Autoimmunity to Human Heat Shock Protein 60, Chlamydia pneumoniae Infection, and Inflammation in Predicting Coronary Risk
Abstract— Heat shock protein 60 (Hsp60) and Chlamydia pneumoniae infection have both been associated with cardiovascular diseases. Our aim was to study the role of Hsp60 antibodies as coronary risk predictors and their association with C pneumoniae infection and inflammation. This was a prospective, nested, case-control study. The cases consisted of 239 middle-aged Finnish men who developed myocardial infarction or coronary death during the follow-up. Baseline levels of IgA and IgG antibodies to human-specific and C pneumoniae-specific Hsp60 were measured by enzyme immunoassay. Human Hsp60 IgA, but not IgG or C pneumoniae Hsp60, antibodies were a significant risk factor for coronary events (odds ratio 2.0, 95% CI 1.1 to 3.6, when the fourth and first quartiles are compared). When an elevated human Hsp60 IgA antibody level (above the second quartile) was present simultaneously with a high C pneumoniae IgA antibody level (the third quartile) and an elevated C-reactive protein level (the second quartile), compared with all factors at low levels, the risk was 7.0 (95% CI 2.6 to 19.1) without adjustment and 5.0 (95% CI 1.8 to 14.2) when adjustment was made for age and smoking. In conclusion, an elevated human Hsp60 IgA antibody level was a risk factor for coronary events, especially when it was present together with C pneumoniae infection and inflammation.
Heat shock proteins (Hsps) are a class of evolutionarily highly conserved proteins that are produced by cells in response to stress or injury, including infection, to protect themselves from unfavorable conditions. Bacterial Hsps have a high sequence homology with their human counterparts, and this has been suggested, in the course of bacterial infections, to lead to autoimmune processes predisposing to autoimmune diseases.1
Atherosclerosis is now considered an inflammatory disease, and humoral and cellular immune mechanisms may both be involved in the development of atherosclerotic lesions.2 Studies from Wick’s laboratory3–11⇓⇓⇓⇓⇓⇓⇓⇓ have proposed the hypothesis, based on animal experiments3,4⇓ and human studies,5–10⇓⇓⇓⇓⇓ that the host immune response to microbial Hsp60 may lead to the development of atherosclerosis.11 Chlamydia pneumoniae infection has also been linked with atherosclerosis. Several lines of evidence speak for the association: there is a seroepidemiological association between C pneumoniae infection and cardiovascular diseases,12 C pneumoniae is present in >50% of the lesions but only seldom in healthy areas,12 and C pneumoniae infection induces atherosclerotic lesions in experimental animals.13–15⇓⇓ Hsp60 seems to play a key role in chronic chlamydial infections: it is overproduced in persistent infections, and it has been associated with the hypersensitivity and immunopathology seen in these infections.16 In human atherosclerotic plaques, chlamydial Hsp60 is found colocalized with human Hsp60, and the expression of chlamydial and human Hsp60 is increased in atherosclerotic compared with nonatherosclerotic tissues.17 Chlamydial Hsp60 protein has also been found to induce a positive response in 71% of the C pneumoniae-specific T-cell lines generated from tissue specimens of carotid atheroma.18
In our previous study,19 we showed that C pneumoniae-specific IgA antibodies and immune complexes as markers of a chronic, active, or recurrent infection were associated with an increased risk of coronary events and that the simultaneous presence of a high C-reactive protein (CRP) level, a marker of inflammation, substantially increased the risk. In the present study, our purpose was to investigate the role of antibodies against human Hsp60 in the same population and to determine whether these antibodies are associated with inflammation and C pneumoniae infection and whether they predict the risk of coronary events alone or in combination with other risk factors.
The participants for this prospective nested case-control study were selected from the trial cohort of the Helsinki Heart Study, which has been described in detail previously.20,21⇓ Briefly, the Helsinki Heart Study was a double-blind, randomized, placebo-controlled, coronary primary prevention trial with gemfibrozil in dyslipidemic (non-HDL cholesterol >5.2 mmol/L) middle-aged men in Finland, beginning in the year 1980. Subjects with a history of coronary heart disease or any other major illness were excluded. Baseline sera were collected from all 4081 participants. Data on smoking habits and body mass index and blood pressure were recorded. Total and HDL cholesterol and triglyceride levels were determined from baseline sera as described elsewhere.20,21⇓ During the 8.5-year follow-up, 241 of the 4081 participants suffered a coronary event, either a nonfatal myocardial infarction or coronary death. The controls were chosen from participants who completed the study period without coronary events and were individually matched with 241 cases for treatment (gemfibrozil/placebo) and region of residence. The geographical matching was considered necessary because of the regional exposure of some infectious agents studied. For the present study, baseline sera were available from 239 case-control pairs. All sera were not available for all measurements. Characteristics of the study population are presented in Table 1.
