Correspondence: Jürgen Rödel, Institute of Medical Microbiology, Friedrich Schiller University of Jena, Semmelweisstr. 4, D-07740 Jena, Germany. Tel.: +49 3641 933905; fax: +49 3641 933474; e-mail: firstname.lastname@example.org
Chlamydia pneumoniae infection may play a role in the pathogenesis of atherosclerosis. In this study, an oligonucleotide microarray was utilized to examine the transcriptional response of human aortic smooth muscle cells (AoSMC) to C. pneumoniae infection. Alteration of mRNA expression in 71 out of 780 genes was detected at 24 h after infection. Among the down-regulated genes, platelet-derived growth factor receptor-β (PDGFR-β) was identified as a target for further analysis because the PDGF system is involved in the fibroproliferative response of SMC in atherogenesis. Reverse transcriptase PCR analysis demonstrated that C. pneumoniae inhibits the up-regulation of PDGFR-β mRNA occurring in AoSMC after mock infection. PDGFR-β protein synthesis was examined by immunoblotting and fluorescence-activated cell sorting. Compared with mock-infected cells, the amount of receptor protein was reduced at 24, 48, and 72 h after infection. Diminished PDGFR-β synthesis in infected cultures was accompanied by the suppression of AoSMC growth following PDGF-BB stimulation. The interference of C. pneumoniae with PDGFR-β expression may result in decreased SMC proliferation in atherosclerotic plaques, thereby affecting the development and stability of advanced lesions.
Atherosclerosis is characterized by lipid accumulation in the arterial vessel wall, local inflammation, and fibrosis (Ross, 1999). The importance of inflammatory processes that are involved in the formation of early atherosclerotic lesions and in complications resulting in acute ischemic syndromes is compatible with the hypothesis that infectious agents may function as risk factors in atherosclerotic diseases. Chlamydia pneumoniae has been intensively investigated in this context but its etiological involvement remains controversial. In several cases, C. pneumoniae could be isolated in culture from atherosclerotic plaque samples but PCR-based diagnostic studies produced inconsistent results and therefore the prevalence of vascular chlamydial infection is unknown (reviewed by Ieven & Hoymans, 2005). Large antibiotic trials on the secondary prevention of clinical events in patients with coronary heart disease have failed to show any benefit but it should be considered that the results of these therapeutic studies cannot rule out a pathogenic role of chlamydiae (Cannon et al., 2005; Grayston et al., 2005). Animal models did also not result in a general consensus but experiments on mice demonstrated that C. pneumoniae can accelerate hypercholesteremia-induced atherosclerosis (reviewed by de Kruif et al., 2005). Numerous in vitro studies on the infection of endothelial cells, macrophages/monocytes, and smooth muscle cells (SMC) have offered plausible mechanisms by which the pathogen may contribute to endothelial dysfunction, low-density lipoprotein (LDL) accumulation in early lesions, and fibrous plaque development (Belland et al., 2004).
The activation of SMC that migrate from the media to the intima or differentiate from mesenchymal stem cells plays an important role in the progression of atherosclerosis. Neointimal SMC proliferate and synthesize interstitial collagens, thereby forming a fibrous cap that covers the necrotic lipid-rich core of the plaque (Ross, 1999). In previous works, it has been shown that the infection of these cells by C. pneumoniae induces the production of IL-6, basic fibroblast growth factor (bFGF), matrix metalloproteinases (MMPs), and prostaglandin E2 (PGE2) (Rödel et al., 2000, 2003, 2004). To further characterize altered gene expression in Chlamydia-infected SMC, an oligonucleotide microarray screening was performed. Among the down-regulated genes, platelet-derived growth factor receptor β (PDGFR-β) was identified, which has not yet been described as a cellular factor modulated by C. pneumoniae infection. Because neointimal SMC proliferation in atherosclerotic plaques is mediated by the mitogenic effects of PDGFs via PDGFR-β signaling, the effect of C. pneumoniae on the expression of PDGFR-β in SMC was further investigated (Sano et al., 2001; Kozaki et al., 2002).
Materials and methods
Chlamydia pneumoniae infection of aortic smooth muscle cells (AoSMC)
Chlamydia pneumoniae strain TW-183 (obtained form the Institute of Ophthalmology, London, UK) was propagated in buffalo green monkey (BGM) cells as described previously (Rödel et al., 2000). Infectivity titers of chlamydial stocks were quantified by titrating the number of inclusion-forming units (IFU) per milliliter in BGM cells. Mycoplasma contaminations were excluded using the MycoDtect DNA array (Greiner Bio-One, Frickenhausen, Germany).
