Abstract
Escherichia coli chaperonin (cpn) 60 (groEL) is a protein-folding oligomer lacking tryptophan residues that copurifies with tryptophan-containing proteins and peptides. Cpn 60 is a major immunogen in infectious diseases, and evidence suggests that groEL and mycobacterial cpn 60s can induce cytokine synthesis, stimulate cytokine-dependent bone resorption, and up-regulate expression of vascular endothelial cell adhesion molecules. Whether such activities are due to the cpn 60 or to the copurifying/contaminating proteins/peptides has not been determined. Here we report a method for removing the protein contaminants of groEL and demonstrate that this, essentially homogeneous, groEL remains a potent inducer of human monocyte IL-1β and IL-6 production. Contaminating peptides had no cytokine-inducing activity and did not synergize with purified groEL. The LPS inhibitor polymyxin B and the CD14-neutralizing Ab MY4 had no inhibitory action on groEL demonstrating that activity is not due to LPS contamination. Heating groEL had no effect on its capacity to stimulate human monocytes to secrete IL-6. Proteolysis of groEL with trypsin, sufficient to produce low molecular mass peptides, also had no inhibitory effect. Thus, we conclude that groEL is a potent inducer of monocyte proinflammatory cytokine production, which acts through the binding of nonconformational peptide domains that are conserved after proteolysis. These data suggest that if groEL was released from bacteria it could induce prolonged tissue pathology by virtue of its cytokine-inducing activity and its resistance to proteolytic inhibition of bioactivity.
Protein denaturation is a major problem for cells and one that was overcome by the early evolution of proteins with the catalytic capacity to fold and refold proteins. These constitutive protein-folding proteins are called molecular chaperones and exist within cells as families of different molecular mass proteins (e.g., 90, 70, 60, and 30 kDa). A subpopulation of the molecular chaperones are products of inducible genes and are known as heat shock or stress proteins. The importance of the molecular chaperones for cell survival can be inferred from the significant conservation of primary sequence found between bacterial and human proteins (1, 2).
In spite of this significant structural homology, bacterial molecular chaperones, and in particular the 60-kDa molecule known as chaperonin 60 (cpn 60)3, are major immunogens in bacterial infections (3), and cross-reactivity between bacterial cpn 60 and the structurally similar human proteins has been suggested to play a role in autoimmune diseases such as diabetes mellitus and rheumatoid arthritis (4). Furthermore, there is increasing evidence that extracellular molecular chaperones have a range of biologic actions (5). Of relevance to immunity are reports that cpn 60 from various bacteria induce the release of certain proinflammatory cytokines (6, 7, 8) and that the Mycobacterium tuberculosis cpn 60 induces the expression of CD54, CD62e, and CD106 on vascular endothelial cells in an IL-1- and TNF-independent manner (9). We have demonstrated that the Escherichia coli cpn 60 (groEL), but not that of M. tuberculosis or Mycobacterium leprae, is a potent stimulator of murine bone resorption in vitro (10) and that such resorption can be inhibited by neutralizing the activity of IL-1 with IL-1ra (11). This suggested that in this murine bone explant system, bone resorption was ultimately due to the capacity of the bacterial cpn 60 to induce cytokine gene transcription. However, the bone cell population responsible for the synthesis of cytokines (myeloid or mesenchymal) has not been identified.
GroEL is the cpn 60 that has received most attention from chemists and biochemists. It has been known for some years that purified natural and cloned groEL harbors contaminating proteins/peptides. This has been determined by the presence of tryptophan-fluorescing material in groEL (whose gene sequence contains no codons for tryptophan), and by the numbers of additional protein bands seen in SDS-PAGE of groEL (12, 13). Until now, it has not proved possible to completely remove this material, therefore the nonfolding activity of groEL could be due to these contaminating species and not groEL itself. In this paper, we report on the use of a combination of dye-binding affinity chromatography and ATP-gradient elution to remove proteins and peptides contaminating groEL preparations and demonstrate that the essentially homogeneous groEL so produced is an active cytokine-inducing molecule with human monocytes. Furthermore, we report that heating or tryptic digestion of groEL has no significant effect on its cytokine-inducing activity, suggesting that nonconformational epitopes in this molecule are responsible for activating monocytes.
