| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Vascular Biology Group, Division of Experimental Pathophysiology and Immunology, Department Biocenter, Innsbruck Medical University, Innsbruck, Austria
1 Correspondence: Vascular Biology Group, Division of Experimental Pathophysiology and Immunology, Department Biocenter, Innsbruck Medical University, Fritz-Pregl-Str. 3/4. OG. 6020 Innsbruck, Austria. E-mail: David.Bernhard{at}uibk.ac.at
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-tubulin. As a consequence of the microtubule collapse, cytoskeletal structures as well as intermediate filaments break down, leading finally to a contraction of vascular endothelial cells. We observed a smoke extract-induced, calpain-dependent degradation of the intracellular form of platelet-endothelial cell adhesion molecule 1/CD31, as well as a release of P-selectin/CD62P, IL-6, and IL-8 from endothelial cells into the supernatant. Increased levels of soluble CD62P and IL-6 are well known to be associated with smoking in humans. Increased permeability of the vascular endothelium is a crucial event in atherogenesis. This work highlights the compounds and mechanisms by which cigarette smoke induces leakiness of the vascular endothelium.—Bernhard, D., Csordas, A., Henderson, B., Rossmann, A., Kind, M., Wick, G. Cigarette smoke metal-catalyzed protein oxidation leads to vascular endothelial cell contraction by depolymerization of microtubules.
Key Words: cigarette smoke extract • endothelial • atherosclerosis • microtubules • oxidative stress • cytokines/chemokines
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
A variety of different models have been proposed to explain the proatherogenic activity of smoking. Smoking has been shown to cause modulation of lipid profiles (increase of LDL, triglyceride and decrease of HDL) (1)
, lipid oxidation and increase of antioxidized LDL antibodies (2)
, up-regulation of inflammation markers (3
, 4)
, modulation of platelet function (5
, 6)
, altered estrogen metabolism (7
, 8)
, increased carbon monoxide levels (6
, 9)
, modulated lymphocyte-endothelial interaction (10)
, modulated platelet-endothelial interaction (11)
, endothelial dysfunction (12
, 13)
, an increased number of infections (14)
, and lung disease (15)
. Taken together, these data suggest that smoking modulates a wide variety of different physiological systems toward an atherosclerosis-causing/accelerating pathophysiological status. These proatherogenic modulations were observed throughout most stages of the disease, suggesting that cigarette smoking acts as a multifactorial noxa.
The broad spectrum of smoking "side effects" can be easily explained by the fact that cigarette smoke contains a large number (
4800) of different compounds, many of which are bioactive. Besides the hydrophobic tar fraction, which comprises numerous mutagens like benzo[a]pyrenes (16)
, the hydrophilic fraction is thought to mainly contain the atherosclerosis-causing agents (17)
, although the identities of these compounds are not known. Hydrophilic fractions of cigarette smoke comprise nicotine, metals, and a large number of different oxidants and free radicals. The latter two have been shown to cause low density lipoprotein oxidation in vitro (18)
as well as in vivo (2)
, suggesting that smoking-mediated oxidative stress is an in vivo-relevant phenomenon.
In addition to signs of oxidative stress, the metals cadmium and lead are increased in smoker’s blood compared with nonsmokers (19)
. Although information on other metals is sparse and contradictory, smoking-increased concentrations of certain metals might multiply the oxidative stress by catalyzing protein oxidation (similar to iron in the Fenton reaction; see the IUPAC Compendium of Chemical Terminology 2003).
In the past, the impact of smoking on the vascular endothelium has mainly been studied by focusing on patient blood samples, and the number of in vitro cell biological studies is low. The limited number of standardized in vitro systems has so far hindered this research. To overcome this problem, we have developed a chemically defined in vitro system (20)
that allowed us to perform standardized in vitro experiments (21)
.
According to the "response to injury" hypothesis of atherosclerosis (22)
, the first event in the genesis of atherosclerosis is the impairment of vascular endothelial function. Previous work, including our own, has highlighted that cigarette smoke constituents cause vascular endothelial damage. Hydrophilic smoke fractions were shown to induce apoptosis of endothelial cells in some systems (23)
and necrosis in others (21
, 24)
.