Subcultures of human AoSMC (C-12533, PromoCell, Heidelberg, Germany) were grown in 35-mm-diameter culture wells (six-well plates). The cells were inoculated with C. pneumoniae at multiplicities of infection (MOIs; IFU per cell) of 0.5, 1, or 5. After centrifugation at 4000 g at 37°C for 45 min, the inoculum was decanted, and the cells were further incubated with minimal essential medium (OptiMEM; Gibco, Invitrogen, Karlsruhe, Germany) containing 5% fetal calf serum (FCS) (Biochrom, Berlin, Germany). Infection of the cells was confirmed by determination of chlamydial inclusions after staining with fluorescein isothiocyanate (FITC)-conjugated antibody to chlamydial lipopolysaccharide (Imagen; Dako, Hamburg, Germany) as described previously (Rödel et al., 2003).
In some experiments, conditioned medium from mock-infected and infected AoSMC cultures was harvested at 48 h after infection, clarified by centrifugation at 20 000 g at 4°C for 30 min, and applied to uninfected AoSMC for 24 h.
Total RNA was prepared using the RNeasy Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. For oligonucleotide microarrays, cells from three 35-mm-diameter wells were collected for each sample. RNA quality was examined by agarose gel electrophoresis. The amount was measured by A260 nm.
For the analysis of SMC gene expression, the Lab-Arraytor human 60-inflammation in-house research array (SIRS-Lab, Jena, Germany) comprising oligonucleotide probes of an average length of 65 bases for 780 genes involved in inflammatory pathways was used. cDNA synthesis, labeling, and hybridization to the arrays were performed according to the manufacturer's protocol.
Fifteen micrograms of total RNA were reverse transcribed using the SuperScript II reverse transcriptase kit (Invitrogen) and oligo (dT)18 primer, substituting a fraction of dTTP in the newly synthesized strands with aminoallyl-dUTP (AA-dUTP). Conversion of RNA/cDNA duplexes into single-strand cDNA was performed by RNA alkaline hydrolysis. cDNAs were labeled with AlexaFluor 555 and AlexaFluor 647 monofunctional dyes (Invitrogen) by coupling to incorporated AA-dUTP. Cohybridization of labeled cDNA to microarrays was performed in a formamid-based hybridization buffer at 42°C for 12 h in an HS400 hybridization station (Tecan, Crailsheim, Germany). After washing and drying, the microarrays were scanned using an Axon 400B fluorescence scanner (Molecular Devices, Ismaning, Germany). Digital images resulting from posthybridization array scanning were quantified with GenePix Pro 4.0 software (spot detection and background subtraction, spot flagging according to defined signal-to-noise threshold values). For the normalization and transformation of signals obtained from both channels, the approach from Huber et al. (2003) was applied. The technical replicates on a microarray (multiple spots of the same probe) were filtered from the corrected and transformed signal intensities depending on their spot quality. Per spot, replicates with the highest quality flag were selected and the corresponding signal intensities were averaged. In order to minimize the number of false positives, probes with predominantly poor spot quality (over all hybridization) were excluded from the analysis. Natural logarithmic (ln) signal ratios of infected/mock-infected cells of ≥0.4 or ≤−0.4 were considered to be indicative of differences in mRNA expression. Only genes whose expression varied in two independent experiments were selected.
Reverse transcriptase (RT)-PCR analysis
RT-PCR was carried out as described previously (Rödel et al., 2003). The specific primers were as follows: 5′-AATGTCTCCAGCACCTTCGT-3′ and 5′-AGCGGATGTGGTAAGGCATA-3′ (688-bp product) for PDGFR-β (Basciani et al., 2002), and 5′-GAAGGAGCCTGGGTTCCCTG-3′ and 5′-TTTCTCACCTGGACAGGTCG-3′ (217-bp product) for PDGF-B (Wang et al., 1989). The sequences of pyruvate dehydrogenase (PDH) primers used as control were 5′-GGTATGGATGAGGACCTGGA-3′ and 5′-CTTCCACAGCCCTCGACTAA-3′ (105-bp product) (Rolfs et al., 1992). PCR products were electrophoresed on 1% agarose gels and visualized with SYBR green staining. The volumes (OD × mm2) of the band images were quantitated with Multi-Analyst PC software (Bio-Rad, Munich, Germany) and normalized against the PDH signal from the same sample.
