Abstract
Evidence is emerging that the two chaperonin (Cpn) 60 proteins of Mycobacterium tuberculosis, Cpn60.1 and Cpn60.2, have moonlighting actions that may contribute to the pathology of tuberculosis. We studied the release of Cpn60.1 from M. tuberculosis and infected macrophagelike cells and compared recombinant Cpn60.1 and Cpn60.2 in a range of cell-based assays to determine how similar the actions of these highly homologous proteins are. We now establish that Cpns are similar as follows: (i) Cpn60.1, as it has been shown for Cpn60.2, is released by M. tuberculosis in culture, and Cpn60.1 is furthermore released when the bacterium is in quiescent, but not activated, macrophagelike cells, and (ii) both proteins only showed a partial requirement for MyD88 for the induction of proinflammatory cytokine production compared to lipopolysaccharide. However, we also found major differences in the cellular action of Cpns. (i) Cpn60.2 proved to be a more potent stimulator of whole blood leukocytes than Cpn60.1 and was the only one to induce tumor necrosis factor alpha synthesis. (ii) Cpn60.1 bound to ca. 90% of circulating monocytes compared to Cpn60.2, which bound <50% of these cells. Both chaperonins bound to different cell surface receptors, while monocyte activation by both proteins was completely abrogated in TLR4−/− mice, although Cpn60.2 also showed significant requirement for TLR2. Finally, an isogenic mutant lacking cpn60.1, but containing intact cpn60.2, was severely inhibited in generating multinucleate giant cells in an in vitro human granuloma assay. These results clearly show that, despite significant sequence homology, M. tuberculosis Cpn60 proteins interact in distinct ways with human or murine macrophages.
Mycobacterium tuberculosis remains a major global pathogen killing an estimated 2 million people each year (32). Young and coworkers have recently stressed the urgent need to understand how this organism causes disease and to identify the key virulence factors (35). One of the earliest described factors from M. tuberculosis, and one potentially involved in the pathogenesis of tuberculosis, was the heat shock protein chaperonin 60 (Cpn60) or heat shock protein 60 (Hsp60). This was initially identified as a “common antigen” to which antibodies were raised in a variety of bacterial infections and later identified as the molecular chaperone homologous to Escherichia coli GroEL and termed Hsp65 (15). The Hsp65 protein of the mycobacteria has been the subject of enormous interest because of its immunomodulatory influence on T-cell immunity (31). These effects were thought to be due to Hsp65 acting as an immunogen. However, Friedland et al. revealed that M. tuberculosis Hsp65 was able to directly activate human monocyte proinflammatory cytokine production (5) and generated the hypothesis, which is still the subject of some controversy, that chaperonin 60 proteins are moonlighting proteins (11) that can be secreted from cells and have cell-cell signaling functions (1, 18).
M. tuberculosis is one of the growing proportion of bacteria that have been found to have more than one gene coding for chaperonin 60 proteins (13). The current nomenclature for these chaperonin proteins is Cpn60.1 for the most recently discovered protein and Cpn60.2 for the Hsp65 protein. These proteins share 61% sequence identity (2). Purified recombinant versions of these proteins, in which lipopolysaccharide (LPS) contamination has been removed and further controlled for, activate human monocyte proinflammatory cytokine production, with the Cpn60.1 protein being both more potent and efficacious than the Cpn60.2 protein. The activity of the Cpn60.1 protein is blocked by neutralizing antibodies to CD14, whereas these antibodies have no effect on the monocyte activation induced by Cpn60.2 (17). This suggested that, if released from bacteria, the Cpn60.1 protein could be a major proinflammatory stimulus in mycobacterial infections.
To gain further insight into the role played by the two chaperonin 60 proteins and the chaperonin 10 protein of M. tuberculosis, attempts were made to inactivate each of these genes in the H37Rv virulent strain of this bacterium (10). These experiments revealed that genes encoding Cpn60.2 and Cpn10 were essential. However, the cpn60.1 gene could be inactivated without any obvious phenotype in culture or when the bacterium was grown within quiescent or activated macrophages. When used to infect mice or guinea pigs, the Δcpn60.1 isogenic mutant grew at the same rate as the wild type and complemented mutant. Strikingly, this mutant failed to induce a granulomatous response in mice or guinea pigs (10). These data support the hypothesis that M. tuberculosis Cpn60.1 is a directly acting virulence factor, driving the process of granuloma formation.
Early studies had shown that Cpn60 proteins from Gram-negative bacteria could stimulate bone breakdown in in vitro explant assays but that the Cpn60.2 proteins from M. tuberculosis and M. leprae were inactive (12). It was later shown that E. coli GroEL was a potent stimulator of the differentiation and growth of the multinucleate cell population responsible for bone breakdown: the osteoclast (27). The M. tuberculosis Cpn60.1 protein was also shown to be inactive in this bone resorbing assay (20). However, further analysis of the activity of the recombinant M. tuberculosis Cpn60 proteins has revealed that the Cpn60.2 protein can neither promote nor inhibit bone resorption or the differentiation/growth of osteoclasts. In contrast, the Cpn60.1 protein is a potent inhibitor of osteoclast formation acting, at least in part, by blocking production of the key osteoclast transcription factor NFATc1 (34).
These findings suggest an important role for M. tuberculosis Cpn60.1 in the pathogenesis of tuberculosis, and it was therefore important to study the biological effects of this protein in more detail and contrast it with the biological actions of the Cpn60.2 protein.
MATERIALS AND METHODS
Protein expression and recombinant protein purification.
The genes encoding the M. tuberculosis Cpn60 proteins were cloned as described previously (29) into TOP10 (Invitrogen, United Kingdom), and the His6 recombinant proteins were expressed by using standard methods. LPS, which could interfere with the analysis of monocyte activation, was removed by washing the solubilized recombinant Cpn60 proteins, when bound to an Ni-NTA affinity chromatography matrix (Invitrogen), with 30 ml of a 1-mg/ml polymyxin B (Sigma-Aldrich, United Kingdom) solution. The proteins were then eluted from the column with imidazole, according to the manufacturer's instructions, and then further purified by anion-exchange chromatography (Perseptive Biosystems, California), taking great care to limit LPS contamination. The purified proteins were then dialyzed against phosphate-buffered saline (PBS), and their purities were assessed by SDS-PAGE using 4 to 12% Bis-Tris gels stained with Simply Blue Safe Stain (all from Invitrogen).
Analysis of LPS contamination.
LPS contamination of the purified recombinant M. tuberculosis Cpn60.1 and Cpn60.2 proteins was determined by using a Limulus amebocyte lysate assay (Associates of Cape Cod, Liverpool, United Kingdom) according to the manufacturer's instructions. All purified chaperonins contain ∼10 pg of LPS/μg of protein. To ensure the cell activation is not due to the contaminating LPS, the proteins were digested with proteinase K, as described previously (29), and used for cytokine stimulation in control experiments.
Growth of M. tuberculosis H37Rv.
M. tuberculosis H37Rv was grown without disturbance in 7H9 medium supplemented with 0.05% Tween and 10% albumin-dextrose complex (Difco Laboratories). Viability was determined by measurement of the CFU on days 1, 4, 6, 10, 12, 16, and 30 of cultures on 7H11 agar medium supplemented with oleic-albumin-dextrose complex. Bacterial culture supernatants were collected on day 0 and on days 6, 10, 12, 16, and 30 of incubation; filtered through a 0.2-μm-pore-size filter (Sartorius); and stored at −20°C before use in enzyme-linked immunosorbent assays (ELISA) for the detection of M. tuberculosis Cpn60.1 or in Western blot analysis for the detection of M. tuberculosis Cpn60.2.
Infection of J774A.1 murine macrophagelike cell line with M. tuberculosis H37Rv.
