|Home | About | Journals | Submit | Contact Us | Français|
High-mobility group box 1 protein (HMGB1), an abundant nuclear protein, was recently established as a proinflammatory mediator of experimental sepsis. Although extracellular HMGB1 has been found in atherosclerotic plaques, its potential role in the pathogenesis of atherothrombosis remains elusive. In the present study, we determined whether HMGB1 induces tissue factor (TF) expression in vascular endothelial cells (ECs) and macrophages. Our data showed that HMGB1 stimulated ECs to express TF (but not TF pathway inhibitor) mRNA and protein in a concentration- and time-dependent manner. Blockade of cell surface receptors (including TLR4, TLR2, and RAGE) with specific neutralizing antibodies partially reduced HMGB1-induced TF expression. Moreover, HMGB1 increased expression of Egr-1 and nuclear translocation of NF-κB (c-Rel/p65) in ECs. Taken together, our data suggest that HMGB1 induces TF expression in vascular endothelial cells via cell surface receptors (TLR4, TLR2, and RAGE), and through activation of transcription factors (NF-κB and Egr-1).
Tissue factor (TF), a 263 amino acid membrane glycoprotein, is one major stimulator of the blood coagulation cascade. The initiation of coagulation is a major step in the pathogenesis of acute coronary syndromes. In line with this, TF protein levels and activity are elevated in atherosclerotic plaques particularly in the lipid-rich core [1–2]. Moreover, the levels of TF in atherectomy specimens from patients with unstable angina or myocardial infarction are higher than those in patients with stable angina , suggesting a potential involvement of TF in the initiation and propagation of acute coronary syndromes. Although the underlying causes for elevated TF expression in atherosclerotic plaques is not fully understood, its expression and activities can be enhanced by various stimuli (such as tumor necrosis factor-α (TNF-α), interleukin-1β (IL-1β), CD40 ligand, oxidized low-density lipoprotein (oxLDL), serotonin, histamine, urotensin II, and Chlamydia pneumoniae) [4–11]. However, the potential role of an abundant nuclear protein, high mobility group box 1 (HMGB1), in the regulation of TF expression was previously unknown.
HMGB1 has recently been established as a proinflammaotry mediator of lethal systemic inflammation (e.g., endotoxemia and sepsis), arthritis, and local inflammation [12–13]. A recent study demonstrated that extracellular HMGB1 levels are elevated in human atherosclerotic plaques, but not in normal arteries . In vitro, exposure of neutrophils, monocytes, or macrophages to HMGB1 leads to nuclear translocation of NF-κB and expression of various proinflammatory cytokines . In addition to active secretion from innate immune cells, HMGB1 can also be released by necrotic  or apoptotic cells , cholesterol-stimulated vascular smooth muscle cells , and oxidative stress-challenged macrophages . Notably, chronic inflammation and local accumulation of low-density lipoprotein (LDL) cholesterol are the two major characteristics of atherosclerosis . Moreover, oxidative stress and cell death occur commonly in human atherosclerotic lesions, and contribute to the pathogenesis of atherosclerosis [20–22]. However, it was previously unknown whether extracellular HMGB1 plays a pathogenic role in atherothrombosis.
There has been an on-going debate regarding the abilities of HMGB1 in inducing TF expression in human peripheral blood mononuclear cells [23, 24]. Here we provided evidence that HMGB1 enhanced TF expression and activities in vascular endothelial cells and macrophages. Furthermore, HMGB1 similarly enhanced Egr-1 expression and NF-κB (p65/c-Rel) nuclear translocation in vascular endothelial cells. Taken together, our experimental data suggest that HMGB1 might be involved in the pathophysiology of atherothrombosis.
Recombinant HMGB1 protein was expressed in E. coli, and purified to homogeneity as described previously . Purified HMGB1 was tested for endotoxin content by the chromogenic Limulus amebocyte lysate assay (Endochrome, Charles River), and contained < 500 pg endotoxin per microgram of rHMGB1.
Primary human coronary artery endothelial cells (HCAECs) (Clonetics) were cultured in EGM-2MV medium (Clonetics) with full supplements and 5% FBS as described . The purity of EC cultures was >99% as determined by immunostaining with the anti-von Willebrand factor monoclonal antibody (Dako, Carpinteria, CA). Twenty-four hours prior to the experiment, ECs were cultured in M199 supplemented with 0.1% human serum albumin. Human umbilical vein endothelial cell line (HUVEC-CS) was purchased from American Type Culture Collection (ATCC) and maintained at 37°C under 5% CO2 in RPMI 1640 medium (Gibco) supplemented with 10% fetal calf serum (Gibco).
