Monocytes Harboring Cytomegalovirus: Interactions With Endothelial Cells, Smooth Muscle Cells, and Oxidized Low-Density Lipoprotein
Possible Mechanisms for Activating Virus Delivered by Monocytes to Sites of Vascular Injury
Abstract Cytomegalovirus (CMV) infection and its periodic reactivation from latency may contribute to atherogenesis and restenosis. It is unknown how CMV is delivered to the vessel wall and is reactivated. We examined the following hypothesis: CMV, present in monocytes recruited to sites of vascular injury, is activated by endothelial cell (EC) or smooth muscle cell (SMC) contact and by oxidized low-density lipoproteins (oxLDLs). The CMV major immediate-early promoter (MIEP) controls immediate-early (IE) gene expression, and thereby viral replication. To determine whether elements of the vessel wall can activate CMV present in monocytes, we transiently transfected the promonocytic cell line HL-60 with a chloramphenicol acetyltransferase reporter gene construct driven by MIEP. MIEP activity increased 1.7±0.5-fold (P=.02) when the transfected HL-60 cells were cocultured with ECs, 4.5±1.5-fold when cocultured with SMCs (P=.03), and 2.0±0.5-fold (P=.01) when exposed to oxLDL. The combination of oxLDL and EC coculture increased MIEP activity over 7-fold. We also found that freshly isolated human monocytes, infected with endothelium-passaged CMV, were capable of transmitting infectious virus to cocultured ECs or SMCs. CMV-related progression of atherosclerosis or restenosis may, at least in part, involve monocyte delivery of the virus to the site of vascular injury, where the vascular milieu, ie, contact with ECs, SMCs, and oxLDL, can contribute to viral reactivation and/or replication by enhancing CMV IE gene expression. The virus may then infect neighboring ECs or SMCs, initiating a cascade of events predisposing to the development of atherogenesis-related processes.
- immediate-early protein
- chloramphenicol acetyltransferase assay
- viral gene expression regulation
Cytomegalovirus is a ubiquitous virus, as evidenced by the fact that the majority of adults throughout the world are seropositive for CMV (reviewed in Reference 11 ). Like other herpesviruses, CMV is retained in a latent state for life.2 Although CMV causes serious disease in immunocompromised subjects (such as neonates, patients with AIDS, or transplant recipients), with the exception of a mononucleosis-like syndrome, it is generally believed that CMV does not cause disease in healthy immunocompetent individuals. This may, however, be incorrect. In particular, an increasing body of information suggests that CMV may participate in the development of atherosclerosis (reviewed in Reference 33 ) and in the development of restenosis following coronary angioplasty.4
If CMV does play such a role, it would be critical to elucidate how CMV is delivered to the site of the vascular lesion and how the virus is reactivated from latency. CMV may reside in the vessel wall in a latent state or a state in which it replicates at a low (and possibly intermittent) but clinically unrecognizable level. An alternative possibility suggested by studies published in the last few years5 6 focuses on the monocyte. Circulating monocytes have been shown by PCR analysis to harbor CMV DNA.5 CMV does not easily infect monocytes, and monocytes circulate in the blood only briefly. However, CMV can infect monocyte precursors present in bone marrow, and the viral genome persists in these cells.7 8 9 This observation, when considered with the fact that clinically important CMV infection develops in bone marrow transplant recipients seronegative for CMV but whose donors are CMV seropositive,10 suggests that myelomonocytic precursor cells may act as a reservoir of CMV and that circulating monocytes may act as a vector, delivering the virus to sites of vascular inflammation or injury.
These studies have also demonstrated that circulating monocytes are nonpermissive for CMV; despite the presence of viral genome, viral gene products are not expressed in such cells (ie, the virus is in a latent state). Expression of IE viral gene products occurs only after differentiation of the monocytes to form macrophages. This has been shown for monocytes derived from seropositive normal subjects and for monocytes infected with CMV under cell culture conditions.6 11 Although the extent of viral activity in macrophages varies and is dependent on the virus strain and methods of differentiation used, it appears that the more differentiated the monocyte, the more permissive it is to viral gene expression. The present study tested the possibility that cells of the vessel wall—ECs and SMCs—might enhance activation of CMV carried in monocytes (ie, act as “differentiating” agents) and thereby facilitate interactions between the virus and vessel wall that might predispose to either atherosclerosis or restenosis. Because oxLDL has been linked to atherogenesis and because it is present in the vascular subintimal space, we also studied the effects of oxLDL on CMV activation.
The strategy we used was based on the fact that the CMV MIEP controls expression of the IE gene products of the virus, which are critical requirements for expression of the early and late gene products of the virus and thereby for progression of the viral life cycle in the host cell.12 Given the important control function of the CMV MIEP, it is also thought that activation of MIEP may be a key step in the reactivation of CMV from latency.13 Therefore, we transfected monocyte-related cells with an MIEP–reporter gene plasmid and determined the effect of cocultured SMCs, ECs, and oxLDL on MIEP activity. The following hypothesis was investigated: CMV, present in monocytes recruited to sites of vascular injury, is activated by EC or SMC contact and by oxLDLs.
