Vol. 62, No. 12

INFECMION AND IMMUNITY, Dec. 1994, p. 5689-5693

0019-9567/94/$04.00+0 Copyright © 1994, American Society for Microbiology

Bacterial Heat Shock Proteins Directly Induce Cytokine mRNA and Interleukin-1 Secretion in Macrophage Cultures CARSTEN RETZLAFF,'t YOSHIMASA

YAMAMOTO,`*

PAUL S. HOFFMAN,2 HERMAN FRIEDMAN,'

AND THOMAS W. KLEIN'

Department of Medical Microbiology and Immunology, University of South Florida College of Medicine, Tampa, Florida 33612,1 and Department ofMicrobiology, Dalhousie University, Halifax, Nova Scotia, Canada2 Received 8 August 1994/Returned for modification 9 September 1994/Accepted 26 September 1994

Bacterial heat shock proteins (hsp) have been shown to be important immunogens stimulating both T cells and B cells. However, little is known concerning the direct interactions between hsp and macrophages. In this study, we demonstrated that treatment of macrophage cultures with purified bacterial hsp, including Legionelia pneumophila hsp6O, Escherichia coli GroEL, Mycobacterium tuberculosis hsp7O, Mycobacterium leprae hsp65, and Mycobacterium bovis BCG hsp65, increased the steady-state levels of cytokine mRNA for interleukin-la (IL-la), IL-10, IL-6, tumor necrosis factor alpha, and granulocyte-macrophage colony-stimulating factor as well as supernatant IL-1 secretion. This effect was shown not to be due to contamination of the hsp preparations with bacterial lipopolysaccharide. However, not all hsp induced cytokines; M. tuberculosis hsplO showed minimal activity in our study. These results suggest that bacterial hsp might modulate immunity by rapidly and directly increasing cytokine production in macrophages. Heat shock proteins (hsp), known as chaperones or stress proteins, are highly conserved proteins with important biological functions in protein biogenesis. Most of these proteins are constitutively expressed at a low level under physiological conditions in all eukaryotic and prokaryotic cells examined to date (4). A variety of stressful stimuli, including heat shock, nutrient deprivation, oxygen radicals, and viral infection, induce a marked increase in intracellular hsp synthesis (9, 21). Although these proteins are predominantly located in intracellular compartments, recent studies suggest that some stress protein determinants can be expressed on the surfaces of mammalian and bacterial cells and can be secreted extracellularly (6, 7, 15, 16). The immunology of hsp has been studied extensively. For example, it has been demonstrated that hsp can participate in immune responses to bacterial infections and in development of autoimmune diseases (3). Members of the hsp60 class serve as immunodominant targets of aI3+ and T+ T lymphocytes as well as stimulating antibodies during infections by many bacterial pathogens, including Mycobacterium, Legionella, Borrelia, and Treponema organisms as common antigens (11, 12, 14, 15, 20, 25, 26). In particular, hsp, including members of the hsp60 family, have been used in attempts to induce immunological protection against infection by mycobacteria or legionellae (1, 18, 28), which can survive and replicate in macrophages. Macrophages generally respond to microbial infections with increased expression and secretion of cytokines such as interleukin-1 (IL-1), IL-6, and tumor necrosis factor alpha (TNF-a) (19, 23, 27, 29, 31). However, the role of hsp in macrophagepathogen interactions is poorly understood, although recent studies have suggested that mycobacterial hsp65 may induce immunostimulatory functions by increasing the production of monokines (10, 33). The present study was designed to exam* Corresponding author. Mailing address: Department of Medical Microbiology and Immunology, University of South Florida College of Medicine, MDC Box 10, 12901 Bruce B. Downs Blvd., Tampa, FL 33612-4799. Phone: (813) 974-3281. Fax: (813) 974-4151. t Present address: Department of Medical Microbiology, University of Leipzig, Leipzig, Germany.

