(Received for publication, January 24, 1997, and in revised form, April 9, 1997)
From the Vascular Medicine and Atherosclerosis Unit,
Cardiovascular Division, Department of Medicine, Brigham and Women's
Hospital, Harvard Medical School, Boston, Massachusetts 02115, the
¶ Geneva Biomedical Research Institute, 14 chemin des Aulx, 1228 Geneva, Switzerland, the
Department of Immunology and Cell
Biology, Research Center Borstel, 23845 Borstel, Germany, and the
** Martin-Luther-University, Ernst-Grube-Straße 40, 06097 Halle/Saale, Germany
Inflammation contributes to a variety
of arterial diseases including atherosclerosis. Interleukin 1
(IL-1
) in its activated mature 17-kDa form may mediate aspects of
vascular inflammation. As shown previously, human vascular wall cells,
such as smooth muscle cells (SMC), express the IL-1
precursor upon
stimulation and the IL-1
-converting enzyme (ICE) constitutively but
do not produce mature IL-1
or express ICE activity. How SMC, the
most numerous cell type in arteries, may release active IL-1
has
therefore remained a perplexing problem. We report here that
stimulation of human vascular SMC and endothelial cells (EC) through
CD40 ligand, a mediator recently localized in human atheroma, induced elaboration of the IL-1
precursor as well as activation of
cell-associated ICE. In addition to the constitutively expressed 45- and 30-kDa immunoreactive ICE proteins, vascular cells incubated with
recombinant human CD40 ligand (rCD40L) (but not IL-1 or TNF) showed an
increase of a 20-kDa immunoreactive ICE protein by Western blot
analysis. Furthermore, SMC and EC stimulated through rCD40L processed
recombinant human IL-1
precursor (pIL-1
), generating a cleavage
product of approximately 17 kDa. Appearance of both the 20-kDa
immunoreactive ICE protein and pIL-1
processing activity required at
least 6 h of stimulation with 0.3 or 1.0 µg/ml rCD40L,
respectively, and was inhibited by pre-incubation of the ligand with an
anti-CD40L antibody. Stimulation of vascular SMC and EC through rCD40L
resulted in the release of biologically active IL-1
, indicating
processing of the native IL-1
precursor induced by the ligand. These
findings establish a novel mechanism of IL-1
activation in human
vascular cells and, moreover, indicate a new pathway of ICE-activation, which could participate in inflammatory aspects of atherogenesis and
other disease states.
Interleukin 1 (IL-1)1 figures
importantly in many physiological and pathological processes, notably
inflammatory diseases including atherosclerosis. Two distinct genes
give rise to the two IL-1 isoforms denoted IL-1 and IL-1
that
bind to common receptors. Interleukin 1
, often membrane-associated,
can act by contact with neighboring target cells (1). Interleukin 1
,
when secreted in its mature form, can act at a distance in a paracrine
manner. Acquisition of biological activity for IL-1
(but not
IL-1
) requires processing into the mature, 17-kDa protein (2-5).
Upon stimulation, monocytes produce a cell-associated 33-kDa precursor
form of IL-1
(2, 3). Maturation of the IL-1
precursor into the
active 17-kDa form results from cleavage at the
Asp116-Ala117 site by a cysteine proteinase
denoted IL-1
-converting enzyme (ICE) (6-10). ICE in turn is
synthesized as a precursor molecule of 45 kDa, which is thought to be
autocatalytically cleaved to form an active homodimeric enzyme of 20- and 10-kDa subunits ((p20/p10)2) (11, 12). ICE was the
prototype of a group of cysteine proteases, now called the
caspase-family (13). In addition to ICE (caspase-1), this protease
family includes pro-apoptotic enzymes, such as human ICH-1 (caspase-2)
or CPP32 (caspase-3). Each of these homologous enzymes share the active
site cysteine and aspartate binding clefts. Studies of the enzymatic
specificity of ICE demonstrated highly selective proteolytic activity,
i.e. requiring aspartic acid in the P1-position
(9, 14). Interleukin 1
and the apoptotic mediator CPP32 are among
the substrates of ICE (15, 16). Although ICE can autoactivate (17), the
initial mechanisms of activation and regulation of ICE-processing
remain unknown.
Most studies investigating expression of ICE or IL-1 activation have
focused on monocytes or monocyte-derived cell lines. However, normal
arteries contain few if any mononuclear phagocytes. IL-1 derived from
vascular smooth muscle (SMC) and endothelial cells (EC) may initiate
local immune and inflammatory responses and induce expression of
adhesion molecules (18-20) and chemoattractant cytokines,
e.g. IL-8 or IL-1 itself (21-25), that can then recruit the
"professional" phagocytes. In particular, inflammatory components of atherogenesis may involve IL-1 (26, 27). Although human atherosclerotic plaques contain both IL-1
and ICE (28, 29), the mechanisms that activate either the cytokine or the enzyme remain
undefined.
