Objective: Monocyte adhesion to endothelial cells and subsequent secretion of matrix metalloproteinases (MMPs) by activated macrophages are key events in arteriosclerosis and restenosis. We tested the hypothesis that interleukin-10 (IL-10), a potent anti-inflammatory cytokine, inhibits monocyte-endothelial cell interactions. Methods: The effect of IL-10 on monocyte/endothelial cell adhesion, as well as on the expression of MMP-9 and the tissue inhibitor of MMP-9, TIMP-1, were first tested in vitro in coculture systems. In addition, we used an ex vivo binding assay to study the inhibitory effect of IL-10 on monocyte adhesion to carotid arteries obtained from either normal, or l-nitro arginine-methyl ester (l-NAME)-treated rats. The effect of IL-10 on the expression of monocyte adhesion molecules (CD18 and CD62-L) was studied by flow cytometry. Results: IL-10 (150 ng/ml) inhibits monocyte adhesion to endothelial cells (by 35%) and to carotid arteries (by 40 and 50%, in normal and l-NAME-treated rats, respectively), via direct modulation of the expression of CD18 and CD62-L. Moreover, IL-10 dose-dependently decreases MMP-9 activity and increases TIMP-1 levels in coculture systems, both at the transcriptional level. Conclusions: Our results suggest that IL-10 is an important modulator of monocyte–endothelial cell interactions.
Time for primary review 23 days.
Monocyte adhesion to endothelial cells, plays a crucial role in the pathogenesis of arteriosclerosis and restenosis . Recruited monocytes transform into activated macrophages, which synthesize various pro-inflammatory cytokines and modulate extracellular matrix (ECM) turnover . For example, macrophage infiltration of the atherosclerotic plaque is associated with an increased risk of plaque rupture , the principal cause of acute coronary syndromes . Conversely, suppression of monocyte adhesion to endothelial cells may be protective .
Migration of monocytes in the subendothelial space requires protease expression and activity. MMPs are a group of zinc-dependent neutral endoproteases able to hydrolyze all the components of the ECM . Increased MMP secretion may facilitate migration of inflammatory cells into the intima [7–10]. In particular, MMP-9 is expressed in the most symptomatic coronary artery lesions , but not in the normal arterial wall . MMP-9 activity is inhibited when MMP-9 binds to its tissue inhibitor TIMP-1, which preserves the extracellular matrix homeostasis in the arterial wall. An imbalance between MMPs and TIMPs is likely to participate in atherogenesis .
The role of monocyte adhesion and the MMP/TIMP balance in vascular disease is best studied by experimental models, in which a direct contact between monocytes and endothelial cells is used to induce monocyte activation. This can be performed in vitro, as well as ex vivo, using monocyte/endothelial cell coculture  and monocyte/whole artery culture  systems, respectively.
Interleukin-10 (IL-10) is a potent anti-inflammatory cytokine, which inhibits the synthesis of pro-inflammatory cytokines by activated monocytes. IL-10 was shown to reduce the adhesiveness of monocytes to stimulated endothelial cells in vitro [14,15], via an inhibitory effect on endothelial cell adhesion molecules ICAM-1 and VCAM-1 . In addition, IL-10 knock-out mice develop severe atherosclerosis in response to hypercholesterolemia [15,16], and systemic administration of recombinant IL-10 prevents intimal hyperplasia after stent implantation in hypercholesterolemic rabbits . In the latter case, the protective effect of IL-10 was mediated by a dramatic decrease in macrophage infiltration of the injured arterial wall . Furthermore, IL-10 inhibits MMP-9 expression and enhances the production of TIMP-1 in human monocytes .
The aims of the present study were: (1) To determine the effects of pharmacological concentrations of IL-10 on monocyte/endothelial cell adhesion, in vitro, as well as on monocyte/arterial wall adhesion ex vivo; and (2) To examine the effects of IL-10 on the MMP-9/TIMP-1 balance in monocyte–endothelial cell coculture systems.
