| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Laboratory of Immunology, INSERM U519, IFRMP23, University of Rouen, France;
* Brain Inflammation and Immunity Group, Department of Medical Biochemistry and Immunology, Cardiff, UK; and
Research Institute of Highly Pure Biopreparations, Saint Petersburg, Russia
1Correspondence: Brain Inflammation and Immunity Group, Department of Medical Biochemistry and Immunology, Tenovus Building, Cardiff, CF144XN, UK. E-mail: gasque{at}cardiff.ac.uk; Tiphaine_Monsinj{at}hotmail.com
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: C3aR • C5aR • MAP kinases
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. One biological consequence of this activation is the release of potent inflammatory molecules, C3a and C5a anaphylatoxins. C anaphylatoxins share several biological properties, including mast cell degranulation, smooth muscle contraction, vasodilation, and recruitment of immune cells to the site of inflammation (for review, see ref 2
). C3a and C5a act through specific receptors, named respectively C3aR and C5aR, that belong to the seven membrane-spanning receptor superfamily and are functionally coupled to a pertussis toxin (PTX) -sensitive G-protein (3
, 4)
. Expression of these receptors, initially thought to be restricted to monocytes/macrophages and polymorphonuclear cells, appears to be extended to several tissues. Indeed, C5aR expression has been demonstrated on nonmyeloid cells of the liver, lung, and brain (5
6
7
8
9
10)
, as well as on endothelial (8
, 11
12
13
14
15
16
17
18)
and epithelial cells (19
20
21
22
23
24)
. C5a has a broad spectrum of activity and can promote superoxide radical production from eosinophils, hydrolytic enzyme release from neutrophils, expression of several cytokines and chemokines by myeloid and nonmyeloid cells, phagocyte activation, and induction of P-selectin expression by human umbilical vein endothelial cells (HUVEC) (for a review, see ref 25
).
The cloning of C3aR cDNA is recent, and relatively little is known about this receptor (26
27
28
29
30
31
32)
. RT-PCR analyses indicated that the C3aR transcript was expressed in several lymphoid and endocrine tissues. Northern blot analyses revealed a 2.2 kb C3aR mRNA transcript particularly abundant in lung and heart; surprisingly, however, and in contrast to C5aR, C3aR expression was largely unaffected by lipopolysaccharide (LPS) challenge (31)
. Binding and functional studies have indicated the presence of C3aRs on human eosinophils, basophils, guinea pig platelets, mast cells, and nerve cells. C3a is known to promote strong chemotaxis (for eosinophils, and human mast cells), cytokine and chemokine production, leukotriene production by basophils, and platelet aggregation (23
, 29
, 33
34
35
36
37
38
39
40
41)
. Although a functional C3aR has recently been characterized on HUVEC, little is known about the role of C3a or even C5a on endothelial cells (18)
.
Activation of C in the vicinity of endothelium is thought to contribute to the inflammatory reaction, leukocyte infiltration, and overall immune response (42)
. Endothelial C fixation at sites of inflammation could generate an endothelium-restricted adhesion signal (C opsonins) whereas C anaphylatoxins could promote the production of cytokines and chemokines to drive leukocyte infiltration after injury. We have already shown that endothelial cells can synthesize in vivo and in vitro all components of the C cascade and nearly all C regulators (43
44
45)
. We reported that IL-1ß and glucocorticoids were able to induce the production of C3a anaphylatoxin in the vicinity of the endothelial cells possibly to drive a C-dependent inflammatory response (45)
. We further investigated the cellular and molecular routes by which activation of C on endothelial cells might influence the expression of inflammatory genes and have made use of the well-validated in vitro HUVEC model.
