Richard D. Diamond

Catalases of Aspergillus fumigatus

ABSTRACT Upon infection of a host, the pathogenic fungus Aspergillus fumigatus is attacked by the reactive oxygen species produced by phagocytic cells. Detoxification of hydrogen peroxide by catalases was proposed as a way to overcome this host response. A. fumigatus produces three active catalases; one is produced by conidia, and two are produced by mycelia. The mycelial catalase Cat1p was studied previously. Here we characterized the two other catalases, their genes, and the phenotypes of gene-disrupted mutants. CatAp, a spore-specific monofunctional catalase, is resistant to heat, metal ions, and detergent. This enzyme is a dimeric protein with 84.5-kDa subunits. The 749-amino-acid polypeptide exhibits high levels of similarity to the Aspergillus nidulans CatA catalase and to bacterial catalase HPII of Escherichia coli. In spite of increased sensitivity to H2O2, killing of ΔcatA conidia by alveolar macrophages and virulence in animals were similar to the killing of conidia by alveolar macrophages and virulence in animals observed for the wild type. In contrast to the Cat1p and CatAp catalases, the mycelial Cat2p enzyme is a bifunctional catalase-peroxidase and is sensitive to heat, metal ions, and detergent. This enzyme, an 82-kDa monomer, is homologous to catalase-peroxidases of several fungi and bacteria. Surprisingly, mycelium of the double Δcat1Δcat2 mutant with no catalase activity exhibited only slightly increased sensitivity to H2O2 and was as sensitive to killing by polymorphonuclear neutrophils as mycelium of the wild-type strain. However, this mutant exhibited delayed infection in the rat model of aspergillosis compared to infection by the wild-type strain. These results indicate that conidial catalase is not a virulence factor and that mycelial catalases transiently protect the fungus from the host.

Damage to Candida albicans Hyphae and Pseudohyphae by the Myeloperoxidase System and Oxidative Products of Neutrophil Metabolism In Vitro

In previous studies, we noted that Candida hyphae and pseudohyphae could be damaged and probably killed by neutrophils, primarily by oxygen-dependent nonphagocytic mechanisms. In extending these studies, amount of damage to hyphae again was measured by inhibition of [(14)C]cytosine uptake. Neutrophils from only one of four patients with chronic granulomatous disease damaged hyphae at all, and neutrophils from this single patient damaged hyphae far less efficiently than simultaneously tested neutrophils from normal control subjects. Neutrophils from neither of two subjects with hereditary myeloperoxidase deficiency damaged the hyphae. This confirmed the importance of oxidative mechanisms in general and myeloperoxidase-mediated systems in particular in damaging Candida hyphae. Several potentially fungicidal oxidative intermediates are produced by metabolic pathways of normal neutrophils, but their relative toxicity for Candida hyphae was previously unknown. To help determine this, cell-free in vitro systems were used to generate these potentially microbicidal products. Myeloperoxidase with hydrogen peroxide, iodide, and chloride resulted in 91.2% damage to hyphal inocula in 11 experiments. There was less damage when either chloride or iodide was omitted, and no damage when myeloperoxidase was omitted or inactivated by heating. Azide, cyanide, and catalase (but not heated catalase) inhibited the damage. Systems for generation of hydrogen peroxide could replace reagent hydrogen peroxide in the myeloperoxidase system. These included glucose oxidase, in the presence of glucose, and xanthine oxidase, in the presence of either hypoxanthine or acetaldehyde. In the presence of myeloperoxidase and a halide, the toxicity of the xanthine oxidase system was not inhibited by superoxide dismutase and, under some conditions, was marginally increased by this enzyme. This suggested that superoxide radical did not damage hyphae directly but served primarily as an intermediate in the production of hydrogen peroxide. The possible damage to hyphae by singlet oxygen was examined using photoactivation of rose bengal. This dye damaged hyphae in the presence of light and oxygen. The effect was almost completely inhibited by putative quenchers of singlet oxygen: histidine, tryptophan, and 1,4-diazobicyclo[2.2.2]octane. These agents also inhibited damage to hyphae by myeloperoxidase, halide, and either hydrogen peroxide or a peroxide source (xanthine oxidase plus acetaldehyde). Myeloperoxidase-mediated damage to hyphae was also inhibited by dimethyl sulfoxide, an antioxidant and scavenger of the hydroxyl radical. These data support the involvement of oxidative mechanisms and the myeloperoxidase-H(2)O(2)-halide system, in particular in damaging hyphae in vitro and perhaps in vivo as well.

