Jump to
Abstract
This study tests the hypothesis that the control of vascular smooth muscle cell (VSMC) apoptosis is regulated by the antagonistic balance between vasoactive substances such as NO and angiotensin II (Ang II). Moreover, it is postulated that the cellular signaling pathways involved in regulating vessel tone are also coupled to the regulation of programmed cell death. Using an in vitro model system, we documented that the addition of NO donor molecules S-nitroso-N-acetylpenicillamine or sodium nitroprusside to VSMC dose-dependently induced apoptosis as documented by DNA laddering and quantified by analysis of cellular chromatin morphology. The mediator role of the guanylate cyclase signaling pathway in NO-induced apoptosis was evidenced by (1) induction of apoptosis by the 8-bromo-cGMP analogue, (2) potentiation of NO-induced apoptosis by cGMP-specific phosphodiesterase inhibition, and (3) the prevention of NO-induced apoptosis by the inhibition of the cGMP-dependent protein kinase Iα. In contrast, Ang II directly antagonized NO donor– and cGMP analogue–induced apoptosis via activation of the type I Ang II receptor. These findings suggest that the countervailing balance between NO and Ang II may determine the overall cell population within the vessel wall by regulating genetic programs determining cell death as well as cell growth.
The vasculature has the intrinsic capacity to undergo dramatic changes in architecture in response to hemodynamic stimuli and vessel injury. We have postulated that this process of vascular remodeling plays an important role in determining the natural history of vascular diseases such as atherosclerosis, restenosis after angioplasty, and hypertension.1 2 Classical paradigms defining the pathobiology of vascular disease have focused on the abnormal regulation of cell growth in response to peptide mitogens such as PDGF and basic fibroblast growth factor. However, it has become increasingly clear that vasoactive substances such as NO and Ang II also have the capacity to induce long-term changes in vessel structure in addition to acute effects on vessel tone. Studies in our laboratory and others using in vitro model systems have documented that vasoconstrictors tend to promote VSMC growth, whereas vasodilators are growth inhibitors.2 3 4 5 Moreover, we have recently used an in vivo gene transfer experimental approach to demonstrate that the endogenous generation of NO inhibits cell growth and neointimal formation after balloon injury.6 Conversely, we have shown that in vivo gene transfer of the angiotensin-converting enzyme gene and the consequent local generation of Ang II within the intact vessel wall stimulate VSMC growth and vessel wall thickening.7 Taken together, these two sets of experiments provide evidence that these countervailing autocrine-paracrine vasoactive substances are important determinants of vascular structure and lesion formation in the intact animal.
In addition to changes in the regulation of cell growth, we have postulated that the regulation of cell death by apoptosis may be another important determinant of vessel structure and lesion formation. In response to a variety of stimuli and circumstances, cells have an intrinsic capacity to activate a gene-directed program that commits the cell to a suicidal death, described as apoptosis. Recent studies of human vascular lesions using indirect techniques to assess apoptotic death suggest that this process occurs within the context of atherosclerosis and restenosis after angioplasty.8 9 10 Moreover, experimental work in animal models has confirmed that cell death by apoptosis appears to occur during the processes of vascular remodeling and lesion formation.11 12 We have postulated that vascular structure is determined in large part by a balance between cell growth and cell death by apoptosis.1 2 However, the autocrine-paracrine factors that regulate this balance between vascular cell death and cell survival remain to be defined.
It has become increasingly clear that the process of cell death by apoptosis is a relatively ubiquitous phenomenon observed in a variety of cell types.13 Studies with in vitro model systems have documented that both endothelial cells and VSMCs undergo apoptosis in response to the removal of mitogens such as PDGF-BB present within serum.14 15 16 It remained to be determined whether vasoactive substances have the capacity to modulate this process of apoptosis in vascular cells.
The present study tested the hypothesis that NO and Ang II may modulate vascular structure by regulating cell death as well as cell growth. Using an in vitro model system, our results indicate that NO is a potent inducer of VSMC apoptosis. Moreover, we have characterized a second messenger signaling pathway involving cGMP that mediates apoptosis in VSMCs. Furthermore, we have documented that Ang II exerts a countervailing influence to NO by inhibiting apoptosis via activation of the AT-I receptor. These findings have important implications for understanding the role of vasoactive substances as modulators of apoptosis during the processes of vascular remodeling and lesion formation.
Materials and Methods
Cell Culture and Experimental Protocols
Rabbit aortic VSMCs were isolated and characterized as previously described,3 and passages 5 to 9 were used for experimentation. Human umbilical artery smooth muscle cells were purchased from Clonetics Corp, and passages 5 to 9 were used for experimentation. The clonal embryonic rat thoracic aorta cell line A7r5 was obtained from American Type Culture Collection. VSMCs were maintained in polystyrene 75-cm2 flasks in DMEM/F-12 HAM (GIBCO Laboratories) supplemented with 20% (human VSMC) or 10% (rabbit and A7r5 VSMC) heat-inactivated FBS (GIBCO), penicillin (100 U/mL), streptomycin (100 μg/mL), and 25 mmol/L HEPES buffer.
