| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
,#,¶
,¶,#,¶,§
*
The Nora Eccles Harrison Cardiovascular Research and Training Institute,
The Eccles Program in Human Molecular Biology and Genetics, and the Departments of
Anatomy,
§
Biochemistry,
¶
Medicine,
||
Pathology, and
#
Pediatrics, University of Utah School of Medicine, Salt Lake City, UT 84112
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, IL-1, and IL-8. Expression of COX-2 in stimulated PMNs was
paralleled by secretion of PGE2. The release of
PGE2 was blocked by a selective nonsteroidal inhibitor of
COX-2, indicating that the enzyme is responsible for the prostanoids
produced, and was inhibited by dexamethasone. The time course of
LPS-induced COX-2 expression and other features were different in
freshly isolated PMNs, monocytes, and macrophages, indicating that
COX-2 expression is differentially regulated in myeloid cells of
different lineages and degrees of maturation. Consistent with this,
IL-4 and IL-10, which suppressed LPS-induced COX-2 expression in
monocytes, had little effect on this response by PMNs. These
experiments demonstrate that PMNs express COX-2 when appropriately
stimulated. Thus, they may actively influence the eicosanoid
composition of the acute inflammatory milieu. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
PMNs synthesize factors that locally alter the functions of other cells in both defensive and un-regulated acute inflammatory responses. The best-known examples are reactive oxygen metabolites and leukotrienes (LT). Oxygen metabolites kill bacteria, injure host cells, and can mediate intercellular signaling (6, 7). LTs, a group of oxygenated products of arachidonic acid that are recognized by receptors on target cells, also mediate intercellular signaling (8, 9). The production of reactive oxygen metabolites and LTs is mediated by enzymes that are constitutively present in PMNs when they exit the bone marrow (10, 11). Until recently, it was thought that PMNs, which are terminally differentiated cells, do not synthesize new enzymes or other proteins in response to inflammatory stimuli. However, recent in vitro experiments indicate that genes for chemokines, cytokines, and other factors can be induced in stimulated PMNs (12, 13, 14). There is also evidence that PMNs have the capacity to synthesize new enzymes that confer the potential to modify the types and/or concentrations of signaling factors present in infected or injured tissues (15, 16). However, the conditions under which synthesis of new enzymes and other factors by PMNs occur are, for the most part, unknown (17). Furthermore, it is largely unknown if synthesis of modulatory enzymes occurs in human neutrophils in inflammatory states in vivo.
Cyclooxygenase (COX), also known as PGH synthase, is the enzyme that catalyzes the committed step in the metabolism of arachidonic acid to PGs and thromboxanes (TXs) (18, 19). Two isoforms of COX exist. The type I enzyme (COX-1) is constitutively present in many cell types, whereas type II (COX-2) is usually absent under basal conditions but can be induced in certain cells by mitogens, cytokines, and other factors (18, 19, 20, 21). Several normal human cell types are known to express the inducible isoform when appropriately stimulated (22, 23, 24, 25, 26), and eicosanoid products of arachidonic acid produced by constitutive activity of COX-2 may influence physiologic functions in certain tissues (27). In contrast, inappropriate expression of COX-2 is a mechanism of neoplasia (28, 29, 30).
In PMNs from normal subjects without inflammatory conditions,
arachidonic acid is dominantly metabolized to LTs by constitutively
present 5-lipoxygenase (8, 11). However, we considered the possibility
that COX-2 is inducible in neutrophils, since other human myeloid cells
can express the enzyme when appropriately stimulated (22, 31, 32, 33, 34, 35), and
that it may catalyze conversion of arachidonate to PGs. Here, we show
that COX-2 is expressed by isolated human PMNs stimulated with LPS,
TNF-, and other inflammatory agonists. We also show for the first
time that COX-2 is present in PMNs in inflamed human tissue. The
synthesis of COX-2 is accompanied by generation of PGE2
when PMNs are stimulated in vitro, and COX-2 expression and consequent
PGE2 release are influenced by chemokines and other factors
that modify acute inflammatory responses. We also found that the
regulation of COX-2 expression in PMNs varies from its control in
monocytes and in myelomonocytic cell lines that are derived from the
same progenitor cells (36). In several respects our findings differ
from those of Niiro et al. (37), who also reported that human PMNs
synthesize COX-2 when stimulated.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The protocols for collecting blood samples and tissue samples for immunohistochemical analysis (see below) were approved by the University of Utah Institutional Review Board. Whole blood was collected from consenting healthy adult donors into syringes containing 10,000 U of heparin/100 ml blood. Each 30 ml of blood was layered over 15 ml Ficoll-Paque (Pharmacia Biotech, Piscataway, NJ) and centrifuged at 800 x g in a Beckman Desktop TJ-6 centrifuge for 30 min. The upper layer and interface were removed and either discarded or saved for monocyte studies. The remaining cell fraction was washed twice with normal saline and pelleted at 400 x g. The cell pellet was then mixed 1:1 with 3% T-500 dextran (Pharmacia, Uppsala, Sweden), prepared in isotonic saline, and allowed to sediment for 1 h. The upper fraction, containing predominantly PMNs, was removed and centrifuged at 400 x g. The resulting pellet was treated for 30 s with 0.2% saline in order to lyse remaining erythrocytes, and the suspension was then made isotonic by adding 1.8% saline (1:1, v/v). The cells were washed, pelleted, and resuspended at a concentration of 5.0 x 106 cells/ml in RPMI 1640 containing 10% FBS, 100 U/ml penicillin, and 100 ng/ml streptomycin. The differential cell count was consistently >99% neutrophils, determined by staining cytospin preparations with Diff-Quik (Baxter Scientific Products, McGaw Park, IL) and counting across the entire slide. The viability of the cells was >99% as determined by trypan blue dye exclusion.