In addition to endothelial cell death, which occurs only at high concentrations and did not start until 12 h of incubation, we observed a dramatic, yet reversible, contraction of endothelial cells already starting at low smoke constituent concentrations within minutes. Since this phenomenon leads to atherosclerosis-relevant endothelial denudation of the blood vessels and to aggregation of endothelial cells, we set out to examine the reasons for both processes. In the present project we analyzed the chemical and biological reasons for endothelial cell contraction and detachment in response to treatment with hydrophilic smoke constituents.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. The cells were routinely passaged in 0.2% gelatine-coated (Sigma, Steinheim, Germany) polystyrene culture flasks (Becton Dickinson, Meylan Cedex, France) in endothelial cell basal medium (CC-3121, BioWhittaker, Inc., Walkersville, MD, USA) supplemented with EGM SingleQuots supplements (including 2% FBS) and growth factors (CC-4133, BioWhittaker, Inc.) in a humidified atmosphere containing 5% CO2. For Western blot/OxyblotTM analyses, 3 x 105 HUVECs per well were seeded onto gelatine-coated 6-well plates (Becton Dickinson). For morphological analyses, 5 x 104 HUVECs per chamber/well were grown in 4-chamber Labtek slides (Nalge Nunc, Naperville, IL, USA) or 24-well plates (Becton Dickinson). The medium was replaced with fresh medium before the start of each experiment. Chemical reagents were purchased from Merck (Darmstadt, Germany) unless stated otherwise and were of analytical grade quality.
Preparation of aqueous cigarette smoke extracts (CSE)
CSE were prepared as described previously (20)
. Briefly, using a "smoking machine" developed and validated in our laboratory, the smoke of two commercially available filter cigarettes (Marlboro, Philip Morris Products S. A., Neuchatel, Switzerland; nicotine: 0.8 mg; tar: 10 mg; carbon monoxide: 10 mg) was "puffed" through 8 mL of prewarmed (37°C) cell culture medium without additives (control: air drawn through the medium). The extracts were routinely chemically fingerprinted by HPLC-MS. Before use, extracts were filtered through a 0.22 µm pore filter unit. The pH of the extracts was routinely checked and adjusted if the pH value differed from 7.2.
Live cell imaging
Live cell imaging was performed under cell culture conditions with the PH2 phase-contrast function of a Zeiss Axiovert 100m confocal laser scanning microscope. Pictures were taken every 15 s for 95 min. Image size was 230 x 230 µm and 512 x 512 pixels. Laser performance was 0.25 mW. A video file was then generated using Metamorph software.
Phase-contrast microscopy
For microscopic screening experiments, endothelial cells were seeded into 24-well plates (see above). Inhibitors were added 30 min before addition of CSE. Cells were analyzed by phase-contrast microscopy 45 and 90 min after the addition of CSE. Pictures were taken with the PH2 phase contrast function of a Nikon ECLIPSE TE 300 microscope attached to a digital imaging system (Nikon, Japan). The following inhibitors and concentrations were tested: aprotinin, 1 and 5 µg/mL; leupeptin, 1 and 5 µg/mL; pepstatin, 1 and 5 µg/mL; epoxomicine, 5, 50, and 250 µM; LLNL, 5, 50, 250, and 500 µM; doxycycline, 200, and 1000 µM; EDTA, 200, 500, and 1000 µM; NAC, 250, 500, and 1000 µM; -tocopherol, 5, 50, and 250 µM; ascorbic acid, 50, 250, and 1000 µM; resveratrol, 10 and 50 µM; atorvastatin, 1, 10, and 100 µM; catalase, 500, and 2000 U/mL; SOD, 500 and 2000 U/mL; aspirin, 10, 100, and 1000 µM; taxol, 1, 10, and 100 nM; 3-aminobenzamide, 2.5 mM; butyrate, 2 and 5 mM; guanosine-5'-O-(2-thiodiphosphate), 100 µM; bis-indoyl-maleimide, 2.5 and 10 µM; bungarotoxine, 20 and 200 nM; di-hydro-erythroidine, 1 and 10 µg/mL. With the exception of 100 nM taxol, we were not able to observe cytotoxic effects of the compounds within 2 h of treatment; EDTA led to CSE-independent cell contraction (only when no Ca2+ ions were supplemented).