Enzyme-linked immunosorbent assays (ELISA)
The IL-8 and RANTES (regulated on activation normal T-cell expressed and secreted) levels in culture supernatants were measured by CytoSet ELISAs (Biosource, Ratingen, Germany) according to the supplier's protocol.
Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and immunoblotting
Polyacrylamide-SDS gel electrophoresis of cell lysates and immunoblot assays were performed according to the protocol of a previous study (Rödel et al., 2000). PDGFR-β was detected using a rabbit polyclonal antibody (958, Santa Cruz Biotechnology, Heidelberg, Germany). As a reference band, Janus kinase 1 (JAK1) was detected with a mouse monoclonal antibody (clone 73, BD Biosciences, Heidelberg, Germany). After incubation with alkaline phosphatase-conjugated goat anti-rabbit (Santa Cruz Biotechnology) or -mouse IgG (Dianova, Hamburg, Germany), the bands were visualized with 5-bromo-4-chloro-3-indolylphosphate toluidine salt-p-nitroblue tetrazolium chloride (Sigma Fast, Sigma-Aldrich, Seelze, Germany).
Cell suspensions were stained with phycoerythrin (PE)-conjugated rabbit antibody to human PDGFR-β (Santa Cruz Biotechnology) or normal rabbit IgG (Santa Cruz Biotechnology) as an isotype control. Labeled cells were analyzed with a FACScan (BD Biosciences) flow cytometer and Cell Quest software. Ten thousand cells were analyzed for each sample.
SMC proliferation assay
Cells were plated onto six-well plates at 105 cells well−1, maintained in complete SMC growth medium 2 (PromoCell) overnight, and then infected with C. pneumoniae at an MOI of 5. After infection, the cells were incubated with OptiMEM containing 5% FCS. The next day, PDGF-BB (Active Bioscience; purchased from Biochrom, Berlin, Germany) was added to infected and mock-infected SMC at a final concentration of 50 ng mL−1. In parallel wells, SMC were restimulated with complete SMC growth medium 2 (PromoCell) and used as proliferation control. After 3 days, cells were trypsinized, resuspended in medium, and counted under the microscope in a Neubauer chamber.
Statistical comparisons were made using Student's t test. P values ≤0.05 were considered to be statistically significant.
Identification of C. pneumoniae-modulated genes by oligonucleotide microarray
To evaluate changes in mRNA expression in human AoSMC in response to C. pneumoniae infection, a microarray screening analysis was performed using the Lab-Arraytor human-60, which was manufactured for internal R&D purposes of SIRS-Lab GmbH Jena. AoSMC were infected at an MOI of 5, which resulted in 29±4% of inclusion-containing cells in the cultures (n=4). With the aim of focusing on differentially transcribed genes of cytokines, chemokines, MMPs, growth factors, and their receptors, RNA was prepared at 24 h after infection because in previous studies the expression of some of these factors was found to be significantly modulated by chlamydial infection at this time point (Rödel et al., 2000, 2003). During the experiments, 47 genes were observed whose expression was enhanced upon chlamydial infection (Table 1). Among these were primarily genes encoding for CC and CXC chemokines. In comparison, only a few interleukin genes were differentially transcribed upon infection (IL-1β, IL-6, IL-11, and IL-25). Genes that were also up-regulated included apoptosis-related factors, signal transduction molecules, transcription regulators, adhesion molecules, and tumor necrosis factor (TNF) receptor superfamily members (Table 1). For the genes depicted in Table 1, the results of two independent microarrays generally confirm each other. However, it should be noted that there are some examples that have to be interpreted with caution because of quantitative differences in the values of the experiments.
Table 1. AoSMC gene expression profile at 24 h following Chlamydia pneumoniae infection
Relative expression [mean ratio (first ratio/second ratio)]†
NCBI GenBank accession number.
Fold increase of mRNA level in infected cells was calculated relative to the expression value in mock-infected cells and is represented on a ln scale. Results of two separate experiments including mean values are presented. Genes of an ln expression ratio ≥0.4 or ≤−0.4 in both experiments were regarded as differentially transcribed genes.