VSMCs were grown in FBS to a near-confluent state in 75-cm2 flasks or six-well polystyrene plates (Nunc, Inc) and rinsed twice with serum-free medium before establishing experimental conditions. After 24 hours, adherent and nonadherent cells were harvested by trypsinization and collected for DNA extraction or stained with fluorescent DNA-binding dyes and analyzed for apoptosis as described below. The 24-hour time point was established on the basis of 48-hour pilot studies, in which detectable levels of apoptosis due to serum deprivation or exposure to NO donors were observed by 8 to 16 hours; these levels increased to a plateau within 24 hours. We observed a similar time course of induction in response to serum withdrawal using time-lapse videomicroscopy.
Reagents
SNAP was purchased from Calbiochem. SNP, NAP, bovine hemoglobin, 8-bromo-cGMP, 8-bromo-cAMP, and human synthetic Ang II were obtained from Sigma Chemical Co. Zaprinast was purchased from Biomol. Rp-8-pCPT-cGMPS and 8-pCPT-cGMP were purchased from BioLog Life Science Institute. Losartan was obtained from DuPont.
Determination of Apoptosis
Concomitant Phase-Contrast and Ultraviolet Light Microscopy
VSMCs were grown in 10% FBS to a near-confluent state in 25-cm2 flasks and incubated for an additional 24 hours in DMEM/F-12 HAM. Flasks were gassed with 95% air/5% CO2 before being sealed. An inverted microscope (Nikon Corp) equipped with a plastic environmental chamber maintained at 37°C with an external heater was used to house the flasks. To capture and store images, a low-light–sensitive monochrome CCD camera (Sony Corp) was connected to a computer (Apple Inc) equipped with a digital frame-grabbing board (Digital Vision, Inc). Light passing through an infrared filter was regulated by a controller (Fisher Scientific) to turn on for 5 seconds every minute. Digitized images were captured and stored at 10-minute intervals. At the end of the 24-hour incubation, H33342 (5 μg/mL, Molecular Probes) was added to the flask for an additional 20-minute incubation. To obtain concomitant phase-contrast and ultraviolet light images, visible and ultraviolet (excitation and emission filters to pass all wavelengths above 400 nm) light sources were balanced to resolve features of the cell membrane as well as nuclear chromatin morphology.
Microscale Analysis of DNA Fragmentation
VSMCs grown in 10% FBS to a near-confluent state in 75-cm2 flasks, as described above, were incubated for an additional 24 hours in DMEM/F-12 HAM supplemented with 0.1% FBS in the presence or absence of 500 μmol/L SNAP. Extraction of DNA from both adherent and nonadherent cells was performed using an adaptation of a previously described protocol.17 Briefly, adherent cells were collected by trypsinization and pooled with nonadherent cells in the media. Cells were pelleted by centrifugation, washed twice with PBS, and incubated at 55°C for 3 hours in a lysis buffer consisting of 100 mmol/L NaCl, 10 mmol/L Tris-Cl (pH 8.0), 25 mmol/L EDTA (pH 8.0), 0.5% SDS, and fresh 0.15 mg/mL proteinase K. The DNA was extracted twice with equal volumes of phenol/chloroform/isoamyl alcohol and precipitated in the presence of 3 mol/L ammonium acetate, 10 mmol/L magnesium chloride, and 2.5 vol of ethanol at −20°C overnight. After RNase treatment (12.5 μg/mL, 1 hour at 37°C), DNA samples were phenol-extracted and ethanol-precipitated as described above. Care was taken in the handling of cells and extracted DNA to minimize nicking or shearing of DNA, which generates random length 3′-OH ends that may compete with oligonucleosomal length fragments in the end-labeling reaction.
Extracted and purified DNA was 3′-end–labeled and size-fractionated by agarose gel electrophoresis for autoradiography as previously described.18 This technique allows for increased sensitivity in determining DNA fragmentation in relatively small quantities of DNA. Reagents for end labeling were obtained from Boehringer Mannheim and include terminal deoxynucleotide transferase enzyme, 5× concentrated reaction buffer (1 mol/L potassium cacodylate, 0.125 mol/L Tris-Cl, and 1.25 mg/mL bovine serum albumin [pH 6.6]), and cobalt chloride (25 mmol/L). Radiolabeled dideoxynucleotide ([α32P]ddATP, 3000 Ci/mmol) was obtained from Amersham. Purified DNA (1 μg) was incubated for 1 hour at 37°C in the presence of 50 μCi [α32P]ddATP, 25 U terminal deoxynucleotide transferase, 1× reaction buffer, and cobalt chloride. The reaction was terminated by the addition of 25 mmol/L EDTA (pH 8.0). Unincorporated radionucleotide was separated from labeled DNA by two successive precipitations with 0.2× vol of 10 mol/L ammonium acetate and 3× vol of 100% ethanol in the presence of 50 μg tRNA. One half of the labeled sample was loaded on a 2% agarose gel and separated by electrophoresis for 3 hours at 50 V in 1× Tris acetate EDTA buffer. The gel was subsequently dried and exposed to Kodak X-Omat film for 2 hours at −70°C.
