), one of the 14-kDa PLA2 enzymes, is reported to contribute to the arachidonic acid metabolism under certain conditions (11). It is released from the secretory granules of inflammatory cells (12), and its production by a wide variety of cells is often induced by proinflammatory stimuli (13-15). The involvement of type IIA sPLA2 in delayed PG generation by certain cell types has been demonstrated by pharmacological, immunochemical, and molecular biological studies (14-16). In vivo studies showed that various sPLA2 inhibitors attenuated inflammation partially (17), type IIA sPLA2 exacerbated edema when injected into inflamed tissues (18), and type IIA sPLA2 transgenic mice displayed more severe inflammatory responses than normal mice when subjected to proinflammatory stimuli (19), providing support for an augmentative, rather than initiatory, role of type IIA sPLA2 in biological responses. Our earlier studies revealed that type IIA sPLA2 association with proteoglycans on plasma membranes is important for this sPLA2 to exert its actions (20). More recently, a novel type V sPLA2 has been shown to be important for stimulus-initiated arachidonic acid release (21, 22). This finding reveals the redundancy of sPLA2 family members and may account for why some inbred mouse strains, in which the type IIA sPLA2 gene is naturally disrupted due to an extra thymidine insertion into exon 3, leading to a frameshift mutation that produces a severely truncated non-functional type IIA sPLA2 protein (23, 24), remain normal in their whole life.

In an attempt to resolve the functional cross-talk between PLA2 and COX enzymes, we have taken advantage of a cell line derived from type IIA sPLA2-deficient C57BL/6J mouse (23, 24). Here we show that in C57BL/6J-derived MC3T3-E1 cells, cPLA2 is functionally linked to COX-2 for cytokine-induced delayed PGE2 generation, and type IIA sPLA2 augments it when enforcedly introduced. Furthermore, we have obtained evidence that PGE2 acts as an autocrine amplifier of these responses by increasing the expression of both cPLA2 and COX-2, thereby revealing a positive feedback loop of the PG biosynthetic route in this particular cell line.


EXPERIMENTAL PROCEDURES

Materials

Antibodies and cDNAs for cPLA2, type IIA sPLA2, COX-1 and COX-2, and the COX inhibitors used are described elsewhere (20). Arachidonic acid, PGE2, and the PGE2 enzyme immunoassay kit were purchased from Cayman Chemical. Arachidonoyl trifluoromethyl ketone (AACOCF3) was purchased from Calbiochem.

Activation of MC3T3-E1 Cells

Mouse osteoblastic MC3T3-E1 cells (Riken Cell Bank) were maintained in alpha -MEM medium (Dainippon Pharmaceutical) supplemented with 10% (v/v) fetal calf serum, 100 units/ml penicillin, 100 µg/ml streptomycin, and 2 mM glutamine. MC3T3-E1 (5 × 104 cells/ml) in 500 µl or 2 ml of culture medium were seeded onto 24- or 6-well plates, respectively, and after 3 days when the cells had reached confluence, 500 µl or 2 ml of fresh culture medium, respectively, with or without 5 ng/ml mouse interleukin (IL)-1beta (Genzyme) and/or 1,000 units/ml mouse tumor necrosis factor alpha  (TNFalpha ; Genzyme), was added to each well to activate the cells. The supernatants were taken for PGE2 enzyme immunoassay, and for immunoblot analysis, the cells were trypsinized, washed once with 10 mM phosphate buffer, pH 7.4, containing 150 mM NaCl (phosphate-buffered saline (PBS)), and resuspended in cell lysis buffer (20) to produce 1 × 107 cells/ml. As described below for the RNA blot analysis, TRIzol (Life Technologies) was added directly to the cell monolayers grown in 6-well plates.

RNA Blotting

Equal amounts (~10 µg) of total RNA, purified using TRIzol reagent, were applied to each lane of 1.2% (w/v) formaldehyde-agarose gels, electrophoresed, and transferred to Immobilon-N membranes (Millipore). The resulting blots were then sequentially probed with cPLA2, COX-2, COX-1, and beta -actin cDNA probes that were labeled with [32P]dCTP (Amersham Life Science, Inc.) by random priming (Takara Biomedicals Inc.). All hybridizations were carried out under high stringency condition as described previously (20).

SDS-Polyacrylamide Gel Electrophoresis/Immunoblotting

Cell lysates were applied to 10% (w/v) SDS-polyacrylamide gels and electrophoresed under reducing condition. The separated proteins were electroblotted onto nitrocellulose membranes (Schleicher & Schuell) using a semi-dry blotter (MilliBlot-SDE system; Millipore), according to the manufacturer instructions. The membranes were probed with antibodies against cPLA2, COX-2, and COX-1 and visualized with the ECL Western blot analysis system (Amersham Life Science, Inc.) as described previously (20).