Staining of cells and fluorescence microscopy
For cell structure analyses, HUVECs were treated as indicated, washed with phosphate-buffered saline (PBS; pH=7.2), and fixed and permeabilized with 99.5% acetone at – 20°C for 2.5 min. Cells were then allowed to dry for 30 min at room temperature (RT), followed by blocking with 0.1% bovine serum albumin (BSA; Sigma-Aldrich, Vienna, Austria) in PBS for 30 min at RT, stained with 10 µg/mL TRITC-phalloidin (for f-actin; Sigma-Aldrich) and SYTOX green (25 nM) (for nuclear DNA; Molecular Probes; Eubio; Vienna; Austria), anti-tubulin antibody (1:100; Calbiochem-Novabiochem Corporation, San Diego, USA), or anti-vimentin antibody (1:200; Sigma, Austria). After washing four times with PBS and applying FITC-labeled secondary antibodies, cells were washed again, embedded into gelatine solution (9+1: gelatine+A. bd.), and analyzed using a Nikon ECLIPSE E 800 fluorescence microscope attached to a digital imaging system (Nikon, Tokyo, Japan).
ELISA and FlowCytomix analyses
Detection of the soluble forms of CD31/PECAM-1 (by ELISA), P-selectin/CD62P (by ELISA), CD40L (by FlowCytomix), tPA (by FlowCytomix), VCAM-1 (by FlowCytomix), IL-6 (by FlowCytomix), IL-8 (by FlowCytomix), and MCP-1 (by FlowCytomix) was performed according to the manufacturers instructions (Bender MedSystems GmbH, Vienna, Austria).
Western blot
Western blot was performed as described previously (26)
. The primary antibodies used were -tubulin (Calbiochem-Novabiochem), CD31 (JC70A; DAKO), and CD62P (1E3; DAKO). The secondary antibody used was HRP-labeled rabbit anti-mouse IgG (DAKO).
Immunoprecipitation of CD31
To separate the cell surface form of CD31 from the intracellular form, 5 x 106 log-phase growing cells were harvested by mild trypsinization. The cells were blocked with 0.1% BSA in PBS for 15 min at RT. Anti-human endothelium-specific CD31 antibody (3 µg; clone: JC70A; DAKO, Denmark) was added to the cells. After washing twice in PBS, the cells were lysed on ice in 500 µL of a standard hypotonic lysis buffer (26)
. After removal of nuclei and nonlysed cells by centrifugation at 4°C, 500 x g for 5 min, 50 µL of protein A/G agarose (Oncogene Research Products, San Diego, CA, USA) was added to the supernatant of the lysed cells and incubated for 1 h rotating at 4°C. After antibody-mediated binding of surface CD31 to the agarose, the agarose was collected by centrifugation. The pellet was washed four times in lysis buffer and stored on ice. The supernatant was then incubated with anti-human endothelium-specific CD31 antibody (3 µg; clone: JC70A) for 1 h rotating at 4°C. After 1 h, 50 µL of protein A/G agarose was added and the supernatant was again incubated for 1 h. Finally, the agarose, including the antibody-attached intracellular CD31, was collected as described above. Laemmli buffer was added to agarose pellets, the mixture was heated (95°C, 3 min), and the samples were analyzed by Western blot.
OxyblotTM
To detect oxidative modifications of proteins, the OxyblotTM kit (Chemicon International, Hofheim, Germany) was used. This assay detects oxidation-dependent carbonyl groups within proteins. After derivatization of these sites with di-nitrophenyl hydrazones, oxidatively modified proteins were detected by an antibody-mediated process. Protein preparation and processing, including detection, was performed according to the manufacturer’s instructions and as outlined in Western blot.
Detection of intracellular oxidative stress
Intracellular oxidative stress was determined by online FACS analysis. Cells were incubated with 100 µM carboxy-H2DCF-DA (Molecular Probes) for 30 min under cell culture conditions. Cells were harvested and transferred into FACS tubes. Immediately after baseline determination, the indicated compounds/concentrations were added to the suspended cells, followed by FACS analysis in an argon laser-equipped FACScan (FL-1 channel; Becton Dickinson, San Jose, CA, USA).
Induced COUPLED PLASMA (ICP) analysis
Elemental analysis was performed with an ICP-OES (PU7000) using standard conditions.