In this study, we observed that HUVEC expressed constitutively both anaphylatoxin receptors and that stimulation by C3a or C5a induced an increase in IL-1ß, IL-8, and RANTES (regulated upon activation on normal T expressed and secreted) mRNA levels. C5a alone down-regulated IL-6 mRNA levels. These effects were blocked by pretreatment with neutralizing anti-receptor antibodies and PTX, which together strongly suggest the involvement of Gi protein-coupled signaling pathways. After C3a and C5a treatments, IL-8 but not IL-6 protein levels were up-regulated in supernatants of HUVEC as early as 3 h poststimulation (probably from storages). The prolonged C3a or C5a stimulation induced a second long lasting wave of IL-8 production. To investigate and better understand the mechanism of action of C anaphylatoxins, we studied the mitogen-activated protein (MAP) kinase transduction signals activated by the action of C3a and C5a on HUVEC.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
or obtained from Clonetics, San Diego, Ca, USA (CC-2519) and cultured in the recommended EGM-2-MV medium (BioWhittaker-Cambrex, Wokingham, UK). The human umbilical cords were treated with 0.1% (w/v) collagenase A (Boehringer Mannheim, Meylan, France) for 10 min at 37°C. Dissociated cells were grown at 37°C in 
-minimum Eagle medium containing 20% heat-inactivated fetal calf serum (FCS; Bioproducts, Gagny, France), 20 µg·mL-1 endothelial cell growth supplement (Sigma, Saint Quentin-Fallavier, France), 2 mM L-glutamine, 90 µg·mL-1 heparin, 2.5 µg·mL-1 fungizone, 50 µg·mL-1 penicillin, and 50 µg·mL-1 streptomycin. Cell culture reagents were from BioWhittaker (Verviers, Belgium) unless otherwise stated. Double immunofluorescence staining protocols (5)
using mouse anti-CD31 antibody (clone MEM 05, generous gift from Dr. V. Horejsi, Praha, Czech Republic), rabbit anti-factor VIII-related protein (von Willebrand factor, Neomarkers, Stratech, Luton, UK), or mouse anti-thrombomodulin (anti-CD141 antibody, Serotec, Kidlington, UK) were performed routinely to identify the endothelial cell population. Cells were grown to near confluence in 25 or 75 cm2 flasks (Costar, ATGC, Noisy le Grand, France) coated with 1% gelatin (Sigma) at 37°C in a 5% CO2 atmosphere. Subculturing of the cells was performed by treating the cell monolayer with trypsin-versene and cells were used for experimental work in the second through fourth passage. In some experiments HUVEC cells were stimulated with recombinant human cytokines (200 IU/mL of IL-1ß, 200 IU/mL IFN-
, or 1000 IU/mL of TNF-) for 24 h. K562 and C3aR-transfected K562 cells were cultured as described (23)
and used to test the specificity of the different anti-C3aR antibodies. THP1 monocyte cell line (C5aRdim) and phorbol myristate acetate (PMA) -stimulated THP1 (C5aRbright) (5)
were used to test the different anti-C5aR antibodies (see Table 1
).
|
Immunostaining, FACS analysis, and monoclonal anti-anaphylatoxin receptors
Immunohistochemistry and FACS analyses were performed essentially as described (47)
. The rhodamine- or FITC-conjugated donkey antibodies against rabbit or mouse immunoglobulins, respectively, were from Jackson (Stratech, Luton, UK). The mouse anti-C3aR antibody (clone BIIG1, produced in house) recognizes the human C3aR second extracellular loop (23)
. Two anti-C5aR antibodies (clones S5/1 and W17/1) were from Serotec (Kidlington, UK).
Chemicals, cytokines, and antibodies
PTX and PMA were purchased from Sigma. Recombinant human IFN-, IL-1ß, tumor necrosis factor (TNF-) were gifts from Hoffmann-La Roche (Nutley, NJ, USA). Human C3a was generated by activation of C using zymosan A and purified as already described (39)
. The purity of C3a after gel filtration was assessed by SDS-PAGE, mass spectrophotometry (UWCM Proteomics Facility, Cardiff) and was confirmed to be free of LPS by limulus assay. We determined that C3a preparations were free of C5a contamination using a C5a radioimmunoassay (39)
. C5a multiple array peptide was synthesized by solid-phase synthesis (40)
.
