|Home | About | Journals | Submit | Contact Us | Français|
Infectious agents have been linked to atherosclerosis and its acute manifestations; however, little is known about their influence in the context of established risk factors.
To elucidate the role of the cytomegalovirus (CMV)-encoded chemokine receptor US28 in myocardial infarction (MI) afflicting patients with or without type II diabetes mellitus (NIDDM) on a molecular level.
In a group of patients (n=112) with a high prevalence of NIDDM and coronary artery disease, CMV serology was performed, and mRNA expression of US28 and immediate early 1 gene as markers of CMV reactivation were analyzed in peripheral mononuclear blood cells by a nested reverse transcription-polymerase chain reaction. Moreover, transendothelial chemotaxis assays using mononuclear cells transfected with or without US28 were performed in vitro.
While the incidence of smoking was higher in nondiabetic patients with MI than in those without MI, significant differences in other risk factors, such as cholesterol, low density lipoprotein, fibrinogen, blood pressure, and Chlamydia pneumoniae immunoglobulin G or CMV immunoglobulin G titres, were not observed. In contrast, the levels of C-reactive protein reflecting inflammation or infection were raised in NIDDM patients with or without MI. Notably, mRNA expression of intermediate early 1 gene and US28 indicative of CMV reactivation was detected in a small subset (four of 21) of NIDDM patients with MI but not in those without MI (P<0.03). Transfection of US28 in mononuclear cells conferred transendothelial chemotaxis to monocyte chemokines, inferring a mechanism for deleterious effects of CMV under permissive conditions.
Results show that MI was associated with mononuclear expression of CMV genes such as functional chemokine receptor US28 in a subset of patients with NIDDM, inferring that this association may predispose to MI. Ongoing infection or inflammation in NIDDM patients as shown by increased C-reactive protein may account for susceptibility to CMV reactivation and MI.
Epidemiological studies have suggested a link between coronary artery disease (CAD) and bacterial or viral infections, ie, Chlamydia pneumoniae or cytomegalovirus (CMV) (1). An association of CMV with vascular disease has been derived from studies correlating relative risks and CMV seropositivity, which were based mainly on restenosis, transplantation or extracoronary lesions (1,2). However, the presence of CMV in plaques is not correlated with serum titres, CMV mRNA has not been detected in atherectomy specimens, and increased CMV antibody levels were not a risk factor for CAD and its acute manifestations in a nested case control study (1,3). Notably, CMV encodes the chemokine receptor homologue US28, which binds CC chemokines (4) and can thereby mediate migration of CMV-infected vascular smooth muscle cells (5). Beyond a molecular key for CMV to vascular disease, this has been speculated to facilitate its dissemination. Because active CMV infection in the vascular wall may be transient, mononuclear cells, a primary source of persistent CMV with high mobility, may be intrinsically better suited as efficient vectors for CMV delivery throughout the body and for re-exacerbation and migration during altered immune responses, atherogenic activation or differentiation (6,7).
To elucidate whether acute myocardial infarction (MI) may be associated with CMV reactivation in the context of established risk factors or inflammatory markers, mononuclear cell expression of CMV-encoded genes (eg, US28), in addition to CMV serology, was studied in a group of patients with a high prevalence of type II diabetes mellitus (NIDDM).
The study was approved by the ethics committee of the Ludwig-Maximilians-Universität, Munich, Germany. Participants in the study were consecutive, unselected and consenting patients with vascular risk who either had suffered MI within the previous 48 h (n=36) or had not suffered MI within six months (n=76). Diagnosis of acute MI was established by enzymatic and electrocardiographic criteria. NIDDM had been previously diagnosed. Blood was taken within 48 h after the onset of symptoms (MI) or at routine visits (no MI). Peripheral blood mononuclear cells (PBMC) were prepared by Ficoll density-gradient centrifugation. Total RNA was isolated from 106 PBMC using TRIzol (Invitrogen, Germany). Serum levels of cholesterol, low density lipoprotein, fibrinogen, C-reactive protein, glycated hemoglobin A1, immunoglobulin (Ig) G specific for CMV or Chlamydia pneumoniae and soluble vascular adhesion molecule-1 were determined by enzymatic, coagulometric, serological or ELISA assays.
Total RNA was treated with DNase, DNA contamination was excluded by polymerase chain reaction (PCR) as outlined below, mRNA was reverse transcribed, and cDNA was used for nested PCR reactions (35 cycles each) with primers TTGAC-TACGACGATGAAGCG and CAGTGACAAAAGGCG-AGTGA (outer) or AGAACTCATGCTCGGTGCTT and GAGCGCGCGCTTGAGTGATT (inner) for US28, and immediate early 1 (IE1) gene primers flanking the exon 3-4 splice junction (6). Amplifications without reverse transcription (RT) revealed no products, excluding fragments derived from CMV DNA. With the use of dilutions of pcDNA3-US28 provided by Pleskoff et al (8) or CMV-DNA, PCR reactions were sensitive to detect fewer than 100 copies. Products were analyzed by 1.5% agarose gel electrophoresis and high performance liquid chromatography (9).
