(Received for publication, November 22, 1996, and in revised form, January 6, 1997)
From the Department of Immunology, The Scripps Research Institute, La Jolla, California 92037
Chemoattractants induce a variety of phagocytic
functions including transendothelial migration, degranulation, and the
generation of superoxide anions. We report here that the prototypic
chemotactic peptide fMet-Leu-Phe (fMLF) stimulates the activation of
nuclear factor-B (NF-
B), a transcription factor that is central
to the regulation of proinflammatory immediate-early gene expression. In freshly prepared peripheral blood mononuclear cells, fMLF (1-100 nM) induced a
B binding activity that was
receptor-dependent and involved the p50 and p65 subunits of
the NF-
B/Rel family of proteins. The activation of NF-
B by fMLF
appeared to be cell-specific and different from the activation of
NF-
B by tumor necrosis factor-
(TNF
). Neutrophil preparations
that responded to fMLF, TNF
, and lipopolysaccharides with
interleukin-8 secretion did not show NF-
B activation, whereas
N-formyl peptide receptor (FPR)-transfected HL-60 cells
were responsive to TNF
but not fMLF for NF-
B activation. Differentiation of FPR-transfected HL-60 cells with dimethyl sulfoxide for 3-5 days conferred the capability of the cells to activate NF-
B
in response to fMLF without a significant increase in the amount of
FPR. These results identify NF-
B as a transcription factor that can
be activated by the prototypic chemotactic peptide and demonstrate that
this function is both highly regulated and dependent on signaling
components specifically expressed during myeloid differentiation.
The coordinated regulation of gene expression by phagocytic
leukocytes is generally considered to be essential to the early inflammatory response in humans. The production of cytokines, chemokines, cell-surface receptors/adhesion proteins, and other molecules by activated monocytes, macrophages, and neutrophils can
serve to orchestrate immune cell development and is essential to host
defense against invading microorganisms. The regulation of gene
expression in these cells is governed by the activities of
transcription factors such as nuclear factor-kappa B
(NF-B),1 NF-IL-6, and AP-1, which are
themselves regulated by external stimuli. NF-
B is of paramount
importance to immune cell function owing to its ability to activate the
transcription of many proinflammatory immediate-early genes (1, 2).
NF-
B, which has been studied in great detail and shown to be dimeric
in structure, is composed of members of the Rel family of
transcriptional activators. In dormant cells NF-
B is normally
complexed with a member of the I
B proteins, which localizes the
transcription factor to the cytosol in an inactive state (3).
Stimulation of the cells with ligands such as lipopolysaccharide (LPS),
tumor necrosis factor-
(TNF
), and IL-1
results in rapid
dissociation of I
B and subsequent entry of the active NF-
B to the
nucleus where it can interact with DNA. Although the signaling
mechanisms leading to activation of NF-
B are not completely
understood, it is widely believed that phosphorylation and degradation
of I
B is pivotal to this process (4-6).
More recent studies have demonstrated activation of NF-B by G
protein-coupled receptors (7-10). The lipid-derived chemoattractants platelet-activating factor (PAF) and leukotriene B4
(LTB4) can activate NF-
B in monocytes (9, 11, 12), and
in transfected cell lines expressing the PAF receptor (10).
Furthermore, this activation of NF-
B in monocytes results in the
transcriptional activation of genes encoding cytokines and growth
factors (9, 12). The ability of the lipid-derived chemoattractants to
modulate gene expression through NF-
B reflects a potentially novel
function of leukocyte chemoattractants in the inflammatory process.
However, the scope of G protein-mediated NF-
B activation in
leukocytes and its downstream signaling components remain to be
determined.