Production of polyclonal Abs against C5aR and C3aR
Production of a specific anti-human C5aR antibody (L67): cDNA encoding the amino-terminal part of the C5aR (29 aa) was cloned in a GST fusion protein system (PGEX 2T) (Amersham Pharmacia-Biotech, Uppsala, Sweden). The amino-terminal part of C5aR cDNA was produced by RT-PCR from PMA-differentiated U937 using two specific oligonucleotides. The sequence TCA was introduced in the downstream primer to create a stop codon. Bam HI and Eco RI sites were introduced in both primers to direct the cloning of C5aR cDNA into the PGEX 2T expression vector. After transfection in Escherichia coli, expression of the fusion protein was induced by isopropyl-ß-D galactoside 0.1 M for 3 h. Cells were pelleted, lysed in 2 M urea, and sonicated. The expressed fusion protein was purified on a glutathione-Sepharose column 4B (GST-C5aRL fusion protein) (Amersham Pharmacia-Biotech) using the batch method as described by the manufacturer and analyzed by Western blot (data not shown). The size of the fusion protein was 34 kDa. Finally, the fusion protein was submitted to high purification by HPLC gel filtration on a Bio-Sil-SEC250 column (300 mmx7.8 mm) (Bio-Rad, Hercules, CA, USA) before being used for rabbit immunization. For the immunization, the recombinant fusion protein (500 µg) was emulsified in complete Freund’s adjuvant and injected subcutaneously at multiple sites along the spinal column of New Zealand rabbits (Charles Rivers, St. Aubin les Elbeuf, France). After 1 month, rabbits were immunized again with the recombinant protein emulsified in incomplete Freund’s adjuvant (IFA). They received two more injections of IFA at 1 wk intervals and finally were bled 1 wk later. IgG from immunized rabbits were purified by perfusion chromatography on a Poros Protein A column using Vision workstation (PerSeptive Biosystems, Framingham, MA, USA).
Two affinity-purified anti-human C3aR antibodies were used. First, the rabbit anti-C3aR peptide (PSGFPIEDHETSPLDNSDAFLSTHLKLFPS) corresponding to amino acid residues 270–300 of the human C3aR loop (named BIIG) was produced as described (47)
. This antibody is against a small domain of the C3aR and does not have blocking activities. In contrast, the rabbit (L755) immunized with the recombinant GST-C3aR loop (full-length second extracellular loop of human C3aR fused to GST and produced as described above) displayed strong blocking activities (this study).
Detection of phosphorylated MAP kinases by Western blot and ELISA
Activation of MAP kinases (p44/p42) was assessed by the phosphoPlus p44/p42 MAP kinase (Thr 202/Tyr 204) antibody kit (New England BioLabs, Beverly, MA, USA) using Western blot and ELISA protocols. For the Western blot, cells were cultured in 6-well plates in serum-free medium ultratome (BioWhittaker). After different periods of stimulation with peptides C3a or C5a, cells were lysed. Lysates were sonicated for 10 s and heated at 95°C. Proteins were separated by electrophoresis on 16% polyacrylamide gels and blotted onto nitrocellulose membranes (Millipore, Bedford, MA, USA). Membranes were saturated with 0.1% Tween-20, 5% w/v skimmed dry milk in TRIS buffered saline (TBS) for 30 min at room temperature. Two primary antibodies were used: rabbit anti-p44/p42 MAPKs Ab or rabbit anti-phosphorylated p44/p42 MAPKs Ab. Incubation was performed overnight at 4°C in the saturating solution. After three washes in 0.1% Tween-20 in TBS, membranes were saturated as described before. Incubation with goat anti-rabbit horseradish peroxidase-conjugated Ab was performed at room temperature for 2 h. After three washes in 0.1% Tween-20 in TBS and a final wash in distilled water, membranes were developed by electrochemiluminescence and exposed on autoradiographic films. For the ELISA, HUVEC were first cultured in 96-well plates (104 cells/well in 100 µL) in EGM-2-MV Clonetics medium for 24 h, then serum starved for 4–5 h before stimulation with anaphylatoxins diluted in serum-free medium Optimem (Invitrogen, Abingdon, UK). In some experiments cells were pretreated with PTX (2 µg/mL for 4 h) or anti-anaphylatoxin receptor antibodies (5 µg/mL). The 96-well plates were washed with serum-free medium and cells were fixed using 0.25% paraformaldehyde, followed by several washings using 0.2M glycine in PBS before permeabilization with cold 90% methanol. This technique was adapted from the flow cytometry protocol provided by the manufacturer for intracellular staining of MAP kinases. Levels of intracellular MAP kinases were assessed by ELISA using rabbit anti-p44/p42 MAP kinases or rabbit anti-phosphorylated p44/p42 MAP kinases (Thr202/Tyr204 at 1/200) (1/200), followed by the peroxidase-conjugated donkey anti-rabbit antibody (Jackson, Stratech). The OD after OPD development was measured at 490nm (see below).