Materials and Methods
Preparation of homogeneous groEL Standard (groELST)
GroEL was cloned, overexpressed in E. coli, and purified as described (14, 15) with the final material eluting from a Q-Sepharose column being ∼95% pure as judged by SDS-PAGE and silver staining. This was designated groELST. In some experiments the groELST was dialyzed against PBS and applied to a Detoxigel (polymyxin B) column (Pierce & Warriner, Chester, U.K.) to remove any E. coli LPS. The eluate was concentrated using Minicon concentrators with PM3 (Amicon, Beverly, MA) membranes to a concentration of 300 μg/ml. The LPS content of the material before, and after, column elution was determined by the Limulus amoebocyte lysate assay as described in Reference (16).
While acting as a functional molecular chaperone, the groELST preparations demonstrated fluorescence emission at a peak of approximately 350 nm, indicating the presence of tryptophan. GroEL does not contain tryptophan and so the fluorescence indicated the presence of contaminating peptides/proteins. The contaminating tryptophan-containing material has been reported to be removed by passing the proteins through a Reactive Red column (17). However, we did not find that this removed enough of the contaminating bands seen on SDS-PAGE, suggesting that not all contaminating proteins contained tryptophan. Modifications of this technique to remove all contaminating proteins were examined and a successful technique was developed. GroELST was dialyzed overnight against 20 mM Tris-HCl, pH 7.6, containing 5 mM KCl, 1 mM β-mercaptoethanol (β-ME) (buffer A). On the day of use, groELST was supplemented with 5 mM MgCl2 and 2 mM ATP and dialyzed for 1 h at room temperature against the buffer B (i.e., buffer A supplemented with 5 mM MgCl2 and 2 mM ATP) before loading onto a Reactive Red 120 agarose column (Sigma) (2.6 × 28 cm). The groELST was allowed to equilibrate on the column for 30 min before column washing with the buffer B. The groELST was then eluted at a flow rate of 1.5 ml/min with a linear gradient of buffer B to buffer A (a gradient volume of 400 ml). GroEL eluted at the end of the gradient. The fractions (15 ml) containing protein were identified by absorbance at 280 nm. The column was washed with various solutions and it was found that deionized water removed the non-groEL-bound protein. To determine which fraction contained the tryptophan, a fluorescence spectrum (excitation at 280 nm) of each protein fraction was monitored using a Shimadzu RF 5001PC spectrofluorimeter (Shimadzu, Milton Keynes, U.K.). The fractions containing the tryptophan-free groEL, designated groEL Reactive Red (groELRR), were combined, concentrated using an Amicon N2-pressure concentrator with 100-kDa membrane, and dialyzed into 50 mM Tris-HCl, pH 7.6, containing 2 mM DTT, and stored at −20°C in 50% glycerol. The aliquot containing the contaminants was concentrated on a 3-kDa cut-off membrane and stored under the same conditions.
Concentrations of groELST and groELRR
The concentrations of these various proteins and preparations of these proteins were estimated for biochemical protein-folding studies as follows: groES by absorption at 280 nm: extinction coefficient ε280 = 4.72 × 103 M−1 cm−1 (15). GroELST by absorption at 280 nm: extinction coefficient ε280 = 2.92 × 104 M−1 cm−1 (15). GroELRR by quantitative amino acid analysis. For studies of the cytokine-inducing actions of the groEL preparations use was made of the commercial Bradford assay (Bio-Rad, Hemel Hempstead, UK).
Measurement of tryptophan fluorescence
Fluorescence was measured by using a Shimadzu 5001PC spectrofluorimeter fitted with a thermostated cuvette holder at 20°C with a 5-nm slit width and a 1-cm cuvette. Measurements were made on 0.5 μM groEL in the presence of 6 M guanidine hydrochloride (GuHCl), pH 7.0, and 30 mM β-ME. Both groEL and the reference substances, N-acetyltryptophanamide and l-tyrosine (also in 6 M GuHCl, pH 7.0, and 30 mM β-ME) were excited at either 295 nm, to selectively monitor the tryptophan fluorescence, or at 280 nm, to monitor both the tryptophan and tyrosine fluorescence. The tryptophan content was calculated from the intensity of light emission at 354 nm and 303 nm, respectively, using calibration curves fitted through the fluorescence of standard solutions of N-acetyltryptophanamide.
