Toll-Like Receptors 2-Deficient Mice Are Protected Against Postischemic Coronary Endothelial Dysfunction

【摘要】  Objectives- Toll-like receptors (TLR) 2 are expressed in cardiac and inflammatory cells, and regulate leukocyte function. Because leukocyte adhesion is a critical event in endothelial injury induced by ischemia/reperfusion (I/R), we assessed whether TLR2 were involved in I/R - induced coronary endothelial injury.

Methods and Results- Ischemia - reperfusion markedly decreased NO-mediated coronary relaxations to acetylcholine assessed ex vivo. In contrast, in TLR2 deficient mice, I/R paradoxically improved the NO-mediated responses to acetylcholine. To precise the cellular compartment expressing TLR2 which is involved in endothelial injury, we developed bone-marrow chimeric mice by transplanting TLR2 -/- bone marrow to WT mice or WT bone marrow to TLR2 -/- mice and submitted them to I/R 5 weeks after transplant. Both chimeric mice displayed similar protection as TLR2 -/- mice against I/R-induced endothelial dysfunction, suggesting a role of TLR2 expressed on both non-bone marrow cells (in our case presumably endothelial cells and/or cardiomyocytes) and cells of bone marrow origin (presumably neutrophils). TLR2 deficiency was also associated with a smaller infarct size, and reduced reperfusion-induced production of reactive oxygen species and leukocyte infiltration.

Conclusions- TLR2 contribute to coronary endothelial dysfunction after I/R, possibly through stimulation of neutrophil- (and free radical-) mediated endothelial injury.

Toll-like receptor 2 deficient mice (TLR2 -/- ) displayed a complete protection against endothelial dysfunction induced by ischemia/reperfusion (unaltered relaxing response to acetylcholine ex vivo). TLR2 -/- mice also had reduced infarct size, production of reactive oxygen species and leukocyte infiltration. TLR2 thus contribute to coronary endothelial dysfunction after reperfusion.

【关键词】  endothelium nitric oxide ischemia leukocytes immune system

Introduction

Cardiac ischemia causes myocardial necrosis but also leads to coronary vascular injury, and especially injury to the endothelium. Indeed, reperfusion is known to induce endothelial dysfunction, characterized by decreased endothelium-dependent coronary vasodilatation. 1,2,3,4 This dysfunction may aggravate the progression of atherosclerosis, favor coronary vasoconstriction/vasospasm and thrombosis, and thus may increase the risk of infarction/re-infarction. Such injury to the endothelium critically depends on the early oxidative burst occurring on reperfusion, 5 and on the acute inflammatory response characterized by an early adhesion of neutrophils to the endothelium. 6 Thus, elucidation of the mechanisms involved in these endothelial-neutrophil interactions after ischemia may help designing new interventions that protect the coronary vasculature during myocardial infarction.

Recent evidence identified Toll-like receptors (TLRs) as important contributors of neutrophil-mediated responses during inflammation. They are activated by many ligands encountered in sepsis (lipopolysaccharides, peptidoglycans), but also by endogenous factors produced on stress or cell damage such heat shock proteins 7,8,9 and high mobility group box 1 protein (HMGB1). 10,11 Among TLRs, TLR2 have been shown to be implicated in the inflammatory response (eg, NF B activation) to cell necrosis 12 or reactive oxygen species (ROS) in inflammatory cells but also in the endothelium and cardiomyocytes. 13,14 Furthermore, these receptors have recently been shown to aggravate ischemia/reperfusion injury in isolated perfused hearts, 15 and modulate chronic ventricular remodeling after myocardial infarction or after doxorubicin. 16,17

Based on the evidenced mentioned above, it is tempting to speculate that TLRs, and especially TLR2 contribute to neutrophil-endothelial interactions and thus to endothelial injury. However, the role of TLR2 in postischemic coronary endothelial injury has not yet been tested.

The purpose of our study was thus to assess the role of TLR2 in postischemic coronary endothelial injury. We first developed and characterized a new model for evaluating coronary endothelial dysfunction after ischemia/reperfusion in mouse coronary arteries, which allowed us to test whether this dysfunction persists in mice genetically deficient for TLR2, and to correlate the functional changes with changes in infarct size, inflammatory response and oxidative stress.

Methods

Experimental Animals

Male TLR2 -/- mice 18 were obtained from the EMMA (European Mouse Mutant Archive) network in Orleans, France. They were compared with their respective wild-type (WT) control mice with the same C57BL/6 background. Body weights at the time of the experiments were in the range of 22 to 26 g.