Materials and Methods
The promyelocytic cell line HL-60 (American Type Culture Collection) was used for transfection experiments. These cells are closely related to early progenitor cells and can be differentiated into granulocytes or monocytes/macrophages.14 These cells have been shown by flow cytometry to express low levels of antigens typically associated with monocytes.14 The HL-60 cells were cultured in Iscove’s medium (GIBCO) containing penicillin-streptomycin (Biofluids) supplemented either with 20% FBS or, in experiments in which cells were incubated with oxLDL, with GMSX (a supplement containing insulin, transferrin, and selenium) (both from GIBCO). All cells in the present study were incubated at 37°C with 5% CO2.*
Freshly isolated monocytes were provided by the Department of Transfusion Medicine, Clinical Center, National Institutes of Health: peripheral blood of normal donors was used to obtain cell suspensions containing mean monocyte proportions of at least 85% obtained by counterflow centrifugal elutriation.15 The presence of CMV antibody was determined by enzyme-linked immunosorbent assay (Becton Dickinson). Monocytes were cultured under the same conditions as HL-60 cells and were differentiated in the experiments indicated, according to the method described by Taylor-Wiedeman et al,6 with 5×10−5 mol/L hydrocortisone (Sigma Chemical Co) for 12 to 24 hours, followed by 5 ng/mL of GM-CSF (Calbiochemicals) for an additional 9 days.
Human umbilical vein ECs and pulmonary artery SMCs, used at passages 2 to 7, were cultured in EGM and SMGM 2, respectively, and supplemented with 10% FBS (Clonetics).
Human foreskin fibroblasts, a gift of P.E. DiCorleto, Cleveland (Ohio) Clinic Foundation, were cultured in DMEM/F-12 medium supplemented with FBS (Biowhitaker).
We investigated the capacity of ECs, SMCs, and oxLDL to stimulate transcription of a CAT reporter construct containing the CMV MIEP and enhancer. The plasmid pHD101CAT3, containing a 2.1-kb region of the human CMV MIEP fused to the CAT reporter gene, was a gift of E.S. Huang, University of North Carolina, Chapel Hill.16
Cells were transfected with DEAE-dextran, according to the method described by Stein et al.14 Briefly, 5 g of plasmid DNA was added to 1 mL of the following buffer (mmol/L): Tris 25 (pH 7.4), NaCl 137, KCl 5, CaCl2 0.7, MgCl2 0.5, and Na2HPO4 0.6, along with 5 mg/mL DEAE-dextran (Pharmacia Biotech, Inc). This DNA preparation was used to resuspend 5×106 HL-60 cells. The cells were then incubated at 37°C in the transfection mixture for 1 hour, collected by centrifugation, resuspended in growth medium at a density of 2.5×106 cells/mL, and incubated at 37°C.
Coculture of Transfected HL-60 Cells With ECs and SMCs
HL-60 cells were collected by centrifugation 18 to 20 hours after transfection and resuspended in EGM or SMGM 2 for coculture with ECs or SMCs, respectively. Each sample of HL-60 cells was cocultured with ECs or SMCs in one well of a six-well plate for 48 hours and then harvested as described below. ECs and SMCs were seeded 3 days in advance in six-well plates (Falcon, Becton Dickinson Labware) at a density of 2.5×105 cells per well. Control samples of noncocultured transfected HL-60 cells were treated in the same manner. As a positive control, PMA (Sigma), prepared in dimethyl sulfoxide, was added at a final concentration of 0.01 μg/mL 24 hours before harvest of the cells for CAT assay.
OxLDL Preparation and Treatment of Cells
Native LDL was isolated from plasma by sequential ultracentrifugation.17 OxLDL was prepared by incubation with CuCl2, as described by Quinn et al.18 Minimally modified LDL was prepared by incubation with FeSO4 as described by Liao et al.19 Extent of oxidation was measured by thiobarbituric acid reactive substance assay.
HL-60 cells were transfected and incubated in HL-60 growth medium for 48 hours. Cells (5×106 per sample) were collected by centrifugation and resuspended in 1.5 mL Iscove’s medium containing 200 μg/mL of oxLDL, native LDL (negative control), or PMA (positive control). Cells were incubated for an additional 24 hours, after which 1 mL of growth medium was added to each sample. After a 4-hour incubation, cells were harvested for CAT assay.
Cell Harvest and CAT Assay
Cells, collected by scraping, were transferred with medium into centrifuge tubes (Corning); residual cells were removed by PBS wash and added to the tubes. A fraction of the HL-60 cells adhered to the underlying cells; therefore, both HL-60 and the underlying monolayer of ECs or SMCs were collected. As a control for these samples, to maintain a common ratio of HL-60 cells to ECs and to SMCs at analysis but without the effects of coculture, ECs or SMCs, plated as described, were added to noncocultured HL-60 cells at the time of harvest. Cells were washed three times in 4 mL PBS and resuspended in 1 mL of the following buffer: 0.04 mol/L Tris-HCl (pH 7.4), 1 mmol/L EDTA, and 0.15 mol/L NaCl; they were then transferred to 1.5-mL microcentrifuge tubes. Cells were collected by centrifugation at 3000 rpm for 2 minutes in a microcentrifuge and then resuspended in 100 μL of 0.25 mol/L Tris-HCl (pH 8). After three freeze-thaw cycles, consisting of 3-minute incubations each in a dry ice–ethanol bath, and incubation at 37°C, cell lysates were clarified by centrifugation at 14 000 rpm for 5 minutes at 4°C. Supernatants were removed, assayed for protein content (BCA kit, Pierce Chemicals), flash-frozen in dry ice–ethanol, and stored at −70°C.
