Mycobacterium tuberculosis heat-shock protein 70 impairs maturation of dendritic cells from bone marrow precursors, induces interleukin-10 production and inhibits T-cell proliferation in vitro

Abstract

In different inflammatory disease models, heat-shock proteins (hsp) and hsp-derived peptides have been demonstrated to possess anti-inflammatory properties. While some studies have shown that hsp can directly interact with antigen-presenting cells, others report that bacterial hsp can induce specific T cells with regulatory phenotypes. Effective characterization of the immunomodulatory effects of hsp 70, however, has historically been confounded by lipopolysaccharide (LPS) contamination. In this study, we compared the effects of LPS-free Mycobacterial tuberculosis hsp 70 (TBhsp70) and its possible contaminants on dendritic cells (DC). We demonstrate herein that LPS-free TBhsp70 inhibits murine DC maturation in vitro, while LPS-contaminated TBhsp70 induces DC maturation. Mock recombinant preparations have no effect. In contrast to LPS, TBhsp70 does not induce tumour necrosis factor-alpha production by DC, but interleukin-10. In vivo, only LPS-contaminated TBhsp70 induces up-regulation of CD86 in splenic mature DC. Finally, TBhsp70 inhibited phytohaemagglutinin-induced T-cell proliferation. Our results support the hypothesis that TBhsp70 does not have inflammatory potential, but rather has immunosuppressive properties.

Figure 1

Kinetics of LPS-free TBhsp70 inhibition of dendritic cell maturation. Dendritic cells were grown from bone marrow and incubated on the fifth day of culture with different stimuli: PBS, 10⁻⁵ m dexamethasone (DEX), 60 μg of either LPS (1 EU/µg), BSA or TBhsp70. Cells were harvested 24 hr and 48 hr later and analysed for CD86 and MHC class II expression. Three populations are typically identified at this time in culture: CD86^hi MHC class II^hi cells (percentage indicated in the upper right quadrants) are the already mature DC; the CD86^lo MHC II^lo population are the still immature DC; and the double-negative population has not yet differentiated.

Figure 2

Cytokines detected in the supernatants harvested 48 hr after addition of stimuli to cultures. (a) TNF-α;(b) IL-10. Stimuli included PBS; 10⁻⁵ m DEX (DEX), 60 μg of either LPS (1 EU/µg), BSA or TBhsp70. This experiment was repeated five times, with comparable results.

Figure 3

Effect of Triton X-114 treatment on recombinant preparations. (a) Endotoxin content of recombinant protein preparations and stock LPS solution (1 μg/ml) measured by the LAL assay, before and after Triton extraction. (b) ATPase activity of TBhsp70 before and after Triton extraction. (c, d) Cytokine content in DC culture supernatant after 48-hr incubation with different doses (10, 20 or 40 μg/ml) of either LPS, TBhsp70, EgAFFP or (microliztre equivalents of) mock preparation, before or after Triton extraction. Open bars, before Triton extraction; black bars, after Triton extraction. This experiment was repeated six times, with similar results.

Figure 4

LPS-free TBhsp70, but not its contaminants, inhibits DC maturation and does not induce TNF-α or NO. Murine DC were incubated with 10, 20 or 40 μg/ml of LPS-contaminated (open shapes) or Triton-extracted (black shapes) preparations. For mock preparations, equivalent μl amounts were used. (a) •, PBS; ▴, clean TBhsp70; ▵, dirty TBhsp70; (b) ○, LPS; ▾, clean EgAFFP; ◊, dirty EgAFFP; (c) •, PBS, ▮, clean mock prep; □, dirty mock preparation; (d), TNF-α production by the cultured DC, incubated with stimuli described above, with LPS contaminated (open bars) or Triton-extracted (black bars) preparations.

Figure 5

Clean TBhsp70 does not induce up-regulation of CD86 in DC in vivo. Mice (regular BALB/c or Rag^–/– BALB/c) were injected intravenously with either PBS, 40 μg LPS, dirty TBhsp70, or clean TBHSp70. Spleens were removed 18 hr later, treated with collagenase D and the single-cell suspension was analysed by flow cytometry. Expression of CD86 was analysed in CD11c⁺ B220^– cells. This experiment was repeated twice, with identical results, and is currently the standard assay in our laboratory for LPS contamination of recombinant proteins because of its sensitivity. A variant of this experiment, with spleen removal after 6 hr of injection, was performed three times. Dark line represents PBS control; grey line represents experimental data.

Figure 6

Clean TBhsp70 inhibits PHA-induced T-cell proliferation in vitro. In this experiment, splenocytes from four mice were incubated with 1% PHA plus either (a) DEX (10⁻⁷ to 10⁻⁵ m); or (b) clean (▴) TBhsp70; or (c) dirty (▮) TBhsp70, for 96 hr. Proliferation/viability is expressed as percentage of PHA-induced proliferation. **P < 0·01; ***P < 0·001. This experiment was performed five times, with comparable results.