Murine Cardiac Ischemia-Reperfusion

Mice were anesthetized by intraperitoneal injection of a mixture of ketamine (150 mg/kg) and xylazine (6 mg/kg). Animals were ventilated with a rodent microventilator (Minivent 885, Hugo Sachs Electronics, March, Germany) and the body temperature maintained by use of a 37°C warm plate. Ischemia was achieved by ligating the left anterior descending artery proximal to its origin, using an 8 to 0 prolene suture which was tied off during 30 minutes and then removed to allow reperfusion for an hour. Successful ischemia was confirmed by visual observation of cyanosis, whereas successful reperfusion was confirmed by visualizing reactive hyperemia. In some experiments, the right jugular vein was catheterized to continuously infuse saline or the ROS scavenger mercaptopropionyl glycine (MPG, 20 mg/kg/h), starting 20 minutes before reperfusion. This was based on our previous experiments which showed that MPG prevented reperfusion injury to the coronary endothelium in rats. 5 MPG was prepared fresh daily, diluted in saline and adjusted to pH 7.4 with NaOH. Different sets of mice were used for the evaluation of endothelial function, infarct size, neutrophil accumulation and production of free radicals.

Generation of Bone-Marrow Chimeric Mice

To precise the cellular localization of TLR2 involved in endothelial injury, we developed bone-marrow chimeric mice. Briefly, at 5 to 6 weeks of age, WT mice or TLR2 -/- mice were sub-lethally irradiated (9 Gy) and injected intravenously the next day with 7 x 10 6 bone-marrow cells from the indicated donor mice (WT or TLR2 -/- ). Five weeks after grafting, mice were subjected to I/R and evaluation of endothelial function as described above.

In Vitro Vascular Studies

Coronary endothelial function was assessed on the basis of vascular studies performed previously by our group in rats. 4,5 At the end of the experiment, the heart was removed and immediately placed in cold, oxygenated Krebs buffer. A 1 mm long segment of the coronary artery distal to the site of occlusion was carefully dissected and mounted in a small vessel myograph for isometric tension recording (JP Trading Aarhus, Denmark). For this purpose, the artery was threaded onto two 25 µm tungsten wires. Normalization procedure was performed after an equilibration period, as previously described. 4 After a 30 minute equilibration period, the vessels were contracted with 10 -5 mol/L serotonin (which does not induce any endothelium-dependent relaxing responses in this model) before applying increasing concentrations of acetylcholine (10 -8 to 3 x 10 -5 mol/L). Arteries were then washed to allow concentration-response curves in response to an NO donor, sodium nitroprusside (SNP; 10 -9 to 10 -5 mol/L), to assess the endothelium-independent relaxations. Some arteries were incubated with the NO synthase inhibitor N G -Nitro L-Arginine (LNNA 10 -4 mol/L for 30 minutes) to assess the contribution of NO to the relaxing responses to acetylcholine.

Assessment of Infarct Size

At the end of reperfusion, the suture was tied again and 2% Evans blue dye was infused in a retroperfused manner to delineate the area at risk. The left ventricle was then removed, gently frozen in cold isopentane and sliced into 1 mm cross sections. The slices were then incubated for 20 minutes with 1% triphenyltetrazolium chloride (TTC) solution ( -Aldrich) at 37°C. The infarcted area was determined according to its pale, white coloration compared with viable (red) myocardium and evaluated, as well as the area at risk, using computerized planimetry.

Myocardial Myeloperoxidase Activity Assay and Leukocyte Infiltration

Myeloperoxidase (MPO) activity was measured as a marker of myocardial neutrophil infiltration. After 60 minutes reperfusion, myocardial samples were isolated from the previously ischemic tissue, delineated as described above for infarct size, and were immediately frozen. Samples were later placed in 0.5 mL of a 50-mmol/L potassium PBS (pH 6.0) containing 0.5% hexadecyltrimethyl ammonium bromide, homogenized, and centrifuged for 30 minutes at 14600 rpm at 4°C. The supernatants were mixed 1:9 (vol/vol) with 50 mmol /L PBS (pH 6.0) containing 0.167 mg/mL o-dianisidine and 0.0005% hydrogen peroxide; the absorbance change was performed at 40 and 100 seconds at a wavelength of 460 nm.

In parallel, leukocyte infiltration was assessed in 4 µm frozen histological sections incubated with an anti CD45 antibody (Ly5, Becton Dickinson) and revealed using N-Histofine simple stain mouse MAXPO (Microm Microtech).

Myocardial Production of Reactive Oxygen Species

Reactive oxygen species (ROS) production was evaluated by electron paramagnetic resonance (EPR) spectroscopy. Briefly, hearts were removed at different time points of reperfusion and quickly washed in a physiological solution. The center of the previously ischemic area (left ventricular free wall) was separated and incubated at 37°C for 60 minutes in Krebs-HEPES buffer containing 5 mmol/L Diethyl dithiocarbamate, 25 mmol/L deferoxamine and the spin probe 1-hydroxy-3-methoxycarbonyl-2,2,5,5-tetramethyl pyrrolidine hydrochloride (CMH, 500 µmol/L, Noxygen, Elzach, Germany). Spectra of the oxidized product of CMH (CM. ) were recorded from frozen samples with a Miniscope MS-200 (Magnetech) with the following acquisition parameters: Bo-field 3350 G; microwave frequency 9.78 GHz; microwave power 1 mW; modulation amplitude 5 G; sweep time, 120 s. Spectra intensity was measured from the height of the central line and expressed in µmol/L CM. produced for 60 minutes, normalized per milligram of protein.