Mouse macrophagelike cell line J774A.1 was either left resting in 24-well tissue culture plates or was incubated with gamma interferon (IFN-γ; 100 U/ml; R&D Systems, Minneapolis, MN) for 24 h, followed by the addition of 200 ng of LPS/ml for 3 h before infection with M. tuberculosis H37Rv at a multiplicity of infection of 1:1 for 4 h, after which the cell line was washed to remove nonadherent bacteria. The cells were left to incubate with bacteria for 1 to 3 days before the cell supernatant was collected. Subsequently, the cells were washed five times with warm Hanks buffered salt solution and lysed with 0.1% Triton X-100, which does not lyse M. tuberculosis. The bacterial numbers in the cell supernatant and the lysed cell suspension were enumerated by counting the CFU, as described above. The supernatant and the cell lysate were then filtered through a 0.2-μm-pore-size filter and stored at −20°C for subsequent assay of M. tuberculosis Cpn60.1 with ELISA and M. tuberculosis Cpn60.2 by Western blotting.
ELISA for detection of M. tuberculosis Cpn60.1.
A monospecific antiserum recognizing M. tuberculosis Cpn60.1 was produced as described earlier (34). This antiserum had minimal cross-reactivity with M. tuberculosis Cpn60.2 and human Cpn60. An ELISA was developed by using 10 μg/ml of unlabeled antiserum to Cpn60.1 and 10 μg/ml of biotin-labeled antiserum to Cpn60.1 to visualize bound Cpn60.1. Recombinant M. tuberculosis Cpn60.1 was diluted in the culture medium to a concentration ranging from 0 to 51.2 ng/ml and used as standard. The lower limit of sensitivity of the assay was 0.4 ng of recombinant Cpn60.1/ml.
Western blotting for detection of M. tuberculosis Cpn60.1 and Cpn60.2.
To determine whether M. tuberculosis Cpn60.1 is present inside M. tuberculosis-infected macrophages, the cell lysates were prepared, as described above and supplemented with protease inhibitor cocktail (all from Santa Cruz Biotechnology, California). The total protein concentration in the cell lysates was measured by using a bicinchoninic acid assay (Thermo Fisher Scientific, Massachusetts), according to the manufacturer's instructions. The samples were read at 592 nm by using a Nanodrop spectrophotometer (Nanodrop Technologies) and diluted to a concentration of 1.5 μg/μl of total protein. To determine whether M. tuberculosis Cpn60.2 is present in the supernatant of M. tuberculosis liquid culture and the M. tuberculosis-infected J774 macrophagelike cell line, the supernatants were collected, as described above, and either used immediately or first concentrated by trichloroacetic acid (TCA) precipitation (supernatants of infected J774 cells only) or by ultrafiltration (Vivaspin; Vivascience, Sartorius, United Kingdom). For TCA precipitation, TCA was added to the supernatants to 10% final concentration and left on ice for 2 h, after which the supernatants were pelleted, and the pellets were washed twice in ice-cold acetone before being resuspended in LDS sample buffer (Invitrogen). The supernatants and the diluted cell lysates were then separated on a NuPAGE 4 to 12% Bis-Tris gel (Invitrogen) and blotted onto Hybond-P membrane (Amersham Pharmacia Biotech, United Kingdom). The blots were blocked with 8% semiskimmed milk in PBS-0.1% Tween and probed with anti-M. tuberculosis Cpn60.1 or anti-Hsp65 (Stressgen). The blots were visualized by using anti-rabbit IgG horseradish peroxidase (HRP) or anti-mouse IgG HRP (Cell Signaling Technology, Massachusetts) and Super Signal West Pico chemiluminescent substrate (Pierce/Thermo Fisher Scientific, United Kingdom). The blots were finally exposed to Kodak BioMax Light film (Sigma Aldrich).
Whole-blood cytokine production assay.
Whole blood was obtained from the National Blood Service Donor Centre, Tooting, South London. The ethical approval for the study was obtained by Wandsworth Research Ethics Committee. To determine the effect of recombinant Cpn60 proteins on the mixed leukocyte populations found in whole blood, various concentrations of recombinant Cpn60 proteins or 10 ng of LPS/ml was added to undiluted human blood in 24-well plates, followed by incubation for 24 h, after which time the cells were sedimented by centrifugation and cytokine levels (IFN-γ, tumor necrosis factor alpha [TNF-α], interleukin-1 [IL-1], IL-4, IL-6, IL-8, IL-10, and IL-12) in plasma estimated by specific cytokine ELISA as described below.
Isolation and culture of human peripheral blood leukocytes.
Buffy coats were obtained from the National Blood Service Donor Centre, Tooting, South London. The peripheral blood leukocytes were obtained by density gradient centrifugation using Histopaque (Sigma-Aldrich), according to the manufacturer's instructions. The mononuclear cell layer was collected, washed in PBS and resuspended in supplemented RPMI 1640 (containing 100 μg of penicillin-streptomycin/ml and 2 mM l-glutamine; all from Invitrogen). The cells were counted by using trypan blue (Sigma-Aldrich) to assess cell viability, which was >95%, and plated in 24-well tissue culture plates (Becton Dickinson) at the concentration of 2 × 106 cells/well for cytokine assays. To obtain a largely monocytic cell population the cells were incubated at 37°C in a 95% air-5% CO2 atmosphere for 1 h, after which the nonadherent cells were removed, and the remaining cells were washed with PBS. Fresh supplemented RPMI medium containing 2 mM l-glutamine and 2% autologous plasma was added to the adherent cells.
Isolation of mouse bone marrow-derived macrophages.
The bone marrow cells from C57/BL6 wild-type, TLR2-deficient, and TLR4-deficient mice were obtained from the Sackler Institute, King's College London. The cells were cultured at 37°C in a 95% air-5% CO2 atmosphere in 25-cm2 flasks in Dulbecco modified Eagle medium (DMEM; Gibco) supplemented with 10% fetal bovine serum, 4 mM l-glutamine, 100 U of penicillin/ml, and 100 μg of streptomycin/ml (all from Invitrogen) and 1 ng of mouse macrophage colony-stimulating factor (M-CSF; R&D Systems)/ml. The nonadherent cells were removed, and the supplemented culture medium was changed every 2 days. On day 8, the cells were harvested and plated at 5 × 105 cells/ml in 24-well plates and then used in stimulation experiments.
Transient transfection of human monocytes with the cDNA constructs.
The dominant-negative MyD88 and empty vectors were obtained from Richard Darveau, University of Washington, Seattle. Transient transfection of human monocytes was obtained by using Lipofectamine Plus reagent (Invitrogen) according to the manufacturer's instructions. A portion (0.5 μg) of DNA was diluted in Opti-MEM medium (Invitrogen) and incubated with the cells for 24 h at 37°C in a 95% air-5% CO2 atmosphere, after which the medium was removed, and fresh medium was added to the cells before stimulation.
Stimulation of human monocytes and mouse bone marrow-derived macrophages.
The M. tuberculosis Cpn60.1, Cpn60.2, LPS (from E. coli O113:H10 [Associates of Cape Cod]), and zymosan (Invivogen, California) were diluted in either supplemented RPMI medium or supplemented DMEM to 10× final concentration. Polymyxin B was added to the chaperonins at a concentration of 20 μg/ml. The proteins, LPS, or zymosan was then incubated with the cells at 37°C in 95% air-5% CO2 atmosphere for 20 h, after which the medium was removed and stored at −20°C. In some experiments, anti-CD14 antibody (clone MΦP9; BD Biosciences Europe) or the isotype control (eBioscience, San Diego, CA) was incubated with the human monocytes before the addition of 10 μg of M. tuberculosis Cpn60.1/ml or 10 ng of LPS/ml as previously described (17).