IL-8 and IL-6 ELISAs
For IL-8, 96-well plates (Costar, Corning, NY, USA) were incubated with monoclonal anti-human IL-8 Ab (T9 g10, 1 µg·mL-1) in 0.05 M borate buffer, pH 8.0, and subsequently blocked with PBS containing 1% BSA. After washing, appropriate dilutions of samples were added, incubated for 1 h at room temperature, washed, and incubated with a monoclonal biotinylated anti-human IL-8 (clone 8 g12 diluted 1:2000) for 1 h at room temperature. After washing, streptavidin–peroxidase (Sigma, dilution: 1:3000) was added. The substrate for the peroxidase was o-phenylenediamine (DAKO, Carpinteria, CA, USA) and the cytokine concentration was calculated relative to IL-8 standards. The assay’s threshold was 20 pg·mL–1. The human IL-6 ELISA kit was used according to the manufacturer’s instructions (Bender Medsystems, Caltag Medsystems, Towcester, UK).
RNA preparation
Total RNA was isolated from cultured cells using isothiocyanate guanidium and ultracentrifugation on a cesium chloride cushion, followed by phenol/chloroform extraction as described previously (8)
. The quality of RNA was determined by electrophoresis on a 1% agarose gel and concentration was determined by absorbance at 260 nm. Fifty µg of the total RNA was treated for 20 min at 37°C with 90 u of RQ-1 RNase-free DNase (Promega, Charbonnières, France) in 100 µL of buffer (40 mM Tris-HCL pH 8.0, 10 mM NaCl, 6 mM MgCl2, and 10 mM CaCl2) and 200 u of RNAsin RNase inhibitor (Promega) to remove all traces of contaminating genomic DNA.
Reverse transcription PCR
For further details on RT-PCR procedure, see ref 5
. The sequences and positions of the primers used in this study are given below.
IL-8 sense: TCT TGG CAG CCT TCC TGA TT (100);
IL-8 antisense: AAC TTC TCC ACA ACC CTC TG (343);
glyceraldehyde 3-phosphate dehydrogenase (GAPDH) sense: TGC CAT CAA CGA CCC CTT CA (153);
GAPDH antisense: TGA CCT TGC CCA CAG CCT TG (702);
IL-6 sense: GAT GGA TGC TTC CAA TCT GGA T (906);
IL-6 antisense: AGT TCT CCA TAG AGA ACA ACA TA (1349);
RANTES sense: CCT CCG ACA GCC TCT CCA CA (1);
RANTES antisense: GTG TAA GTT CAG GTT CAA GGA (332);
IL-1ß sense: CAT ATG AGC TGA AAG CTC TCC A (502);
IL-1ß antisense: GAG GTG CTG ATG TAC CAG TT (808);
C5aR sense: CGGGATCCATGAACTCCTTCAATTATACC (probe);
C5aR antisense: CGGAATTCTCAAGTTTTATCCACAGGGGTGT (probe).
For semiquantitative RT-PCR, PCRs were performed with 25 pmol of each primer, 1.25 pmol of each GAPDH primer and 1 µCi [33P]dATP (Redivue, Amersham, les Ulis, France). Primers were used with a GAPDH: cytokine molar ratio of 1:20. The number of cycles was chosen so as not to reach the plateau for GAPDH (25 cycles). PCR products were loaded onto a 6% acrylamide gel and separated by electrophoresis migration. Gels were dried and exposed onto Biomax Films (Kodak film, Sigma, St. Louis, MO, USA). Autoradiograms were analyzed by densitometry using Lecphor imaging system software (Biocom, Les Ulis, France). The relative amount of each cytokine mRNA was estimated by dividing the peak densitometry area of the cytokine amplicon by that of the GAPDH amplicon.
The level of cytokine mRNA expression from unstimulated cells was set to 1 arbitrary unit and values for the stimulation time were calculated accordingly.
cDNA probe production
cDNA probes for human IL-8, IL-6, IL-1ß, and RANTES were cloned from either U937 RNA or HUVEC RNA after RT-PCR using human cDNA-specific primers (see above).
The RT-PCR product was purified by a Gel extraction Kit (Promega) after the manufacturerÆs instructions, subcloned in the pGEM-T vector system (Promega) and sequenced. The probe was 100% homologous with the reported cDNA sequences (GenBank) reported in EMBL Data Library.