2.1 Cell cultures
Peripheral blood monocytes were obtained from healthy donors by phlebotomy. Monocytes were purified by Ficoll–Hypaque gradient centrifugation (Pharmacia Fine Chemicals), then isolated by adherence to plastic at 37°C for 1 h in the presence of RPMI 1640 medium (BioMedia), as described . In these conditions, adherent cells contained <95% monocytes .
Mono Mac6 (MM6), a human monocytic cell line with characteristics of mature monocytes , were cultured in RPMI 1640, containing 10% fetal calf serum (FCS), 1% PSA (50 UI/ml Penicillin, 50 μg/ml Streptomycin, 25 μg/ml Amphotericin) and 2 mmol/l l-glutamine (all from Sigma).
Human umbilical vein endothelial cells (HUVEC) were harvested and cultured, as described . At confluency, cells were trypsinized then transferred to 24-well cluster plates. In all experiments, endothelial cells were used at second or third passage.
Two culture systems were studied. In the first system, 1 ml of medium containing 5.105 MM6 was seeded onto confluent HUVEC in serum-free RPMI 1640. In the second, MM6 were cultured in a 1 ml conditioned medium obtained from serum-free HUVEC cultures. Cultures were incubated for 18, 24 or 48 h. At each time point, culture medium was collected and centrifuged at 400 g for 5 min to eliminate nonadherent cells, then subjected to SDS–PAGE gelatin zymography, Western blot analysis, or ELISA. In all experiments, protein concentrations in culture media were determined using the Bradford assay (Bio-Rad), and equal amounts of protein were used for each sample.
2.3 In vitro binding assays
MM6 were fluorescently labeled with 0.75 μl/ml PKH2-GL (Green Fluorescent Cell Linker Kit, Sigma) for 10 min at room temperature and washed twice.
Fluorescent MM6 (5.105 cells) were seeded onto confluent HUVEC, with or without recombinant human IL-10 (kindly supplied by Schering-Plough Research Institute, Kenilworth, NJ), then incubated at 37°C for 24 h. Increasing IL-10 concentrations (50, 100, 150 ng/ml; n = 3 experiments for each concentration) were tested. Cocultures were washed four times, and fluorescence was counted with a spectrofluorimeter (Fluo-star) at 480 nm excitation and 510 nm emission. In additional experiments, HUVEC were pre-treated with identical concentrations of IL-10 for 30 min (n = 3 experiments for each concentration), then washed and incubated with fluorescent MM6. Results were expressed as a percentage of the fluorescence measured in 5.105 MM6.
2.4 Ex vivo binding assays
All experimental protocols were performed in accordance with the recommendations of the French Accreditation of Laboratory Animal Care (authorization N° 00577). Male Wistar rats (Iffa Credo) weighing 120–130 g were used. One group (control, n = 12) was fed a normal diet. The second group received the nitric oxide suppressor l-nitro arginine-methyl ester (l-NAME) in the drinking water (50 mg/kg/day for 8 weeks, n = 12). We have shown that this regimen induces chronic hypertension  and increases ex vivo adhesion of monocytes to the endothelium . Rats were euthanized by pentobarbital overdose, and both carotid arteries were excised and used for ex vivo monocyte binding assays. MM6 in HBSS buffer were fluorescently labeled with 0.75 μl/ml PKH26-GL (Red Fluorescent Cell Linker Kit, Sigma), preincubated with or without IL-10 (50, 100, 150 ng/ml) for 30 min, then seeded on carotid arteries (endothelial side up), with or without LPS (1 μg/ml). After 30 min incubation, carotid arteries were washed with HBSS and adherent MM6 were counted as described . For each experimental condition, n = 4 carotid arteries were studied.