Quantitative Analysis of Apoptosis by Fluorescence Microscopy
As an additional assay of cell death by apoptosis, we used fluorescent DNA-binding dyes to define nuclear chromatin morphology as a quantitative index of apoptosis within the cell culture system. Cells to be analyzed for apoptosis by nuclear chromatin morphology were stained with H33342 and PI (Molecular Probes) and viewed under fluorescence microscopy as previously described.19 20 Using both membrane-permeable (H33342) and -impermeable (PI) dyes in the assay allowed for the determination of cell viability, plasma membrane integrity, and an accounting of any nonapoptotic toxic or necrotic death induced in the study groups.
At the conclusion of each experimental protocol, H33342 (5 μg/mL) was added to the culture medium and incubated for 20 minutes at 37°C. The media and a PBS rinse of the culture wells were collected before the brief addition and decantation of trypsin/EDTA. Culture wells were incubated in residual trypsin/EDTA for 5 minutes in humidified air at 37°C to achieve maximal cell detachment before a rinse and collection with PBS. Collected media, rinse, and trypsinized cells were pooled and collected by centrifugation at 1200 rpm 5 minutes at 4°C. Cell pellets were resuspended in a small volume (50 μL) of serum-containing medium with 1 μg/mL H33342 and 5 μg/mL PI. An aliquot (25 μL) was placed on a glass slide, covered with a glass coverslip, and viewed under fluorescence microscopy (Leica Lab-S, 100-W mercury bulb, excitation and emission filters to pass all wavelengths >400 nm). Individual nuclei were visualized at ×400 to distinguish the normal uniform nuclear pattern from the characteristic condensed coalesced chromatin pattern of apoptotic cells.
Although chromatin undergoes condensation during mitosis, these cells can be readily distinguished from apoptotic cells by the uniform and equatorial pattern of chromatin condensation compared with the randomly coalesced pattern typical of apoptotic cells. PI-positive cells with a noncondensed chromatin morphology indicated a nonapoptotic, toxic, or necrotic death with early loss of membrane integrity. Although numerous VSMCs exhibit this pattern of toxin-induced death in response to treatment with 1% sodium azide, no significant amount of necrotic cell death was observed in the study groups. To quantify apoptosis, 400 nuclei from random microscopic fields were analyzed by an observer blinded as to the treatment group. Scoring was performed by one individual, with an intra-assay variability of 5%. The results of a representative set of experiments were further verified by a second blinded individual with an interobserver variability of <10%. Counts are expressed as the ratio of apoptotic nuclei to total nuclei to obtain percent apoptotic nuclei.
The apoptotic index obtained is a cross-sectional and not a cumulative assessment at the time of cell harvest and may be an underestimate of the total apoptotic cell population, since apoptotic cells in a prenuclear condensation phase would be scored as normal and since late-stage apoptotic cells whose cellular membranes have disintegrated may not be detected. In preliminary experiments, we have observed a high concordance between rates of apoptosis observed with this method compared with assays of subdiploid apoptotic populations using FACS analysis and phase-contrast time-lapse videomicroscopy.
Statistical Analysis
Each experimental condition was performed in triplicate, and each experiment was repeated a minimum of three times. Statistical analyses were performed by ANOVA or unpaired two-tailed Student's t test. Data are represented as the mean±SD.
Results
VSMC Apoptosis: Role of Serum Factors
Initially, we established an in vitro model system for assessing the process of apoptosis in VSMCs. Rabbit aortic VSMCs grown to near confluence and maintained in medium containing 10% serum exhibit little detectable apoptosis (1%) as quantified by the nuclear chromatin morphology assay. However, in response to serum withdrawal, there was a ninefold increase in apoptosis within 24 hours (1±0.3% in 10% serum versus 9±0.5% in serum-free medium, P<.001 [n=12]). As shown in Fig 1⇓, we assessed the concordance between morphological analyses of apoptosis provided by phase-contrast and ultraviolet light microscopy. As visualized by phase-contrast microscopy, cells deprived of serum exhibited the characteristic features of cell shrinkage, membrane blebbing, and rounding typical of apoptotic death. Moreover, simultaneous assessment of nuclear chromatin morphology by H33342 staining and ultraviolet light microscopy verified that these cells eventually manifested typical apoptotic condensed and coalesced nuclei. Additionally, as shown below, we have confirmed that the process of apoptosis defined on the basis of nuclear chromatin morphology and phase-contrast microscopy correlates with apoptosis defined on the basis of internucleosomal DNA fragmentation assessed by gel electrophoresis. In pilot titration experiments, we determined that quantities of FBS as low as 0.1% were sufficient to markedly inhibit apoptosis induced by serum withdrawal (9±0.5% serum free versus 3±0.3% in 0.1% FBS, P<.001 [n=12]). These low serum concentrations are commonly used in conventional in vitro studies of cell growth.