Northern blot
Total RNA (20 µg) was separated by electrophoresis onto 0.8% agarose/0.2 M formaldehyde gels at 50 V overnight. RNA was capillary transferred onto Membrane Hybond N-+ (Amersham, Les Ulis, France) in 50 mM NaOH for 3 h. RNA marker (Promega) was used as an RNA ladder. The membrane was washed twice in SSPE 2x and allowed to dry at room temperature. Hybridization with the different probes was performed as described previously (40)
. Labeling of the different probes was performed by a random prime labeling system using the rediprime II kit (Amersham Pharmacia-Biotech, Buckinghamshire, UK). Membranes were exposed onto Kodak Films (Sigma) for 20 days at -80°C.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 47)
and a similar approach was used to ascertain the expression of both anaphylatoxin receptors on endothelial cells. HUVEC were strongly stained using several antibodies against human C3aR and C5aR (Table 1
, Fig. 1
B). PMA-treated THP1 monocyte cell line had a mean fluorescence at least double that of HUVEC for C3aR, whereas C5aR expression on endothelial cells was four- to fivefold lower compared with PMA-THP1 cells. However, we found that HUVEC contained an important intracellular pool of both receptors. Stimulation of HUVEC with IL-1ß or IFN- significantly increased the expression of C3aR but not C5aR at the cell membrane (Fig. 1B
and data not shown). The optimal stimulus turned out to be IL-1ß, then IFN-. We found that recombinant TNF- failed to control C3aR and C5aR expression by HUVEC after 24 h, although ICAM-1 expression was strongly up-regulated (data not shown).
|
Indirect immunofluorescence confirmed that C3aR and C5aR were expressed by the majority of endothelial cells (Fig. 1A
). For instance, HUVEC stained for specific endothelial cell markers (e.g., CD31, CD141) displayed a patchy C3aR on the membrane (Fig. 1Ab
). Again, although most of the staining was associated with the cell membrane, C3aR and C5aR were also detected within cytoplasmic stores. No staining was detected in the nucleus. Control antibodies (mouse anti-rat CD59 and rabbit anti-CHO cells) failed to stain HUVEC cells (Fig. 1Ad
).
Differential expression of cytokine genes in HUVEC stimulated with C3a or C5a
Production of IL-8, Il-1ß, IL-6, and RANTES was studied in endothelial cells stimulated with a range of concentrations of anaphylatoxins and for different periods. Endothelial cells at 
80% confluence were incubated with various concentrations of anaphylatoxins C3a or C5a and expression of IL-8, IL-6, IL-1ß, and RANTES was quantified by semiquantitative RT-PCR. We first checked the expression of these cytokines by unstimulated HUVEC. Data presented in Fig. 2
A show that HUVEC expressed basal levels of IL-8, IL-6, IL-1ß, and RANTES mRNA, which had the correct predicted size according to the primers used.
|
Stimulation of endothelial cells with 10–8 M C3a induced a significant increase in IL-8, IL-1ß, and RANTES mRNA levels but exerted no effect on IL-6 expression. IL-1ß and IL-8 mRNA levels were increased as early as 3 h and returned to nearly basal levels at 9 h, whereas RANTES mRNA levels were not increased until 6 h had elapsed and remained high at 9 h (Fig. 2B
). In contrast, C3a exerted no significant effect on IL-6 expression by HUVEC. These results were confirmed by Northern blots using specific probes for the four cytokine/chemokines (Fig. 3
A, B).
|
C3a is known to bind to the high-affinity C3aR that is functionally coupled to a G-protein. Therefore, we wondered whether the observed effects of C3a on cytokine mRNA levels could be altered by PTX, an inhibitor of certain G-proteins. HUVEC were preincubated with PTX for 4 h, then stimulated for 3 h with C3a (for IL-1ß, IL-8, and IL-6) and for 6 h with C3a (for RANTES) (Fig. 4
). To confirm that PTX had no direct effect on the cells, HUVEC were incubated with PTX alone. The level of IL-8 mRNA after incubation with PTX alone was comparable to that of nonstimulated cells. As a control for the blocking effect of PTX, cells were stimulated with peptides alone (3 h). In C3a-activated HUVEC, pretreatment with PTX reduced IL-1ß, IL-8, and RANTES mRNA levels to levels comparable to those observed in nonstimulated cells. The level of IL-6 mRNA remained unchanged as it was not affected by anaphylatoxin C3a.
|
When HUVEC were activated by 10–8 M C5a, a significant increase in IL-8, IL-1, and RANTES mRNA levels was recorded and a slight yet significant decrease in IL-6 mRNA levels was observed (Fig. 5
A). IL-8 and IL-1ß mRNA levels were increased as early as 3 h and returned to basal levels in up to 6 h whereas RANTES mRNA levels were not significantly increased until 6 h had elapsed and returned to basal levels from 9 h (Fig. 5A
). Conversely, IL-6 mRNA levels were significantly decreased from 1 h until 9 h when HUVEC were stimulated with C5a. We found that pretreatment of HUVEC with PTX before stimulation with C5a blocked the increase in IL-8, IL-1ß, and RANTES expression as well as the decrease in IL-6 expression induced by C5a (Fig. 5B
).
|
As demonstrated by these key experiments, C3a and C5a act in the same way on IL-1, IL-8, and RANTES mRNA levels by increasing expression of these cytokines, but only C5a decreases IL-6 mRNA levels.