CAT activity was measured according to the previously described method.20 Briefly, reaction mixtures containing equal amounts of protein from the cell lysates, 10 μL of 1 mol/L Tris-HCl (pH 8), 4 μL of 14C-labeled chloramphenicol (CAT assay grade, American Radiolabelled Chemicals, Inc), and 5 μL of 5 nmol/L n-butyryl coenzyme A (Sigma), in a total volume of 100 μL, were incubated for 2.5 hours at 37°C. The butyrylated product was quantified by the phase-extraction method.21
Virus Stock Propagation
The endothelium-passaged strain of CMV, VHL/E, a gift of J. Waldman, Ohio State University, Columbus, was propagated in human umbilical vein ECs as described previously.22 23 Virus stock was harvested from heavily infected ECs, which were lysed by repeated freeze-thaw cycles and stored in sucrose buffer at −70°C, as described by Waldman and colleagues.22 23 This virus suspension was diluted in serum-free medium to obtain the multiplicities of infection indicated.
Infection of Monocytes With CMV and Coculture With ECs and SMCs
Monocytes were infected at a multiplicity of infection of 0.01 plaque forming units per cell. Freshly isolated monocytes were counted, collected by centrifugation, and resuspended in the virus preparation described above. At various times after infection, the monocytes were washed, resuspended in EGM, and cocultured with ECs or resuspended in SMGM 2 and cocultured with SMCs. Cultures were monitored for development of CMV-induced CPEs in the underlying ECs or SMCs.
Detection of Viral Gene Transcription in Monocytes by RT-PCR
RNA was prepared from both differentiated and nondifferentiated monocytes 5 days after infection or after mock infection, using Trisol (Life Technologies) according to the manufacturer’s recommendation. Total RNA was used as a template for RT of first-strand cDNA synthesis using the SuperScript Preamplification System (Life Technologies).
Amplification of target cDNA was carried out with primers and probes for nested-intron spanning for CMV IE72, IE84, major early, and late pp28 genes, as described by Taylor-Wiedeman et al.6 The identity of the respective PCR-amplified products was confirmed by Southern transfer of samples in 1% agarose gels to nitrocellulose filters and hybridization with specific oligonucleotide probes as described previously.6
Monocytes were plated in eight-chamber slides (Lab Tek, Nunc Inc) at a density of 1×105 cells/cm2. Cells were differentiated in hydrocortisone and GM-CSF or left untreated (control). Cells were then either infected or not infected with CMV. Slides were washed in PBS, fixed in 50% methanol and 50% acetic acid for 5 minutes, and rehydrated with PBS. A Vectastain Elite ABC kit (Vector Laboratories) was used for indirect horseradish peroxidase immunostaining. Incubations and washes were carried out according to manufacturers’ instructions. Monoclonal antibodies directed against the CMV IE genes, IE72 or IE84 (Vancouver Biotech), were diluted in 0.1% crystalline BSA (Sigma) in PBS to a final concentration of 1 μg/mL. Peroxidase-positive nuclei were detected with a diaminobenzidine substrate reagent set (Kirkegaard and Perry Laboratories, Ltd).
Comparisons among transfection samples were analyzed by a paired two-tailed t test.
To test our hypothesis, two lines of investigation were undertaken: the first was to document the impact of components of the vessel wall on CMV gene activity in monocyte-related cells; the second, to investigate the effect of these vessel wall components on the ability of initially nonpermissive monocytes to transmit infection to ECs and SMCs.
In the first part of the study, the promyelocytic cell line, HL-60, was used. These progenitor cells can be differentiated into macrophages and can therefore be considered monocyte/macrophage precursors.14 IE CMV gene activity was assayed by transfecting HL-60 cells with a reporter-plasmid construct encoding the CMV MIEP. Stein et al14 previously demonstrated that baseline MIEP activity is low in HL-60 cells and that this activity can be enhanced by treatment of the cells with the phorbol ester PMA.14 We confirmed that in our hands PMA-transactivated MIEP was transfected into HL-60 cells (data not shown). Therefore, we chose this system to test the effects of oxLDL and coculture with ECs and SMCs on the baseline activity of MIEP in HL-60 cells, using PMA and native LDL as positive and negative controls, respectively.
MIEP Activity in HL-60 Cells in Response to Coculture With ECs and SMCs
In six separate experiments performed in duplicate or triplicate, coculture of MIEP-CAT–transfected HL-60 cells with ECs increased MIEP-CAT activity 1.7±0.5-fold (P=.02) over baseline activity in noncocultured HL-60 cells (Fig 1A⇓). Coculture of MIEP-CAT–transfected HL-60 cells with SMCs increased MIEP-CAT activity 4.5±1.5-fold (n=4, P=.03) over baseline activity in noncocultured HL-60 cells (Fig 1B⇓).
In order to determine if this effect is specific to vascular cells, the effect of coculture of MIEP-transfected HL-60 cells with primary human foreskin fibro-blasts was tested. Coculture of transfected HL-60 cells with fibroblasts resulted in an increase in MIEP activity similar to the increase observed in samples of transfected HL-60 cells that were cocultured with vascular cells (data not shown).
Effect of Conditioned Medium From SMCs on MIEP Activity in HL-60 Cells
To determine whether the increase in MIEP activity induced by SMC coculture was dependent on direct contact between the SMCs and HL-60 cells or could be observed with the SMC-conditioned medium alone, MIEP-transfected HL-60 cells were exposed to the following conditions, and the effects on MIEP-CAT activity were compared: (1) culture in fresh growth medium, (2) culture in conditioned medium from SMCs, (3) culture in conditioned medium from SMC/HL-60 cell coculture, and (4) coculture with SMCs. The highest activity, a 3.9-fold increase, was observed in samples obtained from HL-60 cells directly cocultured with SMCs. The conditioned medium of SMCs and conditioned medium from SMC/HL-60 cocultures led to smaller increases in MIEP activity of 2.7- and 2.9-fold, respectively.