VSMC apoptosis: Concomitant phase-contrast and ultraviolet light microscopy. Rabbit aortic VSMCs deprived of serum for 24 hours were stained with H33342 and viewed under balanced ultraviolet and phase-contrast microscopy (×200). The cell exhibiting nuclear chromatin condensation (arrow) also exhibits morphological features characteristic of apoptosis, such as cell shrinkage, rounding, and membrane blebbing.
NO Induces VSMC Apoptosis
To test the hypothesis that NO induces apoptosis, we administered two structurally dissimilar NO donor agents, SNAP and SNP. As described above, rabbit VSMCs under low-serum (0.1%) conditions exhibited a low rate of apoptotic cell death. In contrast, the administration of SNAP (500 μmol/L) induced a fourfold increase in the rate of apoptosis compared with vehicle controls (3±0.3% vehicle [n=12] versus 13±0.6% SNAP [n=9], P<.001). As seen in Fig 2⇓, the administration of both NO donors induced apoptosis in a dose-dependent fashion. To assess the onset of NO donor–induced apoptosis, a time-course study was performed (Fig 3⇓). The addition of 500 μmol/L SNAP to rabbit VSMCs in low serum (0.1% FBS) resulted in the onset of apoptotic death within 16 hours, which reached plateau levels within 24 hours. To confirm the broad applicability of this observation, we also assessed the effect of these NO donors on the rat clonal aortic VSMC line A7r5 and on human umbilical artery smooth muscle cells. The administration of 500 μmol/L SNP resulted in a threefold increase in apoptosis (3±0.4% vehicle [n=9] versus 9±0.6% SNP [n=6], P<.001) under low serum (0.1% FBS) conditions in the A7r5 cells, whereas the administration of 500 μmol/L SNAP resulted in an eightfold increase in apoptosis (1±0.1% vehicle [n=9] versus 8±0.3% SNAP [n=6], P<.001) under similar low-serum conditions in the human VSMCs.
Dose-dependent induction of apoptosis by SNAP and SNP. Rabbit aortic VSMCs in 0.1% FBS were exposed to increasing concentrations of SNAP or SNP for 24 hours before harvest. Quantification of apoptotic cell death as measured by chromatin morphology indicates a significant (*P<.001) dose-dependent induction of apoptosis by both NO donors at the concentrations indicated (n=3, representative experiment).
Time course of SNAP-induced apoptosis. Rabbit aortic VSMCs in 0.1% FBS were exposed to 500 μmol/L SNAP, and quantification of apoptotic cell death as measured by chromatin morphology was assessed at baseline and at 8, 16, and 24 hours. Results indicate that SNAP induces apoptosis in a time-dependent manner, with a peak effect at 24 hours (n=6) (*P<.001).
To verify that NO induced cell death by apoptosis, we confirmed the concordance between morphological analyses by phase-contrast and ultraviolet light microscopy that we previously established for cell death induced by serum withdrawal (Fig 1⇑). For further confirmation of apoptotic death, we also extracted DNA for assessment of internucleosomal DNA fragmentation by gel electrophoresis. As shown in Fig 4⇓, DNA extracted from cells maintained in low serum remained intact without significant fragmentation. However, exposure to SNAP induced a characteristic pattern of DNA fragmentation indicative of apoptotic cell death. Similar results were obtained after SNP administration (data not shown).
VSMC apoptosis in response to SNAP: DNA laddering. Autoradiogram shows extracted 3′-end–labeled and electrophoretically separated DNA from rabbit aortic VSMCs cultured for 24 hours in 0.1% FBS (lane 1) or 0.1% FBS with the addition of 500 μmol/L SNAP (lane 2). Oligonucleosomal-length (200-bp integer) DNA fragmentation indicates that SNAP induces VSMC apoptosis.
To confirm that the induction of apoptosis in response to SNAP and SNP administration was related to NO release, we assessed the effect of blocking the action of NO on cell function by coincubation with hemoglobin. It is well established that NO released by SNAP or SNP binds to hemoglobin with high avidity and that NO sequestered by hemoglobin is unable to activate cellular functions.21 Indeed, the administration of SNAP (500 μmol/L) in the presence of hemoglobin (50 μmol/L) failed to induce VSMC apoptosis (Fig 5⇓). The specificity of hemoglobin blockade was confirmed by finding that hemoglobin had no effect on cell viability in low serum (3±0.3% for 0.1% FBS [n=9] versus 3±0.4% for 0.1% FBS+hemoglobin [n=6], P>.05) and failed to inhibit the induction of apoptosis by serum withdrawal (10±0.5% for serum withdrawal [n=9] vs 10±0.4% for serum withdrawal+hemoglobin [n=6], P>.05). Furthermore, we documented that the administration of the parent compound NAP (500 μmol/L), which lacks the capacity to release NO, also failed to induce VSMC apoptosis. Thus, the effect of SNAP and SNP on VSMC cell death represents a specific response to NO.