Isolation of peripheral blood monocytes
Peripheral blood monocytes were isolated by countercurrent elutriation as previously described (38).
Reagents
LPS (Escherichia coli 0111:B4), dexamethasone, PMA,
and calcium ionophore were purchased from Sigma Chemical Co., St.
Louis, MO. Recombinant human IL-10, IL-8, IL-4, IL-1ß, IL-1ra,
ENA-78, TNF-, GM-CSF and mouse anti-human IL-8 neutralizing Ab
were purchased from R&D Systems, Minneapolis, MN. Recombinant human C5a
was a kind gift from John Bohnsack, MD, Department of Pediatrics,
University of Utah, Salt Lake City, UT. PGE2 enzyme
immunoassay kits were purchased from Assay Designs, Ann Arbor, MI.
Arachidonic acid and NS-398 were purchased from Cayman Chemical, Ann
Arbor, MI.
Western blot analysis
Freshly isolated PMNs were placed in 12-well culture dishes (Falcon, no. 3043) and incubated with agonist or vehicle control at 37°C in 5% CO2 for various times as indicated. At the end of the incubation period the cells were scraped from the wells, centrifuged at 1400 rpm, separated from the medium, and lysed in ice-cold buffer (20 mM Tris-HCl, 16 mM 3-([3-cholamidopropyl]dimethylammonio)-1-propanesulfonate, 1 mM EDTA, 1 mM benzamidine hydrochloride, 1 µg/ml leupeptin, and 10 µg/ml soybean trypsin inhibitor). Following incubation on ice for 30 min, the lysates were centrifuged (1000 x g) for 5 min at 4°C and the pellets were discarded. The protein content in the lysate was determined using a microbicinchonic acid assay (Pierce Chemical Co., Rockford, IL) with BSA as standard. Samples containing 100 µg total protein were mixed with Laemmli reagent (final concentration 2% sodium dodecyl sulfate w/v, 10% glycerol v/v, 0.0005% bromphenol blue w/v) under reducing conditions (10% 2-ME) and heated at 95°C for 5 min. The samples were kept on ice prior to loading of gel. SDS-PAGE was performed using 10% acrylamide for the separating gel and 3.3% for the stacking gel. Proteins were transferred onto hybond ECL nitrocellulose membranes (Amersham, Life Science, Arlington Heights, IL) with a semidry transfer unit (Hoefer Scientific Instruments, San Francisco, CA). Transfer was performed in a 25 mM Tris, 192 mM glycine buffer, and 20% methanol for a minimum of 5 h at 500 mA. Blots were saturated for 1 h at room temperature with a 5% fatfree dry milk in Tris-buffered saline (50 mM Tris-HCl, pH 7.5, and 250 mM NaCl) and incubated with specific murine mAbs raised against human COX-1 or COX-2 (kind gifts from J. Maclouf, Lariboisiere Hospital, Federated Institute of Circulation-Lariboisiere, Paris, France) for 12 h at room temperature. The blots were then washed four times for 10 min each in the Tris buffer containing 5% fatfree dry milk and twice for 20 min each. In preliminary experiments, we optimized the concentration of COX Ab, the time of incubation, and the wash conditions in order to achieve the maximal signal. The blots were further incubated with horseradish peroxidase-conjugated goat anti-mouse Ab (Tago Immunologicals, Burlingame, CA) at 1/1000 for 1 h at room temperature. Excess secondary Ab was eliminated by washing in Tris-buffered saline (twice for 20 min followed by twice for 10 min). Chemiluminescence substrates (Amersham) were used to reveal positive bands on Hyperfilm ECL (Amersham) according to the manufacturer’s instructions. Protein bands were quantitated by computer assistance using NIH Image 1.55b software (NIH Public Software, Bethesda, MD) and a Macintosh Power PC. Normalization to total protein loaded was done for all samples.
Assays of eicosanoid and chemokine release by stimulated neutrophils
Supernatants were removed from the cells following scraping and
centrifugation at 1400 rpm. PGE2 concentrations were
measured by commercial enzyme immunoassay using the manufacturer’s
instruction (Assay Designs, Ann Arbor, MI). TNF- and IL-8 were
measured by ELISA as previously described (39).