Amino acid analysis
GroELRR was hydrolyzed in vacuo for 18 h at 110°C in 50 μl of 6 N HCl (analysis grade) (Perkin-Elmer Applied Biosystems Division, Warrington, U.K.). A total of 200 pmol of norleucine (Sigma, Poole, U.K.) was also added before hydrolysis as an internal standard. Hydrolysates were loaded onto a PE-ABD 420 amino acid analyzer with on-line chromatographic separation of the amino acid derivatives (PE-ABD 130A HPLC) followed by data analysis (920A Data Analysis Module). All the instrumentation was operated according to the manufacturers’ instructions. The amino acid analysis was repeated four times to determine assay reproducibility.
Protein-folding assay
Pig heart mitochondrial malate dehydrogenase (mMDH, Boehringer Mannheim, Lewes U.K.), groEL, and groES were dialyzed against 150 mM sodium phosphate (pH 7.6), 2 mM β-ME, and 1 mM EDTA. mMDH was denatured by dilution of an aliquot of stock solution to 0.3 mg/ml (4.3 μM dimer concentration) into denaturing buffer (150 mM sodium phosphate (pH 7.6), 20 mM β-ME, 10 mM EDTA, and 3 M GuHCl). This solution was incubated at 20°C for 2 h to ensure full structural equilibration. Naturation of the mMDH was initiated by diluting the denatured protein to a concentration of 10 μg/ml (143 nM, 30-fold dilution) into phosphate buffer containing 20 mM β-ME, 10 mM MgCl2, 10 mM KCl, 2 mM ATP, and either 858 nM groEL plus 1716 nM groES (oligomer concentrations) (assisted folding buffer) or the same buffer not containing chaperonins (spontaneous folding buffer). The resulting solutions were incubated for several hours at 20°C and assayed for mMDH activity at given time points by determining the enzymatic activity of 20-μl aliquots introduced into an assay buffer aliquot (980 μl) comprised of 150 mM sodium phosphate (pH 7.6), 2 mM β-ME, 0.5 mM oxaloacetate, and 0.2 mM NADH (Sigma) as described (14, 15).
Cell culture
The methods used to prepare and culture human peripheral blood monocytes have been described (18), and in all studies cells were cultured at 37°C in an atmosphere of 5% CO2/air. Human peripheral blood neutrophils were prepared from buffy coat residues as described by Au et al. (19). Two myelomonocytic cell lines were used—U937 (American Type Culture Collection, Manassas, VA) and Mono-Mac-6 cells (kindly provided by Dr. Ziegler-Heitbrock, Munich, Germany). U937 cells were cultured in RPMI 1640 medium (Life Technologies, Paisley, U.K.) containing 2 mM glutamine (Life Technologies), 100 U/ml penicillin/streptomycin (Life Technologies) and 10% FCS (ICN, Thame, U.K.). Mono-Mac-6 cells were cultured as described previously (20).
Stimulation of cells with groEL
All myelomonocytic cells were cultured in 24-well plates at a cell density of 2 × 106 cells/ml. The human neutrophils were used at a cell density of 5 × 104 cells/ml, again in 24-well plates. All cells were exposed to concentrations of groEL in the range of 1 ng/ml to 10 μg/ml. In all experiments a highly purified preparation of LPS (international standard 84/650 from the National Institute for Biologic Standards and Control, London, U.K.) was added to separate cultures to ensure that cells were responsive. This preparation was used in all other studies of LPS reported in this paper. Cells were incubated in the presence of groEL or LPS for 18 h, then the medium was collected into Eppendorf tubes and immediately frozen to −70°C and stored at this temperature until the cytokine content of the media was tested by ELISA.
Cytokine assays
The concentration of IL-1β or IL-6 released into the medium by cultured cells was measured by two-site ELISA as described (20). The lower limit of sensitivity of both assays was 4 pg/ml.