Induction of apoptosis by SNAP is NO dependent. Bar graph shows percentage of rabbit aortic VSMCs undergoing apoptosis in response to a 24-hour incubation in (1) 0.1% FBS (n=12), (2) 0.1% FBS with 500 μmol/L SNAP (n=9), (3) 0.1% FBS with 500 μmol/L SNAP coincubated with 50 μmol/L hemoglobin (Hgb) (n=6), or (4) 0.1% FBS with 500 μmol/L NAP (n=6). Analysis of nuclear chromatin morphology indicates an induction of apoptosis by SNAP (*P<.001), which is effectively quenched by Hgb and is not replicated by the parent compound NAP (P<.001).
Cellular Mechanisms of NO-Induced VSMC Apoptosis
NO induces vascular smooth muscle relaxation via the activation of soluble guanylate cyclase and the consequent increase in intracellular cGMP. Therefore, we tested the hypothesis that the induction of apoptosis by NO is mediated by the cGMP-dependent cell-signaling pathway. In pilot studies, we confirmed previous studies that have documented that the concentrations of SNP and SNAP used in the present study induce dose-dependent elevations in VSMC cGMP levels.22 23 Therefore, we initially examined whether the effects of NO on VSMC cell death could be mimicked by stable analogues of cGMP. As shown in Fig 6⇓, the administration of 8-bromo-cGMP induced VSMC apoptosis in a manner similar to that observed with NO donors. This response appeared to be specific to the cGMP-mediated second messenger pathway, since the administration of cAMP analogues had no effect on VSMC survival.
NO-induced apoptosis is mimicked by cGMP but not cAMP. Bar graph shows percentage of rabbit aortic VSMCs undergoing apoptosis in response to a 24-hour incubation in (1) 0.1% FBS (n=12), (2) 0.1% FBS with 1 mmol/L 8-bromo-cGMP (cGMP) (n=6), or (3) 0.1% FBS with 1 mmol/L 8-bromo-cAMP (cAMP) (n=6). Analysis of nuclear chromatin morphology indicates a significant (*P<.001) induction of apoptosis by cGMP but not cAMP.
To further define the signaling pathway mediating NO-induced apoptosis in VSMCs, we used two additional experimental approaches. We postulated that if cGMP levels regulate VSMC apoptosis, inhibition of intracellular cGMP degradation should enhance NO-induced cell death. Although the cGMP-specific phosphodiesterase inhibitor zaprinast (50 μmol/L)24 25 had minimal effects on baseline cell viability in the absence of agonists, coadministration of zaprinast with SNAP (100 μmol/L) significantly potentiated NO-induced apoptosis compared with submaximal concentrations of SNAP alone (Fig 7⇓). Furthermore, we tested the hypothesis that activation of the cGMP-dependent protein kinase Iα is an essential mediator of NO-induced VSMC death. In the absence of agonist stimulation, the cGMP-dependent protein kinase Iα inhibitor Rp-8-pCPT-cGMPs26 (100 μmol/L) had no effect on baseline apoptosis under low-serum conditions (Fig 8⇓). However, coincubation of the cGMP-dependent protein kinase Iα inhibitor with SNAP (500 μmol/L) abolished NO-induced VSMC apoptosis (Fig 8⇓). As an additional assessment of the specificity of the inhibitory effect, we confirmed that Rp-8-pCPT-cGMPs failed to inhibit apoptosis induced by serum withdrawal (9±0.7% serum free [n=9] versus 8±0.8% serum free+Rp-8-pCPT-cGMPs [n=6], P>.05). In contrast to the effect of inhibiting cGMP protein kinase Iα, activation of this component of the signal transduction pathway with the agonist 8-pCPT-cGMP27 mimicked NO-induced apoptosis, as evidenced by DNA laddering (Fig 9⇓). Taken together, these findings demonstrate that NO induces VSMC apoptosis via a cGMP-dependent pathway that involves the activation of protein kinase Iα.
Induction of apoptosis by NO is potentiated by inhibition of cGMP breakdown. Bar graph shows percentage of rabbit aortic VSMCs undergoing apoptosis as quantified by analysis of nuclear chromatin morphology after a 24-hour exposure to (1) 0.1% FBS (n=12), (2) 0.1% FBS with 100 μmol/L SNAP (n=9), (3) 0.1% FBS with 100 μmol/L SNAP coincubated with 50 μmol/L of the cGMP-specific phosphodiesterase inhibitor zaprinast (ZAP) (n=6), or (4) 0.1% FBS with 50 μmol/L ZAP alone (n=6). ZAP markedly potentiated (P<.001) the induction of apoptosis by submaximal concentrations of SNAP (*P<.001) and had minimal effects on baseline cell viability.