5689

ine whether hsp from different bacteria and different hsp families directly induce cytokine expression and secretion in macrophage cultures. For this purpose, we investigated the cellular steady-state levels of IL-lo, IL-113, IL-6, TNF-ot, and granulocyte-macrophage colony-stimulating factor (GM-CSF) mRNA by a quantitative reverse transcriptase (RT)-PCR and also investigated secretion of soluble IL-1. For these experiments, thioglycolate-elicited BALB/c mouse macrophages were stimulated with Legionella pneumophila hsp60 (Lphsp6O), Escherichia coli GroEL, Mycobacterium tuberculosis hsp70 (Mt-hsp7o) or Mt-hspl0, Mycobacterium leprae hsp65 (Ml-hsp65), or Mycobacterium bovis BCG hsp65 (Mb-hsp65). The experiments show that most but not all of the hsp tested are relatively potent inducers of acute-phase cytokines. Lp-hsp6O was cloned and expressed in E. coli and purified as described previously (15). GroEL, the 60-kDa chaperone of E. coli, was purchased from Sigma Chemical Co., St. Louis, Mo. The mycobacterial hsp were kindly provided by J. D. A. van Embden, National Institute of Public Health and Environmental Protection, Bilthoven, The Netherlands, with support from the UNDPIWorld Bank/WHO Special Program for Research and Training in Tropical Diseases. Thioglycolate-elicited peritoneal macrophages, from 8- to 12-week-old female BALB/c mice (Jackson Laboratories, Bar Harbor, Maine), were obtained as described previously (30), suspended in RPMI 1640 medium supplemented with 10% heat-inactivated fetal calf serum (HyClone Laboratory, Logan, Utah), and allowed to adhere to six-well tissue culture plates (Costar, Cambridge, Mass.). After a washing with Hanks' balanced salt solution, the resulting macrophage monolayers (approximately 3 x 106 cells per well; .99% macrophages as determined by morphology according to Giemsa stain) were treated for 3 h either with an hsp diluted in RPMI 1640 medium containing 16 ,ug of polymyxin B per ml or with 0.5 ,ug of E. coli lipopolysaccharide (LPS) (O111:B4; Sigma) diluted in RPMI 1640 medium. The total RNA was isolated from the macrophage cultures by a single-step method (2) with Tri Reagent (Molecular Research Center, Cincinnati, Ohio) and spectrophotometrically quantified. Reverse transcription of total RNA (1 jug) was done with avian myeloblastosis virus RT (Promega, Madison,

_ oo

o o T.t'

°;. =l _L'

= %AIQ

u _4-

.

II

-

-

:

Wis.) and oligo(dT)15 primers. PCR of the prepared cDNAs (3 ,ul) was performed with primers specific for IL-la, IL-1p, IL-6, TNF-a, and GM-CSF (Stratagene, La Jolla, Calif.) as described previously (31). Primers specific for ,B-actin (Stratagene) and 2 microglobulin (BMG) (5, 32) were used as internal standards. The PCR products were visualized under UV light after electrophoresis in a 2% agarose gel containing ethidium bromide. The identities of the products were validated by their predicted sizes in the gels. Figure 1 shows a representative gel of the cytokine RT-PCR products of RNA extracted from hsp-treated macrophage cultures. Stimulation with Lp-hsp6O, GroEL, Mt-hsp7O, Ml-hsp65, and Mb-hsp65 increased the cellular mRNA levels of IL-la, IL-1j, IL-6, TNF-a, and GM-CSF; however, stimulation with Mt-hsplO induced either no (IL-6 and GM-CSF) or very little (IL-lot, IL-1p, and TNF-ao) mRNA.

