Candida albicans cell surface superoxide dismutases degrade host-derived reactive oxygen species to escape innate immune surveillance

Abstract

Mammalian innate immune cells produce reactive oxygen species (ROS) in the oxidative burst reaction to destroy invading microbial pathogens. Using quantitative real-time ROS assays, we show here that both yeast and filamentous forms of the opportunistic human fungal pathogen Candida albicans trigger ROS production in primary innate immune cells such as macrophages and dendritic cells. Through a reverse genetic approach, we demonstrate that coculture of macrophages or myeloid dendritic cells with C. albicans cells lacking the superoxide dismutase (SOD) Sod5 leads to massive extracellular ROS accumulation in vitro. ROS accumulation was further increased in coculture with fungal cells devoid of both Sod4 and Sod5. Survival experiments show that C. albicans mutants lacking Sod5 and Sod4 exhibit a severe loss of viability in the presence of macrophages in vitro. The reduced viability of sod5Delta/Delta and sod4Delta/Deltasod5Delta/Delta mutants relative to wild type is not evident with macrophages from gp91phox(-/-) mice defective in the oxidative burst activity, demonstrating a ROS-dependent killing activity of macrophages targeting fungal pathogens. These data show a physiological role for cell surface SODs in detoxifying ROS, and suggest a mechanism whereby C. albicans, and perhaps many other microbial pathogens, can evade host immune surveillance in vivo.

Fig. 1

C. albicans induces ROS production in wild-type BMDMs and mDCs. A–C. ROS measurement by luminol-dependent chemiluminescence at 37°C in 2.5 min intervals over a 90 min period [relative luciferase units (RLU) min⁻¹ per 1000 immune cells]. A. Stimulation of BMDMs (a) or mDCs (b) with yeast-form C. albicans (yC.a) at an MOI of 2:1 (equivalent to 2 μg yeast dry weight per well), 5:1 (5 μg/well) or 10:1 (10 μg/well) or with zymosan (20 μg/well). (c) Quantification of the total ROS release between 10 and 50 min (striped area) by calculating the area under the curve (MOI 5:1). The average of three independent experiments is presented. *mDCs produce 5.5 ± 0.35 times more ROS than BMDMs. **P < 0.02. B. Stimulation of BMDMs with hyphae-form C. albicans (hC.a) at 3 μg dry weight/well, 6 μg/well or 12 μg/well. C. Stimulation of gp91phox^−/− and wt BMDMs with yeast-form C. albicans at an MOI of 5:1 or zymosan (20 μg/well). A–C. Results of one experiment per condition are shown. Data were reproduced in at least three independent experiments. Statistical significance was calculated using a two-tailed Student's t-test.

Fig. 2

ROS accumulate when BMDMs or mDCs, but not gp91phox^−/− BMDMs, are infected with sod5Δ/Δ cells. A–C. ROS measurement by luminol-dependent chemiluminescence at 37°C in 2.5 min intervals over a 90 min period [relative luciferase units (RLU) min⁻¹ per 1000 immune cells]. A. (a) Stimulation of BMDMs with either the wild type (CA-IF100) strain or the sod1Δ/Δ (CA-IF003), sod4Δ/Δ (CA-IF015), sod5Δ/Δ (CA-IF019) mutant strains or sod5Δ/SOD5 heterozygous strain (CA-IF017) (MOI 5:1). (b) Stimulation of BMDMs with the sod5Δ/Δ::SOD5 revertant (CA-IF027) (MOI 5:1) or PMA (10 nM). (c) Quantification of the total ROS release between 10 and 60 min (striped area) by calculating the area under the curve (MOI 5:1). The average of three independent experiments is presented. *Infection with sod5Δ/Δ yields 4.3 ± 0.68 times more ROS than with wild-type C. albicans.**P < 0.02. B. (a) Stimulation of mDCs with either the wild type (CA-IF100) strain, or the sod1Δ/Δ (CA-IF003), sod4Δ/Δ (CA-IF015), sod6Δ/Δ (CA-IF023) or sod5Δ/Δ mutant strains. (b) Quantification of the total ROS release between 10 and 60 min (striped area) by calculating the area under the curve. The average of three independent experiments is presented. *Infection with sod5Δ/Δ yields 4 ± 0.64 times more ROS than with wild type cells. **P < 0.05. C. Stimulation of gp91phox^−/− or wild-type BMDMs with either the wild-type (CA-IF100) strain, the sod5Δ/Δ (CA-IF019) mutant strain or sod5Δ/Δ::SOD5 re-integrant (CA-IF027). A–C. Results of one experiment per condition are shown. Data were reproduced in at least three independent experiments. Statistical significances were calculated using a two-tailed Student's t-test.

Fig. 3

Extracellular ROS accumulate in the presence of sod5Δ/Δ cells. A. Superoxides measurement by lucigenin-dependent chemiluminescence at 37°C over a 90 min period [relative luciferase units (RLU) under the curve]. Stimulation of BMDMs with either the wild-type (CA-IF100) strain, or the sod4Δ/Δ (CA-IF015), sod5Δ/Δ (CA-IF019) mutant strain or the sod5Δ/Δ::SOD5 revertant (CA-IF027) (MOI 5:1). *Infection with sod5Δ/Δ yields 3.2 ± 0.21 times more superoxides than with wild-type C. albicans**P > 0.005. B. Extracellular ROS measurement by isoluminol-dependent chemiluminescence at 37°C in 2.5 min intervals over a 90 min period [relative luciferase units (RLU) min⁻¹ per 1000 cells]. Stimulation of BMDMs with either the wild-type (CA-IF100) strain or the sod5Δ/Δ (CA-IF019) or sod6Δ/Δ (CA-IF023) mutant strains (MOI 5:1). Quantification of the total ROS release between 30 and 70 min (striped area) by calculating the area under the curve. *Infection with sod5Δ/Δ yields 10 ± 0.5 times more extracellular ROS than with wild-type cells. **P > 0.001. C. Intracellular ROS production in response to the phorbol ester PMA, wild-type (CA-IF100) strain or sod5Δ/Δ (CA-IF019) mutant strain was measured by FACS analysis using H₂DCF-DA staining of BMDMs after 15 min (a) or 40 min (b) of infection. A–C. Results of one experiment per condition are shown. All data were reproduced in at least three independent experiments. Statistical significances were calculated using a two-tailed Student's t-test.