N-Formylated peptides that may be released from invasive
bacteria have been one of the most extensively studied
chemoattractants. Much work has focused upon structural and functional
analyses of the heptahelical cell-surface receptor for these peptides
using the prototypic agonist
N-formyl-methionyl-leucyl-phenylalanine (fMLF). Stimulation
of monocytes and neutrophils with fMLF induces cellular functions
including directed cell movement, phagocytosis, and the generation of
reactive oxygen intermediates (13). A variety of signaling pathways
leading from receptor activation of pertussis toxin-sensitive G
proteins that contribute to the regulation of these processes have
been documented (14). More recent studies have demonstrated that
peptide chemoattractants such as fMLF and C5a can stimulate cytokine
secretion by monocytes, neutrophils and eosinophils (15, 16). With
eosinophils, the involvement of the transcription factor NF-B was
suggested although no direct evidence for NF-
B activation was
provided (16). We investigated the ability of fMLF to activate NF-
B
in myeloid cells and demonstrate here that this peptide chemoattractant
can activate NF-
B in human monocytes but not in neutrophils or in the promyelocytic HL-60 cells. Differentiation of HL-60 cells with
dimethyl sulfoxide (Me2SO) resulted in the acquisition of fMLF-stimulated activation of NF-
B that was independent of receptor up-regulation. In addition, transfected HL-60 cells expressing functional FPR were responsive to TNF
but not to fMLF for NF-
B activation, suggesting the involvement of separable signaling pathways
leading to transcription factor activation. These findings indicate
that activation of NF-
B may be a general function of chemoattractants but is limited to specific classes of terminally differentiated myeloid cells. Furthermore, this work highlights the
utility of HL-60 cells as a model for a more detailed investigation of
the signaling mechanisms leading to NF-
B activation by different ligands.
LPS isolated from Salmonella minnesota
Re595 was a gift from R. Ulevitch (Scripps Research Institute).
Recombinant murine TNF was kindly provided by V. Kravchencko
(Scripps Research Institute). The recombinant TNF
contains less than
0.277 ng/mg LPS as measured by the Limulus amebocyte lysate
assay (Cap Cod Associates, Woods Hole, MA). Pyrrolidinedithiocarbamate
(PDTC), purchased from Sigma, was kept as a 1 mM stock in
H2O and was added to cultures directly. FITC-fMLF
(Molecular Probes, Eugene, OR) and unlabeled fMLF (Sigma), were diluted
with posphate-buffered saline from concentrated stocks maintained at
20 °C in Me2SO. A double-stranded oligonucleotide probe containing a decameric
B sequence (underlined)
5
-AGTTGAG
CAGG-3
was labeled using
polynucleotide T4-kinase (Life Technologies, Inc.) and
[
-32P]ATP (10 mCi/ml, Amersham Corp.). Rabbit
polyclonal antibodies against a panel of NF-
B/Rel family of proteins
were purchased from Santa Cruz Biotechnology (Santa Cruz, CA).
Anti-CD14 monoclonal antibody was from Coulter Immunology (Miami, FL),
and phycoerythrin conjugated goat anti-mouse IgG was purchased from
Sigma. Detection of IL-8 in culture supernatants was carried out using
a commercially available enzyme-linked immunosorbent assay kit
according to the procedures suggested by the manufacturer (Biosource
International, Camarillo, CA).
Human peripheral blood leukocytes were fractionated using Percoll based on the method of Ulmer and Flad (17). Briefly, blood was collected from healthy donors using acid citrate buffer containing 2% dextrose as an anticoagulant. Erythrocytes were removed by sedimentation with 6% hetastarch (Baxter). Peripheral blood mononuclear cells (PBMC) and neutrophils in the supernatant were further separated by centrifugation at 450 × g at 10 °C for 40 min through Percoll step gradients (70 and 55%). Viability of cells in a routine preparation was approximately 98% as determined by trypan blue exclusion. Following a wash in fresh RPMI 1640 medium (serum-free), cells (5 × 106) were stimulated with various ligands in 1.5-ml microcentrifuge tubes at 37 °C.
PBMCs prepared from several healthy donors were found to contain three
populations of cells as identified by flow cytometry on forward and
side scatter dot plots: platelets, lymphocytes and monocytes (Fig.
1). Using CD14 as a cell-surface marker for the
monocytes, we observed that only these cells, but not platelets or
lymphocytes, express the FPR as detected by FITC-fMLF direct binding
(Fig. 1). Thus, PBMC was used without further fractionation, and any
responses to fMLF could be attributed to binding and stimulation of monocytes in these preparations.