Some of the results were consistent with previous reports on SMC gene expression upon C. pneumoniae infection, including up-regulation of matrix metalloproteinase 1 (MMP-1), monocyte chemotactic protein 1 (MCP-1) and IL-6 (Dechend et al., 2003; Rödel et al., 2003; Yang et al., 2005). As an indication of the accuracy of the microarray data, the levels of IL-8 and RANTES were determined in culture supernatants. The genes of both chemokines were differentially transcribed following infection but IL-8 mRNA was elevated to a higher extent than RANTES mRNA (Table 1). These findings correspond to chemokine levels in culture supernatants. At 72 h after infection the IL-8 level was 14.1-fold increased (from 194±80 pg mL−1 in mock-infected cultures to 3005±395 pg mL−1 in infected cultures), whereas the increase in the RANTES level was only 3.6-fold (from 175±94 pg mL−1 in mock-infected cultures to 505±134 pg mL−1 in infected cultures) (Fig. 1).
The expression of 24 genes was observed to be down-regulated upon infection (Table 1). These genes included apoptosis-related factors, arachidonic acid metabolite receptors, signal transduction molecules, cytokine receptors, and TNF ligand and receptor superfamily members (Table 1). Among the down-regulated genes, PDGFR-β were identified as a new target for further analysis because PDGFR-β is expressed in atherosclerotic plaques and is involved in PDGF-BB-mediated SMC migration and neointimal proliferation (Rubin et al., 1988; Abe et al., 1998; Sano et al., 2001).
Expression of PDGFR-β in C. pneumoniae-infected AoSMC
RT-PCR analysis of PDGFR-β mRNA levels in AoSMC confirmed the microarray results (Fig. 2a and b). In mock-infected cells, PDGFR-β mRNA increased over a 24 h incubation after centrifugation and medium change. In contrast, the accumulation of mRNA was suppressed following chlamydial infection (Fig. 2a). Exposure of AoSMC to different infectious doses demonstrated a dose-dependent inhibitory effect of C. pneumoniae on PDGFR-β mRNA accumulation (Fig. 2b). Neither mock-infected nor infected AoSMC expressed mRNA of the β receptor ligand PDGF-B (data not shown).
Further experiments were performed to determine the production of PDGFR-β protein. Immunoblot analyses showed a specific band of PDGFR-β of c. 180 kDa (Fig. 2c and d). In comparison with mock-infected cells, lower amounts of receptor protein were observed in Chlamydia-exposed cells at 24, 48, and 72 h following infection (Fig. 2c). At 48 h, C. pneumoniae infection downregulated the PDGFR-β protein level in a dose-dependent fashion (Fig. 2d). FACS analysis also revealed a decreased PDGFR-β protein expression in infected AoSMC (Fig. 2e).
The incubation of uninfected AoSMC with conditioned medium from infected cultures did not markedly affect PDFGR-β mRNA expression, indicating that the diminished expression of PDGFR-β in infected cell cultures is due to the direct effects of chlamydiae but not significantly mediated by the induction of soluble factors acting in a paracrine manner (Fig. 3).
Inhibition of PDGF-BB-stimulated AoSMC proliferation by C. pneumoniae
To examine whether decreased PDGFR-β expression upon infection results in a loss of the ability of AoSMC to respond to PDGF-BB stimulation, a proliferation assay was performed. For determination of SMC growth, direct cell counting was used instead of a DNA synthesis assay because it could not be excluded that the incorporation of labeled nucleotides into the DNA of metabolically active chlamydiae inside the cells may affect the test results.
Mock-infected AoSMC did not proliferate when cultured in basal medium containing 5% FCS but no additional supplements. Cell numbers in C. pneumoniae-infected cultures did not significantly differ from those counted in mock-infected cultures after 3 days of incubation (Fig. 4). The addition of 50 ng mL−1 PDGF-BB caused a moderate but significant increase in cell numbers in mock-infected cultures over the 72 h period of time but did not stimulate the proliferation of infected cells (Fig. 4).
Chlamydia pneumoniae has been localized in SMC in atherosclerotic plaques by electron microscopy and immunohistochemistry (Kuo et al., 1993; Shor et al., 1998). This study shows that C. pneumoniae infection of AoSMC in vitro reduces the expression of PDGFR-β and suppresses cell proliferation in the presence of PDGF-BB.
