| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
|
FJ
EXPRESS SUMMARY ARTICLE The Full-length version of this article is also available, published online June 18, 2001 as doi:10.1096/fj.00-0670fje. |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Unit of Critical Care,
Department of Cardiothoracic Surgery, The Royal Brompton and Harefield N.H.S. Trust, Imperial College School of Medicine, London SW3 6NP, UK
2Correspondence: Unit of Critical Care, The Royal Brompton & Harefield N.H.S. Trust, Imperial College School of Medicine, Sydney St., London SW3 6NP, UK. E-mail: j.a.mitchell{at}ic.ac.uk
SPECIFIC AIMS
The colony-stimulating factors (CSF) are important inhibitors of leukocyte apoptosis. We have recently shown that cytokines stimulate human vascular smooth muscle cells (HVSMCs) to release granulocyte macrophage-CSF (GM-CSF) and granulocyte-CSF (G-CSF)-like immunoreactivity and that the inflammatory form of cyclo-oxygenase (COX-2) differentially modulates their release. Using human neutrophil apoptosis as a bioassay, our aim was to investigate the biological relevance of these observations.
PRINCIPAL FINDINGS
1. GM-CSF and G-CSF release from stimulated HVSMCs is
differentially regulated by COX-2
Under basal conditions, HVSMCs cultured from the saphenous vein
released undetectable levels of both GM-CSF and G-CSF, as measured by
ELISA. The release of both CSFs was stimulated in the presence of
IL-1ß (1.0 ng/ml), G-CSF (4375±919 pg/ml, n=8) being
released in higher concentrations than GM-CSF (86±11 pg/ml,
n=8). Treatment of HVSMCs with the COX-1/COX-2 inhibitor
indomethacin (1x10-5 M) or the selective COX-2
inhibitor DFU (1x10-5 M) before stimulation
with IL-1ß significantly potentiated the release of GM-CSF
(indomethacin 408±23, n=8: DFU 421±31 pg/ml,
n=8) while inhibiting the release of G-CSF (indomethacin
1271±205 pg/ml, n=8: DFU 1518±338 pg/ml, n=8).
No significant difference was observed between the effects of
indomethacin vs. DFU on CSF release. These findings, confirming our
previous observations, implicate COX-2 in the regulation of CSF
release.
2. GM-CSF and G-CSF inhibit neutrophil apoptosis
Neutrophil apoptosis, assessed using an ELISA to determine levels
of histone-associated mono- and oligonucleosomes in the cytoplasmic
fraction of cell lysates, was inhibited by both GM-CSF and G-CSF at 1,
6, 24, and 48 h time points. The greatest sensitivity of the
apoptotic response to inhibition by GM-CSF and G-CSF was seen at 6 h. The ability of the two CSFs to inhibit neutrophil apoptosis at the
6 h time point was confirmed by DNA laddering. GM-CSF was found to
be a more potent inhibitor of neutrophil apoptosis than G-CSF
(Fig. 1
). This finding is important in itself since many inflamed tissues
release G-CSF in higher levels than GM-CSF, leading to the suggestion
in the literature that G-CSF is the more important neutrophil-promoting
factor.
|
3. Conditioned medium removed from unstimulated HVSMCs inhibits
neutrophil apoptosis
Despite containing undetectable levels of GM-CSF and G-CSF,
conditioned medium removed from unstimulated HVSMCs inhibited
neutrophil apoptosis (by 41.5±3.2%, n=6). When HVSMCs were
treated with indomethacin (1x10-5 M),
levels of GM-CSF and G-CSF released remained undetectable. No
significant difference was observed between the effects of conditioned
medium removed from unstimulated and indomethacin-treated HVSMCs.
Although completely reversing the effects of exogenous GM-CSF
(0.01 ng/ml) and G-CSF (0.1 ng/ml) on neutrophil apoptosis,
neutralizing anti-GM-CSF (10 µg/ml) and -G-CSF (10 µg/ml)
antibodies had no effect on the level of neutrophil apoptosis seen in
the presence of conditioned medium from unstimulated HVSMCs or from
HVSMCs treated with indomethacin alone. We conclude that the
anti-apoptotic activity of conditioned medium removed from unstimulated
or indomethacin-treated cells is not attributable to either GM-CSF or
G-CSF; the identity of the mediator(s) responsible remains to be
elucidated.
4. Conditioned medium removed from stimulated HVSMCs further
inhibits neutrophil apoptosis
Levels of both GM-CSF and G-CSF were elevated in conditioned
medium removed from HVSMCs stimulated with IL-1ß. In line with this,
conditioned medium taken from IL-1ß-stimulated HVSMCs inhibited
neutrophil apoptosis to a significantly greater extent than conditioned
medium removed from unstimulated HVSMCs (Fig. 2A
). DNA laddering confirmed that conditioned medium removed
from IL-1ß-stimulated HVSMCs inhibited neutrophil apoptosis. Despite
alterations in CSFs levels induced by indomethacin, the ability of
conditioned medium to modulate apoptosis of neutrophils was not
affected by this treatment (Fig. 2B
).
|
5. Anti-GM-CSF and -G-CSF antibodies differentially reverse the
effects of conditioned medium removed from stimulated HVSMCs
Anti-GM-CSF antibody had no significant effect on the reduced
level of neutrophil apoptosis seen in the presence of conditioned
medium removed from HVSMCs stimulated with IL-1ß (Fig. 2A
). Under the same conditions, anti-G-CSF antibody
significantly increased the level of apoptosis. A combination of the
two antibodies further increased the level of apoptosis (Fig. 2A
). In contrast, where neutrophils were treated with
conditioned medium from IL-1ß-stimulated HVSMCs pretreated with
indomethacin, anti-GM-CSF antibody but not anti-G-CSF antibody
significantly increased the level of apoptosis (Fig. 2B
).