C3a and C5a modulate cytokine production after binding to their specific high-affinity receptors on HUVEC
To ascertain that cytokine mRNA modulation was specific to anaphylatoxin action, HUVEC were preincubated for 1 h with an anti-human C3aR antibody (BIIG; 1:500) and the anti-human C5aR antibody (L-67; 1:500), then stimulated for 3 h with C3a or C5a (Fig. 6
A). Similar experiments were performed with the three other cytokines (data not shown). The anti-human C3aR against a small domain of the second extracellular loop (see Materials and Methods) failed to neutralize the IL-8 up-regulation induced by C3a although a slight inhibition was reproducibly observed (Fig. 6A
, lane 2). However, we found that the rabbit polyclonal generated against the full-length human C3aR second extracellular loop (L755), but not the mouse anti-C3aR antibody (clone BIIG1), strongly inhibited C3a activity on endothelial cells (Fig. 6B
). The anti-human C5aR (L-67) or the mouse anti-human C5aR directed against the Nt domain (clones S5/1 and W17/1) blocked the C5a-induced up-regulation of IL-8, with levels comparable to those observed in nonstimulated cells but did not block the stimulatory effect induced by C3a (Fig. 6A, B
).
|
C3a and C5a are endothelial cell secretagogues to promote the release of chemokines from storage organelles
It has been reported that IL-8 is stored in Weibel-Palade bodies in HUVEC and rapidly released within minutes after stimulation with endothelial secretagogues such as PMA or histamine (48)
. Given the C3a- and C5a-mediated strong up-regulation of IL-8 mRNA, we wondered whether they induced rapid secretion of the IL-8 protein by endothelial cells. Layers of endothelial cells were incubated at 37°C for different periods ranging from 1–3 h with 10–8 M C3a or 10–8 M C5a. The supernatants of the anaphylatoxin-stimulated endothelial cells were harvested and the concentration of secreted IL-8 and IL-6 (not stored in WB bodies) were assessed by specific and sensitive ELISAs. C3a and C5a induced a rapid and strong release of IL-8 (but not IL-6) from HUVEC and at 3 h poststimulation the level of IL-8 was twofold higher compared with control cells. A synergistic effect was observed when C3a and C5a were used simultaneously and this effect was significantly abolished by pretreating HUVEC with PTX (Fig. 7
A).
|
C3a and C5a control the constitutive secretion of IL-8 but not IL-6 by HUVEC
HUVEC can be stimulated by diverse stimuli to promote de novo synthesis of cytokines and chemokines, which, after localization in the Golgi, are rapidly and constitutively secreted without being stored (49)
. We here addressed the capacity of C3a and C5a to control this constitutive secretion pathway by stimulating HUVEC over 24–72 h. After 24 h, a basal level of IL-8 produced by endothelial cells was detected (142.3±17.9 pg·mL-1). A significant increase in IL-8 production was recorded as early as 32 h (187±16.01 pg·mL-1; P<0.05) after the addition of C3a and cytokine levels increased continuously through 72 h where they reached a maximum (505±32.4 pg·mL-1; P<0.001) (Fig. 8
). HUVEC were also stimulated with C5a 10–8 M and a significant increase in IL-8 production was detected as late as 72 h (485±65.2 pg·mL-1; P<0.05). These effects were time and dose-dependent (Fig. 8
and Fig. 9
). C3a did not influence IL-6 secretion, although we found that C5a, in contrast, inhibited IL-6 secretion levels. We used cells treated with IL-1ß and TNF- to show that HUVEC were capable of producing very high levels of IL-6 (Fig. 9)
.
|
|
C3aR and C5aR are coupled to Erk MAP kinase signaling
Previous studies have implicated the MAP kinase pathway as one of the transduction mechanisms activated by Gi protein coupled to anaphylatoxin receptors (3
, 4)
. To test the functional activity of anaphylatoxin receptors, we assessed one possible transduction pathway of C3aR and C5aR on HUVEC, the MAP kinase pathway.