Immunohistochemical analysis of COX-2 in inflamed tissues
We examined adult human lung tissue obtained at autopsy from three patients whose death was attributed to the ARDS, three patients with acute bacterial pneumonia, three patients with sepsis, and eight patients who died of nonpulmonary causes. The tissues were fixed in 10% buffered neutral formalin for 2 h at room temperature and stored in 70% ethanol until embedment in paraffin. Tissue sections (4–5 µm) were collected on PLUS slides (VWR, Media, PA) for immunohistochemistry. Briefly, the deparaffinized sections were treated with 3% H2O2 in methanol for 10 min at 37°C to remove endogenous peroxidase. The sections were washed with PBS, blocked with normal horse serum, and then incubated with monoclonal anti-human COX-2 (see above) as the primary Ab at 4°C overnight (the optimal dilution was 1:400). Staining controls included monoclonal anti-human von Willebrand factor and monoclonal anti-human insulin (both at 1:400 dilution) as primary Abs, omission of the primary Ab, and omission of the secondary Ab (biotinylated anti-murine IgG). Ag detection was done by the avidin-biotin-horseradish peroxidase method (ABC Elite kit; Vector Laboratories, Burlingame, CA). We used Gill’s no. 3 hematoxylin to counterstain the sections. Photography was done with the aid of a Zeiss Axiophot microscope.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Human peripheral blood neutrophils exposed to LPS (100 ng/ml)
expressed COX-2 when assayed by Western analysis using a specific mAb.
COX-2 appeared as a doublet (Fig. 1
A), reflecting differences in
glycosylation as previously reported (40). Little or no COX-2 protein
was detected in freshly isolated PMNs or in PMNs exposed to saline
alone, the vehicle for LPS. The COX-2 protein was present by 1 h
after treatment of the cell with LPS, peaked at 3 h, and returned
toward baseline values when incubations were extended to 24 h
(Fig. 1A). This temporal pattern was consistent in 10
experiments in which multiple time points were examined, although the
rate of return to resting levels was somewhat variable. The decrease in
COX-2 protein at 24 h was not due to death of the leukocytes,
since greater than 90% of the cells excluded trypan blue at this time
point in two experiments.
|
C). In parallel
assays using Western analysis with a specific mAb, COX-1 expression did
not change regardless of the concentration of LPS or the time of
incubation (data not shown).
Because our leukocyte isolation method yielded >99% neutrophils, the
purity of the cell suspensions provided clear evidence that the COX-2
protein detected on Western blotting was from this cell type. However,
we further examined the characteristics of expression of COX-2 in
monocytes compared with PMNs in order to exclude a contribution from
trace numbers of mononuclear cells in the PMN preparation. This
analysis was also important because monocytes accumulate with PMNs in
acute inflammatory lesions (1), and cell-cell interactions between the
two leukocyte types might influence the expression of COX-2 in
neutrophils. First, we isolated monocytes by countercurrent
elutriation, exposed them to LPS (100 ng/ml), extracted proteins, and
performed Western analysis on these samples. We found that freshly
isolated monocytes, like PMNs, did not express COX-2 in the absence of
stimulation. COX-2 protein expression was detected at 3 h after
LPS stimulation (Fig. 1B), similar to its induction in PMNs
(Fig. 1A), but the time course of expression thereafter
differed markedly in the two cell types. In monocytes, the expression
of COX-2 increased over the first 12 h of stimulation and was
sustained to 24 h (Fig. 1B). This contrasts with the
more transient time course in PMNs (Fig. 1A), and indicates
that COX-2 is regulated differently in the two classes of myeloid
cells. To further examine the contribution of monocytes to COX-2
protein in cell mixtures subjected to stimulation with LPS, we added
known numbers of monocytes to highly purified neutrophil preparations
and incubated them for 3 h. Enrichment with 2% monocytes did not
yield an increase in COX-2 protein over that in PMNs alone; addition of
6% monocytes was required to see a change in COX-2 protein. These
experiments indicate that PMNs synthesize COX-2 in the absence of
monocytes, although it remains possible that cell-cell interactions
between the two cells enhance the level of expression (for example, by
generation of IL-8 or TNF- by stimulated monocytes—see below).
Expression of COX-2 by LPS-stimulated neutrophils results in secretion of PGE2
To determine if the expression of COX-2 by PMNs results in the
synthesis of PGs, we measured the release of PGE2.
Suspensions of neutrophils or monocytes were stimulated with LPS (100
ng/ml) and then were incubated with exogenous arachidonic acid (20
µM) for 20 min. The supernatants were collected and analyzed for
PGE2. Both neutrophils and monocytes secreted
PGE2 into the medium when treated with LPS. The temporal
pattern of secretion (Fig. 2) paralleled
the expression of COX-2 protein in each cell type (compare Fig. 1 and
Fig. 2). Little or no PGE2 was released from LPS-stimulated
PMNs in the absence of addition of exogenous arachidonic acid,
indicating that induction of COX-2 was not sufficient for this
response. The release of PGE2 from PMNs stimulated under
these conditions was blocked by treatment of the leukocytes with a
selective nonsteroidal inhibitor of COX-2, NS-398 (41) (data not
shown). We also determined if endogenous arachidonate can be
metabolized to PGE2 when COX-2 has been induced. Following
stimulation with 100 ng/ml LPS, 10 µM calcium ionophore was added for
5 min to release arachidonate from membrane lipids. Stimulation of PMNs
under these conditions resulted in enhanced release of PGE2
(Fig. 2C), as with treatment of the cells with exogenous
arachidonic acid (Fig. 2A). Neutrophils exposed to the
vehicle for LPS alone released little PGE2 into the medium
when treated with arachidonate or calcium ionophore in parallel
incubations.