An enzyme immunoassay (EIA) was used to measure immunoglobulin class-specific antibodies to C pneumoniae-specific and human-specific Hsp60 in the baseline sera, as described in detail elsewhere.22 The measurements were done in a blinded fashion. In brief, microtiter plates were coated with a recombinant C pneumoniae Hsp60 protein produced in Bacillus subtilis (U. Airaksinen, unpublished data, 2002) and a recombinant human Hsp60 produced in Escherichia coli (Sigma Chemical Co), both at a concentration of 5 μg/mL, in PBS overnight at 37°C. The hsp60 gene of C pneumoniae was amplified by polymerase chain reaction from a genomic DNA extract from the Kajaani 6 strain of C pneumoniae. The polymerase chain reaction product was cloned into an expression vector, pKTH39, and transformed into B subtilis. A C-terminal histidine tag (His6) allowed purification of the produced protein by Ni2+ chelate affinity chromatography (Qiagen Inc). On the basis of preliminary experiments, baseline sera diluted 1:50 for IgA and 1:200 for IgG antibodies in PBS containing 10% FBS were added, and the plates were incubated for 2 hours at 37°C. The plates were then incubated for 2 hours at 37°C with alkaline phosphatase-conjugated anti-human IgA and anti-human IgG. Finally, the plates were incubated at 37°C for 30 minutes with a substrate solution containing 1 mg p-nitrophenyl phosphate disodium in 1 mL carbonate MgCl2 buffer. Absorbance was measured against blank at 405 nm, and the results were expressed as EIA units (EIU) by multiplying the optical densities by 100.
C pneumoniae IgA and IgG serum antibodies have been analyzed previously19 by the microimmunofluorescence (MIF) method, as described in detail elsewhere,23 with the elementary bodies of Kajaani 6 strain of C pneumoniae used as antigen.
CRP concentrations have been measured previously19 by the use of EIA (UC CRP ELISA, Eucardio Laboratory).
Because of skewed distributions, the variables were categorized by quartiles as follows: 1, ≤7.4 EIU; 2, 7.5 to 10.5 EIU; 3, 10.6 to 14.8 EIU; and 4, ≥14.9 EIU for human Hsp60 IgA; 1, ≤16.1 EIU; 2, 16.2 to 26.6 EIU; 3, 26.7 to 47.3 EIU; and 4, ≥47.4 EIU for human Hsp60 IgG; 1, ≤10 EIU; 2, 10.1 to 15.7 EIU; 3, 15.8 to 24.2 EIU, and 4, ≥24.3 EIU for C pneumoniae Hsp60 IgA; 1, ≤32.3 EIU; 2, 32.4 to 49.3 EIU; 3, 49.4 to 86.1 EIU; and 4, ≥86.2 EIU for C pneumoniae Hsp60 IgG; 1 and 2, 5 MIF titer; 3, 10 to 20 MIF titer; and 4, ≥40 MIF titer for C pneumoniae IgA; 1, 16 MIF titer; 2, 32 MIF titer; 3, 64 to 128 MIF titer; and 4, ≥256 MIF titer for C pneumoniae IgG; and 1, ≤0.93 mg/L; 2, 0.94 to 1.70 mg/L; 3, 1.71 to 3.79 mg/L; and 4, ≥3.80 mg/L for CRP. For C pneumoniae IgA antibodies, the lower 2 quartiles were combined, because all the MIF titers in those 2 quartiles were equal to 5. As cutoff levels for all Hsp60 antibody levels and for the CRP level, we used the third quartiles to define elevated levels, but for C pneumoniae MIF titers, we used the fourth quartiles (a high level), because our previous studies have indicated that only the highest levels of this common pathogen involve an increased risk of coronary events.19 We studied the associations between the putative risk factors by cross tabulations and tested for trend by using logistic regression analysis on the same variables, eg, for human Hsp60 and CRP, the dichotomous variable indicating whether CRP was elevated or not and the variable indicating the quartiles of human Hsp60. In the analysis for trend, this latter explanatory variable was used as continuous.
