Role of Reactive Oxygen Intermediates in Cytomegalovirus Gene Expression and in the Response of Human Smooth Muscle Cells to Viral Infection
Because cytomegalovirus (CMV) may contribute to restenosis and atherosclerosis and because smooth muscle cells (SMCs) are involved in these disease processes, we examined CMV-SMC interactions. Using confocal microscopy to identify a redox-sensitive fluorescent marker, we found that CMV infection of SMCs generates intracellular reactive oxygen intermediates (ROIs). CMV also activated nuclear factor κB (NFκB), a cellular transcription factor, as demonstrated by increased NFκB binding to DNA (electrophoretic mobility shift assay). Antioxidants inhibited activation, suggesting a role of ROIs in CMV-induced NFκB activation. By using antioxidants to assess the role of ROIs in modulating virally mediated effects, we also found that CMV-induced ROIs (1) are critical to the transactivation of the viral major immediate promoter (MIEP) by its immediate-early protein IE72 (determined by cotransfection of an IE72 expression vector and a reporter gene downstream from the MIEP) and (2) are necessary for IE72 expression (determined by immunocytochemistry) and viral replication (determined by viral titer assay on indicator cells) following CMV infection of SMCs. Because ROIs, through activation of NFκB, can also induce expression of cellular genes involved in immune and inflammatory responses, the ROI response to CMV infection may also represent a parallel survival mechanism that has evolved in the host cell to protect against viral infection. We conclude that CMV induces intracellular ROI generation within minutes after infection of SMCs and then uses these ROIs to facilitate its own gene expression and replication. Conversely, antioxidants inhibit CMV immediate-early gene expression and viral replication.
Cytomegalovirus, a member of the herpesvirus group, is capable of infecting many tissues and cell types, including coronary and pulmonary arterial SMCs. After primary infection, the virus persists in certain tissues in a latent state, during which no viral gene products are expressed.
We recently reported the detection by polymerase chain reaction of CMV DNA in coronary restenosis lesions in patients who had previously undergone successful balloon angioplasty.1 On the basis of this and other findings, we hypothesized that CMV contributes to the development of restenosis after reactivation of the virus via angioplasty-induced injury. CMV DNA also has been detected in arterial walls of patients with atherosclerosis, and it is possible that periodic reactivation of the virus also predisposes patients to develop atherosclerosis.2 3 If CMV does play a role in these diseases, it would be important to identify those changes in the cellular environment that facilitate CMV reactivation and expression of its gene products. Therefore, we focused on whether CMV infection of SMCs triggers a cellular response that is conducive to reactivation of the virus and to viral gene expression.
We were particularly interested in identifying a response that could activate NFκB. This pleiotropic transcription factor transactivates the major promoters of several viruses, including HIV and CMV.4 5 Most important, one of the mechanisms by which HIV gene expression is enhanced and by which HIV is reactivated from latency is through the generation of ROIs, which exert their effects, at least in part, by activation of NFκB.4 Several cytokines, upon binding to their receptors, have been shown to activate such a signaling pathway, which ultimately mediates expression of many cellular genes, including those involved in the immune and inflammatory responses.4 Therefore, paradoxically, the same signaling response may both activate latent viruses and constitute an important cellular defense against infecting pathogens.
Because CMV has evolved strategies to coopt cellular mechanisms that enhance its own survival,6 we asked in this investigation whether CMV infection of SMCs generates increased intracellular levels of ROIs and, if so, whether such changes in the redox state of the cell activate NFκB and enhance CMV IE gene expression. In addition, one of the IE gene products of CMV, IE72, is a potent transactivator of its own promoter, the MIEP. Although it has been reported that this action is mediated by NFκB,7 8 9 the interactions between NFκB, IE72, and free radicals are largely unknown. Therefore, we also sought to determine the effect of antioxidants on the IE72-dependent transactivation of the MIEP and on the ability of the virus to replicate and to exert cytopathic effects.
Materials and Methods
Cells and Virus
Rat primary aortic SMCs, passages 4 to 12, were grown in medium 199 supplemented with 10% FBS, 2 mmol/L l-glutamine, 1 unit/mL penicillin-streptomycin, and 1 μg/mL medium (GIBCO/BRL). Human coronary SMCs (passages 4 to 8) were grown from explants of a 1-cm segment of normal human coronary artery, as described elsewhere.1 Human pulmonary and coronary artery SMCs (passages 4 and 5) and their optimal medium (SmGM), including 5% FBS, were purchased from Clonetics. Cells were kept at 37°C in an atmosphere of 5% CO2 and were infected or transfected during the log phase of their growth. Human CMV, Towne strain, was passaged in human embryonic lung fibroblasts (HEL 299, American Type Culture Collection), as described before.10 Coronary SMCs were infected at MOIs of 1, 2, or 5. Immunocytochemistry was performed 16 hours after infection; Western blot analyses of IE72 were performed 16 hours, 48 hours, and 10 days after infection; and ROI studies were performed 3 hours after infection.
The construction of the pHD101CAT3 (MIEP-CAT) plasmid11 and the pRSV72 plasmid, a gift from J.A. Nelson, Oregon Health Sciences University, Portland, Ore,12 has been described previously. The 3X-κB-CAT and 3X-mutκB-CAT plasmids contain three copies of a wild-type κB DNA binding sequence (TGGGGATTCCCCA) or three copies of a mutated sequence (TGCGGCTTCCCGA). They were gifts from A.S. Baldwin, Jr, and are described elsewhere.13 The pRSV-CAT and pRSV-IL2R were gifts from Bruce Howard (National Institutes of Health, Bethesda, MD).