NO-induced apoptosis is dependent on protein kinase Iα activity. Bar graph shows percentage of rabbit aortic VSMCs undergoing apoptosis in response to a 24-hour exposure to (1) 0.1% FBS (n=12), (2) 0.1% FBS with 500 μmol/L SNAP (n=9), (3) 0.1% FBS with 500 μmol/L SNAP coincubated with 100 μmol/L of the cGMP-dependent protein kinase Iα inhibitor, Rp-8-pCPT-cGMPS (Rp-cGMP) (n=6), or (4) 0.1% FBS with 100 μmol/L of Rp-cGMP alone (n=6). Inhibition of the cGMP-dependent protein kinase Iα by Rp-cGMP abolishes the apoptosis induced by maximal concentrations of SNAP but has no effect on basal level of apoptosis (*P<.001).
Apoptosis is induced by activation of protein kinase Iα. Autoradiogram shows extracted 3′-end–labeled and electrophoretically separated DNA from rabbit aortic VSMCs cultured for 24 hours in 0.1% FBS (lane 1) or 0.1% FBS with the addition of 1 mmol/L of the protein kinase Iα agonist 8-pCPT-cGMP (lane 2). Oligonucleosomal-length (200-bp integer) DNA fragmentation indicates that 8-pCPT-cGMP induces VSMC apoptosis.
Countervailing Influence of Ang II on NO-Induced Apoptosis
The homeostatic regulation of vascular tone involves an interaction of cellular signals in which the cGMP-mediated vasorelaxation induced by NO can be counteracted by the action of the vasoconstrictor Ang II. Therefore, we examined whether the cGMP-mediated induction of apoptosis by NO could also be counteracted by the action of Ang II. Despite the administration of high quantities of SNP and SNAP (500 μmol/L) that induced a fourfold increase in VSMC apoptosis, coincubation with Ang II (300 nmol/L) nearly abolished the induction of cell death in response to NO in rabbit VSMCs (Fig 10⇓). To further define signaling mechanisms governing this antagonism between Ang II and NO in the regulation of apoptosis, we tested the hypothesis that the Ang II antiapoptotic signal is downstream from the generation of cGMP. As shown in Fig 10⇓, Ang II inhibited apoptosis induced by the protein kinase Iα agonist 8-pCPT-cGMP.
Countervailing influence of Ang II on NO-induced apoptosis. Bar graph shows percentage of rabbit aortic VSMCs undergoing apoptosis after a 24-hour exposure to (1) 0.1% FBS (n=12), (2) 0.1% FBS with 500 μmol/L SNP (n=9), (3) 0.1% FBS with 500 μmol/L SNP coincubated with 300 nmol/L Ang II (n=6), (4) 0.1% FBS with 500 μmol/L SNAP (n=9), (5) 0.1% FBS with 500 μmol/L SNAP coincubated with 300 nmol/L Ang II (n=6), (6) 0.1% FBS with the addition of 1 mmol/L of the protein kinase Iα agonist 8-pCPT-cGMP (n=6), or (7) 0.1% FBS with 1 mmol/L of 8-pCPT-cGMP coincubated with 300 nmol/L Ang II (n=6). Quantitative analysis of nuclear chromatin morphology indicates that whereas incubation with SNAP, SNP, or 8-pCPT-cGMP induces a fourfold increase in apoptosis (*P<.001), coincubation with Ang II nearly abolishes the NO and protein kinase Iα agonist–induced cell death (P<.001).
To further examine the antiapoptotic properties of Ang II, we also assessed the effect of Ang II administration on the induction of apoptosis triggered by serum withdrawal. As shown in Fig 11⇓, Ang II (300 nmol/L) substantially inhibited cell death in human VSMCs under these conditions. The antiapoptotic effect of Ang II was inhibited by the AT-I–specific receptor antagonist losartan (Fig 11⇓). Similar results were obtained with rabbit VSMCs (data not shown). These data indicate that Ang II has broad efficacy in activating a cellular pathway(s) that prevents apoptosis in response to either NO or withdrawal of the survival factors present within serum.
Antiapoptotic effect of Ang II (AII) is mediated by the AT-I receptor. Bar graph shows percentage of human umbilical artery VSMCs undergoing apoptosis in response to a 24-hour exposure to (1) serum-free medium (SF) (n=6), (2) SF with 300 nmol/L AII (n=6), (3) SF with 300 nmol/L AII coincubated with 10 μmol/L of the AT-I receptor antagonist losartan (n=6), or (4) SF with 10 μmol/L losartan. Quantitative analysis of nuclear chromatin morphology indicates that the apoptosis induced by serum withdrawal is inhibited by relatively low concentrations of AII (*P<.001). The antiapoptotic effect of AII is reversed by selective blockade of the AT-I receptor by losartan (P<.001). Losartan administration has no effect on baseline apoptosis.