Recent work has demonstrated co-expression of CD40 and its ligand CD40L
in human atherosclerotic plaques, indicating a possible role for this
receptor-ligand pair in vascular pathology (30). CD40L, originally
described as a 33-kDa activation-induced transiently expressed
CD4+ T cell surface molecule (31-34), is also expressed on
macrophages, endothelial cells, and smooth muscle cells (30). Previous
studies of the interactions between CD40L and its receptor CD40
concentrated on the role of these leukocyte-surface proteins in T
cell-dependent B cell differentiation and activation (35).
CD40 ligation regulates a variety of activities, including B cell
growth, differentiation, and death (35, 36), cytokine production by
monocytes (37), and expression of leukocyte adhesion molecules on EC
(38-40). Recent reports from several groups linked CD40/CD40L
interaction to the mechanisms of apoptosis (41-43), a process in which
ICE and other caspases play major roles, as reviewed elsewhere (44). We
therefore tested the hypothesis that CD40L modulates the expression
and/or activity of ICE and thus of IL-1 in cells of the vascular
vessel wall, particularly smooth muscle and endothelial cells.
We demonstrate here that recombinant human CD40L (rCD40L) induces
de novo synthesis of the IL-1 precursor and coordinates expression of a 20-kDa immunoreactive ICE protein with the expression of biological ICE-activity in human vascular smooth muscle and endothelial cells. Moreover, supernatants of the rCD40L-stimulated cultures, but not supernatants from cells exposed to a variety of other
mediators, contained biological IL-1
activity.
Human vascular SMC were isolated
from saphenous veins by explant outgrowth (45) and cultured in
Dulbecco's modified Eagle's medium supplemented with 1%
L-glutamine, 1% penicillin/streptomycin (BioWhittaker,
Walkersville, MD), and 10% FCS (Atlanta Biologicals, Norcross, GA).
Human vascular EC were isolated from saphenous veins by collagenase
treatment (1 mg/ml; Worthington Biochemicals, Freehold, NJ) and
cultured in dishes coated with fibronectin (1.5 µg/cm2;
Upstate Biotechnology Incorporated, Lake Placid, NY) as described elsewhere (46). Cells were maintained in medium 199 (M199;
BioWhittaker), supplemented with 1% penicillin/streptomycin, 5% FCS,
100 µg/ml heparin (Sigma), and 50 µg/ml ECGF (endothelial cell
growth factor; Pel-Freez Biological, Rogers, AR). Both cell types were
subcultured following trypsinization (0.5% trypsin (Worthington
Biochemicals), 0.2% EDTA (EM Science, Gibbstown, NJ)) in
75-cm2 culture flasks (Becton Dickinson, Franklin Lakes,
NJ) and used throughout passages 2 to 4. Culture media and FCS
contained less than 40 pg of endotoxin/ml as determined by chromogenic
Limulus amoebocyte assay-analysis (QLC-1000; BioWhittaker). Human
vascular SMC and EC were characterized by immunostaining with an
anti-smooth muscle cell -actin (Dako, Carpinteria, CA) and anti-vWF
antibody (von Willebrand factor; Dako), respectively. In some
experiments, the cells were cultured for 24 h prior to the
experiment in medium lacking FCS. Vascular EC were cultured in M199
supplemented with 0.1% human serum albumin (Immuno-US-Incorporated,
Rochester, MN), and vascular SMC were cultured in IT
(insulin/transferrin) medium as described (47).
Monocytes were isolated from freshly prepared human peripheral blood
mononuclear cells obtained from leukopacs of healthy donors (kindly
provided by Dr. Steve K. Clinton, Dana Farber Institute, Boston, MA) by
density gradient centrifugation as described previously (30). Monocytes
(1 × 106 cells/ml) harvested by scraping were
resuspended in processing buffer (10 mM HEPES, 1 mM dithiothreitol, 10% glycerol; Sigma) and used directly
for processing assays or stored in aliquots at 20 °C. Purity of
monocytes was
95%, as determined by fluorescence-activated cell
sorter analysis (anti-human CD68 monoclonal antibody, fluorescein isothiocyanate conjugated; Pharmingen, San Diego, CA).
For detection of IL-1 activity, the mouse thymocyte cell line
D10.G4.1 (kindly provided by Dr. Andrew Lichtman, Brigham and Women's
Hospital, Boston, MA) and human dermal fibroblasts were isolated and
cultured as described previously (48).