Assessment of Intracellular Redox State
Intracellular generation of ROIs during the first stages of CMV infection was measured using DCFH-DA (Molecular Probes).14 15 16 Fluorescence was monitored and recorded by confocal laser scanning microscopy (Leica TCS4D, Leica Lasertechnik). Excitation and emission wavelengths were 488 and 520 nm, respectively. Images were collected using a 512×512 pixel format and archived for later analysis. The intensity of the fluorescence was quantified with the analysis software provided with the confocal microscopy system. Three different fields (each measuring 0.015 mm2 in area) were analyzed for each time point of each experiment. By means of computer software, the relative fluorescence intensity was calculated by dividing the total intensity of the fluorescence in the measuring field (expressed on a 0- to 255-step gray scale of fluorescence emission) by the percentage of the area of the field occupied by fluorescent cells. This served to compensate for variations in the number of cells in different measuring fields.
Effects of CMV Infection on ROI Generation
Coronary SMCs were grown in four-well chamber slides for 72 hours and infected with 5 MOI of CMV for 2 hours. After removal of the viral supernatant and one wash with HBSS without phenol red (GIBCO), separate wells were treated for 1 hour with the agents described below. Virus-free supernatant was used as a control and was prepared as described before.17 The cells were then washed with HBSS and incubated for 5 minutes at 22°C with fresh HBSS containing freshly prepared 5 μmol/L DCFH-DA. Fluorescence was monitored and recorded by confocal laser microscopy. Mock-infected SMCs without or with drug treatment were used as negative controls for all protocols.
Drugs Used to Elucidate Mechanisms Responsible for CMV-Induced ROI Generation
To determine the mechanisms responsible for ROI generation by viral infection, 2.5×104 coronary SMCs were seeded in four-well chamber slides (NUNC) and grown for 72 hours in their optimal medium (described above). The cells were then starved in 0.5% FBS for 72 hours and treated with 10% FBS for 2 to 4 hours before infection with 5 MOI of CMV in serum-free medium. Appropriate aliquots of virus stock were diluted in serum-free and antibiotic-free medium, and 300 μL of this viral suspension was applied to each well for 2 hours at 37°C and 5% CO2/95% humidity. Free virus was then removed, and the cells were washed twice with serum-free medium and then treated with drug or enzyme.
NAC, PDTC, or Tiron
The pH of a 0.4 mol/L aqueous stock solution of NAC (Janssen Pharmaceuticals) was adjusted to 7.4 with NaOH. PDTC was prepared as a 30 mmol/L and Tiron as a 1 mol/L aqueous stock solution. The solutions were sterilized by filtration and diluted in serum-free medium to give final concentrations of 1 to 30 mmol/L NAC, 1 to 30 μmol/L PDTC, and 1 to 10 mmol/L Tiron. Coronary SMCs at 2 hours after infection were treated with the drugs for 60 minutes. After drug removal, the cells were washed with HBSS, exposed to 5 μmol/L DCFH-DA, and monitored with a confocal microscope.
Using the same protocol, coronary SMCs were treated with 5 to 35 units/mL catalase (Calbiochem No. 219008) diluted in fresh medium.
To determine whether the xanthine/xanthine oxidase system and superoxide generation contribute to the CMV-mediated induction of ROIs and CMV IE gene expression, two separate experiments were performed: infected SMCs were treated (1) with Tiron (see above) or (2) with oxypurinol (Sigma Chemical Co). For the oxypurinol experiments, 100 mmol/L oxypurinol (Sigma Chemical Co) was dissolved in dimethyl sulfoxide and diluted in medium to give final concentrations of 1 to 10 μmol/L.
For each of the above experiments, cells were returned to the incubator and monitored for at least 2 weeks. The cells grew normally, indicating that the measured effects were not due to cell death.
Mobility Shift Assay
Coronary SMCs were grown to 90% confluence in 175-cm2 flasks. The growth medium was removed, and SMCs were pretreated for 1 hour as follows: with serum-free medium and vehicle or with 20 mmol/L NAC in serum-free medium. After 1 hour, the medium was renewed, and SMCs were infected with 5 MOI of CMV for 1, 30, 60, and 120 minutes. Cells were then washed twice with PBS and harvested with a cell scraper, and nuclear lysates were obtained using the buffers and the protocol as described previously.18 Double-stranded wild-type NFκB oligonucleotides containing DNA binding sites of NFκB (Promega) were end-labeled with polynucleotide kinase and [γ-32P]ATP. Approximately 0.5 ng of labeled DNA (50 000 dpm), 3 μg of nuclear extract, and 100 ng of poly(dI-dC) copolymer were mixed with 10 mmol/L HEPES (pH 7.9), 25 mmol/L KCl, 0.2 mmol/L EDTA, 1 mmol/L dithiothreitol, 10% glycerol, and 0.05% Nonidet P-40 in a final volume of 20 μL. This mixture was incubated for 30 minutes at 22°C, and DNA-protein complexes were resolved on 7% native acrylamide gels and then run in 5 mmol/L Tris/38 mmol/L glycine running buffer.
