Chlamydial and human heat shock protein 60s activate human vascular endothelium, smooth muscle cells, and macrophages (original) (raw)
Chlamydial and human HSP 60s induce E-selectin, ICAM-1, and VCAM-1 expression on endothelial cells. Expression of adhesion molecules on endothelium and/or smooth muscle cells regulates the leukocyte trafficking within the vascular wall, likely contributing to atherogenesis (17, 18). We therefore tested whether chlamydial or human HSP 60 can induce the expression of E-selectin on ECs, and of ICAM-1 and VCAM-1 on ECs and SMCs. After six hours, either chlamydial or human HSP 60 induced E-selectin on ECs to an extent similar to that induced by E. coli LPS; nonstimulated cells expressed very low levels of E-selectin (Fig. 1, top; Table 1). After 24 hours, either chlamydial or human HSP 60 induced ICAM-1 on ECs to an extent similar to that induced by E. coli LPS; nonstimulated cells expressed low levels of ICAM-1 (Fig. 2, top; Table 1). Chlamydial or human HSP 60 also induced VCAM-1 on ECs, as did E. coli LPS, although the level of expression was lower than for E-selectin and ICAM-1 (Table 1). In contrast, SMCs showed no increase over basal levels of either ICAM-1 or VCAM-1 (data not shown).
Chlamydial and human HSP 60s induce E-selectin production by endothelial cells. (top) ECs were incubated with medium only (unstimulated control), or with chlamydial HSP 60 (5 μg/ml), human HSP 60 (5 μg/ml), E. coli LPS (1 μg/ml), or inactivated C. pneumoniae (107 U/ml), for 6 h. Cells were harvested by treatment with trypsin-EDTA and then stained with antibody against E-selectin (solid histograms) or with mouse IgGs (as isotype control; open histograms). Chlamydial HSP 60 and human HSP 60 had a similar effect on E-selectin production as E. coli LPS. Inactivated C. pneumoniae did not elicit E-selectin expression by ECs. (bottom) Before incubation, reagents were heat-treated by boiling for 20 min. Heat treatment abolished the effect on E-selectin by chlamydial HSP 60 and human HSP 60, but did not modify the effect of thermostable E. coli LPS. Three independent experiments showed similar results. EC, endothelial cells; HSP, heat shock protein; LPS, lipopolysaccharide.
Chlamydial and human HSP 60s induce ICAM-1 production by endothelial cells. (top) ECs were incubated with medium only (unstimulated control), or with chlamydial HSP 60 (5 μg/ml), human HSP 60 (5 μg/ml), E. coli LPS (1 μg/ml), or inactivated C. pneumoniae (107 μg/ml), for 24 h. Cells were harvested by treatment with trypsin-EDTA and stained with antibody against ICAM-1 (solid histograms) or with mouse IgGs (as isotype control; open histograms). Chlamydial HSP 60 and human HSP 60 had a similar effect on ICAM-1 production as E. coli LPS. Inactivated C. pneumoniae did not elicit ICAM-1 expression by ECs. (bottom) Before incubation, reagents were heat-treated by boiling for 20 min. Heat treatment abolished the effect on ICAM-1 by chlamydial HSP 60 and human HSP 60, but did not modify the effect of thermostable E. coli LPS. Three independent experiments showed similar results. ICAM-1, intercellular adhesion molecule-1.
Percentages of endothelial cells positively stained for E-selectin, ICAM-1, and VCAM-1
Infection of ECs with live C. pneumoniae induces the expression of E-selectin, ICAM-1, and VCAM-1 (12). However, it is not clear whether this effect is mainly due to membrane constituents, such as LPS or the major outer membrane protein (MOMP), or to cytoplasmic constituents, such as the chlamydial HSP 60. Under the conditions used in this study, cells exposed to formalin-inactivated C. pneumoniae, which retain antigenicity (10) but cannot enter host cells, did not show increased expression of E-selectin, ICAM-1, or VCAM-1 (Figs. 1 and 2, top; Table 1). This observation suggests that a cytoplasmic component, such as the chlamydial HSP 60, and not a membrane constituent, may be responsible for the observed effects on ECs after infection with live C. pneumoniae.
Both the chlamydial and human HSP 60s used in this study were prepared by expression in E. coli, therefore contamination with E. coli LPS might account for the observed effect of HSP 60s. To exclude this possibility, we heat-treated the HSP 60s (both chlamydial and human) and the control E. coli endotoxin. Bacterial lipopolysaccharides exhibit thermostability, while most proteins are thermolabile. Heat treatment of either chlamydial or human HSP 60 abrogated their ability to induce E-selectin, ICAM-1, or VCAM-1, but did not alter the effects of E. coli endotoxin (Figs. 1 and 2, bottom; Table 1), thus indicating lack of appreciable contamination with LPS in the HSP 60 preparations.
