Chlamydia pneumoniae and atherosclerosis

Authors


E-mail rbelland@utmem.edu; Tel. (+1) 901 448 8544; Fax (+1) 901 448 3330.

Summary

Exposure to Chlamydia pneumoniae is extremely common, and respiratory infections occur repeatedly among most people. Strong associations exist between C. pneumoniae infection and atherosclerosis as demonstrated by: (i) sero-epidemiological studies showing that patients with cardiovascular disease have higher titres of anti-C. pneumoniae antibodies compared with control patients; (ii) detection of the organism within atherosclerotic lesions, but not in adjacent normal tissue by immunohistochemistry, polymerase chain reaction and electron microscopy and by culturing the organism from lesions; and (iii) showing that C. pneumoniae can either initiate lesion development or cause exacerbation of lesions in rabbit and mouse animal models respectively. The association of this organism with atherosclerosis has also provided sufficient impetus to conduct a variety of human secondary prevention antibiotic treatment trials. The results of these studies have been mixed and, thus far, no clear long-lasting benefit has emerged from these types of investigations. Studies of C. pneumoniae pathogenesis have shown that the organism can infect many cell types associated with both respiratory and cardiovascular sites, including lung epithelium and resident alveolar macrophages, circulating monocytes, arterial smooth muscle cells and vascular endothelium. Infected cells have been shown to exhibit characteristics associated with the development of cardiovascular disease (e.g. secretion of proinflammatory cytokines and procoagulants by infected endothelial cells and foam cell formation by infected macrophages). More detailed analysis of C. pneumoniae pathogenesis has been aided by the availability of genomic sequence information. Genomic and proteomic analyses of C. pneumoniae infections in relevant cell types will help to define the pathogenic potential of the organism in both respiratory and cardiovascular disease.

Introduction

Atherosclerosis and its complications lead to half of all adult deaths in the United States and other western societies (McMillan, 1995; Ross, 1999). Atherosclerosis and cardiovascular disease (CVD) are multifactorial, highly complex diseases with numerous aetiologies simultaneously and sequentially collaborating in subtle ways to affect lesion development, progression and maturation to an advanced, disease-provoking entity. The lesion, or atheroma, is an inflammatory site composed of a necrotic lipid-rich core, modified vascular endothelium, smooth muscle cells, foamy monocyte/macrophages, lymphocytes and a variety of inflammatory mediators (Ross, 1999). A variety of risk factors are known to be associated with atherosclerosis and the pathogenesis of CVD. These include genetic and lifestyle factors such as elevated low-density lipoprotein (LDL) cholesterol levels, high blood pressure, hypoglycaemia, stress, smoking and obesity (Wilson et al., 1998). Traditional risk factors clearly contribute to CVD, but roughly 40% of cases have no well-defined risk factor associated with them. Recent appreciation of atherosclerosis as a chronic inflammatory disease (Ross, 1999) has rejuvenated efforts to examine the role played by infectious agents in CVD. Much of this interest has focused on Chlamydia pneumoniae infection because of its association with atherosclerosis by a variety of epidemiological and experiment-based studies (reviewed by Grayston, 2000 and discussed below).