Transfections and CAT Assays
Rat SMCs were either transfected with MIEP-CAT plasmid or were cotransfected with the MIEP-CAT and pRSV72 via the lipofection method using N-[1-(2,3-dioleoyloxy)propyl]-N,N,N-trimethylammonium methyl sulfate (DOTAP) reagent (Boehringer) as previously described.1 After 18 hours, the medium was renewed for 4 hours and then the cells were transfected in serum-free Opti-MEM (GIBCO) with DOTAP/plasmid DNA for 18 hours. The medium was replaced with medium 10 for 40 hours and then with serum-free medium 199 before treatment with H2O2, NAC, or PDTC. The cells were then scraped and lysed for CAT assays as described before.1
Human coronary SMCs were grown in SmGM supplemented with 5% FBS. For the CAT assay, cells were seeded in 10-cm plates (5×105) containing SmGM and 10% FBS. The cells were transfected 22 hours later (after a medium change to 13.5 mL of SmGM without serum or antibiotics) with 0.1 μg/mL of MIEP-CAT. For the cotransfections, we used 0.1 μg/mL MIEP-CAT (plus 1 μg/mL pRSV72) and DOTAP (Boehringer Mannheim) or lipofectamine (GIBCO/BRL) according to the manufacturers' instructions. CAT assays were performed by phase extraction as described previously.19 A total of three experiments were performed.
Transfections were also performed using human pulmonary artery SMCs. Transfection efficiency was 1% in human SMCs and 3% to 5% in rat SMCs, as assessed by magnetic affinity cell sorting.20
Viral Titer Assay and Cytopathic Effects
Human CMV (CMV Towne, passage 45) was passaged in our laboratory in HEL 299 fibroblasts to titers ranging from 1×106 to 1×109/mL medium as estimated by the Reed-Muench method (TCID50).21 Stock aliquots of 0.1 to 1 mL were frozen at −70°C. Indicator cells (HEL 299) were seeded at 3000 per well in 96-well plates and triplicate wells were infected 24 hours later with 10-fold dilutions of the virus stock (1:100 through 1:109) in 200 μL serum-free medium per well. The inoculum was replaced after 2 hours with growth medium which was renewed every 72 hours thereafter. The cultures were monitored daily under the microscope for 2 to 3 weeks, cytopathic effects were recorded, and the titer (TCID50) was calculated by the Reed-Muench Interpolation Formula.21
Coronary SMCs were seeded in 48-well plates (15 000/cm2) and, for immunocytochemistry of IE72 (72-kD IE CMV region 1 product), in eight-well glass chamber slides (NUNC) for 48 to 72 hours and then infected with CMV at 5 MOI. Two hours after adsorption, the virus-containing supernatant was removed, and duplicate wells were treated with 5, 10, or 20 mmol/L NAC. In preliminary experiments, we determined that 40% of viral activity in infected coronary SMCs was cell-associated and 60% was in the supernatant. Therefore, we harvested infected SMCs (72 hours after infection) by scraping the cells into the medium, then sonicating the suspension, and plating aliquots at dilutions of 10−1 on indicator cells (HEL 299) at 80% to 90% confluence.
Cytopathic effects and plaque formation were assessed 3 to 10 days later. At 10 days after infection, cytopathic changes were evident in 60% to 80% of the SMCs. Additional plates seeded with coronary SMCs were infected or mock-infected, treated with NAC as described above, and harvested for cell counting 24, 48, or 96 hours later to ensure that changes in viral titer were not due to changes in cell number after infection.
Cytopathic effect was assessed by counting the number of foci of infected cells exhibiting cytomegalic changes. Coronary SMCs were examined 96 hours after infection; NAC (10 or 20 mmol/L) was added immediately after removal of free virus and again after renewal of media at 48 hours after infection.
Coronary SMCs or pulmonary SMCs in the eight-well slides were overlaid with CMV (2 MOI) for 2 hours; then, after removal of the viral supernatant, they were treated with NAC or oxypurinol for a total of 16 hours. The cells were then fixed, incubated overnight with 6E1 anti-IE72 antibody (Vancouver Biotech), and stained with the Vectastain Elite kit. The procedure has been described previously in detail.22 A total of three experiments were performed.
Pulmonary SMCs or coronary SMCs were grown in 175-cm2 flasks to 80% confluence and then infected with CMV at 2 MOI for 2 hours. After removal of the viral supernatant, one flask each was treated with 25 mL of medium with 0.5% FBS, with and without the addition of 20 mmol/L NAC. One flask served as a mock-infected control. Cells were harvested with a cell scraper 16 hours, 48 hours, or 10 days after infection. Medium and NAC were renewed at 24 hours and at 4 and 7 days after infection. Cell pellets were flash-frozen and lysed on ice in 100 μL buffer (0.01 mol/L Tris, pH 7.5, 0.144 mol/L NaCl, 0.5% Nonidet P-40, 0.5% SDS, 0.15 unit/mL aprotinin, 1 mmol/L phenylmethylsulfonyl fluoride, and 10 mmol/L iodoacetamide)1 (Bio-Rad).
CMV Infection Changes Intracellular Redox State in Coronary SMCs
We first determined whether SMCs have the capacity to generate ROIs in response to CMV infection. Intracellular ROI levels were assayed by adding to the cells 5 μmol/L DCFH-DA in HBSS (without phenol red). This nonpolar dye enters the cell, where it is deacetylated to the polar compound DCFH.14 15 16 To ascertain whether this method effectively measures ROIs in SMCs, we exposed SMCs to H2O2. Within 10 minutes after adding H2O2, a concentration-related increase in fluorescence intensity was visualized and quantified by confocal microscopy (Fig 1a through 1d⇓).
