Gliotoxin is a virulence factor of Aspergillus fumigatus: gliP deletion attenuates virulence in mice immunosuppressed with hydrocortisone

Abstract

Gliotoxin is an immunosuppressive mycotoxin long suspected to be a potential virulence factor of Aspergillus fumigatus. Recent studies using mutants lacking gliotoxin production, however, suggested that the mycotoxin is not important for pathogenesis of A. fumigatus in neutropenic mice resulting from treatment with cyclophosphomide and hydrocortisone. In this study, we report on the pathobiological role of gliotoxin in two different mouse strains, 129/Sv and BALB/c, that were immunosuppressed by hydrocortisone alone to avoid neutropenia. These strains of mice were infected using the isogenic set of a wild type strain (B-5233) and its mutant strain (gliPDelta) and the the glip reconstituted strain (gliP(R)). The gliP gene encodes a nonribosomal peptide synthase that catalyzes the first step in gliotoxin biosynthesis. The gliPDelta strain was significantly less virulent than strain B-5233 or gliP(R) in both mouse models. In vitro assays with culture filtrates (CFs) of B-5233, gliPDelta, and gliP(R) strains showed the following: (i) deletion of gliP abrogated gliotoxin production, as determined by high-performance liquid chromatography analysis; (ii) unlike the CFs from strains B-5233 and gliP(R), gliPDelta CFs failed to induce proapoptotic processes in EL4 thymoma cells, as tested by Bak conformational change, mitochondrial-membrane potential disruption, superoxide production, caspase 3 activation, and phosphatidylserine translocation. Furthermore, superoxide production in human neutrophils was strongly inhibited by CFs from strain B-5233 and the gliP(R) strain, but not the gliPDelta strain. Our study confirms that gliotoxin is an important virulence determinant of A. fumigatus and that the type of immunosuppression regimen used is important to reveal the pathogenic potential of gliotoxin.

FIG. 1.

HPLC analysis of gliotoxin in the CFs from strains B-5233, gliPΔ, and gliP_R. The arrows indicate the gliotoxin peaks. The inset in the first panel shows the gliotoxin standard. A representative HPLC analysis is shown. mAU, milliabsorbance units.

FIG. 2.

Apoptosis and cell death of EL4 thymoma cells. EL4 cells were incubated for 18 h with diluted CFs from strains B-5233 (squares), gliPΔ (triangles), and gliP_R (circles). The dilutions of CFs tested were 2.5, 5, 10, and 20% (vol/vol). (A) Bak conformational changes were analyzed by FACS using a rabbit polyclonal antibody against the N-terminal sequence (NT; Upstate) and Alexa 647 goat anti-rabbit Ab. (B) Production of superoxide anions was tested with 2-hydroxiethidium staining. (C) Caspase 3 activation was monitored by FACS using a monoclonal Ab against the active enzyme. (D) The mitochondrial-membrane potential was monitored with DiOC₆[3] staining. (E) Membrane integrity was tested by phosphatidylserine translocation (AV) and membrane lysis (PI). The data are given as means ± SEM from three independent experiments.

FIG. 3.

Cell detachment of MEF. MEF were incubated for 4 h with 2.5, 5, and 10% (vol/vol) CFs. The cells incubated with RPMI 1640 (control) were flat and elongated, forming a monolayer adhering to the surface. When the MEF were incubated with 2.5% CFs from strain B-5233 or gliP_R, some cells displayed a rounded shape (arrows). Incubation with 5 and 10% CFs from strain B-5233 or gliP_R caused a majority of the cells to become round and lose anchorage. Changes in MEF morphology or detachment were not observed at any concentration of gliPΔ CFs. Bar, 100 μm.

FIG. 4.

Chemiluminescence of neutrophils incubated with CFs from strains B-5233, gliPΔ, and gliP_R. The cells were incubated in 5% CFs (vol/vol) for 30 min before PMA was added to activate the NADPH oxidase. The data shown represent the means of Sum RLU (±SEM; n = four to six donors).

FIG. 5.

Virulence studies with the mouse strains 129/Sv and BALB/c. Mice were immunosuppressed with hydrocortisone and inoculated intranasally with 5 × 10⁶ conidia. The survival rates of the recipient 129/Sv and BALB/c mice are shown in panels A and B, respectively. The Kaplan and Meier survival with log rank test was used for comparisons of survival levels among the groups: comparison between the groups infected with strains B-5233 and gliPΔ showed a P value of 0.005 for 129/Sv mice and a P value of 0.007 for BALB/c mice. The virulence assay with the 129/Sv strain was repeated three times with similar results. (C and D) Hematoxylin and eosin-stained lung sections prepared 72 h postinfection of 129/Sv mice with strain B-5233 (C) or gliPΔ (D). (E and F) GMS-stained lung sections prepared 96 h postinfection with strains B-5233 (E) and gliPΔ (F). The arrows indicate hyphae. Bar, 20 μm.