We showed here that the stimulation of HUVEC with C3a or C5a resulted in activation of the mitogen-activated forms of the p44/p42 kinases (Fig. 10
). Two different Abs were used here: the anti-MAP kinase Ab that recognizes total MAP kinases and served as an internal control antibody for our analysis and the anti-phosphorylated MAP kinase antibody that specifically detects the activated form of p44/p42 MAP kinases. The maximum activated state of MAP kinases in C5a- or C3a-stimulated HUVEC was reached 30 min poststimulation (Fig. 10A
), but activation was already detectable at 5 min poststimulation by C anaphylatoxins (data not shown). An ELISA technique was performed on fixed/permeabilized HUVEC to assess accurately the activation of MAP kinases after stimulation with different doses of C3a or C5a in the presence or absence of PTX. The results are presented in Fig. 10B
and further confirmed that C3a and C5a binding to their respective receptors engaged a strong activation of Erk kinases. Pretreatment of HUVEC with PTX abolished the C3a- or C5a-induced activation of p44/p42 kinases, which remained at levels comparable to those observed in unstimulated control cells (Fig. 10B
).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Optimal leukocyte emigration requires the production of leukocyte chemoattractants and the up-regulation of different classes of adhesion molecules on the surface of endothelial cells. Specific activation of C in the vicinity of altered endothelium is thought to contribute to the inflammatory reaction by promoting leukocyte infiltration and the overall immune response (1)
. A pioneering study has highlighted that endothelial C fixation at sites of inflammation (e.g., sepsis) could indeed generate endothelium-restricted adhesion signals (C3b, iC3b opsonins) whereas C anaphylatoxins could promote the expression of adhesion molecules and the production of cytokines and chemokines by endothelial cells to drive leukocyte infiltration after injury (42)
.
On the one hand, C anaphylatoxins (e.g., C5a) have been shown to stimulate the affinity maturation of leukocyte integrins (50)
, and Foreman et al. demonstrated that C5a can induce the expression of P-selectin and the secretion of von Willebrand factor on endothelial cells (11)
. On the other hand, C anaphylatoxins produced in an inflammatory focus deep in the parenchyma are postulated to deliver directional cues during the trans-endothelial migration of leukocytes. Hence, and in analogy to chemokines, a prerequisite for the in vivo activity of C chemottractants is that after production in the tissues, they somehow diffuse to the vessel wall and are transported by the endothelial cells to the luminal surfaces to attract leukocytes bearing anaphylatoxin receptors. Much of this paradigm remains largely speculative and lacks convincing evidence.
We have previously shown that endothelial cells can synthesize in vivo and in vitro all the components of the C cascade and nearly all C regulators (44
, 51
, 52)
. We reported that IL-1 and glucocorticoids induce the production of C3a anaphylatoxin by HUVEC cells and postulated that this autocrine anaphylatoxin-dependent pathway could be involved in driving subsequently the release of chemokines and cytokines by endothelial cells to control the inflammatory response (45)
. We further investigated the cellular and molecular routes by which activation of C on endothelial cells might influence the expression of inflammatory genes (IL-8, RANTES, IL-1ß, and IL-6) by deciphering the specific roles of C3a and C5a on HUVEC used here as a model endothelial cell line. This investigation was of particular importance in that several groups identified strong expression of C5aR and C3aR on activated endothelia in acute and chronic inflammatory diseases such as stroke, meningitis, and multiple sclerosis (5
, 8
, 15
, 21
, 47)
. We here show for the first time that both anaphylatoxins are potent and rapid stimuli to promote the expression of chemokines and cytokines by endothelial cells.
We found that HUVEC constitutively expressed both receptors, which, after interaction with their respective ligands, engaged G-protein signaling pathways and led to activation of the Erk MAP kinases. We found that signaling through both receptors was PTX sensitive, implying coupling to Gi to promote cytokine and chemokine release. Our understanding of the cellular signaling pathways engaged by C anaphylatoxins is still in its infancy. Using two cellular models (HUVEC and HMEC). Schraufstatter and colleagues recently confirmed that endothelial cells expressed constitutively C5aR and C3aR albeit at low levels compared with leukocytes or monocytes cell lines (18)
. In contrast to our present findings, they reported that actin reorganization induced by C3a on endothelial cells was not PT sensitive but depended on rho activation possibly after signaling through G12 and/or G13. Although G-protein usage by C3aR and C5aR is not absolute within a given cell type or between different cell types, it remains to be determined whether C3a or C5a binding to the high-affinity receptor engages not only single but several G-protein signaling pathways. A novel anaphylatoxin receptor (named C5L2) has recently been identified; in contrast to C5aR or C3aR, it couples only weakly to Gi-like proteins (53
, 54)
. This receptor has so far been found on granulocytes, fibroblasts, and adipocytes and serves as a promiscuous C fragment binding protein for C3a, C5a, C4a, and their desArg forms. We do not at this time know whether endothelial cells express C5L2 or whether it is a functional receptor to promote chemokine production, and experiments along these lines are now highly warranted.