tn

so

~,

2

=

;

f-actin

IL-1t

In order to analyze further the quantitative aspects of this effect, the levels of IL-13 mRNA were examined in greater detail by the differential PCR technique (32). This technique employs the coamplification of messages for IL-13 and BMG (as an endogenous standard) in a single PCR tube (24). Following amplification, the reaction products were separated by high-pressure liquid chromatography (HPLC) (410 BIO; Perkin-Elmer, Norwalk, Conn.) on a DEAE-NPR anionexchange column (TosoHass, Montgomeryville, Pa.) (32). The cytokine DNA peak areas were normalized for sample-tosample comparison by calculating a ratio of the cytokine DNA versus the respective BMG DNA (internal control) peak area. As shown in Fig. 2, treatment of macrophage cultures for 3 h with 0.5 jig of LPS per ml caused a large increase in the IL-1 PCR product peak area relative to the peak area for BMG. Of interest, treatment with 0.5 jig of either Gro-EL, Mt-hsp70, or Ml-hsp65 per ml increased the peak area to the same extent as LPS treatment; however, treatment with Lphsp6O, Mt-hsplO, and Mb-hsp65 induced much less of an

IL-1l

IL-6

TNFre

GMI-CSF FIG. 1. Hsp increase steady-state levels of cytokine mRNA in macrophages. Cultures were treated for 3 h with either tissue culture medium (control), E. coli LPS (0.5 jig/ml), or bacterial hsp (5 jug/ml) and then subjected to RNA extraction and RT-PCR amplification. Shown is a representative photograph of an ethidium bromide gel of the PCR products.

increase. Treating the cultures with a higher concentration (i.e., 5 jig/ml) of the various hsp increased the IL-1B message even further, especially in the case of Lp-hsp60 and Mb-hsp65, but the response to Mt-hsplO was consistently low. These

Control I F-

LPS

Lp-hsp60 GroEL 0 Mt-hspl 0 -J

-

z

CD

-I

Mt-hsp70 MI-hsp65 Mb-hsp65

WM

I1 0

100

200

300

400

500

IL-113/BMG-RATIO IN % FIG. 2. hsp60 and -70 families increase IL-1l8 mRNA, while hsplO has a minimal effect. Macrophage cultures were treated for 3 h with either tissue culture medium only (control), E. coli LPS (0.5 jig/ml), or bacterial hsp at a concentration of either 0.5 ,ug/ml (solid bars) or 5 jig/ml (hatched bars). PCR products were analyzed by HPLC, and the levels of IL-1 product were expressed as the IL-1,B/BMG ratio [(IL-11 peak area/BMG peak area) x 100]. Each bar represents the mean for three independent experiments; error bars indicate standard deviations. *, value is not significantly (P > 0.05) different from that of the unstimulated control group by the two-tailed Student t test. 5690

NOTES

VOL. 62, 1994 Q

u

U

u

a

Lt

-

-

o vn X 0 a. ov:. o

oN

m;04040

B-actin

IL-113 FIG. 3. Heating Lp-hsp60 to 95°C decreases cytokine-inducing activity. Portions of E. coli LPS (0.5 ,ug/ml) or Lp-hsp60 (5 ,ug/ml) were preheated for 1 h at 95°C and then added to macrophage cultures for 3 h. Other cultures were also treated with either tissue culture medium only (control) or unheated LPS and hsp. The RNA was extracted and amplified. Shown is a representative photograph of an ethidium bromide gel of the PCR products.

results confirm and extend those obtained by others (10, 33) using THP-1 cells stimulated with mycobacterial hsp65. Our results show that primary macrophage populations (i.e., peritoneal macrophages), in addition to the THP-1 cell line, can be stimulated directly by purified hsp. Furthermore, we show that hsp from the 60- and 70-kDa-protein families and from bacteria other than mycobacteria are potent cytokine inducers;

CONTROL LPS ILp-hsp6O z GroEL 0 Mt-hspl 0 in* -j Mt-hsp70 C,, Ml-hsp65 Mb-hsp65 0