The transfection of HL-60 cells with the FPR has been detailed elsewhere (18). Both transfected and untransfected HL-60 cells were maintained at 37 °C with 5% CO2 in RPMI 1640 medium supplemented with 10% heat-inactivated fetal bovine serum, 2 mM L-glutamine, 10 mM HEPES (pH 7.0), penicillin (100 IU/ml), and streptomycin (50 mg/ml). HL-60 cells (5 × 105 cells/ml) were differentiated with Me2SO (1.3% v/v) over a period of 5 days, and the expression of the FPR was determined as described below.
Cell Staining and Flow CytometryAll flow cytometry measurements were performed on a FACScan flow cytometer (Becton Dickinson, Mountain View, CA). Determinations of fMLF receptor (FPR) expression were made by incubating cells (105) on ice for 10 min followed by the addition of FITC-conjugated fMLF (10 nM) for an additional 10 min. Quantitation of receptor expression on the whole cells was accomplished by relating the binding of saturating concentrations of [3H]fMLF to the fluorescence detected by flow cytometry. Subsequent receptor quantities were then determined by converting mean fluorescence by flow cytometry to receptor number. To identify monocytes in mononuclear fractions of human blood, cells were prelabeled with anti-CD14 monoclonal antibody (1:200), washed in posphate-buffered saline, and then incubated with phycoerythrin-conjugated goat anti-mouse IgG (1:500). All incubations were done on ice for 60 min; the cells were washed twice in posphate-buffered saline between each incubation and prior to flow cytometry analysis.
Electrophoretic Mobility Shift AssaysFollowing appropriate
treatment of cells, nuclear extracts were prepared using the method of
Dignam et al.(19) with minor modifications as detailed elsewhere (10).
Prior to separation of DNA·protein complexes, binding reactions were
initiated by adding 1 µl of (approximately 1 × 106
cpm) labeled B probe to a reaction mixture which had been
preincubated for 10 min at room temperature. The mixture contained 2-5
µg of nuclear extract in binding buffer (2.5 mM HEPES, pH
7.8, 2.5 mM MgCl2, 25 mM KCl, 0.25 mM dithiothreitol, 0.2 µg/ml poly(dI-dC), 50 µg/ml
sonicated salmon sperm DNA, and 5% glycerol). After an additional 10 min at room temperature mixtures were loaded onto pre-electrophoresed
(60 min) 6% acrylamide gels and run at 100 V using a 50 mM
Tris borate buffer containing 1 mM EDTA. After separation,
gels were dried on Whatman DE81 paper and exposed to a phosphor imaging
plate (Molecular Dynamics) or x-ray film if necessary. For supershift
analysis, approximately 4 µl of antibodies (1 mg/ml) against NF-
B
subunits were incubated with the binding mixtures for 15 min prior to
the addition of radiolabeled probe.
Whole cell lysates
were solubilized by dilution with 2 × polyacrylamide gel
electrophoresis buffer and 20 µg of protein was separated on 10%
denaturing SDS gels using the method of Laemmli (20). Protein profiles
were then transferred to nitrocellulose membranes (Hybond ECL, Amersham
Corp.) and blocked overnight at 4 °C in PTS buffer (10 mM phosphate buffer, pH 7.0, 0.05% Tween 20, 0.9% NaCl,
and 0.025% NaN3) containing 3% bovine serum albumin. NF-B subunits were detected on the blots by incubation for 1 h
with a 1:1,000 dilution of either anti-p50 or anti-p65 rabbit polyclonal antibodies (Santa Cruz Biotechnology). After 3 washes in PTS
containing 1% bovine serum albumin, blots were probed with 1:5,000
alkaline phosphatase-conjugated goat anti-rabbit IgG for 1 h.
Following extensive washing, immunodetected proteins were visualized by
incubation with alkaline phosphatase substrate (Sigma). Protein
concentrations were determined using the BCA reagents (Pierce).