Chlamydia pneumoniae, like all chlamydiae, is an obligate intracellular prokaryotic pathogen. Unlike C. trachomatis, the other major human chlamydial pathogen, which exhibits an in vivo tropism for mucosal epithelial cells, C. pneumoniae can infect and survive in a wider array of host cell types, including lung epithelium, resident macrophages (alveolar and monocyte derived), circulating monocytes, arterial smooth muscle cells and vascular endothelium (Shemer-Avni and Lieberman, 1995; Gaydos et al., 1996; Jahn et al., 2000). C. pneumoniae can infect and modify the physiology of the various cell types present in the lung, circulation and atheroma itself and may transit from the lung to the atheroma via circulating monocytes (Moazed et al., 1998; Lin et al., 2000) and lymphocytes (Kaul et al., 2000; Haranaga et al., 2001). The C. pneumoniae intracellular developmental cycle resembles that of other chlamydiae in its general features (summarized by Hahn et al., 2002). Chlamydial growth is biphasic, consisting of two alternating functional and morphological forms (Fig. 1). The elementary body (EB) is the metabolically inert, infectious form of the organism that is capable of transient extracellular survival. EBs bind to as yet undefined host cell receptors, are internalized via a pathogen-specified process and are detectable within a membrane-bound vesicle immediately after entry. This vesicle is capable of interacting with post-Golgi secretory vesicles in ways that allow for the incorporation of host phospholipids (Hackstadt et al., 1996; 1997). Chlamydiae also block intracellular host cell responses, such as fusion of the pathogen-containing endosome with lysosomes, and thus avoid host cell factors that would be detrimental to intracellular survival. Soon after entry, chlamydiae differentiate from infectious EB to the intracellular replicative form of the organism, referred to as the reticulate body or RB. This differentiation, which is dramatic in terms of altered chlamydial morphology, must reflect an orchestrated sequence of differential gene expression. Transformation of EB to RB results in loss of the disulphide cross-linking of the outer membrane complex, decondensation of the genome and initiation of DNA, RNA and protein synthesis. RB multiplication results in the formation of an intracellular microcolony (termed the inclusion) of chlamydiae. Chlamydiae remain within the confines of the host cell vesicle throughout their cycle of intracellular development but modify this structure in several ways during their growth phase. For example, chlamydial proteins (Inc category proteins) are secreted and inserted into the inclusion membrane (Bannantine et al., 1999). Chlamydiae also code for genes that are capable of producing a functional type III secretion system (Hsia et al., 1997). Inc proteins and proteins secreted into the host cell cytoplasm may play important roles in various host cell changes (e.g. cytokine and chemokine production, adhesion molecule expression, changes in MHC surface protein levels and changes in cell death-related protein expression) associated with chlamydial infection. The final stages of chlamydial growth during a productive infection involve differentiation of RB back to EB. This is accompanied by lysis of the host cell or direct release of EB.

Figure 1.

The developmental cycle of C. pneumoniae. The normal cycle may be disrupted by the induction of persistence as a result of nutritional deprivation, antibiotic treatment or growth in specific cell types and the appearance of large aberrant bacterial forms. Reactivation of the normal developmental cycle can be seen in many cases when the inducer of persistence is removed.

Productive infections are only one of several possible outcomes of chlamydial interactions within any given host cell. For example, treatment of infected host cells with certain antibiotics fails to eradicate chlamydiae but, instead, results in a chronic or persistent non-productive infection characterized by abnormal appearing RB that fail to mature into infectious EB (Matsumoto and Manire, 1970; Hammerschlag and Vuletin, 1985). Similarly, low nutrient stress can prevent chlamydiae from completing the developmental cycle in permissive host cells without leading to eradication of the organisms (Coles et al., 1993). It is important to distinguish in vitro models of persistence from chronic chlamydial infections in patients. The use of the term ‘persistent’, as defined by Beatty et al. (1994), refers to an aberrant intracellular chlamydial development characterized by a dramatic reduction in the number of infectious forms produced. Chronic infection has been more difficult to define given the contributions of several factors including low-level inapparent infection, repeated reinfection and the multiple conditions known to give rise to persistent infection in vitro. Long-term infections most probably occur through a complex multifactorial process that may be difficult to reproduce in the laboratory. Chronic chlamydial infections have been recognized in vivo for decades (Meyer and Eddie, 1933; Hanna et al., 1968), and physiological mediators of persistence include immune-regulated cytokines (Summersgill et al., 1995) such as gamma-interferon (IFNγ) and tumour necrosis factor alpha (TNFα). For example, intracellular development of C. pneumoniae can be inhibited in respiratory epithelial cells by treatment with IFNγ. The inhibition is augmented in the presence of TNFα (Mehta et al., 1998). The inhibition is not eradicative but rather reversible upon the addition of the essential amino acid tryptophan. The tryptophan decyclizing enzyme, indoleamine 2,3-dioxygenase (Mehta et al., 1998), has been shown to play a similar role to that shown for the development of persistence in C. trachomatis (Beatty et al., 1994). The descriptions of chlamydial persistence as induced by immune regulated cytokines also include links to pathogenesis, especially in chronic diseases, and particularly as a result of heightened expression of a key mediator of proinflammatory responses, GroEL.1 (Kol et al., 1998; Byrne and Kalayoglu, 1999; Mayr et al., 1999). Intracellular persistent forms of the organism are inherently more suited to chronic infections and are more difficult to eradicate by antibiotics. These attributes are consistent with involvement of chronic diseases such as atherosclerosis.