5691

however, the potency is less in proteins from the hsplO family. It might be argued that LPS contamination of the hsp preparations was a significant factor in the induction of cytokine message. However, this is doubtful for several reasons. First of all, the LPS contamination was 1l ng/10 ,ug of hsp per ml as measured by the Limulus amebocyte lysate assay QCL-1000 (Bio Whittaker, Walkersville, Md.). This low LPS concentration did not induce detectable increases in cytokine mRNA expression or enhance the expression when added to hsp preparations (data not shown). A second mitigating condition involved the use of polymyxin B (16 ,ug/ml) in all experiments, which completely neutralized 20 ng of LPS per ml in terms of IL-lp mRNA-inducing activity. Third, inactivation of the hsp by heating (1 h, 95°C) led to a marked reduction of IL-1B message, as shown in Fig. 3 (Lp-hsp60 as a representative example), whereas LPS was fully active under the same conditions. Finally, treatment with the anti-hsp60 antibody neutralized the cytokine-inducing activity of hsp (Fig. 4). To correlate the above-mentioned increases in message level with the induction of cytokine secretion, we next examined IL-1 bioactivity (17) in the culture supernatants of macrophages stimulated with the various hsp (10 ,ug/ml). A significant increase in IL-1 bioactivity was found in supernatants of all hsp-stimulated cultures except those exposed to Mt-hsplO (Fig. 4). GroEL and Mt-hsp7O appeared to induce higher levels of IL-1 bioactivity than did Lp-hsp6O and the other mycobacterial hsp. This induction of supernatant bioactivity could be reduced by pretreatment with antibody to Lp-hsp60 (Fig. 4), showing the high degree of immune cross-reactivity among the members of the hsp60 family (15) as well as supporting our above-mentioned conclusion that the hsp component in these preparations is at least partly responsible for the induction of cytokines.

M--

,~~~~~~~~~~~~~~~~~~~~~~~~~~ Pretreatment

with antibody: - - No

-_ -Yes

5

10

15

20

25

IL-1 ACTIVITY IN CPM (Thousands) FIG. 4. Hsp (with the exception of hsplO) increase supernatant IL-1 bioactivity and are neutralized by antibody to Lp-hsp6O. Cultures were treated for 24 h with either tissue culture medium only (control), E. coli LPS (5 pLg/ml), hsp (10 ,ug/ml), or hsp pretreated (1 h at 4°C) with rabbit anti-Lp-hsp60 serum (1:100 dilution). Each bar represents the mean for at least three independent experiments; error bars indicate standard deviations. The IL-1 content in all stimulated macrophages, except Mt-hsplO, (*), is significantly increased (P < 0.05) compared with that in the unstimulated control group. Preincubation with antiserum resulted in a significant decrease of IL-1 in all groups (P < 0.05) as determined by the two-tailed Student t test.

5692

NOTES

Our findings show that bacterial hsp can directly induce

cytokine mRNA as well as cytokine secretion by macrophages. The mechanisms involved are not clear, but the binding of hsp to cell surface receptors is a possibility. hsp are known to have a high affinity for immature or partially denatured proteins, binding them for the process of protein folding and unfolding (13). Similar protein-protein interactions, therefore, with macrophage surface receptors might fulfill ligand-receptor requirements activating the cell. In such a case, the specificity of binding would be expected to be broad because the amino acid homology between members of the hspl0, -60, and -70 families is low, as shown by the Lipman-Pearson protein alignment (Lasergene; DNASTAR Inc., Madison, Wis.), and consequently a variety of receptors might be activated. The reason for the relatively lower potency of the hsplO preparation in comparison with other hsp preparations is also not clear. GroES, the hsplO cognate in E. coli, has been shown to have unique protein-protein interaction characteristics which differ from those of the other hsp, implying that they might have lower affinities (8). On the other hand, a trivial explanation might be related to the relatively small size of these proteins, which prohibits them from effectively cross-linking and capping receptors following binding. The results of this study suggest a direct role of bacterial hsp in the regulation of acute-phase cytokine release from macrophages. To our knowledge this is the first report showing that such stimulatory activity cuts across boundaries of hsp species and family derivation. We thank Catherine A. Newton for performing the Limulus assays. This work was supported by grant AI 16618 from the National Institute of Allergy and Infectious Diseases.