|
Chemokines and cytokines enhance COX-2 expression induced by LPS in neutrophils
We determined if chemokines, cytokines, and chemotactic peptides
alter the expression of COX-2 protein in PMNs, in addition to LPS. We
found that the rank order of potency of these factors in comparison
with LPS at 3 h of stimulation was: LPS > TNF- > IL-1,
GM-CSF. There was no increase in COX-2 protein expression over that in
PMNs treated with control buffer in response to IL-8, ENA-78, or C5a
(Fig. 3). We further examined TNF-,
IL-8, and IL-1 since they are synthesized by PMNs (12, 13) and could
provide autocrine signals. Treatment of PMNs with TNF- alone induced
COX-2 protein expression at a threshold concentration of 10 pg/ml.
COX-2 expression was maximal at 3 h and persisted for 24 h
(not shown). Thus, TNF- induces prolonged expression of COX-2 by
PMNs, in contrast to the more transient expression triggered by LPS
under these conditions (Fig. 1A). We then examined the
effect of concomitant treatment of PMNs with LPS and TNF-, since
additive or synergistic action of these two factors may be important in
sepsis and endotoxemia. At the peak of the time course triggered by LPS
(3 h), TNF- enhanced COX-2 protein expression (Fig. 4, top). The enhancement depended
on the concentration of TNF- (not shown), and the expression of
COX-2 persisted at maximal levels for 24 h (Fig. 4, top), as in the case of stimulation with TNF- alone. To
determine if this pattern of COX-2 protein expression was accompanied
by eicosanoid synthesis, we measured the release of PGE2.
We found a parallel pattern of enhanced PGE2 release when
the PMNs were incubated with LPS and TNF- in combination prior to
exposure to exogenous arachidonic acid. This enhancement occurred at
each time point over a 24-h incubation period (Fig. 4, bottom).
|
|
) (data
not shown). However, IL-8 in concentrations as low as 1 ng/ml
potentiated the expression of COX-2 in PMNs treated with LPS (100
ng/ml) for 3 h. The enhanced expression of COX-2 was evident
throughout the time course induced by LPS, but the time relationship
was not altered (Fig. 5A). We
found that IL-8 is released from PMNs in response to treatment with
LPS, as previously reported (12, 13), but that the onset of its
expression as measured by ELISA did not begin until after the peak of
COX-2 expression (not shown). In addition, a neutralizing Ab against
IL-8 did not reduce the expression of COX-2 in neutrophils stimulated
with LPS (Fig. 5B; compare lanes 2 and
5). In parallel, exogenous rIL-8 enhanced the expression of
COX-2 in LPS-stimulated PMNs (Fig. 5B, lane 4) as
expected from previous experiments (Fig. 5A). These
experiments indicate that autocrine stimulation of PMNs by IL-8 does
not contribute to expression of COX-2 at the time of its maximal
presence but that exogenous IL-8 can enhance its accumulation.
Similarly, incubating human PMNs with human rIL-1ß in combination
with LPS resulted in a small but consistent enhancement of COX-2
protein accumulation at 3 h (Fig. 3C). The presence of
human rIL-1ra did not inhibit COX-2 protein expression by
LPS-stimulated cells (Fig. 3C; compare lanes 3
and 6). In contrast, the enhanced response noted when PMNs
were coincubated with LPS and exogenous IL-1ß was inhibited in the
presence of IL-1ra (Fig. 5C; compare lanes 3,
4, and 5). These experiments indicate that IL-1,
like IL-8, is not a requisite autocrine factor for COX-2 synthesis but
that it can increase the level of COX-2 in PMNs if delivered
exogenously.
|
To determine if COX-2 is expressed by neutrophils in vivo, we
examined inflamed human lung from subjects with sepsis, ARDS, or
bacterial pneumonia by immunohistochemical analysis. Staining with the
same Ab used for Western analysis demonstrated that PMNs in lung
sections from patients with acute bacterial pneumonia express the
protein (Fig. 6). Neutrophils in sections
of lung tissue from subjects with ARDS or sepsis were also positive
(not shown). Sections of lung from subjects dying of noninflammatory
causes (cerebral neoplasm, head trauma) contained few neutrophils, as
expected, and these cells did not stain for COX-2. Immunostaining for
COX-2 was not uniform in PMNs in sections of lung from subjects dying
of pneumonia, ARDS, or sepsis (Fig. 6) (data not shown). Many
intravascular and extravascular PMNs were unstained or were lightly
stained when other PMNs in the same field were strongly positive for
the brown immunoperoxidase reaction product. Omission of the primary or
the secondary Ab or substitution of an irrelevant primary Ab
(anti-insulin) eliminated the immunostaining of the positive cells,
demonstrating specificity of the analysis (Fig. 6b). The
variable staining of PMNs in the same field may be a feature of the
sensitivity of the immunohistochemical procedure, the length of time
individual PMNs in the inflamed tissues had been exposed to an agonist
such as LPS or TNF- (Figs. 1 and 4), or differential expression of
positive and negative signaling molecules that modify COX-2 expression
by PMNs (Figs. 3, 4, 5, and below) in the local milieu. Certain other
cells in inflamed human lung were also stained by the Ab to COX-2
including macrophages (Fig. 6a), type 1 epithelial cells,
and endothelial cells. Each cell type has been reported to express
COX-2 in vitro. As with PMNs the immune staining of the latter cell
types was not uniform, and individual cells or groups of cells in each
section reacted strongly whereas others were weakly stained or were
negative.