Application of the cation exchanger Serdolit
The cation exchanger Serdolit (1 g, Serva, Germany) was modified from its Na+ salt form to the H+ form by the addition of 10 mL of 2 N HCl for 10 min. Serdolit was washed with double distilled water (ddH2O) until the pH of the supernatant was 7. After removal of residual supernatant, the smoke extract (prepared in ddH2O to avoid cell culture medium cation interaction with the exchanger) and the corresponding control were added to the activated Serdolit. After incubation for 10 min, Serdolit was sedimented (1xg) and the supernatant transferred into a fresh tube. The pH dropped to 1 in the smoke extract—not in the control (control pH=7)—clearly indicating the release and replacement of H+ with cations from the extract. The pH of the extract was adjusted to 7 and the same volume of ddH2O was added to the control to ensure the same dilution. The cation depleted solutions were then adjusted to isotonic conditions by adding 10x PBS before being applied to the cells. The time between CSE preparation and CSE application to the cells was the same within each experiment.
Application of thiopropyl Sepharose (TPS) and EDTA
TPS (0.5 g, Amersham) was allowed to rehydrate for 15 min in 5 mL of ddH2O. After swelling, the supernatant was removed, the smoke extract (prepared in ddH2O), or the corresponding control was added to the Sepharose. After incubation at RT for 15 min, Sepharose was sedimented (300xg; 5 min; RT) and the supernatant was transferred to a fresh tube. The pH was adjusted, the dilution corrected, and isotonic conditions restored by adding 10x PBS before the extracts were applied to the cells. The time between CSE preparation and CSE application to the cells was the same in each experiment.
Similarly, 50 mM EDTA was added to the smoke extract (prepared in ddH2O) and incubated for 10 min at RT. Residual/free EDTA was then blocked by addition of 50 mM CaCl2 to avoid quenching of cations that are crucial for cell-cell interactions. After pH adjustment and the volume correction, 10x PBS was added to ensure isotonic conditions and the extracts were applied to the cells.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Since cigarette smoke is known to contain a variety of different oxidants, we set out to test for the existence of oxidative stress as a potential reason for the phenomenon. The left panel of Fig. 1
demonstrates that proteins of cells treated with CSE (CSE+) are highly oxidatively modified in comparison to control cells (Co+) and that the detection method for oxidized proteins is specific (no derivatization was conducted in Co– and CSE–). To test for and quantify the magnitude of intracellular oxidative stress, we determined the potential of CSE to oxidize the oxidation-sensitive plasma membrane-permeable dye carboxy-H2DCF-DA (right panel, Fig. 1
). In comparison to H2O2 (data not shown), CSE showed a considerably high potential to induce intracellular oxidative stress. Calculation of the oxidative potential of CSE in comparison to H2O2 showed that 100% CSE is equivalent to a 2 mM H2O2 solution. Therefore, the oxidative potential of a 16% CSE solution corresponds to that of a 300 µM H2O2 solution. In experiments where we tried to mimic the CSE-dependent effects with H2O2, no effect was observed until 1 mM H2O2.
|
Since it is well known that some metals catalyze protein oxidation by oxidants and since increased levels of metals can be found in smoker’s blood (19)
, we examined the metal content of CSE. By ICP analysis we could detect the following metals in CSE (ranging from nM to µM concentrations): Al, Si, Ti, V, Cr, Fe, Ni, Co, Cu, Mn, Zn, Sr, Pb, Cd, and Ba.
We have previously shown (21)
that N-acetyl cysteine (NAC) is capable of inhibiting the contraction of vascular endothelial cells in response to CSE treatment. Since thiol groups (as contained in NAC) are known to function as antioxidants (a principle applied by the glutathione system in cells) and to bind metal ions (as used in metal detoxification therapy), we tested whether the application of thiol groups to fresh CSE, followed by removal of the thiol groups (including bound compounds) before CSE addition to the cells, would also inhibit cell contraction. The lower left set of images in Fig. 2
shows that not only NAC is protective for endothelial cells, but also TPS (for experimental details, see Materials and Methods), clearly indicating that thiol groups do not exert their protective potential by interfering with cellular signal transduction but inactivate endothelial cell damaging agents present in CSE. (TPS contains thiol groups that are attached to insoluble agarose beads, which were removed before the TPS-treated CSE was added to the cells.) When an equimolar combination of NAC and TPS was added to CSE, the protective effect disappeared, probably because the thiol groups of NAC and TPS bound to each other and were thereby inactivated.
|
Screening for inhibitors of CSE-mediated vascular endothelial cell contraction
To take a closer look at the efficacy of metal ion binding compounds, antioxidants, protease inhibitors, and inhibitors of cellular signal transduction to inhibit CSE-mediated cell contraction, we performed a compound screen (Table 1
and Fig. 2
).
|
The data from Table 1
show that neither the activation of extracellular proteases (blocked by aprotinin, leupeptin, or pepstatin) nor changes in proteasome or calpain activity (blocked by epoxomicine and LLnL/doxycycline) are crucially involved in CSE-caused cell contraction. Similarly, by the addition of catalase and superoxide dismutase (SOD) to CSE, extracellular (potentially CSE-contained) H2O2 and –OH groups were excluded as important players in this process.
