Induction of macrophage-derived SLPI by Mycobacterium tuberculosis depends on TLR2 but not MyD88

Abstract

Macrophages respond to Mycobacterium tuberculosis by regulating expression of gene products that initiate a host innate response to this micro-organism. In this study, we report that exposure of murine peritoneal macrophages to heat-killed Mycobacterium tuberculosis (HK-Mtb) led to an increase in secretory leucocyte protease inhibitor (SLPI) gene expression and protein secretion in a time- and dose-dependent manner. HK-Mtb-induced SLPI mRNA expression was sensitive neither to a protein synthesis inhibitor, cycloheximide, nor to an actin polymerization blocker, cytochalasin D. Treatment of macrophages with interferon (IFN)-gamma inhibited HK-Mtb-induced SLPI expression. RAW264.7 cells stably expressing SLPI produced a reduced level of tumour necrosis factor (TNF) in response to HK-Mtb as compared with mock transfectants. Aerosol infection of mice with live M. tuberculosis resulted in an induction of SLPI gene expression in infected lungs. Macrophages from Toll-like receptor 4 (TLR4)-/- or MyD88-/- mice responded to M. tuberculosis similarly to wild-type macrophages by exhibiting increased SLPI expression. In contrast, macrophages from TLR2-/- mice were incapable of inducing SLPI in response to M. tuberculosis. This induction signifies the presence of a TLR2-dependent but MyD88-independent M. tuberculosis signalling pathway, suggesting involvement of adaptor protein(s) other than MyD88 in M. tuberculosis-mediated induction of SLPI.

Figure 1

Secretion of secretory leucocyte protease inhibitor (SLPI) by macrophages in response to heat-killed Mycobacterium tuberculosis (HK-Mtb). (a) Dose curve. Peritoneal macrophages (10⁷) were treated with HK-Mtb [multiplicity of infection (MOI) from 1 to 8] for 24 hr. Protein samples from cultured supernatants were trichloroacetic acid (TCA)-precipitated, electrophoresed, and immunoblotted with anti-SLPI antibody. (b) Time–course. Secreted SLPIs from macrophages (10⁷) treated with HK-Mtb (MOI = 4) for the indicated times were prepared and analysed as in (a).

Figure 2

Heat-killed Mycobacterium tuberculosis (HK-Mtb)-induced secretory leucocyte protease inhibitor (SLPI) expression and the effect of cycloheximide or cytochalasin D. (a) Peritoneal macrophages (10⁷) were incubated alone or with HK-Mb [multiplicity of infection (MOI) = 4] for the indicated times. Total RNA was prepared, electrophoresed and hybridized with SLPI cDNA probe. The membrane was rehybridized with β-actin probe as a loading control. (b, c) Macrophages were pretreated with 10 µg/ml cycloheximide (CX) or 5 µg/ml of cytochalasin D (CHD) for 30 min, then incubated with 100 ng/ml lipopolysaccharide (L) or HK-Mtb (M, MOI = 4) for an additional 6 hr. Northern blots were prepared and analysed as in (a).

Figure 3

Effect of interferon (IFN)-γ on heat-killed Mycobacterium tuberculosis (HK-Mtb)-induced secretory leucocyte protease inhibitor (SLPI) expression. Peritoneal macrophages (10⁷) were incubated with 100 ng/ml lipopolysaccharide (LPS), or 10 µg/ml peptidoglycan (PGN), or 4 × 10⁷ HK-Mtb in the presence or absence of 100 U/ml of IFN-γ for 24 hr. SLPI and tumour necrosis factor (TNF) mRNA levels were analysed by northern blots, controlled and expressed as in Fig. 2.

Figure 4

Induction of secretory leucocyte protease inhibitor (SLPI) expression by Mycobacterium tuberculosis depends on Toll-like receptor 2 (TLR2) but not TLR4 or MyD88. Peritoneal macrophages from wild-type (WT), TLR2^–/–, TLR4^–/– or MyD88^–/– mice were incubated (a–c) with either heat-killed Mycobacterium tuberculosis (HK-Mtb) [M, multiplicity of infection (MOI) = 4] or lipopolysaccharide (L, 100 ng/ml) for 24 hr. (a)The expression of the secretory leucocyte protease inhibitor (SLPI) and tumour necrosis factor (TNF) mRNA was assessed by northern blot analysis as described in Fig. 2. (b) SLPI protein expression in the lysate was shown via western blot using tubulin as a loading control. (c) TNF protein in the culture supernatants was determined by enzyme-linked immunosorbent assay (ELISA). KO, knock-out mice.

Figure 5

Induction of secretory leucocyte protease inhibitor (SLPI) expression by live Mycobacterium tuberculosis depends on Toll-like receptor 2 (TLR2) but not TLR4 or MyD88. Peritoneal macrophages from wild-type (WT), TLR2^–/–, TLR4^–/– or MyD88^–/– mice were incubated with live M. tuberculosis[multiplicity of infection (MOI) = 4] for the time indicated. The expression of the SLPI mRNA was assessed by northern blot analysis as described in Fig. 2.

Figure 6

Secretory leucocyte protease inhibitor (SLPI) induction in the lungs after low-dose Mycobacterium tuberculosis aerosol infection. RNA was prepared from lungs of mice that were infected with 100 colony-forming units (CFU) of M. tuberculosis for different times up to 71 days. Real-time quantitative reverse transcriptase–polymerase chain reaction (qPCR) was performed as described in the ‘Materials and methods’ section. The results are mean ± standard deviation of triplicate samples from one of two experiments. The data were analysed using Student's t-test (*P = 0·065; **P = 0·041).

Figure 7

Secretory leucocyte protease inhibitor (SLPI) expression in macrophages leads to a suppression of tumour necrosis factor (TNF) production in response to Mycobacterium tuberculosis. RAW 264·7 cells stably transfected with SLPI (▪) or vector only (□) were treated with heat-killed Mycobacterium tuberculosis (HK-Mtb) at the indicated multiplicity of infection (MOI) for 16 hr. Conditioned media were collected for TNF determination by enzyme-linked immunosorbent assay (ELISA). Results are mean ± standard deviation of triplicates from one of four similar experiments. Student's t-test analysis shows that the differences between TNF production between two cell lines are significant at MOI 4 and 8 (P < 0·05, Student's t-test).