Estrogen-Receptor–Mediated Inhibition of Human Endothelial Cell Apoptosis
Estradiol as a Survival Factor
Background A series of studies was performed to examine the ability of estradiol (E2) to protect endothelial cells from apoptosis.
Methods and Results Light and transmission electron microscopy demonstrated typical features of apoptosis in human umbilical vein endothelial cells (HUVEC) exposed to tumor necrosis factor-α (TNF-α). Northern and Western blot analyses revealed induction of message and protein for the interleukin-1β converting enzyme (ICE), which has been shown to mediate apoptosis induced by TNF-α. Immunofluorescent staining of HUVEC colocalized ICE expression to apoptotic HUVEC. Direct cell counting demonstrated a significant decrease in total endothelial cell number after 24 hours of TNF-α exposure and a dose-dependent reversal of the effect of TNF-α with E2 treatment. This protective effect was abrogated by an estrogen-receptor antagonist. Fluorescence-activated cell sorting analysis revealed 39.3% apoptosis after 24 hours of TNF-α exposure. Treatment with E2 resulted in a 50% decrease in apoptosis. Similarly, viability assays revealed 35±4% cell death after TNF-α exposure. Simultaneous treatment with E2 resulted in a dose-dependent reduction of cell death to a minimum of 18±2%. The protective effect of E2 was nullified by a specific estrogen-receptor antagonist.
Conclusions E2 treatment resulted in a dose-dependent, receptor-mediated inhibition of TNF-α–induced endothelial cell apoptosis. These studies indicate that E2 may also serve a maintenance function in preventing endothelial cell death after noxious stimuli and suggest that the ICE pathway may mediate cytokine-induced apoptosis in endothelial cells. Preservation of endothelial integrity represents another mechanism that may account for the atheroprotective effect of estrogen.
The inhibitory influence of estrogen on the development of atherosclerosis has been suggested by an abundance of human epidemiological and animal experimental data.1 2 3 4 5 6 7 8 9 Although part of the antiatherogenic action of estrogen appears to be due to a salutary effect on serum lipids,10 11 the protective mechanism of estrogen cannot be entirely explained on this basis.3 12
Because endothelial “injury” is a postulated mechanism for initiation of atherogenesis,13 protection of endothelial cells from toxic insults could represent an alternative mechanism for the inhibition of atherogenesis by E2. Endothelial cells have been shown to express a functional estrogen receptor.14 15 Furthermore, numerous studies have now documented a direct influence of E2 on endothelial cell biology in vivo and in vitro.16 17 18
To investigate the potential role of E2 as a survival factor for endothelial cells, we evaluated the ability of E2 to inhibit programmed cell death of human endothelial cells induced by the cytokine TNF-α. There were several reasons for choosing this in vitro model of endothelial cell injury. First, recent data suggest an important functional role for TNF-α in human atherosclerosis and restenosis19 and in experimental atherosclerosis.20 21 22 Second, the induction of apoptosis in endothelial cells by TNF-α has been well characterized in nonhuman species,23 thus providing an established model for examining the kinetics of toxin-induced endothelial cell death. Finally, preliminary studies in our laboratory (Fig 1⇓) indicated that E2 was capable of protecting endothelial cells from apoptosis induced by TNF-α.
However, controversy has existed regarding the induction of apoptosis by TNF-α exposure in human endothelial cells.24 25 The objective of the present study, therefore, was twofold. The first goal was to verify the induction of TNF-α–mediated apoptosis in human endothelial cells and to characterize the molecular mechanisms responsible for this effect. After establishing the validity of this model, we then proceeded to precisely define the inhibition of endothelial cell apoptosis by E2.
HUVEC were isolated according to the method of Jaffe et al,26 plated onto 1.5% gelatin-coated wells, and grown in phenol red free medium 199 (M199) (Gibco) with 20% fetal bovine serum that had been dextran-charcoal stripped to remove steroid hormones (Gemini Bio-Products), endothelial cell–growth supplement (100 μg/mL), and heparin (50 U/mL) (complete medium). HUVEC were used in passages 2 and 3 to avoid “age-dependent” variations in levels of apoptosis.27
HUVEC were subcultured onto tissue culture plates at ≈50% confluence or onto 4-well microscope slides at a density of 1×105 cells per well and allowed to attach overnight. Medium was changed to fresh medium the following day, and the cells were examined under phase-contrast microscopy. When the cells were at 70% confluence, the cultures were divided into three groups: control groups received fresh standard culture medium; groups treated with TNF-α (Genzyme) were exposed to fresh standard culture medium with the addition of 40 ng/mL TNF-α; the TNF-α plus E2 group was treated with fresh standard culture medium with the addition of TNF-α 40 ng/mL plus E2 in a range of concentrations from 10−13 to 10−7 mol/L. The concentration of TNF-α used to induce apoptosis in the present study was chosen on the basis of previously published literature.28 Although circulating levels of TNF-α do not reach the levels used in this in vitro model system, the local concentration of TNF-α, and specifically the effective concentration at the cell-cell interface between endothelial cells and TNF-α–secreting smooth muscle or inflammatory cells, is not measurable and could presumably achieve levels comparable to those used in our system.