The neointimal proliferation of SMC and the synthesis of extracellular matrix proteins by these cells result in the formation of fibroatheromatous lesions during atherogenesis (Ross, 1999). PDGFs are discussed as important factors contributing to SMC migration and increased proliferation (Raines, 2004). This growth factor family consists of homo- and heterodimeric ligands that are composed of polypeptide chains A, B, C, or D. PDGFR-α binds A, B, and C chains whereas PDGFR-β binds B and D chains with high affinity. While PDGF receptors are present in normal arterial tissue at only very low levels, the expression of PDGFR-β is elevated in SMC in atherosclerotic lesions (Rubin et al., 1988). In ApoE−/− mice, the administration of PDGFR-β antibodies has been shown to reduce the SMC content in fibrous plaques (Sano et al., 2001). The findings of this work suggest that C. pneumoniae infection may diminish PDGFR-β expression in SMC during atherogenesis. Because chlamydial infection inhibits the up-regulation of PDGFR-β mRNA naturally occurring in AoSMC following mock infection in culture the pathogen obviously interferes in signaling pathways that are involved in the stimulation of PDGFR-β gene expression.
It is unclear whether C. pneumoniae stimulates the production of PDGF-BB, which represents a primary ligand for PDGFR-β. Coombes et al. (2002) reported the up-regulation of PDGF-B mRNA in Chlamydia-infected endothelial cells. However, elevated amounts of PDGF-BB homodimers were not found in culture supernatants of infected endothelial cells (Prochnau et al., 2004). In this study an induction of PDGF-B mRNA in AoSMC upon infection could not be observed. In the present experiments, the international C. pneumoniae reference strain TW-183 was used. At present, it is unknown whether genetic differences between respiratory and cardiovascular chlamydial isolates may significantly affect host cell responses.
Interestingly, human cytomegalovirus (HCMV), a pathogen that has also been associated with atherosclerotic diseases, up-regulates PDGFR-β expression in SMC, thereby enhancing the proliferative activity of the cells in the presence of PDGF-BB (Reinhardt et al., 2005). These findings indicate that C. pneumoniae and HCMV may have differential effects on neointimal SMC proliferation in atherosclerotic plaques.
The present results seem to be in contrast to the observation that C. pneumoniae stimulates SMC proliferation via p44/p42 mitogen-activated protein kinase activation (Sasu et al., 2001). On the other hand, C. pneumoniae can inhibit the proliferative activity of SMC not only by down-regulation of PDGFR-β synthesis but also by induction of high PGE2 levels suppressing SMC growth in a paracrine manner (Rödel et al., 2004). Therefore, it is possible that a net inhibitory effect of chlamydial infection on neointimal SMC proliferation may occur in vivo. This hypothesis is supported by studies on atherosclerotic plaque morphology in LDL receptor/ApoE−/− and LDL receptor−/− mice in which a significant decrease in the SMC content of the fibrous cap of aortic and carotid plaques following C. pneumoniae infection has been demonstrated (Ezzahiri et al., 2003; Hauer et al., 2006).
The formation of a fibrous cap covering the lipid-rich core determines the stability of atherosclerotic plaques (Lafont, 2003). A thin fibrous cap with few SMC and low amounts of interstitial collagens increases the vulnerability of the plaque and the risk of rupture with subsequent thrombosis, events that are mainly responsible for acute ischemic syndromes such as myocardial infarction or stroke (Libby, 1995; Aidinian et al., 2006). Vulnerable plaques are characterized not only by a thin fibrous cap but also by a significantly increased content of macrophages and T lymphocytes (Aidinian et al., 2006). The results of the microarray experiments show that the infection of SMC with C. pneumoniae stimulates the expression of several chemokines that attract different types of leukocytes and that are known to be produced in atherosclerotic lesions (MCP-1, RANTES, PARC (pulmonary and activation-regulated chemokine), Eotaxin, and IP-10) (Weber et al., 2004).
In conclusion, the findings of this study suggest that C. pneumoniae infection of SMC in atherosclerotic lesions may diminish SMC proliferation via down-regulation of PDGFR-β expression but may enhance leukocyte accumulation, thereby promoting the development of a more vulnerable plaque phenotype. Whether these effects have clinical significance remains unclear because it has been called into question that C. pneumoniae infection is a common phenomenon in atherosclerosis.
This work was supported by Grant B307-04004 from the Interdisciplinary Centre for Clinical Research (ICCR) of Jena and the Research Ministry of Thuringia. The authors thank E. Birch-Hirschfeld (Institute of Virology and Antiviral Therapy, Jena) for providing oligonucleotide primers. The authors are grateful to K. Prager and K. Nüske for excellent technical assistance.