Chlamydial and human HSP 60s induce IL-6 production by ECs, SMCs, and macrophages. IL-6 is an important mediator of the acute phase response and can increase plasma levels of fibrinogen, C-reactive protein (CRP), and serum amyloid A protein (SAA) (36). Increased fibrinogen levels might promote thrombosis on a preexisting atheroma, and levels of IL-6, CRP, and SAA correlate with prognosis in patients with unstable angina (37, 38), and with the risk of myocardial infarction in healthy subjects (39). Infection of human monocytes with C. pneumoniae induces the production of IL-6 (13). We therefore tested whether chlamydial or human HSP 60 can induce IL-6 expression in ECs, SMCs, and macrophages. After 24 hours, chlamydial and human HSP 60s significantly increased the expression of IL-6 by ECs (Fig. 3, top), SMCs (Fig. 4, top), and macrophages (Fig. 5, top), as compared with control levels. While inactivated C. pneumoniae did not increase IL-6 expression by ECs and SMCs, it elicited IL-6 production by monocyte-derived macrophages (Figs. 3–5, top). This apparent discrepancy might relate to the well-described phagocytic properties of macrophages, a characteristic not prominent in either ECs or SMCs. Interestingly, while E. coli LPS significantly increased IL-6 expression in ECs and macrophages, it did not affect IL-6 production by SMCs, further supporting the lack of LPS contamination in the preparations of either HSP 60. Moreover, heat treatment of either chlamydial or human HSP 60 abrogated their ability to induce IL-6 on ECs (Fig. 3, bottom), SMCs (Fig. 4, bottom), and macrophages (Fig. 5, bottom), but did not alter the effect of E. coli endotoxin.
Chlamydial and human HSP 60s induce IL-6 production by endothelial cells. (top) ECs were incubated with medium only (unstimulated control), or with chlamydial HSP 60 (5 μg/ml), human HSP 60 (5 μg/ml), E. coli LPS (1 μg/ml), or inactivated C. pneumoniae (107 U/ml), for 24 h. Samples were collected and analyzed for IL-6 by ELISA. Chlamydial HSP 60 and human HSP 60, as well as E. coli LPS, significantly increased IL-6 production over control levels. Inactivated C. pneumoniae did not elicit IL-6 expression by ECs. (bottom) Before incubation, reagents were heat-treated by boiling for 20 min. Heat treatment abolished the effect on IL-6 by chlamydial HSP 60 and human HSP 60, but did not modify the effect of thermostable E. coli LPS. Results shown represent the mean ± SEM of three independent experiments. Statistically significant vs. control (P ≤ 0.01, two-sided). IL-6, interleukin-6.
Chlamydial and human HSP 60s induce IL-6 production by smooth muscle cells. (top) SMCs were incubated without serum for 48 h and then with medium only (unstimulated control), or with chlamydial HSP 60 (5 μg/ml), human HSP 60 (5 μg/ml), E. coli LPS (1 μg/ml), or inactivated C. pneumoniae (107 U/ml), for 24 h. Samples were collected and analyzed for IL-6 by ELISA. Chlamydial HSP 60 and human HSP 60 significantly increased IL-6 production over control levels. Neither E. coli LPS nor inactivated C. pneumoniae did elicit IL-6 expression by SMCs. (bottom) Before incubation, reagents were heat-treated by boiling for 20 min. Heat treatment abolished the effect on IL-6 by both chlamydial HSP 60 and human HSP 60. Results shown represent the mean ± SEM of three independent experiments. Statistically significant vs. control (P ≤ 0.05, two-sided). SMCs, smooth muscle cells.
Chlamydial and human HSP 60s induce IL-6 production by monocyte-derived macrophages. (top) Macrophages were incubated with medium only (unstimulated control), or with chlamydial HSP 60 (5 μg/ml), human HSP 60 (5 μg/ml), E. coli LPS (1 μg/ml), or inactivated C. pneumoniae (107 U/ml), for 24 h. Samples were collected and analyzed for IL-6 by ELISA. Chlamydial HSP 60 and human HSP 60, as well as E. coli LPS and inactivated C. pneumoniae, significantly increased IL-6 production over control levels. (bottom) Before incubation, reagents were heat-treated by boiling for 20 min. Heat treatment abolished the effect on IL-6 by chlamydial HSP 60, human HSP 60, and inactivated C. pneumoniae, but did not modify the effect of thermostable E. coli LPS. Results shown represent the mean ± SEM of three independent experiments. Statistically significant vs. control (P ≤ 0.01, two-sided).