Recent studies have identified a G protein-coupled signaling
pathway initiated by the LTB4 and PAF receptors that leads
to the activation of NF-B in human monocytes (9, 11, 12). Because
monocytes express receptors for several other chemoattractants including fMLF, we examined whether fMLF could activate a similar pathway in these cells. The addition of a saturating concentration of
fMLF (100 nM) to freshly isolated PBMC resulted in a rapid and transient activation of
B binding activity as measured by EMSA
(Fig. 2A). The DNA·protein complexes were
detected 30 min following addition of fMLF, peaked in amount after 60 min, and subsequently decreased to less than half of the maximum in
4 h. The extent of the induced DNA binding activity was dependent
on the dose of fMLF as the response was detectable with nanomolar concentrations of fMLF and reached a maximum with 50 nM
fMLF (not shown). The effectiveness of nanomolar concentrations of fMLF to induce
B binding activity suggested that it is a primary response to fMLF stimulation mediated by a high affinity receptor. This point
was confirmed by blocking fMLF-induced
B binding activity with
N-cinnamoyl-FLFLF (Fig. 2A), a newly identified
fMLF receptor antagonist.2 Further
fractionation of PBMC by centrifugation through additional Percoll
gradients revealed that the purified monocytes displayed essentially
the same
B binding activity in response to fMLF stimulation (data
not shown).
Because NF-B activation by fMLF has not been previously reported, we
examined the specificity and composition of the resultant
B binding
activity in the stimulated PBMC. Competition with unlabeled
B probe
significantly reduced the major species identified on these gels thus
indicating that the DNA binding activity was specific to the prototypic
B sequence (Fig. 2B). PDTC, an antioxidant inhibitor of
NF-
B that functions by stabilization of the inhibitory protein I
B
(21), was found to block the appearance of the DNA·protein complexes
in fMLF-stimulated cells (Fig. 2B). Gel mobility supershift with specific antibodies is widely used as supportive evidence for the
identification of specific proteins in DNA·protein complexes detected
by EMSA. In order to confirm the identity of the putative NF-
B
species observed in fMLF-treated PBMC, nuclear extracts were
preincubated with equal amounts of antibodies directed against well
documented subunits of the NF-
B/Rel proteins. Gel retardation assays
revealed upward shifting of the DNA·protein complexes only by
antibodies against the p50 and p65 proteins (Fig. 2C),
similar to the supershift pattern observed in other systems (22). These results indicate the presence of p50 and p65 subunits of the
NF-
B/Rel protein family, possibly forming a p50/p50 homodimer and a
p50/p65 heterodimer in the DNA·protein complex induced by fMLF.
Preincubation of nuclear extracts with antibodies specific to p52,
c-Rel, and RelB did not cause a supershift of the DNA·protein complex
(Fig. 2C), but this experiment alone could not completely
rule out the possibility that fMLF may activate other NF-
B/Rel
proteins in these cells. Taken together, results derived from
competition and supershift experiments confirmed fMLF induction of
NF-
B activation in monocytes. The finding that PDTC inhibits the
response suggested that fMLF-stimulated activation of NF-
B involves
I
B degradation similar to the responses reported for other inducers
of NF-
B (4, 21).
It has been well established that monocytes are a major source of
proinflammatory cytokines. More recently it has been reported that
neutrophils also have transcriptional activity and can secrete cytokines including IL-1, IL-6, and IL-8 following activation with
either cytokines such as TNF
or chemoattractants such as C5a and
fMLF (15, 23). Since the expression of IL-1
and IL-8 is regulated to
a great extent by NF-
B at the level of transcription, we reasoned
that stimulation with fMLF might activate NF-
B in these cells. This
notion was examined by comparing the ability of PBMC and neutrophils to
respond to fMLF as well as TNF
and LPS (Fig.
3A). The magnitude of the NF-
B response to
fMLF in PBMC was similar to that of the better characterized inducers such as LPS and TNF
. Surprisingly, none of these agonists were able
to induce detectable NF-
B activation in neutrophils stimulated for
1 h (Fig. 3A) or for several different time periods
from 15 min to 2 h (data not shown).
In order to confirm that the neutrophil preparations were capable of
de novo synthesis and secretion of cytokines, the cells were
treated with fMLF, and the IL-8 protein secreted into the medium was
measured. Stimulation of neutrophils with 100 nM fMLF caused a 4-5-fold increase in the levels of IL-8 protein secreted to
the extracellular medium when compared with unstimulated cells (Fig.