However, the cation exchanger Serdolit as well as the cation chelating compound EDTA (for experimental details, see Materials and Methods) showed partial but clear inhibitory activity (lower right set of images of Fig. 2
), suggesting a crucial involvement of cations (metals) in this process.
Antioxidants showed either no effect at all (vitamin C, resveratrol) or only marginal protection (vitamin E, atorvastatin). There is experimental evidence that vitamin C and resveratrol, in contrast to the two other antioxidants, have pro-oxidative properties in the presence of metal ions (27
, 28)
.
We tested inhibitors of cellular signal transduction and several enzymatic processes for their potential to inhibit cell contraction. According to the data presented, one can exclude nicotinic acetylcholine receptor signaling, deacetylase signaling, and G-protein activation-dependent processes as signaling pathways for endothelial cell contraction. The poly-ADP-ribose polymerase inhibitor 3-aminobenzamide, which has been shown to inhibit H2O2-mediated cell contraction caused by energy depletion (A. Csordas; unpublished observation), as well as aspirin, had no effect. Protein kinase C inhibition exerted a marginal protective effect.
Taxol (paclitaxel) a microtubule stabilizing agent, although toxic for HUVECs at 100 nM and at later time points (>2 h), protected HUVECs from contraction (see lower right image of Fig. 2
).
Taken together, the most powerful agents to inhibit CSE-mediated vascular endothelial cell contraction are thiol group-containing agents, followed by cation inactivators, then microtubule stabilizing agents, followed by antioxidants.
Microtubule depolymerization is a crucial event in CSE-mediated cell contraction
After having defined the compounds within cigarette smoke that mediate vascular endothelial cell contraction, and vectored by the finding that microtubule stabilizers inhibit the phenomenon, we focused on the role of microtubule stability in this process. The data presented in Fig. 3
A demonstrate that not only the cell structure (Fig. 2)
, but also the microtubules of cells pretreated with taxol (10 nM), stay intact after CSE treatment.
|
The Western blot analyses shown in Fig. 3B
demonstrate that -tubulin in CSE-treated cells is degraded, matching the kinetics of endothelial cell contraction, and that CSE-mediated tubulin degradation is inhibited by acetyl-cysteine (1 mM). Epoxomicine (50 µM), a specific inhibitor of the proteasome, potently inhibited tubulin degradation, suggesting that tubulin is targeted to the proteasome and degraded in response to CSE treatment. Experiments where we tried to inhibit cellular contraction by applying epoxomicine to the cells clearly showed that epoxomicine is not capable of inhibiting cell contraction (Table 1)
. In contrast, taxol, which is capable of inhibiting cell contraction, does not interfere with tubulin-degradation.
Figure 3C
demonstrates that stabilization of microtubules by taxol rescues cytoskeletal structure (f-actin) as well as intermediate filament (vimentin) breakdown, clearly placing CSE-mediated microtubule damage upstream of the CSE-effects on the cytoskeleton and the intermediate filaments.
CSE treatment of cells reduces cellular PECAM-1 (CD31) and P-selectin (CD62P) protein levels
To elucidate the effects of CSE on other cellular proteins, we performed a Western blot-based screen for protein modifications/regulations (data not shown).
No effect of CSE treatment could be observed on: integrins 2, 3, 5, v, ß1, or ß3; proteins L1, connexin 26, heat shock protein 60 and 70, or ß actin. Only three proteins showed differences when untreated cells were compared with CSE-treated cells. As expected, we could detect differences not only in tubulin protein levels (see Fig. 3B
) but also in protein levels of P-selectin/CD62P (Fig. 4
) and the 100 kDa form (i.e., the smaller form) of PECAM-1/CD31 (Fig. 5
).
|
|
Figure 4A
shows that CD62P is down-regulated matching the kinetics of endothelial cell contraction/detachment and that acetyl-cysteine (1 mM) inhibits the down-regulation. To determine the fate of CD62P, we performed ELISAs of cell culture supernatants (Fig. 4B
) and detected increased concentrations of soluble CD62P in response to CSE treatment of cells.