2.5 MM6 adhesion molecule expression
Five hundred μl of 106/ml MM6 in HBSS were either left on ice or preincubated with IL-10 (150 ng/ml) or HBSS at 37°C for 30 min before cell stimulation with LPS (1 μg/ml) for an additional 30 min at 37°C (n = 3 experiments). β2 integrin and l-selectin expressions were quantified using FITC-conjugated antibodies directed against CD18 (Becton Dickinson) and CD62-L (Immunotech coultronics), respectively. Non-specific binding was determined with isotypic control antibodies. After a 30 min-incubation on ice, cells were washed with HBSS, resuspended in 1% paraformaldehyde and immediately analyzed by flow cytometry (FACScan, Becton Dickinson). Results were expressed as median fluorescence intensity (MFI), as described .
2.6 SDS–PAGE zymography
Gelatinolytic activities in cell culture media were measured as described . Each sample was loaded with 5 μg of protein. Results of n = 3 experiments are expressed as percentage of gelatinolytic activity in unstimulated monocytes. To discriminate between MMP and serine protease activity, additional gels prepared in presence of inhibitors of MMPs (EDTA 30 mmol/l) or serine proteases (Pefabloc 2 mmol/l), were loaded with the same protein templates and electrophoresed in parallel. Monocytes stimulated with phorbol-12 myristate-13 acetate (PMA, 10 ng/ml) for 18 or 24 h were used as positive controls for MMP-9 activity.
2.7 Western blot
Standard procedures were used . MMP-9 was detected in cell culture media (n = 4 experiments) using an anti-MMP-9 antibody (Ab-1, Calbiochem; 1:2000 dilution). Gels were exposed to X-ray films, and quantification of the autoradiograms was performed by densitometric scanning using the NIH Image 1.61 software.
TIMP-1 levels were measured in culture media by ELISA (Biotrak™ TIMP-1, Amersham).
Total RNA was extracted from MM6 and HUVEC using TRIzol™ (Life Technologies), and reverse transcribed as recommended by the manufacturer and 100-ng cDNA templates were amplified. The primers used to measure IL-10 receptor (IL-10R, 25 PCR cycles), MMP-9 (26 cycles), TIMP-1 (22 cycles) and GAPDH (18 cycles) mRNA levels are listed in Table 1. Semiquantitative analyses were performed using γ-ATP-33P-radiolabeled primers. Results of n = 3 experiments are expressed in arbitrary units and adjusted for GAPDH mRNA levels.
|mRNA||Sense primer||Antisense primer|
|MMP-9||5′GGCGCTCATGTACCCTATGT3′||5′TCA AAG ACC GAG TCC AGC TT3′|
|mRNA||Sense primer||Antisense primer|
|MMP-9||5′GGCGCTCATGTACCCTATGT3′||5′TCA AAG ACC GAG TCC AGC TT3′|
2.10 Statistical analysis
Data are presented as mean±S.D. Comparisons were performed by one-way ANOVA followed by a Fisher test, when appropriate. When two parameters may have influenced MMP-9/TIMP-1 expression – i.e., incubation time, culture system, or IL-10 treatment – a two-way ANOVA was performed to test the effect of each parameter and their interaction on MMP-9/TIMP-1 expression. A value of P<0.05 was considered statistically significant.
3.1 Effect of IL-10 on in vitro monocyte adhesion
The presence of IL-10R on MM6 was demonstrated by RT-PCR analysis (data not shown). In MM6/HUVEC cocultures, 150 ng/ml IL-10 (but not lower IL-10 concentrations) significantly reduced MM6 adhesion to HUVEC from 60 to 40% after 24 h incubation (Fig. 1). Pre-treatment of HUVEC with IL-10 for 30 min had no effect on MM6/HUVEC adhesion.
3.2 Effect of IL-10 on ex vivo monocyte adhesion
Non-stimulated MM6 adhered weakly to rat carotid arteries (Fig. 2). After a LPS challenge, the number of adherent MM6 increased by ∼10-fold, but only by ∼6-fold in presence of IL-10.
l-NAME treatment for 8 weeks induced a significant increase in systolic blood pressure (211±15 vs. 146±10 mmHg, P<0.05). The number of adherent MM6 to the carotid artery of l-NAME-treated rats increased by ∼4-fold, but only by ∼2-fold in presence of IL-10 (Fig. 2).