To maintain the matching of the case-control design when the risks of coronary events were evaluated, conditional logistic regression analysis was used to estimate the odds ratios (ORs) by quartiles of the putative risk factors. The ORs associated with the quartiles were estimated simultaneously in the multivariate model. Even when the ORs were estimated from the matched data, we used the pooled data to calculate the numbers of cases and controls in the different risk factor categories to give an additional description of the risk pattern. When joint effects were studied, a new categorical variable was constructed to represent the different combinations of the original variables. For example, in the case of smoking and human Hsp60, a 4-level variable was used as follows: (1) smoking negative and Hsp negative, (2) smoking negative and Hsp positive, (3) smoking positive and Hsp negative, and (4) smoking positive and Hsp positive.
Associations Between Factors
An elevated level of IgA antibodies to human Hsp60 was present in 54% of the cases and in 45% of the controls (P=0.040). Other Hsp60 antibodies were found equally often in the cases as in the controls. There was a good correlation between the human Hsp60 IgA antibodies and their biological analogue, IgA antibodies against C pneumoniae Hsp60 (r=0.558, P<0.001). When cases and controls were analyzed separately, the correlation was stronger among the cases (r=0.610, P<0.001) than among the controls (r=0.510, P<0.001). Sixty-two percent of the cases and 54% of the controls who had a human Hsp60 IgA antibody level in the first quartile of distribution also had a C pneumoniae Hsp60 IgA antibody level in the first quartile. At the other end, 57% of the cases and 51% of the controls who had a human Hsp60 IgA antibody level in the highest quartile also had a C pneumoniae Hsp60 IgA antibody level in the highest quartile.
An elevated IgA antibody level of the C pneumoniae bacterium was found in 26% of the cases and in 17% of the controls (P=0.024). The corresponding IgG antibodies were almost equally common in the cases (19%) as in the controls (15%). Human Hsp60 IgA antibodies may be partly due to a C pneumoniae infection: 67% of the cases with IgA antibodies to C pneumoniae in the highest quartile also had an elevated IgA antibody level to human Hsp60, whereas the corresponding percentage in the controls was 50% (Table 2). Among the cases, the percentage of persons with an elevated human Hsp60 IgA antibody level increased from 48% in the lowest 2 quartiles of C pneumoniae IgA antibodies to 67% in the highest quartile (P=0.014), whereas no such trend was seen among the controls.
An elevated CRP level was present in 63% of the cases and in 36% of the controls (P<0.001). In the cases, there seems to be an association between the human Hsp60 IgA antibody level and the CRP concentration. The percentage of cases with an elevated CRP level increased from 54% in the lowest quartile of the human Hsp60 IgA antibodies up to 72% in the highest quartile; the trend was close to statistical significance (P=0.072, Table 2).
Univariate Associations With Coronary Risk
An elevated level of IgA antibodies to human Hsp60 was a significant risk factor for coronary events. There was a 50% increase in risk among those in the second quartile compared with those in the first quartile, and the risk increased to 2-fold among those in the third and fourth quartiles (Table 3). No increment in coronary risk was associated with IgA antibodies against C pneumoniae Hsp60 in spite of the good correlation with IgA antibodies to human Hsp60. In case of IgA antibodies to C pneumoniae, the risk was present only among those in the highest quartile, as already published in our previous study.19 None of the studied IgG antibodies was associated with coronary risk (data not shown). When the risk factors included in the present study were studied one at a time, an elevated CRP level unquestionably was found to be the greatest risk factor. With the lowest quartile used as a reference, the risk was 4.5-fold in the highest quartile. The adjustment for age and smoking decreased the risk by 1 unit.
Joint Effects on Coronary Risk
Because the factors studied tended to be associated especially among the cases, we explored the effect of their simultaneous occurrence on the risk of coronary events. A high IgA level of C pneumoniae antibodies and an elevated IgA level of human Hsp60 antibodies were both slightly associated with coronary risk. When these elevated levels were present simultaneously, the risk increased clearly compared with the risk associated with low levels of both antibodies (Table 4). Among those with a low CRP level, no risk was associated with an elevated IgA antibody level to human Hsp60. On the contrary, among those with elevated CRP, an elevated level of antibodies to human Hsp60, compared with a low level of human Hsp60 antibodies, nearly doubled the risk. In cases with elevated levels of both risk factors, compared with those with both factors at low levels, the risk was 4.6-fold without adjustment and 3.6-fold when adjustments were made for age and smoking.