Chlamydial persistence occurs in cells treated with antibiotics or immune-regulated cytokines and cells deprived of essential nutrients and can be elicited as a function of the host cell type that is infected. For example, although productive C. pneumoniae infections occur in nutrient-replete epithelial, endothelial and smooth muscle cells, infection of mononuclear phagocytes leads to an uncharacterized form of persistent growth that can be reversed by co-culture with endothelial cells (Lin et al., 2000) or by treatment of infected monocytes with low levels of cycloheximide (unpublished observations). Persistent C. pneumoniae infection of monocytes, although associated with heightened resistance to antibiotics (Gieffers et al., 2001), remains otherwise uncharacterized.

Association of C. pneumoniae with atherosclerosis

Exposure to C. pneumoniae is extremely common, and infections occur repeatedly among most people. Antibody titres to C. pneumoniae are not common in individuals less than 5 years old, but up to 50% of individuals are seropositive by the age of 20 (Grayston, 2000). The prevalence of antibody titres continues to rise in populations of adults, reaching a peak of 80% of males and 70% of females being seropositive by the age of 65 (Grayston, 2000). Population prevalence studies demonstrate that infections are not restricted geographically and that reinfections occur frequently.

Evidence for the presence of the organism in atherosclerotic lesions has emerged from nearly 40 studies conducted by several different groups of investigators (e.g. Kuo et al., 1993a,b; 1995). Direct detection of organisms by immunohistochemistry, electron microscopy, in situ hybridization or amplification of chlamydial DNA by polymerase chain reaction (PCR) has shown that the organism is present in atheroma. In addition, viable organisms have been detected by amplifying mRNA transcripts from atheromas or by culturing infectious organisms from atherosclerotic tissue. Shor et al. (1992) were the first to demonstrate that C. pneumoniae was present in atherosclerotic lesions. These investigators used transmission electron microscopy (TEM) to detect C. pneumoniae in macrophage foam cells, followed by confirming studies on the same tissue using immunohistochemistry to detect C. pneumoniae antigens and PCR to detect C. pneumoniae-specific DNA. In a subsequent study (Kuo et al., 1995), histological and PCR-based evidence for C. pneumoniae in atherosclerotic lesions was found by evaluating autopsy tissue taken from individuals between the ages of 15 and 34 years. In another study, Muhlestein et al. (1996) examined atherosclerotic plaques for the presence of C. pneumoniae from patients undergoing coronary atherectomy and compared the results with two sets of control groups. Some 79% of the 90 atherectomy specimens had immunohistochemical evidence of C. pneumoniae. Only one control sample (4.2%) out of 12 normal coronary specimens and 12 transplant-induced arterial disease specimens showed evidence of C. pneumoniae. Strikingly, it was found that the presence of atherosclerotic lesions predicted the presence of C. pneumoniae with an odds ratio of 3. Additional studies have also shown that not only is C. pneumoniae frequently present in atheromatous lesions from diseased individuals, it is notably absent from non-diseased tissue or tissue unrelated to the site of disease in a given individual (Jackson et al., 1997).

Evidence for C. pneumoniae viability in atherosclerotic lesions was first provided by Ramirez (1996). This study examined excised hearts from individuals undergoing heart transplants. Three separate laboratories successfully cultured C. pneumoniae from one patient's coronary artery, and other investigators (Blasi et al., 1996; Muhlestein et al., 1996; Ong et al., 1996) have isolated C. pneumoniae from atheromas. However, successful isolation and propagation of the organism from atherosclerotic lesions remains an uncommon event. One possible explanation is that C. pneumoniae is present in the lesion in a non-cultivatable, persistent state (Beatty et al., 1994). Interestingly, Kol et al. (1998) have localized chlamydial heat shock protein 60 (GroEL.1), an inflammatory antigen abundantly expressed by persistent chlamydiae (Beatty et al., 1994), to macrophages present in human atherosclerotic lesions. Although the implication of this finding is not clear, persistent infection of the lesion results in the chronic production of GroEL.1 with subsequent inflammation and lesion progression.