Cytokine ELISA.
The matched antibody pairs and recombinant cytokines for the detection of human IL-1, IL-4, IL-6, IL-8, IL-10, IL-12, IFN-γ, TNF-α, and mouse IL-6 were obtained from the National Institute of Biological Standards and Controls (Potters Bar, United Kingdom) and R&D Systems. ELISA plates (96-well; Nunc Maxisorp, Denmark) were coated with the capture antibodies and incubated overnight at 4°C. The following day, the plates were washed with PBS-0.05% Tween and blocked with 1% bovine serum albumin (BSA) in PBS (all from Sigma-Aldrich). All of the standard cytokines were diluted in the culture medium to a concentration ranging from 0 to 2,000 pg/ml. The standards and samples were left to incubate on the plates for 2 h, after which the plates were washed, and the biotinylated detection antibodies were added. After washing, the avidin-HRP (Sigma-Aldrich) or streptavidin-HRP (R&D Systems) were added for 20 min, and the color reaction was developed by the addition of o-phenylenediamine (Sigma-Aldrich) and stopped by using 1 N H2SO4. The cytokine concentration in the negative control was subtracted from the test samples. All of the data are represented as the mean cytokine production ± the standard deviation.
Fluorescent labeling of Cpn60 proteins.
The recombinant Cpn60 proteins were labeled with Alexa Fluor 488 by using a commercial protein labeling kit (Molecular Probes/Invitrogen). A total of 50 μl of 1 M sodium bicarbonate was added to 500 μl of each protein at a concentration of 2 mg/ml to raise the pH of the solution. The proteins were then incubated with Alexa Fluor 488 for 1 h at room temperature in the dark, followed by an overnight incubation at 4°C. The unincorporated dye was removed from the labeled proteins by gel filtration. The degree of labeling was determined by measuring the absorbance of the proteins at 280 and 494 nm. The labeled proteins had an average of two dye molecules bound to each protein molecule.
Binding and competition assays with recombinant M. tuberculosis Cpn60 proteins.
The peripheral blood leukocytes were adjusted to a concentration of 106 cells in 200 μl of supplemented RPMI medium. The cells were either left untreated or incubated with increasing concentrations of labeled M. tuberculosis Cpn60.1 and Cpn60.2 (0.25 to 1 μM) for 45 min on ice. The cells were then centrifuged, resuspended in fresh medium, and incubated with either mouse IgG2a labeled with phycoerythrin (PE; for isotype control; BD Biosciences Europe) or incubated with anti-CD14 antibody labeled with PE (clone MΦP9; BD Biosciences Europe) for another 30 min on ice. Cells were then washed with cold PBS containing 1% BSA and fixed with PBS containing 2% paraformaldehyde. For the competition studies, the cells were incubated with 10-fold molar excess of unlabeled M. tuberculosis Cpn60.1 for 30 min on ice, after which the labeled M. tuberculosis Cpn60.1 or Cpn60.2 was added, and the cells were left on ice for another 45 min. After the incubation period, the cells were washed with cold PBS-1% BSA and fixed with PBS-2% paraformaldehyde. The binding of chaperonins was determined by measuring the fluorescence intensity by FACScan (BD Biosciences).
Influence of Cpn60.1 on MGC formation.
In vitro granulomas were obtained as previously described (23). Briefly, 106 freshly isolated PBMC were incubated with 104 viable M. tuberculosis (wild-type organism, isogenic mutant, or complemented mutant) for 15 days. The culture medium was RPMI 1640 × glutamate (Difco), containing 7.5% human AB serum (Sigma-Aldrich). The evolution of granulomatous structures was observed every 2 days. At day 15 after infection, granuloma cells were collected and plated onto glass coverslips with a cytospin (Thermo Shandon) fixed for 30 min in PBS-4% paraformaldehyde. For phagocyte quantitation, cells were recovered from 50 granulomas and stained with May-Grünwald-Giemsa, and the proportions of multinucleate cells (MC) and multinucleate giant cells (MGC) were manually determined by using a light microscope (Nikon TE 300).
RESULTS
Release of Cpn60.1 from M. tuberculosis H37Rv in culture.
M. tuberculosis H37Rv were grown for 30 days, during which a sample of the culture supernatant was removed on days 0, 6, 10, 12, 16, and 30, filtered, and analyzed by ELISA for the presence of Cpn60.1. An isogenic mutant of M. tuberculosis not expressing cpn60.1 (10), but growing at the same rate as the wild-type organism, was used as a control in this experiment. To determine whether there were factors in the culture medium interfering with the ELISA, known concentrations of Cpn60.1 were added to the medium and assayed by using the ELISA. This showed that there was full recovery of the added Cpn60.1. This experiment revealed that on day 10 there was release of Cpn60.1 into the medium of cultured M. tuberculosis and that the medium from the Δcpn60.1 isogenic mutant and the bacterium-free medium contained no Cpn60.1 (Fig. 1 A). To assess whether Cpn60.1 may be released from viable or dead bacilli, we measured its concentration in early and late M. tuberculosis cultures, specifically on days 0, 6, 12, 16, and 30 (Fig. 1B). The culture on day 0 contained 106 organisms, and the increase in CFU counts (Fig. 1C) is presented on a linear scale to clearly demonstrate that the culture was growing at subsequent time points. This experiment showed that ∼0.5 ng of Cpn60.1/ml was already present on day 6, and it increased to ∼3 ng of Cpn60.1/ml on day 30. We therefore conclude that Cpn60.1 is being released from live bacilli. The same supernatants of M. tuberculosis liquid culture were also assayed for the presence of Cpn60.2. No Cpn60.2 was detected in these samples by Western blotting (results not shown); however, it has recently been reported that Cpn60.2 is released by M. tuberculosis (8).
FIG. 1.
Cpn60.1 is released from M. tuberculosis in liquid culture. (A) M. tuberculosis H37Rv (M. tuberculosis WT) and the isotype strain lacking the cpn60.1 gene (M. tuberculosis cpn60.1 KO) were grown for 10 days in liquid culture, after which the supernatant was collected and filtered. The amount of Cpn60.1 was assessed by ELISA. The results are the means and standard deviations from two independent experiments assayed in triplicate. The asterisks represent a statistically significant difference in the concentration of Cpn60.1 between M. tuberculosis WT and cpn60.1 KO (**, P < 0.005 [Student's t test]). (B) M. tuberculosis H37Rv was grown for 30 days in liquid medium, during which the supernatant was collected and filtered on days 0, 6, 12, 16, and 30. The amount of Cpn60.1 in the medium was assessed by ELISA. The results are the means and standard deviations of two independent experiments assayed in triplicate. (C) The CFU counts were performed on M. tuberculosis liquid culture on days 0, 6, 12, 16, and 30 to assess the viability of the culture. The growth curve started from the baseline of 106 organisms. The data are means and standard deviations of two independent experiments, each performed in triplicate.
Release of Cpn60.1 from M. tuberculosis H37Rv that have invaded murine macrophages.
The mouse macrophagelike cell line J774 was infected with M. tuberculosis H37Rv, and the release of Cpn60.1 into the cell supernatant was monitored over a 3-day period. The J774 cell line lysate and the cell supernatant were analyzed for the presence of Cpn60.1 by Western blotting and ELISA, respectively. In addition, IFN-γ-activated J774 cells were also used in this experiment to determine whether the activation of macrophagelike cells affects the release of Cpn60.1. Bacteria were found both in the medium and in the cultured macrophagelike cells (Fig. 2 A). The culture medium from quiescent macrophagelike cells contained measurable amounts of immunoreactive Cpn60.1 (Fig. 2B.). However, this protein could not be detected when lysates of these cells were Western blotted with this same antibody. In contrast, no immunoreactive Cpn60.1 was detected in any of the three separate experiments performed with J774 cells exposed to IFN-γ (results not shown). The supernatants of infected resting and IFN-γ-activated J774 cells were also assayed for the presence of Cpn60.2 by Western blotting. No Cpn60.2 was detected in these supernatants (results not shown).