We next determined the effect of CMV infection of SMCs on ROI generation (Fig 1e through 1h⇑). ROIs increased markedly upon CMV infection, appearing as quickly as 15 to 30 minutes and reaching maximal levels 2 hours after infection (Fig 1f⇑ to 1h). Since our virus stock was prepared from supernatants of infected fibroblasts in cell culture, we prepared a virus-free supernatant for use as a control.17 Treatment of SMCs with this control medium and DCFH-DA did not induce fluorescence.
Mechanisms Responsible for CMV-Induced DCFH Fluorescence
To confirm that DCFH fluorescence following CMV infection of SMCs was caused by ROIs, SMCs were infected with CMV (5 MOI) for 2 hours. Free virus was then removed, and the antioxidant NAC, which effectively scavenges ·OH and H2O2 but not O2−,23 was added to the media immediately after removal of free virus. NAC at 1, 2.5, and 5 mmol/L caused a concentration-dependent decrease in fluorescence (Fig 1i through 1l⇑), a result confirmed by quantitative measurements of RF units (Table⇓).
Role of H2O2 in CMV-Induced ROI Generation
Human catalase, when added to coronary SMCs for 1 hour after infection, reduced ROIs, similar to the NAC results depicted in Fig 1⇑. Catalase treatment of infected cells at 5 or 35 U/mL caused a concentration-dependent reduction in RF units (Table⇑) and also in two additional experiments in which SMCs were exposed to serum-free medium before adding catalase (serum contains variable amounts of catalase). Because catalase specifically degrades H2O2 to H2O and molecular oxygen, these results indicate that ROI generation in coronary SMCs involves production of H2O2. A total of five experiments were performed.
Role of Xanthine/Xanthine Oxidase System in the Generation of CMV-Induced ROI Generation
CMV interacts with at least two host-cell membrane receptors and induces multiple signaling events before synthesis of IE proteins is detectable.24 Because SMCs of large vessels are an important source of xanthine oxidase,25 which leads to the production of O2− with subsequent generation of H2O2 and ·OH, we determined whether this enzyme system plays a role in the CMV/receptor signaling pathway. We found that Tiron, a potent metal chelator and scavenger of O2− (generated by hypoxanthine/xanthine oxidase) and ·OH,26 or oxypurinol, a specific inhibitor of xanthine oxidase,27 inhibited H2O2 generation induced by CMV infection of SMCs in a concentration-dependent manner (Table⇑). Thus, it appears that this enzyme system plays a role in the generation of ROIs in response to CMV infection of SMCs. Three separate experiments were performed with oxypurinol and with Tiron.
Effect of CMV Infection on NFκB Binding to DNA
Because CMV infection of fibroblasts activates NFκB7 and increases NFκB binding to DNA,28 we next determined whether a similar effect is induced by CMV infection of SMCs and whether such an effect is mediated by the CMV-induced generation of ROIs. As determined by gel-shift assay (see “Materials and Methods”), CMV infection of human coronary SMCs caused increased NFκB binding to a labeled oligonucleotide containing an NFκB binding element. This effect occurred as early as 1 minute after infection (Fig 2⇓, lane 2) and further increased at 30 and 60 minutes. By 2 hours after infection, binding returned to control levels.
That this CMV-induced effect is mediated by the generation of ROIs was suggested by the finding that the increased NFκB binding observed at 60 minutes was abolished by treating cells with 10 mmol/L NAC for 1 hour before infection. To rule out the possibility of a nonspecific effect of NAC on protein/DNA binding, in additional experiments we added 10 mmol/L NAC to lysates of infected cells for 1 hour before gel-shift assay. No effect was observed. Thus, CMV infection of SMCs does increase NFκB DNA binding. That this effect occurs within 1 minute of infection and is abolished by antioxidants suggests that the increased binding is caused by NFκB activation rather than synthesis of new protein, that it is mediated by a signaling event triggered by viral interaction with a cell-surface receptor, and that activation involves, at least in part, ROIs.
Effects of H2O2 on Expression of MIEP-CAT in SMCs
The MIEP of CMV contains four NFκB binding sites, perhaps as an adaptive response of the virus to facilitate viral gene expression in a cellular environment in which activation of NFκB is an intrinsic part of multiple cellular response pathways. Recently, the MIEP of murine CMV was shown to be constitutively active in bovine SMCs. This baseline activity is in part related to NFκB activity and is inhibited by antioxidants.29 Therefore, we determined whether the increased ROI levels generated by the SMC response to CMV infection enhances viral gene expression.
Rat SMCs were transfected with the major IE promoter of CMV fused to a CAT reporter gene (MIEP-CAT). When H2O2 (50 μmol/L) was added to the cells in serum-free medium for 20 hours (48 hours after transfection), CAT expression increased 10-fold. Pretreating cells with NAC inhibited this increase (Fig 3⇓). Therefore, these results demonstrate that in SMCs, ROIs transactivate the CMV MIEP, the first step in the cascade of CMV gene expression that ultimately leads to viral replication.
Effect of Antioxidant Treatment on IE72-Induced Increase in MIEP-CAT Expression
IE72 transactivates its own promoter (the MIEP), thereby leading to a positive-feedback loop.30 To determine whether this activity is redox sensitive, we assessed the effects of antioxidants on the transactivational capacity of IE72. Cotransfection of MIEP-CAT with pRSV-IE72 for 60 hours increased CAT expression in several cell types; the magnitude of the effect was cell-type specific. The increase was 10-fold in HeLa cells (data not shown), 8-fold in rat SMCs (Fig 3⇑), and 2-fold in human coronary SMCs (Fig 4⇓) and in human pulmonary artery SMCs. NAC or PDTC treatment for 8 to 10 hours (68 to 70 hours after cotransfection) inhibited the activation by IE72 (Figs 3 and 4⇑⇓). (Because of toxicity in human coronary SMCs, PDTC was used only in rat SMCs.) The possibility that this effect was due to inhibition of the RSV promoter was excluded by showing that NAC or PDTC did not inhibit RSV-CAT activity in SMCs (Fig 3⇑, bars 12 to 16; data shown are from rat SMCs). Thus, the ability of IE72 to transactivate its own promoter, presumably through activation of NFκB (see below), is at least partly ROI dependent.
Role of NFκB on IE72-Induced Increase in MIEP-CAT Expression in Coronary SMCs
To determine whether transactivation of the CMV MIEP promoter by IE72 is dependent on NFκB binding sites, as previously suggested,8 9 we cotransfected coronary SMCs with an IE72 expression plasmid and with a CAT reporter gene construct containing a minimal promoter and three NFκB sites in tandem (3XκB-CAT).13 Cotransfection increased baseline activity 2-fold compared with that measured after cotransfection with the plasmid lacking IE72. This increase was inhibited by NAC.
To confirm that the IE72-induced transactivation of the MIEP promoter occurred through NFκB activation, we tested a construct of the 3XκB-CAT plasmid in which each of the NFκB sites was mutated (3XmκB-CAT). The mutant construct abolished the transactivation capacity of IE72 in human coronary SMCs, indicating that the effect of IE72 was mediated by the NFκB sites (Fig 4⇑).
Effect of Antioxidant Treatment on Viral Titer and Viral Cytopathic Effect
Once we found that ROIs transactivated the CMV MIEP and that antioxidants inhibited MIEP transactivation by ROIs and by IE72, we determined whether such inhibition impaired the capacity of the virus to replicate and to produce cytopathic effects. Infected SMCs (2 MOI) treated with NAC exhibited a concentration-dependent decrease in viral titer (Fig 5⇓). Treatment with 20 μmol/L PDTC reduced viral titer by 50%. Cell counts (Fig 5⇓) indicated that the NAC-induced decrease in viral titer was not due to cell death. Treatment of infected coronary SMCs with 20 mmol/L NAC also decreased the number of cytopathic foci. Thus, the inhibition of CMV gene transcription induced by antioxidants impairs the capacity of the virus to replicate and to exert cytopathic effects.
Locus of Action of Antioxidants on the Cascade of Viral Gene Expression: Effect of NAC and PDTC on IE72 Expression in Coronary SMCs After CMV Infection
The previous experiment demonstrated that antioxidants inhibit the ability of CMV to replicate. To ascertain at what step of the cascade of viral gene expression antioxidants exert their inhibitory effects, we treated CMV-infected coronary SMCs with NAC, PDTC, or oxypurinol and determined, by immunohistochemistry and by immunoblotting, whether these agents blocked the expression of IE genes or whether the effects were exerted further downstream.
Coronary SMCs were stained for IE72 at 6 and 16 hours after infection, and positive nuclei were counted in three fields for a total of at least 50 to 75 cells. NAC treatment caused a concentration-dependent reduction of IE72-immunoreactive cells expressing nuclear IE72 (Fig 6⇓). Similar inhibitory effects were obtained with PDTC or with oxypurinol (Fig 6⇓, bottom).
Immunoblotting of lysates (20 μg of protein per lane) of CMV-infected cells with an IE72-specific antibody demonstrated that NAC treatment reduced steady state levels of IE72 protein at 16 hours, 48 hours, and 10 days after infection (Fig 7⇓).
These experiments demonstrate that antioxidants and a drug that inhibits an enzyme capable of generating ROIs block expression of IE72. Because IE72 is essential for the expression of the early and late genes of CMV, it is not possible to determine from these experiments whether a reduction of ROI levels also interferes with the expression of these downstream CMV genes independently of its effects on IE72 expression.
The results of the present investigation indicate that generation of ROIs is an essential component of the processes triggered by CMV infection of SMCs. Thus, SMC ROI levels increase as early as 15 to 30 minutes after infection and persist for at least 3 hours (Fig 1⇑). Our results also demonstrate that ROIs are essential to viral gene expression, viral replication, and virus-induced cytopathic effect.
Because SMCs of large vessels are an important source of xanthine oxidase,25 we focused on this enzyme system as a potential mechanism by which CMV infection might cause ROI generation. Increased xanthine oxidase activity, which leads to the production of O2− with subsequent generation of H2O2 and ·OH, has been implicated in the tissue injury occurring, for example, after experimental myocardial ischemia/reperfusion and, particularly relevant to our investigation, during influenza virus infection in mice.31 In both of these models, it was proposed that tissue injury is caused by a cellular inflammatory response in which T cells and monocytes, or neutrophils, are recruited to the site of injury and mediate release of free radicals through the adenosine deaminase–xanthine oxidase system32 ; in both models, beneficial effects followed the administration of a xanthine oxidase inhibitor.
Our results also implicate this enzyme system in the generation of ROIs in response to CMV infection but, most important, indicate that inflammatory cells are not essential to the activation of this enzyme and to the resulting generation of ROIs. We found that Tiron, thought to be a specific inhibitor of O2− (although some recent reports indicate that it also has effects on ROIs due to its potent metal chelating properties), or oxypurinol, a potent inhibitor specific for xanthine oxidase,26 27 diminished the increase in intracellular ROIs induced by CMV infection of SMCs (Table⇑), indicating that these cells have the capacity to respond to viral infection by generating free radicals through activation of cellular xanthine oxidase. We also have evidence that CMV infection of SMCs generates ROIs through the arachidonic acid cascade.33
Certain cytokines, upon binding to their cell surface receptors, activate signal transduction pathways that lead to ROI generation and then to NFκB activation.5 Because increased ROIs and NFκB activation have been associated with activation of latent HIV and enhancement of HIV gene expression,4 34 we determined whether the increase in ROIs induced by CMV infection of SMCs plays a similar role in CMV and, if so, whether such a role is subserved by ROI-induced activation of NFκB. The latter seemed possible, given the fact that CMV infection of fibroblasts had been shown to activate NFκB and increase NFκB binding to DNA.7 28
By transfecting cells with a plasmid containing the CMV MIEP upstream from the CAT reporter gene, we found that H2O2 increases MIEP-CAT transcriptional activity, an effect inhibited by the antioxidant NAC (Fig 3⇑). Moreover, gel-shift analysis revealed that NFκB binding to its DNA binding element rapidly increases after CMV infection of human coronary SMCs (Fig 2⇑), an effect mediated by the generation of ROIs, as suggested by the finding that the increased NFκB binding is abolished by NAC. These results therefore demonstrate that in SMCs, ROIs, by activating NFκB, stimulate transcription of the IE genes of CMV. Interestingly, this ROI-induced mechanism of transactivating the CMV MIEP parallels an ROI-mediated mechanism found in HIV, in that H2O2 transactivates the long-terminal repeat of HIV type 1.34
The IE gene product of CMV, IE72, is a potent transactivator of its own promoter, the MIEP. To determine the interactions between NFκB, IE72, and free radicals, we cotransfected SMCs with an IE72 expression vector and the MIEP-CAT reporter gene construct. As expected, IE72 significantly increased MIEP-CAT activity. The finding that the increased transactivation of MIEP by IE72 was inhibited by either NAC or PDTC (Figs 3 and 4⇑⇑) and the fact that this promoter has multiple NFκB binding elements are consistent with the suggestion that the ability of IE72 to augment transcription of MIEP is mediated through NFκB.7 8 9 That such a mechanism is operative was suggested by our finding that when cotransfected into coronary artery SMCs, IE72 transactivates a CAT-reporter construct containing only three NFκB binding elements, an effect abolished both by NAC and by mutation of the NFκB binding elements.
Given the data indicating an important role of ROIs in CMV gene expression, we next determined whether reducing ROIs by antioxidants inhibits the ability of the virus to replicate in SMCs and attenuates its cytopathic effects. NAC induced a concentration-dependent decrease in viral titer of infected coronary artery SMCs (Fig 5⇑), an effect also observed when cells were treated with PDTC. When cytopathic effect was assessed 96 hours after infection, we found that antioxidant treatment protected SMCs from developing cytopathic changes.
To ascertain at what step of the cascade of viral gene expression the antioxidants exert their inhibitory effects, we treated CMV-infected coronary SMCs with NAC, PDTC, or oxypurinol and determined, by immunohistochemistry and by immunoblotting, whether the antioxidants blocked the expression of IE genes or whether the effects were exerted further downstream. As might have been predicted from the fact that ROIs increase NFκB activity and that NFκB is critical for IE gene transcription, we found that NAC, PDTC, or oxypurinol inhibited IE gene expression, as reflected by the lower percentage of IE72-immunopositive SMCs 16 hours after infection (Fig 6⇑). The NAC-induced immunohistochemical results were confirmed by Western blot (Fig 7⇑). Because expression of IE gene products are ultimately necessary for the expression of all other viral gene products,35 we cannot ascertain whether antioxidants have additional inhibitory effects downstream from the IE gene products. ROIs can therefore be considered essential for viral replication and viral cytopathic effect.
We found in SMCs that NFκB is activated within 1 minute after CMV infection but that DCF fluorescence is not consistently detectable at this time (Fig 1⇑). Furthermore, although DCF fluorescence is prominent 2 hours after CMV infection, no NFκB is detected. If our concept that CMV-induced increase in ROIs contributes to the activation of NFκB, the following observations must be reconciled.
Apparent Early Disparity
We have evidence that the DCF/confocal assay for ROIs is not consistently capable of detecting low levels of ROIs, even though such levels may be of biological significance. In contrast, the gel-shift assay determining activation of NFκB is a sensitive assay, which enabled us to detect NFκB activation as early as 1 minute after viral infection. What our results therefore demonstrate is that CMV infection leads to the generation of ROIs very rapidly (detectable by 15 minutes), that the infection also activates NFκB very rapidly, and that activation of NFκB following viral infection was prevented by prior treatment with antioxidants.
The discordance at 2 hours (NFκB activation is no longer present, whereas ROIs are still present) is probably best explained by the results of two recent studies, suggesting that a compensatory negative-feedback pathway is activated very rapidly in response to induction of NFκB activation. Thus, NFκB is retained in the cytoplasm in an inactive state, bound to IκB. After cell stimulation by any of several stimuli, IκB is phosphorylated and degraded within minutes, thereby activating NFκB and causing its translocation to the nucleus. Most important, however, the cytoplasmic pool of IκB can be replenished within 20 to 90 minutes by IκB gene transcription and protein synthesis. After the reappearance of IκB, NFκB disappears from the nucleus.36 37
Although the precise mechanism for the disparity between the loss of NFκB and high H2O2 levels that we observed 2 hours after infection is not known, we propose the following to explain the disparity. Virion interaction with the host cell membrane generates ROIs, which rapidly modify cellular signaling molecules. The ROIs, probably in conjunction with other membrane-induced signals triggered by CMV, activate NFκB. The nuclear translocation of NFκB probably causes a compensatory negative-feedback loop,36 37 leading to the rapid expression of high levels of IκB and subsequent deactivation of NFκB. Although increased levels of ROIs persist 2 hours after infection, the lack of other CMV-induced membrane signals (at 2 hours, CMV has been endocytosed and probably is no longer generating membrane-dependent signals), in combination with regeneration of IκB, causes deactivation of NFκB. The interactions between an infecting virus and its targeted host cell are complex and reflect the survival mechanisms each has evolved over millions of years. One of the primary mechanisms by which cells defend themselves against viral and microbial invasion is through the production of free radicals. Although some multicellular organisms have developed specialized phagocytes that secrete large pulses of free radicals and thereby destroy virus-harboring cells,38 plants do not have specialized cells. Instead, each infected cell can rapidly generate H2O2, which triggers programmed cell death,39 thereby contributing to pathogen limitation.
This plant defense strategy is remarkably similar to the programmed cell death of eukaryotic cells (apoptosis), which is critical to many processes involved in growth and remodeling. Interestingly, certain apoptotic pathways are believed to require ROIs.40 For example, oxidative stress induces apoptosis in many cell types, and antioxidants, such as PDTC and NAC, can inhibit this process.40 On the other hand, data relating to the role of ROIs in mediating apoptosis in vascular SMCs are conflicting. Thus, Tsai et al41 showed that NAC or PDTC actually induce apoptosis in proliferating SMCs. In contrast, in our laboratory, Johnson et al,42 by overexpressing p53 in SMCs, demonstrated that the resulting p53-induced apoptosis was related to ROI generation and that NAC, PDTC, or catalase inhibited both ROI production and apoptosis. Therefore, it appears that apoptosis occurring in mammalian cells can be mediated by ROI-dependent pathways that can be blocked by antioxidants but that certain cellular conditions may activate ROI-inhibitable apoptosis, causing cell death to be induced by antioxidants.
Although animal species have the specialized inflammatory cell defense directed against pathogen invasion noted above, indirect evidence suggested,43 and now we have shown, that ROIs can also be activated in noninflammatory cells in response to viral infection. In addition to contributing to the induction of apoptosis in certain cellular contexts and thereby preventing the infecting virus from replicating and infecting neighboring cells, a cellular program that generates ROIs in response to viral infection would also contribute to the activation of NFκB, which mediates expression of many cellular genes, including those involved in the immune and inflammatory responses.4 If this formulation is correct, then viruses seem to have adapted to the cellular defense mechanism involving the generation of ROIs and resulting activation of NFκB by incorporating NFκB sites into their promoter that directs expression of viral IE genes, thereby achieving the capacity of using this cellular mechanism to activate their own genetic programs.
Conclusions and Implications
It appears that CMV has evolved to the point that it can successfully use one of the primary signal transduction pathways used by the cell for multiple purposes (the generation of ROIs) to facilitate expression of its own genetic program. Moreover, since CMV DNA is present in the wall of atherosclerotic vessels and in restenosis lesions,1 2 since ROIs are generated in response to injury,44 and since early atherosclerotic changes in the vessel wall impair the capacity to scavenge free radicals,45 our finding that ROIs are conducive to CMV gene expression and to viral replication provides additional evidence compatible with the hypothesis that CMV plays a contributory role in the development of restenosis and of atherosclerosis. If this hypothesis proves correct, the results of this investigation may provide new strategies for inhibiting the contribution of CMV to these disease processes.
Selected Abbreviations and Acronyms
|IκB||=||inhibitor (protein) κB|
|MIEP||=||major IE promoter|
|MOI||=||multiplicity of infection|
|NFκB||=||nuclear factor κB|
|ROI||=||reactive oxygen intermediate|
|SMC||=||smooth muscle cell|
- Received June 4, 1996.
- Accepted September 26, 1996.
Speir E, Modali R, Huang ES, Leon MB, Shawl F, Finkel T, Epstein SE. Potential role of human cytomegalovirus and p53 interaction in coronary restenosis. Science. 1994;265:391-394. Comments.
Gao JL, Murphy PM. Human cytomegalovirus open reading frame US28 encodes a functional beta chemokine receptor. J Biol Chem.. 1994;269:28539-28542.
Kowalik TF, Wing B, Haskill JS, Azizkhan JC, Baldwin AS Jr, Huang ES. Multiple mechanisms are implicated in the regulation of NF-kappa B activity during human cytomegalovirus infection. Proc Natl Acad Sci U S A.. 1993;90:1107-1111.
Cherrington JM, Mocarski ES. Human cytomegalovirus ie1 transactivates the alpha promoter-enhancer via an 18-base-pair repeat element. J Virol.. 1989;63:1435-1440.
Huang ES. Human cytomegalovirus, III: virus-induced DNA polymerase. J Virol.. 1975;16:298-310.