Electron microscopy was performed to confirm that the ultrastructural features of apoptosis were present in cells exposed to TNF-α in the present study. Endothelial cells exposed to the conditions outlined above were fixed in 2.5% glutaraldehyde (pH 7.3) buffered with 0.1 mol/L sodium cacodylate overnight at 4°C and then washed with 0.1 mol/L sodium cacodylate buffer for 15 minutes before postfixation with 1% osmium tetroxide buffered with 0.1 mol/L sodium cacodylate for 1 hour on ice. After another wash with 0.1 mol/L sodium cacodylate buffered for 15 minutes, cells were dehydrated with increasing concentrations of alcohol (30%, 50%, 70%, 80%, 90%, and 100%; three times at each concentration) for 10 minutes each. Next, cells were infiltrated with propylene oxide for 15 minutes, followed by 1:1 propylene oxide:epoxy resin for 1 hour, 1:2 propylene oxide:epoxy resin for 2 hours, and finally 100% epoxy resin for 2 hours. Cells were embedded with fresh epoxy resin into molds and placed in a 60°C oven for 2 hours. Ultrathin sections were stained with uranyl acetate and lead citrate and were examined with the use of a Philips 300 electron microscope.
Northern Blot Analysis
Total RNA was isolated from HUVEC by β-mercaptoethanol (Fisher) and guanidinium isothiocyanate denaturation followed by centrifugation in silica-based spin columns (RNeasy kit, QIAGEN, Inc). Twenty micrograms of RNA per sample was resolved on 1.2% (wt/vol) denaturing agarose gels (containing 2.2 mol/L formaldehyde and 0.5 μg/mL ethidium bromide) and then transferred to a nylon membrane (Hybond-N, Amersham) by capillary blotting. Equal loading was documented by an ultraviolet transillumination photograph of the 28s band. The cDNA fragments were labeled with [32P]dCTP (New England Nuclear) by use of a random priming labeling kit (Boehringer Mannheim) to a specific activity of 5 to 9×108 cpm/μg. Purification from unincorporated nucleotides was achieved with the Bio-Spin 6 minicolumn system (BioRad). After 40 minutes’ prehybridization at 65°C, heat-denatured probes were annealed during a 3-hour period at the same temperature with the use of Rapid-hyb buffer (Amersham) and washed in 2×SSC containing 0.1% (wt/vol) SDS at room temperature for 10 minutes, followed by 1×SSC/0.1% (wt/vol) SDS at 65°C for 5 minutes. The membranes were exposed to film (HyperFilm, Amersham) with intensifying screens at −70°C for 12 hours.
Expression of ICE was determined with the use of a 1.4-kb DNA fragment containing the full-length human ICE cDNA (generously supplied by Michael B. Widmer, PhD, Immunex Corp, Seattle, Wash).
The probe for c-myc expression consisted of a 1.7-kb fragment of the rat c-myc cDNA, corresponding to nucleotides 4735 to 6438 of the rat sequence, with 85% homology to the human sequence.
The probe for p53 expression consisted of a 990-bp fragment of murine p53 with 84% homology to the human cDNA, corresponding to nucleotides 263 to 1256 of the human sequence.
Western Blot Analysis
Cells were washed three times in cold PBS and then lysed for 30 minutes at 4°C in buffer containing 0.1% Nonidet P-40, 0.5% (wt/vol) sodium deoxycholic acid, 0.1% (wt/vol) SDS PBS, pH 7.4, and the following protease inhibitors: 0.5 mmol/L PMSF, 1 μg/mL aprotinin, 1 μg/mL leupeptin, and 0.5 μg/mL pepstatin A. After centrifugation at high speed, the supernatant was collected and the protein content of all samples was determined by use of the Bradford assay (Bio-Rad) with γ-globulin as the standard. Electrophoresis was performed on 12% SDS-polyacrylamide gels, loading 100 μg protein per lane. After transfer to a 0.2-μm PVDF membrane (Bio-Rad), membranes were blocked in 10% (wt/vol) nonfat dry milk (in PBS, pH 7.5) and incubated for 2.5 hours with a rabbit polyclonal (ICE p20 1:200; estrogen receptor 1:200) antibody (Santa Cruz Biotechnology). Detection was performed by use of a secondary horseradish peroxidase–linked anti-rabbit (1:7500 in PBS) antibody (Amersham) and the enhanced chemiluminescence system (Amersham). All steps of the immunoblot were performed at room temperature, and each antibody incubation period was followed by 1 hour of washing the membrane in 0.1% Tween-20 in PBS.