Effect of OxLDL on MIEP Activity in HL-60 Cells
HL-60 cells were treated with oxLDL at concentrations of 50 to 400 μg/mL 48 hours after transfection and 24 hours before harvest. Treatment of cells with 400 μg/mL oxLDL had a toxic effect. Concentrations of 50 to 200 μg/mL oxLDL activated MIEP, in a concentration-dependent manner (Fig 2⇓). The greatest effect was found in cells treated with 200 μg/mL, where a mean increase of 2.0±0.5-fold (P=.01) in CAT activity was detected in six additional experiments that were carried out in duplicate or triplicate (Fig 1C⇑). A similar stimulatory effect was found with minimally modified LDL (data not shown). Exposure of transfected HL-60 cells to native LDL did not significantly increase MIEP activity, averaging 1.2±0.3 times baseline activity.
MIEP Activity in Response to Coculture With ECs Plus OxLDL
HL-60 cells, transfected with MIEP-CAT plasmid, were cocultured with ECs, and oxLDL was added. This combination led to a further increase in MIEP activity over each individual intervention, with a 7.1-fold increase over baseline (Fig 3⇓).
The above results demonstrate that MIEP activity in the HL-60 precursor cells can be enhanced by contact of these cells with ECs, SMCs, and oxLDL. These in vitro results suggest mechanisms that may occur in vivo to CMV-harboring monocytes during the inflammatory response to vascular injury. Numerous studies have shown that CMV activity in monocytes increases with monocyte differentiation. In a study by Ibanez et al,11 CMV-infected monocytes, undergoing differentiation when exposed to T lymphocytes, were able to support CMV replication, whereas infected untreated cells were not. In this regard, monocytes, recruited to the site of vascular injury, undergo differentiation to macrophages upon entry into the subintimal space of the vascular wall. Our results suggest that if such monocytes harbor latent CMV, these conditions—differentiating monocytes exposed to ECs, SMCs, and oxLDL—favor the transactivation of the MIEP of CMV, thereby leading to viral gene expression.
Therefore, in the next part of the study, we wanted first to characterize the ability of freshly isolated monocytes, either nondifferentiated or differentiated, to support viral gene expression when infected with the VHL/E strain of CMV and then to examine the ability of infected monocytes to transmit CMV infection to ECs and to SMCs during coculture.
Analysis of CMV Transcriptional Activity in Infected Monocytes
RT-PCR analysis of RNA extracted from infected nondifferentiated monocytes (5 days after infection) was performed to determine whether viral transcripts were expressed in infected monocytes. RT-PCR amplification products representing transcripts of IE72, IE84, early, and late genes were detected as appropriate band sizes on gel electrophoresis and verified by Southern blotting and hybridization with CMV-specific probes (Fig 4⇓).
Expression of CMV IE Gene Products
CMV IE gene expression was characterized in monocytes infected with the VHL/E strain of CMV (EC-passaged) by immunostaining cells fixed at 2, 3, 5, and 7 days after infection for IE72 and IE84. No IE gene products in infected cells could be detected in nondifferentiated monocytes (Fig 5B⇓). However, nuclear expression of the viral IE genes IE72 (Fig 5F⇓) and IE84 (data not shown) was found in monocytes differentiated by exposure to hydrocortisone and GM-CSF. IE gene expression was detected in 10% of the cells at days 2 to 7, with peak expression at day 5 after infection (Fig 5F⇓).
We next examined whether infected nondifferentiated monocytes could transmit infectious virus to CMV-naive cocultured ECs and SMCs and whether, under these conditions, the virus could produce CPE in the cocultured cells.
Coculture of Infected Monocytes With ECs and SMCs
The coculture of infected undifferentiated monocytes (5 to 14 days after infection) with ECs caused foci of ECs with CPE 7 to 14 days after coculture initiation (Fig 6A⇓). No CPE foci were detected in cocultures of noninfected monocytes and ECs or in ECs cultured alone (Fig 6B⇓). Thus, infected monocytes are capable of transmitting virus to cocultured ECs and SMCs. However, when infected monocyte cultures were maintained for 1 month and then cocultured with ECs, no CPE foci were detected.
The lysate of infected monocytes, produced by repetitive freeze-thaw cycles before coculturing with ECs, also led to the appearance of foci of CPE in ECs, although the number of such foci was reduced. This demonstrates that nonviable infected monocytes are also capable of transmitting virus to ECs, although at a decreased level compared with viable infected monocytes. However, cell association of the virus is necessary for infection of the cocultured ECs, as evidenced by the fact that when the same virus inoculum is maintained in culture under the same conditions and time intervals used for infecting monocytes and subsequently transferred to ECs, no CPE develops in the latter. Thus, it appears that in the cellular environment of the monocyte, the infecting virus remains in an infectious state without undergoing replication, as suggested by our inability to detect IE proteins under these conditions. Alternatively, it may undergo low-level replication, as suggested by the fact that mRNAs of IE, early, and late genes were identified by RT-PCR. In either case, it would appear likely that with coculture, viral gene expression in the monocyte is enhanced, thereby augmenting its infectious capacity.