FIG. 2.
(A) M. tuberculosis log CFU counts in the supernatant (upper panel) and cell lysate (lower panel) of resting (gray line) and IFN-γ-activated (black line) macrophagelike cell line J774. The macrophagelike cells were infected on day 0, and the supernatant or cell lysate was collected at each time point and transferred to the agar medium for CFU determination. The experiment was repeated three times independently, each time in triplicate. The data are represented as mean log CFU counts ± the standard deviations from all three independent experiments. (B) M. tuberculosis Cpn60.1 is found in the cell medium of M. tuberculosis-infected resting macrophagelike cells. The cells were infected with M. tuberculosis H37Rv over a 3-day period, and the cell medium was collected at each time point, filtered, and analyzed for the presence of Cpn60.1 with ELISA. The experiment was repeated three times independently, each time in triplicate. The data are represented as mean protein concentrations ± the standard deviations from all three independent experiments. The asterisks represent statistically significant differences in the concentration of Cpn60.1 between day 0 and day 2 or between day 0 and day 3 (*, P < 0.05; ***, P < 0.0005 [Student's t test]).
Influence of recombinant Cpn60 proteins on whole-blood cytokine synthesis.
Exposure of whole blood to graded concentrations of M. tuberculosis Cpn60.1 or Cpn60.2 produced results discrepant from the findings seen with isolated monocytes (17). Of the eight cytokines measured, only three were produced by contact with these chaperonins (Fig. 3). Unexpectedly, Cpn60.2 was the most potent in the induction of IL-1 and IL-6 production and was the only one of the two chaperonins to induce the synthesis of TNF-α. Neither Cpn60 protein was able to induce the synthesis of IFN-γ, IL-4, IL-8, IL-10, or IL-12.
FIG. 3.
Cytokine production in whole blood in response to M. tuberculosis Cpn60.1 and Cpn60.2 proteins. Whole blood was incubated with various amounts of Cpn60.1 or Cpn60.2 for 24 h, after which the supernatant was collected and the cytokine concentrations (IL-1, ⧫; IL-6, •; TNF-α, ▴) were determined by cytokine ELISA. The experiment was repeated three times independently, each time in triplicate. The data represent the means ± the standard deviations of all three independent experiments.
Binding of M. tuberculosis Cpn60 proteins to human CD14-positive monocytes.
To determine the binding of both M. tuberculosis Cpn60 proteins to peripheral blood leukocytes, the proteins were labeled with Alexa Fluor 488. Flow cytometry revealed that both proteins bound preferentially to the monocyte population. To confirm this, we conducted double labeling experiments in which an antibody (clone MΦP9) to CD14 was used to identify monocytes. We further checked to see that this antibody does not block the ability of the chaperonins to activate human monocyte proinflammatory cytokine synthesis (results not shown). Adding 0.5 or 1.5 μM Alexa Fluor 488-labeled Cpn60 protein to CD14-labeled monocytes revealed a significant difference in the binding of M. tuberculosis Cpn60.1 and Cpn60.2. At 1.5 μM Cpn60.1, virtually all of the CD14 monocytes bind this recombinant protein. However, at the same molar concentration, only 48% of the CD14 monocytes bind Cpn60.2 (Fig. 4).
FIG. 4.
Differential binding of M. tuberculosis Cpn60 proteins to CD14+ human monocyte population. The PBMC were incubated with 0, 0.5, or 1.5 μM concentrations of labeled M. tuberculosis Cpn60.1 (A to C) or 0, 0.5, or 1.5 μM concentrations of labeled M. tuberculosis Cpn60.2 (D to F) and then separated according to the presence of CD14 molecule. The percentages in the upper left and right quadrants refer to the percentages of the gated cell population.
To determine whether the Cpn60.1 and Cpn60.2 proteins bind to the same cell surface receptor a 10-fold-higher concentration of unlabeled Cpn60.1 was preincubated with cells for 30 min, after which fluorescently labeled Cpn60.1 or Cpn60.2 was added, and the amount of relative fluorescence per cell was measured by flow cytometry. These results clearly indicated that unlabeled Cpn60.1 competed for the binding of labeled Cpn60.1 (Fig. 5 A and B). In contrast, there was no competition between unlabeled Cpn60.1 and labeled Cpn60.2, revealing that these two proteins bind to different receptors on the monocyte cell surface (Fig. 5C and D).
FIG. 5.
M. tuberculosis Cpn60.1 specifically binds to the surface of the human monocytes and does not share its binding site with M. tuberculosis Cpn60.2. (A) CD14+ cells were gated and plotted on a histogram. The fluorescence of the cell population incubated with the labeled and unlabeled Cpn60.1 is less than the fluorescence of the cell population incubated with the labeled protein only. (B) The geometric mean fluorescence of the cell population incubated with the labeled and unlabeled Cpn60.1 is significantly less than the mean fluorescence of the cell population incubated with the labeled protein only. The experiment was repeated three times. The star represents a significantly lower geometric mean (*, P < 0.05 [Student's t test]). The data in panel A show representative results from three replicate experiments. The data in panel B show the average geometric means ± the standard deviations of all three experiments. (C) The fluorescence of the CD14+ cell population incubated with the labeled Cpn60.2 and the unlabeled Cpn60.1 is not less that the fluorescence of the CD14+ cell population incubated with the labeled Cpn60.2 only. (D) The geometric mean fluorescence values of the two cell populations are not significantly different (Student's t test). The experiment was repeated three times. The data in panel C show a representative results from three replicate experiments. The data in panel D show the average geometric means ± the standard deviations of all three experiments.
Proteinase K treatment of recombinant Cpn60 proteins blocks activity.
There is still controversy about the potential role of LPS contamination in the activation of monocytes by recombinant Cpn60 proteins expressed in E. coli. In an attempt to rule this out, we treated both mycobacterial Cpn60 proteins with proteinase K. This completely proteolyses both proteins (Fig. 6 A). Monocytes were exposed to intact Cpn60 proteins, proteins which had been proteinase K treated, and to proteins which had been proteinase K treated and to which polymyxin B (20 μg/ml) had been added. As can be clearly seen, proteinase K treatment inhibits cytokine stimulating activity by 92 to 96%, and this residual activity is abolished by polymyxin B (Fig. 6B).
FIG. 6.
(A) M. tuberculosis Cpn60.1 and Cpn60.2 are completely degraded by proteinase K. The chaperonins were either left untreated or treated with the proteinase K for 18 h at 37°C. Lanes show intact M. tuberculosis Cpn60.1 (lane 1), proteinase K-treated Cpn60.1 (lane 2), intact M. tuberculosis Cpn60.2 (lane 3), and proteinase K-treated Cpn60.2 (lane 4). (B) The proinflammatory activity of M. tuberculosis Cpn60.1 and Cpn60.2 does not depend on LPS. Human monocytes were incubated with intact Cpn60.1 or Cpn60.2 in the presence of polymyxin B (▨) or proteinase K-treated proteins with (▪) or without polymyxin B (░⃞). Most TNF-α production is abrogated in response to digested proteins, and the small amount remaining was eliminated by the addition of polymyxin B, suggesting that LPS could not account for the observed proinflammatory activity. The experiment was repeated three times independently, each time in triplicate. The data represent mean values of TNF-α from all three independent experiments ± the standard deviations. The asterisks represent statistically significant difference in TNF-α production between monocytes stimulated with untreated proteins and proteinase K-treated proteins with or without polymyxin B (**; P < 0.005, ***; P < 0.0005 [Student's t test]).