In parallel, myocardial content of superoxide anions was visualized with the fluoroprobe dihydroethidine (DHE, Acros Organics, France). 19 Frozen cardiac sections taken at 2 minutes reperfusion were incubated with DHE (2.10 -6 mol/L) for 30 minutes at 37°C in a light-protected humidified chamber, then rinsed with PBS (1X, 5 minutes). Ethidium Bromide (the product of oxidation of DHE by superoxide anions) was excited at 488 nm with an emission spectrum of 610 nm. Images were obtained using a microscope equipped with a 585 nm long-pass filter fluorescence.

Cytokine Gene Expression

Total RNA was extracted from samples isolated from the previously ischemic tissue, delineated as described above for infarct size. Real time polymerase chain reaction (PCR) was performed with Light Cycler System (Roche). All PCR experiments were performed using the Light Cycler-FastStart DNA Master SYBR Green I kit (Roche). Quantitative PCR was performed in a total reaction volume of 20 µL per capillary for the LightCycler format. The number of cycles at which the best-fit line through the log-linear portion of each amplification curve intersects the noise band is inversely proportional to the log of copy number. The resulting 18S values were used as standard for presentation of the mRNA data of the different transcripts.

Oligonucleotide primers were designed according to the published sequences: MCP1: sense 5'-CCCAATGAGTAGGCTGGAGA-3'; antisense 5'-GCTGAAGACCTTAGGGCAGA-3'; tumor necrosis factor (TNF) : sense 5'-CTGGGACAGTGACCTGGACT-3'; antisense 5'-GCACCTCAGGGAAGAGTCTG-3'; IL6: sense 5'-AGTTGCCTTCTTGGGACTGA-3'; antisense 5'-TCCACGATTT- CCCAGAGAAC-3'; IL1ß: sense 5'-GCCCATCCTCTGTGA- CTCAT-3'; antisense 5'-AGGCCACAGGTATTTTGTCG-3'; TLR2: sense 5'-CTCCCACTTCAGGCTCTTTG-3'; antisense 5'-AGGAACTGGGTGGAGAACCT-3'; 18S: sense 5'-GTGGA- GCGATTTGTCTGGTT-3'; antisense 5'-CGCTGAGCCAGT- CAGTGTAG-3'.

Statistical Analysis

Data are expressed as mean±SEM. In in vitro experiments, n represents the number of animals from which the arteries were taken. Responses were compared using one-way analysis of variance (ANOVA) or 2-sample t test assuming equal variances. Differences were considered statistically significant when P <0.05.

Results

Coronary Vascular Function

Vascular Diameters and Contractile Responses

Normalized internal diameters of the arteries were not significantly different between groups (µm: WT: Sham 169±9, n=13, I/R 166±4, n=15; TLR2 -/- : Sham 175±3, n=10, I/R 173±5, n=10). Also, no significant differences were observed in the levels of precontraction obtained before the administration of acetylcholine (mN/mm : WT: Sham 1.33±0.12, n=13, I/R 1.28±0.11, n=15; TLR2 -/- : Sham 1.35±0.14, n=10, I/R 1.15±0.16, n=10).

Postischemic Endothelial Dysfunction in WT Mice

The relaxing responses to increasing concentrations of acetylcholine in normal mice are shown in Figure 1. In arteries isolated from sham-operated mice, acetylcholine induced concentration-dependent relaxations. These relaxations were virtually abolished by the NOS inhibitor LNNA, suggesting that they were entirely mediated by NO.

Figure 1. Relaxing responses induced by increasing concentrations of acetylcholine in the absence (A) or the presence (B) of the NOS inhibitor N G -nitro L-Arginine (LNNA 10 -4 M), or the NO donor sodium nitroprusside (SNP) (C) in serotonin-precontracted coronary arteries isolated from WT (circles) or TLR2 -/- (triangles) mice, either sham-operated (open symbols) or subjected to ischemia/reperfusion (I/R; solid symbols). The responses to acetylcholine or SNP are presented as a percent of the precontraction to serotonin. Data are mean±SEM. ** P <0.01 I/R WT vs sham WT; # P <0.05 and ## P <0.01 I/R TLR2 -/- vs I/R WT; $ P <0.05 I/R TLR2 -/- vs Sham TLR2 -/-.

In coronary arteries isolated from mice subjected to I/R, we observed marked and significant decreases in the NO-mediated relaxing responses to acetylcholine. In contrast, relaxing responses to the NO-donor SNP were not affected by previous I/R. This suggests that the responsiveness of the smooth muscle to NO was unaffected, and thus that I/R induced a selective dysfunction of the endothelial cells.

This I/R-induced endothelial dysfunction was partly prevented by in vivo infusion with the ROS scavenger MPG, which in contrast did not affect the relaxing response to SNP ( Figure 2 ). MPG also did not modify the vascular function of coronary arteries isolated from sham-operated mice.