Disparate effects of interferon-gamma and tumor necrosis factor-alpha on early neutrophil respiratory burst and fungicidal responses to Candida albicans hyphae in vitro.

We examined effects of priming with recombinant human interferon-gamma (IFN) or tumor necrosis factor-alpha (TNF) on neutrophil responses to Candida albicans hyphae. Both cytokines increased early superoxide generation after hyphal stimulation. The more pronounced effects of TNF were accompanied by an augmented surface membrane depolarization rate and were insensitive to both pertussis toxin and calcium ion chelation, but were negated by concomitant incubation with puromycin or cycloheximide during priming. IFN augmented hyphal killing despite its only minor enhancement of early respiratory burst responses, but TNF reduced neutrophil fungicidal activity to nearly 40% below those by unprimed control cells even though it enhanced early superoxide responses more dramatically. Though TNF-primed neutrophils killed hyphae at normal initial rates, IFN-primed or even unprimed cells manifested more fungicidal sustained activity. These disparate consequences of cytokine priming on hyphal destruction were paralleled by differences in late generation of potentially candidacidal oxidants, hydrogen peroxide, and hypochlorous acid. IFN added during priming failed to correct TNF-associated functional defects in neutrophil anti-Candida responses. Thus, augmentation of early respiratory burst responses to oxidant-sensitive organisms need not necessarily reflect concomitant salutary effects on microbicidal activity.

A subcutaneous reservoir for intrathecal therapy of fungal meningitis.

Abstract Subcutaneous cerebrospinal-fluid reservoirs were inserted into 15 patients with cryptococcal meningitis, four with coccidioidal meningitis, and two with chronic meningitis of undetermined cause. Complications associated with 17 reservoirs in 12 patients necessitated reservoir removal or prevented later use for intrathecal therapy. Nine of these complications were related to reservoir insertion, and eight to problems related to later usage. Insertion may have been a contributing cause of death in three patients and of neurologic deterioration in another four. In nine patients there were 10 cerebrospinal-fluid bacterial infections apparently related to reservoir usage or insertion. Reservoirs were cleared of bacterial infection and continued to be functional in four of the five patients who completed a course of treatment with antibiotics alone. Successful anti-fungal therapy was clearly attributable to intraventricular therapy in three patients with cryptococcosis and three with coccidioidomycosis...

The Role of Late Complement Components and the Alternate Complement Pathway in Experimental Cryptococcosis

Summary Duration of survival was comparable in normal and C4D guinea pigs infected iv with cryptococci. Survival was shortened in animals depleted of late complement components by treatment with CVF. Some animals received a single treatment with CVF to allow recovery of complement components after challenge with cryptococci. This resulted in lower counts of fungi in peripheral blood, lung, and liver, but no improvement in levels of fungi in brain, or in survival when compared with animals continuously depleted of C3–9. Late complement components were therefore important in clearance of cryptococci from extraneural sites. However, once infection was established in brain, complement levels in the central nervous system may have been too low to aid in destruction or organisms. The authors wish to thank Miss Margret Huber for expert technical assistance and Dr. David Alling for help with statistical evaluation.

Successful treatment of Candida osteomyelitis with fluconazole. A noncomparative study of two patients.

We describe two patients with osteomyelitis due to Candida spp. treated with fluconazole, a new triazole antifungal. One patient had extensive involvement of ribs and costochondral regions of the anterior chest, and the other had vertebral infection. Both were cured with courses of 10 and 14 months, with greater than or equal to 1 year of follow-up after fluconazole was discontinued. Fluconazole is an attractive agent for the treatment of Candida osteomyelitis and deserves to be studied more extensively for this indication.

A rapid fluorescent assay to distinguish attached from phagocytized yeast particles.

In studies of phagocytosis and its consequences for cell activation, it is important to distinguish those particulate stimuli which are completely ingested and internalized from those which are only attached to phagocytic surfaces. Ingestion can be profoundly influenced by both the type of opsonins on the surface of the stimulus and the expression and activation of receptors on the phagocytes for these opsonins. We report a new fluorescent assay which facilitates rapid and reproducible quantitation of attached versus fully internalized live or dead yeast particles by phagocytes. The assay employs the fluorescent dye diaethanol (Uvitex 2B) which stains chitin on the cell wall of fungi and is excluded from live phagocytes. The diaethanol assay and a standard, previously published methylene blue dye exclusion assay yielded comparable results using human neutrophils or monocytes incubated with heat-killed, serum-opsonized Candida albicans. The diaethanol assay proved useful in distinguishing differences in effects of various opsonins, as human neutrophils selectively opsonized with either pooled human serum (PHS), IgG (heat-inactivated PHS) or complement (IgG-depleted PHS) completely internalized 69.5%, 91.3% and 52.5% of cell-associated zymosan particles respectively. Finally, the new assay permitted comparisons of differing macrophage populations, as resident murine peritoneal macrophages internalized only 10.6% of complement-opsonized, cell-associated zymosan particles compared with 41.7% when the macrophages were elicited with thioglycolate. The assay should prove useful to investigators studying both fungal phagocytosis and killing, as well as to those performing general studies of receptor expression, regulation and function.