M. tuberculosis Cpn60 proteins and activation of TLR2- and TLR4-deficient macrophages.
We next examined the role played by Toll-like receptors (TLRs) in monocyte activation by M. tuberculosis Cpn60 proteins. The cell signaling activity of Cpn60.2 has previously been linked with TLR4 binding, and we therefore wondered whether the M. tuberculosis Cpn60.1 also used the same receptor for activating cell signaling pathways. We used bone marrow-derived macrophages from two knockout mouse strains (C57/BL6 TLR2− and C57/BL6 TLR4−). We first examined the responsiveness of mouse macrophages from the wild-type strain to the mycobacterial chaperonins. The cells released the proinflammatory cytokine IL-6 in a dose-dependent manner in response to both Cpn60.1 and Cpn60.2 from M. tuberculosis (results not shown). Next, the macrophages from TLR2- and TLR4-deficient mice were incubated with 10 μg/ml of either Cpn60.1 or Cpn60.2. zymosan (a TLR2 ligand) and LPS (a TLR4 ligand) were used as controls. The secretion of IL-6 in response to both Cpn60.1 and Cpn60.2 was completely abolished in TLR4-deficient macrophages, confirming the requirement of TLR4 for the cell signaling activity of the two chaperonins. While the production of IL-6 was not affected in TLR2-deficient macrophages in response to M. tuberculosis Cpn60.1, the effects of Cpn60.2 were significantly dependent on TLR2. These data suggest that the cell signaling activity of Cpn60.1 depends wholly on the participation of TLR4. In contrast, Cpn60.2 can signal either through TLR2 or through the use of TLR4 (Fig. 7).
FIG. 7.
IL-6 production from bone marrow-derived macrophages from C57/BL6 wild-type mice and from TLR4-deficient (A) and TLR2-deficient (B) mice in response to M. tuberculosis Cpn60.1 and Cpn60.2. The cells were incubated with 10 μg of chaperonin/ml for 24 h before quantification of IL-6 in the cell medium. The asterisks represent a statistically significant difference in IL-6 production between wild-type (▪) and TLR-deficient cells (░⃞) (**, P < 0.005; ***, P < 0.0005 [Student's t test]). The experiment was repeated three times in triplicates. The data represent the mean cytokine production from all three independent experiments ± the standard deviations.
Role of MyD88 in M. tuberculosis Cpn60-induced monocyte activation.
The MyD88-dependent pathway is the most common pathway for cell activation in response to TLR ligands (21). To determine whether monocyte activation in response to Cpn60.1 and Cpn60.2 depends on the adaptor protein MyD88, we utilized a plasmid expressing a dominant-negative form of the molecule that we transfected into human monocytes. LPS was used as a control of the transfection efficiency and efficacy. The MyD88 dominant-negative mutant, as expected, virtually abolished the response of human monocytes to LPS. In contrast, there was ca. 50% inhibition of Cpn60.1 activity, while the degree of inhibition of Cpn60.2 was just barely significant (Fig. 8).
FIG. 8.
M. tuberculosis Cpn60.1 and Cpn60.2 only partially depend on MyD88 for maximal production of TNF-α from human monocytes. Human monocytes were transfected with either empty plasmid (▪) or the plasmid encoding a dominant-negative form of MyD88 (░⃞) before incubation with 10 μg of either Cpn60.1 or Cpn60.2/ml or 10 ng of LPS/ml. The experiment was repeated three times, and each time it was performed in triplicates. The data represent the mean values of TNF-α from all three independent experiments ± the standard deviations. The asterisks represent a significantly lower production of the cytokine (*, P < 0.05; **, P < 0.005 [Student's t test]).
Role of Cpn60 proteins in granuloma formation.
We have previously generated an isogenic mutant lacking the cpn60.1 gene (Δcpn60.1) and have complemented this mutant, allowing recovery of wild-type activity (10). We incubated the wild-type H37Rv virulent strain of M. tuberculosis and the isogenic mutant and complemented strain with normal human peripheral blood mononuclear cells (PBMC) in a human granuloma formation assay (24, 25). In a previous study we have demonstrated that while MC (one to two nuclei per cell) can be induced within granulomatous structures by any mycobacterial species, independently of their virulence, only highly virulent species, such as M. tuberculosis, were able to also induce the formation of large MGC (>5 nuclei per cell), also called Langhan's giant cells (16). In order to assess the role of cpn60.1 in such a cellular differentiation process, we measured the numbers of MC and MGC produced within the granulomatous structures, respectively, induced by the three strains under study: wild-type and complemented M. tuberculosis strains, which have both cpn60.1 and cpn60.2 genes, and the isogenic mutant, where only cpn60.2 is present. The numbers of giant cells (one to two nuclei per cell) generated in the presence of the three strains of M. tuberculosis were the same. However, exposure to the wild-type and complemented M. tuberculosis strains resulted in 30% of the giant cells being highly multinucleated. In contrast, in blood cells infected with the Δcpn60.1 mutant only 3.2% of the macrophages were multinucleated (Table 1).
TABLE 1.
MGC formation in human in vitro granulomas exposed to M. tuberculosis (H37Rv), the isogenic mutant lacking the cpn60.1 gene, and the complemented strain
| Duration of culture (days) | Mean % MGC ± SDa |
||
|---|---|---|---|
| Wild type | Δcpn60.1 mutant | Δcpn60.1 complemented mutant | |
| 3 | 0 | 0 | 0 |
| 6 | 0 | 0 | 0 |
| 9 | 30.3 ± 4.5 | 3.2 ± 1.6 | 28.7 ± 1.5 |
Results are expressed as the means of three individual experiments.
DISCUSSION
M. tuberculosis exerts a profound influence on the human immune system, and it has been almost 25 years since the discovery that T cells responsive to M. tuberculosis could produce, or protect against, the development of the chronic destructive adjuvant arthritis model in the rat (9, 30, 31). Surprisingly, the antigen to which these immune cells were raised was the Hsp65/Cpn60.2 protein of M. tuberculosis, and this protein was later implicated in the pathogenesis of models of other human autoimmune diseases such as diabetes (3) and atherosclerosis (33). In these early studies the Hsp65/Cpn60.2 protein was assumed to act simply as a very powerful T-cell immunogen which, in itself, is surprising, given the significant sequence homology with the human Hsp60 protein (3). However, in 1993 Friedland et al. revealed that Hsp65/Cpn60.2 could also act as a direct stimulating signal for human monocytes inducing the production of proinflammatory cytokines (5). This was the first evidence that Cpn60 proteins had protein moonlighting activity and, indeed, predated the term (11). Thus, it was possible that Hsp65/Cpn60.2 was, like IFN-γ or LPS, inducing a classically activated macrophage state able to present antigens and kill bacteria (19). However, while the exposure of human monocytes to M. tuberculosis Hsp65/Cpn60.2 induced the production of the same amount of cytokines as released by cells exposed to IFN-γ together with LPS, cells stimulated with Hsp65/Cpn60.2 did not show the expected increased expression of Fcγ receptors, major histocompatibility complex class II proteins or the release of reactive oxygen intermediates. Thus, M. tuberculosis Hsp65/Cpn60.2, unlike IFN-γ, is not inducing a classically activated state in human monocytes (22).