RT–PCR
Human peripheral blood monocytes, prepared by density gradient centrifugation and adherence as described (19), were plated at a density of 5 × 106 cells/ml in 24-well culture dishes (Life Technologies) and allowed to adhere for 2 h. Cells were washed and exposed to 1 μg/ml intact or trypsinized groEL or to 10 ng/ml LPS for various times ranging from 1 to 18 h. At each time point the media were collected for cytokine assay and then the total RNA was extracted from each of the cell cultures by the method of Chomczynski and Sacchi (21). The cDNA preparation and semiquantitative PCR for IL-6 and IL-1β were performed essentially as described (18) but using a protocol of 30 cycles of PCR. Glyceraldehyde-3-phosphate dehydrogenase amplification was used as an internal mRNA standard (22), again using a semiquantitative PCR protocol of 25 cycles. The number of cycles was set to produce submaximal levels of PCR product, that is, the reaction was terminated during the exponential phase.
Role of CD14 in groEL-induced cytokine synthesis
To determine whether groEL bound to the CD14 receptor to stimulate cytokine synthesis, human monocytes were preincubated with the anti-CD14-blocking mAb MY4 (Coulter, Luton, U.K.) at 5 μg/ml for 1 h before the addition of groEL or highly purified LPS. Cells were cultured overnight and then the supernatant was removed and assayed for IL-6 as described. The LPS-binding antibiotic polymyxin B was also used, at a concentration of 20 μg/ml, to determine the possible role of contaminating E. coli LPS on the biologic activity of groEL.
Influence of heating and trypsinization on groEL-induced IL-6 synthesis
Purified groEL was heated to 80°C for 1 h and was then tested for IL-6-inducing activity on human monocytes and compared with a similar aliquot of groEL maintained at room temperature for 1 h. To determine the sensitivity to proteolytic enzymes, the GroELST (50 μg/ml) was digested with soluble trypsin, both standard grade and sequencing grade (Sigma), at a 20:1 (w/w) ratio, in 50 mM Tris, pH 7.8 (23). Controls included trypsin alone, trypsin and trypsin inhibitor, or untrypsinized groEL. Aliquots were removed at various times and the reaction was terminated by the addition of 4 mM PMSF. Trypsin-bound beads (Sigma) were also used in these experiments to allow the rapid removal of the proteolytic enzyme from the reaction by centrifugation. GroELST and groELRR at a concentration of 40 μg/ml were incubated with trypsin beads (2.5 U/ml) for 4 h. Digestion of groEL was determined by SDS-PAGE and the trypsinized material was then tested for its IL-6-inducing activity with human monocytes. Lower concentrations of groEL were also trypsinized for equivalent times but were not analyzed by SDS-PAGE.
Effect of a groEL-derived peptide
The peptide GENEEQNVGIK, which encompasses residues 431–442 in the groEL of Actinobacillus actinomycetemcomitans, was synthesized (Genosys, Pampisford, U.K.) (and assessed as being 95% pure) and was tested at concentrations up to and including 10 μg/ml for its ability to stimulate human monocyte cytokine synthesis.
Results
Preparation of groELST
The standard preparations (14, 15) of recombinant groELST that we produced initially exhibited tryptophan fluorescence and were estimated to contain 0.75 mol tryptophan/mol groELST (Fig. 1⇓). On overloaded gels of the groELST a large number of additional protein bands with molecular masses in the range of >60 to <14 kDa were visible (Fig. 2⇓). After passage through a Reactive Red column in the presence of ATP and elution from the column, the groELRR contained an estimated <0.1 mol tryptophan/mol groEL (Fig. 1⇓) and on overloaded SDS-PAGE the groELRR had lost the presence of the supernumerary bands (Fig. 2⇓). Elution of the column with deionized water resulted in the elution of the contaminating peptides/proteins, which had a wide range of molecular masses (Fig. 2⇓). The essentially homogeneous groELRR retained its full protein-folding activity as compared with the standard preparation (Fig. 3⇓). The purity of the groEL after elution from the Reactive Red column was established by analysis of the amino acid composition. The percentage of each amino acid was found to be almost precisely that predicted from the amino acid sequence (Table I⇓).
Fluorescence spectrum of groEL before (○) and after (▵) elution from a Reactive Red column. The absorption spectrum of N-acetyltryptophanamide (+) and of tyrosine (×) is shown for comparison. All samples were denatured in 6 M GuHCl, pH 7.0, with 30 mM β-ME as described in Materials and Methods.