3B). Furthermore, preincubation of cells with actinomycin D
inhibited this increase suggesting that IL-8 protein synthesis in
response to fMLF was dependent on transcription. The same neutrophil preparations were also responsive to TNF and LPS, as both agonists stimulated IL-8 secretion by 8-10-fold, and this increase could be
completely inhibited by actinomycin D (Fig. 3B). These
results demonstrated that the neutrophils used in these experiments
were not deleteriously altered by the fractionation process and
responded similarly to TNF
, LPS, and fMLF.
Due to the lack of NF-B activation in neutrophils following
stimulation with several agonists, it was not clear whether NF-
B subunits were present in these cells. In order to clarify this issue,
polyclonal antibodies against the p50 and p65 subunits of the
NF-
B/Rel protein family were used to probe Western blots prepared
from whole cell lysates of PBMC, neutrophils, and the promyelocytic
cell line HL-60 (Fig. 4). The antibodies identified major species migrating at 50 and 65 kDa, respectively, in PBMC and
HL-60 cells. In contrast, proteins migrating at parallel positions in
lanes containing extracts from neutrophils were barely detectable. This
finding suggested that the lack of NF-
B activation in neutrophils may be attributed to insufficient quantities of the p50 and p65 proteins in these cells.
As a preliminary effort directed at the elucidation of the mechanism of
NF-B activation by fMLF, our attention focused on HL-60 cells as a
potential model. The HL-60 cell is a promyelocytic cell line that can
be induced to differentiate into both monocytic and granulocytic
lineages (24). Treatment of HL-60 cells with Me2SO has been
shown to induce granulocytic differentiation of the promyelocytic cells
(25). Although other reagents such as phorbol myristate acetate and
dibutyryl cyclic AMP can also initiate differentiation of HL-60 cells,
they are capable of activating NF-
B (26) and therefore were not
chosen for our studies. Following Me2SO (1.3% v/v)
treatment, HL-60 cells attained the ability to respond to fMLF with
NF-
B activation, while the undifferentiated cells were unresponsive
to saturating concentrations of fMLF (Fig. 5A). As early as 3 days following
Me2SO treatment, fMLF-dependent NF-
B
activation could be detected, and this response was maximal in cells
allowed to differentiate for 5 days. Me2SO treatment induced very little spontaneous NF-
B activation as the background remained low. Subunit analysis of fMLF-induced NF-
B composition by
supershift assay indicated the presence of both p50 and p65 proteins
similar to that observed in fMLF-stimulated monocytes (Fig.
5A; compare with Fig. 2C). Thus,
Me2SO-differentiated HL-60 cells are fully capable of
activating NF-
B in response to fMLF.
A number of granulocytic functions are acquired by HL-60 cells upon
treatment with Me2SO, including chemotactic and phagocytic competence, inducible superoxide generation, and the increased expression of several chemoattractant receptors. As measured by FITC-fMLF binding, the expression of FPR on the cell surface increased dramatically in Me2SO-treated HL-60 cells after the 3rd day
(Fig. 5B). In order to determine whether expression of FPR
was the essential component responsible for fMLF-inducible NF-B
activation in differentiated cells, we tested FPR cDNA-transfected
HL-60 cells with or without additional Me2SO treatment. As
reported previously, the stably transfected HL-60 cells express
functional FPR without differentiation (18). Stimulation of the
transfected cells with fMLF induces several G protein-coupled functions
including calcium mobilization and actin polymerization (18). However,
in our experiments the transfected HL-60 cells did not exhibit
significant levels of NF-
B activation when stimulated with fMLF
(Fig. 6A). Similarly, only a weak NF-
B response was
observed in HL-60 cells stably transfected to express the C5a receptor,
which responded well to C5a with mobilization of Ca2+ (Fig.