Following on from this early work by Friedland and Peeterman, a number of other groups reported that other bacterial Cpn60 proteins were able to activate monocytes (6). Henderson and coworkers reported that a potent osteolytic protein from an oral bacterium was the Cpn60 protein (12) and that this protein stimulated bone destruction by acting as a growth factor for the generation of the multinucleate osteoclast population of bone (27). However, it was found that the Hsp65/Cpn60.2 proteins of M. tuberculosis and Mycobacterium leprae were unable to promote bone breakdown (12). This led us to clone and express the gene encoding the second Cpn60 protein of M. tuberculosis, which was also found to be unable to promote osteoclast formation and bone breakdown (20). Further analysis of the recombinant Cpn60.1 protein revealed that it was a more potent and efficacious monocyte agonist than Hsp65/Cpn60.2 (17). However, it has only been in the past few years that major differences between the biological actions of the Cpn60.1 and the Hsp65/Cpn60.2 proteins have begun to appear. The major differences in these proteins appeared when we attempted to inactivate the genes encoding the Cpn10 and Cpn60 proteins in the virulent H37Rv strain of M. tuberculosis. This revealed that cpn60.2 and cpn10 were essential genes that could only be inactivated if a plasmid-based copy of each gene was provided for the bacterium. However, the cpn60.1 gene could be inactivated without any apparent phenotype (10). This suggested that Cpn60.1 did not function as a molecular chaperone, and this was confirmed by complementation studies in E. coli which revealed that M. tuberculosis cpn60.1 could not replace E. coli groEL. In contrast, the cpn60.2 gene could complement (10); however, it was recently found that neither mycobacterial cpn60 gene can complement an E. coli strain with a temperature-sensitive groEL44 allele, apparently due to their inability to form higher oligomeric forms (14, 26). Unexpectedly, M. tuberculosis H37Rv lacking a functional cpn60.1 gene grew in animals at the normal rate but failed to induce a granulomatous response, suggesting that this protein is directly associated with the ability of the tubercle bacilli to create granulomas with their associated giant cells (10). It was therefore surprising, when a more detailed analysis was made of the response of osteoclast precursor cells to the two M. tuberculosis 60-kDa chaperonins, that the Hsp65/Cpn60.2 protein has no stimulatory or inhibitory effects on these cells. In contrast, Cpn60.1 is a potent inhibitor of osteoclast formation acting over an extended period of the 7-day process of murine osteoclast formation. Inhibition is not associated with NF-κB or mitogen-activated protein kinase inhibition but is related to the inhibition of transcription of the key osteoclast transcription factor NFATc1 (34). This is not simply an in vitro artifact. Administration of small doses of Cpn60.1 to rats at the onset of adjuvant arthritis completely blocked the massive osteoclastic driven bone resorption seen in rats with this disease. Surprisingly, inhibition of bone resorption was not associated with inhibition of joint inflammation. In comparison, Hsp65/Cpn60.2 failed to inhibit joint inflammation or bone resorption.
These findings suggested that Cpn60.1 is a virulence factor of M. tuberculosis. The recent finding that Hsp65/Cpn60.2 is involved in the uptake of M. tuberculosis into macrophages (8) also suggests that this protein is also a virulence factor. The aim of the present study was to determine more about the biology of these two moonlighting proteins and to determine what differences in activity existed between them.
The first question we addressed is whether Cpn60.1 is released from M. tuberculosis. Using a specific ELISA for Cpn60.1, we have shown that M. tuberculosis in culture releases low but detectable levels of Cpn60.1. We used the Δcpn60.1 mutant as a control, and this gave no signal, showing the antibody is not cross-reacting with Hsp65/Cpn60.2. Low amounts of Cpn60.1 are already present in the supernatant of an early M. tuberculosis culture, suggesting that this protein may be released from live bacilli. We also found that macrophagelike cells infected with M. tuberculosis but not stimulated with IFN-γ released ng/ml quantities of Cpn60.1. There was no detectable Cpn60.1 in lysates of these J774 cells, suggesting that Cpn60.1 is rapidly transported to the exterior of the cell. When macrophagelike cells were exposed to IFN-γ, there was no recordable Cpn60.1 in these cells or in the medium supporting them. This may suggest that IFN-γ has an effect on the synthesis of Cpn60.1 in M. tuberculosis residing intracellularly. Similar findings have been observed by Schnappinger et al., who were studying changes in gene expression of M. tuberculosis in vitro and during infection of naive or IFN-γ-activated macrophages (28). In that study, significant changes in gene expression in M. tuberculosis were observed as soon as 4 h postinfection in naive macrophages, and the expression of Cpn60.1 at 24 h postinfection was increased ∼2.8 times compared to growth in vitro. In contrast, the synthesis of Cpn60.1 in M. tuberculosis residing inside a IFN-γ-activated macrophage was suppressed and did not significantly change for up to 48 h postinfection.
In contrast to Cpn60.1, we failed to detect Cpn60.2 in supernatants of M. tuberculosis liquid culture and M. tuberculosis-infected J774 cells. It should, however, be noted that Hickey et al. reported recently that Cpn60.2 is released from M. tuberculosis cultures, and it associates with the capsular material of this bacterium, where it binds to macrophage surface receptor CD43 (7, 8).
Turning to a comparison of the actions of both mycobacterial Cpn60 proteins on leukocytes, we first examined the ability of both proteins to activate proinflammatory cytokine production by whole human blood. We already knew the response of human monocytes to both proteins (they induced the synthesis of IL-1β, TNF-α, IL-6, IL-8, IL-10, IL-12, and granulocyte-macrophage colony-stimulating factor [17]) and were interested to see how a more complex leukocyte population would respond. In the present study, we monitored the production of IFN-γ, TNF-α, IL-1β, IL-4, IL-6, IL-8, IL-10, and IL-12. Exposure of whole blood to graded concentrations of M. tuberculosis Cpn60.1 or Cpn60.2 produced results discrepant from the findings with isolated monocytes (17). Of the eight cytokines measured, only three were produced by contact with these chaperonins. Unexpectedly, Hsp65/Cpn60.2 was the most potent in the induction of IL-1β and IL-6 production and was the only one of the two chaperonins to induce the synthesis of TNF-α. Neither Cpn60 protein was able to induce the synthesis of IFN-γ, IL-4, IL-8, IL-10, and IL-12. This suggests that these chaperonins interact with leukocytes, other than monocytes, generating networks of cell activation distinct from that produced by purified monocytes themselves. It also shows a major difference in cytokine induction in that only Hsp65/Cpn60.2 is able to induce TNF-α synthesis from this complex of leukocytes. This response is completely at variance with the response of whole blood and isolated blood monocytes to LPS in which the proinflammatory cytokines (such as IL-1β and TNF-α) are either produced at a higher level or at an equal level in whole blood compared to the production by isolated monocytes (4). We have previously shown that depletion of CD3+ T lymphocytes had no effect on IL-6 or IL-8 production by human monocytes (17). It remains to be determined whether the differences in cytokine production is due to some cell in whole blood, other than the T lymphocyte, which controls the responses to these chaperonins.
Up to this point our assumption had been that the mycobacterial chaperonins interacted with the same monocyte populations in the blood. To determine that this was the case, we used double labeling with antibodies to CD14 and Alexa Fluor 488-labeled mycobacterial Cpn60 proteins to determine that both proteins bound to monocytes. Surprisingly, the Cpn60.1 protein bound to virtually all circulating CD14+ monocytes. In contrast, the Hsp65/Cpn60.2 protein bound <50% of these cells. In the human, two major monocyte populations are proposed to be present in the circulation: CD14++ CD16− and CD14+ CD16+. The CD14+ CD16+ cells are characterized as proinflammatory since they produce TNF-α but little IL-10 when stimulated (36). From the flow cytometry, it would appear that Cpn60.1 interacts with both CD14-bearing monocyte populations, whereas the Hsp65/Cpn60.2 only interacts with one of these subsets.