We then combined all the 3 factors: CRP concentration and IgA antibody levels of C pneumoniae and human Hsp60. Table 5 shows clearly that IgA antibodies to C pneumoniae and to human Hsp60 are not risk factors unless the CRP level is elevated. Compared with the risk with all 3 factors at low levels, the risk for an elevated CRP level alone is 2.5. The risk does not change when a high C pneumoniae IgA antibody level is taken along, but an elevated human Hsp60 IgA antibody level increases the risk to 4-fold. When all the 3 risk factors are present simultaneously, compared with all factors at low levels, the risk is 7-fold without adjustment and 5-fold when adjustments are made for age and smoking.
We also compared the risks of coronary events among those having 1, 2, or 3 of the factors considered (a high level of IgA antibodies to C pneumoniae, an elevated level of IgA antibodies to human Hsp60, and an elevated level of CRP concentration). There was a clear dose-response pattern with an increasing number of risk factors studied. When those with all the 3 factors at low levels were used as a reference, the ORs were 1.75 (95% CI 1.02 to 3.01), 2.82 (95% CI 1.50 to 5.33), and 7.15 (95% CI 2.69 to 18.99) for those with 1, 2, and 3 simultaneous risk factors present, respectively. Adjustment for age and smoking decreased the ORs to 1.44 (95% CI 0.82 to 2.54), 2.15 (95% CI 1.10 to 4.21), and 4.51 (95% CI 1.64 to 12.37), respectively. Among those with only 1 factor, in 45% of the persons it was CRP, in 7% it was C pneumoniae IgA antibodies, and in 49% it was human Hsp60 IgA antibodies. Among those with 2 factors present simultaneously, in 20% of the persons they were CRP and C pneumoniae IgA antibodies, in 66% they were CRP and human Hsp60 IgA antibodies, and in 14% they were C pneumoniae and human Hsp60 IgA antibodies.
Role of Smoking
Adjustment for smoking in the statistical analyses decreased the ORs considerably (Tables 3 through 5⇑⇑). Therefore, we studied the possible modifying effect of smoking on the risk associated with elevated levels of IgA antibodies to human Hsp60 (Table 6). Among nonsmokers, the OR for an elevated level of human Hsp60 antibodies was 1.81 (95% CI 0.99 to 3.28), whereas the OR for smoking alone was 2.82 (95% CI 1.57 to 5.07). Thus, the actual joint effect, OR 3.97 (95% CI 2.15 to 7.31), was somewhat closer to being additive than multiplicative.
In the present study, we showed that the presence of an elevated level of IgA antibodies against human Hsp60 protein predicts a coronary event several years before the coronary event actually occurs. Researchers at Wick’s laboratory have studied the role of microbial Hsp60 in the development of atherosclerosis. Their studies indicate that immunization with mycobacterial Hsp65 induces atherosclerosis in laboratory animals.3,4⇓ In humans, they have found that immunity to mycobacterial Hsp65 is associated with the pathogenesis of carotid5 and coronary6 atherosclerosis. Hsp65 antibodies were found more often in patients with atherosclerotic lesions than in persons without such lesions,5 and the presence of these antibodies also predicted carotid atherosclerosis.7,8⇓ It has been shown that these antibodies cross-react with E coli Hsp60, chlamydial Hsp60, and human Hsp60 antibodies and are cytotoxic to endothelial cells.9 In the present study, we confirmed these findings and extended them to apply to coronary events: autoimmunity to human Hsp60 predicts myocardial infarctions and coronary deaths.