Fifteen percent SDS-PAGE (silver stained) of material examined for tryptophan fluorescence. Lane 1 is the m.w. marker. Lane 2 is the original groELST preparation loaded at a concentration of 8 μg/ml. Lane 3 is the same loading of groELRR eluted from the Reactive Red column. Lane 4 is the peptide material associated with 100 mg groELST binding to the Reactive Red column after removal of the groEL and elution from the column by deionized water.
Time course of groEL-assisted and spontaneous refolding of denatured mMDH. The spontaneous folding occurred in the absence of groEL (▴). The extent of folding in the presence of groELST (•) and groELRR (○) was increased by both “forms” of groEL, with the same yield of refolded mMDH being obtained in both cases.
Amino acid composition of groELRR compared with predicted composition
The LPS content of both the purified groELST and groELRR was low (3–6 pg/μg protein) and at the limit of the Limulus amoebocyte assay used. Passage through a Detoxigel column had only a minimal effect on the apparent LPS content of the groEL. It was noted, however, that the Detoxigel column had the capacity to bind groEL.
Induction of cytokine synthesis by groELST and groELRR
Human peripheral blood monocytes exposed to groELST showed a reproducible (five experiments) dose-dependent secretion of both IL-1β and IL-6. Indeed, such induction of cytokine release was found at concentrations as low as 10 ng/ml (12 pM) of groELST (Fig. 4⇓). GroELRR produced essentially the same response as the starting material (groELST at concentrations of 1 and 10 μg/ml). However, the homogenous groEL was reproducibly less active than the starting material at lower concentrations (Fig. 4⇓). In contrast to its ability to stimulate human monocytes, groELST did not induce the production of IL-6 or IL-1β by the two myelomonocytic cell lines or by human neutrophils (results not shown).
Dose response of groEL-induced cytokine secretion from human peripheral blood monocytes. A. IL-6 secretion from human peripheral blood monocytes induced by groELST (○) and groELRR (▪). Results are expressed as the mean and SD of three replicate cultures. B. As A, but showing IL-1β secretion.
The contaminating peptides removed from the groELST were concentrated and tested for their ability to stimulate human monocyte cytokine synthesis. Preparations of these contaminating peptides failed to show any activity with a number of different isolates of human monocytes (three experiments: results not shown). The possibility that these contaminating peptides may act in synergy with the groEL was tested by incubating various concentrations of the contaminating peptides with a suboptimal concentration, (0.1 μg/ml), of groELRR. Such additions also failed to induce cytokine-stimulating activity; the residual activity in C being due to contaminating groEL (Fig. 5⇓).
Determination of possible synergistic interactions between groELRR and the peptides associated with groELST in terms of the induction of human monocyte IL-6 synthesis. IL-6 produced by cells exposed to: 1 μg/ml groELST (lane A), 0.1 μg/ml groELST (lane B), 0.1 μg/ml groELRR plus contaminating peptides from 1.1 μg groELST (lane C), 0.1 μg/ml groELRR plus contaminating peptides from 0.26 μg groELST (lane D), 0.1 μg/ml groELRR plus contaminating peptides from 0.05 μg groELST (lane E), 0.1 μg/ml groELRR (lane F), and 1 μg/ml groELRR (lane G). The results are expressed as the mean and SD of three replicate cultures.
Effect of polymyxin B and anti-CD14 on groELST activity
The IL-6-stimulating activity of E. coli LPS was completely blocked by both polymyxin B and by the anti-CD14 mAb MY4. In contrast, polymyxin B had no inhibitory effect on groEL-induced IL-6 synthesis. The anti-CD14 mAb MY4 also had no effect on the activity of groELST. However, a surprising finding was that in a number of assays (three out of six) the anti-CD14 mAb caused an elevation of IL-6 synthesis by human monocytes (Fig. 6⇓).
The effect of addition of polymyxin B (PB) or MY4, an anti-CD14 mAb (αCD14) on the secretion of IL-6 by human monocytes stimulated either with groELST or with E. coli LPS. Results are expressed as the mean and SD and are representative of three replicate experiments.