Given that the M. tuberculosis Cpn60 proteins appear to bind to different subsets of human circulating monocytes, it could be assumed that they bind to different cell surface receptors. To confirm this, we conducted competition binding experiments with Alexa Fluor 488-labeled chaperonins. This clearly showed that unlabeled Cpn60.1 could compete for the binding of fluorescently labeled Cpn60.1. However, binding of the labeled Hsp65/Cpn60.2 protein could not be inhibited by excess unlabeled Cpn60.1. To ascertain whether the receptor for binding involves interaction with either of the two major proinflammatory TLRs, we used macrophages from TLR-deficient mice. Just to confirm that the results with these cells are not due to LPS contamination, we showed that proteinase K treatment of both Cpn60 proteins abolished their monocyte-stimulating activity. TLR4− macrophages failed to respond to either Cpn60.1 or Hsp65/Cpn60.2 or to LPS, but they did respond to the TLR2 agonist, zymosan. This seemed a clear-cut response. However, when the chaperonins were added to TLR2− macrophages there was, as would be expected, no significant change in stimulation by Cpn60.1. However, the response to Hsp65/Cpn60.2 was decreased by 80%. The LPS and zymosan controls showed the expected response. Thus, these two chaperonin 60 proteins differ in their cell surface receptors and in the TLR-dependent transduction mechanism producing cell stimulation.
It has been established that the MyD88-dependent pathway is the most common pathway for cell activation in response to TLR ligands such as LPS (21), and it was assumed that both M. tuberculosis Cpn60 proteins would activate it. However, using a dominant-negative MyD88 system it was found that Cpn60.1-dependent cell activation was partially dependent on MyD88, whereas Hsp65/Cpn60.2 showed only a minimum use of this adaptor protein. In contrast, LPS-induced macrophage activation was wholly MyD88 dependent. This reveals an important difference between these two chaperonins and between the chaperonins and LPS. If the activity of these chaperonins was solely due to LPS contamination, it could not account for this result.
The Δcpn60.1 mutant failed to generate a granulomatous response in mice, suggesting that the Cpn60.1 protein is necessary to induce the classical proinflammatory process induced by mycobacterial infections, which culminates with the formation of cellular aggregates called granulomas. One of the main signs of M. tuberculosis-induced granulomas is the presence of MGC. We have now tested this hypothesis in a human in vitro granuloma assay which allows the measurement of both MC and MGC formation. Incubation of human blood monocytes with the wild-type, complemented strain or the cpn60.1 isogenic mutant resulted in similar levels of production of multinucleated cells. However, when the state of multinucleation of these cells was measured, it was found that in the absence of the cpn60.1 gene almost no multinucleated giant cells were generated. This can be interpreted as suggesting that Cpn60.1 is a promoter of giant cell multinucleation, which is a hallmark of virulent mycobacterial species (16). Clearly then, Cpn60.2, which is present in all three strains of M. tuberculosis and is now known to be released by the bacterium (8), has no role in this multinucleation process. This is another example of distinct differences between these two chaperonins (Table 2). The M. tuberculosis Cpn60 proteins are clearly important virulence factors involved in signaling to a variety of immune cells. The findings reported here extend our understanding of the differences in these two highly homologous proteins as modulators of leukocyte activation. It is clear that the human and murine immune systems have evolved to discriminate between these highly homologous Cpn60 proteins. The reasons for this, and the influence it has on the pathogenesis of tuberculosis, still requires more detailed investigation.
TABLE 2.
Comparison of the moonlighting actions of the M. tuberculosis Cpn60 proteins
| Activity | Observationa |
|
|---|---|---|
| Cpn60.1 | Cpn60.2 | |
| Induction of monocyte cytokine synthesis | ++ | + |
| Induction of whole-blood cytokine synthesis | + | ++ |
| Induction of classic macrophage activation | ? | - |
| Induction of bone breakdown in vitro | - | - |
| Inhibition of bone breakdown | ++ | - |
| Inhibition of osteoclast formation | ++ | - |
| Release from bacterium | + | + |
| % Binding to circulating monocytes | 86 | 47 |
| Role of TLR2 in monocyte activation | - | Si |
| Role of TLR4 in monocyte activation | ++ | ++ |
| Involvement of Myd88 | St | Wk |
St, strong; Wk, weak; Si, significant.
Acknowledgments
B.H. and A.R.M.C. acknowledge funding from the MRC in the form of an industrial studentship to A.C. We also thank Helperby Therapeutics, plc, for cofunding this study.
Editor: A. Camilli
Footnotes
Published ahead of print on 26 April 2010.
REFERENCES
- 1.Coates, A., A. Cehovin, and Y. Hu. 2008. Chaperonin 60 and macrophage activation, p. 160-172. In D. J. Chadwick and J. Goode (ed.), The biology of extracellular chaperones. Novartis Foundation Symposium 291. John Wiley & Smith, Chichester, United Kingdom. [DOI] [PubMed]
- 2.Coates, A. R., T. M. Shinnick, and R. J. Ellis. 1993. Chaperonin nomenclature. Mol. Microbiol. 8:787. [DOI] [PubMed] [Google Scholar]
- 3.Cohen, I. R., and D. B. Young. 1991. Autoimmunity, microbial immunity and the immunological homunculus. Immunol. Today 12:105-110. [DOI] [PubMed] [Google Scholar]
- 4.De Groote, D., P. F. Zangerle, Y. Gevaert, M. F. Fassotte, Y. Beguin, F. Noizat-Pirenne, J. Pirenne, R. Gathy, M. Lopez, I. Dehart, et al. 1992. Direct stimulation of cytokines (IL-1β, TNF-α, IL-6, IL-2, IFN-γ, and GM-CSF) in whole blood. I. Comparison with isolated PBMC stimulation. Cytokine 4:239-248. [DOI] [PubMed] [Google Scholar]
- 5.Friedland, J. S., R. Shattock, D. G. Remick, and G. E. Griffin. 1993. Mycobacterial 65-kD heat shock protein induces release of proinflammatory cytokines from human monocytic cells. Clin. Exp. Immunol. 91:58-62. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Henderson, B., E. Allan, and A. R. Coates. 2006. Stress wars: the direct role of host and bacterial molecular chaperones in bacterial infection. Infect. Immun. 74:3693-3706. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Hickey, T. B. 2009. Identification of Cpn60.2 as a surface ligand of Mycobacterium tuberculosis that facilitates bacterial association with macrophages via CD43. Ph.D. thesis. University of British Columbia, Vancouver, Canada.