Reverse Transcription-Polymerase Chain Reaction (RT-PCR)

To amplify mouse type IIA sPLA2 cDNA, the sense primer 5'-ATGAAGGTCCTCCTCCTGCTAG-3' and antisense primer 5'-TCAGCATTTGGGCTTCTTCC-3' were used (20). The RT-PCR was carried out using an RNA PCR kit (AMV) Version 2 (Takara), according to the manufacturer instructions, with 1 µg of total RNA from MC3T3-E1 cells as a template. Equal amounts of each RT product were PCR-amplified with ex Taq polymerase (Takara) by 25 cycles consisting of 1 min at 94 °C, 1 min at 57 °C, and 2 min at 72 °C. The amplified cDNA was resolved on 2% (w/v) agarose gels, transferred onto an Immobilon-N membrane, and probed with mouse type IIA sPLA2 cDNA prelabeled with [32]dCTP.

Genomic Analysis of the Mouse sPLA2 Gene

To obtain genomic DNA, the cell lysate was incubated with 100 µg/ml proteinase K (Sigma) overnight at 55 °C. The DNA was extracted with phenol/chloroform, precipitated with ethanol, and reconstituted with 10 mM Tris-HCl, pH 8.0, containing 1 mM EDTA. The DNA obtained was digested with BamHI (Takara) and a portion was separated on 1% (w/v) agarose gel, transferred to an Immobilon-N membrane, and probed with mouse sPLA2 cDNA prelabeled with [32P]dCTP. In accordance with the type IIA sPLA2 genotype (23, 24), DNA obtained from MC3T3-E1 cells yielded a single 8.2-kb band, whereas DNA from BALB/cJ mouse brain, used as a positive control that expresses normal sPLA2, yielded a 2.5- and a 5.7-kb band after BamHI digestion (data not shown).

Preparation of Recombinant Mouse sPLA2

Recombinant mouse type IIA sPLA2 expressed by Sf9 cells (Invitrogen) transfected with sPLA2 cDNA using a baculovirus system was purified using sequential heparin-Sepharose and anti-sPLA2 antibody column chromatography, as described previously (25). The purity of the recombinant enzyme was checked with a silver staining kit (Wako), and the amount was determined using a BCA protein assay kit (Pierce).

Transfection Analysis

Approximately 1 µg of mouse type IIA sPLA2 cDNA subcloned into pCRTM3 (Invitrogen) (20) was mixed with 5 µl of CellFection (Life Technologies) in 200 µl of Opti-MEM medium (Life Technologies) for 15 min and then added to MC3T3-E1 cells that had attained 60-80% confluence in 6-well plates and had been supplemented with 0.8 ml of Opti-MEM. After incubation for 6 h, the medium was replaced with 2 ml of fresh culture medium, the cells were cultured overnight, the medium was replaced again with 2 ml of fresh medium, and cultured for a further 2 days. For the study of transient expression, the cells were then cultured with or without IL-1beta /TNFalpha . To establish stable sPLA2 transformants, transfected cells were cloned by limiting dilution in 96-well plates in culture medium supplemented with 400 µg/ml G418 (Life Technologies). After culture for 21 days, wells containing a single colony were chosen and sPLA2 expression was assessed by RT-PCR. The established clones were expanded and used for the experiments.


RESULTS

Regulation of Delayed PGE2 Generation and Expression of cPLA2 and Two COX Isoforms

When MC3T3-E1 cells were stimulated with the combination of 5 ng/ml IL-1beta and 1,000 units/ml TNFalpha , PGE2 was generated, reaching a peak within 3-6 h followed by a plateau phase up to 24 h with a second increasing phase from 24 to 48 h (Fig. 1A). When cultured with IL-1beta /TNFalpha , the time course of PGE2 generation was parallel to that of COX-2 protein expression, which was undetectable before stimulation, induced markedly within 3-6 h, and followed by a plateau phase and a further increase over 24-48 h (Fig. 1B). COX-2 mRNA, which was undetectable in the absence of IL-1beta /TNFalpha , increased to reach its maximal level at 3 h, decreased from 3 to 6 h, and then increased again to reach a second peak over 24-48 h (Fig. 1C). The cPLA2 protein was constitutively expressed in MC3T3-E1 cells and increased markedly during stimulation with IL-1beta /TNFalpha for 12-48 h (Fig. 1B), accompanied by a modest increase in expression of its mRNA (Fig. 1C). COX-1 protein, which was expressed constitutively, also increased after stimulation with IL-1beta /TNFalpha for 12-48 h (Fig. 1B), in parallel to the increased expression of COX-1 mRNA (Fig. 1C).


Fig. 1. Time courses of PGE2 generation (A) and the expression of cPLA2, COX-1, and COX-2 proteins (B) and steady-state transcripts (C) in MC3T3-E1 cells after activation with IL-1beta and TNFalpha . The cells were cultured for the indicated periods with or without 5 ng/ml IL-1beta and 1,000 units/ml TNFalpha , as described under "Experimental Procedures." A representative result of three independent experiments is shown.
[View Larger Version of this Image (29K GIF file)]

Dose-response experiments, assessed after 48 h, revealed that optimal PGE2 generation occurred when >1.2 ng/ml IL-1beta (EC50 between 0.3 and 0.6 ng/ml) was added to the culture in the presence of TNFalpha (Fig. 2A). The dose-dependence of PGE2 generation on IL-1beta showed close correlations with the increased expression of cPLA2 and COX-2 proteins, which reached maximal levels in the presence of 1.2 ng/ml IL-1beta (Fig. 2B), whereas the COX-1 protein level attained its peak with as little as 0.3 ng/ml IL-1beta (Fig. 2B).