Animal models have been used to define further the role of C. pneumoniae as a risk factor in atherosclerosis. Two such models have been studied so far. The first (Hu et al., 1999; Campbell et al., 2000) involves the use of hyperlipidaemic mouse strains that either spontaneously develop atherosclerosis as a result of hyperlipidaemia (ApoE–/– mice) or develop lesions when fed an atherogenic diet (LDL receptor–/– mice). When hyperlipidaemic mice from each strain were infected intranasally with C. pneumoniae, atheromas were more pronounced, and organisms could be recovered from the lesions. Furthermore, persistent infection was established in infected animals. In non-hyperlipidaemic animals, the organism was able to infect and induce inflammatory changes in the aorta, but the animals did not develop typical changes indicative of advanced atheromas. These results suggest that C. pneumoniae may exhibit a tropism for atherosclerotic tissue and that infection accelerates the development of disease in hyperlipidaemic animals. Importantly, atherosclerotic lesions in mice do not develop after infection with the sexually transmitted pathogen Chlamydia trachomatis, indicating that the effects are species specific to C. pneumoniae. The second animal model uses New Zealand white rabbits (Fong et al., 1997; Laitinen et al., 1997; Muhlestein, 2000): animals that do not develop atherosclerosis spontaneously unless fed a hyperlipidaemic diet were found to develop aortic changes consistent with early atherosclerotic lesions after respiratory infection with C. pneumoniae. The atypical bacterial pathogen Mycoplasma pneumoniae, which causes lung pathology similar to C. pneumoniae, fails to cause inflammatory changes or induce atherosclerotic lesions in rabbits (Fong et al., 1997). Taken together, these studies strongly suggest that C. pneumoniae can target vasculature, induce inflammation and initiate or promote lesion development in animal models of atherosclerosis. Both murine and rabbit models have been extended to include antibiotic treatment of infected animals. Muhlestein et al. (1998) first reported that azithromycin treatment prevented accelerated atherosclerosis in hyperlipidaemic rabbits infected with C. pneumoniae. Recent investigations by Fong (2000) on non-hyperlipidaemic rabbits showed that azithromycin and clarithromycin were also highly effective in preventing atherosclerosis-like changes

Direct association of C. pneumoniae infection with atherosclerosis can be assessed by treating CVD patients with appropriate antibiotics and determining a difference in CVD-related events in the treated versus control populations. Three small-scale secondary prevention trials have been conducted on individuals with CVD. Gupta et al. (1997) administered a short course of placebo or azithromycin (500 mg day−1) to male survivors of an acute myocardial infarction (MI) who had high antibody titres to C. pneumoniae. They observed that, after a mean follow-up of 18 months, patients with high antibody titres were fourfold more likely to suffer adverse cardiovascular events and that azithromycin treatment of these individuals significantly reduced the occurrence of these events (28% versus 8%, P = 0.03). This pilot trial was expanded in the ACADEMIC study (Muhlestein et al., 2000), which randomized more CVD patients (300 versus 60) into longer (3 months) placebo versus azithromycin treatment groups. This study was designed to detect marked (>50%) reductions in cardiovascular events. The authors observed a trend towards benefit after 12–18 months of follow-up, but no significant difference between the control and treatment groups. A third treatment trial (ROXIS; Gurfinkel et al., 1997) randomized 202 patients presenting with unstable angina or non-Q-wave MI to placebo or roxithromycin (150 mg twice daily for 30 days). The authors observed a statistically significant reduction in recurrent angina, MI and death in patients treated with roxithromycin after 1 month of follow-up (2% versus 9%, P = 0.032). However, these differences were less pronounced after 6 months follow-up (8.7% versus 14.6%, P = 0.26). The designs of these trials were not perfect (Grayston, 1998; Anderson et al., 1999), and the studies recruited too few patients to detect moderate differences after treatment. In addition, in light of new evidence suggesting that C. pneumoniae within monocytes is refractory to standard antichlamydial therapy (Gieffers et al., 2001), the antibiotic dosing and regimen used in these studies may be suboptimal to eradicate persistent infection. The antibiotics may also have anti-inflammatory properties (Grayston, 1998; Anderson et al., 1999; Anderson and Muhlestein, 2000) that confound interpretation of the data. However, success in reducing secondary events using antibiotics with known efficacy against C. pneumoniae does add to the body of data implicating C. pneumoniae as an emerging risk factor in atherosclerosis. Large antibiotic treatment trials in patients with CVD are being conducted to address several caveats raised in the above studies. The Azithromycin and Coronary Events Study (ACES) is sponsored by the National Institutes of Health and is recruiting 4000 patients with coronary artery disease (CAD; Jackson, 2000). The Weekly Intervention with Zithromax for Atherosclerosis and Related Disorders (WIZARD) trial is testing the effect of azithromycin versus placebo in 3500 seropositive patients with CAD (Dunne, 2000). The Azithromycin in Acute Coronary Events Study (AZACS) trial of 1400 patients and the Pravastatin or Atorvastatin Evaluation and Infection Therapy (PROVE-IT) trial recruiting 4200 patients are also under way. Data gathered from these large clinical trials may strengthen the association between C. pneumoniae and atherosclerosis; however, pathogenic mechanisms must be better defined to design trials with the optimal antibiotic regimen, dosing, treatment period and follow-up.