The transmission of virus resulting in CPE foci in underlying SMCs was similar to that found with the EC coculture experiments (data not shown).
Coculture of Monocytes From CMV-Immunopositive Subjects With ECs
Monocytes from immunopositive subjects were cocultured with ECs and SMCs, in an attempt to reactivate latent CMV in the monocytes. The cocultured samples were maintained in culture and monitored for the development of CPE in the underlying vascular cells for 1 month. No CPE foci were observed.
In the present investigation, we determined whether biologically relevant vascular mechanisms exist that are compatible with the concept that the circulating monocyte can act as a vector for CMV, targeting virus delivery to arterial lesions, where the virus is then activated by components of the vessel wall, thereby contributing to CMV-induced atherogenic-related processes.
Our results demonstrate that there are in fact interactions between cellular components of the vessel wall and monocyte-related cells that can lead to CMV activation. Thus, a critical requirement for progression of the intracellular viral life cycle is the expression of the IE gene products of CMV, which are controlled by the major IE promoter of the virus.12 To determine whether vessel wall components can activate the CMV MIEP within monocytes, we transfected HL-60 cells, a monocyte-related cell line, with an MIEP reporter plasmid and determined the effect on MIEP activity of coculturing the transfected HL-60 cells with either SMCs or ECs or of exposing the transfected HL-60 cells to oxLDL.
As expected, and in accordance with the study reported by Stein et al,14 MIEP activity was low under control conditions. However, coculture of the MIEP-transfected HL-60 cells with ECs or SMCs consistently and significantly increased MIEP activity, by ≈2-fold in ECs and by 4.5-fold in SMCs (Figs 1⇑ and 3⇑). MIEP activity also increased 2-fold when the MIEP-transfected HL-60 cells were exposed to oxLDL. Moreover, the combination of oxLDL exposure and coculture of HL-60 cells with ECs caused an even greater increase in MIEP activity, averaging >7-fold. The effect of coculture on MIEP activity is not limited to vascular cells, since coculture of transfected HL-60 cells with human foreskin fibroblasts also resulted in an increase in MIEP activity.
We next determined whether infected undifferentiated monocytes would be capable of transmitting virus to either ECs or SMCs or whether the virus would be entirely cell-associated and incapable of infecting neighboring uninfected cells. In this regard, Waldman et al24 recently reported that CMV-infected ECs were capable of transmitting the virus to cocultured monocytes. They also found that these monocytes could transmit the virus to uninfected ECs. Their study therefore established a paradigm for CMV-induced transplant rejection, in which immune responses are believed to be targeted largely to ECs—the concept being that CMV-infected ECs elicit a greater immune response than do uninfected ECs. The results of the present study extend their findings and conclusions to a paradigm more related to injury-induced restenosis or atherosclerosis, in which local vascular injury evokes an inflammatory response, resulting in delivery of the virus to the site of vascular injury by primarily infected monocytes. Once there, the virus is activated by interactions between the monocyte vector and constituents of the vessel wall, resulting in transmission of infection to ECs and SMCs.
Thus, we found that when monocytes are primarily infected by CMV, monocyte-EC and monocyte-SMC transmission does occur. Five to 14 days after fresh human undifferentiated monocytes were infected with CMV, coculture with either ECs or SMCs led to the transmission of virus and development of CPE in the target vascular wall cells. The source of the transmitted viral particles could have been either residual inoculum or the result of a very low level of virus production in the nondifferentiated monocytes. The latter may be at least partly responsible for the presence of infectious virus in the monocyte, because mRNA of IE, early, and late viral genes was detectable in the monocytes by RT-PCR (Fig 4⇑). However, viral proteins could not be detected by the less sensitive technique of immunostaining (Fig 5⇑). The absence of IE proteins in the nondifferentiated monocytes, determined by immunostaining, versus the positive RT-PCR results for IE gene expression (as well as early and late CMV gene expression) is most likely the result of the greater sensitivity of detecting mRNA by RT-PCR versus detecting protein by immunostaining. If there was a posttranscriptional block so that no IE proteins were expressed, we would not have expected transcription of early and late viral genes, which are dependent on the expression of IE proteins.
Thus, coculture of abortively infected monocytes, expressing viral transcripts but little or no viral protein, with ECs and SMCs led to an increase in viral activity in the monocytes and, ultimately, to the transmission of infectious virus to the underlying vascular cells. We also attempted to establish a true latency model (as rigorously defined by the expression of neither viral proteins nor gene transcripts) in vitro by infecting the monocytes and maintaining them in culture for longer intervals. We found, however, that at longer intervals after infection (up to 1 month), the virus could not be reactivated, as evidenced by the absence of CPE development in cocultured ECs.
In an additional attempt to reactivate latent virus in vitro, we cocultured ECs with uninfected monocytes from seropositive donors (represented by ≈50% of the samples we tested). This did not lead to CPE in the cocultured ECs. We also failed to detect IE proteins by immunostaining after chemical differentiation of monocytes from seropositive donors. This result is in agreement with the report of Taylor-Wiedeman and colleagues,6 in which unsuccessful attempts to recover infectious CMV from monocytes of donors by coculture of the undifferentiated monocytes with permissive fibroblasts were described. The reason for the difficulty in reactivating truly latent CMV in vitro is unknown. It may be due to the low number of cells that harbor the CMV genome or to some other inhibitory state that requires the interaction of a complex array of factors only found under certain circumstances in vivo.