Preincubation of carotid arteries (from normal or l-NAME-treated rats) with IL-10 did not significantly alter the number of adherent MM6 (data not shown).
3.3 IL-10-induced modulation of adhesion molecule expression by MM6
As shown in Table 2, MM6 expressed CD18 and CD62-L at baseline. After LPS stimulation, CD18 expression increased ∼2.5-fold and CD62-L expression decreased ∼11-fold, as expected . In contrast, the expressions of CD18 and CD62-L were not significantly different from the baseline when MM6 were preincubated with IL-10 before LPS stimulation.
|Anti CD18 (MFI)||Anti CD-62L (MFI)|
|LPS (1 μg/ml)||106±4*||5±1*|
|LPS+IL-10 (150 ng/ml)||30±4†||50±10†|
|Anti CD18 (MFI)||Anti CD-62L (MFI)|
|LPS (1 μg/ml)||106±4*||5±1*|
|LPS+IL-10 (150 ng/ml)||30±4†||50±10†|
MFI, median of fluorescence intensity; n = 3 experiments.
P<0.001 vs. baseline.
P<0.001 vs. LPS.
3.4 Effect of MM6/HUVEC interaction on MMP-9 secretion
In these experiments, both freshly isolated human monocytes and MM6 were used. Stimulation of MM6 with PMA (10 ng/ml) resulted in ∼4- and ∼9-fold increase in MMP-9 activity at 18 and 24 h, respectively (Fig. 3A, lanes 8 and 9), and served as positive controls. Similar results were obtained with human monocytes (data not shown).
MM6/HUVEC and monocyte/HUVEC cocultures, as well as incubation of MM6 and monocytes with HUVEC-conditioned medium, time-dependently increased MMP-9 activity (two-way ANOVA: P<0.05), with a greater effect of cocultures (Fig. 3). There was no interaction between the effects of time and culture systems on MMP-9 activity (interaction term P = 0.5). In contrast, incubation of HUVEC with MM6 and monocyte-conditioned medium had no effect on MMP-9 activity (data not shown). MMP-2 activity remained unaffected.
3.5 Effect of IL-10 on MMP-9 in coculture systems
IL-10 reduced by ∼50% (two-way ANOVA: P<0.05) MMP-9 activity induced either by MM6/HUVEC and monocyte/HUVEC cocultures or cultures of MM6 and monocytes in HUVEC-conditioned medium for 24 h (Fig. 4A and B). In addition, Western blot analyses indicated a similar reduction in MMP-9 protein levels in both systems (Fig. 4C and D). The effect of IL-10 on MMP-9 was independent of the culture system (interaction term: P = 0.6).
Addition of IL-10 to MM6 and monocytes reduced by ∼25% basal MMP-9 activity, whereas IL-10 did not affect MMP-9 activity in HUVEC (data not shown).
3.6 Effect of IL-10 on TIMP-1
Basal levels of TIMP-1 were detected by ELISA, both in MM6-, monocyte- and HUVEC-conditioned medium. MM6/HUVEC and monocyte/HUVEC cocultures for 24 h, as well as culture of MM6 and monocytes in HUVEC-conditioned medium, significantly increased TIMP-1 levels (Fig. 5).
IL-10 increased significantly TIMP-1 levels in MM6 and monocytes, but not in HUVEC. In MM6/HUVEC and monocyte/HUVEC cocultures, as well as in cultures of MM6 and monocytes in HUVEC-conditioned medium for 24 h, the high TIMP-1 levels were further increased by IL-10 (two-way ANOVA: P<0.05). The effect of IL-10 on TIMP-1 was independent of the culture system (interaction term: P = 0.5).