These studies add credence to the notion that C. pneumoniae infection is linked to the development of atherosclerosis but have specific limitations that must be considered (Quinn, 1998). The high prevalence of C. pneumoniae infection makes it exceedingly difficult to identify true differences in seropositivity between cases and controls. Indeed, most patients with cardiovascular disease and their age-matched controls are in an age group where C. pneumoniae prevalence nears 80%. Most of these studies detected anti-C. pneumoniae antibodies using the microimmunofluorescence test, known to have poor reproducibility from laboratory to laboratory (Peeling et al., 2000). A number of studies used chlamydial immune complexes or cLPS to detect infection. In these studies, cross-reactions with other antigens, such as cardiolipin, itself associated with CVD (Rebic et al., 1993), may explain in part the observed association. Measurement of immune responses to specific chlamydial antigens, especially members of the heat shock protein family, has provided more convincing seroepidemiological data to implicate the possibility that exposure to chlamydial GroEL1 represents an independent risk factor for coronary atherosclerosis and disease development (Mayr et al., 1999; Ciervo et al., 2002; Fong et al., 2002; Huittinen et al., 2002; Mahdi et al., 2002; Biasucci et al., 2003). Risk factors for C. pneumoniae infection are not known with certainty; therefore, residual confounding variables may explain why only some studies show a positive association. For example, several retrospective studies did not adjust for smoking, an important CVD risk factor and a potentially important risk factor for C. pneumoniae infection (Hahn and Golubjatnikov, 1992), and many prospective studies did not consider socioeconomic status. Finally, the enrolment of high-risk patients in these studies may be problematic if successful antibiotic intervention is dependent on elimination of C. pneumoniae at earlier stages of disease progression.

Pathogenic mechanisms

Chlamydia pneumoniae has the capacity to initiate and propagate inflammation in ways that contribute to atherosclerosis. The pathogen probably accesses the vasculature during local inflammation in a lower respiratory tract infection. C. pneumoniae DNA has been detected in CD3+ T lymphocytes (Kaul et al., 2000) and monocytes (Maass et al., 2000) recovered from peripheral blood, and both cell types may serve as vehicles of transport (Moazed et al., 1997). The organism disseminates systemically but exhibits tropism to arterial vasculature (Gaydos, 2000). The pathogen can infect and multiply within all cell types commonly found in the atheroma, including coronary artery endothelial cells, macrophages and aortic artery smooth muscle cells (Kaukoranta-Tolvanen et al., 1996; Fryer et al., 1997; Moazed et al., 1997; Dechend et al., 1999; Gaydos, 2000; Maass et al., 2000). Interestingly, a monocyte cell line infected with C. pneumoniae is capable of transmitting the pathogen to coronary artery endothelial cells in culture (Gaydos, 2000), although the mechanism of transfer is not clear.