|
Our examination of inflamed human tissues together with experiments in vitro (see above) indicated that the expression of COX-2 in PMNs and in other cells of myeloid lineage can be modified by positive and negative signaling factors. To further explore this issue we first compared freshly isolated PMNs and monocytes with other myeloid cells including monocyte-derived macrophages (38) and the HL60, U937, and THP-1 cell lines, which have been used to model responses of the primary cell types. Each cell type expressed mRNA or protein for COX-2 assayed by reverse transcriptase-PCR and/or by Western analysis when treated with LPS or PMA. However, each of the cell lines was much less consistent in its responses than were freshly isolated PMNs, monocytes, or cultured monocyte-derived macrophages (not shown). Therefore, we focused on the primary myeloid cells in subsequent studies of COX-2 regulation.
We next examined the effect of dexamethasone on COX-2 in PMNs compared
with other myeloid cells, since glucocorticoids have been reported to
inhibit its expression by both transcriptional and post-transcriptional
mechanisms (reviewed in Refs. 19 and 23). Isolated neutrophils
pretreated for 2 h with dexamethasone (2 to 10 µM) and
subsequently stimulated with LPS for 3 h contained less COX-2
protein than did control PMNs. In parallel, pretreatment of PMNs for
2 h with dexamethasone inhibited the enhanced release of
PGE2 induced by LPS, providing further evidence that
induction of COX-2 is required for this response. As previously
reported (20), dexamethasone also suppressed the expression of COX-2 in
monocyte-derived macrophages. These experiments indicate that the
expression of COX-2 in PMNs and other myeloid cells can be negatively
modified, that endogenous glucocorticoids may regulate the proteins,
and that the local balance of positive (e.g., IL-8 or TNF—see
above) and negative signaling molecules in an inflammatory lesion may
determine the level of COX-2 expression by PMNs (Fig. 6).
We then examined the effect of IL-4 and IL-10, which modulate acute
inflammatory responses and are reported to inhibit the synthesis of
COX-2 by human monocytes (42, 43, 44). We found that exogenous IL-10
suppressed the expression of COX-2 in monocytes treated with LPS for 3
to 12 h, confirming this observation (Fig. 7B). In contrast, IL-10 had no
inhibitory effect on the peak expression of COX-2 in neutrophils
stimulated with LPS, and little effect on its level at later times
(Fig. 7, A and C). We found that IL-10 inhibited
IL-8 release by LPS-stimulated PMNs (not shown), as expected (45, 46, 47),
indicating that it is recognized by this cell type and that it can
modify certain inflammatory responses. Similarly, we found that IL-4
completely suppressed COX-2 protein expression in monocytes treated
with LPS (not shown). In contrast, IL-4 only partially attenuated the
expression of COX-2 protein in PMNs at the time of maximal stimulation
with LPS. The magnitude of inhibition of COX-2 accumulation in PMNs
treated with IL-4 was variable; the maximal decrease is shown in Figure 7C. Together, these experiments indicate that COX-2 is
regulated differently by modifiers such as IL-4 and IL-10 in PMNs
compared with monocytes, although the two are related cells of the same
myeloid lineage.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). Because COX-2 is considered to be the critical target of
nonsteroidal anti-inflammatory drugs that are administered to
inhibit tissue-damaging inflammation (19), our findings suggest that
the neutrophil is a key cell type altered when agents of this class are
given to interrupt acute inflammatory syndromes (51).
During preparation of this manuscript other investigators reported that
human PMNs express COX-2 in response to LPS (37). Although our study
yielded the same result, our observations on the regulation of COX-2 in
stimulated PMNs differ substantially from those in the earlier report.
We found maximal accumulation of COX-2 at 3 h after stimulation
with LPS with waning levels of the enzyme thereafter (Fig. 1) in the
absence of concomitant treatment with a chemokine or cytokine (Fig. 4, 5). In contrast, Niiro et al. reported that COX-2 progressively
accumulates over a 24-h incubation in response to LPS (37), a time
course similar to that we observed in stimulated monocytes (Fig. 1B). The reasons for this difference are unknown but may
involve variables in the isolated cell preparations, the incubation
conditions used, or differences in the Abs used. We employed a mAb
against human COX-2 whereas Niiro et al. used an Ab raised against rat
COX-2. We also observed that IL-4 and IL-10 were much less effective
inhibitors of COX-2 than was reported by Niiro et al. (37), a
difference that may be related to the specific times at which
inhibition was assayed in our experiments compared to theirs (see
below). Regardless of the specific reasons contributing to these
differences, our findings demonstrate that the temporal expression and
other aspects of the regulation of COX-2 in human PMNs can vary
considerably from the patterns reported earlier (37).