Fig. 2. Dose-dependent effects of IL-1beta on PGE2 generation (A) and cPLA2, COX-1, and COX-2 protein expression (B). MC3T3-E1 cells were cultured for 48 h with the indicated concentrations of IL-1beta in the presence of 1,000 units/ml TNFalpha . A representative result of three independent experiments is shown.
[View Larger Version of this Image (22K GIF file)]

IL-1beta /TNFalpha -induced PGE2 generation was suppressed almost completely by 1 ng/ml NS-398 (COX-2 inhibitor (26)) but was largely resistant to 10 µg/ml valeryl salicylate (COX-1 inhibitor (27)) (Fig. 3A). Stimulation of cells in the co-presence of 10 µg/ml aspirin, which inactivated both COX-1 and -2 (28), resulted in complete suppression of PGE2 generation, whereas pretreatment with 10 µg/ml aspirin, which irreversibly inactivated pre-existing COX-1 only, did not affect PGE2 generation (Fig. 3A). PGE2 generation was also suppressed almost completely by 1 µM AACOCF3 (cPLA2 inhibitor (29)) (Fig. 3A). The IC50 values of NS-398 and AACOCF3 were approximately 0.1 ng/ml and 0.02 µM, respectively (Fig. 4).


Fig. 3. Effects of COX and cPLA2 inhibitors on PGE2 generation (A) and the expression of cPLA2 and COX-2 proteins (B) and transcripts (C) in MC3T3-E1 cells after activation with IL-1beta and TNFalpha . MC3T3-E1 cells were cultured for 48 h with 5 ng/ml IL-1beta and 1,000 units/ml TNFalpha in the presence of either 10 µg/ml valeryl salicylate (VS), 10 µg/ml aspirin, 1 ng/ml NS-398, 1 µM AACOCF3 or vehicle. In the aspirin pretreatment experiment, the cells were preincubated for 5 h with 10 µg/ml aspirin, washed, and then activated with IL-1beta /TNFalpha . In panels B and C, 100 µM arachidonic acid (AA) or 10 µg/ml PGE2 was added, as required for the experiments. A representative result of three independent experiments is shown.
[View Larger Version of this Image (30K GIF file)]


Fig. 4. Dose-dependent effect of cPLA2 and COX-2 inhibitors on IL-1beta /TNFalpha -induced PGE2 generation and cPLA2 and COX-2 protein expression. MC3T3-E1 cells were cultured for 48 h with 5 ng/ml IL-1beta and 1,000 units/ml TNFalpha in the presence of the indicated concentrations of AACOCF3 or NS-398. A representative result of four independent experiments is shown.
[View Larger Version of this Image (37K GIF file)]

Amplification of the Delayed Response by PGE2

We found that when MC3T3-E1 cells were stimulated for 48 h with IL-1beta /TNFalpha in the presence of NS-398 or AACOCF3, induction of proteins (Fig. 3B) and transcripts (Fig. 3C) for cPLA2 and COX-2 was suppressed simultaneously to almost basal levels. The inhibition of cPLA2 and COX-2 expression was dependent upon the concentrations of these inhibitors and showed good correspondence with the reduction of PGE2 generation (Fig. 4). In contrast, these inhibitors did not affect COX-1 expression (Fig. 5), and the COX-1 inhibitor, valeryl salicylate, had no effect on the induction of cPLA2 and COX-2 expression (Fig. 3, B and C). Notably, when exogenous PGE2 (10 µg/ml) was added to IL-1beta /TNFalpha -stimulated cells cultured in the presence of AACOCF3 or NS-398, induction of both cPLA2 and COX-2 expression was restored completely (Fig. 3, B and C). Furthermore, adding exogenous PGE2 increased cPLA2 and COX-2 expression in cells stimulated with IL-1beta /TNFalpha in the absence of these inhibitors even more, although adding PGE2 alone to unstimulated cells had no effect (Figs. 3C and 5A). The concentration of exogenous PGE2 required for the recovery of cPLA2 and COX-2 induction in NS-398-treated cells was 1-10 µg/ml (Fig. 5B), comparable with the amounts of PGE2 produced by IL-1beta /TNFalpha -stimulated cells in the absence of NS-398 (Fig. 1A). These observations led us to hypothesize that PGE2 produced by IL-1beta /TNFalpha -stimulated cells acted as an autocrine positive regulator of the induction of both cPLA2 and COX-2 expression.