- 8.Hickey, T. B., L. M. Thorson, D. P. Speert, M. Daffe, and R. W. Stokes. 2009. Mycobacterium tuberculosis Cpn60.2 and DnaK are located on the bacterial surface, where Cpn60.2 facilitates efficient bacterial association with macrophages. Infect. Immun. 77:3389-3401. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Holoshitz, J., Y. Naparstek, A. Ben-Nun, and I. R. Cohen. 1983. Lines of T lymphocytes induce or vaccinate against autoimmune arthritis. Science 219:56-58. [DOI] [PubMed] [Google Scholar]
- 10.Hu, Y., B. Henderson, P. A. Lund, P. Tormay, M. T. Ahmed, S. S. Gurcha, G. S. Besra, and A. R. Coates. 2008. A Mycobacterium tuberculosis mutant lacking the groEL homologue cpn60.1 is viable but fails to induce an inflammatory response in animal models of infection. Infect. Immun. 76:1535-1546. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Jeffery, C. J. 2009. Moonlighting proteins: an update. Mol. Biosystems 5:345-350. [DOI] [PubMed] [Google Scholar]
- 12.Kirby, A. C., S. Meghji, S. P. Nair, P. White, K. Reddi, T. Nishihara, K. Nakashima, A. C. Willis, R. Sim, M. Wilson, et al. 1995. The potent bone-resorbing mediator of Actinobacillus actinomycetemcomitans is homologous to the molecular chaperone GroEL. J. Clin. Invest. 96:1185-1194. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Kong, T. H., A. R. Coates, P. D. Butcher, C. J. Hickman, and T. M. Shinnick. 1993. Mycobacterium tuberculosis expresses two chaperonin-60 homologs. Proc. Natl. Acad. Sci. U. S. A. 90:2608-2612. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Kumar, C. M., G. Khare, C. V. Srikanth, A. K. Tyagi, A. A. Sardesai, and S. C. Mande. 2009. Facilitated oligomerization of mycobacterial GroEL: evidence for phosphorylation-mediated oligomerization. J. Bacteriol. 191:6525-6538. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Lamb, J. R., V. Bal, J. B. Rothbard, A. Mehlert, P. Mendez-Samperio, and D. B. Young. 1989. The mycobacterial GroEL stress protein: a common target of T-cell recognition in infection and autoimmunity. J. Autoimmun. 2(Suppl.):93-100. [DOI] [PubMed] [Google Scholar]
- 16.Lay, G., Y. Poquet, P. Salek-Peyron, M. P. Puissegur, C. Botanch, H. Bon, F. Levillain, J. L. Duteyrat, J. F. Emile, and F. Altare. 2007. Langhans giant cells from Mycobacterium tuberculosis-induced human granulomas cannot mediate mycobacterial uptake. J. Pathol. 211:76-85. [DOI] [PubMed] [Google Scholar]
- 17.Lewthwaite, J. C., A. R. Coates, P. Tormay, M. Singh, P. Mascagni, S. Poole, M. Roberts, L. Sharp, and B. Henderson. 2001. Mycobacterium tuberculosis chaperonin 60.1 is a more potent cytokine stimulator than chaperonin 60.2 (Hsp65) and contains a CD14-binding domain. Infect. Immun. 69:7349-7355. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Maguire, M., A. R. Coates, and B. Henderson. 2002. Chaperonin 60 unfolds its secrets of cellular communication. Cell Stress Chaperones 7:317-329. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Martinez, F. O., L. Helming, and S. Gordon. 2008. Alternative activation of macrophages: an immunologic functional perspective. Annu. Rev. Immunol. 27:451-483. [DOI] [PubMed] [Google Scholar]
- 20.Meghji, S., P. A. White, S. P. Nair, K. Reddi, K. Heron, B. Henderson, A. Zaliani, G. Fossati, P. Mascagni, J. F. Hunt, M. M. Roberts, and A. R. Coates. 1997. Mycobacterium tuberculosis chaperonin 10 stimulates bone resorption: a potential contributory factor in Pott's disease. J. Exp. Med. 186:1241-1246. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.O'Neill, L. A., and A. G. Bowie. 2007. The family of five: TIR-domain-containing adaptors in Toll-like receptor signaling. Nat. Rev. Immunol. 7:353-364. [DOI] [PubMed] [Google Scholar]
- 22.Peetermans, W. E., C. J. Raats, J. A. Langermans, and R. van Furth. 1994. Mycobacterial heat-shock protein 65 induces proinflammatory cytokines but does not activate human mononuclear phagocytes. Scand. J. Immunol. 39:613-617. [DOI] [PubMed] [Google Scholar]
- 23.Peyron, P., J. Vaubourgeix, Y. Poquet, F. Levillain, C. Botanch, F. Bardou, M. Daffe, J. F. Emile, B. Marchou, P. J. Cardona, C. de Chastellier, and F. Altare. 2008. Foamy macrophages from tuberculous patients' granulomas constitute a nutrient-rich reservoir for Mycobacterium tuberculosis persistence. PLoS Pathog. 4:e1000204. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Puissegur, M. P., C. Botanch, J. L. Duteyrat, G. Delsol, C. Caratero, and F. Altare. 2004. An in vitro dual model of mycobacterial granulomas to investigate the molecular interactions between mycobacteria and human host cells. Cell Microbiol. 6:423-433. [DOI] [PubMed] [Google Scholar]
- 25.Puissegur, M. P., G. Lay, M. Gilleron, L. Botella, J. Nigou, H. Marrakchi, B. Mari, J. L. Duteyrat, Y. Guerardel, L. Kremer, P. Barbry, G. Puzo, and F. Altare. 2007. Mycobacterial lipomannan induces granuloma macrophage fusion via a TLR2-dependent, ADAM9- and β1 integrin-mediated pathway. J. Immunol. 178:3161-3169. [DOI] [PubMed] [Google Scholar]
- 26.Qamra, R., V. Srinivas, and S. C. Mande. 2004. Mycobacterium tuberculosis GroEL homologues unusually exist as lower oligomers and retain the ability to suppress aggregation of substrate proteins. J. Mol. Biol. 342:605-617. [DOI] [PubMed] [Google Scholar]
- 27.Reddi, K., S. Meghji, S. P. Nair, T. R. Arnett, A. D. Miller, M. Preuss, M. Wilson, B. Henderson, and P. Hill. 1998. The Escherichia coli chaperonin 60 (groEL) is a potent stimulator of osteoclast formation. J. Bone Miner. Res. 13:1260-1266. [DOI] [PubMed] [Google Scholar]
- 28.Schnappinger, D., S. Ehrt, M. I. Voskuil, Y. Liu, J. A. Mangan, I. M. Monahan, G. Dolganov, B. Efron, P. D. Butcher, C. Nathan, and G. K. Schoolnik. 2003. Transcriptional adaptation of Mycobacterium tuberculosis within macrophages: insights into the phagosomal environment. J. Exp. Med. 198:693-704. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Tormay, P., A. R. Coates, and B. Henderson. 2005. The intercellular signaling activity of the Mycobacterium tuberculosis chaperonin 60.1 protein resides in the equatorial domain. J. Biol. Chem. 280:14272-14277. [DOI] [PubMed] [Google Scholar]
- 30.van Eden, W., J. E. Thole, R. van der Zee, A. Noordzij, J. D. van Embden, E. J. Hensen, and I. R. Cohen. 1988. Cloning of the mycobacterial epitope recognized by T lymphocytes in adjuvant arthritis. Nature 331:171-173. [DOI] [PubMed] [Google Scholar]
- 31.van Eden, W., R. van der Zee, and B. Prakken. 2005. Heat-shock proteins induce T-cell regulation of chronic inflammation. Nat. Rev. Immunol. 5:318-330. [DOI] [PubMed] [Google Scholar]
- 32.World Health Organization. 2008. Global tuberculosis control: surveillance, planning, financing: WHO Report 2008. World Health Organization, Geneva, Switzerland.
- 33.Wick, G., M. Knoflach, and Q. Xu. 2004. Autoimmune and inflammatory mechanisms in atherosclerosis. Annu. Rev. Immunol. 22:361-403. [DOI] [PubMed] [Google Scholar]
- 34.Winrow, V. R., J. Mesher, S. Meghji, C. J. Morris, M. Maguire, S. Fox, A. R. Coates, P. Tormay, D. R. Blake, and B. Henderson. 2008. The two homologous chaperonin 60 proteins of Mycobacterium tuberculosis have distinct effects on monocyte differentiation into osteoclasts. Cell Microbiol. 10:2091-2104. [DOI] [PubMed] [Google Scholar]
- 35.Young, D. B., M. D. Perkins, K. Duncan, and C. E. Barry III. 2008. Confronting the scientific obstacles to global control of tuberculosis. J. Clin. Invest. 118:1255-1265. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Ziegler-Heitbrock, L. 2007. The CD14+ CD16+ blood monocytes: their role in infection and inflammation. J. Leukoc. Biol. 81:584-592. [DOI] [PubMed] [Google Scholar]