Human vascular
SMC or EC cultured in 75-cm2 flasks were washed twice and
incubated for 24 h prior to the experiment in the absence of
serum. The medium was replaced by medium lacking methionine and
cysteine but supplemented with rCD40L (49) in the presence of 60 µCi/ml [L-35S]protein labeling mix (NEN
Life Science Products). All experiments were performed in the presence
of the endotoxin inhibitor polymyxin B (1 µg/ml; Sigma). After
24 h at 37 °C, cells were harvested by scraping in
immunoprecipitation buffer (50 mM Tris-HCL, 0.1% SDS,
0.1% sodium deoxycholate, 1% Nonidet P-40, 150 mM NaCl, 5 mM EDTA, 20 µg/ml soybean trypsin inhibitor, 0.1%
mM phenylmethylsulfonyl fluoride, 0.2 units/ml aprotinin,
0.025% sodium azide). Cell-extracts were centrifuged (30 min, 4 °C,
10,000 × g) and supernatants were precleared with
non-immune rabbit-serum (18 h, 4 °C; Vector, Burlingame, CA). After
centrifugation (10 min, 4 °C, 10,000 × g), proteins of the supernatants were immunoprecipitated (2 h, 4 °C) with the IL-1-specific polyclonal rabbit antibody (Upstate Biotechnology Inc.). After adding protein A-agarose (1.5 h, 4 °C; Life
Technologies, Inc.), precipitates were centrifuged (2 min, 4 °C,
300 × g), and the pellet was resuspended in 50 µl of
SDS-PAGE sample buffer (0.2 M Tris (pH 6.8), 5% glycerol,
0.1% SDS, 3%
-mercaptoethanol, 0.1 mg/ml bromphenol blue, final
concentrations). After heating for 10 min at 95 °C, the samples were
separated by SDS-PAGE, transferred on polyvinylidene difluoride
membranes (Millipore, Bedford, MA), and exposed to autoradiography film
(NEN Life Science Products).
Cell extracts, equalized in total protein,
were separated by standard SDS-PAGE under reducing conditions, and
transferred to polyvinylidene difluoride membranes using a semi-dry
blotting apparatus (3.0 mA/cm2, 30 min; Bio-Rad, Hercules,
CA). Blots were blocked (2 h), and dilution of first and second
antibody was made in 5% defatted dry milk, PBS, 0.1% Tween 20. After
1 h of incubation with the respective primary antibody (1:1,000
polyclonal rabbit anti-IL-1 (Upstate Biotechnology Inc.), 1:200
polyclonal goat anti-IL-1
-converting enzyme (M19, Santa Cruz
Biotechnology, Santa Cruz, CA)) blots were washed four times (15 min in
PBS, 0.1% Tween 20) and the secondary peroxidase-conjugated goat
anti-rabbit antibody (1:10,000; Jackson Immunoresearch, West Grove, PA)
was added for another hour. Finally, after 4 times washing (20 min,
PBS, 0.1% Tween 20), immunoblots were developed using the Western blot
chemiluminescence system (NEN Life Science Products) or the chromogenic
system adding diaminobenzidine (50 µg/ml; Sigma) in substrate buffer
(17 mM citric acid, 65 mM
NaH2PO4, 0.1% H2O2,
0.01% (w/v) Thimerosal) to the blots. Independently produced
antibodies directed against IL-1
(rabbit polyclonal anti-IL-1
(Santa Cruz) and the mouse monoclonal antibody Fib3 (50)) as well as
ICE (rabbit polyclonal
-ICEP20 antibody (51)) yielded
similar results in the experiments performed, indicating that the
reagents employed specifically recognize the intended proteins.
Cultured human vascular SMC and EC as well
as monocytes were harvested by scraping in processing buffer (10 mM HEPES, 1 mM dithiothreitol, 10% glycerol;
final concentrations; Sigma). After 3 freeze-thaw cycles, 30 µl of
cell extract (containing equal amounts of total protein) were incubated
for the designated times at 37 °C with 50 ng of recombinant human
IL-1 precursor (pIL-1
; Cistron, Pine Brook, NJ). All assays were
performed in a final volume of 50 µl. The processing was stopped by
adding 10 µl SDS-PAGE (5 ×) sample buffer and heating the samples
(10 min, 95 °C). Finally, the samples were separated by SDS-PAGE and
were analyzed by immunoblotting as described above. Specificity of the
processing was analyzed by pre-incubation (10 min, 37 °C) of cell
extracts with 100 µM ICE-inhibitor (Ac-Tyr-Val-Ala-Asp-H
(aldehyde); Peptide Institute, Osaka, Japan) (8) prior to addition of
the precursor.