Similarly, we analyzed the time, dose, and acetyl cysteine (1 mM) dependencies of CD31 regulation. Even though the 100 kDa form of CD31 showed good correlation to the kinetics of cell contraction/detachment (Fig. 5A
), no increase in supernatant contained soluble CD31 was observed (ELISA depicted in Fig. 5B
).
Therefore, we set out to determine the cellular location of the two forms of CD31. Figure 5C
shows that the 130 kDa form (i.e., the form that is not affected by CSE) of CD31 is present on the cellular surface whereas the 100 kDa (i.e., regulated) form of CD31 is an intracellular protein (see Materials and Methods section). We continued to determine the fate of the intracellular 100 kDa form of CD31. Addition of taxol (10 nM) or epoxomicine (50 µM) to the cells did not block the down-regulation of intracellular CD31, clearly separating the effects of CSE on the microtubule system from this phenomenon. However, when we applied acetyl-leucin-leucin-norleucinal (LLNL; 250 µM) or MDL28170(15 µM) as inhibitors of calpains to the cells prior to CSE treatment, the reduction of intracellular CD31 concentration was blocked. Since calpains are known to be regulated by Ca2+, we tested whether Ca2+-deprived HUVECs down-regulate the 100 kDa form of CD31 in response to CSE treatment. When we performed the experiments under Ca2+-free conditions (in PBS and in the presence of the intracellular Ca2+ chelator BAPTA), CSE-mediated reduction of the 100 kDa CD31 form was blocked. Ca2+-free conditions led to a massive CSE-independent contraction and detachment of endothelial cells after 1 h. Therefore, incubation periods were reduced (incubation with BAPTA for 5 min; CSE (16.6%) treatment for 30 min).
Low CSE concentrations induce IL-6 and IL-8 secretion of HUVECs whereas higher CSE concentrations inhibit cytokine as well as chemokine secretion
Smoking is known to induce a systemic inflammatory response. To analyze whether endothelial cells exposed to cigarette smoke constituents contribute to this phenomenon, we measured cytokine and chemokine concentration in supernatants of HUVECs treated with CSE concentrations ranging from 2% to 32% after various time points. The data depicted in Fig. 6
demonstrate that neither soluble VCAM nor CD40L can be found in HUVEC culture supernatants in the presence or absence of CSE. Human tPA (tissue plasminogen activator) showed a tendency to be up-regulated in untreated cells and at low CSE concentrations, but these findings were not significant. MCP-1, IL-6, and IL-8 were up-regulated under normal cell culture conditions (0%). However, 2% and 4% CSE enhanced IL-6 and IL-8 secretion by HUVECs up to 3-fold compared with the control. Surprisingly, the secretory activity of HUVECs for MCP-1, IL-6, and IL-8 was completely inhibited at concentrations greater than 4% CSE.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. In the present project, we reveal the chemical and biological basis for vascular endothelial cell contraction and denudation in response to the addition of cigarette smoke constituents. We focused on this aspect because it is well known from animal models and from human surgery that vascular endothelial cell denudation by mechanic, chemical, or biological injury causes intima hyperplasia, thrombus formation, and atherosclerosis (29
30
31)
. Basically, our data suggest a crucial involvement of metals, which are constituents of cigarette smoke (see Results), in the observed phenomena since metal chelating agents and cation exchangers are inhibitory. The method used to detect protein oxidation is highly sensitive for metal-catalyzed protein oxidation.
Although antioxidants (vitamins C and E, resveratrol, and atorvastatin) showed a minor protective effect, we hypothesize that oxidants/free radicals also play an important role in this process for the following reasons. Oxidative modifications of cellular proteins as well as intracellular oxidative stress could be detected (Fig. 1)
. Preliminary microarray analyses demonstrate that CSE-treated cells switch on a transcriptional response between 3 and 7 h as induced by oxidative damage (B. Henderson; unpublished observation; up-regulation of heme-oxygenase, superoxide dismutase, chaperones, etc.). Thiol groups (which bind metal ions and have antioxidant properties) comprise a higher protective potential than cation exchangers or metal chelating agents.
Due to the short duration of CSE treatment necessary until the first effects become manifest, we exclude complex signal transduction pathways (via transcription and/or translation) from being involved in this process. The fundamental participation of signaling pathways involving activation of extracellular proteases, deacetylases, G-protein signaling, protein kinase C-like signaling, nicotinic acetylcholine receptor signaling, PARP-dependent processes, cyclooxygenase (COX) I or II-dependent signaling pathways (inhibited by aspirin, and resveratrol), or classical oxidant-dependent processes/pathways (32
33
34)
seems to be unlikely based on the data presented in Table 1
.