Primary peritoneal macrophages were isolated from Balb/C mice as previously described . Murine macrophages were pre-cultured in RPMI 1640 medium (Gibco) supplemented with 10% FBS (Gibco) and 2 mmol/L glutamine. All experiments with recombinant HMGB1 were performed in the presence of 1 μg/mL polymyxin B (PMB) to neutralize activities of contaminating LPS.
Murine macrophage-like RAW264.7 cells, obtained from the American Type Culture Collection (ATCC, MD), were cultured in RPMI medium 1640 (Life Technologies, NY) supplemented with 10% heatinactivated FBS (GIBCO/BRL), 2mM glutamine (GIBCO/BRL).
Cells were lysed in 2× sodium dodecyl sulfate (SDS) buffer, and resolved by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). The proteins were transferred to a PVDF membrane (Millipore) by semidry transfer, and equal loading was confirmed by Ponceau S staining. Antibodies to human TF and TF pathway inhibitor (TFPI) (both from American Diagnostica Inc.) were used at 1:1000 dilution. Blots were normalized to GAPDH expression (1:2000 dilution; Sigma).
Total RNA was extracted using the TRIzol reagent (Invitrogen Corp., Carlsbad, CA), according to the manufacturer’s instructions. Reverse transcription was performed as described earlier . Total RNA (1 μg) was incubated with DNaseI, and real-time quantitative PCR was performed as described previously .
The actichrome TF activity assay kit (American Diagnostica Inc.) was used to determine the TF activity of cell lysates of ECs following HMGB1 (1 μg/mL) stimulaiton. Protein samples (30 μg) were assayed for the cellular TF activity in accordance with the manufacturer’s instructions.
For measuring the cell surface TF activity, ECs at the same density of 104 cells/well were seeded onto 96-well plates. Following overnight incubation, cells were treated with 1 μg/mL HMGB1 for the indicated time periods. Cell monolayers were washed twice with PBS, and the reagents were directly added to measure TF activity.
Murine macrophages were solubilized with 15 mmol/L octyl-D-glucopyranoside (Sigma) at 37°C for 15 min, and the procoagulant activities were evaluated using a one-step clotting assay, as described previously .
Primary ECs were pretreated for 30 min at 37°C with mouse anti-human TLR4 (eBioscience, Catalog Number: 16–9917; 20 μg/mL), mouse anti-human TLR2 (eBioscience, Catalog Number: 16–9922; 20 μg/mL), mouse anti-human RAGE (R&D Systems, Catalog Number: MAB11451; 20 μg/mL), or control IgG2a antibody (eBioscience, Catalog Number: 16–4724; 60 μg/mL). The cells were harvested at 5 h after stimulation for obtaining cell lysates and at 3 h after stimulation for RNA extraction.
Nuclear protein was isolated using the Nuclear Extraction Kit (Chemicon). NF-κB (c-Rel/p65 heterodimer) nuclear translocation and Egr-1 expression were measured by western blotting. Antibodies against human c-Rel, p65, and Egr-1 (all from Cell Signaling) were used at 1:1000 dilutions. Blots were normalized to PCNA expression (1:10000 dilution; Sigma).
The wild-type human TF promoter (−227) and a construct containing a mutation of the NF-κB (NF-κBm) or Egr-1 (Egr-1m) site have been described elsewhere . For transfection, the HUVEC-CS cell line (ATCC) was seeded at a density of 105 cells/well in a 24-well culture plate one day before the experiment. The cells were transfected with 400 ng DNA/well for 5 h using the Lipofectamine reagent (Gibco), according to the manufacturer’s protocol. All samples were co-transfected with equal amounts of a pRL-TK construct encoding Renilla luciferase (Promega) to compensate for variations in transfection efficiencies. Twenty-four hours after transfection, the cells were incubated with HMGB1 (10 μg/mL) for 12 h. The cells were harvested, and luciferase assays were performed using the dual luciferase reporter system (Promega) according to the manufacturer’s protocol.
Data are presented as mean ± SD. Statistical analysis was performed by the Student t test or ANOVA as appropriate. P < 0.05 was considered to be statistically significant.
To examine whether HMGB1 leads to endothelial procoagulation, TF expression levels were determined after exposure to various concentrations of HMGB1 for different time periods. Western blot analysis revealed that stimulation of primary ECs with HMGB1 (0.01 to 10 μg/mL) led to a concentration- and time-dependent induction of TF expression. The maximal effect (up to 24-fold increase) was observed 5–7 h following stimulation with HMGB1 (10 μg/mL) (Fig. 1A, 1B). Furthermore, real-time quantitative RT-PCR revealed that TF mRNA was induced in a time-dependent manner, indicating that HMGB1 induced TF expression at the transcriptional level. The maximal up-regulation of TF mRNA was observed after 1 h and then declined within 3 h of HMGB1 stimulation (Fig. 1C).