Human vascular SMC or EC
were incubated for 24 h with the respective stimuli (None, rCD40L,
or IL-1) in the absence or presence of the anti-CD40L antibody (5 µg/ml). The culture supernatants were added in the absence or
presence of the neutralizing IL-1
antibody (1 µg/ml; Endogen,
Cambridge, MA) to (i) the murine thymocyte cell line D10.G4.1 or (ii)
subconfluent fibroblast cultures (5000 cells/cm2), and the
IL-1 assay was performed as described previously (48, 52). Briefly,
after 72 h of stimulation, cells were pulsed for the final 24 h with tritiated thymidine ([3H]thymidine, 5 µCi//well,
NEN Life Science Products) in 96-well plates and harvested, and
[3H]thymidine incorporation (disintegrations per minute
per culture ± S.D.) was determined. The mean of triplicate
cultures was determined. Alternatively, fibroblasts were fixed with
paraformaldehyde (2%), stained with crystal violet (10% in methanol),
and lysed by incubation with 100 µl of SDS (1%), and finally,
absorbancy was measured at 550 nm.
Human
vascular SMC and EC express CD40 protein and respond to its ligand
CD40L (30, 38-40). Stimulation of vascular cells with rCD40L induced
concentration-dependent de novo synthesis of the
33-kDa IL-1 precursor, as shown for human vascular SMC by metabolic
labeling and immunoprecipitation (Fig. 1). Induction of
the protein required at least 1 µg/ml rCD40L. The precipitated IL-1
protein migrated at approximately 33 kDa as expected for the
precursor form (53-55). Smaller forms of IL-1
, i.e. the
biologically active mature form with a molecular mass of 17 kDa, were
neither detected in cell extracts nor culture supernatants. Similar
results were obtained with human vascular EC (data not shown).
Recombinant Human CD40L Increases Expression of a 20-kDa Immunoreactive ICE Protein
Human vascular SMC and EC produce the
IL-1 precursor (1, 21, 56) but do not release mature forms of
IL-1
upon stimulation with IL-1
, IL-1
, TNF, endotoxin etc.
(51). We therefore further analyzed the effect of rCD40L on the
expression and/or activation of the ICE, the enzyme responsible for
production of biologically active, mature IL-1
(6-10). We first
investigated whether or not stimulation of vascular cells affected the
expression of ICE proteins. In monocytes or monocytic cell lines, this
enzyme exists as a 45-kDa zymogen, an intermediate form of 30 kDa, and
active subunits of 20 (p20) and 10 kDa (p10) (8, 17). An antibody raised against the p20 subunit detects a 45-, 30-, and 20-kDa band in
vascular SMC and EC (51). Neither regulation of these immunoreactive
ICE proteins nor biologically active ICE forms have been previously
found in vascular cells. However, stimulation of vascular cells through
rCD40L increased the expression of the 20-kDa immunoreactive ICE
protein, as illustrated here for SMC (Fig. 2). This
increase did not occur in cells cultured in the presence of serum, an
unphysiologic condition for SMC (47). Thus, the following experiments
were performed using vascular cells cultured in the absence of serum.
Howard et al. (57) showed that activation of ICE requires
co-expression with IL-1
. We therefore further explored the influence
of recombinant human mature IL-1
(rIL-1
) or rIL-1
/rCD40L
co-stimulation on ICE-expression. However, rIL-1
either alone or in
combination with rCD40L did not alter the expression of the 20-kDa
immunoreactive ICE protein (Fig. 2). Appearance of the 20-kDa
immunoreactive ICE protein depended on the rCD40L concentration (Fig.
3A). Furthermore, the increase of the 20-kDa
immunoreactive ICE protein depended on the time of stimulation with
rCD40L, first detected after 2 h (Fig. 3B). The early
detection of the ICE protein could be due to processing rather than
de novo synthesis of the constitutively expressed ICE
precursor. The increasing strength of the 20-kDa band, compared with
the weak zymogen band, may be due to additional de novo
synthesis and subsequent processing of the ICE precursor, which is
highly autocatalytic. In addition to the p20 subunit, active ICE
contains the p10 subunit (12). We did not detect the p10 subunit in
these analyses because the anti-p20 antibody used does not recognize p10.