Effect of heat or proteolysis on the cytokine-inducing activity of groELST
Heating groELST had no significant effect on the ability of this protein to stimulate IL-6 synthesis (results not shown). A similar finding was made when groELST was exposed to soluble trypsin. Trypsin-coated beads caused a rapid breakdown of groELST and groELRR, with the products being of low molecular mass as assessed by SDS-PAGE. However, in spite of a complete loss of the 60-kDa band on SDS-PAGE, there was no significant decrease in the IL-6-stimulating activity of groELST or groELRR following trypsin digestion (Fig. 7⇓). Trypsinization of lower concentrations of groEL than used in these studies utilizing SDS-PAGE also failed to show significant loss of activity. Comparison of the kinetics of IL-1β and IL-6 mRNA transcription (as indirectly assessed by PCR amplification of intracellular mRNA and protein secretion) from cells treated with either intact or trypsinized groELST or groELRR (incubated with trypsin beads for 2 h) showed that, at the concentration chosen, there was essentially no difference in the rates of cytokine induction by the two grades of groEL (Fig. 8⇓). As has been reported, mRNA for IL-1β was always found in unstimulated monocytes; mock-treated cells showed no induction of IL-6 mRNA (result not shown).
The effect of trypsinization of groELST on its ability to stimulate human monocytes to secrete IL-6. A, IL-6-inducing activity of groEL (1.0 μg/ml), undigested (0) and trypsin digested for 15, 60 and 120 min. The term “media” shows the IL-6 production by unstimulated cells. Results are typical of three replicate experiments. Results are expressed as mean and SD of three replicate cultures. B, Time course of trypsinization on the SDS-PAGE profile of groELST (0.25 μg). The numbers represent the time of exposure to trypsin in minutes: 0 min (untreated), 15, 60, and 120 min.
Comparison of the kinetics of IL-6 protein synthesis and release and the intracellular mRNA levels for IL-1β and IL-6. A. Upper panel, kinetics of release of immunoreactive IL-6 from human monocytes stimulated with intact groELST (○) or groELRR (▪) at a concentration of 1 μg/ml. Results are expressed as the mean and SD of three replicate cultures. Lower panel, RT-PCR showing the kinetics of appearance of the mRNA for IL-1β or IL-6. IL-1β message was routinely present in cells at zero time. B, As A, but with monocytes stimulated with trypsinized groELST (○) groELRR (▪).
Activity of a groEL peptide
The peptide GENEEQNVGIK, which contains residues 431 to 442 of the cpn 60 of A. actinomycetemcomitans, was synthesized and tested for cytokine-inducing activity, but proved to be negative in all assays.
Discussion
In a recent review, chaperonins have been claimed to be “one of the most potent stimulators of the immune system” (3). This is clearly seen by the fact that the cpn 60 molecules of bacteria such as E. coli and M. tuberculosis are potent immunogens giving rise to significant titers of serum Abs (24). In addition, in mice immunized with M. tuberculosis, 20% of the T lymphocytes reacting to the bacteria responded to the cpn 60 (25). Are the 60-kDa chaperonins of bacteria anything more than potent immunogenic proteins? Recent evidence has suggested that the mycobacterial cpn 60 can stimulate monocytes to produce various proinflammatory cytokines (6, 7, 8) and can also induce vascular endothelial cells to express adhesion receptors in a noncytokine-dependent manner (9). Thus, cpn 60 may play a positive role in the induction of immune responses and be a mediator of immunopathology.