Supported by the detection of rapid intracellular oxidative stress (data from Fig. 1
), we hypothesize that the "contractive agents" cause direct chemical damage to cellular structures and/or induce the activation of intracellular proteases (proteasome and calpain) that finally lead to cell contraction and detachment.
Microtubule stability plays a crucial role in CSE-mediated cell contraction. Stabilization of microtubules not only inhibited microtubule collapse (Fig. 3A
), but also cytoskeletal and intermediate filament breakdown (Fig. 3C
), entailing preservation of cellular shape (Fig. 2)
. The finding that taxol inhibited the cellular collapse but could not block tubulin degradation and that epoxomicine exerted the exact opposite effect may be explained by the following hypothesis: microtubules (i.e., tubulin) are modified by CSE constituents; this modification causes the loss of tubulin function (resulting in the breakdown of polymerized microtubules) and "labels" free tubulin (50% of the cellular tubulin exists in the free monomeric form) as ready for proteasome-dependent degradation. Taxol only inhibits the depolymerization whereas epoxomicine inhibits the degradation of monomeric tubulin via the proteasome pathway.
Concerning the mechanism for detachment of endothelial cells in response to CSE treatment (see supplementary video file), we would like to suggest a model similar to that for cell contraction involving (surface) protein modification and loss of protein function. However, the data shown in Fig. 3
clearly demonstrate that cell detachment does not rely on depolymerization of the microtubule system. Although CSE-mediated contraction was inhibited by taxol, cell detachment was not. This finding separates both phenomena and highlights their mutual independence. As a mechanism for cell detachment, we propose a model where adhesion molecules are inactivated by protein oxidation, resulting in protein destruction or release of adhesion proteins as observed for CD62P. For an exact definition of the molecules and mechanisms involved, a proteomics approach is required.
The generation of soluble CD62P seems to depend on two processes. Stone et al. (35)
reported that in response to CSE exposure of HUVECs, CD62P is transported to the cell surface. In a second, not clearly understood, step CD62P is released into the supernatant. Potential mechanisms for this process include shedding of the proteins from the membrane and the production of secretion-specific isoforms of CD62P (36)
.
Despite the effects of CSE on CD62P, we were able to observe that the 100 kDa form of CD31/PECAM-1 was down-regulated in response to CSE treatment in a calpain- (inhibited by LLnL and MDL28170 and Ca2+-dependent fashion (see Fig. 5
). CD31 processing seems to be closely linked to Ca2+ signaling. Recently, Wong et al. (37)
have shown that processing of CD31 from its 130 kDa form to its cleaved (100 kDa) form depends on calmodulin-dissociation from the protein. The protective calmodulin-CD31 bond is stabilized by Ca2+ ions. In contrast to this mechanism, reduction/degradation of the 100 kDa form of CD31 described herein seems to depend on increased intracellular Ca2+ levels, since Ca2+-free culture conditions inhibited 100 kDa CD31 processing.
Cytokines, chemokines, and other soluble factors are important players in atherogenesis. In our analyses of the effects of CSE on HUVEC cytokine and chemokine secretory activity (Fig. 6)
, we observed that IL-6 and IL-8 are up-regulated at low CSE concentrations. Up-regulation of endothelial IL-8 secretion has been shown to play a role in monocyte recruitment to the vessel wall (38
, 39)
, and IL-6 secretion might reflect endothelial proinflammatory signaling (39)
. Both events are known to contribute to atherogenesis and may highlight a mechanism by which smoking induces vessel wall infiltration and inflammation. Higher CSE concentrations led to a stop in IL-6, IL-8, and MCP-1 secretion. Since IL-8 and MCP-1 are known mediators of matrix metalloproteinase activation, endothelial cell tube formation, and angiogenesis (40)
, this finding might also help explain the adverse effects of smoking. Although the cellular response to cigarette smoke constituents is not well understood, there is evidence that modulation of chemokine/cytokine production is mediated via protein kinase signaling (41)
. To explain the chemokine/cytokine-secretion inhibiting activity of endothelial cells after CSE treatment, there is a very good correlation with the observed amount of intracellular oxidative stress caused by CSE (Fig. 1)
. Cellular damage caused by oxidative stress might activate the cells’ repair machinery, and this might lead to a shutoff of energy consuming protein synthesis.