Remarkably, C3a and C5a increased synergistically the secretion (probably from a preformed pool contained in storage depots) as well as biosynthesis of IL-8, the prototype member of the C-X-C subfamily of chemokines known to strongly activate integrin-mediated adhesion of neutrophils (55)
. IL-8 as well as RANTES (also up-regulated by C3a and C5a) belongs to the chemokine family that predominantly mediates the primary attraction of inflammatory cells. In contrast, it is known that increased levels of IL-8 may be necessary for restoration of the integrity of the endothelium by promoting the formation of a fibrin clot and inducing thrombogenesis as well as proliferation and structural reorganization of endothelial cells in the process of angiogenesis (55)
. Overall, our results agree with previous reports demonstrating C anaphylatoxin-dependent immunoregulatory activities through, for example, the induction (TNF-, IL-1, IL-6, and IL-8) and suppression (IL-12) of cytokines by leukocytes and epithelial cells (23
, 29
, 33
34
35
36
37
38
39
40
41)
.
Although C3a failed to control IL-6 biosynthesis, conversely, C5a was shown to have a slight yet significant inhibitory effect on IL-6 secretion. Thus, the effect of C5a on IL-6 in endothelial cells differs from that observed in mononuclear cells. Moreover, it has been demonstrated that C3a exhibited anti-inflammatory properties by suppressing LPS-induced TNF-, IL-ß, and IL-6 secretion from isolated peripheral blood mononuclear cells (29
, 33
34
35
36
37
38)
. We found that C3a was a robust stimulus to induce IL-1ß.
These different findings highlight the extraordinary multipotent activities of C anaphylatoxins that have long been ignored or neglected. Comprehensive studies are urgently needed to decipher the complex yet remarkable pleiotropic activities of these innate cytokine-like molecules. We have learned that C anaphylatoxins probably preceded the emergence of cytokines and chemokines and could overall represent the ancestral molecular signatures of immunoregulatory proteins.
Down-regulation of IL-6 expression by C5a is interesting given the recent finding that IL-6 and soluble IL-6R (sIL-6R) are critical regulators of the transition from neutrophil to monocyte recruitment during inflammation (56)
. We do not now know whether C3a or C5a controls the release of sIL-6R by leukocytes, but it is tantalizing to hypothesize that C anaphylatoxins may be involved in upstream control of the transition from neutrophil to mononuclear cell infiltration, a hallmark of acute inflammation. The proposed paradigm is supported by the finding that C anaphylatoxins are critically involved in several acute inflammatory diseases (57
, 58)
. We propose that during acute inflammation, sustained production of C3a and C5a in the deeply inflamed tissue could promote neutrophil infiltration given the increased expression of IL-8 by endothelial cells, which themselves become coated with C opsonins. However, the subsequent down-regulation of IL-6 production by endothelial cells under the control of C5a could prevent the accumulation of monocytes involved in the clearance of cell debris, including the apoptotic neutrophils. It will be interesting to ascertain whether the proposed paradigm can be emulated in vivo by injection of C anaphylatoxins together with neutralizing antibodies. Better knowledge of the intricate cellular and molecular cross-talk that involves C anaphylatoxins, chemokines, and cytokines could help our understanding of how and why mediators of acute inflammation (aiming to promote healing) lead, in some circumstances, to chronic inflammation.
In summary, we have provided evidence that anaphylatoxin receptors are expressed by endothelial cells, confirming previous findings that these receptors are also expressed in nonmyeloid cells. Our results suggest that binding of C3a and C5a to endothelial cells may induce via cytokine release several biological actions that could be of pathophysiological importance. Indeed, we demonstrated that anaphylatoxin stimulation increased IL-8, IL-1ß, and RANTES mRNA levels in HUVEC in a time- and dose-dependent manner. This work strengthens the concept that anaphylatoxins are of prime importance in inflammation.
|
ACKNOWLEDGMENTS |
|---|
Received for publication July 31, 2002. Accepted for publication February 14, 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
 |