6A). Notably, we observed that TNF
was able to stimulate a potent NF-
B activity in the transfected cells (Fig. 6A)
as well as in untransfected HL-60 cells (see below), indicating that the signaling pathways utilized by TNF
for NF-
B activation were fully functional in these cells. This finding is in line with the
Western blot data indicating that p50 and p65 subunits are present in
the undifferentiated HL-60 cells (Fig. 4). These results suggest that
TNF
stimulation of NF-
B activation utilizes signaling mechanisms
that may be different from the ones employed by the G protein-coupled
FPR in HL-60 cells, which apparently lack the response to fMLF despite
the expression of the FPR by transfection.
The above data indicate that untransfected HL-60 cells lack not only
the FPR but also a component that is necessary for efficient NF-B
activation in response to fMLF. This component may be acquired by
differentiation of the cells with Me2SO, which at the same time also induces the expression of the FPR. To test this hypothesis, we treated the FPR-transfected HL-60 cells with Me2SO for 5 days and measured the capability of the differentiated cells to
activate NF-
B in response to fMLF. As shown in Fig.
7, Me2SO-treated cells displayed marked
NF-
B activation following fMLF stimulation while the background
B
binding activity remained low. A similar result was obtained in C5a
receptor-transfected HL-60 cells, which responded after
Me2SO differentiation (data not shown). To examine whether Me2SO induced a significant increase in the FPR expression
level that contributed to the NF-
B response to fMLF, the cells were subjected to flow cytometry analysis for quantitation of receptor expression. Without Me2SO treatment, FPR-transfected HL-60
cells expressed on average 148,000 receptors/cell based on mean
fluorescence measurement against the FITC standard. Following
differentiation with Me2SO for 5 days, the cells expressed
on average 170,000 receptors/cell (n = 3) (Fig. 7).
This is comparable with untransfected HL-60 cells, which expressed
approximately 151,000 receptors/cells after a similar induction with
Me2SO. Thus, the small increase (15%) in the numbers of
FPR following Me2SO treatment cannot explain the marked
gain in responsiveness to fMLF simulation.
In order to determine whether the increase in responsiveness of NF-B
to fMLF in HL-60 cells after differentiation was specific to the
chemoattractant but not a result of a general increase in activable
NF-
B, HL-60 cells at different stages following treatment with
Me2SO were analyzed for p50 and p65 content by Western blot
and for responsiveness to TNF
by EMSA (Fig. 8). In
these experiments NF-
B activated by TNF
did not increase after
differentiation but appeared to decrease in HL-60 cells treated with
Me2SO for 3 and 5 days. The relative amounts of p50 and p65
proteins visible on Western blots also did not increase in HL-60 cells
following Me2SO induction (Fig. 8). In fact, there appeared
to be a slight decrease in the levels of the p65 protein. Thus, our
data suggest the presence of a myeloid differentiation-associated factor that is necessary for fMLF- but not TNF
-stimulated NF-
B activation in HL-60 cells.
Results from this study provide direct evidence for the activation
of NF-B by fMLF at concentrations comparable with those that
stimulate other phagocyte cellular responses. This function of the well
characterized chemoattractant may be of physiological relevance since
chemoattractants are among the first and major activating factors that
stimulate a migrating phagocyte. The physiological consequences of
NF-
B activation as relevant to gene transcription were not the focus
of this study. However, owing to both qualitative and quantitative
similarities in the fMLF-induced NF-
B response to those elicited by
either TNF
or LPS in monocytes, it is suggested that gene expression
would result as has been well documented for these other ligands.
Studies from this and other laboratories indicate that the
chemoattractants LTB4 and PAF may also stimulate NF-
B-dependent gene expression suggesting that this
might be a general function of leukocyte chemoattractants (9-12).
Leukocyte chemoattractants bind receptors with 7 transmembrane domains
that signal by activating heterotrimeric G proteins. Proximal signaling
events mediated by G proteins differ from those triggered by cytokines
suggesting that chemoattractants might utilize signaling mechanisms
different from those of TNF and IL-1
for activating NF-
B. It
is predicted that both cytokine- and chemoattractant-initiated pathways
converge at some point. As shown here TNF
could activate NF-
B in
undifferentiated HL-60 whereas fMLF could not. The requirement of
differentiation for fMLF responsiveness could thus be explained by the
up-regulation of signaling components that function upstream of the
point of convergence. It is also possible that
chemoattractant-initiated pathways might involve one of the several
signaling mechanisms utilized by TNF
for NF-
B activation;
however, the slight decrease in TNF
-responsive NF-
B during HL-60
differentiation does not support this idea. Thus, results shown here
indicate the presence of specific signaling components that may be
unique to the G protein-mediated pathway(s) leading to NF-
B
activation, reflecting a potentially novel mechanism for activating
this transcription factor. Further study is necessary to identify the
myeloid differentiation-associated factor and to determine whether
myeloid differentiation is required for other chemoattractants to
activate NF-
B.