Ghazal P, Young J, Giulietti E, DeMattei C, Garcia J, Gaynor R, Stenberg RM, Nelson JA. A discrete cis element in the human immunodeficiency virus long terminal repeat mediates synergistic trans activation by cytomegalovirus immediate-early proteins. J Virol.. 1991;65:6735-6742.
Finco TS, Baldwin AS Jr. Kappa B site-dependent induction of gene expression by diverse inducers of nuclear factor kappa B requires Raf-1. J Biol Chem.. 1993;268:17676-17679.
Bass DA, Parce JW, Dechatelet LR, Szejda P, Seeds MC, Thomas M. Flow cytometric studies of oxidative product formation by neutrophils: a graded response to membrane stimulation. J Immunol.. 1983;130:1910-1917.
Ohba M, Shibanuma M, Kuroki T, Nose K. Production of hydrogen peroxide by transforming growth factor-beta 1 and its involvement in induction of egr-1 in mouse osteoblastic cells. J Cell Biol.. 1994;126:1079-1088.
Alcami J, Barzu T, Michelson S. Induction of an endothelial cell growth factor by human cytomegalovirus infection of fibroblasts. J Gen Virol.. 1991;72:2765-2770.
Staal FJ, Roederer M, Herzenberg LA. Intracellular thiols regulate activation of nuclear factor kappa B and transcription of human immunodeficiency virus. Proc Natl Acad Sci U S A.. 1990;87:9943-9947.
Balew HC, Forrester FT, Lyerk HC, Velleca WM, Bird BR, Roberts JD. Basic Laboratory Methods in Virology. Atlanta, Ga: US Department of Health, Education, and Welfare; Center for Disease Control, Virology Branch; 1975.
Speir E, Epstein SE. Inhibition of smooth muscle cell proliferation by antisense oligodeoxynucleotide targeting the messenger RNA encoding proliferating cell nuclear antigen. Circulation. 1992;86:538-547.
Albrecht T, Boldogh I, Fons MP, Valyi-Nagy T. Activation of proto-oncogenes and cell activation signals in the initiation and progression of human cytomegalovirus infection. In: Becker Y, Darai G, Huang ES, eds. Frontiers of Virology 2. New York, NY: Springer-Verlag; 1993:384-411.
Yurochko AD, Kowalik TF, Huong SM, Huang ES. Human cytomegalovirus upregulates NF-kappa B activity by transactivating the NF-kappa B p105/p50 and p65 promoters. J Virol.. 1995;69:5391-5400.
Bellas RE, Lee JS, Sonenshein GE. Expression of a constitutive NF-kappa B-like activity is essential for proliferation of cultured bovine vascular smooth muscle cells. J Clin Invest.. 1995;96:2521-2527.
Stenberg RM, Fortney J, Barlow SW, Magrane BP, Nelson JA, Ghazal P. Promoter-specific trans activation and repression by human cytomegalovirus immediate-early proteins involves common and unique protein domains. J Virol.. 1990;64:1556-1565.
Akaike T, Ando M, Oda T, Doi T, Ijiri S, Araki S, Maeda H. Dependence on O2- generation by xanthine oxidase of pathogenesis of influenza virus infection in mice. J Clin Invest.. 1990;85:739-745.
Grisham MB, Hernandez LA, Granger DN. Xanthine oxidase and neutrophil infiltration in intestinal ischemia. Am J Physiol.. 1986;251:G567-G574.
Speir E, Shibutani T, Yu ZX, Epstein SE. Aspirin, by inhibiting free radical generation induced by cytomegalovirus infection of human smooth muscle cells, inhibits cytomegalovirus gene expression and cytomegalovirus replication. Circulation. 1995;92(suppl I):I-229. Abstract.
Wathen MW, Stinski MF. Temporal patterns of human cytomegalovirus transcription: mapping the viral RNAs synthesized at immediate early, early, and late times after infection. J Virol.. 1982;41:462-477.
Brown K, Park S, Kanno T, Franzoso G, Siebenlist U. Mutual regulation of the transcriptional activator NFκB and its inhibitor, IκB-α. Proc Natl Acad Sci U S A.. 1993;90:2532-2536.
Arenzana-Seisdedos F, Thompson J, Rodriguez MS, Bachelerie F, Thomas D, Hay R. Inducible nuclear expression of newly synthesized IκB-α negatively regulates DNA-binding and transcriptional activities of NFκB. Mol Cell Biol.. 1995;15:2689-2696.
Segal AW. The electron transport chain of the microbicidal oxidase of phagocytic cells and its involvement in the molecular pathology of chronic granulomatous disease. J Clin Invest.. 1989;83:1785-1793.
Tsai JC, Jain M, Hsieh CM, Lee WS, Yoshizumi M, Patterson C, Perella MA, Cooke C, Wang H, Haber E, Schlegel R, Lee ME. Induction of apoptosis by pyrrolidinedithiocarbamate and N-acetylcysteine in vascular smooth muscle cells. J Biol Chem.. 1996;271:3667-3670.
Johnson TM, Yu ZX, Ferrans VJ, Lowenstein RA, Finkel T. Reactive oxygen species are downstream mediators of p53-dependent apoptosis. Proc Natl Acad Sci U S A.. 1996;93:11848-11852.
Ohara Y, Peterson TE, Harrison DG. Hypercholesterolemia increases endothelial superoxide anion production. J Clin Invest.. 1993;91:2546-2551.
Baenerle PA, Baltimore D. A 65-kD subunit of active NFκB is required for inhibition of NFκB by IκB. Genes Dev. 1989;3:1689-1698.