Discussion
We have postulated that the process of vascular remodeling in atherosclerosis, restenosis after angioplasty, and hypertension plays an important role in the natural history of these diseases.1 2 A rapidly emerging body of evidence suggests that vascular remodeling and lesion formation is determined in large part by the balance between cell growth and cell death by apoptosis.8 9 10 11 Our laboratory and others have previously shown that vasodilators such as NO inhibit cell growth, whereas vasoconstrictors such as Ang II promote cell growth.2 3 5 6 In the present study, we have extended this paradigm by demonstrating that these vasoactive substances have similar countervailing influences on the regulation of VSMC cell death by apoptosis.
The present study has documented that the administration of either SNP or SNAP induces apoptosis of VSMCs. We used three well-described and complementary methodological approaches for characterizing the process of apoptosis–phase-contrast microscopy,15 nuclear chromatin morphology,19 and gel electrophoresis for internucleosomal DNA fragmentation.18 In additional studies, we have confirmed that these methods are concordant with other assays of apoptosis, such as in situ end labeling of DNA28 and FACS analysis of subdiploid populations19 29 in our in vitro model system.30 Based on these various biochemical and morphological criteria, it is clear that these NO donor agents induce a form of VSMC cell death that can be described as apoptosis, as conventionally defined.31 32
Overall, the present study indicates that the effect of SNP and SNAP administration on VSMC apoptosis is mediated by NO according to the following lines of evidence: (1) two structurally dissimilar NO donor agents produce similar results, (2) the effect is dose dependent, (3) the induction of apoptosis is blocked by the NO-quenching agent hemoglobin, (4) the parent compound NAP has no effect alone, and (5) the induction of apoptosis is triggered by an agonist that activates a cGMP-dependent signaling pathway and is therefore likely to be NO. Given the short half-life of NO (5 to 20 seconds) and the kinetics of NO release from these donor agents over a 24-hour period, it is difficult to determine the precise level of NO that induces VSMC apoptosis. However, previous studies suggest that the quantities of NO released by these agents in vitro is quite comparable to levels of NO endogenously generated by endothelial cells and VSMCs under stimulated conditions as assayed by the NO electrode technique or measurement of the resultant cGMP levels.33 34 Furthermore, in recent studies, we have observed that the administration of natriuretic peptides (atrial natriuretic peptide or C-type natriuretic peptide) in nanomolar concentrations is also capable of activating cGMP-dependent signaling pathways that induce VSMC apoptosis.22 Taken together, these data indicate that the quantities of NO generated under physiological and/or pathophysiological circumstances have the capacity to induce VSMC apoptosis. We speculate that the capacity of NO to induce apoptosis in vivo may be dependent on the coexpression of other proapoptotic mediators present with the cellular microenvironment. For example, it was recently reported that cytokines, such as tissue necrosis factor-α and interleukin-1β, can induce VSMC apoptosis when administered in a combinatorial manner and that the induction of cell death is mediated in part via the induction of autocrine-paracrine NO generation.35 The present study provides complementary evidence that NO is sufficient to induce apoptosis in rat, rabbit, and human VSMCs in the absence of costimulation with cytokines. Our data are in accord with other studies that have shown that NO has the capacity to induce apoptosis in nonvascular cell types.36 37 However, it is noteworthy that NO has been shown to be a bifunctional modulator of cell fate capable of either stimulating or inhibiting cell death, depending on the cell type.38 The present study provides the first characterization of the effect of NO on the regulation of apoptosis in VSMCs.
The regulatory mechanisms and signaling pathways that determine cell death versus cell survival in VSMCs are poorly defined. On the basis of prior evidence indicating that NO regulates VSMC tone via a cGMP-dependent pathway, we postulated that this same signal transduction mechanism may also mediate the induction of apoptosis by NO. The essential role of a cGMP-dependent pathway in mediating NO-induced apoptosis was demonstrated by several observations: (1) the concentrations of NO donors used in the present study induce significant dose-dependent increases in intracellular levels of cGMP capable of modulating VSMC function as previously described23 and confirmed in our laboratory,22 (2) stable cGMP analogues mimic the effect of the NO donor agents, (3) the effect of NO on cell death was potentiated by the cGMP-specific phosphodiesterase inhibitor zaprinast, and (4) the cGMP-dependent protein kinase Iα inhibitor abolished the induction of apoptosis in response to NO. Given that cGMP can modulate phosphodiesterase activity and thereby modify cAMP levels,39 we also confirmed that cAMP analogues had no effect on cell viability. Our recent observation that natriuretic peptides also induce apoptosis is consistent with the role of cGMP as a cellular mediator of a proapoptotic signal.22 Overall, these findings suggest that conventional second messenger systems such as the cGMP pathway are linked to the regulation of apoptosis as well as the regulation of the contractile apparatus and cell growth in these vascular cells.