Fig. 5. Effects of exogenous PGE2 and arachidonic acid on cPLA2, COX-2, and COX-1 protein expression. A, MC3T3-E1 cells were cultured with or without 5 ng/ml IL-1beta and/or 1,000 units/ml TNFalpha in the presence or absence of 10 µg/ml PGE2. B, cells were cultured with IL-1beta /TNFalpha with or without 1 ng/ml NS-398 in the presence of various concentrations of PGE2. C, cells were cultured with IL-1beta /TNFalpha with or without 1 µM AACOCF3 in the presence of various concentrations of arachidonic acid. After incubation for 48 h, immunoblotting using specific antibodies was carried out. A representative result of three independent experiments is shown.
[View Larger Version of this Image (23K GIF file)]

In support of this speculation, supplementing AACOCF3-treated cells with exogenous arachidonic acid, which bypassed cPLA2-mediated liberation of arachidonic acid from the endogenous pool, also restored the induction of cPLA2 and COX-2 expression (Fig. 3, B and C), and the concentration of exogenous arachidonic acid to do this was 10-100 µM (Fig. 5C), under which conditions 1-10 µg/ml PGE2 was consistently produced. Although NS-398 inhibited COX-2, and thereby preventing metabolism of endogenous arachidonic acid to PGE2, excess exogenous arachidonic acid was utilized by COX-1 and metabolized to PGE2, which reached 1~2 µg/ml, by NS-398-treated cells. Consequently, the addition of exogenous arachidonic acid to NS-398-treated cells restored cPLA2 and COX-2 expression partially (Fig. 3B), which was suppressed considerably by valeryl salicylate (data not shown). In contrast to the marked suppression of the induction of cPLA2 and COX-2 expression by AACOCF3 and NS-398 after IL-1beta /TNFalpha stimulation for 48 h (Fig. 3, B and C), the induction of COX-2 mRNA expression observed as early as 3 h was not affected by these inhibitors (data not shown). This result, together with the finding that an increase in the cPLA2 level was evident only after 12-48 h of culture (Fig. 1, B and C), suggests that potentiation of IL-1beta /TNFalpha -induced cPLA2 and COX-2 induction by endogenous PGE2 was limited to the later phase (12-48 h) of cell activation. COX-1 expression was affected neither by exogenous PGE2 nor by arachidonic acid (Fig. 5), providing further evidence that the mechanism responsible for the increase in COX-1 expression differs from that for cPLA2 and COX-2.

Effect of sPLA2 on PGE2 Generation

To assess whether sPLA2 affected PGE2 generation by type IIA sPLA2-deficient MC3T3-E1 cells, two experiments were carried out. First, MC3T3-E1 cells were exposed to exogenous mouse recombinant type IIA sPLA2 in the presence or absence of IL-1beta /TNFalpha . As shown in Fig. 6A, the addition of recombinant type IIA sPLA2 to IL-1beta /TNFalpha -stimulated cells resulted in an increase in PGE2 generation of about 2-fold (Fig. 6A), and this increase was accompanied by increased expression of cPLA2 and COX-2, but not COX-1 (Fig. 6B). In contrast, neither an appreciable increase in PGE2 generation (Fig. 6A) nor increased cPLA2 and COX-2 expression (Fig. 6B) occurred when type IIA sPLA2 was added to unstimulated cells.


Fig. 6. Effects of exogenous sPLA2 on PGE2 generation and enzyme expression. MC3T3-E1 cells were cultured for 48 h with or without 5 ng/ml IL-1beta and 1,000 units/ml TNFalpha in the presence or absence of 10 µg/ml recombinant mouse type IIA sPLA2, and PGE2 generation (A) and cPLA2, COX-2, and COX-1 protein expression (B) were assessed. A representative result of three independent experiments is shown.
[View Larger Version of this Image (19K GIF file)]

In the second experiment, we transfected mouse type IIA sPLA2 cDNA into MC3T3-E1 cells with the aim of revealing the role of endogenous sPLA2 in PGE2 generation. Cells transiently expressing type IIA sPLA2 produced 1.7 times more PGE2 than control cells in response to cytokine stimulation, whereas such transient expression had no appreciable effect in the absence of cytokines (Fig. 7A, left). When PGE2 biosynthesis by two independent stable sPLA2 transformants (E2 and A10), in which the sPLA2 transcript that control (E1) cells did not possess was detectable by RT-PCR (Fig. 7A, inset), was assessed, there were 1.6- and 2.2-fold increases in IL-1beta /TNFalpha -induced PGE2 generation by E2 and A10, respectively, relative to that by E1 (Fig. 7A, right). PGE2 generation by both sPLA2 transformants was suppressed almost completely by AACOCF3 or NS-398, implying that cPLA2 and COX-2 are involved in the enhanced PGE2 generation resulting from ectopic sPLA2 expression. In accordance with PGE2 regulation of the enzyme expression, the IL-1beta /TNFalpha -induced expression levels of proteins (Fig. 7B) and mRNA (Fig. 7C) for COX-2 and cPLA2 in sPLA2 transformants were higher than those in control cells, whereas the COX-1 expression levels did not correlate with their abilities to produce PGE2.


Fig. 7. PGE2 generation by type IIA sPLA2-transfected MC3T3-E1 cells. sPLA2 expression in sPLA2 transformants (E2 and A10) and control cells (E1), assessed by RT-PCR, is shown in the inset of panel A. Cells transfected with mouse type IIA sPLA2 cDNA or control vector were activated with IL-1beta and TNFalpha for 48 h, and PGE2 generation (A) and the expression of COX-2, COX-1, and cPLA2 proteins (B) and mRNA (C) was examined. Either 1 ng/ml NS-398 or 1 µM AACOCF3 was added, as required for the experiments. A representative result of three independent experiments is shown.
[View Larger Version of this Image (45K GIF file)]


DISCUSSION

In this study, we obtained evidence that the murine osteoblastic cell line MC3T3-E1, which originates from C57BL/6J strain that is genetically deficient in functional type IIA sPLA2 (23, 24), exhibits delayed PGE2 generation in response to IL-1beta and TNFalpha that is dependent upon cPLA2 and COX-2, with PGE2 acting as an autoamplifier. The findings that COX-2-dependent delayed PGE2 generation occurred even in the absence of type IIA sPLA2 but was augmented significantly following the enforced introduction of type IIA sPLA2 implies that type IIA sPLA2 is dispensable, but capable of participating, as an enhancer, in the COX-2-dependent delayed PGE2 generation. Moreover, the enhancing effect of type IIA sPLA2 on delayed PGE2 generation was mediated predominantly by increased cPLA2 and COX-2 expression, thereby revealing an unexplored and unambiguous cross-talk between PG biosynthetic enzymes to provide an optimal response in this cell line.

Cytokine-induced PGE2 generation was suppressed completely not only by NS-398 (IC50 ~ 0.1 ng/ml) but also by AACOCF3 (IC50 ~ 0.02 µM), implying the involvement of COX-2 and cPLA2 in the delayed response. An unexpected finding was that treating cells with either NS-398 or AACOCF3 led to concordant reduction of the expression of cPLA2 and COX-2, and supplementing IL-1beta /TNFalpha -stimulated cells with exogenous PGE2 or arachidonic acid abrogated the inhibitory actions of these inhibitors on the expression of both enzymes. Notably, the induction of COX-2 expression was biphasic as evidenced by steady-state expression of its transcript; during the first phase, occurring by 3 h, it was independent of PGE2, whereas during the second phase, proceeding over 12-48 h, it was subjected to autocrine amplification by PGE2. Unlike COX-2, increased cPLA2 expression was evident only after 12-48 h, occurring in parallel with the second phase of COX-2 induction. These results suggest that endogenous PGE2 is a prerequisite for the cytokine-induced increase in cPLA2 expression. Since the increase in cPLA2 protein was more obvious than that in cPLA2 mRNA, a significant post-transcriptional regulation of the expression of its protein must occur, as has been demonstrated in cytokine-stimulated mast cells (30, 31).

PGE2 generation occurred as early as 3 h after the initiation of activation when COX-2 expression had already been induced, but cPLA2 expression remained unchanged (Fig. 1). These results, together with those of several previous studies (7, 8), indicate that increased cPLA2 expression is not the sole factor, but post-translational modification of this enzyme is essential for its function. Yamamoto and co-workers (32) showed that transient arachidonic acid release preceded the initial rise in PGE2 generation by TNFalpha -stimulated MC3T3-E1 cells. We also observed a significant increase in arachidonic acid release following IL-1beta /TNFalpha stimulation that reached a peak within 1-2 h (data not shown). It is therefore likely that this early arachidonic acid release promoted the initial phase of COX-2-dependent PGE2 generation. Nevertheless, as both IL-1beta and TNFalpha are poor Ca2+-mobilizers, cPLA2 activation throughout the culture period might occur through a Ca2+-independent pathway. Although both IL-1beta and TNFalpha are known to activate several mitogen-activated protein kinase family members (33), which in turn phosphorylate cPLA2 at Ser-505 and increase its catalytic activity to some extent (6, 34), accumulated evidence has revealed that this phosphorylation alone is insufficient for cPLA2 activation in vivo (6-8, 34). A recently described novel phosphorylation site on cPLA2 (Ser-727) might account, at least in part, for the Ca2+-independent regulation of cPLA2 (35), although this has not been proved directly.

Although COX-1 expression during culture increased in parallel with cPLA2 and the second phase of COX-2 induction, the failure of NS-398 and AACOCF3 to suppress the increase in COX-1 expression and the inability of exogenous PGE2 and arachidonic acid to increase it suggest that the induction of COX-1 expression is regulated independently of PGE2. COX-1 has been shown to have the capacity to metabolize excess exogenous arachidonic acid to PGs (2). Consistent with this, we found that exogenous arachidonic acid was converted to PGE2 through the COX-1 pathway by NS-398-treated MC3T3-E1 cells. Alternatively, and more physiologically, COX-1 has been implicated in immediate PG biosynthesis utilizing endogenously derived arachidonic acid, such as IgE-dependent PGD2 generation by mast cells (3), thromboxane generation by thrombin-stimulated platelets (36), and PGE2 generation by A23187-stimulated COS-7 cells overexpressing cPLA2 and COX-1 (37). Increased COX-1 expression in mast cells leads to priming for increased immediate PGD2 synthesis (30). Therefore, changes in COX-1 expression may reflect the particular stage of cell differentiation, enabling the differentiated cells to produce more PGs rapidly when they are exposed to exogenous arachidonic acid or encounter any stimulus that initiates immediate PG biosynthesis involving functional linkage of cPLA2 and COX-1.

Exogenous addition or enforced expression of type IIA sPLA2 led to significant enhancement of IL-1beta /TNFalpha -initiated PGE2 generation by type IIA sPLA2-deficient MC3T3-E1 cells, whereas sPLA2 alone had no appreciable effect on PGE2 generation. These observations are in line with the currently proposed hypothesis that type IIA sPLA2 contributes to the enhancement, rather than initiation, of biological responses (11, 17-20). Of particular importance are the findings that the introduction of type IIA sPLA2 was accompanied by increased expression of both cPLA2 and COX-2 and that enhanced PGE2 generation was suppressed by AACOCF3 and NS-398. The most likely explanation for these observations is that sPLA2 hydrolyzes plasma membrane phospholipids, leading to the production of a small amount of PGE2, which in turn amplifies the cPLA2/COX-2-dependent PGE2 biosynthetic pathway. An alternative possibility is that sPLA2 elicits the induction of cPLA2 and COX-2 expression through the sPLA2 receptor, independently of its phospholipid-hydrolyzing activity (38). However, the only sPLA2 receptor on MC3T3-E1 cells identified so far is type I sPLA2-specific (39). Nevertheless, the almost complete abrogation of sPLA2-enhanced PGE2 generation by AACOCF3 suggests that cPLA2-derived arachidonic acid accounts for the majority of the arachidonic acid released (i.e. sPLA2 is an upstream modulator of cPLA2) or that, as demonstrated by Balsinde and Dennis (40), functional cPLA2 is required for sPLA2 to exert its action (i.e. cPLA2 is an upstream modulator of sPLA2).

The fact that type IIA-deficient inbred mouse strains display no abnormality (23, 24) suggests that some other sPLA2s could substitute for type IIA sPLA2 function. In view of this standpoint, type V sPLA2 has been recently shown to be an alternative effector of arachidonate metabolism (21, 22). In our preliminary study, any sPLA2 activity, assessed under the standard sPLA2 assay condition (20), was not detected in MC3T3-E1 cells (data not shown). Therefore the contribution of type V sPLA2 to PGE2 generation in this cell line appears to be negligible, even though the possiblity that a pretty low level of endogenous type V sPLA2 still participates in it as a modifier cannot be ruled out. Moreover, it would be possible that exogenous type V sPLA2 compensates for type IIA sPLA2 to augment PGE2 generation in a similar way.

Nonetheless, not only did this study reveal functional coupling of cPLA2 and sPLA2 with COX-2, in which cPLA2 and sPLA2 act as an initiator and an enhancer, respectively, of delayed PGE2 generation, but it also revealed a mechanism whereby sPLA2 augments the delayed response by inducing cPLA2 and COX-2 expression. However, the pathway shown here, which utilizes a metabolite as a signal amplifier (PGE2 in this case), may not be always applicable to other systems, but rather limited to particular cell types. For instance, the induction of COX-2 expression in cytokine-stimulated mast cells was not affected by several PLA2 and COX inhibitors, indicating that the regulatory mechanism was PG-independent (41). Arachidonic acid up-regulated, whereas PGE2 down-regulated, the induction of COX-2 in uterine stromal cells (42). IL-1beta -induced arachidonic acid release and subsequent COX-2-dependent PGE2 generation were augmented significantly in sPLA2-transfected compared with normal CHO-K1 cells, without accompanying changes in cPLA2 and COX-2 expression (20). In mouse calvaria cells, cPLA2 did not couple with IL-1alpha -induced COX-2 to produce PGE2 unless a rapid increase in the intracellular Ca2+ levels was triggered by a secondary Ca2+-mobilizing stimulus (43). These observations imply the mechanisms responsible for PG biosynthesis by discrete PLA2s differ according to the cell type.


FOOTNOTES

*   This work is supported by Grants-in-Aid for Scientific Reseach from the Ministry of Education, Science and Culture of Japan.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Dagger    To whom reprint requests should be addressed: Dept. of Health Chemistry, School of Pharmaceutical Sciences, Showa University, 1-5-8 Hatanodai, Shinagawa-ku, Tokyo 142, Japan. Tel.: 81-3-3784-8196; Fax: 81-3-3784-8245; E-mail: kudo{at}pharm.showa-u.ac.jp.
1   The abbreviations used are: PG, prostaglandin; cPLA2, cytosolic phospholipase A2; sPLA2, secretory phospholipase A2; COX, cyclooxygenase; IL, interleukin; TNFalpha , tumor necrosis factor alpha ; AACOCF3, arachidonyl trifluoromethyl ketone; PCR, polymerase chain reaction; RT-PCR, reverse transcriptase-PCR; kb, kilobase(s).

ACKNOWLEDGEMENTS

We thank Drs. J. D. Clark, W. L. Smith, J. Trzaskos, and J. P. Arm for providing cDNAs, antibodies, and inhibitors. We thank T. Tamaoki for technical assistance.


REFERENCES

  1. Smith, W. L., Garavito, R. M., and DeWitt, D. L. (1996) J. Biol. Chem. 271, 33157-33160 [Free Full Text]
  2. Reddy, S. T., and Herschman, H. R. (1994) J. Biol. Chem. 269, 15473-15480 [Abstract/Free Full Text]
  3. Murakami, M., Matsumoto, R., Austen, K. F., and Arm, J. P. (1994) J. Biol. Chem. 269, 22269-22275 [Abstract/Free Full Text]
  4. Morham, S. G., Langenbach, R., Loftin, C. D., Tiano, H. F., Vouloumanos, N., Jennette, J. C., Mahler, J. F., Kluckman, K. D., Ledford, A., Lee, C. A., and Smithies, O. (1995) Cell 83, 473-482 [Medline] [Order article via Infotrieve]
  5. Clark, J. D., Lin, L.-L., Kriz, R. W., Ramesha, C. S., Sultzman, L. A., Lin, A. Y., Milona, N., and Knopf, J. L. (1991) Cell 65, 1043-1051 [Medline] [Order article via Infotrieve]
  6. Lin, L.-L., Wartmann, M., Lin, A. Y., Knopf, J. L., Seth, A., and Davis, R. J. (1993) Cell 72, 269-278 [Medline] [Order article via Infotrieve]
  7. Hoeck, W. G., Ramesha, C. S., Chang, D. J., Fan, N., and Heller, R. A. (1993) Proc. Natl. Acad. Sci. U. S. A. 90, 4475-4479 [Abstract]
  8. Wu, T., Levine, S. J., Lawrence, M. G., Logun, C., Angus, C. W., and Shelhamer, J. H. (1994) J. Clin. Invest. 93, 571-577 [Medline] [Order article via Infotrieve]
  9. Lin, L.-L., Lin, A. Y., and DeWitt, D. L. (1992) J. Biol. Chem. 267, 23451-23454 [Abstract/Free Full Text]
  10. Roshak, A., Sathe, G., and Marshall, L. A. (1994) J. Biol. Chem. 269, 25999-26005 [Abstract/Free Full Text]
  11. Murakami, M., Kudo, I., and Inoue, K. (1995) J. Lipid Mediat. Cell Signal. 12, 119-130 [CrossRef][Medline] [Order article via Infotrieve]
  12. Murakami, M., Kudo, I., Suwa, Y., and Inoue, K. (1992) Eur. J. Biochem. 209, 257-265 [Abstract]
  13. Oka, S., and Arita, H. (1991) J. Biol. Chem. 266, 9956-9960 [Abstract/Free Full Text]
  14. Murakami, M., Kudo, I., and Inoue, K. (1993) J. Biol. Chem. 268, 839-844 [Abstract/Free Full Text]
  15. Pfeilschifter, J., Schalkwijk, C., Briner, V. A., and van den Bosch, H. (1993) J. Clin. Invest. 92, 2516-2523 [Medline] [Order article via Infotrieve]
  16. Barbour, S. E., and Dennis, E. A. (1993) J. Biol. Chem. 268, 21875-21882 [Abstract/Free Full Text]
  17. Tanaka, K., Kato, T., Matsumoto, K., and Yoshida, T. (1993) Inflammation 17, 107-119 [Medline] [Order article via Infotrieve]
  18. Murakami, M., Kudo, I., Nakamura, H., Yokoyama, Y., Mori, H., and Inoue, K. (1990) FEBS Lett. 268, 113-116 [CrossRef][Medline] [Order article via Infotrieve]
  19. Chapdelaine, J. M., Ciofalo, V. B., Grass, D. S., Felkner, R. H., Wallace, R. E., and Swanson, M. E. (1995) Arthritis Rheum. 38, S293
  20. Murakami, M., Nakatani, Y., and Kudo, I. (1996) J. Biol. Chem. 271, 30041-30051 [Abstract/Free Full Text]
  21. Balboa, M. A., Balsinde, J., Winstead, M. V., Tischfield, J. A., and Dennis, E. A. (1996) J. Biol. Chem. 271, 32381-32384 [Abstract/Free Full Text]
  22. Reddy, S. T., Winstead, M. V., Tischfield, J. A., and Herschman, H. R. (1997) J. Biol. Chem. 272, 13591-13596 [Abstract/Free Full Text]
  23. MacPhee, M., Chepenik, K. P., Liddell, R., Nelson, K. K., Siracusa, L. D., and Buchberg, A. M. (1995) Cell 81, 957-966 [Medline] [Order article via Infotrieve]
  24. Kennedy, B. P., Payette, P., Mudgett, J., Vadas, P., Pruzanski, W., Kwan, M., Tang, C., Rancourt, D. E., and Cromlish, W. A. (1995) J. Biol. Chem. 270, 22378-22385 [Abstract/Free Full Text]
  25. Ashraf, M. M. D., Murakami, M., Shimbara, S., Amakasu, Y., Atsumi, G., and Kudo, I. (1996) Biochem. Biophys. Res. Commun. 229, 726-732 [CrossRef][Medline] [Order article via Infotrieve]
  26. Futaki, N., Yoshikawa, K., Hamasaka, Y., Arai, I., Higuchi, S., Iizuka, H., and Otomo, S. (1993) Gen. Pharmacol. 24, 105-110 [Medline] [Order article via Infotrieve]
  27. Bhattacharyya, D. K., Lecomte, M., Dunn, J., Morgans, D. J., and Smith, W. L. (1995) Arch. Biochem. Biophys. 317, 19-24 [CrossRef][Medline] [Order article via Infotrieve]
  28. Medae, E. A., Smith, W. L., and DeWitt, D. L. (1993) J. Biol. Chem. 268, 6610-6614 [Abstract/Free Full Text]
  29. Street, I. P., Lin, H.-K., Laliberte, F., Ghomashchi, F., Wang, Z., Perrier, H., Tremblay, N. M., Huang, Z., Weech, P. K., and Gelb, M. H. (1993) Biochemistry 32, 5935-5940 [Medline] [Order article via Infotrieve]
  30. Murakami, M., Matsumoto, R., Urade, Y., Austen, K. F., and Arm, J. P. (1995) J. Biol. Chem. 270, 3239-3246 [Abstract/Free Full Text]
  31. Murakami, M., Austen, K. F., Bingham, C. O., III, Friend, D. S., Penrose, J. F., and Arm, J. P. (1995) J. Biol. Chem. 270, 22653-22656 [Abstract/Free Full Text]
  32. Takahashi, Y., Taketani, Y., Endo, T., Yamamoto, S., and Kumegawa, M. (1994) Biochim. Biophys. Acta 1212, 217-224 [Medline] [Order article via Infotrieve]
  33. Liu, Z.-G., Hsu, H., Goeddel, D. V., and Karin, M. (1996) Cell 87, 565-576 [Medline] [Order article via Infotrieve]
  34. Kramer, R. M., Roberts, E. F., Um, S. L., Borsch-Haubold, A. G., Watson, S. P., Fisher, M. J., and Jakubowski, J. A. (1996) J. Biol. Chem. 271, 27723-27729 [Abstract/Free Full Text]
  35. de Carvalho, M. G. S., McCormack, A. L., Olson, E., Ghomashchi, F., Gelb, M. H., Yates, J. R., III, and Leslie, C. C. (1996) J. Biol. Chem. 271, 6987-6997 [Abstract/Free Full Text]
  36. Langenbach, R., Morham, S. G., Tiano, H. F., Loftin, C. D., Ghanayem, B. I., Chulada, P. C., Mahler, J. F., Lee, C. A., Goulding, E. H., Kluckman, K. D., Kim, H. S., and Smithies, O. (1995) Cell 83, 483-492 [Medline] [Order article via Infotrieve]
  37. Huang, Z., Payette, P., Abdullah, K., Cromlish, W. A., and Kennedy, B. P. (1996) Biochemistry 35, 3712-3721 [CrossRef][Medline] [Order article via Infotrieve]
  38. Lambeau, G., Ancian, P., Barhanin, J., and Lazdunski, M. (1994) J. Biol. Chem. 269, 1575-1578 [Abstract/Free Full Text]
  39. Tohkin, M., Kishino, J., Ishizaki, J., and Arita, H. (1993) J. Biol. Chem. 268, 2865-2871 [Abstract/Free Full Text]
  40. Balsinde, J., and Dennis, E. A. (1996) J. Biol. Chem. 271, 6758-6765 [Abstract/Free Full Text]
  41. Murakami, M., Bingham, C. O., III, Matsumoto, R., Austen, K. F., and Arm, J. P. (1995) J. Immunol. 155, 4445-4453 [Abstract]
  42. Prigent-Tessier, A., Pageaux, J-F., Fayard, J-M., Lagarde, M., Laugier, C., and Cohen, H. (1996) Eur. J. Biochem. 241, 872-878 [Abstract]
  43. Chen, Q-R., Miyaura, C., Higashi, S., Murakami, M., Kudo, I., Saito, S., Hiraide, T., Shibasaki, Y., and Suda, T. (1997) J. Biol. Chem. 272, 5952-5968 [Abstract/Free Full Text]

©1997 by The American Society for Biochemistry and Molecular Biology, Inc.