Most of the work on cpn 60 has concentrated on the E. coli protein groEL, and has focused on the mechanisms by which groEL folds proteins. Although groEL has been crystallized, and its structure analyzed by x-ray crystallography to a resolution of <3Å (26), it is established that preparations of this protein (natural or recombinant) are contaminated with a variety of proteins and peptides. This is seen by the finding that preparations of groEL (which contains no tryptophan residues) exhibit tryptophan fluorescence and, in SDS-PAGE gels, this protein produces a series of silver-stained bands with molecular masses ranging from >60 to <14 kDa (12, 13). These proteins are believed to be the population of molecules that have associated with groEL within the cytoplasm of E. coli. While chemists and biochemists were principally interested in the protein-folding properties of groEL, the presence of such low levels of contaminating proteins/peptides was not perceived to be a major problem, as even in their presence groEL assisted the folding and refolding of proteins. However, the presence of these contaminating proteins could contribute to any nonfolding actions reported for groEL. A number of workers have attempted to develop methods for removing these bound proteins (17, 27, 28). In these methods, use has been made of the technique of dye-ligand chromatography, which was introduced in the mid-1970s (29) and utilizes the capacity of commercially available dyes to selectively bind proteins. This technique initially appeared to be selective for nucleotide-binding proteins. However, in recent years, a wider range of proteins have been found to bind to the dyes used. We have modified a methodology described by Clark et al. (17) by using a Reactive Red dye-binding column in the presence of ATP-containing buffers. Application of recombinant groEL (groELST), which contained the expected tryptophan-containing contaminants, to a dye-binding column containing Reactive Red in the presence of ATP, resulted in the sequestration of these contaminating proteins and peptides on the dye matrix and allowed the elution of a homogeneous preparation of groELRR as determined by amino acid analysis. Analysis of groELST and groELRR for LPS revealed that such preparations contained very low levels of contaminating LPS.
Using this new method of preparing groEL, which is: 1) free of contaminating proteins; 2) contains biologically insignificant amounts of LPS, and 3) retains full protein-folding activity, we have tested the capacity of this protein to stimulate the synthesis of the proinflammatory cytokines IL-1β and IL-6 by human myelomonocytic cells. Human peripheral blood monocytes responded to groELST at concentrations as low as 10 ng/ml and maximal responses occurred at between 1 and 10 μg/ml. RT-PCR has been used to follow intracellular levels of mRNA for both cytokines. Addition of groEL to cells caused an elevation in intracellular mRNA for IL-6 although it cannot be determined if the effect was due to elevation of gene transcription or stabilization of transcripts. As expected, unstimulated cells contained IL-1β mRNA. Addition of equivalent concentrations of homogeneous groELRR to human monocytes produced similar maximal responses (at 1 and 10 μg/ml) in terms of release of cytokines and elevations in IL-6 mRNA, but at lower concentrations the protein appeared less potent. This suggested that the contaminating proteins could be contributing to the activity of groELST. However, the isolated contaminating peptides had no ability to stimulate cytokine synthesis and, when added at various concentrations to a suboptimal dose of groELRR, failed to provide any evidence of synergistic interactions with groEL. It is established that groEL is a protein containing highly flexible segments and undergoes significant conformational changes when it binds to proteins (23). Such changes in conformation have been shown by changes in the pattern of peptide products produced when groEL, under various conditions, including association with proteins, is exposed to trypsin (30), and trypsin has been used as a probe for conformational changes in groEL (31). It is therefore possible that removal of all the peptides binding to groEL induces major conformational changes in the protein structure such that the normally flexible and accessible peptide regions responsible for cytokine induction are in some way masked and are only able to display biologic activity at relatively high concentrations.
While groELST and groELRR stimulated human peripheral blood monocytes they failed to stimulate two myelomonocytic cells lines (U937 and Mono-Mac-6) or human blood neutrophils to secrete IL-1β or IL-6. This supports the earlier work of Friedland and coworkers who reported that M. tuberculosis cpn 60 (hsp 65) failed to activate the human myeloid cell line THP-1 (6).
The content of LPS in the groEL preparations was very low, but to ensure that the cytokine-inducing activity was not due to these small amounts of contaminating material, use was made of polymyxin B, which binds and inactivates LPS and the anti-CD14 mAb, MY4, which inhibits LPS activation of cells. While both polymyxin B and anti-CD14 inhibited the activity of E. coli LPS, they failed to inhibit the activity of groEL. MY4 was tested in six separate experiments using monocytes prepared from different individuals. In three of these experiment this Ab appeared to enhance the cytokine-inducing activity of the groEL. The possibility of some interaction between CD14 and whatever the cellular receptor is for groEL needs to be considered, and the role of such interactions has been addressed in recent reviews (32, 33, 34).