To determine the effects of HMGB1 on TF activity, primary ECs were stimulated with HMGB1 (1 μg/mL) for the indicated time periods (h), and the cellular TF activity was evaluated using an assay kit. Stimulation of primary ECs with 1 μg/mL HMGB1 resulted in significant increase in cellular TF activity within 5 –9 h of stimulation (Fig. 1D), which was in parallel with HMGB1-induced elevation of TF protein levels. Since intracellular TF and encrypted TF are functionally inactive , we tested the effects of HMGB1 on cell surface TF activity in a subsequent experiment. Our data indicated that HMGB1 (1 μg/mL) also dramatically enhanced the cell surface TF activity in ECs (Fig. 1E).
TFPI is the direct physiological inhibitor of the TF/FVIIa complex. Interestingly, stimulation with HMGB1 did not affect TFPI expression in ECs (Fig. 1F). As controls, the same lysates were blotted for TF expression, and it was found to be enhanced in response to HMGB1 (data not shown).
Primary murine macrophages were stimulated with HMGB1 (1 μg/mL), and the resulting changes in the levels of procoagulant activity, TF protein, and TF mRNA were determined using a clotting assay and real-time quantitative RT-PCR. Incubation with HMGB1 (1 μg/mL) increased procoagulant activity by 3 h after stimulation, and reached a maximum level at 7 h after stimulation (Fig. 2A). The TF mRNA levels were increased by 13-fold within 2 h after stimulation (Fig. 2B). In addition, we tested whether HMGB1 stimulated TF expression and activity in murine macrophage-like RAW 264.7 cells, and found that HMGB1 similarly increased TF expression in these cells (data not shown).
Because TF expression is very sensitive to LPS stimulation, we wish to exclude the possibility that contaminating LPS in the HMGB1 preparation (<500 pg LPS per microgram of HMGB1) contributes to the observed increase in TF expression in ECs. Accordingly, an effective inhibitor of LPS, polymyxin B (PMB), was employed in parallel experiments. When given at a low concentration (500 pg/ml), LPS induced a slight increase in TF expression in vascular endothelial cells (ECs). This LPS-mediated slight elevation of TF expression was completely abrogated by co-incubation with PMB (1 μg/mL) (Fig. 3A). In contrast, the elevated TF expression induced by rHMGB1 (1 μg/mL) was not inhibited by PMB (1 μg/mL) (Fig. 3A).
In macrophage cultures, LPS also increased TF expression even when given at a low concentration (500 pg/ml) (Fig. 3B). Similarly, co-incubation with PMB (1 μg/mL) completely blocked LPS-mediated increase in TF expression, but did not affect the expression of TF induced by rHMGB1 (1 μg/mL) (Fig. 3B). Collectively, these data indicated that HMGB1-induced TF expression was not likely due to contamination of LPS.
HMGB1 is known to exert most of its actions through three cell surface receptors: TLR4, TLR2, and RAGE. To evaluate the potential involvement of these receptors in HMGB1-induced TF expression, primary ECs were pre-treated with TLR4-, TLR2-, RAGE-specific neutralizing antibodies. Pre-incubation with each of the three antibodies uniformly attenuated HMGB1-induced TF mRNA expression in ECs (Fig. 4A). Consistent with these observations, the HMGB1-induced increase in TF surface activity in ECs was also partly inhibited by these specific neutralizing antibodies (Fig. 4B). In contrast, the irrelevant control antibody did not alter HMGB1-induced TF expression even when given at similar or 3-fold higher concentrations (data not shown). Taken together, these data indicated that TLR4, TLR2, and RAGE are involved in HMGB1-induced TF expression in ECs.
Two transcription factors, i.e., NF-κB (c-Rel/p65) and Egr-1, are responsible for regulating TF expression in response to various stimuli [6–7, 10–11, 28–30]. Since NF-κB (c-Rel/p65) activation requires nuclear translocation of the p65 and c-Rel subunits and Egr-1 transcriptional activation depends on its transient expression, we determined nuclear levels of p65, c-Rel, and Egr-1 by Western blotting analysis. Following stimulation with HMGB1 (1 μg/mL), nuclear levels of p65 and c-Rel were significantly increased within 1–2 h (Fig. 5A and 5B). Interestingly, HMGB1 (1 μg/mL) also significantly induced Egr-1 protein and mRNA expression in a time-dependent manner (Fig. 5C and 5D).