Again, a combination of the two antibodies further increased the level
of apoptosis (Fig. 2B
). This finding indicates not only that
levels of CSFs released by HVSMCs are biologically important, but that
when COX-2 activity is blocked, the balance of biological activity for
the CSFs shifts from G-CSF to GM-CSF.
CONCLUSIONS
CSFs are responsible for the proliferation and differentiation of
cells in the bone marrow and modulate the function of circulating,
mature leukocytes, including neutrophils. GM-CSF and G-CSF promote the
activation and survival of mature neutrophils. Thus, understanding how
the release of these two cytokines is regulated has important
implications for the inflammatory and immune process as well as
pathogenesis of immunological diseases. In the current study, we have
confirmed our previous findings that stimulated HVSMCs corelease
GM-CSF-like and G-CSF-like immunoreactivity. We again demonstrate that
the release of these two CSFs is differentially regulated by COX-2
activity, which simultaneously inhibits GM-CSF release and promotes
G-CSF release from stimulated cells. We have used neutrophil survival
as a bioassay to investigate the biological activity of the CSFs
produced by HVSMCs and how this activity might be effected by the
inhibition of COX. Here we show for the first time that 1)
levels of CSFs released by HVSMCs are biologically important, with
G-CSF being predominantly active; 2) GM-CSF and G-CSF act
synergistically to inhibit neutrophil apoptosis; and 3) when
COX-2 activity is blocked the balance of biological activity for the
CSFs shifts from G-CSF to GM-CSF (Fig. 3
).
|
IL-1ß-stimulated HVSMCs released both GM-CSF and G-CSF, levels of G-CSF released being anything from 10- to 50-fold higher than GM-CSF. Reports in the literature showing that, under inflammatory conditions, G-CSF is produced in higher quantities than GM-CSF have led to the general assumption that G-CSF is the prime CSF responsible for modulating neutrophil survival. However, few studies have compared the relative potencies of GM-CSF vs. G-CSF. In this study, we found GM-CSF to be a more potent inhibitor of neutrophil apoptosis than G-CSF. Therefore, even though G-CSF is released in higher levels, GM-CSF may be equally or, under some conditions, more important in determining neutrophil survival.
Even though conditioned medium taken from unstimulated HVSMCs contained undetectable levels of CSFs, it inhibited neutrophil apoptosis. This property was not attributable to either GM-CSF or G-CSF, and the identity of the mediator(s) responsible remains to be elucidated. However, in line with our observations of CSF immunoreactivity, conditioned medium removed from stimulated HVSMCs contained a greater level of anti-apoptotic activity than that removed from unstimulated cells. This could be inhibited with anti-G-CSF (but not anti-GM-CSF) polyclonal antibodies, suggesting that in these settings (i.e., when G-CSF is present in higher levels) and despite the differences in potency mentioned above, G-CSF is more important than GM-CSF. However, we did observe an apparent synergistic effect of GM-CSF and G-CSF in the conditioned medium on neutrophil survival.
Both here and in previous studies, we have demonstrated that the release of GM-CSF-like and G-CSF-like immunoreactivity from HVSMCs is differentially regulated by COX-2. GM-CSF release is suppressed and G-CSF release potentiated by COX-2 activity. Thus, we suggest that during inflammatory events, inhibition of COX-2 may tip the balance of CSF production in the favor of GM-CSF. To support this, we found that despite no significant overall difference in the level of anti-apoptotic activity, the effect of conditioned medium removed from stimulated HVSMCs pretreated with indomethacin was reversed by anti-GM-CSF, but not anti-G-CSF, polyclonal antibodies. Nevertheless, GM-CSF levels in the conditioned medium were still lower than those of G-CSF, further supporting our observation that GM-CSF is a more potent inhibitor of neutrophil apoptosis than G-CSF. Again, a combination of antibodies to the two CSFs had a synergistic effect on apoptosis. In addition to our observations here using neutrophils (which respond to both GM-CSF and G-CSF), we suggest that our data may be extrapolated to include other leukocytes (e.g., monocytes and eosinophils). Indeed, we predict that the ability of COX-2 inhibitors to increase GM-CSF and decrease G-CSF release will result in a change in the dynamics of the leukocyte population present at the site of inflammation.
Thus, we have demonstrated that HVSMCs release factors that inhibit neutrophil apoptosis and that, when stimulated with IL-1ß, this property is explained by the combined release of GM-CSF and G-CSF. When COX-2 is inhibited, the reduction in G-CSF is counterbalanced by an increase in GM-CSF and neutrophil survival is preserved. Under these conditions, however, GM-CSF, which normally is not associated with an active role in neutrophil survival, is the active factor.
These observations are important because they validate the newly emerging hypothesis that COX activity directly regulates the release of CSFs. This has clear implications for our understanding of the immunological role of COX-2. Our data also implicate an important role for the vascular smooth muscle in the release of CSFs, a function previously attributed to the endothelium. Although our observations are made using vascular cells, we believe that the phenomenon we report is ubiquitous and may have implications for the effects of COX-2 inhibitors in patients with immunologically driven diseases.
FOOTNOTES
1 To read the full text of this article, go
to http://www.fasebj.org/cgi/doi/10.1096/fj.00-0670fje ; to
cite this article, use FASEB J. (June 18, 2001)
10.1096/fj.00-0670fje 
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