Jurkat T cells were transfected with c-myc tagged pcDNA3-US28 (8) or pcDNA3 vector alone, as previously described (10). Expression of US28 in clones selected by resistance to geneticin was confirmed by immunoblotting or flow cytometry using 9EG10 monoclonal antibody (Boehringer Mannheim, Germany). Transmigration assays were performed as described (10) using filters coated with human umbilical vein endothelial cells and monocyte chemotactic protein-1 (MCP-1), RANTES or fractalkine at 1 to 1000 nmol/L in the lower chamber.
Patient data were analyzed by ANOVA and Scheffe’s F test for parametric data, or by χ2 test for non-parametric data. In vitro data were analyzed by signed-rank tests (Kruskal-Wallis, Wilcoxon).
Because the role of endothelial activation and infectious agents in CAD manifestations may be affected by NIDDM due to sustained inflammation or a modified immune control, we compared CMV- or Chlamydia pneumoniae-specific IgG and CMV gene expression in a group of patients with high prevalence of NIDDM and CAD (Table 1). At the time of analysis, patients either had suffered acute MI in the previous 48 h or had not suffered MI within six months. Whereas smoking status differed between nondiabetic patients with and those without MI, no significant differences in age, sex, hypertension, fibrinogen, lipid parameters, C pneumoniae IgG or CMV IgG titres were found. Because CMV IgG titres do not discriminate between true latent and recurrent CMV infection, we tested mRNA expression of IE1 gene and early US28 gene in PBMC from patients by a nested RT-PCR with high sensitivity. IE1 gene and US28 mRNA indicative of CMV reactivation were not detected in patients without MI or NIDDM but were in a small subset (four of 21) of NIDDM patients with MI (Table 1), all of whom were positive for CMV IgG (data not shown). Detection of CMV mRNA correlated with a severe glucose imbalance or suboptimal therapeutic control of NIDDM, as was evident from the increased percentage of glycated hemoglobin A1 (8.5±1.1% versus 6.7±1.2% in all other patients with NIDDM) and increased concentrations of the endothelial activation marker soluble vascular cell adhesion molecule-1 in this subgroup (data not shown). Notably, C-reactive protein as an important inflammatory marker was substantially higher in patients with NIDDM than in those without NIDDM, being surprisingly low for NIDDM patients with MI. However, differences between patients with and those without MI were not observed. This is in accordance with a report that high C-reactive protein concentrations increase the risks associated with high antimicrobial antibody concentrations, such as those directed to herpes simplex virus-1 (3), and may reflect an increased incidence of infectious complications or chronic inflammation in NIDDM that may cause or confound susceptibility to CMV reactivation and MI. Alternatively, impaired viral immune responses, as exemplified by altered cellular and humoral reactions to vaccination (11), may predispose to CMV reactivation in NIDDM. Thus, our results emphasize the potential importance of CMV mRNA detection.
US28 has been shown to mediate migration of arterial smooth muscle cells (5). To identify mechanisms by which US28 in mononuclear cells may contribute to CAD manifestations, US28 was expressed in Jurkat T cells to study transendothelial chemotaxis (Figure 1). Indeed, MCP-1, RANTES or fractalkine induced significant migration of cells transfected with US28 but not vector alone. Hence, functional expression of US28 may facilitate recruitment of infected mononuclear cells into atherosclerotic lesions, promoting instability and thrombogenicity by secretion of matrix proteases or tissue factor (12). This may also be relevant to the increased risk for thrombotic events after coronary stent placement associated with previous CMV infection (13).
Notably, CMV mRNA indicative of active infection may not be detectable in atherectomy samples from primary or restenotic lesions (1). Moreover, CMV infection in rats has been shown to increase neointimal responses to injury without consistent evidence for CMV in the vascular wall (14), suggesting that it may be due to systemic immune reactions. US28 expression and function in mononuclear cells not only may propagate CMV dissemination but also may explain how CMV recurrence may affect (re)stenosis without initially and directly affecting smooth muscle cells in the vessel wall.
In conclusion, our data suggest that CMV reactivation involving expression of the chemokine receptor US28 (which can occur within 48 h of infection to precede MI) may be associated with increased risk for MI in NIDDM. Thus, CMV gene expression may be predictive of acute manifestations of CAD when it coincides with certain established risk factors. On the other hand, noninflammatory stress, such as that due to acute MI, has been described to directly stimulate CMV reactivation with IE gene transcription by way of catecholamine production (15). However, it appears unlikely to be responsible for the CMV reactivation observed in our study because it did not occur in nondiabetic patients with MI. Novel paradigms by which CMV is implicated as a culprit in vascular disease may be verified in animal studies (eg, by knocking in US28) and warrant prospective studies in humans (16,17), which are underway to further substantiate the notion that infections may be important in acute MI. In accordance with studies on other infections (3), conditions of inflammatory stress in NIDDM as shown by increased C-reactive protein may prime or confound susceptibility to CMV reactivation and MI.
This study was supported by Deutsche Forschungsgemeinschaft (grant We-1913/2, GrK-438). This work in part fulfills requirements for the doctoral dissertation of C von Stülpnagel.