Finally, we would like to sum up some aspects that support the in vivo relevance of the following observations. 1) Oxidized LDL can be found in smoker’s blood (12)
, an indicator for the existence of oxidative stress in vivo. 2) Metals like Cd and Pb are increased in smoker’s blood (19)
. 3) Increased levels of soluble CD62P and IL-6 have been reported in numerous of studies (4
, 42
43
44
45)
, whereas only very few publications have reported an association of smoking status with soluble CD31 levels in smoker’s blood (although a large number of projects investigated a potential association). The role of soluble CD62P in atherosclerosis is not yet clearly understood (reviewed in ref 46
), but increased sCD62P levels are correlated with smoking and the risk for cardiovascular diseases (47)
. 4) As an indicator for endothelial damaging activity of cigarette smoke, increased permeation of albumin across an endothelial layer was reported in a pulmonary endothelium in vitro system (48)
.
In summary (Fig. 7
), we would like to suggest the following model by which CSE causes endothelial stress and damage. 1) Metals contained in cigarette smoke catalyze the oxidation of vascular endothelial cell structures (microtubule system and surface proteins) by cigarette smoke oxidants/free radicals. 2)This modification leads to functional impairment, finally resulting in the collapse of the structural equipment of the cell (i.e., microtubules, cytoskeleton, and intermediate filaments) and consequently in the collapse of the cellular shape. 3) This collapse and CSE-mediated detachment of endothelial cells (by functional impairment of the adhesive system of the cells) leads to denudation of the blood vessels. We certainly do not think that cigarette smoke leads to complete denudation of the vessel wall in vivo, but even the slightest shift from a closed endothelium to a more "open" endothelium by spatial opening of gaps between endothelial cells (via cell contraction) or by a small increase in the number of endothelial cell detachments would effectively disturb the functionality of the endothelium. Disturbed function/dysfunction of the vascular endothelium is currently thought to represent the most important step toward atherogenesis, and our data highlight a novel potential mechanism by which smoking induces/accelerates atherosclerosis.
|
|
ACKNOWLEDGMENTS |
|---|
Received for publication September 29, 2004. Accepted for publication January 27, 2005.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  D. Aradi, A. Konyi, L. Palinkas, T. Berki, T. Pinter, T. Tahin, I. Horvath, L. Papp, and A. Komocsi Thienopyridine Therapy Influences Late Outcome After Coronary Stent Implantation Angiology, May 1, 2008; 59(2): 172 - 178. [Abstract] [PDF]
|
|
|
|
 |
|
 S. S. Barbieri, L. Ruggiero, E. Tremoli, and B. B. Weksler Suppressing PTEN Activity by Tobacco Smoke Plus Interleukin-1{beta} Modulates Dissociation of VE-Cadherin/{beta}-Catenin Complexes in Endothelium Arterioscler. Thromb. Vasc. Biol., April 1, 2008; 28(4): 732 - 738. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Scalia, Y. Gong, B. Berzins, L. J. Zhao, and K. Sharma Hyperglycemia Is a Major Determinant of Albumin Permeability in Diabetic Microcirculation: The Role of {micro}-Calpain Diabetes, July 1, 2007; 56(7): 1842 - 1849. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. S. Barbieri and B. B. Weksler Tobacco smoke cooperates with interleukin-1{beta} to alter {beta}-catenin trafficking in vascular endothelium resulting in increased permeability and induction of cyclooxygenase-2 expression in vitro and in vivo FASEB J, June 1, 2007; 21(8): 1831 - 1843. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. M. Cantin, J. W. Hanrahan, G. Bilodeau, L. Ellis, A. Dupuis, J. Liao, J. Zielenski, and P. Durie Cystic Fibrosis Transmembrane Conductance Regulator Function Is Suppressed in Cigarette Smokers Am. J. Respir. Crit. Care Med., May 15, 2006; 173(10): 1139 - 1144. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Bernhard, A. Rossmann, B. Henderson, M. Kind, A. Seubert, and G. Wick Increased Serum Cadmium and Strontium Levels in Young Smokers: Effects on Arterial Endothelial Cell Gene Transcription Arterioscler. Thromb. Vasc. Biol., April 1, 2006; 26(4): 833 - 838. [Abstract] [Full Text] [PDF]
|