Neutrophils are the primary effector cells found at sites of acute
inflammation. In spite of the relative scarcity of ribosomes and
endoplasmic reticulum, neutrophils are capable of mRNA synthesis and protein production (27). Neutrophils respond to a large number of
proinflammatory stimuli including fMLF and other classical chemoattractants. It has been demonstrated that the chemoattractants fMLF, C5a, and LTB4 could synergize the effect of LPS on
IL-8 synthesis by neutrophils, although the study did not demonstrate that these chemoattractants themselves could induce IL-8 production (28). Cassatella et al. (15) independently reported release of IL-8 by neutrophils after stimulation with fMLF alone. They observed
de novo IL-8 protein synthesis, accompanied by an
accumulation of the IL-8 mRNA. Furthermore, only nanomolar
concentrations of fMLF were required for stimulation of IL-8 release,
in agreement with our findings. These studies provided direct evidence
for chemoattractant stimulation of cytokine production in neutrophils, but by themselves did not elucidate the transcriptional mechanisms for
chemoattractant-induced cytokine production. Although NF-B is a
primary regulator of IL-8 gene expression in other cell types (29), the
lack of
B binding activity in neutrophils as demonstrated by this
work suggests that NF-
B is unlikely to play a major role in the
transcription of cytokine genes in these terminally differentiated granulocytes. Thus, future investigations of gene transcription in
neutrophils should be focused on the activation of other transcription factors such as NF-IL-6 and AP-1. Activation of the mitogen-activated protein kinases has been demonstrated by others to occur in neutrophils in response to fMLF (30-32). Since the activation of AP-1 can be regulated by the mitogen-activated protein kinase pathways (33), this
transcription factor may have a role in neutrophil gene expression.
Recently Miyamasu et al. (16) reported that fMLF and C5a
could cause secretion of cytokines by eosinophils. Although this response required pretreatment of the cells with cytochalasin B, it was
dependent on transcription. Based on the finding that the antioxidant
PDTC could block IL-8 generation, the authors postulated the
involvement of NF-B in fMLF- and C5a-stimulated response. However,
no direct evidence was shown that supports NF-
B activation in the
stimulated eosinophils. In light of our finding that neutrophils
contain very little p50 and p65 and respond poorly to a number of
agonists that induce NF-
B activation, additional work is necessary
to determine whether eosinophils, a type of terminally differentiated
granulocyte, differ from neutrophils in terms of NF-
B protein
content and responsiveness to chemoattractants.
The finding that activation of NF-B by fMLF in HL-60 cells requires
granulocytic differentiation suggests that the final differentiation
product more closely resembles a transitory state between promyelocyte
and neutrophil. This notion is supported by the recent finding that
certain granulocytic functions are enhanced in HL-60 cells when
retinoic acid is added to the culture medium in the last day of
differentiation with Me2SO (34). Furthermore, our results
indicate that there is no increase in the content of the p50 protein
and a visible decrease of the p65 protein in Me2SO-treated
HL-60 cells undergoing granulocytic differentiation (Fig. 8). Thus,
granulocytic differentiation may eventually result in the depletion of
the NF-
B subunits as was observed here in neutrophils. Recent
studies indicate a function for NF-
B in preventing apoptosis
(35-37), raising the possibility that the loss of NF-
B and the
commitment of neutrophils to cell death may not be just incidental.
Whether chemoattractant-activated NF-
B occurs transiently during
granulocytic differentiation in vivo is presently unknown, but the idea that NF-
B might serve a functional role during the differentiation of granulocytes is intriguing and warrants further study.
We thank Wade Diehl for assistance in blood cell preparation.