It is important to note that cells respond differently to different laboratory strains of CMV: cells that may be permissive to one strain may be nonpermissive to another. To infect monocytes, we used an EC-passaged strain of CMV, developed by Waldman and colleagues,22 23 to which the EC is permissive. It appears that ECs are permissive to recently isolated clinical strains of CMV, which was the reason we used this particular CMV strain.
There are several potential mechanisms that could lead to the EC/SMC-induced activation of CMV MIEP we observed. In vivo, monocytes are recruited to the vessel wall as part of the inflammatory response to injury. The cascade of events leading to monocyte recruitment is complex, involving the expression of many gene products, which include adhesion molecules, chemokines, and cytokines.25 Upon adhesion and entry to the subintimal space, monocytes differentiate into macrophages. Therefore, it is possible that in our coculture system, interaction between the HL-60 and vascular wall cells causes expression of differentiation-dependent factors, which, in turn, cause CMV activation. We also determined that the activation is in part related to direct cell-to-cell contact and in part to some secreted factor(s). Thus, the greatest MIEP activation was observed with direct coculture. However, the conditioned medium of SMCs and conditioned medium from SMC/HL-60 cocultures also increased MIEP activity.
The activation of CMV MIEP by oxLDL might be mediated by cellular transcription factors activated by oxLDL. Several groups have shown that the binding activity of the cellular transcription factor NF-κB is increased as a result of oxLDL treatment of unstimulated vascular SMCs, ECs, fibroblasts, and macrophages.26 27 28 Minimally modified LDL also induces this activity in ECs.29 This is of relevance, since CMV MIEP contains four known NF-κB binding sites and NF-κB is a potent activator of the MIEP.1 30 However, it has also been shown that oxLDL exerts an inhibitory effect on lipopolysaccharide-induced NF-κB binding activity.26 31 32 Thus, a possible role for this transcription factor in our system needs to be investigated further. Additional transcription factors may be involved in the activation of MIEP by oxLDL; eg, the DNA binding activity of the transcription factor AP-1 has been shown to be induced by oxLDL in SMCs.31
Since the CMV MIEP enhancer region contains response elements for AP-1, the activation of this transcription factor may be involved in the induction of MIEP activity by oxLDL in our system.30 However, it should be noted that AP-1 activity was also shown to be negatively regulated by oxLDL in macrophages.32
The component of oxLDL that causes activation of the CMV MIEP has not been elucidated in the present study. Several groups have shown that lysophosphatidylcholine, an oxidation product of oxLDL, causes changes in gene expression that may contribute to a proatherogenic phenotype.18 33 Future experiments will be needed to determine whether lysophosphatidylcholine is involved in the MIEP activation in our system.
The peroxide content of oxLDL may also be relevant in the activation of MIEP. Indeed, it has been shown by Scholz et al34 and by our group35 that CMV replication and IE gene expression are enhanced in response to an increase in reactive oxygen intermediates in SMCs and ECs.34 35 Moreover, MIEP activation is a redox-sensitive NF-κB–mediated process.35 This may be of importance if future experiments show that NF-κB plays a role in oxLDL–mediated activation of MIEP.
Of additional pathophysiological relevance is the fact that LDL, present in plasma, enters the subintimal space, where it can be oxidized by free radicals produced by ECs and/or SMCs. Thus, once the monocyte adheres to the vessel wall at the site of vascular injury and enters the subintimal space, it is subjected to several stimuli that individually, and in combination, could lead to the activation of CMV residing in the monocyte.
Increasing evidence implicates CMV in the development of atherosclerosis and of restenosis (reviewed in References 3 and 363 36 ): CMV DNA sequences have been detected in the wall of atherosclerotic vessels, and CMV proteins have been identified in cultures of SMCs derived from atherosclerotic vessels37 38 39 ; accelerated coronary atherosclerosis is more common in cardiac transplant patients exposed to CMV than in those who have had no such exposure40 ; and atherosclerotic lesions develop in chickens infected with Marek’s disease virus, an avian herpesvirus.41 In addition, there is an association between the risk of occurrence of postatherectomy restenosis and prior infection with CMV.42
The case for a causal role of CMV in atherosclerosis/restenosis is further suggested by the fact that herpesviruses exert biological effects compatible with the development of these disease processes. Thus, CMV infection can lead to cellular proliferation,43 44 45 an effect that could be caused, at least in part, by CMV-induced secretion of growth factors and cytokines (which may act either in an autocrine or paracrine fashion46 ) or by the inhibition of p53 (a potent inhibitor of cell cycle progression) activity.4 47 48 In addition, inhibition of apoptosis by CMV47 49 might contribute to the accumulation of SMCs in the developing atherosclerotic lesion. CMV infection induces cellular expression of adhesion molecules,50 51 thereby increasing leukocyte adhesion to ECs. It also induces changes that are procoagulant52 53 and activates NF-κB,35 54 55 a cellular transcription factor that stimulates many genes, including those involved in inflammatory and immune responses. Infection of ECs with CMV could also ultimately lead to cell injury and death, thereby compromising the integrity of the intimal barrier. Moreover, an avian herpesvirus and herpes simplex virus have been shown to cause cholesterol accumulation in SMCs,56 57 and we have recently shown that CMV increases expression and activity of the scavenger receptor, which is critical for increased cellular uptake of oxLDL and for the development of foam cells.58 The latter are the characteristic cell type of the fatty streak, the earliest manifestation of atherosclerosis.