HUVEC were plated onto glass four-well slides coated with fibronectin and were allowed to attach overnight. Recombinant human TNF-α (40 ng/mL) was then added to the cells for 20 hours. Cells were washed once with PBS before fixing with 4% formalin for 30 minutes at room temperature. The cells were rinsed once with PBS and then permeabilized with 0.5% NP-40, followed by a second wash with PBS, each lasting 5 minutes. Cells were then incubated at room temperature for 60 minutes in blocking solution comprising 0.5% NP-40, 5% normal goat serum, and PBS in a ratio of 1:2:2. The primary antibody, ICE p10 (Santa Cruz Biotechnology), was added to the cells at a dilution of 1:50 (in blocking solution) and incubated overnight at 4°C. The next day, the cells were washed once with blocking solution, and the secondary antibody, a rhodamine-labeled anti-rabbit IgG (1:2000 dilution in blocking solution, Kirke- gaard and Perry Laboratories Inc), was added to the cells for 60 minutes in the dark. Cells were rinsed twice with PBS and the In Situ Cell Death Detection Kit, AP (Boehringer Mannheim) was used to label apoptotic cells. Per the kit specifications, the TUNEL reaction mixture was prepared (450 μL label solution, 50 μL enzyme solution) and 50 μL of solution was added to each well of the slide (after slide chambers had been removed). Slides were covered with parafilm coverslips, with the mixture spread evenly across the wells, and incubated in a humidity chamber at 37°C for 60 minutes away from light. Cells were then rinsed once with PBS and once with 0.01% Triton X-100 in PBS, followed by the application of Hoechst stain (Hoechst 33342, 1:1000 dilution in distilled water of 0.2 mg/mL stock; Boehringer Mannheim) for 2 minutes at room temperature to counterstain nuclei. Cells were then rinsed twice with PBS and once with distilled water, and slides were coverslipped with the use of a mounting medium for fluorescence (Kirkegaard and Perry Laboratories, Inc).
Quantification of apoptotic and viable cells was accomplished by a multiparameter assay measuring forward light scatter and fluorescence of propidium iodide. HUVEC were stimulated for 24 hours with TNF-α (40 ng/mL) alone or in combination with E2 (10−13 to 10−7 mol/L) and prepared for analysis. Floating cells were collected and adhesive cells were trypsinized (0.05% [wt/vol] trypsin in 0.02% [wt/vol] EDTA), incubated for 5 minutes at 37°C, and harvested. After washing twice in PBS and after slow centrifugation (350g), the pellet was resuspended in ice-cold ethanol (70%) and fixed overnight at 4°C. The cell pellet was stained in PBS, pH 7.4, with addition of 0.1% (wt/vol) Triton X-100 (Sigma Chemical Co), 0.5 mmol/L EDTA, pH 7.4, 0.05 mg/mL RNase A (50 U/L, Sigma) and 50 μg/mL of the intercalative DNA-binding dye propidium iodine (Boehringer Mannheim) at 4°C for 4 hours.29 A final concentration of 1×106 cells/mL staining solution was achieved. DNA content was analyzed from 104 cells (events) per group within the fluorescence gate (excitation with the 488-nm line of an argon laser, and detection at 620 to 700 nm) by use of a Becton Dickinson flow cytometer in combination with the CytoFlow program version 2.2. Apoptotic cells were defined as hypodiploid, having a DNA content of ≥5% that of the diploid cells; materials with less fluorescence were regarded as debris or artifacts.
In addition to phase-contrast microscopy, HUVEC were also examined by fluorescence microscopy after staining with acridine orange as previously described.28 Briefly, cells were fixed in 70% ethanol for 10 minutes, rinsed in PBS, and incubated in the dark with 3 mg/mL acridine orange for 1 hour. Cellular morphology was assessed by fluorescence microscopy. Apoptosis was identified by the findings of condensation and fragmentation of chromatin and blebbing of the cytoplasm (Fig 6A⇓).30 A total of 10 random microscope fields were examined under each experimental condition, with the total number of cells in each field counted.
HUVEC were plated onto 24-well culture plates at a density of 2×104 cells per well. Cells were allowed to attach overnight before being exposed to the conditions outlined. In addition to control conditions, TNF-α alone, and TNF-α in combination with E2, a subset of these cells was also exposed to TNF-α in combination with E2 (10−9 mol/L) and an equimolar concentration of the specific estrogen-receptor antagonist ICI 182780 (generously supplied by Dr A.E. Wakeling, Zeneca Pharmaceuticals, Cheshire, England). These studies were performed to evaluate whether the protection from apoptosis conferred by E2 was mediated by the estrogen receptor.
The CellTiter 96 AQ nonradioactive cell-proliferation assay (Promega) was used to assess cell viability. The assay is composed of the tetrazolium compound MTS and an electron coupling reagent, PMS. MTS is reduced by viable cells to formazan, which can be measured with a spectrophotometer by the amount of 490-nm absorbance. Formazan production is time dependent and proportional to the number of viable cells.
Endothelial cells were cultured in 0.1 mL HUVEC media in 96-well flat-bottomed, fibronectin-coated culture plates (Becton Dickinson). Cultures were seeded at 1×104 cells/well and allowed to attach overnight. After the indicated time of incubation with the appropriate medium, 20 μL MTS/PMS (1:0.05) mixture was added per well, and cells were incubated 1 hour before measuring absorbance at 490 nm. Background absorbance from the control wells (same media, no cells) was subtracted. Eight duplicate studies were performed for each experimental condition.
Cell counts for endothelial cells under various conditions represent duplicate studies from independently performed experiments. Data are presented as mean±SD. ANOVA was used to evaluate the statistical significance of differences between experimental groups with the Newman-Keuls method applied to analyze differences between individual means. Statistical significance was assigned at the level of P<.05.