Infected cells subsequently upregulate adhesion molecule expression and produce inflammatory cytokines. For example, infected endothelial cells augment the expression of adhesion molecules that may promote leucocyte adherence, migration and intimal inflammation (Liuba et al., 2000). There is also direct evidence that infection of human endothelial cells stimulates transendothelial migration of neutrophils and monocytes (Molestina et al., 1999). Endothelial cell infection triggers secretion of the chemokine interleukin (IL)-8 and the procoagulant tissue factor, IL-6 and plasminogen activator inhibitor-1 (Fryer et al., 1997; Dechend et al., 1999), suggesting that C. pneumoniae may facilitate recruitment of inflammatory cells and modulate procoagulant activity respectively. Smooth muscle cells respond to endothelial infection by proliferating (Coombes and Mahoney, 1999; Miller et al., 2000; Summersgill et al., 2000), and direct infection of smooth muscle cells induces secretion of cytokines such as IL-6 and basic fibroblast growth factor (bFGF) that may alter atheroma biology (Coombes and Mahoney, 1999). Infected human macrophages secrete enhanced levels of inflammatory cytokines such as TNFα, IL-1β, IL-6, monocyte chemoattractant protein-1 (MCP-1), macrophage inflammatory protein 1 alpha (MIP-1α) and IL-12 that may promote lesion progression (Miller et al., 2000; Netea et al., 2000), and IL-10 that may prevent apoptosis to perpetuate inflammation (Geng et al., 2000; Kothe et al., 2000). Recent studies suggest that C. pneumoniae triggers specific cell-mediated immunity within plaques, as evidenced by the detection of chlamydia-specific T lymphocytes in atherosclerotic lesions (Halme et al., 1999; de Boer et al., 2000; Mosorin et al., 2000). These cells appear to be primarily CD4+ Th1 subtypes and may contribute to plaque destabilization by T-cell cytokine production. Plaque destabilization may proceed through more direct mechanisms as well; C. pneumoniae enhances the production of matrix-degrading metalloproteinases by infected macrophages (Kol et al., 1998; Vehmaan-Kreula et al., 2001). Another way in which C. pneumoniae may influence atheroma biology is by modulating macrophage–lipoprotein interactions. Infected macrophages ingest excess lipoprotein to become foam cells, the hallmark of early lesions in atherosclerosis (Kalayoglu and Byrne, 1998a,b; Kalayoglu et al., 1999a,b). In addition, C. pneumoniae induces monocytes to oxidize lipoproteins and make them atherogenic (Kalayoglu et al., 1999b; 2000).

Genomic analysis of C. pneumoniae

Genomic sequence information is available for three strains of C. pneumoniae from patients with respiratory disease [Kalman et al., 1999 (CWL029); Read et al., 2000 (AR39); Shirai et al., 2000 (J138)]. Sequence comparisons between isolates indicate a remarkably high level of synteny (i.e. gene order and content is highly conserved) with an overall sequence identity of > 99.9%. Comparisons between AR39 and CWL029 (Read et al., 2000) indicate a difference of ≈ 300 single-nucleotide polymorphisms (SNPs), a small number of deletions/insertions and the presence in AR39 of a small, circular single-stranded (ss)DNA bacteriophage. Comparison of J138 with CWL029 (Shirai et al., 2000) indicates a similar number of SNPs and eight regions of difference in which one of the strains has a deletion/insertion not found in the other strain. Despite this between-strain homogeneity, a relatively high level of polymorphism was reported in strain AR39 (no mention of polymorphism was made in the other two publications). The C. pneumoniae AR39 genome has 304 areas of polymorphism, which range from SNPs (which constitute the majority of variant sequences) to a duplicated region of 1649 bp (encoding one copy of the tyr P gene). A list of reported polymorphisms is provided in Table 1. As the AR39 and CWL029 genomes differ by approximately the same number of SNPs as found in the AR39 population alone, genotypic identification based on SNPs alone would be problematic. This type of intrastrain gene polymorphism has not been observed for sequenced C. trachomatis strains (Kaulman et al., 1999; Read et al., 2000). The reason for this striking difference between C. pneumoniae and C. trachomatis strains is unknown. A distinct possibility exists that C. pneumoniae infections result from polyclonal mixtures of bacteria that arise because of the extremely high incidence of infection. This could explain in part the observed genome polymorphisms. These observations do, however, provide a rationale for analysing clonal variants with a goal of understanding whether genotypic distinctions reflect phenotypic changes that can be associated with differences in the pathogenic potential of these organisms.