Given that CMV may play a role in atherogenesis/restenosis, the data reported in this investigation, when considered together with recently reported findings of this and other laboratories, suggest one possible mechanism by which CMV eventually becomes localized to the vessel wall, where it can then influence disease processes. In this scheme, myelomonocytic precursor cells act as a reservoir of CMV, leading to a small subpopulation of released circulating monocytes that harbor latent CMV. When the monocyte is recruited to sites of vascular injury, as it differentiates into a macrophage upon contact with the vessel wall, its contained virus is reactivated by EC/SMC and oxLDL-induced transactivation of the CMV MIEP. The reactivated virus begins replicating and subsequently infects neighboring ECs and SMCs. In addition, one of the early events that occurs when CMV infects an SMC is an increase in intracellular H2O2.35 Because MIEP activity is reactive oxygen species dependent, an increase in intracellular H2O2 could then contribute to a further increase in MIEP activity and, ultimately, enhanced CMV gene expression.35
Although it appears that this mode of CMV delivery to sites of vascular injury is biologically relevant, it is quite possible that other delivery mechanisms contribute to the vascular pathology that (it appears increasingly likely) CMV causes. If CMV does contribute to the development of vascular disease, then a further understanding of viral latency and reactivation from latency will be of critical importance in the ultimate development of effective treatment strategies.
Selected Abbreviations and Acronyms
|EGM||=||endothelial growth medium|
|MIEP||=||major IE promoter|
|PCR||=||polymerase chain reaction|
|PMA||=||phorbol 12-myristate 13-acetate|
|SMC||=||smooth muscle cell|
|SMGM 2||=||SMC growth medium 2|
1 When oxLDL was added to the coculture system and GMSX rather than serum was used to supplement the medium, we found that although HL-60 cells tolerated these conditions, the ECs and SMCs did not survive. Serum was therefore added to the cocultures. Since the initial experiments with oxLDL (the noncoculture experiments) were conducted in medium supplemented with GMSX, the effect of oxLDL on MIEP activity in transfected HL-60 cells was studied, in parallel, in cells cultured in GMSX and cells cultured in FBS. We found that although the level of the counts per minute increased slightly in the cultures with serum, the fold increase caused by oxLDL remained the same in the different culture conditions.
- Received October 22, 1996.
- Accepted April 7, 1997.
- © 1997 American Heart Association, Inc.
Huang ES, Kowalik T. The pathogenicity of human cytomegalovirus: an overview. In: Beker Y, Darai G, eds. Molecular Aspects of Human Cytomegalovirus Diseases. Berlin, Germany: Springer-Verlag;1993:1-45.
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.
Taylor-Wiedeman J, Sissons JP, Borysiewicz LK, Sinclair JH. Monocytes are a major site of human cytomegalovirus in peripheral blood mononuclear cells. J Gen Virol. 1991;72:2059-2064.
Taylor-Wiedeman J, Sissons JP, Sinclair JH. Induction of endogenous cytomegalovirus gene expression after differentiation of monocytes from healthy carriers. J Virol. 1994;68:1597-1604.
Maciejewski JP, Bruening EE, Donahue RE, Mocarski ES, Young NS, St Jeor SC. Infection of hematopoietic progenitor cells by human cytomegalovirus. Blood. 1992;80:170-178.
Minton EJ, Tysoe C, Sinclair JH, Cessans JPG. Human cytomegalovirus infection of the monocyte/macrophage lineage in bone marrow. J Virol. 1994;68:4017-4021.
Kondo K, Kaneshima H, Mocarski ES. Human cytomegalovirus latent infection of granulocyte-macrophage progenitors. Proc Natl Acad Sci U S A. 1994;91:11879-11883.
Meyers JD, Flournoy N, Thomas ED. Risk factors for cytomegalovirus infection after human marrow transplantation. J Infect Dis. 1986;153:478-488.
Ibanez CE, Schrier R, Ghazal P, Wiley C, Nelson JA. Human cytomegalovirus productively infects primary differentiated macrophages. J Virol. 1991;65:6581-6588.
Mocarski ES. Cytomegalovirus biology and replication. In: Roizman B, Whitley R, Lopez C, eds. The Human Herpesviruses. New York, NY: Raven Press Publishers; 1993:173-226.
Boom R, Geelen JL, Sol CJ, Raap AK, Minnaar RP, Klaver BP, Van der Noorda J. Establishment of a rat cell line inducible for the expression of human cytomegalovirus immediate-early gene products by protein synthesis inhibition. J Virol. 1986;58:851-859.
Stein J, Volk HD, Liebenthal C, Kruger DH, Prosch S. Tumor necrosis factor a stimulates the activity of the human cytomegalovirus major immediate early enhancer/promoter in immature monocytic cells. J Gen Virol. 1993;74:2333-2338.
Abrahamson TG, Carter CS, Read EJ, Rubin M, Goetzman HG, Lizzio EF, Lee YL, Hanson M, Pizzo PA, Hoffman T. Stimulatory effect of counterflow centrifugal elutriation in large-scale separation of peripheral blood monocytes can be reserved by storing the cells at 37°C. J Clin Apheresis. 1991;53:48-53.
Havel RJ, Eder HA, Bragdon JH. The distribution and chemical composition of ulticentrifugally separated lipoproteins in human serum. J Clin Invest. 1955;43:1345-1353.
Quinn MT, Parthasarathy S, Steinberg D. Lysophosphatidylcholine: a chemotactic factor for human monocytes and its potential role in atherogenesis. Proc Natl Acad Sci U S A. 1988;85:2805-2809.