Figure 2. Relaxing responses induced by increasing concentrations of acetylcholine (left) or the NO donor sodium nitroprusside (SNP; right) in serotonin-precontracted coronary arteries isolated from wild-type mice subjected to a sham surgery (open symbols) or ischemia/reperfusion (I/R; solid symbols), and treated either with saline (circles) or the ROS scavenger MPG (20 mg/kg/h; squares). The responses to acetylcholine or SNP are presented as a percent of the precontraction to serotonin. Data are mean±SEM. * P <0.05 and ** P <0.01 I/R vs Sham; # P <0.05 I/R MPG vs I/R.

Absence of Postischemic Endothelial Dysfunction in TLR2 Deficient Mice

The effect of I/R on the relaxing responses of coronary arteries isolated from TLR2 -/- mice are shown in Figure 1. In contrast to the marked impairment observed in WT mice, we observed not only an absence of endothelial dysfunction after I/R in TLR2 -/- mice, but also the fact that I/R paradoxically improved the relaxing responses to acetylcholine, and this increase was statistically significant at the concentration of 10 -6 mol/L. In arteries from TLR2 -/- mice subjected to I/R, the responses to acetylcholine were abolished by LNNA, suggesting that the increased relaxations induced by I/R could also be accounted for by an increased NO production. I/R did not affect the responses to SNP in TLR2 -/- mice.

Respective Role of TLR2 Expression in Non-Bone Marrow Versus Bone Marrow-Derived Cells

The effect of I/R on the relaxing responses of coronary arteries isolated from bone-marrow chimeric mice are shown in Figure 3. In control WT mice grafted with WT bone marrow cells, I/R induced as expected a marked reduction of the response to acetylcholine, which was absent in TLR2 -/- mice grafted with TLR2 -/- bone marrow (although the increased response to acetylcholine in this case was slightly less marked than that observed in ungrafted TLR2 -/- mice; see Figure 1 ). Both TLR2 -/- mice grafted with WT bone marrow and WT mice grafted with TLR2 -/- bone marrow also displayed similar increased responses to acetylcholine, which were not different from those observed in TLR2 -/- mice grafted with TLR2 -/- bone marrow.

Figure 3. Relaxing responses induced by increasing concentrations of acetylcholine (left) or the NO donor sodium nitroprusside (SNP; right) in serotonin-precontracted coronary arteries isolated from mice subjected to ischemia/reperfusion 5 weeks after irradiation and bone marrow grafting. Solid circles : WT mice grafted with WT bone marrow; solid squares: TLR2 -/- mice grafted with TLR2 -/- bone marrow: solid up triangles: TLR2 -/- mice grafted with WT bone marrow; solid down triangles: WT mice grafted with TLR2 -/- bone marrow. The responses to acetylcholine or SNP are presented as a percent of the pre-contraction to serotonin. Data are mean±SEM. * P <0.05 and ** P <0.01 vs WT/WT.

Infarct Size

The size of the area at risk was not different between wild-type and TLR2 -/- mice (% of left ventricle: 43±2, n=7 and 42±5, n=8, respectively). However, compared with wild type, infarct size was significantly smaller in TLR2 -/- mice (% of left ventricle: 13±1, n=7 and 7±2, n=8, respectively, P <0.05; % of area at risk: 31±4, n=7 and 16±3, n=8, respectively; P <0.05).

Cardiac Neutrophil Infiltration

Figure 4 (top) shows myocardial levels of MPO, a specific marker of activated neutrophils, in the ischemic tissue of mice subjected to sham surgery or I/R. Compared with sham-operated mice, I/R significantly increased MPO activity in WT mice. This increase was totally abolished in TLR2 -/- mice.

Figure 4. Top Neutrophil infiltration after I/R assessed by measuring myeloperoxidase (MPO) activity in cardiac tissue homogenates. MPO activity was expressed as units per 100 mg tissue and measured in sham-operated mice or mice subjected to I/R, either WT or TLR2 -/-. Data are mean±SEM. ** P <0.01 I/R WT vs Sham WT; ## P <0.01 I/R TLR2 -/- vs I/R WT. Bottom, Typical examples of leukocyte infiltration in the previously ischemic tissue, assessed in wild-type sham-operated mice, or in wild type or TLR2 -/- mice subjected to 30 minutes ischemia followed by 60 minutes of reperfusion.

Figure 4 (bottom) shows leukocyte infiltration in the previously ischemic zone. I/R was associated with a marked increase in leukocyte infiltration, which was especially concentrated around coronary arteries, both in the center and the periphery of the infarcts. This infiltration was reduced in TLR2 -/- mice.

Myocardial ROS Production After I/R

Levels of ROS within the area at risk, evaluated by EPR at different times during reperfusion, are shown in Figure 5 (top). In WT mice, reperfusion was associated with an early increase in ROS levels, which was not detectable at 1 minute but was significant at 2 and 5 minutes post reperfusion. This increase was however transient, because it was not observed after 60 minutes reperfusion. This increased ROS production, assessed at 2 minutes reperfusion, was abolished by in vivo infusion with the free radical scavenger MPG (data not shown), and was absent in TLR2 -/- mice, in which ROS levels were not different from those observed in the absence of ischemia. TLR2 deficiency also tended to decrease ROS levels in sham-operated mice, although this modest decrease was not statistically significant.