Chlamydial and human HSP 60s trigger NF-κB activation. We have recently shown that either chlamydial or human HSP 60 induce expression of TNF-α in mouse macrophages (10), and we have extended these findings to human macrophages (Kol, A., et al., unpublished observations). The present results show that either HSP 60 can also induce E-selectin, ICAM-1, VCAM-1, and IL-6. Because NF-κB transcriptional complexes contribute importantly to the inducible expression of the genes encoding E-selectin, ICAM-1, VCAM-1, IL-6, and TNF-α (23), we sought evidence that HSP 60s trigger NF-κB activation. In particular, because IL-6 production was induced by HSP 60s in all three cell types examined in this study, we chose to detect NF-κB activation by a gel-shift assay using an oligonucleotide probe that contains the NF-κB–binding site of the IL-6 promoter. HSP 60s did indeed activate NF-κB: a complex was evident by 30 minutes after stimulation and appeared maximal and sustained from two to six hours (Fig. 6). The time course of response to chlamydial or human HSP 60 resembled that of LPS, a well-known activator of NF-κB (23). Heat treatment of chlamydial or human HSP 60 nearly abolished their ability to activate NF-κB, whereas the same treatment did not affect the response to LPS, thus rendering highly unlikely the possibility that LPS contamination in our HSP 60 preparations accounted for the observed effect on NF-κB activation (Fig. 7). Formation of binding complexes from nuclear extracts of cells treated with human HSP 60 were prevented by competition with unlabeled oligonucleotides containing the IL-6 or consensus NF-κB site, but not by an oligomer containing two point mutations in the IL-6 NF-κB site, verifying the specificity of the induced complexes for binding of NF-κB proteins (Fig. 8). Supershift analysis with a panel of anti-Rel antibodies identified p65 (Rel A) and p50 as components of human HSP–induced binding complexes, while antibodies to cRel, RelB, or nonimmune rabbit IgG did not supershift any complex (Fig. 9).
Chlamydial and human HSP 60s induce NF-κB activation in a time-dependent fashion. Endothelial cells were incubated with medium only (unstimulated control), or with chlamydial HSP 60 (2 μg/ml), human HSP 60 (2 μg/ml), or E. coli LPS (1 μg/ml) for 0.5, 2, and 6 h. Nuclear extracts were prepared and subjected to electromobility shift assay with an oligonucleotide containing the NF-κB–binding site of the IL-6 promoter. NF-κB complex formation was evident by 30 min and appeared maximal and sustained from 2 to 6 h. The time course of response to chlamydial or human HSP 60 was similar to that of LPS. NF-κB, nuclear factor-κB.
Chlamydial and human HSP 60s induce NF-κB activation: effect of heat treatment. Endothelial cells were incubated with medium only (unstimulated control), or with chlamydial HSP 60 (2 μg/ml), human HSP 60 (2 μg/ml), or E. coli LPS (1 μg/ml) for 2 h. A portion of each reagent was heat-treated before incubation by boiling for 20 min. Nuclear extracts were prepared and subject to electromobility shift assay with an oligonucleotide containing the NF-κB–binding site of the IL-6 promoter. Heat treatment abolished the induction of NF-κB complex formation by chlamydial and human HSP 60, but did not modify the effect of thermostable E. coli LPS.
Human HSP 60 induces NF-κB activation: specificity of DNA-binding complexes. Endothelial cells were incubated with medium only (unstimulated control), or with human HSP 60 (2 μg/ml) for 2 h. Nuclear extracts were prepared and subject to electromobility shift assay with an oligonucleotide containing the NF-κB–binding site of the IL-6 promoter. Competition of DNA-binding activity in the presence of excess unlabeled competitor, consensus, or mutant NF-κB oligonucleotides demonstrate specificity of HSP 60–induced NF-κB DNA-binding complexes.
Human HSP 60 induces NF-κB activation: identification of p65 and p50 Rel proteins as components of DNA-binding complexes. Endothelial cells were incubated with medium only (unstimulated control), or with human HSP 60 (2 μg/ml) for 2 h. Rel proteins were identified by performing electromobility shift assay on nuclear extracts, with an oligonucleotide containing the NF-κB–binding site of the IL-6 promoter, in the presence of the indicated Rel antisera or with nonimmune rabbit IgG. NF-κB and supershifted complexes are indicated on the left: p65 (Rel A) and p50 were identified as components of human HSP–induced binding complexes, while antibodies to cRel, RelB, or nonimmune rabbit IgG did not shift any binding complex. Probe only indicates electromobility shift assay performed without nuclear extract.