We have previously shown that groEL is a potent inducer of murine bone resorption in vitro (10). In these experiments, we demonstrated that heating groEL to 100°C blocked its bone-resorbing activity by 60 to 70% and that exposure to trypsin for 1 h resulted in 80 to 90% loss of bioactivity. We were therefore surprised that when we repeated these experiments and added the denatured groEL to human monocytes there was no loss in the cytokine-inducing activity of the groELST. We have now repeated the trypsinization experiments ad nauseam using a range of different concentrations of groELST or groELRR and different prepartions of trypsin, including highly purified sequencing grade material, and trypsin beads. It is possible to follow the breakdown of the groEL by SDS-PAGE and, within 2 h, under the conditions used, there is complete proteolysis to peptides with molecular masses < 10 kDa. In spite of this, there is very little loss of bioactivity. There is perhaps somewhat greater loss of bioactivity with the groELRR, and the greatest loss of bioactivity on trypsinization that we have encountered is shown in Figure 8⇑. However, even here the trypsinized material is still able to induce the production of large amounts of proinflammatory cytokine.
It is therefore concluded that groEL is a potent inducer of cytokine synthesis and that the contaminating proteins and peptides that copurify with this oligomer do not contribute to this activity. Cytokine-inducing activity is not due to contaminating LPS. Proteolysis of groEL fails to significantly inhibit its proinflammatory cytokine-inducing activity. This latter finding is unexpected, but is not inconsistent with recent findings about the biologic actions of bacterial chaperonins. We have recently reported that the M. tuberculosis cochaperone, cpn 10, is a potent inducer of bone resorption. Using a series of N- and C-terminal truncated synthetic peptides derived from the primary sequence of this protein, we have found that the bone-resorbing activity is mimicked by three peptides. Mapping these peptides onto the crystallographic model of the E. coli groES (35), we have ascertained that these peptides form part of the flexible loop and a smaller loop region around a conserved tyrosine at position 71 in the chaperonin. Interestingly, these two regions in groES are the parts of this molecule that bind to groEL. Thus, individual peptide domains of cpn 10 appear to be responsible for osteolytic activity (36). Further evidence that cpn 60-derived peptides have biologic activity has come from a recent paper in which a 14-residue peptide (VLGGGSALLRSIPA) sharing sequence similarity with groEL and a newly discovered neurotrophic factor (activity-dependent neurotorphic factor) was reported to be active at femtomolar concentrations as a neuroprotective agent (37).
The identify of the bioactive groEL peptides has not been defined but is currently under study. Recently Muhlradt et al. (38) reported that a lipopeptide from Mycoplasma fermentans was a potent inducer of macrophage nitric oxide production. This peptide, GNNDESNISFKEK, showed some similarity to one of the peptides produced by trypsin digestion of the cpn 60 of A. actinomycetemcomitans (GENEEQNVGIK). We had this peptide synthesized, but it proved inactive. We are currently attempting to isolate the active groEL peptide from the plethora of peptides produced by trypsinization.
Release of groEL from cells represents a powerful proinflammatory stimulus due to its ability to induce human monocyte IL-1β and IL-6 synthesis. At sites of bacterial infection, it is likely that there will be proteases present that could proteolytically inactivate proinflammatory proteins. However, the work presented in this paper suggests that proteolysis of groEL (at least by enzymes with trypsin-like actions) would fail to inhibit the inflammogenic actions of groEL and support the hypothesis that this major intracellular protein could have a role in the inflammation associated with E. coli and possibly other bacterial infections.
Footnotes
-
↵1 This work was supported by the Medical Research Council of Great Britain (P.T.), the Sir Jules Thorn Charitable Trust (K.R.), and Shire Pharmaceuticals (S.K.).
-
↵2 Address correspondence and reprint requests to Dr. Peter Tabona, Maxillofacial Surgery Research Unit, Eastman Dental Institute, University College London, 256 Gray’s Inn Road, London WC1X 8LD, United Kingdom.
-
↵3 Abbreviations used in this paper: Cpn 60, chaperonin 60; β-ME, β-mercaptoethanol; GuHCl, guanidine hydrochloride; mMDH, mitochondrial malate dehydrogenase; groEL, E. coli cpn 60; groELRR, groEL Reactive Red; groES, E. coli cpN 10.
- Received October 6, 1997.
- Accepted March 30, 1998.
- Copyright © 1998 by The American Association of Immunologists
References
In this issue