1.

2. 3. 4.

5.

6. 7. 8. 9.

10.

REFERENCES Blander, S. J., and M. A. Horwitz. 1993. Major cytoplasmic membrane protein of Legionella pneumophila, a genus common antigen and member of the hsp 60 family of heat shock proteins, induces protective immunity in a guinea pig model of Legionnaires' disease. J. Clin. Invest. 91:717-723. Chomczynski, P., and N. Sacchi. 1987. Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal. Biochem. 162:156-159. Cohen, I. R, and D. B. Young. 1991. Autoimmunity, microbial immunity and the immunological homunculus. Immunol. Today 12:105-110. Craig, E. A., B. D. Gambill, and R. J. Nelson. 1993. Heat shock proteins: molecular chaperones of protein biogenesis. Microbiol. Rev. 57:402-414. Ehlers, S., M. E. A. Mielke, T. Blankenstein, and H. Hahn. 1992. Kinetic analysis of cytokine gene expression in the livers of naive and immune mice infected with Listeria monocytogenes. The immediate early phase in innate resistance and acquired immunity. J. Immunol. 149:3016-3022. Ensgraber, M., and M. Loos. 1992. A 66-kilodalton heat shock protein of Salmonella typhimurium is responsible for binding of the bacterium to intestinal mucus. Infect. Immun. 60:3072-3078. Evans, D. J., Jr., D. G. Evans, L. Engstrand, and D. Y. Graham. 1992. Urease-associated heat shock protein of Helicobacter pylon. Infect. Immun. 60:2125-2127. Ezzell, C. 1994. How chaperonins monitor their protein charges. J. NIH Res. 6:31-34. Fayet, O., T. Ziegelhoffer, and C. Georgopoulos. 1989. The groES and groEL heat shock gene products of Escherichia coli are essential for bacterial growth at all temperatures. J. Bacteriol. 171:1379-1385. 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.

INFECF. IMMUN. 11. Hansen, K., J. M. Bangsborg, H. Fjordvang, N. S. Pedersen, and P. Hinderssen. 1988. Immunochemical characterization of and isolation of the gene for a Borrelia burgdorfen immunodominant 60-kilodalton antigen common to a wide range of bacteria. Infect. Immun. 56:2047-2053. 12. Haregewoin, A., G. Soman, R C. Hom, and R W. Finberg. 1989. Human rb+ T cells respond to mycobacterial heat-shock protein. Nature (London) 340:309-312. 13. Hemmingsen, S. M., C. Woolford, S. M. van der Vies, K. Tilly, D. T. Dennis, C. P. Georgopoulos, R. W. Hendrix, and R J. Ellis.

14.

15.

16.

17.

18.

19. 20.

21.

1988. Homologous plant and bacterial proteins chaperone oligomeric protein assembly. Nature (London) 333:330-334. Hindersson, P., J. D. Knudson, and N. H. Axelsen. 1987. Cloning and expression of Treponema pallidum common antigen (Tp-4) in E. coli K-12. J. Gen. Microbiol. 133:587-596. Hoffman, P. S., L. Houston, and C. A. Butler. 1990. Legionella pneumophila htpAB heat shock operon: nucleotide sequence and expression of the 60-kilodalton antigen in L. pneumophila-infected HeLa cells. Infect. Immun. 58:3380-3387. Jarjour, W., V. Tsai, V. Woods, W. Welch, W. Pierce, M. Shaw, H. Mehta, W. Dillmann, N. Zvaifler, and J. Winfield. 1989. Cell surface expression of heat shock proteins. Arthritis Rheum. 32:44. Katz, E. D., and M. W. Dong. 1990. Rapid analysis and purification of polymerase chain reaction products by high-performance liquid chromatography. BioTechniques 8:546-555. Kaufmann, S. H. E., B. Schoel, J. D. A. van Embden, T. Koga, A. Wand-Wurttenberger, M. E. Munk, and U. Steinhotf. 1991. Heat shock protein 60: implications for pathogenesis of and protection against bacterial infections. Immunol. Rev. 121:67-90. Klein, T. W., Y. Yamamoto, S. Wilson, C. Newton, and H. Friedman. 1992. Legionella pneumophila infection and cytokine production. Adv. Exp. Med. Biol. 319:97-104. Koga, T., A. Wand-Wurttenberger, J. DeBruyn, M. E. Munk, B. Schoel, and S. H. E. Kauffmann. 1989. T cells against bacterial heat shock protein recognize stressed macrophages. Science 245: 1112-1115. Lindquist, S., and E. A. Craig. 1988. The heat-shock proteins. Annu. Rev. Genet. 22:631-637.