Figure 5. A, Cardiac ROS levels in sham-operated mice or mice subjected to 30 minutes ischemia followed by different time points of reperfusion (I/R; 1, 2, 5 and 60 minutes). B, Cardiac ROS levels in wild-type or TLR2 -/- mice, either sham-operated or subjected to 30 minutes ischemia followed by 2 minutes of reperfusion. ROS levels were assessed by EPR and expressed as µM CM. per mg proteins in one hour. * P <0.05 and ** P <0.01 vs Sham WT; # P <0.05 vs IR2' WT. C, Typical images (left; enlargement x 20) and semi-quantitative analysis (right) of DHE fluorescence within the previously ischemic zone, assessed in wild-type sham-operated mice, or in wild-type or TLR2 -/- mice subjected to 30 minutes ischemia followed by 2 minutes of reperfusion. * P <0.05 vs Sham.

Figure 5 (bottom) shows the detection of superoxide anion content within the previously ischemic tissue, assessed using DHE at 2 minutes reperfusion. I/R markedly increased DHE fluorescence, especially around coronary arteries, ie, in regions with marked leukocyte infiltration, suggesting that leukocytes are a major source of ROS in this setting. DHE fluorescence was partly reduced in TLR2 -/- mice.

Changes in Gene Expression

Figure 6 shows that I/R increased expression of TNF, IL6 and IL1ß, without affecting MCP-1. TLR2 deficiency abolished the increased IL1ß expression, but did not affect TNF, IL6 or MCP-1.

Figure 6. mRNA expression of TNF, IL6, IL1ß and MCP-1 in wild type or TLR2 -/- mice subjected to 30 minutes ischemia followed by 60 minutes of reperfusion. Data are expressed as a percent of the values obtained in wild-type, sham-operated mice. * P <0.05 vs Sham; # P <0.05 vs WT.

I/R did not affect TLR2 gene expression (TLR2/18S ratio: sham-operated mice: 1.31±0.16, n=8; I/R 1.23±0.15, n=9).

Discussion

The main results of our study are:

I/R induced a marked coronary endothelial dysfunction in mice, which was partly prevented by a ROS scavenger.

This dysfunction was abolished in mice deficient for TLR2. In fact in TLR2 deficient mice I/R paradoxically improved the coronary endothelial responses.

Coronary endothelial responses were also improved to the same extend in TLR2 -/- mice grafted with WT bone marrow and WT mice grafted with TLR2 -/- bone marrow mice.

The restored endothelial function was associated with smaller infarcts, reduced accumulation of neutrophils in the previously ischemic tissue, lower cardiac ROS levels, and reduced expression of the proinflammatory cytokine IL1ß.

Our study demonstrates that reperfusion after ischemia in mice is associated with an impairment of coronary endothelial responses, in agreement with what has been previously described in other species, including dogs, pigs and rats. The transposition of this model to mice is obviously important in the context of the studies in transgenic mice. Despite the difficulty of mounting and studying isolated mouse coronary arteries, we chose to perform our studies in isolated coronary artery segments after in vivo I/R, rather than in isolated buffer perfused hearts subjected to I/R in vitro, because circulating inflammatory cells are a central component of endothelial injury during I/R, and this cannot be properly mimicked in vitro. It should be noted however that our model assessed endothelial dysfunction in epicardial arteries, and the response to I/R may differ from that observed in whole hearts in which the vasodilatory responses that may be studied concern mostly the small size coronary arterioles.

At present, there have been only a limited number of studies evaluating endothelial function in mouse arteries. Most of those used perfused arteries 20,21,22,23 and very few used ring preparation similar to ours. 24 We chose to study endothelial function by assessing the responses to acetylcholine in arteries precontracted by serotonin, in an experimental design similar to the one we used before in rat coronary arteries. 4,5 In preliminary experiments we tested precontractions using the thromboxane analogue U46619 but found that this was associated with markedly reduced relaxations (compared with relaxations in serotonin-precontracted arteries). We also found that the relaxations to acetylcholine were entirely mediated by NO, in agreement with previous studies. 20,21,24 Interestingly, we found that the impairment observed in mouse coronary arteries was quantitatively similar to that we previously observed in coronary arteries isolated from rats subjected to a similar in vivo I/R protocol. 4,5 Using electron microcopy, we previously demonstrated in rats that this functional impairment reflected severe structural injury to the endothelium, 25 but at present we have not assessed this in mice I/R.