Certain features of the expression of COX-2 in neutrophils are similar
to those in other cell types in which it has been found (18, 19). It is
not present under resting conditions, but the enzyme is rapidly
synthesized when the PMNs are stimulated. The agonists that we found to
induce or enhance expression of COX-2 include cytokines (TNF-, IL-1,
and GM-CSF) a chemokine (IL-8), and a phorbol ester in addition to LPS.
This spectrum is consistent with the stimuli that induce COX-2 in other
cell types (18, 19), as well as with a role for COX-2 in neutrophils as
an effector enzyme in the biochemical pathways of acute inflammation.
As in a number of other cell types (19), the rapid expression of COX-2
in PMNs stimulated with LPS was followed by a rapid return toward
baseline in our experiments (Fig. 1). This temporal pattern is
presumably a result of regulated interruption of the synthetic response
and turnover of the newly synthesized protein, and is similar to that
in the inflammatory exudate in an in vivo model of pleural inflammation
during the period dominated by PMN influx (52). The transient pattern
of COX-2 expression in neutrophils treated with LPS as a single agonist
(Fig. 1A) is quite different from that in monocytes
stimulated under identical conditions (Fig. 1B), indicating
differential regulation of the enzyme in these two cell types of the
myeloid lineage. Similarly, we found differences in the temporal
pattern, the degree of expression at the protein level, and other
features when we examined several different primary human myeloid cell
types and cell lines (see Results) (C. G. Maloney and
W. Kutchera, unpublished observations). Others have also observed
differences in regulation of COX-2 in different, but related, myeloid
cells (32, 53). The factors that account for these differences are not
yet clear. Messenger RNA for COX-2 is reported to be present in PMNs
from subjects with allergic asthma (54). In certain passed cells and
cell lines, the regulation of COX-2 occurs by both transcriptional and
post-transcriptional mechanisms (23) (D. Dixon et al., manuscript in
preparation). Whether such mechanisms apply to neutrophils remains to
be determined. We have begun to examine transcription and handling of
mRNA for COX-2 in these cells (C. G. Maloney, unpublished
observations). In this study, however, we focused on its expression at
the protein level since in myeloid leukocytes mRNA for COX-2 and other
inducible factors may be transcribed without progression to translation
of the corresponding protein (55). Of interest, we found that
concomitant stimulation of PMNs with LPS and TNF- caused the
temporal pattern of COX-2 protein expression to be altered (Fig. 4)
from that in neutrophils treated with LPS alone (Fig. 1A),
and to approximately the pattern seen in monocytes treated with LPS as
a single agonist (Fig. 1B). This suggests that regulatory
mechanisms similar to those in monocytes are inducible in PMNs. Such
mechanisms may account for the prolonged time course of COX-2
expression in PMNs reported by Niiro et al. (37).
The factors that negatively modify expression of inducible enzymes in
myeloid leukocytes, including NSAIDs, glucocorticoids, and endogenous
signaling molecules such as IL-4 and IL-10 (Fig. 7) may be critical
variables that influence the course of acute inflammatory syndromes. We
found that exogenous IL-4 and IL-10 completely inhibited COX-2
synthesis by monocytes. In contrast, IL-10 had no effect and IL-4 only
partially attenuated COX-2 protein expression in LPS-stimulated PMNs
(Fig. 7). These results also indicate differential regulation of COX-2
in the myeloid cells. The lack of inhibition of COX-2 accumulation by
IL-4 and IL-10 in our studies compared to those of Niiro et al. (37)
may be due to the fact that we measured the protein at 3 h (Fig. 7), the maximal time point of expression, rather than at 18 to 24
h when its levels are falling under our conditions (Fig. 1).
It has been considered for some time that PMNs release PGE2
and/or TXA2 into inflammatory exudates, which may account
for neutrophil-dependent alterations in vasomotor tone, in platelet
activation and aggregation, and additional physiologic alterations in
endotoxemia and other inflammatory syndromes (51). However, in other
studies arachidonic acid was dominantly metabolized to products of the
lipoxygenase pathway in freshly isolated human PMNs, with no PGs or TXs
produced (56). This implies that constitutively present COX-1 may
contribute little to the eicosanoid output of PMNs under these
conditions. Our observation that generation of PGE2 by
stimulated PMNs is temporally correlated with expression of COX-2
(Figs. 3, 4 and 6), and similar observations by others (37), indicate
that induction of this isoform is a critical variable. In this regard
the neutrophil illustrates the point that COX-1 and COX-2 may have
differential access to arachidonate pools (14) and phospholipases, but
is different from certain other cell types in which induction of COX-2
does not substantially alter prostanoid output (27). Depending on the
time of incubation, the nature of the stimulus, and other variables
(see Results) expression or modulation of the
"downstream" enzymes PGE2 synthase and/or TX synthase
may also be involved in prostanoid synthesis by PMNs. Our preliminary
experiments suggest that TX synthase is induced in neutrophils
stimulated with LPS (W. A. Kutchera and C. G. Maloney, unpublished
observations).