22. Mehra, V., B. R. Bloom, A. C. Bajardi, C. L. Grisso, P. A. Sieling, D. Alland, J. Convit, X. Fan, S. W. Hunter, P. J. Brennan, T. H. Rea, and R L. Modlin. 1992. A major T cell antigen of Mycobacterium leprae is a 10-kD heat-shock cognate protein. J. Exp. Med. 175:275-284. 23. Nathan, C. F. 1987. Secretory products of macrophages. J. Clin. Invest. 79:319-326. 24. Noonan, K. E., and I. B. Roninson. 1988. mRNA phenotyping by enzymatic amplification of randomly primed cDNA. Nucleic Acids Res. 16:10366. 25. Ottenhoff, T. H. M., B. Kale, J. D. A. van Embden, J. E. R. Thole, and R. Kiessling. 1988. The recombinant 65-kilodalton heat shock protein of Mycobacterium bovis bacillus Calmette-Guerin/M. tuberculosis is a target molecule for CD4+ cytotoxic T lymphocytes that lyse human monocytes. J. Exp. Med. 168:1947-1952. 26. Plikaytis, B. B., G. M. Carlone, C.-P. Pau, and H. W. Wilkinson. 1987. Purified 60-kilodalton Legionella protein antigen with Legionella-specific and nonspecific epitopes. J. Clin. Microbiol. 25:20802084. 27. Retzlaff, C., Y. Yamamoto, T. W. Klein, and H. Friedman. Unpublished data. 28. Silva, C. L., and D. B. Lowrie. 1994. A single mycobacterial protein (hsp65) expressed by a transgenic antigen-presenting cell vaccinates mice against tuberculosis. Immunology 82:244-248. 29. Widen, R., T. W. Klein, C. A. Newton, and H. Friedman. 1989. Induction of interleukin 1 by Legionella pneumophila in murine peritoneal macrophage cultures. Proc. Soc. Exp. Biol. Med. 191:

304-308.

30. Yamamoto, Y., T. W. Klein, C. A. Newton, R Widen, and H. Friedman. 1988. Growth of Legionella pneumophila in thioglycolate-elicited peritoneal macrophages from A/J mice. Infect. Immun.

56:370-375.

31. Yamamoto, Y., S. Okubo, T. W. Klein, K. Onozaki, T. Saito, and H.

VOL. 62, 1994 Friedman. 1994. Binding of Legionella pneumophila to macrophages increases cellular cytokine mRNA. Infect. Immun. 62: 3947-3956.

32. Yamamoto, Y., C. Retzlaff, P. He, T. W. Klein, and H. Friedman. Quantitative reverse transcriptase-PCR analysis of Legionella pneumophila-induced cytokine mRNA in different macrophage

NOTES

5693

populations by high-pressure liquid chromatography. Clin. Diagn. Lab. Immunol., in press. 33. Zhang, Y., M. Doerfler, T. C. Lee, B. Guillemin, and W. Rom. 1993. Mechanisms of stimulation of interleukin-1 and tumor necrosis factor-a by Mycobacterium tuberculosis components. J. Clin. Invest. 91:2076-2083.