Phase-Contrast Microscopy Shows Reduced Evidence of TNF-α–Induced Apoptosis in HUVEC Treated With E2
Adherent HUVEC demonstrated typical cobblestone morphology under control conditions (Fig 1A⇑). After exposure to TNF-α, the cells became rounded and partially detached and demonstrated the lobulated appearance of apoptotic cells (Fig 1B⇑). Moreover, the density of the adherent cells was decreased compared with control conditions, indicating detachment of HUVEC exposed to TNF-α. When E2 was present simultaneously with TNF-α, fewer endothelial cells demonstrated the morphological features of apoptosis. The percentage of cells remaining attached was also increased, which was reflected by the increased cell density (Fig 1C⇑). These findings were further corroborated by direct cell counting (see below). These observations suggest that TNF-α induces apoptosis in HUVEC and that E2 inhibits the development of programmed cell death in these cells.
Electron Microscopy Documents Characteristic Ultrastructural Features of Apoptosis in Human Endothelial Cells Exposed to TNF-α
Electron microscopy was performed in the present study to specifically document that the light microscopic features of apoptosis seen in TNF-α–treated human endothelial cells were accompanied by the appropriate ultrastructural morphology. Under control conditions, normal cellular anatomy was identified on examination of endothelial cells with transmission electron microscopy (Fig 2A⇓). In contrast, cells exposed to TNF-α demonstrated the characteristic morphological features of apoptosis. Condensation of chromatin was noted at the periphery of the nucleus, and blebbing and fragmentation of the cytoplasm was seen (Fig 2B⇓).
Thus, the characteristic light microscopic and ultrastructural features of apoptosis were induced in human endothelial cells exposed to high concentrations of TNF-α. To further define the advent of apoptosis in HUVEC, we analyzed the expression of molecular markers for TNF-α–induced and other forms of apoptosis.
Northern Blot Analysis
TNF-α Induces Expression of mRNA for ICE in Human Endothelial Cells
The ICE gene, the mammalian homologue of the C elegans ced-3 cell death gene,14 has been shown to mediate apoptosis induced by TNF-α.31 ICE is also known to induce apoptosis in rat fibroblasts32 and is expressed in human atheroma.33 To provide evidence that HUVEC exposure to TNF-α results in apoptosis by an established molecular mechanism, Northern blot analysis was performed to demonstrate induction of ICE expression.
Northern blot analysis revealed an increase in the expression of the mRNA for ICE after exposure of human endothelial cells to TNF-α (Fig 3A⇓). All three previously described transcripts for ICE34 are strongly induced in HUVEC exposed to TNF-α. This is consistent with TNF-α–induced apoptosis in other cell types in which cell death is mediated by ICE.
TNF-α–Induced Apoptosis in Human Endothelial Cells Is Not Associated With Changes in Expression of the c-myc Proto-oncogene
The c-myc proto-oncogene has been shown to play a functional role in apoptosis of smooth muscle cells35 and fibroblasts.36 Furthermore, c-myc has been demonstrated to function in the induction of apoptosis by the estrogen-receptor antagonist tamoxifen.37 Accordingly, we investigated the regulation of c-myc expression in HUVEC during induction of apoptosis by TNF-α. The expression of c-myc was not changed after 24 hours of TNF-α exposure (Fig 3B⇑). Thus, although it is possible that the kinetics of c-myc regulation prevented detection of altered expression at the 24-hour time point studied, it appears unlikely to play a functional role in the induction of apoptosis by TNF-α within this time frame.
p53 Expression Is Not Modulated by TNF-α in HUVEC
The tumor-suppressor gene p53 has been shown to be involved in growth arrest and apoptosis under the control of various growth factors.38 39 Moreover, p53-dependent apoptosis induced by synthetic retinoids has been demonstrated in an estrogen-receptor–positive breast cancer cell line.24
However, p53 expression is not altered in HUVEC by TNF-α exposure (Fig 3B⇑). Cytokine-induced apoptosis in human endothelial cells, therefore, does not appear to be associated with regulation of p53 expression. This conclusion is tempered by consideration that the time course of p53 regulation may have prevented detection of its regulation in the present study.
Western Blot Analysis Confirms ICE Expression by HUVEC Exposed to TNF-α and Documents Stable Expression of the Estrogen Receptor Under All Conditions
Western blot analysis revealed an increase in the expression of ICE protein after exposure of human endothelial cells to TNF-α. This is consistent with the regulation of ICE expression also seen at the transcriptional level after TNF-α exposure (Fig 4⇓).
To establish the stable expression of the estrogen receptor by cultured endothelial cells, Western blot analysis was also performed with the use of an antibody to the carboxy terminus of the human receptor. As shown in Fig 4⇑, estrogen-receptor expression was maintained under all conditions used in the present study. These findings are consistent with work by others15 and are a prerequisite for the studies detailed below.
Immunofluorescence Microscopy Localizes ICE Expression to Apoptotic HUVEC
To provide further evidence of a functional role for ICE in human endothelial cell apoptosis, ICE protein expression was examined by immunofluorescence microscopy. Cells were simultaneously examined for biochemical features of apoptosis by use of the TUNEL method to fluorescently label DNA strand breaks. Finally, fluorescent counterstaining (Hoechst) of HUVEC nuclei was also performed to permit the examination of cells for morphological features of apoptosis.
As shown in Fig 5⇓, expression of ICE protein, identified by red fluorescence, localized to cells that also exhibited the morphological and biochemical features of apoptosis. The nucleus of the cell that stained positively for ICE expression was shrunken and deformed. In addition, this cell showed evidence of DNA strand breaks manifested as green fluorescence, indicative of terminal deoxy transferase labeling of 3′ DNA ends. The absence of ICE expression in the morphologically normal–appearing cells, which also showed no biochemical evidence of apoptosis, suggests that ICE protein expression was limited to cells undergoing apoptosis.
Having established that the typical light microscopic, ultrastructural, and molecular features of TNF-α–induced apoptosis were present in HUVEC after exposure to TNF-α, we next attempted to more precisely define the protective effect of E2.
Fluorescence Microscopy Demonstrates Inhibition of TNF-α–Induced HUVEC Apoptosis
Because condensation of nuclear chromatin occurs in apoptotic cells, fluorescent staining of chromatin is a convenient way to visualize the degree of apoptosis under varying conditions. Human endothelial cells induced to undergo apoptosis by TNF-α demonstrated typical features of chromatin condensation and fragmentation and cytoplasmic blebbing when viewed at high power (Fig 6A⇓). The intensified fluorescence exhibited by these cells permitted their identification at low power when examined by fluorescence microscopy.
Under control conditions, a background level of fluorescence was noted after staining with acridine orange. Few of the cells demonstrated the bright fluorescence resulting from staining of condensed chromatin characteristic of apoptosis (Fig 6B⇑). After exposure to TNF-α, a significant percentage of endothelial cells demonstrated bright fluorescence consistent with apoptosis (Fig 6C⇑). When endothelial cells were exposed simultaneously to E2 and TNF-α, the percentage of cells exhibiting intense fluorescence was decreased compared with the cells exposed to TNF-α alone, consistent with inhibition of apoptosis in these cells (Fig 6D⇑).
The subjective impression of increasing apoptosis of HUVEC exposed to TNF-α and inhibition of apoptosis after E2 treatment was then quantified by direct cell counting.
Cell Counting Demonstrates Reversal of TNF-α–Induced Reduction of Endothelial Cell Population by E2 Treatment
We performed cell counting using two techniques. Endothelial cells exposed to conditions on microscope slides were evaluated by counting 10 random microscope fields. HUVEC exposed to similar conditions in 24-well culture plates were counted with the use of a hemocytometer.
On control slides, the mean cell count per field was 150.6±23.6 (range, 114 to 180). On the TNF-α–treated slides, mean total cell count per field was 82.2±12.6 (P<.0001 versus control) (range, 69 to 105). When E2 was applied simultaneously with TNF-α, the total cells counted per field were 114.3±15.5 (P<.001 versus TNF-α alone) (range, 75 to 132) (Fig 7A⇓). The decrease in the number of attached cells after treatment with TNF-α is consistent with increased degrees of apoptosis as demonstrated by phase-contrast microscopy (see above) and with the findings of FACS analysis (see below), which indicate that detached/floating cells are apoptotic. The increase in the number of attached (counted) cells that occurs when endothelial cells are incubated simultaneously with E2 at the time of exposure to TNF-α is therefore consistent with inhibition of apoptosis by E2.
Cell counting of HUVEC exposed to study conditions in 24-well culture plates revealed similar findings. As shown in Fig 7B⇑, total cell count decreased from 10.1±2.8×103 cells per well under control conditions to 3.5±1.3×103 cells per well after 24 hours of TNF-α exposure (P<.0001). Simultaneous treatment with E2 resulted in dose-dependent protection from TNF-α–induced apoptosis. E2 at concentrations of 10−13 and 10−9 mol/L did not result in a statistically significant increase in cell count. Significant protection from apoptosis was noted at an E2 concentration of 10−7 mol/L, with total cell count increasing to 5.5±1.8×103 cells (P<.001 versus TNF-α alone). Cell counts in the wells treated with the highest E2 dose were also significantly higher than in wells treated with the two lowest doses, and the linear trend of the dose-response relationship was significant (P<.001).
In an independently performed experiment, the role of the estrogen receptor expressed by endothelial cells in the protective effect of E2 was evaluated (Fig 7C⇑). Once again, total cell count decreased from 18.7±3.7×103 cells per well under control conditions to 5.9±1.4×103 cell per well after 24 hours of TNF-α exposure. Treatment with E2 (10−9 mol/L) resulted in increased viability and an increase in cell count to 11.2±4.0×103 cells per well (P<.001 versus TNF-α alone). When cells were simultaneously exposed to TNF-α and E2 (10−9 mol/L) in the presence of the specific estrogen-receptor antagonist ICI 182780 (10−9 mol/L), including 30 minutes’ preincubation with the ICI compound, the protective effect of E2 was abolished, with total cell count decreasing to 6.3±1.6×103 cells per well (P<.001 versus TNF-α plus E2, P=NS versus TNF-α alone). In a separate series of experiments, the ICI compound alone was shown to be nontoxic in concentrations ≤10−5 mol/L (data not shown). These findings suggest that the protective effect of estrogen, inhibiting TNF-α–induced apoptosis of endothelial cells, is mediated by the estrogen receptor as it is blocked by the specific estrogen-receptor antagonist.
The increase in cell number documented with E2 treatment of HUVEC induced to undergo apoptosis by TNF-α exposure could be due to inhibition of apoptosis by E2 but might also be explained by the previously described mitogenic action of E2 on endothelial cells.17 To distinguish between these potential mechanisms, HUVEC apoptosis was evaluated by FACS analysis to quantify apoptosis on the basis of the degree of DNA fragmentation and by viability assays that measure mitochondrial function and thereby provide a sensitive and early indication of loss of viability in addition to relative cell number.16
FACS Analysis Demonstrates Inhibition of TNF-α–Induced DNA Fragmentation by E2
FACS analysis of human endothelial cells was used to quantify apoptosis by measuring the percentage of cells with a subdiploid DNA content, thus providing verification of the induction and inhibition of apoptosis suggested by phase-contrast and fluorescence microscopy and cell counting (Fig 8⇓).
In control wells, 6.5% of cells were shown to have a hypodiploid DNA content by FACS. TNF-α exposure of endothelial cells resulted in apoptosis of 39.3% of cells after 24 hours. When cells were treated with E2 simultaneous to TNF-α exposure, evidence of HUVEC apoptosis was decreased to 16% of total cells counted. Thus, evidence for nucleosomal DNA fragmentation, one of the hallmarks of apoptosis, was significantly reduced when TNF-α–exposed HUVEC were treated with E2.
When the cell culture medium, containing floating/detached cells, was collected and analyzed separately, all of the cells demonstrated a subdiploid DNA content (data not shown). This finding is consistent with apoptosis in most or all of the detached cells. Therefore, the results of the direct cell counting, demonstrating an increase in the number of attached cells when TNF-α exposure was accompanied by E2 treatment, are a clear reflection of a decrease in the rate of cell detachment that occurs with HUVEC apoptosis.
Viability Assays Verify Enhanced HUVEC Survival After E2 Treatment of TNF-α–Exposed Cells
Viability assays, which measure mitochondrial function, are capable of detecting cell death earlier than other techniques.16 In the present study, the MTS viability assay was used to confirm the protective effect of E2 against TNF-α–induced apoptosis.
TNF-α alone resulted in 35±4% cell death after 24 hours (Fig 9⇓). Simultaneous treatment with E2 demonstrated a dose-dependent increase in cell survival. At an E2 concentration of 10−13 mol/L, no significant survival benefit was conferred. At an E2 concentration of 10−11 mol/L, cell death decreased to 19±3% (P=.001 versus TNF-α alone). When the E2 concentration in the culture medium was increased to 10−9 mol/L, the percentage of dying cells decreased to 18±2% (P=.0004 versus TNF-α alone). The specific estrogen-receptor antagonist ICI 182780 completely abrogated the protective effect of E2. In these experiments, HUVEC were preincubated with the ICI compound for 30 minutes before the simultaneous addition of TNF-α and E2. This provides further evidence that the dose-dependent protective effect of E2 is mediated by the estrogen receptor.
Treatment with E2 alone for 24 hours did not result in a significant increase in cell number, which would be manifested as a net negative percentage cell death. This suggests that the increase in cell number demonstrated with direct cell counting was not the result of E2-induced proliferation. Previous studies have shown that the most pronounced effects of E2 on endothelial cell proliferation occur after 2 to 3 days in culture.17
Estrogen has long been assumed to protect against the development of atherosclerosis by amelioration of serum lipid profiles. The recent demonstration of estrogen-receptor expression in vascular smooth muscle cells40 41 and endothelial cells15 suggests that direct action of estrogen on vascular tissues could represent an alternative mechanism for the atheroprotective effect of estrogen.
The present study demonstrates a direct, receptor-mediated effect of estrogen on human endothelial cells, inhibiting cytokine-induced apoptosis of these cells. Arterial injury has been proposed as an inciting event in the initiation of atherosclerosis.13 In the injury model, disturbance of the anatomic and functional integrity of the endothelial cell monolayer is the sine qua non. The ability of estrogen to inhibit cytokine-induced apoptosis of endothelial cells therefore has interesting implications, suggesting another potential mechanism of the atheroprotective effect of estrogen. The fact that the cytokine used to induce apoptosis in the present study is known to be secreted by inflammatory cells and smooth muscle cells present within atheromatous plaque suggests that the findings of the present study have practical as well as theoretical import.
The findings of the present study also suggest a functional role for the ICE gene in TNF-α–induced endothelial cell apoptosis. A previous study33 documented expression of ICE mRNA in human atheroma by use of reverse transcription–polymerase chain reaction. In that study, immunohistochemical staining suggested ICE expression by macrophages and by smooth muscle cells. In the present study, ICE expression by apoptotic human endothelial cells is also documented. The increased expression of ICE at the mRNA and protein levels in endothelial cells committed to apoptosis by TNF-α exposure and the colocalization of ICE protein with the morphological and biochemical markers of apoptosis in individual endothelial cells provide strong evidence of a functional role for this protein in human endothelial cells.
The mechanism of the survival effect of E2 on endothelial cells remains to be clarified. Estrogen has previously been shown to promote angiogenesis both in vitro and in vivo.17 In that study, estrogen was shown to increase both proliferation and migration of endothelial cells. Previous work18 has shown that estrogen enhances TNF-α–induced expression of adhesion molecules. It is conceivable, therefore, that enhanced endothelial cell adhesion to matrix is responsible for improved survival of the endothelial cells in the present study. Attachment to matrix has previously been shown to be an important determinant of microvascular endothelial cell survival.42
The ability of TNF-α to both participate in the induction of adhesion molecule expression, which is necessary for microvascular endothelial cell survival during angiogenesis, and, at higher doses, to induce macrovascular endothelial cell apoptosis suggests that complex mechanisms control the fate of different endothelial cells under different conditions. TNF-α has been studied primarily as an “inflammatory” cytokine for many years. The possibility that TNF-α could also negatively influence endothelial cell survival is an extension of this previous work.43 44
A potential role for TNF-α in the pathobiology of the arterial wall is suggested by data from human patients as well as animal models. The demonstrated upregulation of TNF-α expression in animal models of arterial injury20 and in human atherosclerosis and restenosis19 provides evidence that regulation of the expression of this cytokine is functionally important. Expression of TNF-α has been demonstrated in human coronary atherectomy specimens,19 with increased expression noted in specimens from restenotic coronary arteries. In a heterotopic transplant model, smooth muscle cell expression of TNF-α has been demonstrated in association with proliferation of these cells,22 and blockade of TNF-α was shown to inhibit neointimal formation.21 The latter study provides the strongest evidence of a functional role for TNF-α in a pathological process in the artery wall. Thus, it would appear that the ability of E2 to inhibit the toxic effects of locally secreted TNF-α on endothelial cells may represent a valuable protective mechanism. In fact, a large number of prior studies in animal models6 7 8 9 have demonstrated the ability of E2 to inhibit the formation of neointimal lesions induced by mechanical injury or hyperlipidemia.
Recent data suggest that accelerated reendothelialization after arterial injury is associated with inhibition of neointimal formation,45 implying that reestablishment of the integrity of the endothelial cell monolayer may serve a protective function, truncating the cascade of events that leads to smooth muscle cell proliferation and neointimal formation. By analogy, the ability of E2 to protect endothelial cells from cytokine-induced injury, thereby preserving endothelial integrity, suggests another possible mechanism of the atheroprotective effect of estrogen. The present in vitro study, however, does not establish the benefit of inhibiting endothelial cell apoptosis in vivo. Such a conclusion will await the results of extensive future study.
Selected Abbreviations and Acronyms
|FACS||=||fluorescence-activated cell sorting|
|HUVEC||=||human umbilical vein endothelial cells|
|ICE||=||interleukin-1β converting enzyme|
|TNF||=||tumor necrosis factor|
- Received October 14, 1996.
- Accepted November 6, 1996.
- Copyright © 1997 by American Heart Association
Williams JK, Adams MR, Klopfenstein HS. Estrogen modulates responses of atherosclerotic coronary arteries. Circulation. 1990;81:1680-1687.
Chen SJ, Li H, Durand J, Oparil S, Chen YF. Estrogen reduces myointimal proliferation after balloon injury of rat carotid artery. Circulation. 1996;93:577-584.
Adams MR, Kaplan JR, Koritnik DR, Clarkson TB. Pregnancy-associated inhibition of coronary artery atherosclerosis in monkeys: evidence of a relationship with endogenous estrogen. Arteriosclerosis. 1987;7:378-383.
Adams MR, Kaplan JR, Manuck SB, Koritnik DR, Parks JS, Wolfe MS, Clarkson TB. Inhibition of coronary artery atherosclerosis by 17-beta estradiol in ovariectomized monkeys: lack of an effect of added progesterone. Arteriosclerosis. 1990;10:1050-1057.
Kushwaha RS, Lewis DS, Carey KD, McGill HC. Effects of estrogen and progesterone on plasma lipoproteins and experimental atherosclerosis in the baboon. Arterioscler Thromb. 1991;11:23-31.
Bush TL, Barrett-Connor E, Cowan LD. Cardiovascular mortality and non-contraceptive estrogen use in women. Circulation. 1987;75:1002-1009.
Thornberry NA, Bull HG, Calaycay JR, Chapman KT, Howard AD, Kostura MJ, Miller DK, Molineaux SM, Weidner JR, Aunins J, Elliston KO, Ayala JM, Casano FJ, Chin J, Ding G, Egger LA, Gaffney EP, Limjuco G, Palyha OC, Raju SM, Rolando AM, Salley JP, Yamin TT, Lee TD, Shively JE, MacCross M, Mumford RA, Schmidt JA, Tocci MJ. A novel heterodimeric cysteine protease is required for interleukin-1β processing in monocytes. Nature. 1992;356:768-774.