Even though we showed that there is a good correlation between the antibodies to human-specific and C pneumoniae-specific Hsp60, only IgA antibodies to human Hsp60 were associated with the risk of future coronary event. However, high levels of C pneumoniae IgA antibodies were also associated with the risk of coronary events, and there was an association between human Hsp60 and C pneumoniae IgA antibodies. This finding suggests that the antibodies to human Hsp60 may have been induced, at least partly, by C pneumoniae Hsp60 during chronic and repeated C pneumoniae infections, but the risk for coronary events is present only when autoimmunity to human Hsp60 has developed. Because of the high sequence homology between bacterial, chlamydial, and human Hsps (75% at the amino acid level), it is naturally possible that other bacterial agents, especially Helicobacter pylori and dental bacteria, which have been associated with coronary heart disease, have contributed to the development of autoimmunity to human Hsp60.24
Very little is known of the role of Hsp60 in the pathogenesis of C pneumoniae infections, but there is much evidence of the association between antibodies to Chlamydia trachomatis Hsp60 and the development of immunopathological damage (blinding trachoma, pelvic inflammatory disease, ectopic pregnancy, and tubal infertility) after repeated and persistent C trachomatis infections.25 Furthermore, it has been shown earlier that antibodies to the conservative peptide of C trachomatis Hsp60 are associated with autoimmunity to human Hsp60 and with infertility as chronic sequelae of C trachomatis infection.26 Unfortunately, the role of the corresponding peptide of C pneumoniae Hsp60 in autoimmunity and in the development of atherosclerosis has not been studied.
In our previous study,19 no risk was associated with a high level of C pneumoniae IgA MIF antibodies unless they were present together with a high CRP level, a marker of systemic inflammation. In the study of Kiechl et al,8 it was shown that among subjects with chronic infections, the risk for atherosclerosis increased clearly when CRP levels exceeded 1 mg/L. In the present study, a similar joint effect pattern was demonstrated for CRP and human Hsp60 IgA antibodies. IgA antibodies to C pneumoniae and to human Hsp60 were not risk factors unless the CRP level was elevated. C pneumoniae antibodies are common in middle-aged and elderly populations, and they may have been induced by past, silent persistent, or active chronic infections. Our findings suggest that IgA antibodies to C pneumoniae, together with an elevated CRP level and an elevated level of antibodies to human Hsp60, as a marker of an autoimmunity reaction, indicate active C pneumoniae infection.
Interestingly, no multiplicative joint effect was seen between herpes simplex type 1 virus antibody levels, which, in our previous study,19 were a significant risk factor for coronary events, and human Hsp60 antibody levels (data not shown). This supports the interpretation that an Hsp60 response is essential for the C pneumoniae infection in the process leading to coronary heart disease, whereas in viral infections, an elevated Hsp60 antibody level is a coexisting phenomenon. This is in agreement with the study of Mayr et al,10 who found that IgA antibodies to C pneumoniae were associated with carotid and femoral atherosclerosis and were correlated with antibodies to mycobacterial Hsp65, whereas they found no association between antibodies to cytomegalovirus and to Hsp65. In the present study, the associations between C pneumoniae antibodies, human Hsp60 antibodies, and CRP were strong, especially among the cases. The fact that the associations were stronger among the cases than among the controls lends some support to the hypothesis that these factors might contribute to the process leading to coronary events.
Several recent reports have suggested that IgG antibodies to C pneumoniae are not associated with the risk of coronary heart disease.27–29⇓⇓ Similarly, we did not find any risk to be associated with IgG antibodies against C pneumoniae, C pneumoniae Hsp60, or human Hsp60, suggesting that the presence of IgG antibodies is not a good marker for chronic C pneumoniae infection. Our present and previous reports have indicated that a high level of IgA antibodies is a better marker for chronic C pneumoniae infection in connection with coronary disease.19,30⇓ The half-life of IgA antibodies is short, whereas IgG antibodies may still be found several years after the infection. On the other hand, previous data indicate that the presence of IgA antibodies in serum samples collected 1 to 2 years apart is an even better marker for chronic infection than the presence of IgA antibodies in a single serum sample.30–32⇓⇓ However, the testing of paired sera at such a long interval may not be feasible in clinical settings. Xu et al,7 who found that mycobacterial Hsp65 antibody titers predicted 5-year mortality, measured total antibodies (all immunoglobulin classes together) to Hsp65; thus, we do not know whether IgA antibodies, rather than IgG antibodies, were also an important risk factor in that study.