We show that endothelial dysfunction after I/R was absent in TLR2 -/- mice. To the best of our knowledge, this is the first demonstration of a central deleterious role of TLR in post-ischemic endothelial injury. Based on what we know on the mechanisms of endothelial injury after I/R, it is likely that the endothelial protection associated with TLR deficiency is the consequence of the decreased reperfusion-induced oxidative stress and acute neutrophil adhesion. This hypothesis is supported by our experiments showing that TLR2 -/- mice display less leukocyte infiltration, reduced ROS levels and reduced expression of IL1ß within the previously ischemic tissue. Once stimulated, TLR may activate various inflammatory signals in the heart (such as NF B 14 ) and trigger neutrophil activation, adhesion and accumulation, 13 and there is substantial evidence that neutrophil-endothelial interactions are essential mediators of endothelial injury during ischemia/reperfusion. 26,27 Interestingly, recent experiments suggests that TLR2 activation in cultured human endothelial cells induces a rapid Weibel-Palade body exocytosis, especially characterized by externalization of P-selectin, 28 which has been shown to be an essential component of reperfusion injury to the endothelium. 29,30

In parallel to the endothelial protection that may occur secondary to the decreased adhesion, it is likely that the endothelial protection may per se contribute to the decreased neutrophil adhesion because NO is a potent inhibitor of leukocyte function. 31 Thus, TLR2 deficiency may interfere with the ?vicious circle? of the neutrophil/endothelial interactions during reperfusion, ie, with the fact that endothelial dysfunction favors neutrophil adhesion which in turn aggravates endothelial dysfunction/injury.

Although the mechanisms by which acute ischemia/reperfusion activates the TLR2 pathway are still unclear, the short time course of the experiments excludes a possible upregulation of TLR2 expression (as supported by the lack of changes in TLR2 mRNA expression observed after reperfusion) and suggests that endogenous, constitutively expressed TLR2 in the heart may be rapidly recruited early during ischemia/reperfusion. Indeed there is recent evidence that TLR2 are activated within minutes by endogenous ligands such as HSP60/70, 7,8,9 and High Mobility Group Box 1 protein (HMGB1). 10,11 Furthermore, HMGB1 has been shown to be upregulated early after hepatic ischemia/reperfusion. 32 Once activated, TLR2 may then contribute to ROS-induced acute inflammatory response 12 and thus act as central actors of the ?amplification loop? between oxidative stress and neutrophil accumulation, ultimately leading to cardiomyocyte and coronary endothelial injury.

One intriguing and unexpected result of our study is that, in TLR2 -/- mice, I/R induced a paradoxical increase in the NO-mediated responses to acetylcholine. At present, the mechanisms of this surprising result are unknown. One possible explanation is that, in this situation of abolished acute inflammatory response (ie, abolished neutrophil accumulation) and thus in the absence of this major deleterious aspect of reperfusion, I/R may become a ?preconditioning? stimulus, leading to increased release of NO. Indeed, ischemic preconditioning has been shown to increase eNOS activity and to induce an early increase in NO release from the endothelium. 33

To precise the cellular localization of TLR2 involved in endothelial injury, we performed a series of experiments in bone marrow chimeric mice, which revealed similar protection in TLR2 -/- mice transplanted with WT bone marrow and WT mice transplanted with TLR2 -/- bone marrow mice. This suggests that TLR2 expressed in cells of non-bone marrow origin (in our case presumably endothelial cells and cardiomyocytes) or bone marrow-derived cells (in our case presumably neutrophils) both play a deleterious role on coronary endothelial function, and reinforces the strong protective effect of TLR2 inhibition in this setting.

The respective role of TLR2 expressed on bone marrow versus non-bone marrow derived cells has been previously assessed in the context of renal ischemia 34 and in atherosclerosis 35 (LDR -/- mice); in both cases protection was observed in TLR2 -/- mice but not in mice with loss of TLR2 expression only in bone marrow cells, suggesting that only TLR2 expression in non-bone marrow cells contributes to the deleterious effects of these receptors. There might be many explanations for this difference, for example because of tissue specificity or to the identity of the bone marrow cells involved: although our acute reperfusion study mostly involves neutrophils, atherosclerosis presumably mostly involves monocytes/macrophages or lymphocytes. In the renal ischemia/reperfusion model, the much longer duration of reperfusion (24 hours versus 1 hour in our experiments) may also recruit cells other than neutrophils, and especially monocytes.

The preserved endothelial function in TLR2 deficient mice was associated with a significant decrease in the extent of necrosis. This is consistent with recently published results obtained in isolated hearts showing that TLR2 deficient mice are protected against myocardial ischemia/reperfusion injury. 15 The parallel endothelial and myocardial protections may suggest that these are causally related. Indeed, the decreased necrosis per se may lead to decreased neutrophil adhesion (because necrosis is a potent inflammatory stimulus) and decreased oxidative stress, and thus indirectly induce endothelial protection. Conversely endothelial protection may favor perfusion and decrease inflammation and thus lead to reduced necrosis. At present, and in the absence of ?selective? treatments of endothelial dysfunction in I/R, it is difficult to clarify these complex interactions between vascular and myocardial injury during reperfusion.

Conclusion

This study demonstrates for the first time the deleterious role of TLR2 (present both in cardiac tissue and circulating cells) in coronary endothelial injury during cardiac ischemia and reperfusion. Development of pharmacological interventions that interfere with the expression and/or activity of these receptors may lead to new treatments of endothelial dysfunction in I/R, but also possibly in other diseases associated with inflammation or oxidative stress.