Fig. 4

A sod4Δ/Δ deletion in a sod5Δ/Δ background boosts ROS accumulation. A. Northern analysis of SOD4, SOD5 and ACT1 mRNA. The clinical C. albicans SC5314 strain was grown at 30°C, 37°C and 37°C plus 10% FCS. B. ROS measurement by luminol-dependent chemiluminescence at 37°C in 2.5 min intervals over a 90 min period [relative luciferase units (RLU) min⁻¹ per 1000 BMDMs]. Stimulation of BMDMs with either the wild-type (CA-IF100) strain or sod5Δ/Δ (CA-IF019), sod5Δ/Δ::SOD5 (CA-IF027), sod4Δ/Δsod5Δ/Δ (CA-IF039) and sod4Δ/Δsod5Δ/Δsod6Δ/Δ (CA-IF070) mutant strains (MOI 5:1) (a). Quantification of the total ROS release between 10 and 50 min (striped area) by calculating the area under the curve (MOI 5:1) and calculating the fold differences. The average of four independent experiments is presented. Infection with sod4Δ/Δsod5Δ/Δ yields 1.36 ± 0.18 times more ROS than by a sod5Δ/Δ strain. *P < 0.05, sod4Δ/Δsod5Δ/Δsod6Δ/Δ triple mutant yields 1.43 ± 0.09 times more ROS than by a sod5Δ/Δ strain. *P < 0.02; and the sod4Δ/Δsod5Δ/Δsod6Δ/Δ triple mutant yields 1.09 ± 0.1 times more ROS than sod4Δ/Δsod5Δ/Δ. ***P > 0.09 (b). C. Superoxides measurement by lucigenin-dependent chemiluminescence at 37°C over a 90 min period [relative luciferase units (RLU) under the curve]. Stimulation of BMDMs with either zymosan (20 μg/well), the wild-type (CA-IF100) strain, the sod5Δ/Δ (CA-IF019), sod4Δ/Δsod5Δ/Δ (CA-IF039) and sod4Δ/Δsod5Δ/Δsod6Δ/Δ (CA-IF070) mutant strains (MOI 5:1). A and B. Results of one experiment per condition are shown. All data were reproduced in two independent experiments. D. Intracellular ROS production in response to wild-type (CA-IF100) strain or sod4Δ/Δ (CA-IF015), sod4Δ/Δsod5Δ/Δ (CA-IF039) mutant strains (MOI 5:1) or zymosan (100 μg ml⁻¹). ROS was measured by FACS analysis using H₂DCF-DA-staining of BMDMs after 45 min of infection. A–C. Results of one experiment per condition are shown. All data were reproduced in at least three independent experiments. Statistical significances were calculated using a two-tailed Student's t-test.

Fig. 5

sod5Δ/Δ and sod4Δ/Δsod5Δ/Δ strains are hypersensitive to killing by BMDMs. A and B. Survival of C. albicans and isogenic mutant cells was determined using the end-point dilution assay. Mean and standard deviation of three independent experiments are presented. A. Wild-type BMDMs in medium without (filled bars) or with 10 U commercial erythrocyte SOD (white bar) were coincubated with either wild-type (wt) C. albicans strain or strains lacking SOD4 (s4Δ), SOD5 (s5Δ blue), the restored SOD5 (s5ΔS5), SOD6 (s6Δ) or strains lacking both SOD4 and SOD6 (s4/6Δ), SOD4 and SOD5 (s4/5Δ green) or lacking all three SOD4, SOD5 and SOD6 (s4/5/6 brown) (a), or with the clinical isolate SC5314 and the sod5Δ/Δ mutant in the SC5314 background (SC5314 s5Δ blue) (b) for 48 h at 37°C with 5% CO₂. B. gp91phox^−/− BMDMs were infected with the wild type (wt) or strains lacking SOD4 (s4Δ), SOD5 (s5Δ blue), the restored SOD5 (s5ΔS5) or strains lacking both SOD4 and SOD5 (s4/5Δ green), the clinical isolate SC5314 or the sod5Δ/Δ mutant in the SC5314 background (SC5314 s5Δ). The percentage of survival for each strain was determined as follows (colonies in absence of BMDMs versus colonies in presence of BMDMs × 100).

Fig. 6

Model for Sod4 and Sod5-mediated protection against respiratory burst. Upon contact with BMDMs and mDCs, Sod4 and Sod5 anchored at the C. albicans (C.a) surface (left) degrade superoxide anions (O₂⁻) to hydrogen peroxide (H₂O₂). The lack of the Sod4 and Sod5 (right) causes ROS accumulation in the medium and perhaps inside the phagosomes (phago), which results in enhanced killing of C. albicans. Production of mitochondrial ROS (Mit) is unaffected.