C pneumoniae infection seems to be more common in smokers than in nonsmokers, suggesting that smoking predisposes one to the development of chronic C pneumoniae infection.32,33⇓ In identical twins, C pneumoniae-specific IgA levels were higher, and cell-mediated immunity (measured as a lymphoproliferation response) was lower in the smoking twins than in their nonsmoking cotwins, indicating that chronic C pneumoniae infections in smokers are linked to lowered cell-mediated protective immunity.34 In our earlier study,19 smoking significantly increased the risk associated with markers of chronic C pneumoniae infection, and its effect with high levels of CRP was more than multiplicative, showing a high degree of synergy. In the present study, the joint effect of smoking and human Hsp60 IgA antibodies was not negligible, but no synergistic effect was seen. Rather, our findings suggest that these 2 factors may, at least partly, act independently.
The limitations of the present study were that no paired sera were available. We tested a post hoc hypothesis in dyslipidemic middle-aged white males, and the results must be confirmed in other races and in women. The case-control status is naturally time dependent, inasmuch as some of the present controls are likely to be future cases, and the same process may well be ongoing in both groups, but only to different extents.
In conclusion, our present study showed that autoimmunity to human Hsp60 is a risk factor for coronary events. Our present and previous findings also indicate that the presence of C pneumoniae IgA antibodies alone is not a serious risk factor, but the risk increases when factors related to autoimmunity and inflammation, such as an elevated level of human Hsp60 IgA antibodies and an elevated CRP concentration, are present simultaneously. The results support the autoimmune hypothesis proposed by Wick et al.11 The speculative pathway in the case of C pneumoniae infection is initiated by a chain reaction in which CRP and Hsp60 play a role: a chronic C pneumoniae infection increases the expression of its own Hsp60, which leads to an increased expression of host Hsp60 and, thus, gradually results in autoimmunity with systemic inflammation and elevated CRP and, finally, in clinical manifestations of coronary heart disease.
This study was supported in part by grants from the Paavo Nurmi Foundation and the Sigrid Juselius Foundation.
Received August 6, 2001; revision accepted November 28, 2001.
- ↵Zügel U, Kaufmann SHE. Role of heat shock proteins in protection from and pathogenesis of infectious diseases. Clin Microbiol Rev. 1999; 12: 19–39.
- ↵Xu Q, Dietrich H, Steiner HJ, Gown AM, Schoel B, Mikuz G, Kaufmann SHE, Wick G. Induction of arteriosclerosis in normocholesterolemic rabbits by immunization with heat shock protein 65. Arterioscler Thromb. 1992; 12: 789–799.
- ↵George J, Shoenfeld Y, Afek A, Gilburd B, Keren P, Shaish A, Kopolovic J, Wick G, Harats D. Enhanced fatty streak formation in C57BL/6J mice by immunization with heat shock protein-65. Arterioscler Thromb Vasc Biol. 1999; 19: 505–510.
- ↵Xu Q, Kiechl S, Mayr M, Metzler B, Egger G, Oberhollenzer F, Willeit J, Wick G. Association of serum antibodies to heat-shock protein 65 with carotid atherosclerosis: clinical significance determined in a follow-up study. Circulation. 1999; 100: 1169–1174.
- ↵Kiechl S, Egger G, Mayr M, Wiedermann CJ, Bonora E, Oberhollenzer F, Muggeo M, Xu Q, Wick G, Poewe W, et al. Chronic infections and the risk of carotid atherosclerosis: prospective results from a large population study. Circulation. 2001; 103: 1064–1070.
- ↵Mayr M, Metzler B, Kiechl S, Willeit J, Schett G, Xu Q, Wick G. Endothelial cytotoxicity mediated by serum antibodies to heat shock proteins of Escherichia coli and Chlamydia pneumoniae: immune reactions to heat shock proteins as a possible link between infection and atherosclerosis. Circulation. 1999; 99: 1560–1566.
- ↵Mayr M, Kiechl S, Willeit J, Wick G, Xu Q. Infections, immunity, and atherosclerosis: associations of antibodies to Chlamydia pneumoniae, Helicobacter pylori, and cytomegalovirus with immune reactions to heat-shock protein 60 and carotid or femoral atherosclerosis. Circulation. 2000; 102: 833–839.
- ↵Fong IW, Chiu B, Viira E, Fong MW, Jang D, Mahony J. Rabbit model for Chlamydia pneumoniae infection. J Clin Microbiol. 1997; 35: 48–52.
- ↵Laitinen K, Laurila A, Pyhälä L, Leinonen M, Saikku P. Chlamydia pneumoniae infection induces inflammatory changes in the aortas of rabbits. Infect Immun. 1997; 65: 4832–4835.