Liao F, Berliner JA, Mehrabian M, Navab M, Demer LL. Minimally modified low density lipoprotein is biologically active in vivo in mice. J Clin Invest. 1991;87:2253-2257.
Goreman C, Moffat LF, Howard BH. Recombinant genomes which express chloramphenicol acetyltransferase in mammalian cell. Mol Cell Biol. 1982;2:1044-1051.
Waldman WJ, Knight DA, Huang EH, Sedmak DD. Bidirectional transmission of infectious cytomegalovirus between monocytes and vascular endothelial cells: an in-vitro model. J Infect Dis. 1995;171:263-272.
Faruqui RM, DiCorleto PE. Mechanisms of monocyte recruitment and accumulation. Br Heart J. 1994;69(suppl):S19-S29.
Shackelford RE, Misra UK, Florine-Casteel K, Thai SF, Pizzo SV, Adams DO. Oxidized low density lipoprotein suppresses activation of NfkB in macrophages via a pertussis toxin-sensitive signaling mechanism. J Biol Chem. 1995;270:3475-3478.
Peng HB, Ravjavashisth TB, Libby P, Liao JK. Nitric oxide inhibits macrophage-colony stimulating factor gene transcription in vascular endothelial cells. J Biol Chem. 1995;270:17050-17055.
Parhami F, Fang ZT, Fogelman AM, Andalibi A, Territo MC, Berliner JA. Minimally modified low density lipoprotein-induced inflammatory responses in endothelial cells are mediated by cAMP. J Clin Invest. 1993;92:471-478.
Ghazal PJ, Nelson JA. Transcription factors and viral regulatory proteins as potential mediators of human cytomegalovirus pathogenesis. In: Beker Y, Darai G, eds. Molecular Aspects of Human Cytomegalovirus Diseases. Berlin, Germany: Springer-Verlag; 1993:360-383.
Ares MPS, Kailin B, Eriksson P, Nilsson J. Oxidized LDL induces transcription factor activator protein-1 but inhibits activation of nuclear factor-κB in human vascular smooth muscle cells. Arterioscler Thromb Vasc Biol. 1995;15:1584-1590.
Ohlsson BG, Englund MCO, Karlsson ALK, Knutsen E, Erixon C, Skribeck H, Liu Y, Bondjers G, Wiklund O. Oxidized low density lipoprotein inhibits lipopolysaccharide-induced binding of NfκB to DNA and the subsequent expression of tumor necrosis factor-α and interleukin-1β in macrophages. J Clin Invest. 1996;98:78-89.
Kume N, Cybulsky MI, Gimbrone MA. Lysophosphatidylcholine, a component of atherogenic lipoproteins, induces mononuclear leukocyte adhesion molecules in cultured human and rabbit arterial endothelial cells. J Clin Invest. 1992;90:1138-1144.
Speir E, Shibutani T, Yu Z-X, Ferrans V, Epstein SE. Role of reactive oxygen intermediates in cytomegalovirus gene expression and in the response of human smooth muscle cells to viral infection. Circ Res. 1996;79:1143-1152.
Epstein SE, Speir E, Zhou YF, Guetta E, Leon M, Finkel T. The role of infection in restenosis and atherosclerosis: focus on cytomegalovirus. Lancet. 1996;348:s13-s17.
Hendrix MGR, Dormans PH, Kitslaar P, Bosman F, Bruggeman CA. The presence of cytomegalovirus nucleic acids in arterial walls of atherosclerotic and nonatherosclerotic patients. Am J Pathol. 1989;134:23-28.
Grattan MT. Accelerated graft atherosclerosis following cardiac transplantation: clinical perspectives. Clin Cardiol. 1991;14(suppl II):16-20.
Albrecht T, Boldogh I, Fons M, Lee CH, AbuBakar SN, Russell JM, Au WW. Cell-activation responses to cytomegalovirus infection: relationship to the phasing of CMV replication and to the induction of cellular damage. In: Harris JR, ed. Subcellular Biochemistry. New York, NY: Plenum Publishing Corp; 1989:157-202.
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.
Gonczol E, Plotkin S. Cells infected with human cytomegalovirus release a factor(s) that stimulates cell DNA synthesis. J Gen Virol. 1984;65:1833-1837.
Muganda P, Mendoza O, Hernandez J, Quian Q. Human cytomegalovirus elevates levels of cellular protein p53 in infected fibroblasts. J Virol. 1994;68:8208-8034.
Zhu H, Shen Y, Shenk T. Human cytomegalovirus IE1 and IE2 proteins block apoptosis. J Virol. 1995;69:7960-7970.
Pryzdial ELG, Wright JF. Prothrombinase assembly on an enveloped virus: evidence that cytomegalovirus surface contains procoagulant phospholipid. Blood. 1994;84:3749-3757.
Kowalik TF, Wing B, Haskill JS, Azizkhan JC, Baldwin AS, Huang ES. Multiple mechanisms are implicated in the regulation of NFκB activity during human cytomegalovirus infection. Proc Natl Acad Sci U S A. 1993;90:1107-1111.
Yurochko AD, Kowalik TF, Huong SH, Huang ES. Human cytomegalovirus upregulates NF-κB p105/p50 and p65 promoters. J Virol. 1995;69:5391-5400.
Hajjar DP, Pomerantz KB, Falcone DJ, Weksler BB, Grant AJ. Herpes simplex virus infection in human arterial cells: implications in atherosclerosis. J Clin Invest. 1987;80:1317-1321.