3.7 Dose-effect of IL-10 on MMP-9 and TIMP-1 in MM6/HUVEC cocultures
Zymography analyses (Fig. 6A) indicated that the inhibitory effect of IL-10 on MMP-9 activity in MM6/HUVEC cocultures was dose-dependent (one-way ANOVA, P<0.01), with maximal MMP-9 inhibition at 150 ng/ml for 24 h incubation. IL-10 also stimulated TIMP-1 secretion in a dose-dependent manner (one-way ANOVA, P<0.01; Fig. 6A).
To investigate whether IL-10 modulates MMP-9 and TIMP-1 gene expression, we performed RT-PCR analyses. Basal MMP-9 mRNA levels were found in MM6, but not in HUVEC. MMP-9 mRNA levels increased by ∼3-fold in MM6/HUVEC cocultures (Fig. 6B). Addition of IL-10 resulted in a dose-dependent decrease in MMP-9 expression (one-way ANOVA, P<0.01), which returned to baseline levels for IL-10 concentrations ≥50 ng/ml.
TIMP-1 mRNA levels were ∼4-fold higher in MM6/HUVEC cocultures than in MM6 (Fig. 6B). IL-10 further increased TIMP-1 mRNA levels in MM6/HUVEC cocultures in a dose-dependent manner (one-way ANOVA, P<0.01).
In the present study, we have demonstrated that IL-10 inhibits monocyte/endothelial cell interactions at two levels. First, IL-10 reduces the adhesion of monocytes to endothelial cells in vitro and to the arterial wall ex vivo. Second, IL-10 regulates the MMP-9/TIMP-1 balance at the transcriptional level, with lower MMP-9 activity and higher TIMP-1 protein levels in IL-10-treated monocyte/endothelial cell cocultures.
IL-10 is a potent anti-inflammatory cytokine, with protective effects against atherosclerosis [15,16] and post-injury intimal hyperplasia . The mechanisms of IL-10 vasculoprotective effect [15–17,28], however, remain unclear. IL-10 expression is found in the most unstable atherosclerotic plaques in humans, and colocalizes with plaque areas showing low-grade inflammation . Moreover, transgenic expression of IL-10 reverses  or prevents  the development of atherosclerosis in hyperlipidemic IL-10 knock-out mice. More recently, we demonstrated that IL-10 inhibits macrophage infiltration after stent implantation in hypercholesterolemic rabbits , resulting in a dramatic reduction of intimal growth. Altogether, these results suggest a direct anti-inflammatory effect of IL-10 on the arterial wall in vivo, but do not provide a mechanistic insight on IL-10 effects on the interaction between inflammatory and vascular cells.
Data on the impact of IL-10 on monocyte/endothelial cell adhesion are scarce. In previous studies, protracted (<5 h) pre-treatment of endothelial cells with pharmacological concentrations of IL-10 were shown to dose-dependently reduce the adhesion of monocytes to stimulated endothelial cells in vitro [14,15]. In contrast with these previous studies, no effect of IL-10 on MM6/HUVEC adhesion was observed in the present study after a brief (30 min) pre-treatment of HUVEC with IL-10. Furthermore, in ex vivo binding assays, pre-treatment of MM6, but not of carotid arteries, with IL-10 inhibited MM6 adhesion to the arterial wall, suggesting a direct effect of IL-10 on MM6. Flow cytometry analysis suggest that both the downregulation of β2 integrins (CD18) and the prevention of l-selectin (CD62-L) shedding in monocytic cells, may be involved in this protective effect. Importantly, acute – LPS activation – or chronic – l-NAME treatment [24,27] – inflammatory stimuli were included in the design of these ex vivo experiments, in order to mimic the inflammatory reaction associated with severe arteriosclerosis. The ∼50% inhibitory effect of IL-10 on ex vivo monocyte adhesion to the arterial wall suggests that the anti-inflammatory effect of IL-10 on the arterial wall may operate in vivo, as well. It is unlikely that the high concentrations of IL-10 required to inhibit monocyte–endothelial cell adhesion in vitro and ex vivo (150 ng/ml) may be achieved spontaneously in the vicinity of the atherosclerotic plaque in vivo. However, such high concentrations of IL-10 have been measured in live animals after systemic injections of IL-10 . Therefore, our results may be relevant to the pharmacological modulation of arterial inflammation by IL-10, which has been demonstrated both in established atherosclerotic plaques  as well as in in-stent restenosis . The results of these ex vivo binding assays should be interpreted cautiously since the interaction between the human monocytic cell line MM6 and rat carotid arteries may not fully replicate the adhesion of human monocytes to human arteries. In addition, a direct inhibitory effect of IL-10 on the arterial wall cannot be ruled out by our ex vivo experiments, since the effect of human IL-10 on rat carotid arteries may be ‘sub-optimal’ due to species specificity.