- ↵Moazed TC, Kuo C-C, Grayston JT, Campbell LA. Murine models of Chlamydia pneumoniae infection and atherosclerosis. J Infect Dis. 1997; 175: 883–890.
- ↵Morrison RP, Lyng K, Caldwell HD. Chlamydial disease pathogenesis: ocular hypersensitivity elicited by a genus-specific 57-kD protein. J Exp Med. 1989; 169: 663–675.
- ↵Kol A, Sukhova GK, Lichtman AH, Libby P. Chlamydial heat shock protein 60 localizes in human atheroma and regulates macrophage tumor necrosis factor-alpha and matrix metalloproteinase expression. Circulation. 1998; 98: 300–307.
- ↵Mosorin M, Surcel H-M, Laurila A, Lehtinen M, Karttunen R, Juvonen J, Paavonen J, Morrison RP, Saikku P, Juvonen T. Detection of Chlamydia pneumoniae-reactive T lymphocytes in human atherosclerotic plaques of carotid artery. Arterioscler Thromb Vasc Biol. 2000; 20: 1061–1067.
- ↵Roivainen M, Viik-Kajander M, Palosuo T, Toivanen P, Leinonen M, Saikku P, Tenkanen L, Manninen V, Hovi T, Mänttäri M. Infections, inflammation, and the risk of coronary heart disease. Circulation. 2000; 101: 252–257.
- ↵Frick MH, Elo O, Haapa K, Heinonen OP, Heinsalmi P, Helo P, Huttunen JK, Kaitaniemi P, Koskinen P, Manninen V, et al. Helsinki Heart Study: primary-prevention trial with gemfibrozil in middle-aged men with dyslipidemia: safety of treatment, changes in risk factors, and incidence of coronary heart disease. N Engl J Med. 1987; 317: 1237–1245.
- ↵Huittinen T, Hahn D, Anttila T, Wahlström E, Saikku P, Leinonen M. Host immune response to Chlamydia pneumoniae heat shock protein 60 is associated with asthma. Eur Respir J. 2001; 17: 1078–1082.
- ↵Birnie DH, Holme ER, McKay IC, Hood S, McColl KE, Hillis WS. Association between antibodies to heat shock protein 65 and coronary atherosclerosis: possible mechanism of action of Helicobacter pylori and other bacterial infections in increasing cardiovascular risk. Eur Heart J. 1998; 19: 387–394.
- ↵Witkin SS, Askienazy-Elbhar M, Henry-Suchet J, Belaisch-Allart J, Tort-Grumbach J, Sarjdine K. Circulating antibodies to a conserved epitope of the Chlamydia trachomatis 60 kDa heat shock protein (hsp60) in infertile couples and its relationship to antibodies to C. trachomatis surface antigens and the Escherichia coli and human HSP60. Hum Reprod. 1998; 13: 1175–1179.
- ↵Ridker PM, Kundsin RB, Stampfer MJ, Poulin S, Hennekens CH. Prospective study of Chlamydia pneumoniae IgG seropositivity and risks of future myocardial infarction. Circulation. 1999; 99: 1161–1164.
- ↵Danesh J, Whincup P, Walker M, Lennon L, Thomson A, Appleby P, Wong YK, Bernardes-Silva M, Ward M. Chlamydia pneumoniae IgG titres and coronary heart disease: prospective study and meta-analysis. BMJ. 2000; 321: 208–213.
- ↵Laurila A, Bloigu A, Näyhä S, Hassi J, Leinonen M, Saikku P. Chronic Chlamydia pneumoniae infection is associated with a serum lipid profile known to be a risk factor for atherosclerosis. Arterioscler Thromb Vasc Biol. 1997; 17: 2910–2913.
- ↵Karvonen M, Tuomilehto J, Pitkäniemi J, Naukkarinen A, Saikku P. Importance of smoking for Chlamydia pneumoniae seropositivity. Int J Epidemiol. 1994; 23: 1315–1321.
- ↵von Hertzen L, Surcel H-M, Kaprio J, Koskenvuo M, Bloigu A, Leinonen M, Saikku P. Immune responses to Chlamydia pneumoniae in twins in relation to gender and smoking. J Med Microbiol. 1998; 47: 441–446.