To determine the receptors involved in HMGB1-induced nuclear translocation of NF-κB (c-Rel/p65) and expression of Egr-1, ECs were pretreated for 30 min with blocking antibody (AB) against TLR4 (20 μg/mL) and TLR2 (20 μg/mL) or RAGE (20 μg/mL), or irrelevant AB (60 μg/mL) before stimulating with HMGB1 (10 μg/mL). Interestingly, pre-incubation with anti-TLRs antibody (20 μg/mL of anti-TLR4 plus 20 μg/mL of anti-TLR2) almost completely abolished HMGB1-induced nuclear translocation of p65, but only slightly reduced HMGB1-induced expression of Egr-1. In contrast, pre-incubation with anti-RAGE antibody significant reduced HMGB1-induced expression of Egr-1, but did not reduce HMGB1-induced nuclear translocation of p65 significantly (Fig. 5E and 5F).
To further investigate the role of NF-κB and Egr-1 in the regulation of TF expression, we transiently transfected HUVEC-CS cell line (ATCC) with plasmids containing the TF promoter region linked to a luciferase reporter. The reporter constructs contained either wild-type or mutated NF-κB and Egr-1 binding sites. HMGB1 stimulation led to a 5.1-fold increase in luciferase reporter activity in ECs tranfected with plasmid containing wild-type NF-κB and Egr-1 binding sites (Fig. 6). In contrast, HMGB1 stimulation led to a 1.0–1.5-fold increase in luciferase reporter activity in ECs tranfected with plasmid containing mutated NF-κB or EGr-1 binding site (Fig. 6), supporting an important role for NF-κB (c-Rel/p65) and Egr-1 in the regulation of HMGB1-mediated TF expression in ECs.
Emerging evidence has suggested a potential role for HMGB1 in the pathogenesis of atherosclerosis . Here we demonstrated that HMGB1 induces TF expression and activation in vascular endothelial cells and macrophages in a time-dependent fashion. In response to HMGB1 stimulation, transcription factors NF-κB (c-Rel/p65) and Egr-1 are immediately activated (within 1 h), which is followed by sequential increase in TNF mRNA and protein levels after 1 and 5–7 h, respectively. In light of the abundant accumulation of extracellular HMGB1 in human atherosclerotic lesions , we propose that extracellular HMGB1 may contribute to elevated TF expression levels in atherosclerotic plaques.
The intricate molecular mechanisms underlying regulation of TF expression are complex. For instance, TF expression is potentially counter-regulated by tissue factor pathway inhibitor (TFPI). However, several inducers (e.g., TNF-α, serum amyloid A, and histamine) for TF expression exhibit distinct influence on TFPI expression in endothelial cells [9, 25]. Notably, HMGB1 stimulation did not affect TFPI expression in ECs, suggesting that HMGB1-induced TF expression may favor pronounced procoagulative changes.
Extracellular HMGB1 is known to interact with three cell surface receptors: TLR4, TLR2, and RAGE. Recent studies have suggested that TLR 2 and TLR 4 are predominantly involved HMGB1-mediated activation of innate immune cells; whereas RAGE may play an important role in HMGB1-mediated haptotactic cell migration . Our experimental data suggested that all three receptors contributed to HMGB1-induced TF expression, supporting an important role for RAGE in atherosclerosis. These observations were consistent with previous notions that many RAGE ligands are effective inducers for TF expression in macrophage and endothelia cells [33–35].
Another interesting finding of this study was that HMGB1 could stimulate expression of Egr-1, a zinc finger transcription factor implicated in the regulation of TF expression [6, 11, 28]. Indeed, mutation of Egr-1 binding sites in the TF gene promoter significantly reduced its promoter activity, supporting an important role for Egr-1 in the regulation of HMGB1-mediated TF gene expression. Notably, Egr-1 is abundantly expressed in human atherosclerotic lesions , and contributes atherosclerosis by regulating expression of growth factors (e.g., platelet-derived growth factor and basic fibroblast growth factor), cytokines (e.g., TNF-α), and adhesion molecules (e.g., ICAM) . Thus, the HMGB1-mediated Egr-1 activation is likely to have broader effects on initiation and progression of atherosclerosis.
In addition, HMGB1 also induced nuclear translocation and activation of NF-κB (c-Rel/p65), another transcription factor involved in the regulation of TF expression in response to various stimuli [6, 10, 11, 30]. Similarly, mutation of the NF-κB binding site in the TF gene promoter also significantly reduced its promoter activity, supporting an important role for NF-κB in the regulation of HMGB1-mediated TF expression.
In conclusion, our data indicated that HMGB1 induces TF expression via three HMGB1 receptors (TLR4, TLR2, and RAGE), and potentially through activation of transcription factors such as NF-κB (p65/c-Rel) and Egr-1. Together, these experimental data provide evidence for the hypothesis that HMGB1 might play an important role in the pathogenesis of atherothrombosis.
This work was supported by grants from National Natural Science Foundation of China (30770905), Mittal Research Foundation of Central South University (06MX25), and in part by the National Institute of General Medical Sciences (R01GM063075 and R01GM070817 to HW).