Monocyte-mediated serum-independent damage to hyphal and pseudohyphal forms of Candida albicans in vitro.

Human peripheral blood monocytes attached to Candida albicans hyphae in the absence of serum and damaged the hyphae without completely ingesting them. Attachment and damage was not augmented by the addition of serum. Damage to hyphae was quantitated by a previously developed metabolic assay that measured leukocyte-induced reduction in uptake of [(14)C]cytosine by the hyphae. Use of cells from patients with hereditary disorders of leukocyte function, chronic granulomatous disease, and myeloperoxidase deficiency indicated that myeloperoxidase-independent and nonoxidative mechanisms could sometimes damage hyphae where oxidative mechanisms were impaired. Damage to hyphae by normal monocytes was inhibited by concentrations of sodium azide and sodium cyanide that primarily affect myeloperoxidase activity, as well as by halide-free conditions, catalase, and putative antagonists of hypochlorous acid or singlet oxygen. Iodination of hyphae, a myeloperoxidase and hydrogen peroxide-dependent process of monocytes, was similarly inhibited by sodium azide, sodium cyanide, and catalase. Under anaerobic conditions, damage to hyphae was reduced by 64.0-68.4%. In contrast, inhibitors of potential nonoxidative antifungal mechanisms, iron salts to saturate iron chelators, and polyanionic amino acid polymers to neutralize cationic proteins did not block damage to hyphae by monocytes. Preparations rich in lysosomal granules from fractionated normal monocytes also did not damage hyphae. Overall, it appeared that oxidative mechanisms were most important for damage to hyphae by normal monocytes. Electron microscopy confirmed that Candida hyphae were damaged and probably killed by monocytes, but monocytes appeared to sustain significant damage in the process. In the absence of serum, monocyte cell membranes became closely approximated to Candida cell walls. It appeared that some Candida could escape this partial engulfment, as they were seen floating free with vesicular trilaminar membrane remnants covering hyphal surfaces. In general, monocytes appeared to be damaged by interactions with Candida hyphae more than neutrophils had been in previous studies.

Mechanisms and Target Sites of Damage in Killing of Candida albicans Hyphae by Human Polymorphonuclear Neutrophils

Target sites of fungal cell damage were studied to define mechanisms of neutrophil-mediated killing of Candida albicans hyphae. Neutrophils induced hyphal cell wall damage, as evidenced by release of cell wall glycoproteins and confocal microscopic changes. Damage occurred in the presence of neutrophil granule extracts and did not require oxidants. However, oxidation of hyphal surface glycoproteins correlated strongly with parallel increments in fungicidal activity, suggesting that oxidants did contribute to maximal cell wall damage. Neutrophil oxidants also induced hyphal DNA fragmentation, primarily single-strand breakage, as shown by increased electrophoretic migration after nuclease-S1 DNA digestion at single-strand break sites. The onset of damage to hyphal cell walls and DNA preceded detectable neutrophil-mediated fungicidal effects. Likewise, hyphal amino acid and nucleotide turnover as well as ATP initially rose, then declined as lethal effects became detectable. Thus, preceding detectable fungal cell death, neutrophil oxidative and oxygen-independent mechanisms damaged defined targets.

Temporal association of calcium mobilization, inositol trisphosphate generation, and superoxide anion release by human neutrophils activated by serum opsonized and nonopsonized particulate stimuli

We have investigated the involvement of phospholipase C mediated polyphosphoinositide turnover in activation of polymorphonuclear leukocytes by particulate stimuli. Results showed that stimulation of leukocytes with serum opsonized zymosan (ingestible particle), serum opsonized Candida albicans hyphae (noningestible particle), or nonopsonized hyphae was followed by a transient rise in cellular inositol phosphates as has been described for neutrophil activation via the formyl peptide receptor. The kinetics of inositol trisphosphate generation paralleled both the time course of changes in cytosolic calcium concentration and the onset of superoxide anion generation, suggesting that these may be related events.