Acknowledgments

We thank Christophe Arnoult for his care and preparation of the mice. We also thank Marc Isabelle for his help in setting up the DHE staining, Sylvanie Renet for PCR and Françoise Lallemand for her help with histology.

Sources of Funding

This work was supported in part by a Servier Research grant.

Disclosures

Philippe Musette, Christian Thuillez and Vincent Richard were recepient of a Servier research grant (see funding).

【参考文献】
  Ku DD. Coronary vascular reactivity after acute myocardial ischemia. Science. 1982; 218: 576-578.

Van Benthuysen KM, McMurtry IF, Horwitz LD. Reperfusion after acute coronary occlusion in dogs impairs endothelium-dependent relaxation to acetylcholine and augments contractile reactivity in vitro. J Clin Invest. 1987; 79: 265-274.

Pearson PJ, Schaff HV, Vanhoutte PM. Long-term impairment of endothelium-dependent relaxations to aggregating platelets after reperfusion injury in canine coronary arteries. Circulation. 1990; 81: 1921-1927.

Richard V, Kaeffer N, Tron C, Thuillez C. Ischemic preconditioning protects against coronary endothelial dysfunction induced by ischemia and reperfusion. Circulation. 1994; 89: 1254-1261.

Kaeffer N, Richard V, Thuillez C. Delayed coronary endothelial protection 24 hours after preconditioning. Role of free radicals. Circulation. 1997; 96: 2311-2316.

Laude K, Beauchamp P, Thuillez C, Richard V. Endothelial protective effects of preconditioning. Cardiovasc Res. 2002; 55: 466-473.

Asea A, Rehli M, Kabingu E, Boch JA, Bare O, Auron PE, Stevenson MA, Calderwood SK. Novel signal transduction pathway utilized by extracellular HSP70: role of toll-like receptor (TLR) 2 and TLR4. J Biol Chem. 2002; 277: 15028-15034.

Vabulas RM, Ahmad-Nejad P, da Costa C, Miethke T, Kirschning CJ, Hacker H, Wagner H. Endocytosed HSP60s use toll-like receptor 2 (TLR2) and TLR4 to activate the toll/interleukin-1 receptor signaling pathway in innate immune cells. J Biol Chem. 2001; 276: 31332-31339.

Vabulas RM, Ahmad-Nejad P, Ghose S, Kirschning CJ, Issels RD, Wagner H. HSP70 as endogenous stimulus of the Toll/interleukin-1 receptor signal pathway. J Biol Chem. 2002; 277: 15107-15112.

Park JS, Svetkauskaite D, He Q, Kim JY, Strassheim D, Ishizaka A, Abraham E. Involvement of toll-like receptors 2 and 4 in cellular activation by high mobility group box 1 protein. J Biol Chem. 2004; 279: 7370-7377.

Park JS, Gamboni-Robertson F, He Q, Svetkauskaite D, Kim JY, Strassheim D, Sohn JW, Yamada S, Maruyama I, Banerjee A, Ishizaka A, Abraham E. High Mobility Group Box 1 protein (HMGB1) interacts with multiple Toll like receptors. Am J Physiol Cell Physiol. 2006; 290: C917-C924.

Li M, Carpio DF, Zheng Y, Bruzzo P, Singh V, Ouaaz F, Medzhitov RM, Beg AA. An essential role of the NF-kappa B/Toll-like receptor pathway in induction of inflammatory and tissue-repair gene expression by necrotic cells. J Immunol. 2001; 166: 7128-7135.

Fan J, Frey RS, Malik AB. TLR4 signaling induces TLR2 expression in endothelial cells via neutrophil NADPH oxidase. J Clin Invest. 2003; 112: 1234-1243.

Frantz S, Kelly RA, Bourcier T. Role of TLR-2 in the activation of nuclear factor-KB by oxidative stress in cardiac myocytes. J Biol Chem. 2001; 276: 5197-5203.

Sakata Y, Dong JW, Vallejo JG, Huang CH, Baker JS, Tracey KJ, Tacheuchi O, Akira S, Mann DL. Toll-like receptor 2 modulates left ventricular function following ischemia-reperfusion injury. Am J Physiol Heart Circ Physiol. 2007; 292: H503-H509.

Shishido T, Nozaki N, Yamaguchi S, Shibata Y, Nitobe J, Miyamoto T, Takahashi H, Arimoto T, Maeda K, Yamakawa M, Takeuchi O, Akira S, Takeishi Y, Kubota I. Toll-like receptor-2 modulates ventricular remodeling after myocardial infarction. Circulation. 2003; 108: 2905-2910.

Nozaki N, Shishido T, Takeishi Y, Kubota I. Modulation of doxorubicin-induced cardiac dysfunction in Toll-like receptor-2-knockout mice. Circulation. 2004; 110: 2869-2874.

Werts C, Tapping RI, Mathison JC, Chuang TH, Kravchenko V, Saint Girons I, Haake DA, Godowski PJ, Hayashi F, Ozinsky A, Underhill DM, Kirschning CJ, Wagner H, Aderem A, Tobias PS, Ulevitch RJ. Leptospiral lipopolysaccharide activates cells through a TLR2-dependent mechanism. Nat Immunol. 2001; 2: 346-352.