Our data suggest that, in addition to its effect on cell adhesion, IL-10 profoundly modulates the MMP-9/TIMP-1 balance. IL-10 is a known inhibitor of MMP-9 and activator of TIMP-1 in LPS-stimulated monocytes . However, the influence of IL-10 on the MMP-9/TIMP-1 balance has never been investigated in monocyte/endothelial cell coculture systems, which recapitulate more closely the process of monocyte activation associated with the first stage of atherogenesis. Overall, both MMP-9 activity and TIMP-1 levels increased in monocyte/HUVEC cocultures. IL-10 reduced MMP-9 activity and increased TIMP-1 levels in this system, suggesting that IL-10 modulates the protease–antiprotease balance involved in the penetration of inflammatory cells within the arterial wall. Interestingly, MMP-9 activity and TIMP-1 levels also increased in monocytes cultured in HUVEC-conditioned medium, suggesting that direct monocyte/endothelial cell contact is not mandatory for MMP-9/TIMP-1 activation in monocytes. Rather, a soluble factor may be released by endothelial cells to stimulate MMP-9/TIMP-1 expression in monocytes . In addition, the inhibitory effect of IL-10 on MMP-9, as well as the stimulatory effect of IL-10 on TIMP-1, were of similar magnitude in monocyte/HUVEC cocultures versus monocytes cultured in HUVEC-conditioned medium. Hence, monocytes are likely the principal cellular target of IL-10 in our coculture system. Finally, our findings that the effects of IL-10 on the MMP-9/TIMP-1 balance are similar when MM6 or freshly isolated human monocytes are used, indicate that these effects are not specific to the transformed MM6 cell line.
Overall, our data provide two distinct mechanisms – i.e., inhibition of monocyte adhesion and modulation of MMP-9/TIMP-1 balance – which may be relevant for the protective effect of IL-10, at pharmacological dosage, against inflammatory vascular diseases, e.g., atherosclerosis  and restenosis . By reducing monocyte recruitment in the arterial wall, IL-10 may limit the inflammatory burden of the plaque, the main cellular component associated with plaque vulnerability [4,29], as well as the exaggerated intimal hyperplasia after angioplasty . In addition, IL-10-induced modulation of the MMP-9/TIMP-1 balance towards reduced MMP-9 and increased TIMP-1 expressions, is likely to prevent excessive extracellular matrix degradation , a deleterious biological event implicated in plaque rupture , as well as intimal growth . It must be underscore, however, that these potentially protective effects of IL-10, even at the highest concentrations, were only partial. Their clinical relevance is therefore uncertain.
In conclusion, the present study demonstrates that the anti-inflammatory cytokine IL-10 is a potent inhibitor of monocyte/endothelial cell interactions. Both the reduction in monocyte adhesion to endothelial cells and the favorable modulation of the MMP-9/TIMP-1 balance induced by IL-10, may participate to the protective effect of IL-10 against arteriosclerosis and restenosis.
The authors are grateful to Alexandre Lebeaut, MD (Schering-Plough, Kenilworth, NJ), for providing recombinant human IL-10. This study was supported in part by grants from the Fondation de l'Avenir (ET6-167) and the Fondation de France (97003880 and 98004139).