Table 1. . Inter- and intrastrain genetic polymorphisms in C. pneumoniae (from Kalman et al., 1999; Read et al., 2000; Shirai et al., 2000).
GeneFunctionPolymorphism
tyr P/yccATyrosine permeaseDeletion of one unit of a tandem 1649 bp repeat (AR-39, J-138)
pmp 6Polymorphic outer membrane proteinLoss of a 393 bp iterated segment (AR-39, J-138)
udk (upstream)Promoter of uridine kinase gene236 bp inverted sequence (AR-39)
CPn 0007Hypothetical protein330 bp deletion (J-138)
htrBLauroyl acyl transferase40 bp insertion (J-138)
pfkAPyrophosphate-fructose 6-phosphate
1-phosphotransferase
89 bp deletion (J-138)
CPn 3050/troAIntergenic85 bp insertion (J-138)
Cpn 0912Hypothetical protein27 bp insertion (J-138)
Cpn 0978Hypothetical proteinWhole gene deletion(J-138)
SNPs670 total (AR-39, CWL029, J-138)
SNPs286 (AR-39)

A focus-forming assay for C. pneumoniae has been described recently (Gieffers et al., 2002). This procedure differs somewhat from the plaque assay described previously for C. trachomatis, but the growth characteristics that require the more complicated focus-forming assay to isolate clones are unknown. In fact, differences exist between C. pneumoniae clonal isolates in that some have the ability to form plaques while others do not (and therefore can only be isolated using the focus-forming assay; unpublished observations). These differences are probably related to the lysis of host cells at the conclusion of the developmental cycle. Presumably, this phenotypic difference is the result of one or more of the reported genotypic differences. The establishment of clonal isolates (by either focus-forming or plaque assays) is of paramount importance in understanding the differences between specific disease-associated isolates. Clonal isolates should be isolated and characterized before using genome-based analysis methods on the isolate, e.g. genomic sequencing or microarray analyses.

Gieffers et al. (2002) have shown recently that mixed populations can be used to isolate individual clones that differ in the number of tandemly duplicated tyrP genes. Furthermore, extending their analysis of this polymorphic allele, they have shown in a preliminary study that CVD C. pneumoniae isolates have a single copy of this gene, whereas respiratory isolates have mixed genotypes containing one, two or three tandem copies of the gene (Gieffers et al., 2003). They were unable definitively to associate this genotype with a growth phenotype, but the differences may be too subtle to detect in vitro, perhaps requiring a more sensitive measure of bacterial growth and persistence (e.g. an animal model system). Although the comparison of respiratory and CVD isolates needs to be expanded (more clinical isolates from geographically diverse sources) and examined more closely, it suggests that perhaps not all C. pneumoniae clones are associated with CVD. This difference in disease-causing potential could arise as a result of selective bottlenecks associated with dissemination of the organism to cardiovascular tissue or establishment of infection in cardiovascular tissues (Fig. 2). One growth characteristic likely to be associated with long-term disease in cardiovascular tissue is the tendency or ability of the isolate to establish persistent infections. The application of genomic analyses to clinical isolates should provide useful information on the development of therapeutic approaches that target isolates most likely to be associated with CVD, and should help in the development of diagnostic tools that indicate the presence of isolates associated with the development of CVD.

Figure 2.

The polyclonal nature of C. pneumoniae strains causing respiratory disease can be demonstrated by the presence of distinct mixtures of bacterial genotypes (as described in Table 1) in strain isolates. Certain genotypes may exhibit a selective advantage in the ability to disseminate or establish infection in cardiovascular sites.

Future directions

The idea that infections contribute to atherosclerotic plaque development and CVD is not a new concept, but it is one that is extremely difficult to prove categorically. As with other multifactorial, chronic disease processes, the situation is immensely complicated, and probably no single experiment will prove or disprove the hypothesis. A large body of associative and correlative evidence exists for the involvement of C. pneumoniae infection in some aspects of the acceleration of CVD. This accumulation of evidence is perhaps the best answer that can be found at this time. The involvement of an infectious agent in the progression of a disease rather than the initiation aetiology represents a subtle shift in the infectious disease paradigm but is probably an accurate assessment of the role of C. pneumoniae in CVD. Clearly, well-defined studies on bacterial pathogenesis will add to this body of evidence in important ways. The identification and implementation of immunological or therapeutic strategies that result in eradication of the organism in both respiratory and cardiovascular sites may be the only way to answer the question irrefutably.

Ancillary