Venkov CD, Rankin B, Vaughan DE. Identification of authentic estrogen receptor expression in cultured endothelial cells: a potential mechanism for steroid hormone regulation of endothelial function. Circulation. 1996;94:727-733.
Tepper CG, Jayadev S, Liu B, Bielawska A, Wolff R, Yonehara S, Hunna YS, Seldin MF. Role of ceramide as an endogenous mediator of fas-induced cytotoxicity. Proc Natl Acad Sci U S A. 1995;92:8443-8447.
Morales DE, McGowan KA, Grant DS, Maheshwari S, Bhartiya D, Cid MC, Kleinman K, Schnaper HW. Estrogen promotes angiogenic activity in human umbilical vein endothelial cells in vitro and in a murine model. Circulation. 1995;91:755-763.
Cid MC, Kleinman HK, Grant DS, Schnaper HW, Fauci AS, Hoffman GS. Estradiol enhances leukocyte binding to tumor necrosis factor (TNF)-stimulated endothelial cells via an increase in TNF-induced adhesion molecules E-selectin, intercellular adhesion molecule type 1, and vascular cell adhesion molecule type 1. J Clin Invest. 1994;93:17-25.
Clausell N, Correa de Lima V, Molossi S, Liu P, Turley E, Gotlieb AI, Adelman AG, Rabinovitch M. Expression of tumor necrosis factor and accumulation of fibronectin in coronary artery restenosis lesions retrieved by atherectomy. Br Heart J. 1995;73:534-539.
Tanaka H, Sukhova G, Schwartz D, Libby P. Proliferating arterial smooth muscle cells after balloon injury express TNF-α but not interleukin-1 or basic fibroblast growth factor. Arterioscler Thromb Vasc Biol. 1996;16:12-18.
Clausell N, Molossi S, Sett S, Rabinovitch M. In vivo blockade of tumor necrosis factor-α in cholesterol-fed rabbits after cardiac transplant inhibits acute coronary artery neointimal formation. Circulation. 1994;89:2768-2779.
Jaffe EA, Nachman RL, Becker CG, Minick RC. Synthesis of antihemophilic factor antigen by cultured human endothelial cells. J Clin Invest. 1973;52:2745-2756.
Duan H, Chinnaiyan AM, Hudson PL, Wing JP, He WW, Dixit VM. ICE-LAP3, a novel mammalian homologue of the caenorhabditis elegans cell death protein ced-3 is activated during fas- and tumor necrosis factor-induced apoptosis. J Biol Chem. 1996;271:1621-1625.
Cerretti DP, Kozlosky CJ, Mosley B, Nelson N, Ness KV, Greenstreet TA, Cannizzaro LA, Huebner K, Black RA. Molecular cloning of the interleukin-1β converting enzymes. Science. 1992;256:97-99.
Bennett MR, Evan GI, Newby AC. Deregulated expression of the c-myc oncogene abolishes inhibition of proliferation of rat vascular smooth muscle cell by serum reduction, interferon-α, heparin, and cyclic nucleotide analogues and induces apoptosis. Circ Res. 1994;74:525-536.
Kang Y, Cortina R, Perry RR. Role of c-myc in tamoxifen-induced apoptosis in estrogen-independent breast cancer cells. J Natl Cancer Inst. 1996;88:279-284.
Canman CE, Gilmer TM, Coutts SB, Kastan MB. Growth factor modulation of p53-mediated growth arrest versus apoptosis. Genes Dev. 1995;9:600-611.
Abrahamson J, Lee JM, Bernstein A. Regulation of p53-mediated apoptosis and cell cycle arrest by Steel factor. Mol Cell Biol. 1995;15:6953-6960.
Losordo DW, Kearney M, Kim EA, Jekanowski J, Isner JM. Variable expression of the estrogen receptor in normal and atherosclerotic coronary arteries of premenopausal women. Circulation. 1994;89:1501-1510.
Karas RH, Patterson BL, Mendelsohn ME. Human vascular smooth muscle cells contain functional estrogen receptor. Circulation. 1994;89:1943-1950.
Re F, Zaneti A, Sironi M, Polentarutti N, Lanfrancone L, Dejana E, Colotta F. Inhibition of anchorage-dependent cell spreading triggers apoptosis in cultured human endothelial cells. J Cell Biol. 1994;127:537-546.
Bradley JR, Johnson DR, Pober JS. Four different classes of inhibitors of receptor-mediated endocytosis decrease tumor necrosis factor-induced gene expression in human endothelial cells. J Immunol. 1993;150:5544-5555.
Karmann K, Hughes CC, Schechner J, Fanslow WC, Pober JS. CD40 on human endothelial cells: inducibility by cytokines and functional regulation of adhesion molecule expression. Proc Natl Acad Sci U S A. 1995;92:4342-4346.
Asahara T, Bauters C, Pastore C, Kearney M, Rossow S, Bunting S, Ferrara N, Symes JF, Isner JM. Local delivery of vascular endothelial growth factor accelerates reendothelialization and attenuates intimal hyperplasia in balloon-injured rat carotid artery. Circulation. 1995;91:2793-2801.