Documentation that human neutrophils can synthesize new enzymes, here shown for COX-2, expands understanding of their biochemical potentials in acute inflammation. This finding also suggests the possibility of interactions in which the product of one pathway affects the output of another. For example, reactive oxygen species, the products of constitutive enzymes in PMNs (10, 7), modulate the expression of COX-2 in certain cells (57) and may do so in myeloid leukocytes (22). Also, endogenously generated nitric oxide modulates expression of COX-2 in murine macrophages (58) and may have a similar effect in human neutrophils. Reactive oxygen species can directly alter the eicosanoid metabolites of COX-2 in neutrophils (59). Our characterization of COX-2 expression in neutrophils in vitro provides a basis on which to explore such complex interactions, and to further define the role of the enzyme in acute inflammation in vivo.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Guy A. Zimmerman, University of Utah, CVRTI, 95 S 2000 E BACK, Salt Lake City, UT 84112-5000.
3 Abbreviations used in this paper: PMN, polymorphonuclear leukocyte; COX, cyclooxygenase; ARDS, acute respiratory distress syndrome; LT, leukotriene; GM-CSF, granulocyte-macrophage CSF; TX, thromboxane; ECL, enhanced chemiluminescence.
Received for publication May 20, 1997. Accepted for publication October 16, 1997.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and lipopolysaccharide. J. Clin. Invest. 95:1669.
This article has been cited by other articles:
|
|
 |
|
  S. Negrotto, E. Malaver, M. E. Alvarez, N. Pacienza, L. P. D'Atri, R. G. Pozner, R. M. Gomez, and M. Schattner Aspirin and Salicylate Suppress Polymorphonuclear Apoptosis Delay Mediated by Proinflammatory Stimuli J. Pharmacol. Exp. Ther., November 1, 2006; 319(2): 972 - 979. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Vega, P. Chacon, G. Alba, R. El Bekay, J. Monteseirin, J. Martin-Nieto, and F. Sobrino Modulation of IgE-dependent COX-2 gene expression by reactive oxygen species in human neutrophils J. Leukoc. Biol., July 1, 2006; 80(1): 152 - 163. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. Wang, A. Bergh, and J.-E. Damber Cyclooxygenase-2 Expression Correlates with Local Chronic Inflammation and Tumor Neovascularization in Human Prostate Cancer Clin. Cancer Res., May 1, 2005; 11(9): 3250 - 3256. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J.-S. Cadieux, P. Leclerc, M. St-Onge, A.-A. Dussault, C. Laflamme, S. Picard, C. Ledent, P. Borgeat, and M. Pouliot Potentiation of neutrophil cyclooxygenase-2 by adenosine: an early anti-inflammatory signal J. Cell Sci., April 1, 2005; 118(7): 1437 - 1447. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W.-S. Cho and C. Chae Expression of Nitric Oxide Synthase 2 and Cyclooxygenase-2 in Swine Experimentally Infected with Actinobacillus pleuropneumoniae Vet. Pathol., November 1, 2004; 41(6): 666 - 672. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. F. Liberty, L. Raichel, Z. Hazan-Eitan, I. Pessach, N. Hadad, F. Schlaeffer, and R. Levy Cytosolic phospholipase A2 is responsible for prostaglandin E2 and leukotriene B4 formation in phagocyte-like PLB-985 cells: studies of differentiated cPLA2-deficient PLB-985 cells J. Leukoc. Biol., July 1, 2004; 76(1): 176 - 184. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. J. Trouba and D. R. Germolec Micromolar Concentrations of Sodium Arsenite Induce Cyclooxygenase-2 Expression and Stimulate p42/44 Mitogen-Activated Protein Kinase Phosphorylation in Normal Human Epidermal Keratinocytes Toxicol. Sci., June 1, 2004; 79(2): 248 - 257. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. W. Lindemann, C. C. Yost, M. M. Denis, T. M. McIntyre, A. S. Weyrich, and G. A. Zimmerman Neutrophils alter the inflammatory milieu by signal-dependent translation of constitutive messenger RNAs PNAS, May 4, 2004; 101(18): 7076 - 7081. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Cipollone, B. Rocca, and C. Patrono Cyclooxygenase-2 Expression and Inhibition in Atherothrombosis Arterioscler Thromb Vasc Biol, February 1, 2004; 24(2): 246 - 255. [Abstract] [Full Text]
|
|
|
|
 |
|