Induction of IL-1
To analyze whether the increase of the
20-kDa immunoreactive ICE protein in rCD40L-stimulated vascular cells
correlates with the induction of biologically active ICE, we performed
processing assays in which cell extracts were incubated with
recombinant human IL-1 precursor (pIL-1
). We monitored processing
of exogenous pIL-1
into smaller fragments of the cytokine by Western
blot analysis. Extracts of vascular cells stimulated with rCD40L, but not extracts from unstimulated or IL-1
-stimulated cultures,
processed pIL-1
to an IL-1
immunoreactive protein of
approximately 17 kDa (Fig. 4A). The cleavage
product obtained in the processing assay using lysates of
rCD40L-stimulated vascular cells comigrated with the band observed with
monocyte extracts. Monocyte extracts, used here as a positive control,
express biologically active ICE (8), which cleaves the IL-1
precursor into the active mature 17-kDa form. Recombinant human CD40L
coordinately induced the IL-1
processing activity and the increase
of the 20-kDa immunoreactive ICE protein. Detection of IL-1
processing activity required extracts of SMC or EC stimulated with at
least 1 µg/ml rCD40L (Fig. 4B) for 2-6 h (Fig.
4C), a concentration- and time-dependence described above
for the induced increase of the 20-kDa immunoreactive ICE protein. A
selective ICE-inhibitor (8) inhibited the processing of pIL-1
(Fig.
4, B and C). Extracts of SMC or EC cultures
stimulated with rCD40L (5 µg/ml) processed the precursor within 1-2
min (Fig. 5A) and required material derived
from at least 10,000 cells/µl (Fig. 5B). Compared with the
original source of ICE, the monocyte, extracts of human vascular SMC
and EC contained approximately 30-fold less ICE activity. Specific ICE
activity was obtained by using monocyte extracts of 0.3-1.6 mg/ml
total protein, whereas SMC or EC extracts of 10-50 mg/ml total protein
were required. However, SMC comprise the most numerous cell type in
arteries, which normally contains few if any monocytes. Therefore,
activation of ICE by CD40 ligation on human vascular SMC may be crucial
to initiate the inflammatory response in the normal vessel wall. Thus
SMC, rather than monocytes, may be the biologically relevant cell type
for local IL-1
precursor processing at sites of initiation of
vascular inflammatory responses.
Human Vascular SMC and EC Release Biologically Active IL-1
As demonstrated above, rCD40L induced
IL-1 precursor expression in vascular cells and also activated ICE.
However, we did not detect native, mature immunoreactive IL-1
in
extracts or supernatants of SMC and EC cultures stimulated with rCD40L
in radioimmunoprecipitation (Fig. 1) or Western blot experiments (data
not shown). These findings may result from a concentration of mature
IL-1
produced by SMC or EC below the detection limit of these
techniques. Thus, we employed bioassays to detect IL-1
activity in
the supernatants of rCD40L-stimulated vascular cells. Supernatants of
rCD40L-stimulated SMC markedly enhanced the IL-1-dependent proliferation of the mouse thymocyte cell line D10.G4.1 (Fig. 6). The induction of the proliferation was inhibited by
either presence of an
-CD40L antibody during conditioning of media
or the presence of an
-IL-1
antibody during the incubation of
D10.G4.1 cells with the supernatant of rCD40L-stimulated SMC (Fig.
6A) or EC (data not shown), demonstrating the
rCD40L-dependent release of biologically active IL-1
.
The
-CD40L antibody did not affect the proliferation induced by
supernatants of rIL-1
-stimulated vascular cells. Induction of
IL-1
activity in rCD40L-stimulated vascular cells depended on the
concentration and required
1 µg/ml rCD40L (Fig. 6B).
Thus, the concentration dependence of the induction of biological
IL-1
activity in the supernatant correlated with the concentration
required for both, the induction of ICE activity and the expression of
the IL-1
precursor. Both of the conventional IL-1 bioassays employed
(proliferation of D10.G4.1 cells or human dermal fibroblasts) yielded
similar results (data not shown).
The presented data indicate that CD40/CD40L interaction regulates
IL-1 activity in the vessel wall by both induction of IL-1
precursor expression and activation of the IL-1
-converting enzyme. This pathway may have particular relevance for atherogenesis, as cells
in atheroma express CD40, its ligand CD40L, as well as IL-1
, a
cytokine implicated in regulation of many aspects of vascular
pathology. As other stimuli tested previously (e.g. IL-1
, IL-1
, TNF
, IL-8, and endotoxin) do not cause release of active IL-1
from vascular wall cells, CD40 ligation may prove important in
initiating and sustaining inflammatory and host defense processes involving blood vessels. Moreover, inhibition of CD40L/CD40 signaling may represent a novel therapeutic target in arterial inflammation and
atherosclerotic diseases.
We thank Maria Muszynski, Curran Murphy, and Elissa Simon-Morrissey (Brigham & Women's Hospital) for skillful assistance.