Fungal metabolite gliotoxin inhibits assembly of the human respiratory burst NADPH oxidase

Abstract

Reactive oxygen species are a critical weapon in the killing of Aspergillus fumigatus by polymorphonuclear leukocytes (PMN), as demonstrated by severe aspergillosis in chronic granulomatous disease. In the present study, A. fumigatus-produced mycotoxins (fumagillin, gliotoxin [GT], and helvolic acid) are examined for their effects on the NADPH oxidase activity in human PMN. Of these mycotoxins, only GT significantly and stoichiometrically inhibits phorbol myristate acetate (PMA)-stimulated O2- generation, while the other two toxins are ineffective. The inhibition is dependent on the disulfide bridge of GT, which interferes with oxidase activation but not catalysis of the activated oxidase. Specifically, GT inhibits PMA-stimulated events: p47phox phosphorylation, its incorporation into the cytoskeleton, and the membrane translocation of p67phox, p47phox, and p40phox, which are crucial steps in the assembly of the active NADPH oxidase. Thus, damage to p47phox phosphorylation is likely a key to inhibiting NADPH oxidase activation. GT does not inhibit the membrane translocation of Rac2. The inhibition of p47phox phosphorylation is due to the defective membrane translocation of protein kinase C (PKC) betaII rather than an effect of GT on PKC betaII activity, suggesting a failure of PKC betaII to associate with the substrate, p47phox, on the membrane. These results suggest that A. fumigatus may confront PMN by inhibiting the assembly of the NADPH oxidase with its hyphal product, GT.

FIG. 1.

Chemical structures of GT and dimethyl-GT (an S-methylated metabolite of GT).

FIG. 2.

Stoichiometric GT-induced inhibition of O₂⁻ generation in PMN. (A) PMN were pretreated with the indicated concentrations of GT at either 1.0 × 10⁷ or 0.25 × 10⁷ cells/ml for 7 min at 37°C. After they were washed, their O₂⁻ generation was evaluated by SOD-inhibitable cytochrome c reduction following PMA stimulation (100 ng of PMA/10⁶ cells/ml). (B) PMN (0.05 × 10⁷ cells/ml) were pretreated in a cuvette for 5 min at 37°C with either helvolic acid or fumagillin and directly subjected to the cytochrome c assay without being washed. The data represent the means ± standard deviations of five (A) and two (B) experiments done in duplicate and are expressed as percentages of the O₂⁻ generated by control PMN pretreated with the respective solvents.

FIG. 3.

Correlation between GT-induced inhibition of O₂⁻ generation and membrane translocation of cytosolic phox components. PMN (0.25 × 10⁷ cells/ml) were pretreated with the indicated concentrations of GT for 7 min at 37°C, washed, and stimulated with 250 ng of PMA/2.5 × 10⁶ cells/ml for 7 min at 37°C, and finally, their membranes were fractionated. (A) O₂⁻ generation was initiated by adding 0.2 mM NADPH to the membrane aliquots (0.05 × 10⁷ cell equivalents) and determined as SOD-inhibitable cytochrome c reduction. The data are the means ± standard deviations of four experiments done in duplicate. (B and C) Aliquots (0.05 × 10⁷ cell equivalents) of the same membrane fractions were subjected to immunoblotting as described in Materials and Methods. Reacted HRP-conjugated secondary antibodies were developed with o-dianisidine (B) and ECL-plus (C). The data are representative of three (B) and four (C) experiments.

FIG. 4.

Postaddition effect of GT on PMA-stimulated membranes. Membranes from PMA-stimulated PMN not treated with GT (0 μg/ml in Fig. 3) were used to examine the postaddition effect of GT. Aliquots (0.05 × 10⁷ cell equivalents) were pretreated in cuvettes with the indicated concentrations of GT for 5 min at room temperature. The effect on O₂⁻ generation was then determined as described for Fig. 3A. The data show the means ± standard deviations of two experiments done in duplicate.

FIG. 5.

Effects of the GT analogue, dimethyl-GT, on O₂⁻ generation and membrane translocation of cytosolic phox components. PMN were pretreated with either GT or its analogue, dimethyl-GT, at 10 μg/0.25 × 10⁷ cells/ml for 10 min at 37°C. After being washed, a portion of the PMN (10⁶ cells) were analyzed for O₂⁻ generation (A, top). The remaining PMN were subjected to subcellular fractionation as described in Materials and Methods. The O₂⁻ generation of membrane fractions (A, bottom) and membrane translocation of cytosolic phox components (B) were then evaluated in the same way as for Fig. 3. The data in panel A show the means ± standard deviations of two experiments done in duplicate. Panel B is representative of two reproducible experiments.

FIG. 6.

Effect of GT on PMA-stimulated phosphorylation of cytosolic phox components. After a 10-min pretreatment with 10 μg of GT/0.25 × 10⁷ cells/ml at 37°C, ³²PO₄-loaded PMN were stimulated with 250 ng of PMA/2.5 × 10⁶ cells/ml for 5 min at 37°C. The PMN were then lysed and immunoprecipitated with rabbit antibodies against cytosolic phox components (see Materials and Methods). (A) They were individually immunoprobed with primary antibodies produced in a goat (p47^phox) and rabbits (p67^phox and p40^phox), followed by the corresponding HRP-conjugated secondary antibodies, and developed with o-dianisidine. (B) The same PVDF sheets were analyzed using a bio-image analyzer. The radioactivities of the respective immunoblot bands are expressed as PSL counts. The data are the means ± standard deviations of two experiments done in duplicate. (C) ³²PO₄-loaded PMN were similarly pretreated with the indicated concentrations of GT and subjected to the same procedures as in panels A and B for p47^phox detection. The PSL counts in p47^phox bands from PMN stimulated (+) with PMA and unstimulated (−) and the net values (Net) are shown.

FIG. 7.

Effect of GT on the distribution of cytosolic phox components to Sol and Skl fractions. PMN were pretreated with 10 μg of GT/0.25 × 10⁷ cells/ml for 10 min at 37°C. After being washed, the PMN were stimulated (+) with 1 μg of PMA/10⁷ cells/ml for 7 min at 37°C or left unstimulated (−) and were then subjected to Sol and Skl fractionation (see Materials and Methods). Equivalent amounts (10⁶ cell equivalents) of both fractions were analyzed by immunoblotting plus developing with o-dianisidine as described for Fig. 3. The data are representative of five reproducible experiments.

FIG. 8.

Effects of GT on PKC βII activity and its membrane translocation. (A) In the two left lanes, rp47^phox (10 pmol) was incubated for 30 min at 37°C in the absence (−) or presence (+) of rPKC βII (0.106 pmol) in a 10-μl cocktail containing 1 mM ATP (4 μCi of [γ-³²P]-ATP), as described in Materials and Methods. The reactions were stopped with Laemmli sample buffer and subjected to bio-image analysis after SDS-PAGE. For the analysis of the GT or PAO effect, the mixtures of rp47^phox and rPKC βII were pretreated for 5 min at 37°C with the indicated amounts of GT or PAO before the phosphorylation was started. (B) PMA-stimulated subcellular fractions were prepared from GT-treated PMN in the same way as for Fig. 3. Aliquots of membrane (10⁶ cell equivalents) and cytosol (10⁶ cell equivalents) fractions were subjected to immunoblotting and ECL-plus developing (for details, see Materials and Methods). The levels of membrane translocation of PKC βII decreased to 55, 46, 34, 19, and 22% of the control (0 μg of GT/ml), respectively, as the GT concentration increased. The experiments were repeated four (A) and three (B) times with similar results.