 E. F.J. de Vries, A. van Waarde, A. R. Buursma, and W. Vaalburg Synthesis and In Vivo Evaluation of 18F-Desbromo-DuP-697 as a PET Tracer for Cyclooxygenase-2 Expression J. Nucl. Med., October 1, 2003; 44(10): 1700 - 1706. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W.-S. Cho and C. Chae Expression of Cyclooxygenase-2 in Swine Naturally Infected with Actinobacillus pleuropneumoniae Vet. Pathol., January 1, 2003; 40(1): 25 - 31. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Chen, H. He, E. F. Mondejar, and G. Hedenstierna Cyclooxygenase inhibitor blocks rebound response after NO inhalation in an endotoxin model Am J Physiol Heart Circ Physiol, January 1, 2003; 284(1): H290 - H298. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. L. Podolin, B. J. Bolognese, J. J. Foley, D. B. Schmidt, P. T. Buckley, K. L. Widdowson, Q. Jin, J. R. White, J. M. Lee, R. B. Goodman, et al. A Potent and Selective Nonpeptide Antagonist of CXCR2 Inhibits Acute and Chronic Models of Arthritis in the Rabbit J. Immunol., December 1, 2002; 169(11): 6435 - 6444. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. R. Clark, M. J. Coffey, R. M. Maclean, P. W. Collins, M. J. Lewis, A. R. Cross, and V. B. O'Donnell Characterization of Nitric Oxide Consumption Pathways by Normal, Chronic Granulomatous Disease and Myeloperoxidase-Deficient Human Neutrophils J. Immunol., November 15, 2002; 169(10): 5889 - 5896. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Pouliot, M.-E. Fiset, M. Masse, P. H. Naccache, and P. Borgeat Adenosine Up-Regulates Cyclooxygenase-2 in Human Granulocytes: Impact on the Balance of Eicosanoid Generation J. Immunol., November 1, 2002; 169(9): 5279 - 5286. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Murakami, D. Takahashi, T. Kinoshita, K. Koshimizu, H. W. Kim, A. Yoshihiro, Y. Nakamura, S. Jiwajinda, J. Terao, and H. Ohigashi Zerumbone, a Southeast Asian ginger sesquiterpene, markedly suppresses free radical generation, proinflammatory protein production, and cancer cell proliferation accompanied by apoptosis: the {alpha},{beta}-unsaturated carbonyl group is a prerequisite Carcinogenesis, May 1, 2002; 23(5): 795 - 802. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. S. Whittaker, K. S. Bahjat, L. L. Moldawer, and M. J. Clare-Salzler Autoregulation of Human Monocyte-Derived Dendritic Cell Maturation and IL-12 Production by Cyclooxygenase- 2-Mediated Prostanoid Production J. Immunol., October 15, 2000; 165(8): 4298 - 4304. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Murakami, Y. Nakamura, T. Tanaka, K. Kawabata, D. Takahashi, K. Koshimizu, and H. Ohigashi Suppression by citrus auraptene of phorbol ester- and endotoxin-induced inflammatory responses: role of attenuation of leukocyte activation Carcinogenesis, October 1, 2000; 21(10): 1843 - 1850. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Notoya, D. V. Jovanovic, P. Reboul, J. Martel-Pelletier, F. Mineau, and J.-P. Pelletier The Induction of Cell Death in Human Osteoarthritis Chondrocytes by Nitric Oxide Is Related to the Production of Prostaglandin E2 Via the Induction of Cyclooxygenase-2 J. Immunol., September 15, 2000; 165(6): 3402 - 3410. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. F. Pereira, L. A. Boven, J. Middel, J. Verhoef, and H. S. L. M. Nottet Induction of cyclooxygenase-2 expression during HIV-1-infected monocyte-derived macrophage and human brain microvascular endothelial cell interactions J. Leukoc. Biol., September 1, 2000; 68(3): 423 - 428. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. N. M. Khan, K. Stanfield, D. Trajkovic, and R. K. Harris Cyclooxygenase-2 Expression in Inflammatory Lung Lesions of Nonhuman Primates Vet. Pathol., September 1, 2000; 37(5): 512 - 516. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Yang, D. Sun, Y. G. Huang, A. Smart, J. P. Briggs, and J. B. Schnermann Differential regulation of COX-2 expression in the kidney by lipopolysaccharide: role of CD14 Am J Physiol Renal Physiol, July 1, 1999; 277(1): F10 - F16. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. J. Czermak, V. Sarma, N. M. Bless, H. Schmal, H. P. Friedl, and P. A. Ward In Vitro and In Vivo Dependency of Chemokine Generation on C5a and TNF-{alpha} J. Immunol., February 15, 1999; 162(4): 2321 - 2325. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. H. Pillinger, C. Capodici, P. Rosenthal, N. Kheterpal, S. Hanft, M. R. Philips, and G. Weissmann Modes of action of aspirin-like drugs: Salicylates inhibit Erk activation and integrin-dependent neutrophil adhesion PNAS, November 24, 1998; 95(24): 14540 - 14545. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Pouliot, C. Gilbert, P. Borgeat, P. E. Poubelle, S. Bourgoin, C. Créminon, J. Maclouf, S. R. Mccoll, and P. H. Naccache Expression and activity of prostaglandin endoperoxide synthase-2 in agonist-activated human neutrophils FASEB J, September 1, 1998; 12(12): 1109 - 1123. [Abstract] [Full Text]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