Isabelle M, Vergeade A, Moritz F, Dautreaux B, Henry JP, Lallemand F, Richard V, Mulder P, Thuillez C, Monteil C. NADPH oxidase inhibition prevents cocaine-induced up-regulation of xanthine oxidoreductase and cardiac dysfunction. J Mol Cell Cardiol. 2007Jan 9; .

Lamping KG, Wess J, Cui Y, Nuno DW, Faraci FM. Muscarinic (M) receptors in coronary circulation: gene-targeted mice define the role of M2 and M3 receptors in response to acetylcholine. Arterioscler Thromb Vasc Biol. 2004; 24: 1253-1258.

Lamping KG, Nuno DW, Shesely EG, Maeda N, Faraci FM. Vasodilator mechanisms in the coronary circulation of endothelial nitric oxide synthase-deficient mice. Am J Physiol Heart Circ Physiol. 2000; 279: H1906-H1912.

Liu JQ, Zelko IN, Folz RJ. Reoxygenation-induced constriction in murine coronary arteries: the role of endothelial NADPH oxidase (gp91phox) and intracellular superoxide. J Biol Chem. 2004; 279: 24493-24497.

Zhang C, Xu X, Potter BJ, Wang W, Kuo L, Michael L, Bagby GJ, Chilian WM. TNF- contributes to endothelial dysfunction in ischemia/reperfusion injury. Arterioscler Thromb Vasc Biol. 2006; 26: 475-480.

Chataigneau T, Feletou M, Huang PL, Fishman MC, Duhault J, Vanhoutte PM. Acetylcholine-induced relaxation in blood vessels from endothelial nitric oxide synthase knockout mice. Br J Pharmacol. 1999; 126: 219-226.

Kaeffer N, Richard V, François A, Lallemand F, Henry JP, Thuillez C. Preconditioning prevents chronic reperfusion-induced coronary endothelial dysfunction in rats. Am J Physiol Heart Circ Physiol. 1996; 271: H842-H849.

Ma XL, Weyrich AS, Lefer DJ, Buerke M, Albertine KH, Kishimoto TK, Lefer AM. Coronary endothelial and cardiac protective effects of a monoclonal antibody to intercellular adhesion molecule-1 in myocardial ischemia and reperfusion. Circulation. 1992; 86: 937-946.

Banda MA, Lefer DJ, Granger DN. Postischemic endothelium-dependent vascular reactivity is preserved in adhesion molecule-deficient mice. Am J Physiol. 1997; 273: H2721-2725.

Into T, Kanno Y, Dohkan JI, Nakashima M, Inomata M, Shibata KI, Lowenstein CJ, Matsushita K. Pathogen recognition by Toll-like receptor 2 activates Weibel-Palade body exocytosis in human aortic endothelial cells. J Biol Chem. 2007[Epub ahead of print, PMID: 17227763].

Weyrich AS, Ma XY, Lefer DJ, Albertine KH, Lefer AM. In vivo neutralization of P-selectin protects feline heart and endothelium in myocardial ischemia and reperfusion injury. J Clin Invest. 1993; 91: 2620-2629.

Lefer AM, Weyrich AS, Buerke M. Role of selectins, a new family of adhesion molecules, in ischaemia-reperfusion injury. Cardiovasc Res. 1994; 28: 289-294.

Kubes P, Suzuki M, Granger DN. Nitric Oxide: an endogenous modulator of leukocyte adhesion. Proc Natl Acad Sci U S A. 1991; 88: 4651-4655.

Tsung A, Sahai R, Tanaka H, Nakao A, Fink MP, Lotze MT, Yang H, Li J, Tracey KJ, Geller DA, Billiar TR. The nuclear factor HMGB1 mediates hepatic injury after murine liver ischemia-reperfusion. J Exp Med. 2005; 201 (7): 1135-1143.

Xuan YT, Tang XL, Qiu Y, Banerjee S, Takano H, Han H, Bolli R. Biphasic response of cardiac NO synthase isoforms to ischemic preconditioning in conscious rabbits. Am J Physiol Heart Circ Physiol. 2000; 279: H2360-H2371.

Leemans JC, Stokman G, Claessen N, Rouschop KM, Teske GJ, Kirschning CJ, Akira S, van der Poll T, Weening JJ, Florquin S. Renal-associated TLR2 mediates ischemia/reperfusion injury in the kidney. J Clin Invest. 2005; 115: 2894-2903.

Mullick AE, Tobias PS, Curtiss LK. Modulation of atherosclerosis in mice by Toll-like receptor 2. J Clin Invest. 2005; 115: 3149-3156.

作者单位：Inserm U644 & Rouen University Hospital (J.F., K.L., J.-P.H., C.T., V.R.) and Inserm U519 (P.M.), Institute for Biomedical Research and IFRMP 2 University of Rouen, France; and Inserm U868 (V.D.-E., J.-F.A.), Toulouse France.