It is of interest that the homeostatic regulatory mechanisms controlling vascular tone and structure appear to involve countervailing stimulatory and inhibitory factors. On the basis of evidence that Ang II opposes the effects of NO on vessel tone and cell growth, we tested the hypothesis that Ang II may prevent NO-induced apoptosis. Indeed, Ang II is a potent antiapoptotic factor capable of reversing the effects of NO as well as the response to serum withdrawal. The capacity of Ang II to inhibit serum withdrawal–induced apoptosis was noted at concentrations as low as 10 nmol/L. The cellular signaling pathway that mediates the antiapoptotic effect of Ang II remains to be explored. The observation that Ang II inhibits cGMP analogue–induced apoptosis suggests that the survival signal is downstream from the activation of cGMP protein kinase. Given the pleiotropic effects of Ang II on a variety of second messenger systems, such as calcium, protein kinase C, mitogen-activated protein kinase, and other tyrosine kinases, there are many potential mechanisms by which angiotensin may stimulate a cell survival signal.40 It is intriguing that Ang II binds to two major subtypes of receptors and that the AT-I type confers an antiapoptotic signal, whereas the AT-II type receptor, which is primarily expressed during ontogeny, appears to promote apoptosis.41 Delineation of the divergent signaling pathways coupled to these two receptor types is worthy of further investigation.
The present data indicate that Ang II may promote cell accumulation by preventing apoptosis in addition to stimulating cell cycle progression. It is conceivable that the influence of Ang II on cell survival may help reconcile conflicting data concerning whether Ang II induces hypertrophy versus hyperplasia in various contexts. For example, in contexts in which there is significant apoptosis, an antiapoptotic factor such as Ang II may induce an increase in cell number by blocking cell death rather than enhancing mitogenesis. Given the recent evidence that suggests a substantial rate of apoptosis in atherogenesis and restenosis after angioplasty,8 9 10 it is conceivable that weak mitogens such as Ang II that inhibit cell death may nevertheless play a substantial role in lesion formation that is not reflected in conventional assays of DNA replication. Further studies are necessary to define the effect of modulating apoptosis on the pathogenesis of vascular lesion formation by antiapoptotic factors such as Ang II.
The regulation of cell death by apoptosis appears to involve a delicate balance between proapoptotic versus antiapoptotic mediators. In accordance with previous observations,15 16 we observed that serum withdrawal induced cell death by apoptosis in VSMCs. These data indicate that factors within serum not only promote cell replication but are also necessary to maintain cell viability. The recent observation that VSMCs derived from atherosclerotic plaques have an increased propensity to undergo apoptosis compared with nonplaque smooth muscle cells indicates that variances in cell phenotype modulate the cell death program.16 In the present study, we confirmed that the qualitative effects of NO and Ang II on the regulation of apoptosis in VSMCs were consistent across species and were verified in different cell isolates including a clonal cell line. We conclude that NO and Ang II are regulators of VSMC apoptosis and that the effect of these agonists on cell fate is likely to be modulated by cell phenotype and other mediators present within the cellular milieu.
In summary, we have shown that vasoactive substances such as NO and Ang II have the capacity to regulate apoptosis in VSMCs. In accordance with the countervailing effects of these factors in the regulation of vessel tone and cell growth, NO induces apoptosis, whereas Ang II inhibits apoptosis. We have shown that the induction of apoptosis by NO is mediated by a cGMP-dependent pathway and that the AT-I type receptor mediates the antiapoptotic effect of Ang II. We speculate that the balance between NO and Ang II may play an important role in determining vessel wall cellularity and thereby modulate vascular structure. It is intriguing that many forms of vascular disease, such as atherosclerosis, restenosis after angioplasty, and hypertension, are characterized by decreased bioactivity of NO and/or increased generation of Ang II.2 We speculate that an imbalance between the expression of these vasoactive substances may perturb the homeostatic balance that regulates vessel tone, cell growth, and cell death. These findings have important implications for understanding the role of vasoactive substances as determinants of vascular remodeling and lesion formation.
Selected Abbreviations and Acronyms
| 8-pCPT-cGMP | = | 8-(4-chlorophenylthio)-guanosine-3′,5′-cyclic monophosphate |
| Ang II | = | angiotensin II |
| AT-I receptor | = | type I Ang II receptor |
| FACS | = | fluorescence-activated cell sorter |
| H33342 | = | Hoechst 33342 |
| NAP | = | N-acetylpenicillamine |
| PDGF | = | platelet-derived growth factor |
| PI | = | propidium iodide |
| Rp-8-pCPT-cGMPS | = | 8-(4-chlorophenylthio)-guanosine-3′,5′-cyclic monophosphorothioate, Rp-isomer |
| SNAP | = | S-nitroso-N-acetylpenicillamine |
| SNP | = | sodium nitroprusside |
| VSMC | = | vascular smooth muscle cell |
Acknowledgments
This study was supported by the Baxter Foundation and National Institutes of Health grant HL-48638. Dr Gibbons is a Pew Biomedical Research Scholar. Dr Pollman is a recipient of an American Heart Association Postdoctoral Fellowship Award (California Affiliate, Inc).
- Received May 21, 1996.
- Accepted July 12, 1996.
References
This Issue
Jump to
Article Tools
Share this Article
- Vasoactive Substances Regulate Vascular Smooth Muscle Cell ApoptosisPermalink:
