(Received for publication, May 12, 1997, and in revised form, June 3, 1997)
From the Department of Health Chemistry, School of Pharmaceutical Sciences, Showa University, Shinagawa-ku, Tokyo 142, Japan
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)-1 and tumor necrosis factor
(TNF
) 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-1
/TNF
-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.
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.
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 CellsMouse osteoblastic MC3T3-E1
cells (Riken Cell Bank) were maintained in -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)-1
(Genzyme) and/or 1,000 units/ml
mouse tumor necrosis factor
(TNF
; 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.
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 -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).
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.
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 sPLA2Recombinant 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 AnalysisApproximately 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-1/TNF
. 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.
When MC3T3-E1 cells were
stimulated with the combination of 5 ng/ml IL-1 and 1,000 units/ml
TNF
, 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-1
/TNF
, 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-1
/TNF
, 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-1
/TNF
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-1
/TNF
for 12-48 h (Fig. 1B), in parallel to the increased
expression of COX-1 mRNA (Fig. 1C).
Dose-response experiments, assessed after 48 h, revealed that
optimal PGE2 generation occurred when >1.2 ng/ml IL-1
(EC50 between 0.3 and 0.6 ng/ml) was added to the culture
in the presence of TNF
(Fig.
2A). The dose-dependence of
PGE2 generation on IL-1
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-1
(Fig.
2B), whereas the COX-1 protein level attained its peak with
as little as 0.3 ng/ml IL-1
(Fig. 2B).
IL-1/TNF
-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).
Amplification of the Delayed Response by PGE2
We
found that when MC3T3-E1 cells were stimulated for 48 h with
IL-1/TNF
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-1
/TNF
-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-1
/TNF
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-1
/TNF
-stimulated cells in the absence of NS-398
(Fig. 1A). These observations led us to hypothesize that
PGE2 produced by IL-1
/TNF
-stimulated cells acted as
an autocrine positive regulator of the induction of both cPLA2 and COX-2 expression.
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-1/TNF
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-1
/TNF
-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.
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-1/TNF
. As shown in Fig.
6A, the addition of
recombinant type IIA sPLA2 to IL-1
/TNF
-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.
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-1/TNF
-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-1
/TNF
-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.
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-1 and TNF
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-1/TNF
-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 TNF-stimulated MC3T3-E1 cells. We also observed a
significant increase in arachidonic acid release following
IL-1
/TNF
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-1
and TNF
are poor Ca2+-mobilizers, cPLA2 activation
throughout the culture period might occur through a
Ca2+-independent pathway. Although both IL-1
and TNF
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-1/TNF
-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-1-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-1
-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.
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.